Citation
Mixed-Metal Molecular Complexes: Single-Molecule Nanomagnets and Bioinorganic Models of the Water Oxidizing Complex of Photosystem II

Material Information

Title:
Mixed-Metal Molecular Complexes: Single-Molecule Nanomagnets and Bioinorganic Models of the Water Oxidizing Complex of Photosystem II
Creator:
MISHRA, ABHUDAYA ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Atoms ( jstor )
Ground state ( jstor )
Hysteresis ( jstor )
Ions ( jstor )
Magnetic fields ( jstor )
Magnetism ( jstor )
Magnetization ( jstor )
Magnets ( jstor )
Molecules ( jstor )
Signals ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Abhudaya Mishra. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
8/31/2006
Resource Identifier:
649814479 ( OCLC )

Downloads

This item has the following downloads:

mishra_a ( .pdf )

mishra_a_Page_152.txt

mishra_a_Page_061.txt

mishra_a_Page_080.txt

mishra_a_Page_155.txt

mishra_a_Page_094.txt

mishra_a_Page_099.txt

mishra_a_Page_208.txt

mishra_a_Page_146.txt

mishra_a_Page_092.txt

mishra_a_Page_013.txt

mishra_a_Page_030.txt

mishra_a_Page_098.txt

mishra_a_Page_001.txt

mishra_a_Page_175.txt

mishra_a_Page_081.txt

mishra_a_Page_213.txt

mishra_a_Page_069.txt

mishra_a_Page_131.txt

mishra_a_Page_221.txt

mishra_a_Page_164.txt

mishra_a_Page_058.txt

mishra_a_Page_039.txt

mishra_a_Page_105.txt

mishra_a_Page_104.txt

mishra_a_Page_051.txt

mishra_a_Page_148.txt

mishra_a_Page_109.txt

mishra_a_Page_119.txt

mishra_a_Page_015.txt

mishra_a_Page_161.txt

mishra_a_Page_021.txt

mishra_a_Page_194.txt

mishra_a_Page_172.txt

mishra_a_Page_017.txt

mishra_a_Page_009.txt

mishra_a_Page_116.txt

mishra_a_Page_036.txt

mishra_a_Page_049.txt

mishra_a_Page_182.txt

mishra_a_Page_008.txt

mishra_a_Page_132.txt

mishra_a_Page_075.txt

mishra_a_Page_086.txt

mishra_a_Page_035.txt

mishra_a_Page_087.txt

mishra_a_Page_178.txt

mishra_a_Page_096.txt

mishra_a_Page_128.txt

mishra_a_Page_157.txt

mishra_a_Page_199.txt

mishra_a_Page_068.txt

mishra_a_Page_134.txt

mishra_a_Page_220.txt

mishra_a_Page_103.txt

mishra_a_Page_032.txt

mishra_a_Page_089.txt

mishra_a_Page_187.txt

mishra_a_Page_125.txt

mishra_a_Page_154.txt

mishra_a_Page_212.txt

mishra_a_Page_206.txt

mishra_a_Page_184.txt

mishra_a_Page_073.txt

mishra_a_Page_176.txt

mishra_a_Page_122.txt

mishra_a_Page_201.txt

mishra_a_Page_170.txt

mishra_a_Page_057.txt

mishra_a_Page_002.txt

mishra_a_Page_133.txt

mishra_a_Page_091.txt

mishra_a_Page_185.txt

mishra_a_Page_204.txt

mishra_a_Page_102.txt

mishra_a_Page_163.txt

mishra_a_Page_130.txt

mishra_a_Page_191.txt

mishra_a_Page_107.txt

mishra_a_Page_189.txt

mishra_a_Page_060.txt

mishra_a_Page_121.txt

mishra_a_Page_215.txt

mishra_a_Page_074.txt

mishra_a_Page_159.txt

mishra_a_Page_196.txt

mishra_a_Page_210.txt

mishra_a_Page_005.txt

mishra_a_Page_047.txt

mishra_a_Page_149.txt

mishra_a_Page_025.txt

mishra_a_Page_123.txt

mishra_a_Page_042.txt

mishra_a_Page_010.txt

mishra_a_Page_033.txt

mishra_a_Page_029.txt

mishra_a_Page_200.txt

mishra_a_Page_012.txt

mishra_a_Page_106.txt

mishra_a_Page_011.txt

mishra_a_Page_153.txt

mishra_a_Page_139.txt

mishra_a_Page_023.txt

mishra_a_Page_053.txt

mishra_a_Page_151.txt

mishra_a_Page_041.txt

mishra_a_Page_183.txt

mishra_a_Page_218.txt

mishra_a_Page_207.txt

mishra_a_Page_211.txt

mishra_a_Page_110.txt

mishra_a_Page_067.txt

mishra_a_Page_088.txt

mishra_a_Page_071.txt

mishra_a_Page_016.txt

mishra_a_Page_052.txt

mishra_a_Page_022.txt

mishra_a_Page_168.txt

mishra_a_Page_165.txt

mishra_a_Page_190.txt

mishra_a_Page_063.txt

mishra_a_Page_144.txt

mishra_a_Page_085.txt

mishra_a_Page_044.txt

mishra_a_Page_202.txt

mishra_a_Page_174.txt

mishra_a_Page_143.txt

mishra_a_Page_028.txt

mishra_a_Page_145.txt

mishra_a_Page_056.txt

mishra_a_Page_167.txt

mishra_a_Page_034.txt

mishra_a_Page_003.txt

mishra_a_Page_180.txt

mishra_a_Page_066.txt

mishra_a_Page_216.txt

mishra_a_Page_205.txt

mishra_a_Page_083.txt

mishra_a_Page_055.txt

mishra_a_Page_024.txt

mishra_a_Page_004.txt

mishra_a_Page_117.txt

mishra_a_Page_135.txt

mishra_a_Page_171.txt

mishra_a_Page_115.txt

mishra_a_Page_198.txt

mishra_a_Page_108.txt

mishra_a_Page_129.txt

mishra_a_Page_093.txt

mishra_a_Page_077.txt

mishra_a_Page_031.txt

mishra_a_Page_156.txt

mishra_a_Page_222.txt

mishra_a_Page_124.txt

mishra_a_Page_050.txt

mishra_a_Page_166.txt

mishra_a_Page_114.txt

mishra_a_Page_127.txt

mishra_a_Page_045.txt

mishra_a_Page_095.txt

mishra_a_Page_101.txt

mishra_a_Page_070.txt

mishra_a_Page_019.txt

mishra_a_Page_014.txt

mishra_a_Page_209.txt

mishra_a_Page_097.txt

mishra_a_Page_195.txt

mishra_a_Page_113.txt

mishra_a_Page_054.txt

mishra_a_Page_203.txt

mishra_a_Page_100.txt

mishra_a_Page_197.txt

mishra_a_Page_192.txt

mishra_a_Page_214.txt

mishra_a_Page_177.txt

mishra_a_Page_072.txt

mishra_a_Page_040.txt

mishra_a_Page_181.txt

mishra_a_Page_064.txt

mishra_a_Page_046.txt

mishra_a_Page_217.txt

mishra_a_Page_188.txt

mishra_a_Page_020.txt

mishra_a_Page_082.txt

mishra_a_Page_147.txt

mishra_a_Page_006.txt

mishra_a_Page_076.txt

mishra_a_Page_142.txt

mishra_a_pdf.txt

mishra_a_Page_169.txt

mishra_a_Page_027.txt

mishra_a_Page_090.txt

mishra_a_Page_111.txt

mishra_a_Page_062.txt

mishra_a_Page_179.txt

mishra_a_Page_026.txt

mishra_a_Page_158.txt

mishra_a_Page_037.txt

mishra_a_Page_193.txt

mishra_a_Page_173.txt

mishra_a_Page_140.txt

mishra_a_Page_138.txt

mishra_a_Page_118.txt

mishra_a_Page_048.txt

mishra_a_Page_065.txt

mishra_a_Page_186.txt

mishra_a_Page_141.txt

mishra_a_Page_007.txt

mishra_a_Page_160.txt

mishra_a_Page_150.txt

mishra_a_Page_078.txt

mishra_a_Page_079.txt

mishra_a_Page_162.txt

mishra_a_Page_043.txt

mishra_a_Page_126.txt

mishra_a_Page_120.txt

mishra_a_Page_137.txt

mishra_a_Page_136.txt

mishra_a_Page_018.txt

mishra_a_Page_084.txt

mishra_a_Page_059.txt

mishra_a_Page_038.txt

mishra_a_Page_112.txt

mishra_a_Page_219.txt


Full Text











MIXED-METAL MOLECULAR COMPLEXES: SINGLE-MOLECULE
NANOMAGNETS AND BIOINORGANIC MODELS OF THE WATER OXIDIZING
COMPLEX OF PHOTOSYSTEM II













By

ABHIUDAYA MISHRA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2006
































Copyright 2006

by

ABHUDAYA MISHRA



































I dedicate this document to my mom, dad and two brothers, for their love and confidence
in me as I pursue the endless journey of life
















ACKNOWLEDGMENTS

I would like to take this opportunity to thank my research advisor, Professor

George Christou, for all his help, guidance, inspiration, and constant encouragement

rendered to me as I pursued my career goals in his research group. I would also like to

thank my other committee members, Dr. C. R. Martin, Dr. M. J. Scott, Dr. D. R. Talham,

and Dr. S. Hill, for some stimulating discussions, and the inspiration which I drew from

their research work, which in turn kindled my own creativity.

In today's world, research is highly interdisciplinary and I would like to take this

opportunity to thank the many research scientists with whom I have collaborated during

my doctoral studies. These collaborators include Dr. Khalil A. Abboud and his staff at

UJFCXC for solving the X-ray crystal structures, and Dr. Wolfgang Wernsdorfer, who

provided cryogenic magnetic measurements (below 1.8 K) on the compounds mentioned

in this dissertation. Additionally, I would like to express my heartfelt appreciation to Dr.

Junko Yano and Dr. Vittal K. Yachandra, at Lawrence Berkeley National Labs, for

performing XAS measurements on complexes which were relevant to the bioinorganic

research. In the course of the collaborative research performed, Khalil, Wolfgang, and

Junko have indeed become wonderful friends.

And speaking of friends, what would research be without any camaraderie between

one's colleagues! I would like to thank all the Christou group members (past and present)

who made Gainesville worth living, for the past five years. Special thanks go to Dolos for

her support and friendship. Alina completes the trio for the awesome time all three of us









have spent together laughing, talking, and of course drinking. I would also like to thank

Tasos, Carol and Monica who helped me get started in this lab and also for the fun

companionship which they provided. The Indian gang of Parul, Ranjan and Rashmi also

deserve special mention. Last, but not least, I would like to acknowledge the

unconditional support of my two brothers, Abhishek and Animesh, and my mother and

father, who have always been there for me whenever I needed something. I am forever

indebted to them for their constant pride, encouragement, love, and unwavering

confidence in me as I undertook the daunting task of my doctoral studies. Indeed, my

family has made this wonderful journey much more meaningful.




















TABLE OF CONTENTS


page

ACKNOWLEDGMENT S .............. .................... iv


LI ST OF T ABLE S ........._..... .............._ ix......_....


LI ST OF FIGURE S .............. .................... xi


AB S TRAC T ......_ ................. ............_........x


CHAPTER


1 GENERAL INTRODUCTION .............. ...............1.....


2 SINGLE-MOLECULE MAGNETS: A NOVEL FAMILY OF Mn" / Ce'V
COMPLEXES WITH A [MnsCesOs]12+ CORE ................. ................. ....__. 16

2. 1 Introducti on ........._..... ...._... ...............16...
2.2 Results and Discussion ........._.._.. ...._... ...............17...
2.2. 1 Syntheses ........._.._... ......._. ..............._ 17..
2.2.2 Description of Structures ................. .......... .. ............... 21....
2.2.2. 1 X-ray crystal structures of complexes 4 7............... ...................21
2.2.2.2 Structural Comparison of complexes 4 7 ................... ...............25
2.2.3 Magnetochemistry of Complexes 4, 5 and 7 ................ ......._._.........28
2.2.3.1 DC studies .............. ...............28....
2.2.3.2 AC studies ................ ..... ... ..............3
2.2.3.3 Hysteresis studies below 1.8 K .............. ...............34....
2.3 Conclusions............... ..............3
2.4 Experimental ........._.___..... ._ __ ...............39....
2.4.1 Syntheses .............. .. ...............3 9...
2.4.2 X-ray Crystallography ................. ......... ...............41. ...

3 SINGLE-MOLECULE MAGNETS: SYNTHESES AND MAGNETIC
CHARACTERIZATION OF A NOVEL FAMILY OF HETEROMETALLIC
MANGANESE-LANTHANIDE COMPLEXES .............. ...............44....


3 .1 Introducti on ..... .................. ...............44.....
3.2 Results and Discussion .............. ...............46....
3.2.1 Syntheses .............. ...............46...
3.2.2 Description of Structures ................. ...............49........... ...











3.2.2. 1 X-ray crystal structure of complexes 11 and 13............... ................49
3.2.2.2 Structural Comparison of Complexes 8-13 ................. ........._..._... 53
3.2.2.3 Structural descriptions of complexes 14 and 15............... ................55
3.2.3 Magnetochemistry of Complexes 8-13, and 15............... ...................5
3.2.3.1 DC studies of complexes 9, 10, 11, and 13 ........._._.... ........._.........58
3.2.3.2 DC studies of [Mn2 202(O2CPh)6(OMe)4(MeOH)4] (15).............. .63
3.2.3.3 AC studies of complexes 9, 10, 11, and 13 ........._._..........._.._.....64
3.2.3.4 Hysteresis studies below 1.8 K on complexes 8-13 .........................68
3.3 Conclusions............... ..............7
3.4 Experimental ........._.___..... ._ __ ...............76....
3.4.1 Syntheses .............. .. ...............76...
3.4.2 X-ray Crystallography ................. ...............80................

4 HIGH NUCLEARITY COMPLEXES: HOMOVALENT
[Th6MnloO22(OH)2(O2CPh)16(NO3)2(H20)s] AND MIXED-VALENT
[Mn70s(OR)2(O2C~h)9(terpy)] (R = Me, CH2Ph) DISPLAYING SLOW-
MAGNETIZATION RELAXATION ........._._........___...... .............8

4. 1 Introducti on ........._._ ...... .. ...............83..
4.2 Results and Discussion .............. ...............86....
4.2.1 Syntheses .............. ... ...............86...
4.2.2 Description of Structures ...._.._.._ ..... .._._. ......... ............8
4.2.2.1 X-ray crystal structure of
[MnloTh6022(OH)2(O2CPh)16(NO3)2(H20)s] (16) ............... ... ........._.._..88
4.2.2.2 X-ray crystal structures of [Mn70,(OR)2(O2CPh)9(terpy)]
complexes (17, 18)............... ......... .... .........9
4.2.3 Magnetochemistry of Complexes 16, 17 and 18 ................. ........._......99
4.2.3.1 DC studies .............. ...............99....
4.2.3.2 AC studies ................ .... ... .............10
4.2.3.3 Hysteresis studies below 1.8 K ........._._........__. ................111
4.3 Conclusions............... ..............11
4.4 Experimental ................. ...............114......... ......
4.4. 1 Syntheses .............. ... ................. 114........ ....
4.4.2 X-ray Crystallography ................. ...............116................

5 THE FIRST STRONTIUM-MANGANESE CLUSTER: SINGLE-MOLECULE
MAGNETISM AND Sr-EXAFS COMPARISON WITH THE WATER
OXIDIZING COMPLEX OF PHOTO SYSTEM II ................ .......................118

5 .1 Introducti on ................. ...............118........... ...
5.2 Results and Discussion ................ ...............120........... ...
5.2. 1 Syntheses .............. .... ................. 120........ ...
5.2.2 Description of Structures ................. ...............121........... ...
5.2.3 Magnetochemistry of Complex 19 .............. .....................124
5.2.3.1 DC studies of 19 ................. ...............124....._.. .
5.2.3.2 AC studies of 19 ....._._._ ... ... .......... ...............128.
5.2.3.3 Hysteresis studies below 1.8 K .............. ...............130....












5.2.4 X-ray Absorption Spectroscopy of Complex 19 .............. ................... 132
5.2.4.1 Sr EXAFS of 19 .............. ...............133....
5.2.4.2 Mn EXAFS of 19 .............. ...............138....
5.3 Conclusions............... ..............14
5.4 Experimental ........._.___..... .___ ...............141....
5.4.1 Synthesis............... .. ..............14
5.4.2 X-ray Crystallography ................. ...............142...............
5.4.3 XAS studies ................. ...............143...............


THE FIRST FAMILY OF HETEROMETALLIC CALCIUM-MANGANESE
COMPLEXES: Ca-EXAFS AND -XANES COMPARISON WITH THE
WATER OXIDIZING COMPLEX OF PHOTOSYSTEM II ................ ...............145


6. 1 Introducti on ................. ...............145........... ...
6.2 Results and Discussion ................ ...............147...............
6.2. 1 Syntheses .................. ... ......... ........... .... ..........4
6.2.2 Description of the Structures of complexes 20-22 ................. ................150
6.2.3 Magnetochemistry of Complexes 20-22 .............. .....................156
6.2.4 Calcium XAS studies of complexes 20-22 ................. ............ .........162
6.2.4. 1 Calcium EXAFS of complexes 20-22 ................. .....................164
6.2.4.2 Calcium XANES of complexes 20-22 .............. .....................169
6.3 Conclusions............... ..............17
6.4 Experimental ............ ......__ ...............175..
6.4. 1 Syntheses ............. ... .._ .............. 175...
6.4.2 X-ray Crystallography ................. ...............177................
6.4.3 XAS Studies .............. ...............178....


APPENDIX


A BOND DISTANCES AND ANGLES................ ...............180


B LIST OF COMPOUNDS............... ...............18


C PHYSICAL MEASUREMENT S .............. ...............190....


LIST OF REFERENCES ........._.._ ........... ...............193....


BIOGRAPHICAL SKETCH .............. ...............206....

















LIST OF TABLES


Table pg

2-1 Crystallographic data for 4, 5, 6, and 7 .............. ...............27....

2-2 Comparison of the magnetic parameters of complexes 4, 5, and 7 ................... .......3 8

3-1 Crystallographic data for 10, 11 and 13 .............. ...............52....

3-2 Bond Valence Sums for the Mn atoms of complexes 15 (Y) and 14 (Yb) ..............57

3-3 Bond Valence Sums for the O atoms of complexes 15 (Y) and 14 (Yb) ........._......57

3-4 Selected bond distances (A+) and angles (o) for complexes 14 and 15.....................58

3-5 Comparison of the SMM parameters of complexes 10, 11 and 13. .........................74

4-1 Bond valence sum calculations for the Mn ions in complex 16............... ................92

4-2 Bond valence sum calculations for the Mn ions in complexes 17 and 18 ...............97

4-3 Crystallographic data for 17 and 18 .............. ...............98....

4-4 Comparison of the magnetic susceptibility parameters of complexes 17 and 18 ..109

5-1 Crystallographic data for [SrMnl4011(OMe)3(O2CPh)ls(MeCN)2] .......................124

5-2 Selected bond and interatomic distances (8+) for complex 19. .............. ..... ........._.135

5-3 Least-squares fits ofFourier-fi1tered peaks I and II of Sr EXAFS data on
Complex 19 and the Sr-substituted WOC in the S1 state ................. ................ 136

6-1 Bond valence sum calculations for the Mn and O atoms of complex 20 ...............152

6-2 Crystallographic data of 20, 21, and 22 ........._... ........ ......_. ...........155

6-3 Selected interatomic and bond distances (A+) for the Ca and Mn atoms of
complexes 20-22 .............. ...............166....

6-4 Least-squares fits of Fourier-fi1tered peaks I and II of Ca EXAFS data on
complexes 20 and 21, and the WOC in the S1 state ................ ............ .........167









A-1 Selected interatomic distances (A+) and angles (o) for
[MnsCeOs(O2CMe)12(H20)4] (4) and [MnsCeOs(O2CMe)12 97)4] (5) ..................180

A-2 Selected interatomic distances (A+) and angles (o) for
[MnsCeOs(O2CCHPh2)12(H20)4] (7)............... ...............181..

A-3 Selected interatomic distances (A+) and angles (o) for
[MnllGd40s(OH)s(O2CPh) 16(N O3)5(H 20)3] (10). ............. ...............182

A-4 Selected interatomic distances (A+) and angles (o) for
[Mnl 1Dy40s(OH)6(OMe) 2(O2CPh) 16(NO3)5(H20)3] (11). ................ .................1 83

A-5 Selected interatomic distances (A+) and angles (o) for
[MnllTb40s(OH)6(OCH2Ph)2(O2CPh)16(NO3)5(H03 (13)........ ...............1841s

A-6 Selected interatomic distances (A+) and angles (o) for
[MnloTh6022(OH) 2(O2CPh) 16(NO3)2(H20)s] (16) ................. .......................185

A-7 Selected interatomic distances (A+) and angles (o) for
[Mn70,(OMe)2(O2CPh)9(terpy)] (17). ............. ...............186....

A-8 Selected interatomic distances (A+) and angles (o) for
[Mn70,(OCH2P>(,Ph)2(O2CPh)(trp) (18). ............. ...............187....
















LIST OF FIGURES


Figure pg

1-1 Representations of magnetic dipole arrangements in (i) paramagnetic, (ii)
ferromagnetic, (iii) antiferromagnetic, and (iv) ferrimagnetic materials. ..................3

1-2 Typical hysteresis loop of a magnet, where M is magnetization, H is the applied
magnetic field and 2M is the saturation value of the magnetization. ........._...............5

1-3 Representative plots of the potential energy versus (a) the magnetization
direction and (b) the ms sublevels .............. ...............8.....

1-4 In-phase XM'T, and out-of-phase XM". Possible tunneling mechanisms (top) for
SMMs and (bottom) typical hysteresis loops with steps, for a Mnl2 SMM.............10

1-5 The S state scheme as proposed by Kok for the oxidation of water. Arrangement
of 4 Mn and 1 Ca atoms in the two latest crystal structures of the WOC of PS II...12

1-6 Crystal structure of the WOC at 3.0 A+, and the 3.5 A+ Mn4Ca cubane-containing
crystal structure. Possible Mn4CaOx topologies suggested by biophysical studies. 13

2-1 PovRay representation at the 50% probability level of the X-ray crystal structure
of 4 ................ ...............22........ ......

2-2 PovRay representation at the 50% probability level of the X-ray crystal structure
of 5 (left) and 6 (right)............... ...............23

2-3 PovRay representation at the 50% probability level of the X-ray crystal structure
of 7 ................ ...............24........ ......

2-4 Comparison of the [MnsCeOs]12+ COre, with the [Mnl2012 16+ COre. The common
core of complexes 4, 5, 6 and 7 ............... ...............25....

2-5 Plots of XMT vs T for complexes 4, 5 and 7. ................ ............. ........ .......29

2-6 Magnetization (M) vs field (H) and temperature (T) data, plotted as reduced
magnetization (M/NCLB) vs H/T, for (left) complex 4, and (right) for complex 7 ....30

2-7 Magnetization (M)1 vs applied magnetic field (H) hysteresis loops: (left) for 4
and (right) for 5. M~is normalized to its saturation value, 2, for both plots...........35










2-8 Magnetization (M)1 vs applied magnetic field (H) hysteresis loops for 7. M~is
normalized to its saturation value, Ms, for both plots............... .................3

3-1 PovRay representation at the 50% probability level of the X-ray crystal
structures of 11 (left) and 13 (right) ................ ...............50..............

3-2 PovRay representation at the 50% probability level of the
[Mnl 1Dy40s(OH)6(OMe)2 21+ COre of 11 (left) and the
[Mnl 1Tb40s(OH)6(OCH 2Ph) 2 21+ COre of 13 (right) ................. .......................51

3-3 PovRay representation at the 50% probability level of the [Mnl 1Gd40s(OH)821+
core of complex 10 ............... ...............54....

3-4 PovRay representations of the crystal structures of complexes 14 (left) and 15
(right). Comparison of the cores of complexes 14 (left) and 15 (right),
emphasizing their near superimposibility .............. ...............56....

3-5 XMT vs T plots for complexes 9 (*), 10 (0), 11 (V), and 13 (A). ............................59

3-6 Magnetization (M) vs field (H1) and temperature (T) data, plotted as reduced
magnetization (M/NCLp) vs H/T, for complex 10 .............. ...............62....

3-7 Plot of XMT(solid circles, *) vs. Tfor complex 15. The solid line in the XMTys T
plot is the fit of the data; see the text for the fit parameters .............. ..............63

3-8 Ac susceptibility of complex 11. (Top) in-phase signal (17') plotted as 3Cu'Tyvs T;
and (bottom) out-of-phase signal gs," vs T............... ...............66...

3-9 Ac susceptibility of complex 13. (Top) in-phase signal (gs,') plotted as 3Cu'Tyvs T;
and (bottom) out-of-phase signal gs," vs T............... ...............67...

3-10 Magnetization (M)1 vs dc field (H) hysteresis loops for single crystals of 8 (top;
left), 9 (top; right), 10 (bottom; left), and 12 (bottom; right) ................ ................69

3-11 Magnetization (M) vs dc field (H) hysteresis loops for single crystals of 11 (top),
and 13 (bottom) .............. ...............70....

3-12 Magnetization vs time decay plots for crystals of complex 13 at the indicated
temperatures. Arrhenius plot using the resulting relaxation time (r) versus T data.72

3-13 Arrhenius plot of the relaxation time (r) versus 1/T, constructed from Dc
magnetization decay data for complexes 11 (left), and 10 (right) ...........................73

4-1 PovRay representation of the X-ray crystal structure (left) of 16, and (right) the
2:3:6:3:2 (Mn:Th:Mn:Th:Mn) layer structure of 16. Bottom depicts the labeled
[MnloTh6022(OH )2 18+ COre of complex 16 .............. ...............89....










4-2 PovRay representation at the 50% probability level of the X-ray crystal structure
of 17, and (bottom) its labeled [Mn705(OMe)2(terpy)]9+ COre ................. ...............94

4-3 PovRay representation at the 50% probability level of the X-ray crystal structure
of 18 (top) and its stereopair (bottom) .............. ...............96....

4-4 Plot of guTys Tfor complex 16. Magnetization (M) vs field (H) and
temperature (T) data, plotted as reduced magnetization (M~NpB) VS H T, for 16 ..100

4-5 Plots of XMT vs T for complexes 17 and 18. ................ .............. ........ .....101

4-6 Magnetization (M) vs field (H) and temperature (T) data for (left) complex 17,
and (right) for complex 18. Two-dimensional contour plot of the error surface
for the D vs g fit for complexes 17 (left), and 18 (right) ................. ................ ..103

4-7 Schematic representation and justification of the spin alignment in the Mn7
complexes based on a fused cubane/butterfly arrangement ................. ................106

4-8 Plots of the in-phase (as XM'T) ac susceptibility signals vs T for complex 16
(left), and complexes 17 and 18 (right). The measurement is in a 3.5 G ac Hield
oscillating at a 500 Hz frequency. ............. ...............108....

4-9 Magnetization (M)1 vs applied magnetic field (H) hysteresis loops for single-
crystals of 16. Hysteresis loops at temperatures of 0.7 and 0. 1 K for single
crystals of 18 ................. ...............112...............

5-1 PovRay representation at the 50% probability level of the X-ray crystal structure
of 19, and its stereopair ................. ...............122.......... ...

5-2 The labeled [Mnl4SrOnl(OMe)3] 18+ COre of complex 19 ................. ................ ..123

5-3 Plot of XMTys T for complex 19. Magnetization (M) vs field (H) and
temperature (T) data, plotted as reduced magnetization (M/NCLB) vs H/T, for 19..125

5-4 Two-Dimensional contour plot of the error surface for the reduced
magnetization (M/NCLB) vs H/T fit for complex 19. Three-Dimensional mesh plot
of the error vs D vs g for the same fit for 19. ................. ............... 127..........

5-5 Plot of the in-phase (as XM' T) and out-of-phase (XM") AC susceptibility signals
vs temperature for complex 19............... ...............129..

5-6 Magnetization (M)1 vs applied magnetic field (H) hysteresis loops for single-
crystals of 19 at a fixed sweep rate of0.14 T/s and at 0.04 K .............. ................130

5-7 Plot of the magnetization versus time decay data of 19 ................ ................ ...13 1

5-8 k3-weighted Sr K-edge EXAFS spectra of the Mnl4Sr compound (red) and Sr
reactivated PS II samples in the S1 state (black) ................ .........................134










5-9 PovRay representation of two open-cubane containing sub-units within 19. Left
depicts a sub-unit which resembles the 3.0 A+ crystal structure of the WOC26 Of
PS II............... ...............137..

5-10 k3-weighted Mn K-edge EXAFS spectra of the Mnl4Sr compound (red) and PS
II samples in the S1 state (black) ................ ...............139........... ..

6-1 PovRay representation of the crystal structure of 20, and its labeled
[Mn13 Ca2010(OH)2(OMe)2~ 18+ COre ................. ...............151.......... .

6-2 Schematic representation of the three bridging modes found in complex 20. .......15 1

6-3 PovRay representation of the crystal structure of the doubly charged anion of 21,
and its labeled [MnllCa4010(OH)2(OMe)2 18+ COre.................. ...............15

6-4 PovRay representation of the crystal structure of 22, and its labeled
[MnsCaO6(O2P(Ph)2)4(OMe)2(MeOH)2] 10+ COre ........._..._... ....................._.154

6-5 Plots of XMTys T for Mnl3Ca2 (*) and MnllCa4 t0) COmplexes ................... .........157

6-6 Magnetization (M) vs Hield (H) and temperature (T) data, plotted as reduced
magnetization (M/NCLB) vs H/T, for complex 20 ................. .........................158

6-7 Magnetization (M) vs Hield (H) and temperature (T) data, plotted as reduced
magnetization (M/NCLB) vs H/T, for complex 22 ................. .........................160

6-8 XMTys T plot for 22 and for each of the two cubanes of the MnsCa complex.
Model used to fit the exchange interactions ................. .............................161

6-9 k3-weighted Ca K-edge EXAFS spectra of the Ca/Mn complexes and PS II S1
state, and Fourier transforms of the EXAFS spectra............_ .. ......._ ........165

6-10 k3-weighted Ca K-edge EXAFS spectra of the Mnl3Ca2 and MnllCa4 COmplexes
and PS II S1 state, and Fourier transform of the EXAFS spectra............._._._.........168

6-11 Ca K-edge XANES spectra of Ca/Mn complexes 20-22 and PS II in the S1 state,
and their second derivative spectra. (Inset) The pre-edge like peaks ................... ..170

6-12 Ca K-edge EXAFS spectra of the Mnl3Ca2 COmplex (red) and PS II S1 state
(black), and Sr K-edge EXAFS spectra of the Mnl4Sr complex (red) and Sr-
reactivated PS II samples in the S1 state (black). Mn4Ca and Mn4Sr interactions
present in complexes 20 and 19 .............. ...............173....

6-13 PovRay representations of units/sub-units present in the Mn/Ca complexes 20-
22, and in the Mnl4Sr complex 19. Sub-units of complex 20 in the first row
resemble the 3.0 A+26 (left) and 3.5 A+22 (center) crystal structures of the WOC.....174
















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

MIXED-METAL MOLECULAR COMPLEXES: SINGLE-MOLECULE
NANOMAGNETS AND BIOINORGANIC MODELS OF THE WATER OXIDIZING
COMPLEX OF PHOTOSYSTEM II

By

Abhudaya Mishra

August 2006

Chair: George Christou
Major Department: Chemistry

The current burgeoning re search in high nuclearity mangane se-containing

carboxylate clusters is primarily due to their relevance in areas as diverse as magnetic

materials and bioinorganic chemistry. In the former, the ability of single molecules to

retain, below a critical temperature (TB), their magnetization vector, resulting in the

observation of bulk magnetization in the absence of a field and without long-range

ordering of the spins, has termed such molecules as Single-Molecule Magnets (SMMs),

or molecular nanomagnets. These molecules display superparamagnet like slow

magnetization relaxation arising from the combination of a large molecular spin, S, and a

large and negative magnetoanisotropy, D. Traditionally, these nanomagnets have been

Mn containing species. An out of the box approach towards synthesizing SMMs is

engineering mixed-metal Mn-containing compounds. An attractive choice towards this

end is the use of Lanthanides (Ln), which possess both a high spin, S, and a large D. A

family of related Mn 'sCe'v SMMs has been synthesized. However, the Ce ion of these









complexes is diamagnetic (Ce '). Thus, further investigation has led to the isolation of a

family of Mn 'llLn '4 COmplexes in which all but the Ln = Eu complex function as

single-molecule nanomagnets. The mixed-metal synthetic effort has been extended to

include actinides with the successful isolation of a Mn 10oTh'V6 COmplex, albeit this

homovalent complex is not a SMM.

In the bioinorganic research, the Water Oxidizing Complex (WOC) in Photosystem

II (PS II) catalyzes the oxidation of H20 to 02 in green plants, algae and cyanobacteria.

Recent crystal structures of the WOC confirm it to be a Mn4CaOx cluster with primarily

carboxylate ligation. To date, various multinuclear Mn complexes have been synthesized

as putative models of the WOC. On the contrary, there have been no synthetic MnCa(Sr)

mixed-metal complexes. Thus, in this bioinorganic modeling research of the WOC,

various synthetic methods have been developed to prepare a variety of heterometallic

MnCa(Sr) complexes, namely, Mnl3Ca2, MnllCa4, MnsCa and Mnl4Sr; these are the first

of their kind. X-ray absorption spectroscopy has been performed on all of these

complexes and the results compared with analogous data on the WOC of PS II. In

particular, Ca, Sr, and Mn, EXAFS and XANES reveal a distinct similarity between the

sub-units within these complexes and the Mn4CaOx site of the WOC. The data strongly

suggest that a single-atom O bridge exists between the Mn atoms and the Ca atom of the

WOC.















CHAPTER 1
GENERAL INTRODUCTION

Magnetism is a topic which is highly underrated since although it has a profound

impact on our day to day life it is much less researched on in academia. It all started with

the ancient Greeks, originally those near the city of Magnesia, and also the early Chinese

who knew about strange and rare stones, possibly chunks of iron ore struck by lightning,

with the power to attract iron in a magical way. A steel needle stroked with such a

"lodestone" became "magnetic" as well, and the Chinese found that such a needle, when

freely suspended, pointed north-south. This led to the discovery and subsequent

exploitation of magnets and magnetism, and human civilization has tremendously

benefited from it ever since. Magnets are so essential and ubiquitous to a plethora of

devices in our daily life that sometimes they are taken for granted. Be it the simple

speakers or headphones, or the complicated motors or telecommunication devices,

magnets find an application almost everywhere. Modern day magnetic materials include

magnetic alloys and oxides, particularly ferrites such as MgFe204, which can function in

transformer cores, magnetic recording or information storage devices. The global market

for magnetic materials is valued at $50 billion and has a proj ected growth rate of 10%

annually. Additionally, with the advent of GNR (genetics, nanotechnology and robotics)

this burgeoning field of magnetism and magnetic materials will undoubtedly expand as

the 21s~t century progresses, since magnets will be crucial to the development of the so-

called "smart materials" and "smart systems."









The behavior of any magnetic material is essentially dependent on the presence of

unpaired electrons, or more precisely the spin associated with the unpaired electrons. The

magnetic field associated with a magnetic substance is the result of an electrical charge in

motion, specifically the spin and orbital angular moment of electrons within atoms of a

material. Hence, the number of unpaired electrons and the net interactions of these

electron spins govern the response or susceptibility X of the material to an applied field.

The presence of unpaired electrons on a material classifies it as paramagnetic and the

absence makes a material diamagnetic. The electron pairs of a diamagnet interact with an

applied field, generating a repulsive field that weakly repels the diamagnet from the

applied field; the sign of X is negative. In contrast, a paramagnet is attracted to an applied

magnetic field; the sign of X is positive. In the presence of an applied magnetic field the

electron spins in paramagnetic materials try to align parallel to the applied field and this

effort is opposed by the entropically favored randomizing effect of thermal energy. Thus,

removal of the field results in the randomization of the spins. As a result, a paramagnet

has a net zero magnetization and may not act as a magnet. However, as the number of

metal centers increase, spin-coupling allows cooperative or non-cooperative interactions

that result in parallel or anti-parallel alignment of spins, respectively. Therefore on

removal of the applied field if the spins remain aligned parallel, thereby possessing a net

magnetic moment, the material is ferromagnetic. Based on the alignment of the magnetic

dipole moment of the spins, materials can be classified as paramagnetic, ferromagnetic,

antiferromagnetic and ferrimagnetic as shown in Figure 1-1. Besides these prototypes,

magnetic materials can also show spin glass, metamagnetic and canted ferromagnetic /

antiferromagnetic behavior.










To remain ferromagnetic after the Hield is removed (i.e., remnant magnetization),

the system must be below a critical temperature, Tc. Above Tc (curie temperature), the

thermal energy (kT) is large enough to cause the electron spins to orient randomly and

the material behaves as a simple paramagnet. Assuming the material is still below Tc and

the spins are aligned parallel, applying a Hield in the reverse direction induces the spins to

align in the opposite direction. At zero magnetization, a positive Hield strength







Paramagnetic Ferromagnetic Antiferromagnetic Ferrmagnetic
Figure 1-1. Representations of magnetic dipole arrangements in (i) paramagnetic, (ii)
ferromagnetic, (iii) antiferromagnetic, and (iv) ferrimagnetic materials.

can be seen, known as the coercive Hield or the hardness of the magnet. This phenomenon

can be seen as a hysteresis loop in a plot of the magnetization M versus the applied Hield,

B (Figure 1-2).1 Antiferromagnetism and ferrimagnetism are the opposite situations to

ferromagnetism. An antiferromagnet has its spins aligned antiparallel (opposing)

producing a net zero magnetization below a critical temperature, known as the Neel

temperature, TN. If the spins align antiparallel but a non-zero magnetization results, the

material is described as ferrimagnetic. This is due to spin centers with different magnetic

moment magnitudes and hence the net magnetization does not cancel out even though the

spin vectors are ordered in an antiparallel fashion. An example is the naturally occurring

magnet, magnetite (Fe304). CrO2 and LaFeO3 TepreSent common examples of

ferromagnetic and antiferromagnetic materials, respectively.

At all temperatures, ferro-, antiferro- and ferri-magnets are composed of domains,

or tiny regions in which all the spins are aligned parallel or antiparallel. The transition









from independent to cooperative behavior in these materials is associated with a curie

temperature, T,. Above T,, there is enough thermal energy to cause a random alignment

of each domain with respect to its neighbor, maximizing the entropy while minimizing

the magnetization of the system. The application of a strong magnetic Hield induces the

alignment of all of the domains with the field and hence with each other, imparting a net

magnetization to the ferro- or ferri-magnetic material. As alignment occurs, the

interaction of spins becomes strong enough to overcome entropy considerations that

maintain the random alignment of the domains.1, 2 When a magnetic field is applied and

then removed at a temperature below T,, the magnetization induced by the Hield does not

entirely disappear, and in some cases can remain equal to the Hield-induced

magnetization. This is in contrast to the behavior observed for paramagnetic systems in

which the spins immediately (>10-9 S) randomly reorient following removal of the Hield.

For suppression of the remnant magnetization, a coercive field in the opposite direction is

applied, inducing the realignment of the spins in the opposite direction and resulting in a

hysteresis loop (Figure 1-2).

Richard Feynman, the Nobel Prize-winning physicist, once gave a lecture called,

"There's Plenty Of Room At The Bottom." This seminal lecture laid the foundation of

nanotechnology and ever since, in accordance with Moore's Law, electronic devices have

become progressively smaller. In this scenario it becomes increasingly important to

understand the magnetic properties of smaller bistable particles for the storage of

information. For information storage a small coercive field (high permeability) with a

relatively rectangular shaped hysteresis loop is required so that the two magnetic

orientations of the spin can represent zero (spin-up) and one (spin-down) in the binary









digit (bit) system used by current technology. The requirement for information storage is

that the system remains at a temperature at which the material exhibits hysteresis while

the removal of the stored information involves heating to a temperature above Te.1 Thus a

considerable amount of research is being concentrated on making smaller and smaller

materials that behave similar to permanent magnets.




tttttttt






-3 -11 2 3

o.s C B (Tesirsa)








Figure 1-2. Typical hysteresis loop of a magnet, where M is magnetization, H is the
applied magnetic field and Mr is the saturation value of the magnetization.

One idea is to fragment a ferromagnet or ferrimagnet to a size smaller than a

single domain (20-200 nm); therefore all the spins within the particle always remain

parallel. These particles, known as superparamagnets, are composed of randomly

oriented spins unless induced by an applied magnetic field. Superparamagnets retain

their magnetization when their magnetic relaxation is slowed below a blocking

temperature, TB. Problems with this approach include a wide distribution of shapes and

sizes.2 Additionally, there is a distribution of barrier heights for the interconversion of the

spins and these materials are insoluble in organic solvents and thus unsuitable for some









applications and studies. Yet another approach to obtaining small magnetic materials in

the superparamagnet range is to build molecule-based magnets. Rather than using solids

of extended lattices such as oxides, this approach uses 3-D lattices of molecular building

blocks, which are synthesized from single molecules and selected bridging groups.

Considerable research focusing on purely organic magnets,3 hybrid organic/inorganic

magnets (Awaga et al.),4 and inorganic cyanide-based network-structured magnets

(Miller and Epstein et al.)" has been done in this area. Advantages of this approach

include low density, solubility, biocompatibility, transparency and high magnetizations.

The first example of molecule-based magnets was discovered in 19666 but it took twenty-

seven more years of research for scientists to realize that a molecule can behave as a

magnet by itself, rather than through long-range ordering. Finally, in 1993, the first

example of a molecule, [Mnl2012z(O2CMe)16(H20)4], in Short Mnlzac, able to behave as a

magnet by itself was discovered.7 This discovery led to a totally new approach to

nanoscale magnets in which the magnetism was intrinsic to the molecule and not due to

interactions between molecules. Magnetic studies in a polyethylene matrix of Mnl280

proved the latter hypothesis true and showed the absence of any long-range three-

dimensional interactions.8 Ever since, polynuclear metal complexes with magnetic

behavior similar to Mnlzac and exhibiting superparamagnetic-like properties have been

called single-molecule magnets (SMMs).9 The name itself is somewhat misleading

because in traditional magnetism, to have a magnet it is necessary to have an infinite

number of coupled spin centers, but is evocative, and can be used provided the above

caveat is taken into consideration.10









Since the discovery of SMMs, there has been a great interest in the chemistry and

physics communities in understanding this new magnetic phenomenon of single-

molecule magnetism and synthesizing new SMMs. This interest has paid rich dividends

with the discovery of many more manganese clusters, as well as vanadium, iron, cobalt

and nickel clusters which behave as SMMs. However of the SMMs known to date, the

[Mnl2012(O2CR)16(H20)x] (R= -Me ,-Ph, ,-C6H4-2-C1, ,-CHCl2 ,-C6Fs,-C6H4-2-Br,-

CH2But ,-CH2Ph with x = 4 and R= -Et -CH2Cl with x = 3) family, also known as Mnl2

complexes,?, to still possess the best structural and electronic properties for this

phenomenon, inasmuch as they display single-molecule magnetism behavior at the

highest temperatures and behave as a magnet below 4 Kelvin.9, 11

Single-molecule magnets are molecules that can function as nanoscale magnets

below a certain blocking temperature. These molecules exhibit slow magnetization

relaxation (reorientation) rates, which result in magnetization hysteresis loops. Since

SMMs display hysteresis, like any classical magnet, they have potential applications in

future magnetic storage devices where one bit of information can be stored on a single

molecule, thereby greatly increasing the data density of information storage devices.

Progress towards this end involves the use of smaller materials of nano- and subnano-

scale dimensions (like SMMs) that behave as "permanent" magnets, albeit with

functional temperatures in the practical range for technological use. Also, due to sub-

nanoscale sizes and monodisperse behavior, these molecules show quantum tunneling of

the magnetization (QTM) at the macroscopic level,12 and thus act as a bridge between the

quantum and classical understanding of magnetism. Since QTM is an inherent property

of these molecules and they show quantum coherence,13 SMMs are possible candidates as










qubits (quantum bits) in future quantum computers.14 In Order to display SMM behavior a

molecule should possess a large spin ground state (S) and a large negative magnetic

anisotropy, gauged by the zero-field splitting (ZFS) parameter D (i.e., Ising or easy-axis-

type anisotropy). The presence of both a high spin ground state and a large, negative ZFS

parameter in a single molecule is rather rare and thus interesting magnetic properties are

generated. For

(a) (b)
0 m.=O



7 -7




10 1. -10 = ms + ms=+0 ms= -lo

Magnetization Direction m,
Figure 1-3. Representative plots of the potential energy versus (a) the magnetization
direction and (b) the ms sublevels, for a Mnl2 COmplex with an S = 10 ground
state, experiencing zero-field splitting.

example, for the aforementioned family of Mnl2 COmplexes the S = 10 always, and the D

varies from -0.3 cm-l to -0.5 cm- The combination of these S and D values lead to a

barrier to magnetization reversal (reorientation) given by U = S2|D| for integer spin

systems and U = (S2-%4)|D| for half-integer spin systems. A better appreciation of the

system is obtained by the realization that the ground state S = 10 spin is split into 21

sublevels (depicted in Figure 1-3) to give a double well potential in zero applied magnetic

field. This splitting is a consequence of the first term (D~S, axial ZFS term), in the

Hamiltonian (eq 1-1) based on the "giant spin model" for molecular systems."

1 = DS + E(. + .)+ p,H-g-S ~+4 r,+~' 11









Because the value of the axial ZFS parameter D for a SMM is negative (-0.50 cml

for Mnlzac), the ms = 10 sublevels lie lowest in energy while the ms = 0 sublevel lies

highest (Figure 1-3). Consequently, there is a potential energy barrier between the "spin-

up" (ms = -10) and "spin-down" (ms = +10) orientations of the magnetic moment. The

energy of each sublevel is given as E (ms) = ms2D, giving rise to a barrier whose

magnitude is given by the difference in the energy between the ms = 10 and ms = 0

energy levels. Thus, to reverse the spin of the molecule from along the z (spin up) to the

+ : (spin down) axis of the molecule, a potential energy barrier, U = S2|D| 50 cm-l (as S

= 10, D = -0.5 cm l) for Mnl28C, must be overcome. For this reason, SMMs exhibit slow

magnetization relaxation at low temperatures. Experimental evidence for this behavior is

supported by the appearance of concomitant frequency-dependent in-phase (XM') and out-

of-phase (XMI") signals in AC magnetic susceptibility measurements, as shown in Figure

1-4 (left). Frequency-dependent ac signals are necessary but not sufficient proof that an

SMM has been obtained. The observance of hysteresis loops in magnetization versus DC

field scans, with coercivities increasing with decreasing temperature and increasing scan-

rates, is an unambiguous confirmation that an SMM has been obtained.9 Sometimes the

hysteresis loops have steps (Figure 1-4 right (bottom)), which are due to quantum

tunneling, and each step corresponds to a sudden increase in the magnetization relaxation

rate. The QTM can be of different kinds (Figure 1-4 right (top)) and these will be

discussed later.

From a magnetic point of view, SMMs are preferred over classical nanoscale

magnetic particles because they have a single, well defined ground state spin S which is a

true quantum spin system, and highly ordered assemblies of SMMs can be obtained in







10


crystalline form. At the same time, from a synthetic chemist' s point of view, these SMMs

can be easily synthesized at room temperature by solution methods and we can obtain a

60 Thermally assised





30 .~~ 4. 9f M,= -9
i I ~~M,= 10 Y Pure nantum" Myh~~ -10




1; Ilf i b) --~ Ict


--36 28Kf~






T(K) tuno(T)
Fiur 1-.(et n-hs 'T(tpadot-fpaey (otmA ssetblt
sigal fo aMnl Cmpex.(Rght Pssbletunelngmehansm (tp)fo
SM sad(otm yia ytrss op ihses o n2SM
coletin f ruy ondipesepatils f anscledienios hih av tu

solbiit (rte tha coloi fomtin inogncslet.Adiinly heta










conidered1-. SLevera In-ew strategies and SMts will be discussed in th en suingchapters.









Manganese carboxylate cluster chemistry finds importance in bioinorganic

chemistry as well, mainly because of the ability of Mn to exist in a range of oxidation

states (-3 to +7), with the most common ones being II, III and IV. Mixed-valency in Mn

ions is found both in coordination compounds as well as Nature. Thus, oxidation state

variability in Mn makes it well suited as the active site for redox reactions in a number of

metalloproteins and enzymes.16 Redox enzymes containing Mn can be classified into

different groups depending on their nuclearity. For example, mononuclear sites

containing one manganese ion are found in manganese superoxide dismutases and

dinuclear sites are found in manganese catalases.l7 A tetranuclear manganese cluster,

called the water oxidizing complex (WOC), comprising primarily oxide and carboxylate

peripheral ligation, resides at the active site of photosystem II (PS II).l It is a complex

aggregate of electron transfer proteins embedded in the thylakoid membrane of

chloroplasts in green plants, algae and cyanobacteria. As the name suggests, it catalyses

the light-driven oxidation of water to dioxygen. This reaction is responsible for

generation of almost all the oxygen on this planet (eq. 1-2). In addition to the synthesis of

dioxygen, WOC also releases protons, which are used to create a proton concentration

gradient across the membrane that drives the synthesis of ATP by ATP-synthase. The

electrons produced in the reaction are transferred through a series of e- carriers to

photosystem I (PS I), where they are eventually used for the fixation of CO2.

2 H20 4 e- 02 +4 H (1-2)

It is believed that four Mn per PS II are essential for water oxidizing activity and

they appear to be in close proximity to each other. The Mn complex is capable of cycling

between four distinct oxidation levels, labeled So through S4 in the pioneering work of









Kok;19 the Sn states contain Mn in various combinations of higher metal oxidation states

(III-IV). The catalytic cycle thus involves the four electron oxidation of the Mn

aggregate, and the latter can be thought of as a biological capacitor, storing not only

charge but oxidizing equivalents, with discharge of capacitor occurring during S4 to So

transition on oxidation of substrate to dioxygen (Figure 1-5 (left)).


or ~ ~ ~ ~ ~ y (IlsIV Mn(IIB Gin 165

EPR: MLS S, S, ParaelW polarization i 9

Asp 170
hv e"-A 4





EP signaligal
FigureM 1-5 (Left The Sl stt ceea pooe yKk o hxdtino ae.7
(Riht Aragmn f4M n aaom sona al)i h w
latestPR crsa stutue (se lae)o teWCofP I2
The Ststte (Mn 2 n2 ICS thrmlymsstbeithdakndstusrfrd
toasth ar-aape sat. oevra udestnin o temehaisi dtal
rearin this caayi prcs sgetyhnee u o h bec fpeiesrcua







figrst5 generatio crysta strutre shm s confirmed the locfrth action of th erncla ncuter.7










but the exact orientation and structure of the Mn complex and its intermediate ligand

environment was not clear. Very recently another crystal structure appeared at 3.5 A+


b o

Gin 165 ~
CP43-Arg 35: Tyr

Ala 344 L
CP43-GIU5 Ca .2.5 1 r
Aspl701. 'r 7Glu 189 o
2.4IF--- 2. 2.1- &t~ C
S2.5\ 2.9 2.2 2

Asp 6Glu 342 3324
His 337 1~5



~16 --H190 Normal
Tyrz
2H2O XI C-terrn
X Ca
Possible reaction E8 04
siteE19 0 D4
D170 -Mn4 05 4 M~n;

W, n3 2 H332

E342Z ,061 H337 r,
J 65 Fiur 10 Clusteer arrangeent ofr 4 Mn and I Ca brridged by
H'Lumen Oaroms and conlsistent with the Mal- and Sr-EXA f-S dichroism.
Figure 1-6. (Left) Crystal structure of the WOC at 3.0 A~ (top)" with the 4 Mn labeled 1-4
in pink, and the 3.5 A+ Mn4Ca cubane-containing crystal structure (bottom).22
(Right) Possible Mn4CaOx topologies suggested by biophysical studies.27a

which assigns most of the amino acids in the protein, identifies 4Mn and ICa ions and

proposes the metal coordination number and geometry. In particular, the authors

conclude that the WOC is a [Mn3CaO4] cubane with the fourth Mn (extrinsic Mn)

attached to one of the cubane oxygen atom (Figure 1-6 (left (bottom)). This crystal

structure by the Barber group, postulating a Mn3CaO4 cubane-like core for the WOC, is

not accepted unanimously because it is believed that radiation damage to the crystals took









place during X-ray data collection.23 However a very interesting development in this

particular report, when compared to the earlier structural elucidations of the WOC, was

the detection of a Ca atom intimately associated with the tetranuclear Mn cluster. It had

long been known that the WOC requires Ca2+ for activity (it acts as a cofactor),24 and

calcium EXAFS (extended X-ray absorption fine structure) studies of the WOC had

revealed a Mn---Ca separation of ~3.4 A+.25 The most recent and "complete" crystal

structure of the WOC by the Zouni group26 aIt 3 .0 A concurs with the fact that the WOC

is a pentanuclear heterometallic site and suggests that the Mn ions are arranged in a "3 +

1" fashion (Figure 1-6 (left (top)), as proposed by the first two studies. The Mn4Ca

structure (Figure 1-6 (left (top)) has Mn1 and Mn3 (one is III and another IV oxidation

state), along with Mn2 (III oxidation state; ligated by the carboxy-terminal carboxylate of

Ala 344) and Ca form a trigonal pyramid, to which is attached the extrinsic fourth Mn4

(IV oxidation state; ligated by Asp 170). Various biophysical studies have also suggested

similar Mn4Ca ClUSter topologies (Figure 1-6 (right)).27 At the current resolution, smaller

ligands such as C, H, O, H20, N and Cl cannot be confidently located. However, both the

studies at 3.0 A+ and 3.5 A+ have identified Ca as being part of the Mn complex using

anomalous diffraction data. Thus, although there is still an ambiguity about the Mn4Ca

structure obtained from crystallography due to radiation damage during X-ray data

collection,23 there is little doubt that the WOC is a heterometallic [Mn4CaOx] cluster.22' 26

The availability of inorganic Ca/Mn complexes to act as synthetic models of the

WOC would represent an important step forward in understanding the magnetic and

spectroscopic properties of the native site and the mechanism of its function. Many

groups have in the past applied the Synthetic Analogue Approach28 to the WOC and a










plethora of Mn4 COmplexes have been synthesized. However, new bioinorganic modelling

approaches to the WOC involve synthesizing heterometallic Mn/Ca complexes as

opposed to homometallic Mn complexes. Therefore, as part of our ongoing interest in

obtaining synthetic models of the WOC and its various modified forms, we wanted to

investigate mixed Ca/Mn chemistry extensively. Additionally, Mn clusters have been the

primary source of SMMs, and it has become ever-increasingly important to devise new

synthetic routes towards clusters with interesting magnetic properties and higher blocking

temperatures. A totally new approach towards SMMs is synthesizing mixed-metal

complexes. This approach has the benefits of the spin as well as the anisotropy not

canceling, thereby increasing the probability of the resulting polynuclear complex

displaying SMM behavior.

Thus, the primary goal of the research featured in this dissertation is the

development of new synthetic routes aimed at the preparation of novel heterometallic

complexes incorporating Mn, whose relevance span the magnetic materials as well as the

bioinorganic research areas. Chapters II-IV report the syntheses of Mn containing

heterometallic complexes of lanthanides and actinides. The magnetic properties of these

are described in detail. The subsequent chapters, V and VI are relevant to the

bioinorganic modeling of the WOC of PS II, and they detail the relvance of mixed Sr/Mn

and Ca/Mn complexes. The X-Ray Absorption Spectroscopy of these complexes and the

comparisons thereof with the WOC are also reported in these chapters.














CHAPTER 2
SINGLE-MOLECULE MAGNETS: A NOVEL FAMILY OF Mn" / Ce'V
COMPLEXES WITH A [MnsCesOs]12+ CORE

2.1 Introduction

One of the motivating themes in polynuclear cluster chemistry research is the

design of high-nuclearity manganese carboxylate clusters which can function as

nanoscale magnetic materials. Since these species are molecular in nature, they fall in the

nanoscale regime and since some of them display superparamagnet-like slow

magnetization relaxation these are termed as single-molecule magnets (SMMs). Thus, an

SMM represents a molecular approach to nanomagnets.9 Such molecules thus behave as

magnets below their blocking temperature (TB), exhibiting hysteresis in magnetization

versus dc field scans. This behavior results from the combination of a large ground spin

state (S) with a large and negative Ising (or easy-axis) type of magnetoanisotropy, as

measured by the axial zero-field splitting parameter D. This leads to a significant barrier

(U) to magnetization reversal, its maximum value given by S2|D| or (S2 %4)|D| for integer

and half-integer spin, respectively. 9, 29 However, in practice, quantum tunneling of the

magnetization (QTM) through the barrier via higher lying M~s levels of the spin S

manifold results in the actual or effective barrier (Gerr) being less than U. The interest in

SMMs for scientists in various disciplines is stimulated by not only their aesthetically

pleasing structures but their ability to display classical magnetic bistability as well as

quantum properties 29 The first SMM discovered was [Mnl2012(O2CMe)16(H20)4 ,29' 9

(hereafter referred to as "Mnl2-acetate") which possesses an S = 10 ground state; together









with the continually growing family of [Mnl2012(O2CR)16(H20)4] (Mnl2; R = various)

molecules, these clusters are still the best and most thoroughly studied SMMs to date.30

Ever since, several types of SMMs have been discovered, most of them containing

primarily Mn"' ions.31 However, there have been only a few isostructural families of

SMMs studied to date, most notably the Mn4 defect-dicubane,32 Mn4 cubanes,33 and

Mnl2 wheel complexes.34

As part of our continuing search for new synthetic routes towards novel structural

types which can function as SMMs, we have joined on-going efforts in mixed-metal

cluster chemistry. Recently, there has been a spurt of research activity in the scientific

community towards heterometallic systems which can behave as SMMs.35 We and our

coworkers had contributed to this relatively nascent field with the successful

characterization of the MnsCe,36 Mn11Dy4,37 Mn2Dy2,38 and Fe2Dy239 SMMs. Indeed, we

had earlier reported the template synthesis of the MnsCe SMM which possessed an S =

16 ground state spin. Although the Ce ion (Ce'V) in the complex was diamagnetic, it

acted as a template around which 8 ferromagnetically coupled Mn ions wrapped. Since

the ground-state spin was then the largest for any Mn species, we carried out a detailed

investigation towards obtaining similar structural types with interesting magnetic

properties. We herein report the successful synthesis, structure, magnetic characterization

and reactivity of four MnsCe complexes and demonstrate that they are new SMMs with

spin-variability within the family.

2.2 Results and Discussion

2.2.1 Syntheses

[MnsCeOs(O2 Me)12 H20)4] -4H20 (4) was originally obtained serendipitously

from a solution of [Mn6CeO9(O2CMe)9(NO3)(H20)2 40 in MeCN/Et20 that had been left









undisturbed for some time. However, once the identity of 4 was know it became crucial

to develop a rational synthetic procedure towards obtaining 4 in high yield. One very

attractive reaction strategy was to use the linear polymer

{ [Mn(OH)(O2CMe)2]- (MeCO2H)-(H20)},,,41 which might wrap around the oxophilic Ce4

ion as Ce4+-OH contacts develop. Thus, with a Ce:Mn ratio of 1:8 and OH

deprotonation, this encirclement could in principle give 4, since complex 1, whose

structure is shown below, provides all the required components needed to obtain 4. Thus,

CH3 CH3 CH3 CH3
O 5C'bO O ,C O O C OO ,C
OH-M~i- -OH-M -OH-M -OH


CH3 CH3 CH3 CHI3

the feasibility of the above-stated hypothesis was tested by reacting 1 with

(NH4)2Ce(NO3)6 in MeCN as depicted in eq 2-1 below.

(8/n)[Mn(OH)(O2CMe)2] n C4+ + 4 H20 [MnsCeOs(O2CMe)12(H20)4]

+ 4 MeCO2H + 4 H+ (2-1)

Indeed, the reaction resulted in the formation of 4 in 55% isolated yield. The

magnetism studies of 4 revealed that it had an S = 16 ground state spin. However, the

complex had considerable intermolecular interactions and a small D value. Thus, we

decided to synthesize derivates of 4 with the obj ectives being two-fold: i) Block the

intermolecular interactions primarily occurring via the bound H20 molecules by

replacing the terminal ligation with other chelates. ii) Obtain derivatives of 4 with bulky

carboxylates that would, besides blocking the intermolecular interactions, help in

flattening the MnsOs loop to a more planar configuration thereby increasing the

magnitude of the small D value which complex 4 possessed (primarily because the JT









axes were perpendicular videe infra)).

Among the various Mn" sources which could be explored to obtain the 8Mn"

compound, [Mn30O(O2CMe)6 97)3]i (2Mn ", I Mn") and(NBun4) [Mn402(O2CPh)9(H20)]

(4Mn"') seemed particularly attractive, as they have been known to give higher nuclearity

complexes.42, 43 Hence, the reaction of [Mn30(O2CMe)6 97)3] with Ce4+ in MeCN in an

approximately 3:1 ratio gave [MnsCeOs(O2 Me)12 P)4]-3 C4H202 (5) as depicted in eq

2-2 below. However, it should be noted that varying the Mn3:Ce ratio from 2: 1 to 5:1

gave the same product, although the yield was optimized when the ratio of 8:3 was used

as stated in eq 2-2.

8 [Mn30(O2CMe)6 97)3] + 3 Ce4+ + 16 H20 3 [MnsCeOs(O2CMe)12 97)4] + 12 py +

12 MeCO2H + 32 H + 20 e- (2-2)

The synthetic strategy shown above proved successful in isolating a MnsCe

complex in which the 4 terminal water molecules had been replaced by pyridine (py)

molecules, thereby reducing intermolecular interactions. In order to flatten the MnsOs

loop, synthesis of the benzoate version of the MnsCe complex was sought. This task was

achieved by reacting 3 (a tetranuclear Mn"' complex) with (N\H4)2Ce(NO3)6 in a 1:1 ratio.

Varying the Mn4:Ce ratio to 1:2 also led to the isolation of the MnsCe benzoate complex

( [MnsCeOs(O2CPh)12 MO N)4] [MnCeOs(O2CPh)12(dioxane)4] ) (6). COmplex 6 is

isostructural with complexes 4 and 5, the difference being that the unit cell has 2 MnsCe

complexes; one having 4 MeCN as terminal ligands and the other with 4 1,4-dioxane

ligands providing terminal ligation. The manifestation of two complexes co-crystallizing

indicates the complexity of this reaction with several species likely to be in equilibrium in

the reaction solution. Thus, factors such as relative solubility, lattice energies,









thermodynamics, crystallization kinetics and others undoubtedly determine the identity of

the isolated product.

The successful isolation of complex 6 encouraged us to seek the syntheses of

MnsCe complexes with even bulkier carboxylate groups. For this purpose, we decided to

employ diphenylacetic acid (Ph2CHCO2H). Such a bulky carboxylate would not only

separate the molecules thereby reducing intermolecular interactions, but would also cause

strain in the central MnsOs core thus flattening the loop. The synthetic technique used to

obtain [MnsCeOs(O2CCHPh2)12 H20)4] (7) WAS our standard ligand substitution

reaction,"1 which has been successful in Mnl2 chemistry. The ligand substitution reaction

(eq 2-3) is an equilibrium that must be driven to completion by (i) using a carboxylic acid

with a much lower pKa than that of acetic acid (4.75); and/or (ii) using an excess of

RCO2H; and/or (iii) removing the acetic acid as its toluene azeotrope. Hence, 20 equiv. of

diphenylacetic acid were reacted with complex 4 and the free acetic acid removed under

vacuum as its toluene azeotrope. Although, the reaction proceeded with 12 equiv. of the

acid too, as summed up in eq 2-3; the extra acid is generally needed to ensure complete

[MnsCeOs(O2CMe)12(H20)4] + 12 Ph2CHCO2H [MnsCeOs(O2CCHPh2)12(H20)4]

12 MeCO2H (2-3)

carboxylate substitution." The presence of the extra incoming acid group ensures (by Le

Chatelier' s principle) that the equilibrium of the reaction in eq 2-3 is broken and pushed

forward and a pure product with complete ligand substitution is obtained. Also, the pKa

of diphenylacetic acid (3.94) is lower than that of acetic acid (4.75) thereby facilitating

the carboxylate substitution reaction and isolation of complex 7.









2.2.2 Description of Structures

2.2.2.1 X-ray crystal structures of complexes 4 7

PovRay representations of the labeled crystal structures of 4 (Figure 2-1), 5 (Figure

2-2, left), 6 (Figure2-2, right), and 7 (Figure 2-3) are depicted in the indicated figures.

The common [MnsCeOs] core present in complexes 4-7 is shown in Figure 2-4 (left). The

crystallographic data and structure refinement details for complexes 4 7 are collected in

Table 2-1. Selected bond distances and angles for complexes 4 and 5 are listed in Table

A-1. The bond distances and angles for complex 7 are listed in Table A-2.

Complex 4-4H20 crystallizes in the tetragonal space group IT with crystallographic

S4 Symmetry. The cluster contains one Ce'V and eight Mn ions bridged by eight CL3-02-

and twelve CH3CO2- grOups. The structure of 4 can be described as a non-planar saddle-

like [Mn" s(CL3-O)s]s+ loop attached to a central Ce'V ion via the triply bridging oxides of

the loop. Peripheral ligation around this central [MnsCeOs]12+l00p is provided by 8 syn,

syn, doubly- and 4 triply-bridging acetate groups. Four H20 molecules (O6 and its

symmetry counterparts in Figure 2-1) terminally ligate on four of the Mn"' (Mn2 in Fig.

2-1) ions. The central Ce ion is octa-coordinated, with the Ce-O bond lengths (2.29-2.37

A+) being typical for eight-coordinate Ce"V.44c All the Mn ions are hexa-coordinate with

near-octahedral geometry and display Jahn-Teller (JT) elongation axes videe infra), with

the JT bonds being at least 0. 1 0.2 A+ longer than the other Mn" -0 bonds, as expected

for high-spin Mn"' ions. Nevertheless, the metal oxidation states were also verified by

bond-valence sum calculations (BVS; see chapter III for theory on BVS), and charge

considerations.44 The complex contains four unbound water molecules as solvent of

crystallization. There are strong inter- and intra-molecular hydrogen-bonding interactions

involving the lattice and bound water molecules as well as O atoms from










oxide and carboxylate ligands.

Complex 5-3C4HsO2 CryStallizes in the tetragonal space group P42/n with a

crystallographic S4 aXiS. The cluster is isostructural with complex 4, with the difference

being that terminal ligation on four of the Mn"' ions is provided by pyridine (py)

molecules rather than H20, as is the case in complex 4. Additionally, there are three 1,4-

dioxane (C4HsO2) mOlecules as solvents of crystallization. Thus inter/intra-molecular

interactions which were present in 4 because ofbound/unbound water have been nullified

in complex 5. This occurs as a consequence of the absence of any H atoms attached to an

electronegative atom; the py as well as the dioxane do not have any hydrogen connected

to N / O atoms. Thus, H-bonding does not occur in complex 5.










(. Ce O Mn


Mn1 0











Figure 2-1. PovRay representation at the 50% probability level of the X-ray crystal
structure of 4. Color scheme: Mn green, Ce cyan, O red, C grey. Hydrogen
atoms have been omitted for clarity. The complex has a four-fold inversion
center.









Complex 6-12C4HsO2-4MeOH crystallizes in the tetragonal space group P4n2 and

the unit cell contains two MnsCe clusters; one with four MeCN molecules providing

terminal ligation and the other with four 1,4-dioxane (C4HsO2) mOlecules providing

terminal ligation. Hence, complex 6 is formulated as { [MnsCeOs(O2CPh)12(MeCN)4]

[MnsCeOs(O2CPh)12 C4HsO2)4]}). Figure 2-2 (right) depicts the MnsCe benzoate sub-

cluster with MeCN providing the terminal ligation. For each of the co-crystallizing

MnsCe clusters of 6, peripheral ligation is provided by eight doubly- and four triply-

bridging benzoate groups. The central [MnsCeOs]12+ COre found in 4 and 5, is also

retained in complex 6.










Figure -2. Povay reprsentaton at te 50% pobabilty ee o hXrycrsa

strutur of 5 lf)ad6(ih).Clrshm:M ren ecaNdr
blue,~- O eCge. tm a e"S benoitdfo lrt.Th opee
haeafu-od ymtyai wit an ivrio cener

Cope -42-C22-MC rstlie n h oe smerymncic
spce gru 2/,adteaymti nt cotisteetr n Ce cuter ope
isfomlaeda [nCe'(2C~212H04 n thus is ~isostrutrlwtope
4, ~e, onain acetrl e" onwhchisboud y 0-0-~ to8 n inswhc










together manifest themselves as a non-planar saddle-like loop. Terminal ligation is

provided by 4 water molecules, as is the case with complex 4. However, peripheral

ligation in 7 is provided by eight doubly- and four triply-bridging diphenylacetate groups

(Ph2CHCO2 ) (See Fig. 2-3).


[c.


Figure 2-3. PovRay representation at the 50% probability level of the X-ray crystal
structure of 7. Color scheme: Mn green, Ce cyan, O red, C grey. H atoms
have been omitted for clarity.

The presence of the big, fat phenyl groups of the bulky carboxylate group in 7

causes a greater degree of separation between individual clusters. In fact, the MnsCe

clusters of 7 are very far apart from each other as was evidenced from the packing

diagram of 7. The fact that individual clusters are aloof and separated from neighbors

might help in improving the magnetic properties of this complex (see later). However,





















Figure 2-4. (Top) Comparison of the [MnsCeOs]12+ COre (left), with the [Mnl2012 16+ COTO
(right). (Bottom) The common core of complexes 4, 5, 6 and 7 depicting the
near perpendicular alignment of the Jahn-Teller (JT) pairs. Thick black bonds
denote JT elongation axes. Color scheme: Mn green, Mn'V purple, Ce cyan,
O red, C gray.

there are some hydrogen bonding interactions amongst the water molecules.

2.2.2.2 Structural Comparison of complexes 4 7

The structures of complexes 4-7 are overall very similar, differing only in the

nature of one or two peripheral bridging ligands videe supra). Hence, complex 4 is

formulated as [MnsCeOs(O2CMe)12(H20)4] and a variation in the terminal ligation from

H20 to py gives complex 5. The benzoate version of complex 4 (complex 6) has

MeCN/dioxane providing terminal ligation. Finally, complex 7 is the diphenylacetate

version of complex 4. One-common structural motif which is conserved through 4-7 is

the [MnsCeOs]12+ COre. This core depicted in Figure 2-4 (left) consists of eight [MnO2 0]

rhombs which inter-connect with each other via the eight shared triply bridging oxides.

Thus, within this description the central core looks like a space-shuttle with eight "flaps"

and the non-planar arrangement of these "wings" give the saddle-like loop structure

arrangement to the core. The eight Ce'V-O bonds are undoubtedly crucial in the

formation of the [MnsOs] loop which engulfs the central Ce ion. Thus, the Ce ion acts as

a template around which forms the non-planar metal-oxo loop. The four Mn2 atoms

(Figure 2-1) occupy the corners of an almost perfect tetrahedron (Mn2-Ce-Mn2' =









109.430), whereas the four Mn1 ions form a severely distorted (flattened) tetrahedron

(Mnl-Ce-Mnl' = 92.020). Within this description, the Ce'V ion occupies the centre of

both the tetrahedra.

Interestingly, the [MnsOs] loop found in these complexes is very similar to that in

[Mnl2012(O2CMe)16(H20)4] (Mnl2-ac)" which has a central [Mn404] cubane instead of a

smaller Ce atom (Figure 2-4 (center)). Therefore, both of them have a non-planar ring of

8 Mn ions which are linked by eight CL3-02- to the central unit. Besides the difference in

the central unit, Mnl2-ac has four more carboxylates providing peripheral ligation than

the MnsCe complexes. A much clearer view of the cores of both of these complexes is

seen in Figure 2-4 whereby the greater folding in the MnsCe complex is accentuated. The

eight Ce-O bonds which connect the Ce to the Mn atoms are indisputably crucial to the

formation of these complexes and cause a greater folding of the [MnsOs] ring in the

MnsCe complexes, than in Mnl2. Thus, a loop is created in the MnsCe complexes rather

than the non-planar ring found in Mnl2 COmplexes.

The relative alignment of the Jahn-Teller (JT) axes is very important with respect

to the magnetic properties displayed by the complex; it determines the anisotropy in the

molecule, or in other words the magnitude of the ZFS term D. As already stated earlier,

some of the JT axes are nearly perpendicular. Figure 2-4 (right) depicts the orientation of

the JT axes which are shown as thick black bonds. For example the Mn2-04-Mn1 angle

in Figure 2-4 (right) is 83.90. Indeed, all the eight JTs occur in sets of two with the Mn-

O-Mn angle in the range of 800-840 for all the MnsCe complexes. Within the four sets of

perpendicular JTs, each set shares the doubly bridging oxygen of the four triply bridging

carboxylates as a common vertex. Thus, all the eight JT axes originate from the doubly












Table 2-1. Crystallographic data for 4-4H20, 5-3C4HsO2, 6-12C4HsO2*4MeOH, and 7-4H20-2CH2C 2-3MeCN.
4 5 6 7
Formula C24H52040MnsCe C56HsoN4038MnsCe C236H2640100Mnl6Ce2 C176H161C 4N3040MnsCe
fw, g/mol 1560.30 1996.87 5859.88 3679.52
Space group Ti P42/n P4n2 P21/n
a, A 23.947(6) 12.7367(6) 22.5647(6) 19.4171(12)
b, A+ 23.947(6) 12.7367(6) 22.5647(6) 31.3568(19)
c, A+ 9.953(5) 23.485(2) 26.3631(13) 27.5856(17)
ao 90 90 90 90

/7, 90 90 90 95.884(2)
Y, 0 90 90 90 90

V, A3 5708(4) 3809.8(4) 13423.2(8) 16707.2(18)
Z 4 2 2 4

T, K 100(2) 173(2) 173(2) 173(2)
Radiation, 8+a 0.71073 0.71073 0.71073 0.71073
peaic, g/cm3 1.816 1.697 1.784 1.463
C1, mml 2.584 1.953 1.168 0.993
R1 b~c 0.0899 0.0497 0.0367 0.0891
wR2 d 0.2163 0.1165 0.1083 0.2136
a Graphite monochromator. blI > 20(I. R1 = 100C(||Fo| |F,||)/|Fol. d wR2 = 100[C[w(Fo2 Fe2 2]/ C[w(Fo2 2 11 2, 1[2 Fo2)
[(ap)2 +bp], where p = [max (Fo2, O) + 2Fe2]/3.









bridging oxygens of the four triply bridging carboxylates and terminate in four cases on

the singly bridging oxygen of another triply bridging carboxylate group. In the remaining

four cases, the JTs terminate on the atoms (O/N) which provide terminal ligation.

Therefore, considerable magnetoanisotropy is expected to get cancelled for these

complexes, consequently resulting in a small D value, as was initially reported for

complex 4.36

2.2.3 Magnetochemistry of Complexes 4, 5 and 7

2.2.3.1 DC studies

Solid-state variable temperature magnetic susceptibility measurements were

performed on vacuum-dried microcrystalline samples of complexes 4, 5 and 7, which

were suspended in eicosane to prevent torquing. The dc magnetic susceptibility (XM) data

were collected in the 5.0-300 K range in a 0.1 T magnetic field and are plotted as XMT vs

T in Figure 2-5. For 4, the XMT value of 39.36 cm3mOl-1K at 300 K remains more or less

constant until 70 K. Then, it steadily increases with decreasing temperature to reach

69.28 cm3mOl-1K at 5.0 K indicating a large ground-state spin for 4. For 5, the XMT value

of 28.31 cm3mOl-1K at 300 K remains steady till 70 K. Then, it starts decreasing with

decreasing temperature to finally reach a value of 17.08 cm3mOl-1K at 5.0 K indicating a

small ground-state spin for 5. For 7, XMT fractionally increases from 28.00 cm3mOl-1K at

300 K to 32.47 cm3mOl-1K at 70 K, and then decreases to a minimum value of 19.85

cm3mOl-1K at 5.0 K indicating a relatively small ground state for 7. The spin-only value

of eight non-interacting Mn" ions is 24.00 cm3mOl-1K assuming g = 2 (Ce'V is

diamagnetic, fo). For 4 the XMT value of 39.36 cm3mOl-1K at 300 K is much higher than

the spin-only value and XMT increases with decreasing temperature suggesting

ferromagnetic interactions within the molecule and a large ground-state spin value.










Conversely, for both 5 and 7, the XMT value decreases with decreasing temperature,

suggesting the presence of overall strong, predominantly antiferromagnetic exchange

interactions within these molecules and consequently a small ground-state spin for them.


80
*Complex 4
~O Complex 5
V Complex 7
S60-


E




rx 20 O




0 50 100 150 200 250 300

Temperature (K)

Figure 2-5. Plots of XT vs T for complexes 4, 5 and 7. XM is the dc molar magnetic
susceptibility measured in a 1.0 kG field.

With eight Mn" centers in 4, 5 and 7, total spin values range from 0 to 16.

However, due to the size and low symmetry of the molecules, a matrix diagonalization

method to evaluate the various Mn pairwise exchange parameters (Jij) within the MnsCe

molecules is not easy. Similarly, application of the equivalent operator approach based on

the Kambe vector coupling method45 iS not possible. Therefore, we focused only on

identifying the ground state S as well as the ZFS term D values, as in any case these

would dominate the low temperature studies we performed videe infra). Hence,

magnetization (M)1 data were collected in the magnetic field and temperature ranges 0.1-7

T and 1.8-10 K in order to determine the spin ground states of complexes 4, 5 and 7. In











the nomenclature, N is Avogadro's number, CLB is the Bohr magneton, T is temperature

and H is the applied magnetic field. The obtained data are plotted as M/NCLB (reduced


magnetization) vs H/T in Figure 2-6 for complexes 4 (left) and 7 (right). For a system

occupying only the ground state and experiencing no ZFS, the various isofield lines

would be superimposed and M/NCLB WOuld saturate at a value of gS. The non-

superposition of the isofield lines in Figure 2-6 is indicative of the presence of strong

ZFS.










34- 02T
26 i 0 3 T
srH 05T
*6T A 06T
v7T 2-* 07T
24 -1 6ting 0 8 T
tting

10 15 20 25 30 3501 2345
H/T [kG/K] HIT [kGlK]


Figure 2-6. Magnetization (M) vs field (H) and temperature (T) data, plotted as reduced
magnetization (M/NCLB) vs H/T, for (left) complex 4 at applied fields of 3, 4, 5,
6 and 7 T and in the 1.8 10 K temperature range, and (right) for complex 7 at
applied fields of0. 1- 0.8 T range and in the 1.8 10 K temperature range. The
solid lines are the fit of the data; see the text for the fit parameters.

The data were fit, using the program MAGNET,46 by diagonalization of the spin

Hamiltonian matrix assuming only the ground state is populated, incorporating axial

anisotropy (Di- ) and Zeeman terms, and employing a full powder average. Thus, the


complexes are modeled for the magnetization fit as a "giant-spin" with Ising-like

anisotropy. The corresponding Hamiltonian is given by eq 2-4, where D is the anisotropy

H = Di- + gfBfl0SH (2-4)









constant, flB iS the axial Bohr magneton, S=is the easy-axis spin operator, g is the

electronic g factor, puo is the vacuum permeability, and His the applied longitudinal field.

The last term in eq 2-4 is the Zeeman energy associated with an applied magnetic field.

The "giant-spin" model and the same MAGNET program is used in the magnetization

fits for complexes reported in the ensuing chapters, and for the sake of brevity, the whole

process will not be repeated again. Instead, just the spin, D value and the isotropic g

values will be mentioned. For 4, the fit (solid lines in Fig. 2-6 (left)) gave S= 16, D = -

0. 10 cm-l and g = 1.98. Thus, ferromagnetic couplings within complex 4 aligns the 8

Mn spins parallel (as has been observed in Mnlzac, vide supra), leading to the second

highest spin ground state for a Mnx species yet reported. The largest spin yet observed for

a Mn complex is the S = 51/2 TepOrted for a Mn25 COmplex.47 The D value of -0. 1 cml is

consistent with the complex having the JT axes perpendicular as stated earlier, and g is <

2, as expected for Mn. When data collected at fields < 3.0 T were included, a satisfactory

fit could not be obtained, which is understandable as the complex has weak

intermolecular interactions, thus higher fields are needed to overcome those interactions.

For complex 5, attempts were made to fit the magnetization data collected in the

0.1 7 T and 1.8-10 K temperature ranges. A satisfactory fit could be obtained only

when data collected in the 0. 1-2 T applied field range were used. The fit of the data gave

a ground state of S = 5, D = -0.30 cm-l and g = 1.83. For complex 7, attempts were made

to fit the magnetization data collected in the 0.1 0.8 T range and 1.8-10 K temperature

ranges. A best fit of the data (Figure 2-6 (right)) yielded a ground state of S = 6, D = -

0.34 cm-l and g = 1.89. The ground states obtained for complexes 5 and 7 are in

agreement with the dc magnetic susceptibility data. However, there are low-lying









excited-states (relative to kT) which complicate the fitting; this was also confirmed from

the sloping nature of the in-phase ac susceptibility data videe infra). We have found that

poor quality fits of the magnetization versus H and T plots are a common problem in

manganese chemistry when the Mnx species is of high nuclearity and there is thus a high

density of spin states resulting from the exchange interactions amongst the many

constituent Mn ions. Thus, low-lying excited states are populated, even at these relatively

low temperatures, and/or the M~s levels from nearby excited states with S greater than that

of the ground state are being sufficiently stabilized by the applied dc field that they thus

approach or even cross the ground state levels; note that the fitting model assumes

population of only ground state levels. Population of the excited states will thus be

difficult to avoid even at the lowest temperatures normally employed. However, the spin

ground states obtained from the fittings were confirmed by the more reliable ac

susceptibility studieS48 described later.

The large ground state spin value obtained for complex 4 suggested that it may

have a barrier to magnetization reversal. Also, the combination of the S and D values

obtained for complexes 5 and 7 may enable them to display slow magnetization

relaxation characteristic of SMMs. The S and D values obtained for complexes 4, 5, and 7

suggest an upper limit to the energy barrier (U) to magnetization reversal of U= S |D| =

25.6 cml for 4, 7.5 cml for 5 and 12.2 cml for 7 respectively. However, the effective

barrier Genf, might be a little bit smaller because of quantum tunneling through the barrier.

Hence, ac susceptibility measurements were performed to investigate whether these

MnsCe complexes functioned as single-molecule magnets.









2.2.3.2 AC studies

In an ac susceptibility experiment, a weak field (typically 1 5 G) oscillating at a

particular frequency (u) is applied to a sample to probe the dynamics of the magnetization

(magnetic moment) relaxation. An out-of-phase ac susceptibility signal (XM") is observed

when the rate at which the magnetization of a molecule relaxes is close to the operating

frequency of the ac field, and there is a corresponding decrease in the in-phase (Xhl' T)

signal. At low enough temperature, where the thermal energy is lower than the barrier for

relaxation, the magnetization of the molecule cannot relax fast enough to keep in-phase

with the oscillating field. Therefore, the molecule will exhibit a frequency-dependent XMa

signal indicative of slow magnetization relaxation. The increase in the frequency-

dependent, imaginary XM" signal is accompanied by a concomitant decrease in the real

Xh' T signal. Frequency-dependent XM" signals are an important indicator of S101Vs.

Alternating current magnetic susceptibility studies were performed on vacuum-

dried microcrystalline samples of 4, 5 and 7 in the temperature range 1.8-10 K with a

zero dc field and a 3.5 G ac field oscillating at frequencies between 5-1000 Hz. The in-

phase (Xhl') component of the ac susceptibility was plotted as Xhl'T vs T. Similarly, the

out-of-phase Xhl" component was plotted as Xhl" vs T. Indeed, frequency-dependent

signals were seen in the in-phase as well as the out-of-phase ac susceptibility plots for all

of the three above-mentioned complexes. The strength of the signals varied and the

intensity in decreasing order was 4 >7 >5. Although a rise in the out-of-phase signal was

accompanied by a concomitant decrease in the in-phase signal, only tails, of peaks which

lie below the operating minimum (1.8 K) of our SQUID magnetometer were seen.

Nevertheless, the in-phase Xhl'T data proved useful for confirming the spin ground state









obtained from magnetization fits. For example, the Xhl'T value of complex 4 is 63.17

cm3mOl-1K at 10 K and increases steadily to 94.05 cm3mOl-1K at 1.8 K and seems to keep

on rising steeply to higher values below 1.8 K. Extrapolation of the data to 0 K indicates

a Xhl'T of ~ 125 which is in agreement with the ~133 cm3mOl-1K value obtained by

applying the formula Xhl'T = (g2/8)S(S+1) and using the g and S values obtained from the

magnetization fits videe supra). For complexes 5 and 7 the presence oflow-lying excited

spin states within close separation of the ground state was confirmed by the sloping

nature of the Xhl'Tys Tplots. The in-phase Xhl'Tvalue drops sharply and reaches 15.39

cm3mOl-1K for 5 and 18.15 cm3mOl-1K for 7 at 1.8 K. These values indicate an S of ca. 5

or 6 for these two complexes which is satisfyingly consistent with the dc magnetization

fit values.

Since all three complexes display frequency dependent ac signals which are an

indication of the superparamagnet-like slow magnetization relaxation of a SMM we

decided to investigate them further. Note however that these signals are necessary but not

sufficient proof that an SMM has been obtained because intermolecular interactions and

phonon bottlenecks can also lead to such signals.49 Thus, lower temperature (<1.8 K)

studies were carried out to explore this possibility further.

2.2.3.3 Hysteresis studies below 1.8 K

If complexes 4, 5 and 7 were indeed SMMs, they should exhibit hysteresis below

their blocking temperatures, Ty3, in a magnetization versus dc field plot. To investigate

this, magnetization vs. applied dc field data down to 0.04 K were collected on single

crystals of 4-4H20, 5-3C4H202 and 7-4H20-3MeCN-2C~CH2 2USing a micro-SQUID

apparatus.'o The observation of hysteresis loops in such studies represents the diagnostic











property of a magnet, including SMMs and superparamagnets below their blocking


temperature (TB). The observed magnetization responses for complex 4 are shown in


Figure 2-7 (left) at a Eixed sweep rate of 0.004 T/s and at different temperatures. The


hysteresis loops of complex 5 at field sweep rates of 0.280 T/s and 0.017 T/s and a

constant temperature of 0.3 K are shown in Figures 2-7 (right). Finally, the magnetization


responses for 7, at different temperatures and a Eixed Hield sweep rate (0.14 T/s) are

shown in Figure 2-8 (left) and at a Eixed temperature (0.04 K) and varying dc field sweep


rates in Figure 2-8 (right), respectively. In all cases, hysteresis loops are seen, whose


1 I I I I- 1-
0.004 T/s0.K

0.5 I -1 0.5-


0o -0
-0o04 K
-02K
-0.5 03 K -1 -0.5-
-04K
S 0 5K I 0 280 T/s
0 6 K 0 017 T/s
-1 I I I I I -I -1
-0.4 -0.2 0 0.2 0.4 -1.2 -0.8 -0.4 0 0.4 0.8 1.2
CloH (T) Ro H (T)


Figure 2-7. Magnetization (M)1 vs applied magnetic field (H) hysteresis loops: (left) for 4
in the temperature range 0.04-0.6 K and at a 0.004 T/s sweep rate; (right) for
5 at 0.280 and 0.017 T/s sweep rates and at a fixed temperature of 0.3 K. M~is
normalized to its saturation value, Ms, for both plots.

coercivities increase with increasing sweep rate and with decreasing temperature, as


expected for the superparamagnet-like properties of a SMM. The data thus confirm


complexes 4, 5 and 7 as new additions to the family of SMMs.

For 4-4H20 in Figure 2-7 (left) hysteresis loops are evident below 0.6 K, with the


dominating feature in the loops being the two-step profile and its variation with


temperature. This two-step profile and its broadening with decreasing temperatures are

characteristic of a weak intermolecular interaction between molecules, undoubtedly









mediated by the hydrogen bonds and dipolar interactions. Similar behavior has been seen

in the Fel9 SMMsS1 which also display intermolecular interactions. However, the above-

mentioned interaction in the MnsCe complex only perturbs the SMM behavior; it is too

weak to give a classical antiferromagnetically ordered network. An intermolecular

exchange parameter (J) of only 0.0025 K and an interaction energy of 0.65 K can be

calculated from the loops shown in Figure 2-7 (left). Thus, complex 4 behaves as an

SMM and the low temperature at which it shows magnetization hysteresis is clearly due

to the small D value, which is consistent with some of the Mn JT axes, the primary

source of the molecular anisotropy, being nearly perpendicular. For 5-3C4H202 hysteresis

loops at 0.3 K are shown in Figure 2-7 (right). Clearly the coercivity increases with

increasing scan rate, as expected for an SMM; the loops at 0.280 T/s are thicker than the

loops at 0.017 T/s. However, there is very little coercivity at H = 0 and this occurs

because of a fast tunnel transition which changes the magnetization direction very

rapidly. Thus, the effective barrier to magnetization relaxation reduces and hence the

hysteresis behavior is seen at lower temperatures (below 0.3 K). Nevertheless, the

hysteresis data are in agreement with the ac signals which were the weakest for complex

5 and the upper limit to magnetization reversal for 5 was also only 7.5 cml

For 7-4H20-3MeCN-2CH2C 2 in Figure 8, the dominating feature in the

temperature-dependent (top) as well as the sweep-rate-dependent (bottom) hysteresis

loops is the large step (corresponding to a large increase in magnetization relaxation rate)

at zero field due to quantum tunneling of the magnetization (QTM). Steps at other field

positions are only poorly resolved, probably due to broadening effects from low-lying

excited states and a distribution of molecular environments (and thus a distribution of





















-1.2 -0.8 -0.4 0 0.4 0.'8 1.2 -1.2 -0.8 -0.4 0 0.4 0.'8 1.2
Rog H (T) Lo H (T)

Figure 2-8. Magnetization (M)1 vs applied magnetic field (H) hysteresis loops for 7: (left)
in the temperature range 0.04-0.5 K at a 0. 14 T/s sweep rate; (right) in the
0.008-0.140 T/s sweep rate range at 0.04 K. M~is normalized to its saturation
value, Ms, for both plots.

relaxation barriers) caused by disordered lattice solvent molecules and ligand disorder. In

addition, intermolecular interactions (both dipolar and exchange) and population of

excited states can result in step broadening.52 COmplex 7 displays hysteresis below 0.5 K

and the plots in Figure 2-8 show increasing coercivities with decreasing temperature and

increasing sweep rates, as expected for SMMs. There is very little coercivity at H = 0.

However, one should remember that that the y-axis for the hysteresis loops of complexes

5 and 7 scan a range of -1.4 till +1.4 T and that of 4 depicts the range of -0.5 to +0.5 T

CRoH values. Unfortunately, the fast relaxation rate in zero field that results in the large

step at this position also prevents us from collecting magnetization vs time decay data

with which to construct an arrhenius plot and determine the effective barrier to relaxation

(Gerr). It must be stated though, that at a qualitative as well as to some extent a

quantitative level, the various dc, ac and hysteresis data are all consistent. Thus, for

complexes 4, 7 and 5 the upper limit to magnetization reversal was found out to be 25.6

cml >12.2 cml >7.5 cm- The strength of the frequency-dependent ac signals also

followed that same descending order and finally the temperature below which they










displayed hysteresis loops (0.6 K >0.5 K >0.3 K) was also consistent with the earlier

mentioned data. All these magnetic parameters are tabulated in Table 2-2 for comparison.

Table 2-2. Comparison of the magnetic parameters of complexes 4, 5, and 7.

Complex S D g U Ty3 (blocking T)a
(cm )> (cm )> (K)
4 16 -0.10 1.98 25.6 0.6
5 5 -0.30 1.83 7.5 0.3
7 6 -0.34 1.89 12.2 0.5


a Ty3 is the blocking temperature: Temperature below which hysteresis loops were
observed.
2.3 Conclusions

Convenient, high-yield synthetic routes towards obtaining a family of isostructural

MnsCe complexes with a common [MnsCeOs]12+ COre, have been developed. Peripheral

ligation has been varied within this family and new surrogates obtained with the end

obj ective of improving the magnetic properties; by blocking intermolecular interactions

and/or flattening the [MnsOs] loop to a more planar configuration. Complex 4 with an S

16 ground state spin possess the third largest S value reported for a Mnx species; the

highest and second highest being the S = 51/2 for a Mn25 SMM,47 and the S = 22 for a

high-symmetry Mnlo complex.97c COmplexes 5 and 7 possess an S = 5 and an S = 6

ground state, respectively. Although spin variability is present in this family of

complexes, each of them possess a sufficient combination of spin and anisotropy to

display superparamagnet-like slow magnetization relaxation and thereby function as

single-molecule magnets (SMMs). Hysteresis measurements confirm the addition of

complexes 4, 5 and 7 to the growing family of SMMs. Thus, in this chapter it has been









aptly demonstrated that a subtle change in the ligand environment has a huge impact on

the resulting magnetic properties within this family of complexes.

This work lucidly reasserts that synthetic manipulation around the metallic core by

organic groups can help in systematically studying the magnetic properties of SMMs, as

has been seen for Mnl2 COmplexes." It reiterates the advantages of the molecular

approach to nanomagnetism whereby standard chemistry methods give a greater deal of

control, when compared to the classical "top-down" nanoparticle approach.

2.4 Experimental

2.4.1 Syntheses

All manipulations were performed under aerobic conditions using chemicals as

received, unless otherwise stated. { [Mn(OH)(O2CMe)2] -(MeCO2H)-(H20)},, (1)53,

[Mn3(O2CPh)6 py)2(H20)] (2),54 [Mn30(O2CMe)6 97)3 ,54 and

(NBun4)[Mn402(O2C~h)9(H20)] (3),55 were prepared as previously reported.

[MnsCeOs(O2 Me)12 H20)4] -4H20 (4). Method A. To a slurry of 1 (2.00g,

7.46mmol) in MeCN (35ml) was added solid (N\H4)2Ce(NO3)6 (0.51g, 0.93mmol) and left

under magnetic stirrer for 8 h resulting in the formation of a brownish precipitate and a

reddish-brown solution, which were separated by filtration. To the filtrate was added

diethylether (40ml) and left under magnetic stirring for 5 more min. This solution was

filtered and the filtrate concentrated by evaporation to yield reddish brown crystalline

material which was washed with acetone and diethylether. The crystalline material was

identified as 4-4H20 and was obtained in 55% isolated yield. Anal. Calc (Found) for

4-4H20: C24H52040MnsCe: C, 18.47 (18.49); H, 3.36 (3.32) %. Selected IR data (KBr,

cm )~: 3392(s, br), 1576(s), 1539(s), 1444(s), 1029(w), 680(s), 657(m), 619(m), 589(s,

br), 551(m), 496(w), 432(w).









Method B. In MeCN (20ml) was dissolved [Mn6CeO9(O2CMe)9(NO3)(H20)2 40

(0.50g, 0.3 8mmol) and left under magnetic stirring for 20 min., filtered off and the filtrate

layered with diethylether (40ml). After two weeks the resulting solution was filtered and

the filtrate slowly concentrated by evaporation to yield reddish brown crystals of 4 which

were filtered and washed with acetone and diethylether and dried in vacuo. The product

was identified as 4 by IR. The yield was 5%.

[MnsCeOs(O2 Me)12 FT)4]3 C4H202 (5). To a slurry of [Mn30(O2CMe)6 97)3]

(1.00g, 1.29mmol) in MeCN (50ml) was added solid (NH4)2Ce(NO3)6 (0.24g, 0.43mmol)

and left under magnetic stirring for 30 min. The solution was filtered and the filtrate

layered with 50 ml of 1,4-dioxane (C4H202). After a week, nice square crystals of 5 were

obtained. These were washed with 1,4-dioxane, dried in vacuo and isolated in 60% yield.

Anal. Calc (Found) for 5-3C4H202: C56HsoN4038MnsCe: C, 33.68 (33.75); H, 4.04 (4.15);

N, 2.81 (2.54) %. Selected IR data (KBr, cm )~: 3429(s, br), 1576(s), 1540(s), 1444(s),

1119(w), 871(w), 679(m), 656(m), 619(m), 589(s, br), 551(m), 496(w), 430(m).

[MnsCeOs(O2CPh)12 MO N)4] [MnCeOs(O2CPh)12(dioxane)4] -12C4H202-4MeO

H (6). To a slurry of 3 (1.00g, 0.62mmol) in MeCN/MeOH (25ml/1ml) was added solid

(NH4)2Ce(NO3)6 (0.34g, 0.62mmol) and left under magnetic stirring for 20 min. The

solution was filtered and the filtrate layered with 25 ml of 1,4-dioxane (C4H202). After a

week, large dark red crystals of 6 were obtained. These were washed with 1,4-dioxane,

dried in vacuo and isolated in 45% yield. Anal. Calc (Found) for 6-12C4H202-4MeOH:

C236H2640100Mnl6 02: C, 48.37 (48.20); H, 4.54 (4.35) %. Selected IR data (KBr, cm )~:

3430(s, br), 1600(m), 1560(s), 1528(s), 1412(s), 1120(w), 872(w), 718(s), 683(m),

615(m), 573(s, br), 510(w), 422(w).









[MnsCeOs(O2CCHPh2)12 H20)4] -4H20-3MeCN-2CH2 2z (7). The carboxylate

substitution reactions which have been very successful for Mnl2 Systems30 wr

employed to synthesize 7. Therefore, diphenylacetic acid (0.54g, 2.5 mmol) was added to

a slurry of 4 (0.20g, 0. 12mmol) in MeCN (50ml) and stirred overnight. The resulting

solution was concentrated by using a toluene azeotrope to remove the free acetic acid.

The process of removing the free acid under vacuum was repeated thrice and the resulting

powder was re-dissolved in MeCN. Concentration of this solution by evaporation gave

nice black crystals which unfortunately were not suitable for X-ray diffraction. Hence,

recrystallization was achieved by dissolving the crystals in CH2 12 and layering with

heptanes. After a week, nice black crystals of 7-4H20-3MeCN-2C~CH2 2r WoTObtained in

60% yield. These were maintained in the mother liquor for X-ray crystallography and

other single-crystal studies, or collected by filtration, washed with heptanes and dried in

vacuo. The synthesis could also be performed by using 5 instead of 4, as the starting

material. The dried solid analyzed as 7-4H20. Anal. Calc (Found) for 7-4H20:

C168H148040MnsCe: C, 59.58 (59.45); H, 4.40 (4.25) %. Selected IR data (KBr, cm )~:

3429(s, br), 1590(m), 1550(s), 1527(m), 1494(w), 1404(s), 1032(w), 745(m), 697(s),

6 5 0(m), 5 8 0(s, b r), 4 3 3(w).

2.4.2 X-ray Crystallography

Data were collected on a Siemens SMART PLATFORM equipped with a CCD

area detector and a graphite monochromator utilizing Mo-Ku radiation (h = 0.71073 A+).

Suitable crystals of the complexes were attached to glass fibers using silicone grease and

transferred to a goniostat where they were cooled to 100K for complex 4 and 173 K for

complexes 5, 6, and 7 for data collection. An initial search of reciprocal space revealed a

tetragonal cell for 4, 5 and 6, and a monoclinic cell for 7; the choice of space groups I4,









P42/n, P4n2 and P21/n, respectively, were confirmed by the subsequent solution and

refinement of the structures. Cell parameters were refined using up to 8192 reflections. A

full sphere of data (1850 frames) was collected using the co-scan method (0.30 frame

width). The first 50 frames were re-measured at the end of data collection to monitor

instrument and crystal stability (maximum correction on I was < 1 %). Absorption

corrections by integration were applied based on measured indexed crystal faces. The

structures were solved by direct methods in SHELXTL6, 56a and refined on F2 USing full-

matrix least squares. The non-HI atoms were treated anisotropically, whereas the

hydrogen atoms were placed in calculated, ideal positions and refined as riding on their

respective carbon atoms.

The asymmetric unit of 4-4H20 consists of one-fourth of the MnsCe cluster lying

on an inversion centre and one H20 molecule of crystallization. Both of these are located

on a C4 TOtation axis. A total of 342 parameters were included in the structure refinement

using 2958 reflections with I > 2o(I) to yield R1 and wR2 of 8.99 % and 21.63 %,

respectively.

For 5-3 C4H202, the asymmetric unit consists of one-fourth of the MnsCe cluster (on

a 4-fold rotation axis) and a half dioxane molecule (located on a 2-fold rotation axis), and

a 1/ dioxane molecule (located on a 4-fold center). A total of 258 parameters were

included in the structure refinement on F2 USing 23703 reflections with I > 2o(I) to yield

R1 and wR2 of 4.97 % and 11.65 %, respectively.

The asymmetric unit of 6-12C4H202-4MeOH consists of a '/ dioxane MnsCe

cluster, a 1/ acetonitrile MnsCe cluster, three disordered dioxane molecules and one

methanol molecule of crystallization. All of these are on a four-fold rotation axis. The









solvent molecules were disordered and could not be modeled properly, thus program

SQUEEZE,56b a part of the PLATON56c package of crystallographic software, was used

to calculate the solvent disorder area and remove its contribution to the overall intensity

data. A total of 646 parameters were included in the structure refinement on F2 USing

6920 reflections with I > 2o(I) to yield R1 and wR2 of 3.67% and 10.83%, respectively.

The asymmetric unit of 7-4H20-3MeCN-2C~CH2 2COnsists of a MnsCe cluster, four

water molecules, three acetonitrile molecules and two dichloromethane molecules. Most

of the phenyl rings display considerable displacement motions but not large enough of a

disorder to allow to successfully resolve them. Consequently, large thermal parameters

are observed for their carbon atoms. Thus, they were refined with isotropic thermal

parameters only. The cluster has four coordinated water molecules; two on each side of

its plane. The two water molecules on each side along with two solvent water molecules

form a diamond shape as a result of Hydrogen bonding. The solvent water molecules

were located inside cavities created by the phenyl rings. The rest of the solvent molecules

were disordered and could not be modeled properly, thus program SQUEEZE,56b a part of

the PLATON package of crystallographic software, was used to calculate the solvent

disorder area and remove its contribution to the overall intensity data. A total of 1010

parameters were included in the structure refinement on F2 USing 74183 reflections with I

> 2o(I) to yield R1 and wR2 of 8.91% and 21.36%, respectively.

Unit cell data and details of the structure refinements for 4-4H20, 5-3C4HsO2,

6-12C4HsO2-4MeOH and 7-4H20-2CHzC2 2-3MeC are listed in Table 2-1.















CHAPTER 3
SINGLE-MOLECULE MAGNETS: SYNTHESES AND MAGNETIC
CHARACTERIZATION OF A NOVEL FAMILY OF HETEROMETALLIC
MANGANESE-LANTHANIDE COMPLEXES

3.1 Introduction

The current burgeoning research in nanosciences and nanotechnology is primarily

governed by the ideology of taking materials to the extreme limit of miniaturization.

Once we venture beyond the macro-, meso- and the micro-scale dimensions and reach the

nanoscale (and even subnano), interesting phenomenon are observed which cannot be

described by the classical properties of matter. Therefore, there has been great interest in

the scientific community in the study of nanoscale materials. One such interesting class

of nanoscale magnetic material is single-molecule magnets (SMMs). SMMs represent a

molecular approach to nanoscale magnetic particles. Since these molecules are magnets,

they show hysteresis like any classical magnet. However, their behavior is unlike

macroscale magnets because they display quantum tunneling of the magnetization

(QTM), a property seen mostly in the micro-, nano-scale and beyond. The magnetic

behavior in SMMs is due to the intrinsic, intramolecular properties of these species, and

is the result of the combination of a large ground state spin (S) value and a significant

magnetic anisotropy of the easy-axis (or Ising) type, as reflected in a negative value of

the axial zero-field splitting (ZFS) parameter, D. As a result, SMMs possess a significant

barrier to reversal (relaxation) of their magnetization vector, and the upper limit to this

barrier (U) is given by S2|D| and (S2-%4)|D| for integer and half-integer S values,

respectively. However, because of QTM the effective barrier Uegf is generally lower than









U. The slow relaxation results in imaginary (out-of-phase) magnetic susceptibility (yu")

signals in ac studies, and in hysteresis loops in magnetization versus applied de Hield

sweeps." SMMs thus represent a molecular (or bottom-up) approach to nanoscale

magnets, and thus differ significantly from classical (or top-down) nanoscale magnets of

metals, metal alloys, metal oxides, etc. These differences include monodispersity,

crystallinity, true solubility (rather than colloid formation), and a shell of organic groups

that prevents close contact of the magnetic cores with those of neighboring molecules,

and which can be varied using standard chemical methods. Since the initial discovery of

the revered [Mnl2012(O2CR)16(H20)4] family of SMMs,7 a number of other structural

types have been discovered, almost all of them being homometallic transition metal

clusters and the maj ority of them being Mn clusters containing at least some Mn" ions.91~

102b, 78 However, recently we and others have been exploring heterometallic clusters as

routes to novel SMMs. One, big advantage of heterometallic clusters is that the spins as

well as the anisotropy will probably not cancel in the resulting polynuclear complex.

Hence, some heterometal SMMs have been reported.'" Again, within this subclass a very

attractive route to SMMs with higher blocking temperatures is the synthesis of mixed

transition metal-lanthanide SMMs. In general, the presence of a lanthanide ion's (i) large

spin such as the S = 7/2 Of Gd3+ and/or (ii) large anisotropy as reflected in a large D value

could serve to generate SMMs with properties significantly different from their

homometallic predecessors. The Hield of transition metal / lanthanide SMMs has

flourished in the last year or so, with reports of Cu2Tb2, MnllDy4, Mn6Dy6, Mn2Dy2,

CuTb, Dy2Cu, and Fe2Dy2 SMMs amongst others.59 Additionally, single-molecule

magnetism has also been reported in homonuclear Ho, Dy, and Tb phthalocyanates.60









With the field of single-molecule magnetism now firmly established, new synthetic

methodologies and approaches need to be taken to obtain clusters with higher blocking

temperatures, TI3 (the temperature below which the molecule functions as a magnet). The

present work describes one such approach.

We had earlier reported our initial breakthrough in mixed Mn/Ln chemistry with

the synthesis of the [MnllDy40s(OH)6(OMe)2(O2C~h)16(NO3)5(H20)3 SMM which

displayed magnetization hysteresis and quantum tunneling.37 Additonally, in Chapter II

we also discussed a family of MnsCe SMMs, albeit we do not consider them as mixed

Mn/Ln SMMs as the Ln ion involved (Ce ') is diamagnetic. We herein describe the

extension of the earlier work with the successful syntheses, structures and detailed

magnetic characterization of the complete family of this MnllLn4 COmplexes (Ln = Nd,

Gd, Dy, Tb, Ho, and Eu). Within this family, all but the Eu complex behave as SMMs;

this was confirmed by the observation of frequency dependent ac susceptibility signals,

magnetization hysteresis and QTM for the aforementioned complexes.

3.2 Results and Discussion

3.2.1 Syntheses

The synthesis of high-nuclearity Mn clusters incorporating Mn and/or Mn' ions

can be achieved either by oxidizing simple Mn" salts61 Or by a reductive aggregation of

MnV" salts such as MnO4 -62 A yet different approach is the use of preformed metal

clusters such as the oxo-centered trinuclear [Mn30(O2CR)L3 0/+1 (R = Me, Ph, Et etc.; L

= py, MeCN, H20) complexes.54 However, various groups have generally been using

flexible alkoxo chelates tripodall ligands, diols, monols) in reactions employing the

above-mentioned triangular unit. This strategy helps in imposing some geometric

constraints on the resulting cluster; the carboxylates impose little or no geometry









restrictions and bridge multiple metals as oxide bonds are formed. Thus, clusters of

various nuclearities have been obtained in which the alkoxide arms act as bridging

ligands.61 In Our research, we have recently been investigating reactions employing a

preformed Mn cluster and a heterometal atom with the end goal of obtaining

heterometallic clusters which might be important to diverse research areas such as

magnetic materials and bioinorganic modeling. 37, 43 Hence, for the magnetic materials

research we have been utilizing the tri- and tetra-nuclear [Mn30(O2CPh)6 py>2(H20)]

(2)54 and (NBun4)[Mn402(O2CPh)9(H20)] (3)55 complexes as starting materials. The

reaction typically is carried out in a mixed solvent system of MeOH/MeCN (1:20 v/v) in

the presence of a Ln(NO3)3. The solvent mixture is necessary to ensure adequate

solubility of the reagents, especially the Ln(NO3)3 Salts. Indeed, reactions employing 2 or

3 and a Ln(NO3)3 in ratios varying from 1:1 to 1:2 gave the family of [Mn11Ln4 45+

compounds (Ln = Nd, Eu, Gd, Dy, and Ho; complexes = 8, 9, 10, 11, and 12,

respectively). For complexes 8-12, the yield was optimized for ratios as stated in the

Experimental Section (see later). The reaction proceeds in the stoichiometric ratio as

depicted, for example, in eq 3-1 for the Mn/Gd reaction and eq 3-2 for the Mn/Dy

reaction. In all reactions except the one involving Mn and Dy, the MeOH acts as an inert

solvent, inasmuch, it does not end up in the resulting isolated complex either as MeOH or

11 [Mn3(O2CPh)6 py)2(H20)] + 12 Gd(NO3)3 + 35 H20 + 3 H + 24 e -

3 [MnllGd40s(OH)s(O2CPh)16(NO3)5(H20)3] + 18 PhCO2H + 22 py + 21 NO3- (3-1)

as methoxides. However, the M11Dy4 COmplex has two bridging methoxides. Also, if

more methanol was used in the reaction mixture, [Mn2Ln202(O2CPh)6(OMe)4(MeOH)4]

complexes were obtained by methanolysis; they will be described later.59g These









observations clearly indicate that the reactions are very complicated with an intricate mix

of several species likely to be in equilibrium in the reaction solutions; this is perhaps

11 [Mn3(O2CPh)6 py)2(H20)] + 12 Dy(NO3)3 + 6 MeOH + 41 H20 + 3 H + 24 e -

3 [MnllDy40s(OH)6(OMe)2(O2CPh)16(NO3)5(H20)3 + 18 PhCO2H

+ 22 py + 21 NO3- (3 -2)

expected given the amount of H20 molecules, their deprotonation and further H removal

to form hydroxide/oxide brides etc. Although the reactions are simple, one-pot and

straightforward, factors such as relative solubility, lattice energies, crystallization kinetics

and others undoubtedly determine the identity of the isolated product. This said, the

products isolated were obtained in high yields in the 55-60% range and were definitely

thermodynamically most stable.

We wanted to further investigate the reaction system and wondered what would

happen if we used other alcohols instead of methanol. Additionally, the Tb complex had

not been synthesized with the above reaction system and the magnetic properties of the

Dy (which is magnetically similar to Tb) complex, were most interesting, as stated in our

earlier communication.59b Thus, the reaction of 2 with Tb(NO3)3 in a MeCN/PhCH20H

(20ml/5ml) solution was carried out from which complex 13 was successfully isolated.

The reaction is depicted in eq 3-3 and the product contains three deprotonated benzyl

alcohol (PhCH20H) molecules, or in other words phenyl methoxides.

11 [Mn3(O2CPh)6 py)2(H20)] + 12 Tb(NO3)3 + 15 PhCH20H + 23 H20 -

3 [MnllTb40s(OH)6(OCH2Ph)3(O2CPh)20(PhCH20 H)(2)

+ 6 PhCO2H + 22 py + 36 NO3- + 47 H' + 11 e- (3-3)









The small volume of methanol / benzyl alcohol in these reactions is very crucial for

the attainment of these clusters. In the absence of methanol either (i) brown precipitates

of manganese oxides/hydroxides were obtained, or (ii) it did not prove possible to isolate

clean products from the reaction solutions. Also, the relative acidity of methanol (pKa=

15.2) and that of benzyl alcohol (pKa = 15.0) are comparable and hence they provide very

similar reaction conditions from which analogous end products were obtained. When

ethanol (pKa = 16.0) was used as the alcohol, no clean products could be isolated.

A different strategy was used for the Mn2Ln2 reactions. In this case, an excess of

MeOH (10ml) was added to the filtrate, after the reaction of the Mn '4 COmplex 3 with

with Yb(NO3)3.5H20 (or Y(NO3)36H20) had been performed in a 1:2 molar ratio in

MeCN/MeOH (20/5 ml). Thus, the methanolysis of a Mn species in the presence of a

Yb3+ Of Y3+ Source resulted in the isolation of two isostructural complexes in 10-25 %

yields; [Mn2 2z02(O2CPh)6(OMe)4(MeOH)4] (14) or

[Mn2 202(O2CPh)6(OMe)4(MeOH)4] (15). Note that in both these Mn-containing

complexes there are also MeO- groups from the MeOH solvent, in addition to terminal

MeOH molecules. Again, there are likely other products from these complicated

reactions in the colored filtrates, but we have not pursued any further separations.

3.2.2 Description of Structures

3.2.2.1 X-ray crystal structure of complexes 11 and 13

PovRay representations of the crystal structures and labeled cores of complexes 11

(left) and 13 (right) are depicted in Figures 3-1 and 3-2, respectively. Crystallographic

data for 10-15MeCN, 11-15MeCN, and 13-3PhCH20H are listed in Table 3-1.

Complex 11-15MeCN crystallizes in the triclinic space group Plwith the MnllDy4

molecule lying on an inversion center. Although 11 is heterometallic, it is homovalent as













i /


Figure 3-1. PovRay representation at the 50% probability level of the X-ray crystal
structures of 11 (left) and 13 (right). Color scheme: Mn yellow, Dy green, Tb
purple, O red, N blue, C grey. H atoms have been omitted for clarity.

it contains eleven Mn"' and four Dy"' ions. The structure (Figure 3-1, left) consists of a

[Mn11Dy4 45+ metallic unit held together by six C14-02-, two C13-02-, SiX C13-HO-, and two CI-

MeO- ions. Peripheral ligation is provided by twelve CI-, and four CI3-bridging benzoate

groups, five chelating NO3 grOups on the Dy ions, two H20 molecules on Mn4 and Mn4',

and a water molecule on Dy2. The doubly bridging O atoms of the four C13-PhCO2-

groups in one case bridge Mn1 / Dyl', and in the other Mn3 / Mn6, and of course the two

symmetry related counterparts. The core (Figure 3-2, left) consists of two distorted

[DyMn302(OH)2] cubanes (Dy2, 017, Mn3, 016, Mn4, 021, Mn5, OS), each attached

via a Mn/Dy pair (one above (Mn6) and another (Dyl) below the plane) to a central, near

linear and planar [Mn304] unit (Mn1, Mn2, Mnl'). The Mn/Dy pairs are also linked to the

Mn3 linear unit via two triply bridging hydroxides (012, 012'). Within this description,

the remaining four hydroxides (021, 016, 021', 016') lie in the two DyMn3 cubanes.

Also, two doubly bridging methoxides (026, 026') link the Dy"' in the cubanes (Dy2,

Dy2'), to two Mn" ions (Mn6, Mn6'). The position of the methoxide oxygen (026) is the

crucial distinction between the cores of the remaining members of this family of










complexes (see later). The metal oxidation states and the protonation levels of Ol-, HO-

and MeO- ions were established by bond-valence sum calculations,63 charge







026 Mn6 c Dl
0121 12 020 ~f~
075

Dt_ 016 / MMn Mn5' un *4
Mn3r
Mn4 OS I0i 017' Dy2' no 20n
012'hn~
021 026'
Dyl Mn6


Fiue3-.Pvayrpesnainathe5%pobitylvl ofth
[Mx0 4s(H6Oe) 1 6r f1 (left and h









octahedral. geomeray. The Dy ions ar te nine-coord iintean thvel DyO bodslenh

range Mn 2.33-2482 O) 6Oe)1 Ar copral wit vaue rpoted in the ltrtr.9~~ ope


cenroymeticMn "nb0 8b^4 (CHIUhter1 (Fig re 3- 1, right). Theovral strcture: of 13 is

similar t compex, 11pue,th diffreneN being tha they MnHb CtOmplabex possesse four


exta bidgraiong, bsenzioat goup andtwo laessers temnalth H20tfio m oleuls InsteJad oftwo




Hage203 molecules, it ow hrastobleny wtalcohos (ePhCH2H) in th ieatr9ddiio to hvig hee





Table 3-1. Crystallographic data for 10-15MeCN,
10


11-15MeCN and 13-3PhCH20H.
11


Formula
fw, g/mol
Space group

b, A
c, A
a, a
/7, a
Y, a
V, A3
Z
T, K
Radiation, 8a
pealc, g/cm3
C1, mm l
R1 b'c
wR2 d


C140H135066N20Mnl 1iGd4
4430.07
Pl
16.5278(15)
17.4169(16)
18.5303(17)
65.238(2)
67.126(2)
65.125(2)
4245.6(7)

173(2)
0.71073
1.750
2.420
0.0656
0.1622


C142H139066N20Mnl 1 Dy4
4436.07
Pl
16.5705(11)
17.2926(11)
18.6213(12)
64.9700(10)
67.2950(10)
65.6110(10)
4256.6(5)

173(2)
0.71073
1.731
2.610
0.0425
0.1006


C196H169063Mn z iTb4
4772.53
Pl
16.6598(19)
18.647(2)
19.859(2)
72.889(2)
66.854(2)
70.608(2)
5254.5(2)


173(2)
0.71073
1.520
2.042
0.0773
0.1942


a Graphite monochromator. b I> 20(1. R1


100E(||Fo| |Fc||)/E|Fol. d wR2 = 100[C[w(Fo2 Fc2 2]/ C[w(Fo2 2 1/2


w = 1/[G2(Fo2) [P2 +bp], where p = [max (Fo2, O) + 2Fc2]/3.









bridging phenyl methoxides (PhCH20-). Thus, 13 is formulated as

[MnllTb40s(OH)6(OCH2Ph)3(O2CPh)20(PhCH20 H220] the

[Mnl 1Tb40s(OH)6(OCH2Ph)2 21+ COre of which is shown in Figure 3-2 (right). Amongst

the benzoate groups sixteen are bridging as was in complex 11. Two of the remaining

four are monodentate on Tb2 and Tb2', whereas the remaining two are r12 terminally

chelating on Tb 1 and Tbl1'. The two benzyl alcohols provide terminal ligation on Mn4

and Mn4'. The water molecule and a phenyl methoxide reside on Tb2. The remaining two

phenyl methoxides (020, 020' in Figure 3-2, right) bridge two Mn/Tb pairs (Mn6, Tb2

and Mn6', Tb2'), as was the case with the methoxides in complex 11. The metal-

oxo/hydroxo core for the Mn/Tb complex is the same as was for complex 11, with the

oxides and hydroxides occurring in the same positions (Figure 3-2). The only difference

(Figure 3-2) is that 026 is the oxygen of a methoxide in 11, and the corresponding 020 in

13 comes from a phenyl methoxide. All the eleven Mn ions are hexa-coordinated, and in

near octahedral geometry. Tb1 is nine-coordinate, and Tb2 is eight-coordinate. The Tb-O

bonds lie in the range 2.374-2.492 A+, comparable with values reported in the

literature.59a~e~g

3.2.2.2 Structural Comparison of Complexes 8-13

The structures of complexes 8-13 are overall very similar to each other, differing

slightly in the nature of a few bridging ligands in the periphery. Indeed, complexes 8, 9,

and 10 can be formulated as [Mnl 1Ln40s(OH)s(O2CPh)16(NO3)5(H20)3], with Ln = Nd,

Eu, and Gd respectively. The common [MnllLn4080Hs(OH1]21COre to these complexes is

shown for the Ln = Gd complex (10) in Figure 3-3, where 014 is a hydroxide connecting

Gd2 and Mn5. The difference between the cores of these complexes and that of 11 and 13

is that the Dy complex contains a methoxide (026) in the position of 014, and the Tb










complex contains a phenyl methoxide (020) in the corresponding positions. Peripheral

ligation around these cores is the same for the Dy and Gd complexes, whereas the Tb


Gd2Mn5 Gd1' M 6

Mn4 Mn" Mn3
Mn2
Mn3 Mn4'
MMn1


Mn6
OT4f Gd2'
Gd1 Mn5"
Figure 3-3. PovRay representation at the 50% probability level of the
[MnllGd40s(OH)s]21+ COre of complex 10. Color scheme: Mn yellow, Gd
cyan, O red, C grey.

complex differs slightly as already described earlier. The holmium complex,

[MnllHo40s(OH)s(O2CPh)1s(NO3)3(H20)7] (12) differs considerably in the peripheral

ligands when compared to other members of this family of complexes. The core for 12 is

the same as the one depicted for 10 in Figure 3-3. However, peripheral ligation is now

provided by eighteen bridging benzoate groups with sixteen of them bridging as was for

the Dy complex. The additional two benzoate groups bridge monodentate on the two Ho

ions in the two cubanes. The other two Ho ions have three chelating nitrate groups on

them. Additionally, there are five water molecules on the Ho ions and two on the Mn ions

(for example Mn6 in Figure 3-3), providing terminal ligation. Thus, although the metallic

core of all the complexes is very similar, they differ slightly in their periphery. This slight

variation in the structures may indeed consequently have an effect on the magnetic

properties of these clusters. Hence we decided to investigate the single-molecule magnet

properties of these complexes, which are described later.









3.2.2.3 Structural descriptions of complexes 14 and 15

PovRay representations of the centrosymmetric crystal structures (top) and labeled

cores (bottom), of complexes 14 (left) and 15 (right) are depicted in Figure 3-4. Selected

interatomic distances and angles are listed in Table 3-4. The two complexes both

crystallize in the monoclinic space group P21 c with the molecules lying on an inversion

centre; the two complexes are isostructural. The cores of the complexes possess a defect-

dicubane structure (two fused cubanes sharing a face, and each missing an opposite

vertex; Fig. 3-4), with two Mn atoms at the central positions and either two 2 Yb"'

(complex 14) or two Y"' (complex 15) atoms at the end positions. The fully-labelled

cores of 14 and 15 are provided in Fig. 3-4 (bottom), which emphasize the near

superimposibility of the mixed 3d/4d complex 15 with the mixed 3d/4f complex 14.

Peripheral ligation about the cores is provided by four syn, syn bridging benzoate groups

bridging each Y/Mn or Yb/Mn edge of the rhombus, two monodentate benzoate groups,

one on each of the Y/Yb, atoms, and four terminal MeOH molecules, two on each of the

Y/Yb, ions. Within this description, two C13-02- atoms (OS, 08a) cap each triangular sub-

unit, and each of the four edges of the M4 rhombus in 14/15 is bridged by a MeO- ion

(07, 07a, 011, 011a). As a result, the Yb"' and Y"' atoms of 14 and 15, respectively, are

eight-coordinate, and the Mn atoms are six-coordinate. There are intramolecular OH---H

hydrogen bonds between the unbound O atom (Ol) and the terminal MeOH ligand

(01---010 = 2.577A+). In addition, there are intermolecular OH---H hydrogen bonds

between the bound MeOH (09) and interstitial MeOH (012) molecules (09---012 =

2.656 A+), and between this interstitial MeOH and the unbound O atom (Ol) of a

neighboring cluster (012---01 = 2.687 A+). Thus, the hydrogen-bonding through the










interstitial MeOH molecules links adj acent metal clusters in the crystal, which also likely

provides a pathway for superexchange interactions between molecules videe infra).










oi c






Mn1 n
oi 011 07al qi07
OB Ybla 08
Yb1\ 08Loa =l Y1 08a
07 011 07 11a

M hnla Mr( hnla


Figure 3-4. (Top) PovRay representations of the crystal structures of complexes 14 (left)
and 15 (right). (Bottom) Comparison of the cores of complexes 14 (left) and
15 (right), emphasizing their near superimposibility. Color scheme: Mn blue,
Yb orange, Y pink, O red, C grey.

It should be noted that mixed Mn2Ln2 COmplexes with the same kind of defect-

dicubane core have previously been reported for Ln = Dy, Gd and Tb.59d, 64 However, in

all of these previous cases, the Mn atoms were in the Mn oxidation state, so 14 and 15

are the first to instead contain Mn V. The Mn' oxidation level is suggested by overall

charge considerations and inspection of metric parameters; in particular, Mn--O bond

distances all lie in the range 1.84-1.97 A+, as expected for Mn V, and thus do not show the

Jahn-Teller axial distortion expected for Mn"' in near octahedral geometry. The Y--O and

Yb--O bond distances are very similar, lying in the 2.30-2.42 and 2.27-2.38 A+ ranges









(Table 3-4), respectively, consistent with eight-coordinate Y"'/Yb"' centers. The Mn'V

oxidation states and the protonation levels of the Ol-, MeO and MeOH were confirmed

by bond valence sum (BVS) calculations, shown in Tables 3-2 and 3-3. The BVS values

Table 3-2. Bond Valence Sums (BVS)a for the Mn atoms of complexes 15 (Y) and 14
(Yb).
Atom Mn" Mn"' Mn'V Assignment

Mn1 (15) 4.132 3.779 3.967 Mn'V

Mn1 (14) 4.124 3.772 3.959 Mn'V
a The underlined value in bold is the one closest to the charge for which it was
calculated. The oxidation state of a particular metal is the nearest whole number to that
value .

Table 3-3. Bond Valence Sums (BVS)a for the O atoms of complexes 15 (Y) and 14
(Yb).
Atom BVS (15) BVS (14) Assignment
07 1.972 1.972 MeO-
08 1.898 1.859 02-
09 1.210 1.198 MeOH
010 1.339 1.342 MeOH
011 2.038 2.016 MeO-

a The oxygen atoms is 02-,MeO- if BVS 2; MeOH if BVS= 1;i and H0i V=0

for the Mn atoms are clearly ~ 4, confirming the Mn'V oxidation level. Values of ~ 2 are

expected for O atoms in the O-" oxidation level and that have no attached atoms that

cannot be seen in X-ray crystallography (i.e. H atoms). This confirms that 07, 08 and

011 are MeO-, Ol-, and MeO-, respectively. In contrast, if there is a H atom which is not

visible and its contribution to the BVS of that O atom is therefore not included, a lower

BVS value is expected, typically 1 1.5 (depending on the degree of its participation in

hydrogen-bonding); this is clearly the case for 09 and 010, which are therefore

confirmed as MeOH groups.









Table 3-4. Selected bond distances (A+) and angles (o) for complexes 14 and 15.
Complex 15 (Mn2 2> COmplex 14 (Mn2 2z)
Y1 07 2.365(2) Yb1 07 2.336(2)
Y1 08 2.356(2) Yb1 08 2.331(2)
Y1 09 2.416(2) Yb1 09 2.383(3)
Y1- 010 2.351(2) Yb1 -010 2.323(3)
Y1- 011 2.299(2) Yb1 -011 2.269(2)
Mn1 08 1.859(2) Mn1 08 1.866(2)
Mn 011 1.913(2) Mn 011 1.912(3)
Y1Mn1 3.3168(7) YblMn1 3.2886(6)
Mn lMnla 2.7819(9) Mn lMnla 2.7769(11)
Y1-011-Mn1 102.08(9) Ybl-011-Mn1 101.79(11)
Mn1-08-Mnla 97.52(9) Mn1-08-Mnla 97.05(11)
Y1-08-Mn1 103.18(9) Ybl-08-Mn1 102.61(10)

The occurrence of Mn'V in 14 and 15 is noteworthy given that the reaction

employed a Mn "4 Starting material and either Y"' or Yb"', neither of which are good

oxidizing agents. This indicates either the participation of atmospheric 02 gaS as the

oxidizing agent or the disproportionation of Mn"' to Mn'V and Mn". We favor the latter

possibility given the low yields of these complexes, but we have not sought Mn" species

in the filtrates to confirm the same.

3.2.3 Magnetochemistry of Complexes 8-13, and 15

3.2.3.1 DC studies of complexes 9, 10, 11, and 13

Solid-state variable temperature magnetic susceptibility measurements were

performed on vacuum-dried microcrystalline samples of complexes 9, 10, 11, and 13,

each suspended in eicosane to prevent torquing. The dc magnetic susceptibility (XM) data

were collected in the 5.0-300 K range in a 0.1 T magnetic field and are plotted as XhTys

Tin Figure 3-5 for the aforementioned complexes. For 9 the XMT value of 26.33 cm3mOl

1K at 300 K is much lesser than the expected value for 11 Mn3+ (S = 2, L = 2, 5Do; g = 2)

and 4 Eu3+ (S = 3, L = 3, 7Fo) non-interacting ions of 33.00 cm3mOl-1K, consistent with

antiferromagnetic exchange interactions within 9. The XMT value decreases slightly with







59


00





E 0 -~~
1-n



20- *
O Complex 10(Gd)
r Complex 11(Dy)
SComplex 13 (Tb)

0 50 100 150 200 250 300

Temperature (K)
Figure 3-5. XMT vs T plots for complexes 9 (*), 10 (0), 11 (V), and 13 (A).

decreasing temperature to reach 13.63 cm3mOl-1K at 25 K. Thereafter it falls rapidly to a

minimum of 9.46 cm3mOl-1K at 5.0 K, indicating a small ground state spin for complex 9.

For 10, the XMT value of 51.43 cm3mOl-1K at 300 K is less than the expected value for 11

Mn3+ (S = 2, XMT = 33 cm3mOl-1K) and 4 Gd3+ (S = 7/2, L = 0, sS7/2) HOn-interacting ions

of 64.52 cm3mOl-1K. The XMT value decreases slightly with decreasing temperature to

reach 43.38 cm3mOl-1K at 15 K, consistent with dominant antiferromagnetic exchange

interactions within complex 10. Below 15 K, the XMT value decreases steeply to reach

39.03 cm3mOl-1K at 5.0 K indicating a ground state spin for 10, which is definitely larger

than that of 9.

For complexes 11 and 13, the XMT values at all temperatures are higher than the

corresponding ones for 9 and 10 (Figure 3-5). For 11, the XMT value of 74.34 cm3mOl-1K

at 300 K is less than the expected value for 11 Mn3+ (S = 2, g = 2) and 4 Dy3+ (S = 5/2, L

= 5, 6H15/2) HOn-interacting ions of 89.68 cm3mOl-1K, and the XMT decrease steadily with

decreasing temperature to reach 62.17 cm3mOl-1K at 40 K; the behavior being consistent









with antiferromagnetic exchange interactions within 11. After 40 K, the XMT drops

slightly to 57.78 cm3mOl-1K at 10 K, and then remains more or less constant with

decreasing temperature to finally reach 57.09 cm3mOl-1K at 5.0 K, indicating a relatively

large ground-state spin for complex 11. For complex 13, the XMT value of 64.35 cm3mOl

1K at 300 K is less, as anticipated, than the expected value for eleven Mn3+ (S = 2) and

four Tb3+ (S = 3, L = 3, 7F6) HOn-interacting free-ions of 80.28 cm3mOl-1K. The magnitude

of XMT decreases with decreasing temperatures to reach a value of 53. 14 cm3mOl-1K at 40

K. Below 40 K, the XMT drops sharply to reach 46.66 cm3mOl-1K at 10 K after which it

levels off at a value of 45.04 cm3mOl-1K at 5.0 K, indicating a large ground-state spin for

13, which might be comparable to that of 11.

The molar dc magnetic susceptibility values of the above-mentioned complexes are

indicative of the contribution of the lanthanide ion to the net experimental magnetic

susceptibility observed and consequently the resulting magnetic properties of these

complexes videe infra). This is hypothesized because all of these complexes have eleven

Mn" centers, in addition to the Ln ions, and the free-ion XMT value (in cm3mOl-1K) of

individual Ln3+ ions in decreasing order is Dy (14. 17) > Tb (11.82) > Gd (7.88) > Eu

(0.00). Accordingly, the magnitude of the XMT with decreasing temperatures and even at

300 K, follows the same descending pattern, i.e., 11 > 13 > 10 > 9, as that of the free-ion

XMT values of the constituent Ln3+ ions of these complexes. Thus, both the data are

satisfyingly consistent, and the overall nature of the plots for all the complexes is very

similar as can be seen in Figure 3-5. However, due to the size and low symmetry of the

molecules, a matrix diagonalization method to evaluate the various metal pairwise

exchange parameters (Jij) within the MnllLn4 mOlecules is not easy. Similarly,









application of the equivalent operator approach based on the Kambe vector coupling

method45 iS not possible. Additionally, due to the presence of strong spin-orbit coupling

effects (arising from the Ln3+ COmponent) in these lanthanide containing mixed-metal

complexes, the fitting of the magnetization data to obtain the S and the axial zero-field

splitting (ZFS) term D, is far from straightforward. The main problem in these procedures

is the assumption of an isotropic, single electronic g factor; all the systems are two g (one

for Mn3+ and another for Ln3+). However, there is some reprieve provided by the

MnllGd4 System which has Gd3+ (S = 7/2, L = 0, g 2) and therefore the resulting

absence of spin-orbit coupling effects from the Ln3+ COmponent. Thus, we focused on

identifying the ground state S value for complex 10, for which magnetization (M)1 data

were collected in the magnetic field and temperature ranges 0.1-0.5 T and 1.8-10 K; the

data are plotted as reduced magnetization (M~NpB) VeTSus H/T in Figure 3-6. The data

were fit, using the program MAGNET,46 by diagonalization of the spin Hamiltonian

matrix assuming only the ground state is populated, incorporating axial anisotropy (D 2)

and Zeeman terms, and employing a full powder average, as has already been discussed

in Chapter II. The fit parameters were S = 9, g = 1.86 (A 0.00) and D = -0.06 (A 0.00) cm

1. The D value of -0.06 cm-l is consistent with the complex having four isotropic Gd3+

ions, and therefore an attenuation in the effective magnetoanisotropy component; it has

already been stated earlier that the Ln3+ ions play a maj or role in the magnetic properties

displayed by these complexes. Additionally g < 2, as can be expected for a system

comprising eleven Mn"' (g < 2) and four Gd"' (g = 2) ions. Also, the expected XMT for an

S = 9 and g = 1.86 system of 38.92 cm3mOl-1K can be compared with the 5 K dc value of

39.03 cm3mOl-1K, which are in good agreement.











14-



10-



6 -1 O 0.1 T
5 m 0.2 T
A 0.3 T
4 -1 O 0.4T
OO v 0.5 T
2 fitting


0.0 0.5 1 .0 1 .5 2.0 2.5 3.0

HIT (kGlK)
Figure 3-6. Magnetization (M) vs field (H) and temperature (T) data, plotted as reduced
magnetization (M/NCLB) vs H/T, for complex 10 at applied fields of 0.1i, 0.2,
0.3, 0.4 and 0.5 T and in the 1.8 10 K temperature range. The solid lines are
the fit of the data; see the text for the fit parameters.

Although 10 has a small D value of -0.06 cm- the large ground state spin value of

S = 9, which is comparable to the S = 10 of Mnl2 SMMs,l suggested that the barrier to

magnetization relaxation might be large enough for the complex to function as an SMM.

The S and D values obtained for complex 10 suggest an upper limit to the potential

energy barrier (U) to magnetization reversal of U= S2|D| = 4.86 cml = 7.00 K,9 although

the actual, effective barrier (Gerf) was anticipated to be less than this due to quantum

tunneling of the magnetization (QTM). Additionally and more importantly, the ground

state spin of the Dy and Tb complexes can be estimated to lie in the S = 10 + 1 region,

based on comparative Dc magnetic susceptibility data videe supra). Thus, ac

susceptibility measurements (see later) were performed on complexes 9, 10, 11, and 13 to

investigate whether they functioned as single-molecule magnets.











3.2.3.2 DC studies of [Mn2 202(O2CPh)6(OMe)4(MeOH)4] (15)

Solid-state variable temperature dc magnetic susceptibility measurements were


performed on a vacuum-dried microcrystalline sample of complex 15 suspended in

eicosane to prevent torquing. The dc magnetic susceptibility (XM) data were collected in

the 5.0-300 K range in a 0. 1 T magnetic Hield and are plotted as XhTys Tin Figure 3-7 for

15. Complex 15 has a XxiTvalue that steadily decreases almost linearly from 3.13

cm3mOl-1K at 300 K to 1.52 cm3mOl-1K at 50 K, and then more rapidly decreases to 0.49

cm3mOl-1K at 5.0 K (Fig. 3-7). The value at 300 K is less than the expected spin-only (g =


2) value for a complex consisting of two non-interacting Mn'V ions (S = 3 2; the 2 Y ions

are diamagnetic) of 3.75 cm3mOl-1K, indicating an antiferromagnetic exchange

35

30 -1

25-

20-

O 15-

~10-

os -I Dc data
fitting

0 50 100 150 200 250 300

Temperature (K)
Figure 3-7. Plot of X T(solid circles, *) vs. Tfor complex 15. The solid line in the XMTyvs
T plot is the fit of the data; see the text for the fit parameters.

interaction between the Mn'V ions and a resultant S = 0 spin ground state. Attempts to fit


the XhlTys T data using the isotropic Heisenberg-Dirac-van Vleck Hamiltonian described


by Hi= -2 J 3; 3,, where S1 = S2 3 2 and J is the magnetic exchange interaction, gave


poor fits that did not reproduce the data over the whole temperature range, particularly

the higher T data. However, it did suggest that the J value is relatively weak, in the J = -









10 to -15 cm-l range, which in fact is consistent with the relatively acute Mn4+-O2--Mn4+

(Mn1-08-Mnla) angles of 97.520. The best fit of the data shown as a solid line in Figure

3-7, gave J= -13.5 cm- g = 1.88, and paramagnetic impurity, p = 0. 10. Although the g

value is as expected for Mn'V (g < 2), the paramagnetic impurity is very high;

approximately 10%. We had extended the work to Y instead of Ln, as it has proven

useful for understanding the magnetic properties of the isostructural Ln-containing

species 14. As for complex 14, we believe the overall behavior is similar to 15, albeit

fitting of the data becomes really complicated because of the presence of the strong spin-

orbit coupling effects arising from the highly anisotropic Yb"' ions. Additionally, there

are intermolecular exchange interactions, via the hydrogen-bonding network as described

earlier, in complex 14, which are not incorporated in the model. These points need to be

considered further if a more quantitative magnetic understanding of this whole family of

mixed-metal species, as well as others being prepared, is required.

3.2.3.3 AC studies of complexes 9, 10, 11, and 13

Alternating current magnetic susceptibility studies were performed on vacuum-

dried microcrystalline samples of 9, 10, 11, and 13 in the temperature range 1.8-10 K

with a zero de Hield and a 3.5 G ac Hield oscillating at frequencies between 50-1000 Hz.

Typically in an ac susceptibility experiment, a weak Hield (generally 1-5 G) oscillating at

a particular frequency (u) is applied to a sample to probe the dynamics of the

magnetization (magnetic moment) relaxation. A decrease in the in-phase ac susceptibility

signal and a concomitant increase in the out-of-phase signal are indicative of the onset of

the slow, superparamagnet-like relaxation of SMMs.7, 9 This occurs because at low

enough temperatures, where the thermal energy is lower than the barrier for relaxation,

the magnetization of the molecule cannot relax fast enough to keep in phase with the










oscillating field. The obtained data for complexes 11 and 13, plotted as Xhl'T vs. T for the

in-phase (Xhl') component and Xhl" (out-of-phase) vs. T, are shown in Figures 3-8 and 3-

9, respectively. For complex 9, the results of the ac susceptibility measurements did not

display any frequency-dependent signals or even tails of peaks which might lie below the

operating minimum of 1.8 K of our SQUID magnetometer. Additionally for complex 10,

at temperatures below 2.5 K, weak tails of XMI" signals whose peak maxima lie at

temperatures below the operating minimum of our SQUID magnetometer (1.8 K) were

observed. These XMI" signals are accompanied by a concomitant frequency-dependent

decrease in the in-phase (XM'T) signals, albeit at these low temperatures.

The results of the ac susceptibility studies on 11 are depicted in Figure 3-8. As can

be seen, the Dy complex displays strong frequency-dependant signals below 4 K. Indeed,

a frequency-dependent decrease in the in-phase XM'T component (Figure 3-8, top) is

accompanied by a concomitant increase in the out-of-phase Xhl" signals (Figure 3-8,

bottom). The latter merely "tails" of peaks that lie at < 1.8 K, the operating limit of our

SQUID. These signals are indicative of slow magnetization (M)1 relaxation. The XM'T

value for complex 11 at 10 K of 61 cm mOl-1K remains more or less constant till 4 K,

after which it drops rapidly from the slow relaxation effect. The ac data are in agreement

with the dc data and both in conjunction suggest a large ground state spin S for complex

11. The slow-magnetization relaxation effects observed in complex 13 are similar to

those seen for 11, as is illustrated in Figure 3-9 for 13. Both in-phase and out-of-phase

frequency-dependent signals indicative of the onset of slow-magnetization relaxation are

observed at temperatures < 3.4 K. These signals are strong, albeit their peaks still lie







66



below 1.8 K. For 13 the Xhn'T value at 10 K of 49.0 cm3mOl-1K drops slightly to 47.5


cm3mOl-1K, before rising to a maximum of 50.0 cm3mOl-1K at 3.4 K. Thereafter, the



60







6-






-* 1000 Hz
5 -I 500 Hz
-0 250 Hz
-* 50 Hz










ll 2 4 6 8 10
Temperature (K)
Figure 3-8. Ac susceptibility of complex 11 in a 3.5 G field oscillating at the indicated
frequencies. (Top) in-phase signal (p'~) plotted as 3y'Tyvs T; and (bottom) out-
of-phase signal yn" vs T.

superparamagnet-like slow magnetization re-orientation initiates and consequently the


Xhl'T value drops (Figure 3-9, top).


The appearance of out-of-phase XMI" signals suggests that complexes 10, 11, and


13 may indeed have a significant (vs. kT) barrier to magnetization relaxation and thus


may be SMMs. In fact, dc and ac magnetic susceptibility studies were also performed on


complexes 3 and 7, the results of which have not been discussed for brevity's sake.

Nevertheless, both these complexes also showed frequency dependent signals in ac














58-

56-

O 54-
E
S 52-



48-

46-

6 -* 1000 Hz
-0 500 Hz
-0 250 Hz
5 -* 50 Hz

4-











ll 2 4 6 8 10

Temperature (K)

Figure 3-9. Ac susceptibility of complex 13 in a 3.5 G field oscillating at the indicated
frequencies. (Top) in-phase signal (p') plotted as yITyVs T; and (bottom) out-
of-phase signal y" vs T.


studies with the strength of the signals (temperature below which the signals appeared) in


descending order being Dy, Tb > Ho, Gd > Nd (complexes 11, 13, 12, 10, and 8,


respectively). Note that the Eu complex (9) did not show any frequency-dependent Ac


signals. However, frequency-dependent ac signals are necessary but not sufficient


evidence that an SMM has been obtained. Confirmation of this requires magnetization vs


applied dc field sweeps to display hysteresis loops, and this was explored on with studies


at temperatures below 1.8 K on complexes 8-13, as described below.









3.2.3.4 Hysteresis studies below 1.8 K on complexes 8-13

To establish whether complexes 8-13 were SMMs, magnetization vs. applied de

Hield data down to 0.04 K were collected on single crystals (that had been kept in contact

with mother liquor) using a micro-SQUID apparatus.so The observation of hysteresis

loops in such studies represents the diagnostic property of a magnet, including SMMs and

superparamagnets below their blocking temperature (Ty3). The observed magnetization

responses for complexes 8-10, and 12 at a Eixed Hield sweep rate and at the indicated

variable temperatures are shown in Figure 3-10 (8 (top; left), 9 (top; right), 10 (bottom;

left), and 12 (bottom; right)). Hysteresis loops were indeed observed for complexes 8, 10,

and 12; the coercivities of the loops increase with decreasing temperature as expected for

the superparamagnet-like properties of a SMM. The results thus confirm the addition of

complexes 8, 10, and 12 as new members to the growing family of SMMs. For complex 9

(MnllEu4) there is no coercivity even at temperatures down to 0.04 K, which is

satisfyingly consistent with the earlier mentioned absence of any frequency-dependent

signals in ac magnetic susceptibility studies performed on 9. Additionally, the ground

state spin S of 9 was postulated to be very small based on the 9.46 cm3mOl-1K dc value at

5.0 K, and comparative analyses with other complexes, hence precluding any possibility

of 9 possessing a big enough barrier to magnetization relaxation. Thus, complex 9 is not a

SMM. For complex 8, the hysteresis loops slightly open up at 0.5 K and a clear coercivity

is observed at 0.04 K and at the Eixed scan rate of 0. 14 T/s. The overall behavior in

hysteresis studies of complexes 8 and 12 is very similar as shown at a Eixed scan rate of

0.035 T/s in Figure 3-10, bottom (left and right, respectively). In both these cases,

hysteresis loops are seen below 0.5 K, although the coercivities of the loops in 12 are

broader than those observed for 10. Nonetheless, the observance of hysteresis loops with














0.5 -l ~ 0.5-



-0. 0- 2~l 1O 0-
-004
C04 -11
-0. -0. 1 0 K -0.5 .


PloH (T) ILoH (T)

0.035 T/s C 0.035 T/s

0.5 -1 -- 0.5-





-0.5 5 ,: -0.5-- 3K


-1 -1 -
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 -1 -0.5 0 0.5 1
Po H(T) Co H (T)


Figure 3-10. Magnetization (M)1 vs de Hield (H) hysteresis loops at a Eixed Hield sweep rate
and at the indicated variable temperatures for single crystals of 8 (top; left), 9
(top; right), 10 (bottom; left), and 12 (bottom; right). The magnetization is
normalized to its saturation value, M~s.

increasing coercivities with decreasing temperature, unequivocally establish that


complexes 10 and 12 are indeed single-molecule magnets.

The hysteresis loops for complexes 11 (top) and 13 (bottom) are depicted in Figure


3-11. Shown in Figure 3-11 (top, left) are the magnetization responses for 11 at a dc field


sweep rate of 0. 14 T/s with the Hield approximately along the easy axis (z axis) of the

molecule, and at variable temperatures including and below 1.0 K. Figure 3-11 (top,


right) depicts the de Hield sweep rate dependence of the hysteresis loops observed for 11

at 0.04 K. Analogous data collected on complex 13 are presented in Figure 3-11


(bottom), at a sweep rate of 0. 14 T/s (left) and at a fixed temperature of 0.04 K (right).












Hysteresis loops were observed for complexes 11 and 13, and as anticipated, the


coercivities (widths) of these loops increase with decreasing temperature (Figure 3-11i,


left) and increasing sweep rate (Figure 3-11, right), as expected for the superparamagnet-


like properties of a SMM.



1- 0.14 T/s 00


0.5 -( -ll~ 0.5-




-05 1 -02l~ U K -0140T/

_1~6S 0 070 T/s
-06 -5. -02 0 0. 5 K -0.5 -0. 0. 017 T/s


08 0 04T/



F- 0 001K T/



-1 -1-
-0.6 -0.4 -0.2 0 0.2 0.4 0.6 -0.6 -0.4 -0.2 0 0.2 0.4 0.6
C0 H (T) 90 H (T)




0at 14 T/s / lf) ada h niatdfedseprae n i



0 aue 00 8 0-- 0 8 /

Hyseri -010 maneiato -0140T/ssepsi hecasia poetyo

magnet, 0n 30hlosar igoti intr Kls -f 0M 035 spra nT/s.

-0.5 -at 0hu 4nic Kopee 01 017 T/s enwadiin otefaiyo i
Mn/L/ SM s 0n 5nl Kh seon -0.5e -f -h 0al 008 dsl htesi T/s n









other being the Mn2Dy2 COmplex reported recently.59d The blocking temperature (TI3) is ~

1.0 K for 11 and 13, above which there is no hysteresis (Figure 3-11 (left)); i.e., the spin

relaxes faster to equilibrium than the time scale of the hysteresis loop measurement. For

several other SMMs studied to date, the hysteresis loops have not been smooth but have

instead displayed step-like features at periodic field values.7, 61b,c, 65, 12 These steps

correspond to increased magnetization relaxation rates and are due to quantum tunneling

of the magnetization (QTM) through the anisotropy energy barrier." The hysteresis loops

in Figure 3-11 do not show such periodic steps because of a distribution of molecular

environments arising from the presence of severely disordered solvent molecules, as well

as chelates such as nitrates. Therefore, the local environment around the MnllLn4

molecules vary, which in combination with low-lying excited spin states (exchange

interactions (J) of Ln ions are typically weak), lead to a distribution in D values, that is, a

distribution in energy barriers to relaxation; the separations between steps is directly

proportional to D, so a distribution in D would give a distribution in step position as a

consequence of which the steps are broadened to the extent that they are smeared out and

thus not visible. In addition, intermolecular interactions (both dipolar and exchange) and

population of excited states can result in step broadening. It should be noted that although

such disorders in solvents and ligand positions might at first glance appear to be too

trivial to cause a noticeable variation in the molecular D value, there is ample precedent

that this is not the case. Instead, on several occasions we have observed that such

properties of SMMs are acutely sensitive to relatively small changes in the local

environment of the molecules.66, 67 The same smearing out of steps has been observed

previously, especially in high nuclearity Mn complexes such as Mnl6,68 Mn21,69 Mn22,70










Mn25,47 Mn30,52 and Mn84,102b to name but a few, and these were similarly assigned to a

distribution in D values.

Although a justification has been given for the absence of steps in hysteresis loops

through the above discussion, the hysteresis loops still suggest that QTM is occurring in

complexes 11 and 13. This is proposed because at temperatures below 0. 1 K the loops

become temperature independent, but a scan-rate study (Figure 3-11 (right)) shows that

the loops are still time-dependent. Further confirmation of this behavior was sought

through time dependent dc magnetization decay data collected in the 0.04 1.0 K range.

This study is useful in obtaining data which can be used to construct an Arrhenius plot

based on the Arrhenius relationship of eq 3-4, where to is the preexponential factor, USir

is the mean effective barrier to magnetization relaxation, and k is the Boltzmann constant.

r = ro exp(14tr/kT) (3-4)

Typically in a dc magnetization decay study, the sample's magnetization is first saturated

in one direction at ~ 5 K with a large dc field, the temperature is then lowered to a

0.6 .-~~LYLYY LYLLYLL 10 3





~10-6
0. 1 0 10 10
t (s) 1T (1/K

Fiur 312 (ef) agetztin s im dca pot frcrstlsofcoplx 3 t h
iniae tepeaurs (Rgt Areiu lt sn hersltn elxto
tie(r eru Taa Tedahd ieiste i o hedt i heteral
acivte rgon t te rheiu euain.Se te et orth ftpaamtes










chosen value, and then the field removed and the magnetization decay monitored with

time. This provided magnetization relaxation rate (1/z) versus temperature (T) data,

which were fit to the Arrhenius relationship of eq 3-4. The corresponding dc

magnetization decay plot (left) and the resulting Arrhenius plot (right) for complex 13

(MnllTb4) are depicted in Figure 3-12. The Arrhenius plots for complexes 11 (left) and

10 (right) are depicted in Figure 3-13. For complexes 11 and 13, the dc decay data was

collected in the 0.04 1.00 K range, and for complex 10 the relaxation time (z) was

determined for magnetization decay spanning the 0.5 0.04 K temperature range. Fits of

the thermally activated region above ~0.6 K for 13, ~0.5 K for 11, and ~0.3 K for 10

(shown as a dashed line in Figures 3-12 (right), and 3-13), gave to = 5.8 x 10-s s and

14g/ k= 9.9 K for the MnllTb4 COmplex (13). For the MnllDy4 COmplex the parameters

were to = 4 x 10-s s and 14g/ k= 9.3 K, and for the MnllGd4 COmplex to = 7 x 10-13 S and

14g /k = 9.0 K. For the sake of comparison, these parameters are listed in Table 3-5 for


107j 107

105 1061*e(ST

10 3 s =(4 x 10-8)exp(9.3/T) =7e1*e^9T
a : W101
r101 r
10-1
10 -110-
10 -3 10 -s
10" -6. 10 -5
0 5 10 15 20 U 5 10 15 20

Figure 3-13. Arrhenius plot of the relaxation time (r) versus 1/T, constructed from Dc
magnetization decay data for complexes 11 (left), and 10 (right). The dashed
line in green is the fit of the data in the thermally activated region to the
Arrhenius relationship with the indicated parameters.

complexes 10, 11 and 13. Below ~ 0.1 K, the relaxation lifetime becomes essentially

temperature-independent for complexes 10, 11 and 13. Although, the magnetization









relaxation processes in the Tb (13) and Gd (10) complexes are very complicated, the

MnllDy4 COmplex's relaxation phenomenon is easier to comprehend. Thus, a closer

Table 3-5. Comparison of the SMM parameters of complexes 10, 11 and 13.
Complexes to (s) Uef k (K) Ty3 (K)a
(MnllLn4) (Preexponential) (Kinetic barrier) (Blocking temp.)
MnllGd4 (10) 7.0 x 10-13 9.0 ~0.5
MnllDy4 (11) 4.0 x 10- 9.3 ~1.0
MniiTb4 (13) 5.8 x 10- 9.9 ~1.0

a Approximate value of the temperature below which hysteresis loops are observed.

inspection of the Arrhenius plot for complex 11 (Fig. 3-13, left) reveals that below ~ 0. 1

K (1/T = 10), the relaxation lifetime becomes essentially temperature independent at ~106

s, consistent with the purely quantum regime where QTM through the anisotropy barrier

is only via the lowest energy +M~s levels. In other words, in this ground-state QTM region

tunneling is now only between the lowest-energy M~s = + S levels (S is the ground state

spin of the MnllDy4 COmplex), and no longer via a thermally (phonon) assisted pathway

involving higher-energy M~s levels. The crossover temperature to this ground-state

tunneling from the thermally activated relaxation is between 0.1 and 0.2 K.

The results of the dc magnetization decay data study on complex 10 (MnllGd4)

provided an effective (kinetic) energy barrier UAgf= 9.0 K. This 14gf may be compared to

the U= S2|D| = 7.0 K obtained from magnetization fits. Although the Uis generally

slightly bigger than 14gf in most cases, the anomaly here arises due to the reduced

magnetization procedure not being so accurate, mainly due to the assumption of the

population of only the ground state, as well as the application of a very simplified model

which neglects rhombic anisotropy (E) and higher order anisotropy terms. Nevertheless

both the barrier values U and 14gf for complex 10 still fall in the same ballpark. Thus, at a









qualitative level, and to some extent to a quantitative level, the various dc, ac, hysteresis,

and magnetization relaxation data are all consistent. It should be noted that in ac studies

the strength of the signals (temperature below which the signals appeared) in descending

order was Dy, Tb > Ho, Gd > Nd (complexes 11, 13, 12, 10, and 8, respectively).

Correspondingly, the temperature below which hysteresis was observed, the kinetic

barrier USy, as well as the Dc value at 5 K, followed more or less the same decreasing

pattern for all of these MnllLn4 COmplexes.

3.3 Conclusions

In this and earlier published work,59b we have demonstrated the initial observation

of magnetization hysteresis and quantum tunneling, two confirmatory properties of

SMMs, in mixed-metal 3d/4f single-molecule magnets. The observation of hysteresis,

especially in heterometallic lanthanide-containing SMMs is of utmost importance

because it is a well established fact now that the Ln component causes these complexes to

suffer from the disadvantage of fast QTM rates.59d, f-h This represents a diminution of the

effective barrier for magnetization relaxation, and thus the temperature below which the

relaxation is blocked and the complex will function as a SMM. Additionally at zero field,

the tunneling is so fast that many of the reported complexes do not show

superparamagnet-like behavior, in other words, a barrier to magnetization relaxation at H

= 0.59d However, and very interestingly, the MnllLn4 COmplexes reported here do not

suffer from fast QTM. Consequently, important parameters such as the effective barrier

to magnetization relaxation, UEg, were obtained for the Dy and Tb complexes which were

9.3 and 9.9 K, respectively. The preexponential (zo)values of 4.0 x 10-s s (MnllDy4) and

4.8 x 10-s s (MnllTb4), are COmparable and satisfying consistent with those reported for









other SMMs. Both these complexes show frequency-dependent ac susceptibility signals

which in conjunction with the observance of hysteresis loops below 1.0 K confirmed that

11 and 13, irrefutably are SMMs. Magnetization fits for the MnllGd4 COmplex,

containing magnetoisotropic Gd ions, yielded an S = 9 and D = -0.06 cml

The ability to probe both classical (hysteresis) and quantum (QTM) magnetism

behavior as a function of the lanthanide ion as demonstrated, by the extension of a

standard synthetic method to several lanthanides, should provide an invaluable means of

improving our understanding of both the chemistry and physics of this area of

heterometallic molecular nanomagnetism. In addition, the large variation in spin and

anisotropy within the lanthanide ions mentioned earlier, coupled with the site-selective

replacement of the Ln3+ ions in the various MnllLn4 COmplexes, offers the possibility of

"spin-inj section" and/or "anisotropy-inj section" within these SMMs, by the choice of a

suitable Ln3+ ion. Thus, it allows greater control over a family of related SMMs and

investigation of the magnetic properties in a causal manner. Finally, we comment that

these complexes along with the ongoing research efforts in this area provide proof-of-

feasibility of extending the single-molecule magnetism phenomenon to lanthanide-

containing species. This exciting research area thus offers the tantalizing possibility of

being able to raise the blocking temperature (TB) Of SMMs to above that of the Mnl2

family .

3.4 Experimental

3.4.1 Syntheses

All chemicals were used as received unless otherwise stated. All manipulations

were performed under aerobic conditions. [Mn30(O2CPh)6 py>2(H20)] (2)54 and

(NBun4)[Mn402(O2C~h)9(H20)] (3)55 were prepared as described previously.









[MnllNd40s(OH)s(O2CPh)16(NO3)s(H20)3] 21 MeCN (8). Solid Nd(NO3)3- 6

H20 (0.32g, 0.74 mmol) was added to a solution of complex 3 (0.50g, 0.31 mmol) in

MeOH/MeCN (1ml/20ml). After stirring for 20 min. the solution was filtered and the

brown filtrate slowly concentrated by evaporation. Within a week nice dark red crystals

of 8 were obtained in isolated 50% yield. These were collected by filtration and dried in

vacuo. The dried solid analyzed as fully desolvated. Anal. Called (Found) for

C112H94N5O66Mnl 1Nd4: C, 3 5.90 (3 6.13); H, 2. 53 (2.52); N, 1.87 (1.68). Selected IR data

(KBr, cm )~: 3420 (br), 1599 (m), 1562 (m), 1492 (w), 1449 (w), 1392 (s), 1178 (w),

1025 (w), 715 (s), 679 (m), 604 (m), 475 (w).

[MnllEu40s(OH)s(O2CPh)16(NO3)s(H20)3] (9). Solid Eu(NO3)3- 5 H20 (0.32g,

0.74 mmol) was added to a stirring solution of complex 2 (0.50g, 0.45 mmol) in

MeOH/MeCN (1:20 v/v). After stirring for 20 min. the solution was filtered and the

brown filtrate slowly concentrated by evaporation. Within a week, nice orange crystals of

9 were obtained in isolated 50% yield. These were collected by filtration and dried in

vacuo. The dried solid analyzed as solvent free. Anal. Called (Found) for

C112H94N5O66MnllEu4: C, 35.61 (35.45); H, 2.51 (2.45); N, 1.85 (1.71). Selected IR data

(KBr, cm )~: 3419 (br), 1599 (m), 1560 (m), 1491 (w), 1449 (w), 1385 (s), 1178 (w),

1025 (w), 816(w), 715 (s), 679 (m), 604 (m), 474 (w).

[MnllGd40s(OH)s(O2CPh)16(NO3)s(H20)3] 15 MeCN (10). Solid Gd(NO3)3- 6

H20 (0.41g, 0.91 mmol) was added to a solution of complex 2 (0.50g, 0.45 mmol) in

MeOH/MeCN (1:20 v/v). After stirring for 20 min. the solution was filtered and the

brown filtrate slowly concentrated by evaporation. Within a couple of days dark red

crystals of 10 were obtained in isolated 55% yield. These were collected by filtration and









dried in vacuo. The dried solid analyzed as solvent free. Anal. Called (Found) for

C112H94N5O66MnllGd4: C, 35.41 (35.15); H, 2.49 (2.44); N, 1.84 (1.69). Selected IR data

(KBr, cm )~: 3421 (br), 1599 (m), 1562 (m), 1492 (w), 1449 (w), 1392 (s), 1178 (w),

1025 (w), 715 (s), 679 (m), 604 (m), 476 (w).

[MnllDy40s(OH)6(OMe)2(O2C~h)16(NO3)s(H20)3] 15 MeCN (11). Solid

Dy(NO3)3- 5 H20 (0.40g, 0.90 mmol) was added to a stirring solution of complex 2

(0.50g, 0.45 mmol) in MeOH/MeCN (1:20 v/v). After stirring for 20 min. the solution

was filtered and the brown filtrate slowly concentrated by evaporation. Within a couple

of days nice dark black crystals of 11 were obtained in isolated 55% yield. These were

collected by filtration and dried in vacuo. The dried solid analyzed as fully desolvated.

Anal. Called (Found) for C114H98N5O66MnllDy4: C, 35.75 (35.91); H, 2.53 (2.56); N, 1.83

(1.74). Selected IR data (KBr, cm )~: 3421 (br), 1599 (m), 1561 (m), 1492 (w), 1449 (w),

1385 (s), 1178 (w), 1025 (w), 716 (s), 680 (m), 606 (m), 567(w), 477 (w).

[MnllHo408(Hs(OCH)18(O2C3)s(NO3)3(H2) ] 21 MeCN (12). Solid Ho(NO3)3- 5

H20 (0.40g, 0.90 mmol) was added to a solution of complex 3 (0.50g, 0.31 mmol) in

MeOH/MeCN (1:20 v/v). After stirring for 20 min. the solution was filtered and the

brown filtrate slowly concentrated by evaporation. Within a week nice dark red crystals

of 12 were obtained in isolated 55% yield. These were collected by filtration and dried in

vacuo. The dried solid analyzed as solvent-free. Anal. Called (Found) for

C126H104N3064Mnl 1Ho4: C, 3 8.33 (3 8.20); H, 2.66 (2.52); N, 1.06 (1.04). Selected IR data

(KBr, cm )~: 3420 (br), 1600 (m), 1560 (m), 1449 (w), 1386 (s), 1178 (w), 1025 (w), 716

(s), 680 (m), 606 (m), 476 (w).

[MnllTb40s(OH)6(OCH2Ph)3(O2CPh)2o(PhCH20 H) 20]3PhCH20H (13).









Solid Tb(NO3)3- 5 H20 (0.19g, 0.45 mmol) was added to a solution of complex 3 (0.5g,

0.31 mmol) in PhCH20H/MeCN (5ml/20ml). After stirring for 20 min. the solution was

filtered and the brown filtrate slowly concentrated by evaporation. Within a couple of

weeks nice dark red crystals of 13 were obtained in isolated 55% yield. These were

collected by filtration and dried in vacuo. The dried solid analyzed as 13- 2PhCH20H.

Anal. Called (Found) for C189H161062MnllTb4: C, 48.67 (48.51); H, 3.48 (3.35). Selected

IR data (KBr, cm )~: 3415 (br), 1601 (m), 1563 (m), 1449 (w), 1385 (s), 1177 (w), 1023

(w), 718 (s), 680 (m), 606 (m), 562 (w), 473 (w).

[Mn2Yb202(O2CPh)6(OMe)4(MeOH)4]* 2MeOH (14). To a solution of complex 3

(0.5g, 0.31 mmol) in MeCN/MeOH (20/5 mL) was added Yb(NO3)3-5H20 (0.31 g, 0.62

mmol) and stirred for 25 min to give a brown solution. This was filtered and more MeOH

(10 mL) added to the filtrate. The resulting solution was slowly concentrated by

evaporation at room temperature for 5 days, during which time brown crystals of 14*

2MeOH slowly grew. These were collected in 10% yield and dried in vacuum. Anal.

Calc. (Found) for 14: C 40.94 (40.87), H 3.99 (3.96). Selected IR data (KBr, cm )~: 3426

(br), 1600 (m), 1540 (m), 1384 (s), 1026 (w), 718 (s), 682 (m), 624 (s), 545 (m), 476 (w).

[Mn2 202(O2CPh)6(OMe)4(MeOH)4]*' 2MeOH (15). To a solution of complex 3

(0.5g, 0.31 mmol) in MeCN/MeOH (20/5 mL) was added Y(NO3)3-5H20 (0.24 g, 0.62

mmol) and stirred for 25 min to give a brown solution. This was filtered and more MeOH

(10 mL) added to the filtrate. The resulting solution was left undisturbed at room

temperature for a week to slowly produce orange crystals of 15 2MeOH, which were

collected in 25% yield and dried in vacuum. The same product was obtained using

YCl3-6H20 instead of Y(NO3)3-6H20. Anal. Calc. (Found) for 15: C 46.24 (46. 15), H









4.50 (4.46). Selected IR data (KBr, cm )~: 3412(br), 1594(m), 1545 (s), 1448 (w), 1391

(s), 1025 (w), 718 (s), 681 (m), 623 (s), 540 (m), 474 (w).

3.4.2 X-ray Crystallography

Data were collected on a Siemens SMART PLATFORM equipped with a CCD

area detector and a graphite monochromator utilizing MoK, radiation (h = 0.71073 A+).

Suitable crystals of solvated 8, 10, 11, 12, 13, 14, and 15 were attached to glass fibers

using silicone grease and transferred to a goniostat where they were cooled to 173 K for

data collection. An initial search of reciprocal space revealed a triclinic cell with the

choice of space group being Pl for the MnllLn4 COmplexes 8-13, and a monoclinic cell

with the choice of space group being P21/c for the Mn2Ln2 COmplexes 14-15,

respectively. Cell parameters were refined using up to 8192 reflections. A full sphere of

data (1850 frames) was collected using the co-scan method (0.30 frame width). The first

50 frames were re-measured at the end of data collection to monitor instrument and

crystal stability (maximum correction on I was < 1 %). Absorption corrections by

integration were applied based on measured indexed crystal faces. The structures were

solved by direct methods in SHELXTL6,56a and refined on F2 USing full-matrix least

squares. The non-HI atoms were treated anisotropically, whereas the hydrogen atoms were

placed in calculated, ideal positions and refined as riding on their respective carbon

atoms.

The asymmetric unit of 8-21MeCN consists of half the MnllNd4 ClUSter lying on an

inversion centre, along with 10.5 MeCN molecules as solvents of crystallization. A total

of 897 parameters were included in the structure refinement using 9582 reflections with I

> 2o(I) to yield R1 and wR2 Of 8.63% and 12.70%, respectively.









For 10- 15MeCN, the asymmetric unit consists of half the MnllGd4 ClUSter located

on an inversion center and 7.5 MeCN molecules as crystallization solvents. All H atoms

of the water ligands could not be located and were not included in the final refinement

cycles. A total of 907 parameters were included in the structure refinement using 10499

reflections with I > 2o(I) to yield R1 and wR2 Of 6.56% and 16.22%, respectively.

The asymmetric unit of 11-15MeCN consists of half the MnllDy4 ClUSter lying on

an inversion centre, along with 7.5 MeCN molecules as solvents of crystallization. A

total of 1041 parameters were included in the structure refinement using 12052

reflections with I > 2o(I) to yield R1 and wR2 Of 4.25% and 10.06%, respectively.

The asymmetric unit of 12-21MeCN consists of half the MnllHo4 COmplex lying on

an inversion centre, along with 10.5 MeCN molecules as solvents of crystallization. A

total of 934 parameters were included in the structure refinement using 14245 reflections

with I > 2o(I) to yield R1 and wR2 Of 5.19% and 14.00%, respectively.

For all of the above mentioned complexes some of the solvent molecules were

disordered and could not be modeled properly, thus the SQUEEZE program,56b a part of

the PLATON56c package of crystallographic software, was used to calculate the solvent

disorder area and remove its contribution to the overall intensity data. All of the

complexes 8, 10, 11, and 12 have one MeCN (50% occupancy) close to a disorder on one

of the Ln ions where one coordination site is 50% NO3- and another 50% H20. The

acetonitrile exists only with the water on one of the four Ln ions in each of the complexes

(for example Dy2 in the MnllDy4 COmplex). Thus, although all complexes are

centrosymmetric, they have odd number of NO3, H20 and MeCN molecules because of









the earlier stated disorder of a NO3 ion related in occupancy with the MeCN/H20

molecules.

The asymmetric unit of 13-3PhCH20H consists of half the MnllTb4 and three

halves of benzyl alcohol solvent molecules (all three were disordered and each were

refined in two parts). Tb 1 has a disorder of a water molecule and benzyl alcohol anion in

the same coordination position having 032 common to both. Thus the C and H atoms of

this ligand were refined with 50% occupancies while the water protons of the water

counterpart were not located and not included in the final refinement. The phenyl groups

of the solvent molecules, and those of C22 and C82 ligands, were refined as rigid bodies

and constrained to maintain a perfect hexagon. The hydroxyl protons and those of the

water molecules were not located and were not included in the final refinement cycles. A

total of 744 parameters were refined in the final cycle of refinement using 29065

reflections with I > 2o(I) to yield R1 and wR2 Of 7.73% and 19.42%, respectively.

The asymmetric unit of 14-2MeOH consists of half the Mn2 b2 ClUSter and a

methanol molecule of crystallization. A total of 378 parameters were included in the final

cycle of refinement using 4821 reflections with I > 2o(I) to yield R1 and wR2 Of 3. 11%

and 7.67%, respectively.

The asymmetric unit of 15-2MeOH consists of half the Mn2 2 ClUSter and a

methanol molecule of crystallization. A total of 378 parameters were included in the final

cycle of structure refinement using 2803 reflections with I > 2o(I) to yield R1 and wR2 Of

3.48% and 7.27%, respectively.














CHAPTER 4
HIGH NUCLEARITY COMPLEXES: HOMOVALENT
[Th6MnloO22(OH)2(O2C~h)16(NO3)2(H20)s] AND MIXED-VALENT
[Mn705(OR)2(O2CPh)9(terpy)] (R = Me, CH2Ph) DISPLAYINTG SLOW-
MAGNETIZATION RELAXATION

4.1 Introduction

We have had a longstanding interest in the development of manganese carboxylate

cluster chemistry, mainly due to its relevance to a variety of areas, including bioinorganic

chemistry," nanoscale magnetic materials -" and catalysis of various oxidation

processes.7 For example, Mn carboxylate clusters are the primary source of single-

molecule magnets (SMMs), individual molecules that retain their magnetization

orientation below a blocking temperature in the absence of an applied field.9 In recent

work, we and others have turned our attention to high-nuclearity mixed 3d/4f clusters of

Mn as a route to potentially interesting new species, and a number of heterometallic

complexes of this type are now available.37, 38, 72 In addition, both mixed 3d/4d and 3d/5d

SMMs containing Mn are now known.73

As already illustrated in Chapters II and III, mixed Mn/Ln systems are a rich source

of SMMs. An obvious extension of the above-stated efforts in high nuclearity mixed-

metal cluster chemistry is to ask whether 3d/5f clusters might be accessible with Mn, and

if yes, whether they would function as SMMs. We have taken up this challenge using Th.

There are relatively few well characterized transition metal/actinide complexes, among

which are the dinuclear metal-metal bonded M-An organometallic complexes (M= Fe, Ru

and An= Th, U)74a and the family of linear trimetallic M2IIU' (M= Co, Ni, Cu, Zn)









complexes containing a hexadentate Schiff base.74b However, only one of these contains

Mn, trinuclear [MnU202L2 97)4] (L- = 1,7-diphenyl-1,3,5,7-heptane-tetronato).

Although thorium is used in a wide array of products and processes, the cluster chemistry

of Th is poorly developed compared to transition metals: Currently there are metal-

organic frameworks76a and organically templated thorium complexes76b known, and the

largest molecular Th complex is Th6.77 We can now report the first mixed Mn/Th

molecular cluster, MnloTh6, which we believe to be the prototype of a potentially large

new area of cluster chemistry.

Additionally, although mixed-metal systems are a very exciting research area, they

still suffer from the fact that in a multi-component paramagnetic system there is more

than one g-tensor value. Hence, magnetic characterization of these systems is very

difficult. Therefore, magnetic parameters such as S, D, and isotropic-g values are difficult

to obtain with the current fitting programs which we possess,46 mainly because of second-

order spin-orbit coupling and higher order anisotropy terms. Thus, in the area of single-

molecule magnetism, research on homometallic Mn containing SMMs is still the primary

focus, with the principal obj ective being the preparation of a SMM that behaves as a

magnet at technologically relevant temperatures, i. e., at least 77 K (that of liquid N2).

This goal has been approached primarily by two means: (i) the preparation of novel 3d

metal carboxylate clusters possessing differing topologies that may behave as SMMs and

(ii) the addition of new derivatives to already existing families of SMMs so that structural

features may be correlated with magnetic properties and ultimately, a rational synthetic

method for improved SMMs may be developed. The first SMM discovered was

[Mnl2012(O2CMe)16(H20)4 ,7 and synthetic manipulation of this complex has provided a




Full Text

PAGE 1

MIXED-METAL MOLECULAR CO MPLEXES: SINGLE-MOLECULE NANOMAGNETS AND BIOI NORGANIC MODELS OF THE WATER OXIDIZING COMPLEX OF PHOTOSYSTEM II By ABHUDAYA MISHRA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by ABHUDAYA MISHRA

PAGE 3

I dedicate this document to my mom, dad a nd two brothers, for their love and confidence in me as I pursue the endless journey of life

PAGE 4

iv ACKNOWLEDGMENTS I would like to take this opportunity to thank my research advisor, Professor George Christou, for all his help, guidance, inspiration, and cons tant encouragement rendered to me as I pursued my career goals in his research group. I would also like to thank my other committee members, Dr. C. R. Ma rtin, Dr. M. J. Scott, Dr. D. R. Talham, and Dr. S. Hill, for some stimulating discus sions, and the inspiration which I drew from their research work, which in turn kindled my own creativity. In today's world, research is highly interd isciplinary and I would like to take this opportunity to thank the many research scientis ts with whom I have collaborated during my doctoral studies. These collaborators in clude Dr. Khalil A. Abboud and his staff at UFCXC for solving the X-ray crystal struct ures, and Dr. Wolfgang Wernsdorfer, who provided cryogenic magnetic measurements (below 1.8 K) on the compounds mentioned in this dissertation. Additionall y, I would like to express my heartfelt appreciation to Dr. Junko Yano and Dr. Vittal K. Yachandra, at Lawrence Berkeley National Labs, for performing XAS measurements on complexes wh ich were relevant to the bioinorganic research. In the course of the collaborati ve research performed, Khalil, Wolfgang, and Junko have indeed become wonderful friends. And speaking of friends, what would resear ch be without any camaraderie between one's colleagues! I would like to thank all the Christou grou p members (past and present) who made Gainesville worth living, for the past five years. Special thanks go to Dolos for her support and friendship. Alina completes the trio for the aw esome time all three of us

PAGE 5

v have spent together laughing, ta lking, and of course drinki ng. I would also like to thank Tasos, Carol and Monica who helped me get started in this lab and also for the fun companionship which they provided. The Indian gang of Parul, Ranjan and Rashmi also deserve special mention. Last, but not least, I would like to acknowledge the unconditional support of my two brothers, A bhishek and Animesh, and my mother and father, who have always been there for me whenever I needed something. I am forever indebted to them for their constant pr ide, encouragement, love, and unwavering confidence in me as I undertook the daunting task of my doctoral studies. Indeed, my family has made this wonderful journey much more meaningful.

PAGE 6

vi TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................ix LIST OF FIGURES...........................................................................................................xi ABSTRACT.......................................................................................................................xv CHAPTER 1 GENERAL INTRODUCTION....................................................................................1 2 SINGLE-MOLECULE MAGNETS: A NOVEL FAMILY OF MnIII / CeIV COMPLEXES WITH A [Mn8Ce8O8]12+ CORE.........................................................16 2.1 Introduction...........................................................................................................16 2.2 Results and Discussion.........................................................................................17 2.2.1 Syntheses....................................................................................................17 2.2.2 Description of Structures............................................................................21 2.2.2.1 X-ray crystal structures of complexes 4 7 ......................................21 2.2.2.2 Structural Comparison of complexes 4 7 .......................................25 2.2.3 Magnetochemistry of Complexes 4 5 and 7 ..............................................28 2.2.3.1 DC studies........................................................................................28 2.2.3.2 AC studies........................................................................................33 2.2.3.3 Hysteresis studies below 1.8 K........................................................34 2.3 Conclusions...........................................................................................................38 2.4 Experimental.........................................................................................................39 2.4.1 Syntheses....................................................................................................39 2.4.2 X-ray Crystallography................................................................................41 3 SINGLE-MOLECULE MAGNETS: SYNTHESES AND MAGNETIC CHARACTERIZATION OF A NOVEL FAMILY OF HETEROMETALLIC MANGANESE-LANTHANI DE COMPLEXES.......................................................44 3.1 Introduction...........................................................................................................44 3.2 Results and Discussion.........................................................................................46 3.2.1 Syntheses....................................................................................................46 3.2.2 Description of Structures............................................................................49

PAGE 7

vii 3.2.2.1 X-ray crystal structure of complexes 11 and 13 ...............................49 3.2.2.2 Structural Comparison of Complexes 8 13 ......................................53 3.2.2.3 Structural descriptions of complexes 14 and 15 ...............................55 3.2.3 Magnetochemistry of Complexes 8 13 and 15 ..........................................58 3.2.3.1 DC studies of complexes 9 10 11 and 13 ......................................58 3.2.3.2 DC studies of [Mn2Y2O2(O2CPh)6(OMe)4(MeOH)4] ( 15 )...............63 3.2.3.3 AC studies of complexes 9 10 11 and 13 ......................................64 3.2.3.4 Hysteresis studies below 1.8 K on complexes 8 13 .........................68 3.3 Conclusions...........................................................................................................75 3.4 Experimental.........................................................................................................76 3.4.1 Syntheses....................................................................................................76 3.4.2 X-ray Crystallography................................................................................80 4 HIGH NUCLEARITY COMPLEXES: HOMOVALENT [Th6Mn10O22(OH)2(O2CPh)16(NO3)2(H2O)8] AND MIXED-VALENT [Mn7O5(OR)2(O2CPh)9(terpy)] (R = Me, CH2Ph) DISPLAYING SLOWMAGNETIZATION RELAXATION........................................................................83 4.1 Introduction...........................................................................................................83 4.2 Results and Discussion.........................................................................................86 4.2.1 Syntheses....................................................................................................86 4.2.2 Description of Structures............................................................................88 4.2.2.1 X-ray crystal structure of [Mn10Th6O22(OH)2(O2CPh)16(NO3)2(H2O)8] ( 16 )....................................88 4.2.2.2 X-ray crystal structures of [Mn7O5(OR)2(O2CPh)9(terpy)] complexes ( 17 18 )....................................................................................93 4.2.3 Magnetochemistry of Complexes 16 17 and 18 ........................................99 4.2.3.1 DC studies........................................................................................99 4.2.3.2 AC studies......................................................................................107 4.2.3.3 Hysteresis studies below 1.8 K......................................................111 4.3 Conclusions.........................................................................................................113 4.4 Experimental.......................................................................................................114 4.4.1 Syntheses..................................................................................................114 4.4.2 X-ray Crystallography..............................................................................116 5 THE FIRST STRONTIUM-MANGANE SE CLUSTER: SINGLE-MOLECULE MAGNETISM AND Sr-EXAFS COMPARISON WITH THE WATER OXIDIZING COMPLEX OF PHOTOSYSTEM II.................................................118 5.1 Introduction.........................................................................................................118 5.2 Results and Discussion.......................................................................................120 5.2.1 Syntheses..................................................................................................120 5.2.2 Description of Structures..........................................................................121 5.2.3 Magnetochemistry of Complex 19 ...........................................................124 5.2.3.1 DC studies of 19 .............................................................................124 5.2.3.2 AC studies of 19 .............................................................................128 5.2.3.3 Hysteresis studies below 1.8 K......................................................130

PAGE 8

viii 5.2.4 X-ray Absorption Spectroscopy of Complex 19 ......................................132 5.2.4.1 Sr EXAFS of 19 .............................................................................133 5.2.4.2 Mn EXAFS of 19 ...........................................................................138 5.3 Conclusions.........................................................................................................140 5.4 Experimental.......................................................................................................141 5.4.1 Synthesis...................................................................................................141 5.4.2 X-ray Crystallography..............................................................................142 5.4.3 XAS studies..............................................................................................143 THE FIRST FAMILY OF HETERO METALLIC CALCIUM-MANGANESE COMPLEXES: Ca-EXAFS AND -XANES COMPARISON WITH THE WATER OXIDIZING COMPLEX OF PHOTOSYSTEM II..................................145 6.1 Introduction.........................................................................................................145 6.2 Results and Discussion.......................................................................................147 6.2.1 Syntheses..................................................................................................147 6.2.2 Description of the St ructures of complexes 20 22 ...................................150 6.2.3 Magnetochemistry of Complexes 20 22 ..................................................156 6.2.4 Calcium XAS studies of complexes 20 22 ...............................................162 6.2.4.1 Calcium EXAFS of complexes 20 22 ............................................164 6.2.4.2 Calcium XANES of complexes 20 22 ...........................................169 6.3 Conclusions.........................................................................................................172 6.4 Experimental.......................................................................................................175 6.4.1 Syntheses..................................................................................................175 6.4.2 X-ray Crystallography..............................................................................177 6.4.3 XAS Studies.............................................................................................178 APPENDIX A BOND DISTANCES AND ANGLES......................................................................180 B LIST OF COMPOUNDS..........................................................................................188 C PHYSICAL MEASUREMENTS.............................................................................190 LIST OF REFERENCES.................................................................................................193 BIOGRAPHICAL SKETCH...........................................................................................206

PAGE 9

ix LIST OF TABLES Table page 2-1 Crystallographic data for 4 5 6 and 7 ....................................................................27 2-2 Comparison of the magnetic parameters of complexes 4 5 and 7 ..........................38 3-1 Crystallographic data for 10 11 and 13 ...................................................................52 3-2 Bond Valence Sums for the Mn atoms of complexes 15 (Y) and 14 (Yb)..............57 3-3 Bond Valence Sums for the O atoms of complexes 15 (Y) and 14 (Yb).................57 3-4 Selected bond distances () and angles () for complexes 14 and 15 ......................58 3-5 Comparison of the SMM parameters of complexes 10 11 and 13 ..........................74 4-1 Bond valence sum calculations for the Mn ions in complex 16 ...............................92 4-2 Bond valence sum calculations for the Mn ions in complexes 17 and 18 ...............97 4-3 Crystallographic data for 17 and 18 .........................................................................98 4-4 Comparison of the magnetic suscep tibility parameters of complexes 17 and 18 ..109 5-1 Crystallographic data for [SrMn14O11(OMe)3(O2CPh)18(MeCN)2].......................124 5-2 Selected bond and interato mic distances () for complex 19 ................................135 5-3 Least-squares fits of Fourier-filter ed peaks I and II of Sr EXAFS data on Complex 19 and the Sr-substituted WOC in the S1 state.......................................136 6-1 Bond valence sum calculations for the Mn and O atoms of complex 20 ...............152 6-2 Crystallographic data of 20 21 and 22 .................................................................155 6-3 Selected interatomic and bond distan ces () for the Ca and Mn atoms of complexes 20 22 ....................................................................................................166 6-4 Least-squares fits of Fourier-filter ed peaks I and II of Ca EXAFS data on complexes 20 and 21 and the WOC in the S1 state...............................................167

PAGE 10

x A-1 Selected interatomic distances () and angles () for [Mn8CeO8(O2CMe)12(H2O)4] ( 4 ) and [Mn8CeO8(O2CMe)12(py)4] ( 5 )..................180 A-2 Selected interatomic distances () and angles () for [Mn8CeO8(O2CCHPh2)12(H2O)4] ( 7 )......................................................................181 A-3 Selected interatomic distances () and angles () for [Mn11Gd4O8(OH)8(O2CPh)16(NO3)5(H2O)3] ( 10 )..................................................182 A-4 Selected interatomic distances () and angles () for [Mn11Dy4O8(OH)6(OMe)2(O2CPh)16(NO3)5(H2O)3] ( 11 )......................................183 A-5 Selected interatomic distances () and angles () for [Mn11Tb4O8(OH)6(OCH2Ph)2(O2CPh)16(NO3)5(H2O)3] ( 13 ).................................184 A-6 Selected interatomic distances () and angles () for [Mn10Th6O22(OH)2(O2CPh)16(NO3)2(H2O)8] ( 16 )..................................................185 A-7 Selected interatomic distances () and angles () for [Mn7O5(OMe)2(O2CPh)9(terpy)] ( 17 )....................................................................186 A-8 Selected interatomic distances () and angles () for [Mn7O5(OCH2Ph)2(O2CPh)9(terpy)] ( 18 )..............................................................187

PAGE 11

xi LIST OF FIGURES Figure page 1-1 Representations of magnetic dipole arrangements in (i) paramagnetic, (ii) ferromagnetic, (iii) antiferromagnetic, and (iv) ferrimagnetic materials...................3 1-2 Typical hysteresis loop of a magnet, wh ere M is magnetization, H is the applied magnetic field and Mr is the saturation valu e of the magnetization...........................5 1-3 Representative plots of the potenti al energy versus (a) the magnetization direction and (b) the ms sublevels..............................................................................8 1-4 In-phase M T and out-of-phase M Possible tunneling mechanisms (top) for SMMs and (bottom) typical hystere sis loops with steps, for a Mn12 SMM.............10 1-5 The S state scheme as proposed by Kok for the oxidation of water. Arrangement of 4 Mn and 1 Ca atoms in the two latest crystal structures of the WOC of PS II...12 1-6 Crystal structure of the WOC at 3.0 and the 3.5 Mn4Ca cubane-containing crystal structure. Possible Mn4CaOx topologies suggested by biophysical studies.13 2-1 PovRay representation at the 50% probabi lity level of the X-ray crystal structure of 4 ............................................................................................................................22 2-2 PovRay representation at the 50% probabi lity level of the X-ray crystal structure of 5 (left) and 6 (right)..............................................................................................23 2-3 PovRay representation at the 50% probabi lity level of the X-ray crystal structure of 7 ............................................................................................................................24 2-4 Comparison of the [Mn8CeO8]12+ core, with the [Mn12O12]16+ core. The common core of complexes 4 5 6 and 7 ................................................................................25 2-5 Plots of MT vs T for complexes 4 5 and 7 ..............................................................29 2-6 Magnetization (M) vs field (H) and temperature (T ) data, plotted as reduced magnetization (M/N B) vs H/T, for (left) complex 4 and (right) for complex 7 ....30 2-7 Magnetization ( M ) vs applied magnetic field ( H ) hysteresis loops: (left) for 4 and (right) for 5 M is normalized to its saturation value, Ms, for both plots...........35

PAGE 12

xii 2-8 Magnetization ( M ) vs applied magnetic field ( H ) hysteresis loops for 7 M is normalized to its saturation value, Ms, for both plots...............................................37 3-1 PovRay representation at the 50% pr obability level of the X-ray crystal structures of 11 (left) and 13 (right).........................................................................50 3-2 PovRay representation at th e 50% probability level of the [Mn11Dy4O8(OH)6(OMe)2]21+ core of 11 (left) and the [Mn11Tb4O8(OH)6(OCH2Ph)2]21+ core of 13 (right).................................................51 3-3 PovRay representation at the 50% probability level of the [Mn11Gd4O8(OH)8]21+ core of complex 10 ...................................................................................................54 3-4 PovRay representations of the crystal structures of complexes 14 (left) and 15 (right). Comparison of th e cores of complexes 14 (left) and 15 (right), emphasizing their near superimposibility................................................................56 3-5 MT vs T plots for complexes 9 ( ), 10 ( ), 11 ( ), and 13 ( ).............................59 3-6 Magnetization (M) vs field (H) and temperature (T ) data, plotted as reduced magnetization (M/N ) vs H/T, for complex 10 ......................................................62 3-7 Plot of MT (solid circles, ) vs T for complex 15 The solid line in the MT vs T plot is the fit of the data; see the text for the fit parameters.....................................63 3-8 Ac susceptibility of complex 11 (Top) in-phase signal ( M ) plotted as M T vs T ; and (bottom) out-of-phase signal M vs T ...............................................................66 3-9 Ac susceptibility of complex 13 (Top) in-phase signal ( M ) plotted as M T vs T ; and (bottom) out-of-phase signal M vs T ...............................................................67 3-10 Magnetization ( M ) vs dc field ( H ) hysteresis loops fo r single crystals of 8 (top; left), 9 (top; right), 10 (bottom; left), and 12 (bottom; right)...................................69 3-11 Magnetization ( M ) vs dc field ( H ) hysteresis loops fo r single crystals of 11 (top), and 13 (bottom)........................................................................................................70 3-12 Magnetization vs time decay plots for crystals of complex 13 at the indicated temperatures. Arrhenius plot usin g the resulting relaxation time () versus T data.72 3-13 Arrhenius plot of the relaxation time () versus 1/ T constructed from Dc magnetization decay data for complexes 11 (left), and 10 (right)...........................73 4-1 PovRay representation of the X -ray crystal struct ure (left) of 16 and (right) the 2:3:6:3:2 (Mn:Th:Mn:Th:M n) layer structure of 16 Bottom depicts the labeled [Mn10Th6O22(OH)2]18+ core of complex 16 ..............................................................89

PAGE 13

xiii 4-2 PovRay representation at the 50% probabi lity level of the X-ray crystal structure of 17 and (bottom) its labeled [Mn7O5(OMe)2(terpy)]9+ core.................................94 4-3 PovRay representation at the 50% probabi lity level of the X-ray crystal structure of 18 (top) and its stereopair (bottom).....................................................................96 4-4 Plot of MT vs T for complex 16 Magnetization (M) vs field (H) and temperature (T) data, plotted as reduced magnetization ( M/N B) vs H/T for 16 ..100 4-5 Plots of MT vs T for complexes 17 and 18 ............................................................101 4-6 Magnetization (M) vs field (H) and temperature (T) data for (left) complex 17 and (right) for complex 18 Two-dimensional contour plot of the error surface for the D vs g fit for complexes 17 (left), and 18 (right)........................................103 4-7 Schematic representation and justifi cation of the spin alignment in the Mn7 complexes based on a fused cubane/butterfly arrangement...................................106 4-8 Plots of the in-phase (as M T ) ac susceptibility signals vs T for complex 16 (left), and complexes 17 and 18 (right). The measurement is in a 3.5 G ac field oscillating at a 500 Hz frequency...........................................................................108 4-9 Magnetization ( M ) vs applied magnetic field ( H ) hysteresis loops for singlecrystals of 16 Hysteresis loops at temperat ures of 0.7 and 0.1 K for single crystals of 18 ..........................................................................................................112 5-1 PovRay representation at the 50% probabi lity level of the X-ray crystal structure of 19 and its stereopair..........................................................................................122 5-2 The labeled [Mn14SrO11(OMe)3]18+ core of complex 19 ........................................123 5-3 Plot of MT vs T for complex 19 Magnetization (M) vs field (H) and temperature (T) data, plotted as reduced magnetization (M/N B) vs H/T, for 19 ..125 5-4 Two-Dimensional contour plot of the error surface for the reduced magnetization (M/N B) vs H/T fit for complex 19 Three-Dimensional mesh plot of the error vs D vs g for the same fit for 19. ........................................................127 5-5 Plot of the in-phase (as M T ) and out-of-phase ( M ) AC susceptibility signals vs temperature for complex 19 ...............................................................................129 5-6 Magnetization ( M ) vs applied magnetic field ( H ) hysteresis loops for singlecrystals of 19 at a fixed sweep rate of 0.14 T/s and at 0.04 K...............................130 5-7 Plot of the magnetization versus time decay data of 19 .........................................131 5-8 k3-weighted Sr K-edge EXAFS spectra of the Mn14Sr compound (red) and Sr reactivated PS II samples in the S1 state (black)....................................................134

PAGE 14

xiv 5-9 PovRay representation of two opencubane containing sub-units within 19 Left depicts a sub-unit which resembles th e 3.0 crystal structure of the WOC26 of PS II........................................................................................................................137 5-10 k3-weighted Mn K-edge E XAFS spectra of the Mn14Sr compound (red) and PS II samples in the S1 state (black)............................................................................139 6-1 PovRay representation of the crystal structure of 20 and its labeled [Mn13Ca2O10(OH)2(OMe)2]18+ core........................................................................151 6-2 Schematic representation of the three bridging modes found in complex 20 ........151 6-3 PovRay representation of the crystal structure of the doubly charged anion of 21 and its labeled [Mn11Ca4O10(OH)2(OMe)2]18+ core................................................153 6-4 PovRay representation of the crystal structure of 22 and its labeled [Mn8CaO6(O2P(Ph)2)4(OMe)2(MeOH)2]10+ core....................................................154 6-5 Plots of MT vs T for Mn13Ca2 ( ) and Mn11Ca4 ( ) complexes............................157 6-6 Magnetization (M) vs field (H) and temperature (T ) data, plotted as reduced magnetization (M/N B) vs H/T, for complex 20 ....................................................158 6-7 Magnetization (M) vs field (H) and temperature (T ) data, plotted as reduced magnetization (M/N B) vs H/T, for complex 22 ....................................................160 6-8 MT vs T plot for 22 and for each of the two cubanes of the Mn8Ca complex. Model used to fit the exchange interactions...........................................................161 6-9 k3-weighted Ca K-edge EXAFS spectra of the Ca/Mn complexes and PS II S1 state, and Fourier transfor ms of the EXAFS spectra..............................................165 6-10 k3-weighted Ca K-edge E XAFS spectra of the Mn13Ca2 and Mn11Ca4 complexes and PS II S1 state, and Fourier transf orm of the EXAFS spectra...........................168 6-11 Ca K-edge XANES spectra of Ca/Mn complexes 20 22 and PS II in the S1 state, and their second derivative spectra. (Inset) The pre-edge like peaks.....................170 6-12 Ca K-edge EXAFS spectra of the Mn13Ca2 complex (red) and PS II S1 state (black), and Sr K-edge EXAFS spectra of the Mn14Sr complex (red) and Srreactivated PS II samples in the S1 state (black). Mn4Ca and Mn4Sr interactions present in complexes 20 and 19 .............................................................................173 6-13 PovRay representations of units/subunits present in the Mn/Ca complexes 20 22 and in the Mn14Sr complex 19 Sub-units of complex 20 in the first row resemble the 3.0 26 (left) and 3.5 22 (center) crystal structures of the WOC.....174

PAGE 15

xv Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MIXED-METAL MOLECULAR CO MPLEXES: SINGLE-MOLECULE NANOMAGNETS AND BIOI NORGANIC MODELS OF THE WATER OXIDIZING COMPLEX OF PHOTOSYSTEM II By Abhudaya Mishra August 2006 Chair: George Christou Major Department: Chemistry The current burgeoning research in high nuclearity manganese-containing carboxylate clusters is primarily due to their relevance in areas as diverse as magnetic materials and bioinorganic chemistry. In the former, the ability of single molecules to retain, below a critical temperature ( TB), their magnetization v ector, resulting in the observation of bulk magnetizati on in the absence of a fi eld and without long-range ordering of the spins, has termed such mo lecules as Single-Molecule Magnets (SMMs), or molecular nanomagnets. These molecules display superparamagnet like slow magnetization relaxation arising from the combination of a large molecular spin, S and a large and negative magnetoanisotropy, D Traditionally, these nanomagnets have been Mn containing species. An out of the box approach towards synthesizing SMMs is engineering mixed-metal Mn-containing compou nds. An attractive choice towards this end is the use of Lanthanides (L n), which possess both a high spin, S and a large D A family of related MnIII 8CeIV SMMs has been synthesized. However, the Ce ion of these

PAGE 16

xvi complexes is diamagnetic (CeIV). Thus, further investigation has led to the isolation of a family of MnIII 11LnIII 4 complexes in which all but the Ln = Eu complex function as single-molecule nanomagnets. The mixed-meta l synthetic effort has been extended to include actinides with the successful isolation of a MnIV 10ThIV 6 complex, albeit this homovalent complex is not a SMM. In the bioinorganic research, the Water Oxidizing Complex (WOC) in Photosystem II (PS II) catalyzes the oxidation of H2O to O2 in green plants, algae and cyanobacteria. Recent crystal structures of the WOC confirm it to be a Mn4CaOx cluster with primarily carboxylate ligation. To date, various multinucle ar Mn complexes have been synthesized as putative models of the WO C. On the contrary, there ha ve been no synthetic MnCa(Sr) mixed-metal complexes. Thus, in this bi oinorganic modeling research of the WOC, various synthetic methods have been develope d to prepare a variety of heterometallic MnCa(Sr) complexes, namely, Mn13Ca2, Mn11Ca4, Mn8Ca and Mn14Sr; these are the first of their kind. X-ray abso rption spectroscopy has been performed on all of these complexes and the results compared with analogous data on the WOC of PS II. In particular, Ca, Sr, and Mn, EXAFS and XANES re veal a distinct similarity between the sub-units within these complexes and the Mn4CaOx site of the WOC. The data strongly suggest that a single-atom O br idge exists between the Mn atoms and the Ca atom of the WOC.

PAGE 17

1 CHAPTER 1 GENERAL INTRODUCTION Magnetism is a topic which is highly underrated since alt hough it has a profound impact on our day to day life it is much less re searched on in academia. It all started with the ancient Greeks, originally those near the city of Magnesia, and also the early Chinese who knew about strange and rare stones, po ssibly chunks of iron ore struck by lightning, with the power to attract iron in a magica l way. A steel needle stroked with such a "lodestone" became "magnetic" as well, and th e Chinese found that such a needle, when freely suspended, pointed north-south. This led to the discovery and subsequent exploitation of magnets and magnetism, and human civilization has tremendously benefited from it ever since. Magnets are so essential and ubiquitous to a plethora of devices in our daily life that sometimes they are taken for granted. Be it the simple speakers or headphones, or the complicated motors or telecommunication devices, magnets find an application almost everyw here. Modern day magnetic materials include magnetic alloys and oxides, particularly ferrites such as MgFe2O4, which can function in transformer cores, magnetic r ecording or information storage devices. The global market for magnetic materials is valued at $50 bill ion and has a projected growth rate of 10% annually. Additionally, with the advent of GNR (genetics, nanotechnology and robotics) this burgeoning field of magnetism and ma gnetic materials will undoubtedly expand as the 21st century progresses, since magnets will be crucial to the development of the socalled smart materials and smart systems.

PAGE 18

2 The behavior of any magnetic material is essentially dependen t on the presence of unpaired electrons, or more precisely the spin associated with the unpaired electrons. The magnetic field associated with a magnetic substa nce is the result of an electrical charge in motion, specifically the spin and orbital angul ar momenta of electrons within atoms of a material. Hence, the number of unpaired el ectrons and the net interactions of these electron spins govern the re sponse or susceptibility of the material to an applied field. The presence of unpaired electrons on a mate rial classifies it as paramagnetic and the absence makes a material diamagnetic. The elec tron pairs of a diamagnet interact with an applied field, generating a re pulsive field that weakly repels the diamagnet from the applied field; the sign of is negative. In contrast, a para magnet is attracted to an applied magnetic field; the sign of is positive. In the presence of an applied magnetic field the electron spins in paramagnetic materials try to align parallel to the applied field and this effort is opposed by the entropically favored randomizing effect of thermal energy. Thus, removal of the field results in the randomiza tion of the spins. As a result, a paramagnet has a net zero magnetization and may not act as a magnet. However, as the number of metal centers increase, spin-coupling allows cooperative or non-c ooperative interactions that result in parallel or anti-parallel alignment of spins, respectively. Therefore on removal of the applied field if the spins re main aligned parallel, thereby possessing a net magnetic moment, the material is ferromagnetic. Based on the alignment of the magnetic dipole moment of the spins, materials can be classified as paramagnetic, ferromagnetic, antiferromagnetic and ferrimagnetic as shown in Figure 1-1. Beside s these prototypes, magnetic materials can also show spin gl ass, metamagnetic and canted ferromagnetic / antiferromagnetic behavior.

PAGE 19

3 To remain ferromagnetic after the field is removed (i.e., remnant magnetization), the system must be below a critical temperature, TC. Above TC (curie temperature), the thermal energy (kT) is large enough to cause the electron spins to orient randomly and the material behaves as a simple paramagne t. Assuming the material is still below TC and the spins are aligned parallel, applying a field in the reverse direction induces the spins to align in the opposite direction. At zero magnetization, a positive field strength ParamagneticFerromagneticAntiferromagnetic Ferrimagnetic Figure 1-1. Representations of magnetic dipole arrangements in (i) paramagnetic, (ii) ferromagnetic, (iii) antiferromagnetic, and (iv) ferrimagnetic materials. can be seen, known as the coercive field or the hardness of the magnet. This phenomenon can be seen as a hysteresis l oop in a plot of the magnetizati on M versus the applied field, B (Figure 1-2).1 Antiferromagnetism and ferrimagnetis m are the opposite situations to ferromagnetism. An antiferromagnet has its spins aligned antiparallel (opposing) producing a net zero magnetization below a critical temperature, known as the Neel temperature, TN. If the spins align antiparallel but a non-zero magnetization results, the material is described as ferrimagnetic. This is due to spin centers with different magnetic moment magnitudes and hence the net magnetiz ation does not cancel out even though the spin vectors are ordered in an antiparallel fashion. An example is the naturally occurring magnet, magnetite (Fe3O4). CrO2 and LaFeO3 represent common examples of ferromagnetic and antiferromagne tic materials, respectively. At all temperatures, ferro-, antiferroa nd ferri-magnets are composed of domains, or tiny regions in which all the spins are ali gned parallel or antipar allel. The transition

PAGE 20

4 from independent to cooperative behavior in these materials is associated with a curie temperature, Tc. Above Tc, there is enough thermal energy to cause a random alignment of each domain with respect to its nei ghbor, maximizing the entropy while minimizing the magnetization of the system. The applica tion of a strong magnetic field induces the alignment of all of the domains with the fiel d and hence with each other, imparting a net magnetization to the ferroor ferri-magne tic material. As alignment occurs, the interaction of spins becomes strong enough to overcome entropy considerations that maintain the random alignment of the domains.1, 2 When a magnetic field is applied and then removed at a temperature below Tc, the magnetization induced by the field does not entirely disappear, and in some cases can remain equal to the field-induced magnetization. This is in contrast to the be havior observed for paramagnetic systems in which the spins immediately (>10-9 s) randomly reorient following removal of the field. For suppression of the remnant magnetization, a coercive field in the opposite direction is applied, inducing the realignment of the spins in the opposite directi on and resulting in a hysteresis loop (Figure 1-2). Richard Feynman, the Nobel Prize-winning physicist, once gave a lecture called, "There's Plenty Of Room At The Bottom." This seminal lecture laid the foundation of nanotechnology and ever since, in accordance with Moores Law, el ectronic devices have become progressively smaller. In this scen ario it becomes increasingly important to understand the magnetic properties of smaller bistable particles for the storage of information. For information storage a small coercive field (high permeability) with a relatively rectangular shaped hysteresis loop is required so that the two magnetic orientations of the spin can represent zer o (spin-up) and one (spin-down) in the bi nary

PAGE 21

5 digit (bit) system used by current technology. The requirement for information storage is that the system remains at a temperature at which the material exhibits hysteresis while the removal of the stored information involves heating to a temperature above Tc.1 Thus a considerable amount of research is being concentrated on making smaller and smaller materials that behave similar to permanent magnets. Figure 1-2. Typical hysteresis loop of a ma gnet, where M is magnetization, H is the applied magnetic field and Mr is the saturation valu e of the magnetization. One idea is to fragment a ferromagnet or ferrimagnet to a size smaller than a single domain (20-200 nm); therefore all the sp ins within the particle always remain parallel. These particles, known as supe rparamagnets, are composed of randomly oriented spins unless induced by an applied magnetic field. Superparamagnets retain their magnetization when their magnetic relaxation is slowed below a blocking temperature, TB. Problems with this approach include a wide distribution of shapes and sizes.2 Additionally, there is a distribution of ba rrier heights for the interconversion of the spins and these materials are insoluble in or ganic solvents and thus unsuitable for some Hc Mr

PAGE 22

6 applications and studies. Yet another appro ach to obtaining small magnetic materials in the superparamagnet range is to build molecule-based magnet s. Rather than using solids of extended lattices such as oxides, this appr oach uses 3-D lattices of molecular building blocks, which are synthesized from single molecules and selected bridging groups. Considerable research focusi ng on purely organic magnets,3 hybrid organic/inorganic magnets (Awaga et al. ),4 and inorganic cyanide-based network-structured magnets (Miller and Epstein et al. )5 has been done in this area. Advantages of this approach include low density, solubili ty, biocompatibility, transpar ency and high magnetizations. The first example of molecule-based magnets was discovered in 19666 but it took twentyseven more years of research for scientists to realize that a molecule can behave as a magnet by itself, rather than through long -range ordering. Finally, in 1993, the first example of a molecule, [Mn12O12(O2CMe)16(H2O)4], in short Mn12ac, able to behave as a magnet by itself was discovered.7 This discovery led to a totally new approach to nanoscale magnets in which the magnetism was in trinsic to the molecule and not due to interactions between molecules. Magnetic studies in a polyeth ylene matrix of Mn12ac proved the latter hypothesis true and showed the absence of any long-range threedimensional interactions.8 Ever since, polynuclear metal complexes with magnetic behavior similar to Mn12ac and exhibiting superparamagnetic-like properties have been called single-molecule magnets (SMMs).9 The name itself is somewhat misleading because in traditional magnetism, to have a ma gnet it is necessary to have an infinite number of coupled spin centers, but is evoc ative, and can be used provided the above caveat is taken into consideration.10

PAGE 23

7 Since the discovery of SMMs, there has been a great interest in the chemistry and physics communities in understanding this new magnetic phenomenon of singlemolecule magnetism and synthesizing new SMMs. This interest has paid rich dividends with the discovery of many more manganese clusters, as well as vanadium, iron, cobalt and nickel clusters which behave as SMMs. However of the SMMs known to date, the [Mn12O12(O2CR)16(H2O)x] (R= -Me ,-Ph, ,-C6H4-2-Cl, ,-CHCl2 ,-C6F5,-C6H4-2-Br,CH2But ,-CH2Ph with x = 4 and R= -Et -CH2Cl with x = 3) family, also known as Mn12 complexes,7, 10 still possess the best structural and electronic properties for this phenomenon, inasmuch as they display singl e-molecule magnetism behavior at the highest temperatures and behave as a magnet below 4 Kelvin.9, 11 Single-molecule magnets are molecules that can function as nanoscale magnets below a certain blocking temperature. Th ese molecules exhibit slow magnetization relaxation (reorientation) rates, which resu lt in magnetization hys teresis loops. Since SMMs display hysteresis, like any classical magne t, they have potential applications in future magnetic storage devices where one bi t of information can be stored on a single molecule, thereby greatly increasing the data density of information storage devices. Progress towards this end involves the use of smaller materials of nanoand subnanoscale dimensions (like SMMs) that behave as permanent magnets, albeit with functional temperatures in the practical rang e for technological use. Also, due to subnanoscale sizes and monodisperse behavior, these molecules show quantum tunneling of the magnetization (QTM) at the macroscopic level,12 and thus act as a bridge between the quantum and classical understa nding of magnetism. Since QT M is an inherent property of these molecules and they show quantum coherence,13 SMMs are possible candidates as

PAGE 24

8 qubits (quantum bits) in future quantum computers.14 In order to display SMM behavior a molecule should possess a large spin ground state ( S ) and a large negative magnetic anisotropy, gauged by the zero-fi eld splitting (ZFS) parameter D (i.e., Ising or easy-axistype anisotropy). The presence of both a high spin ground state and a large, negative ZFS parameter in a single molecule is rather rare and thus interesting magnetic properties are generated. For ms= +10 +9 +8 +7 +6 +5 +4 +3 +2 +1 ms= 0 -1 -2 -3 -4 -5 -6 -7 -8 -9 ms= -10 ms Energy(b) 10-10 = ms9-9 8 -8 7 -7 6-6 5 -5 4 -4 3 -3 2-2 1 -1 0EnergyMagnetization Direction 10-10 = ms9-9 8 -8 7 -7 6-6 5 -5 4 -4 3 -3 2-2 1 -1 0EnergyMagnetization Direction (a) ms= +10 +9 +8 +7 +6 +5 +4 +3 +2 +1 ms= 0 -1 -2 -3 -4 -5 -6 -7 -8 -9 ms= -10 ms Energy(b) 10-10 = ms9-9 8 -8 7 -7 6-6 5 -5 4 -4 3 -3 2-2 1 -1 0EnergyMagnetization Direction 10-10 = ms9-9 8 -8 7 -7 6-6 5 -5 4 -4 3 -3 2-2 1 -1 0EnergyMagnetization Direction (a) Figure 1-3. Representative plots of the pot ential energy versus (a) the magnetization direction and (b) the ms sublevels, for a Mn12 complex with an S = 10 ground state, experiencing zero-field splitting. example, for the aforementioned family of Mn12 complexes the S = 10 always, and the D varies from -0.3 cm-1 to -0.5 cm-1. The combination of these S and D values lead to a barrier to magnetization reversal (reorientation) given by U = S2|D| for integer spin systems and U = (S2-)|D| for half-integer spin syst ems. A better appreciation of the system is obtained by the realization that the ground state S = 10 spin is split into 21 sublevels (depicted in Figure 1-3) to give a double well poten tial in zero applied magnetic field. This splitting is a cons equence of the first term (2S Dz, axial ZFS term), in the Hamiltonian (eq 1-1) based on the giant spin model for molecular systems.15 'B y x zH O S g H ) S S E( S D H 4 2 2 2 (1-1)

PAGE 25

9 Because the value of the axial ZFS parameter D for a SMM is negative (-0.50 cm-1 for Mn12ac), the ms = sublevels lie lowest in energy while the ms = 0 sublevel lies highest (Figure 1-3). Conseque ntly, there is a potential ener gy barrier between the spinup (ms = -10) and spin-down (ms = +10) orientations of the magnetic moment. The energy of each sublevel is given as E (ms) = ms 2D, giving rise to a barrier whose magnitude is given by the difference in the energy between the ms = 10 and ms = 0 energy levels. Thus, to reverse the sp in of the molecule from along the z (spin up) to the + z (spin down) axis of the molecule, a potential energy barrier, U = S2|D| 50 cm-1 (as S = 10, D = -0.5 cm-1) for Mn12ac, must be overcome. For this reason, SMMs exhibit slow magnetization relaxation at low temperatures. Experimental evidence for this behavior is supported by the appearance of concomita nt frequency-dependent in-phase ( M') and outof-phase ( M'') signals in AC magnetic susceptibil ity measurements, as shown in Figure 1-4 (left). Frequency-dependent ac signals are necessary but not sufficient proof that an SMM has been obtained. The observance of hys teresis loops in magnetization versus DC field scans, with coercivities increasing w ith decreasing temperature and increasing scanrates, is an unambiguous confirmati on that an SMM has been obtained.9 Sometimes the hysteresis loops have steps (Figure 1-4 ri ght (bottom)), which are due to quantum tunneling, and each step corresponds to a sudde n increase in the magnetization relaxation rate. The QTM can be of different kinds (Figure 1-4 right (top)) and these will be discussed later. From a magnetic point of view, SMMs are preferred over classical nanoscale magnetic particles because they have a single, well defined ground state spin S which is a true quantum spin system, and highly ordere d assemblies of SMMs can be obtained in

PAGE 26

10 crystalline form. At the same time, from a s ynthetic chemists point of view, these SMMs can be easily synthesized at room temperatur e by solution methods and we can obtain a Figure 1-4. (Left) In-phase MT (top), and out-of-phase M (bottom), AC susceptibility signals for a Mn12 complex. (Right) Possible tu nneling mechanisms (top) for SMMs and (bottom) typical hystere sis loops with steps, for a Mn12 SMM. collection of truly monodisperse particles of nanoscale dimensions which have true solubility (rather than colloid formation) in organic solvents. Add itionally, the central core containing the metals is engulfed in a protective shell of orga nic groups which can be easily varied by ligand substitution, thus providing advantages for certain applications. Due to the strong need for SMMs with even larger S values and mo re negative D values, numerous synthetic strategies aimed at the improvement of these materials have been considered. Several new strategies and SMMs will be discussed in the ensuing chapters. -1 -0.5 0 0.5 1 -1-0.500.51 M/Ms 0H (T) 0.004 T/s 1.3 K 1.8 K 2 K 2.2 K 2.4 K 2.6 K 2.8 K 3 K 3.6 K M/Ms -1 -0.5 0 0.5 1 H0(T) 4 mT/s 1.3 K 1.8 K 2 K 2.2 K 2.4 K 2.6 K 2.8 K 3 K 3.6 K -1 0 0.5 1 -0.5

PAGE 27

11 Manganese carboxylate cluster chemistr y finds importance in bioinorganic chemistry as well, mainly because of the abil ity of Mn to exist in a range of oxidation states (-3 to +7), with th e most common ones being II, III and IV. Mixed-valency in Mn ions is found both in coordination compounds as well as Nature. Thus, oxidation state variability in Mn makes it well suited as the active site for redox reactions in a number of metalloproteins and enzymes.16 Redox enzymes containing Mn can be classified into different groups depending on their nucl earity. For example, mononuclear sites containing one manganese ion are found in manganese superoxide dismutases and dinuclear sites are found in manganese catalases.17 A tetranuclear manganese cluster, called the water oxidizing co mplex (WOC), comprising pr imarily oxide and carboxylate peripheral ligation, resides at the ac tive site of photosystem II (PS II).18 It is a complex aggregate of electron transf er proteins embedded in th e thylakoid membrane of chloroplasts in green plants, algae and cyanob acteria. As the name s uggests, it catalyses the light-driven oxidation of water to di oxygen. This reaction is responsible for generation of almost all the oxygen on this planet (eq. 1-2). In additi on to the synthesis of dioxygen, WOC also releases pr otons, which are used to cr eate a proton concentration gradient across the membrane that drives the synthesis of ATP by ATP-synthase. The electrons produced in the reaction ar e transferred through a series of ecarriers to photosystem I (PS I), where they are ev entually used for the fixation of CO2. 2 H2O 4 e-O2 + 4 H+ (1-2) It is believed that four Mn per PS II are essential for wate r oxidizing activity and they appear to be in close proximity to each other. The Mn complex is capable of cycling between four distinct oxi dation levels, labeled S0 through S4 in the pioneering work of

PAGE 28

12 Kok;19 the Sn states contain Mn in various combin ations of higher metal oxidation states (III-IV). The catalytic cycle thus involves the four electron oxidation of the Mn aggregate, and the latter can be thought of as a biologica l capacitor, storing not only charge but oxidizing equivalents, with discharge of capacito r occurring during S4 to S0 transition on oxidation of substrat e to dioxygen (Figure 1-5 (left)). Figure 1-5. (Left) The S state scheme as proposed by Kok for th e oxidation of water.27a (Right) Arrangement of 4 Mn and 1 Ca atoms (shown as balls) in the two latest crystal structures (see later) of the WOC of PS II.26 The S1 state (MnIII 2MnIV 2) is thermally most stable in the dark and is thus referred to as the dark-adapted state. However, an understanding of the mechanistic details regarding this catalytic process is greatly hinde red due to the absence of precise structural information of the enzyme and the inability to detect any key reaction intermediates. This said, recently, the enigmatic S4 state has been identified.20 The first crystal structure21a of dark adaptive active PS II membrane of the cyanobacterium Thermosynechococcus elongatus was reported at 3.8 in 2001 followed by another21b in 2003 at 3.7 These first generation crystal structures confirmed the location of the tetranuclear Mn cluster

PAGE 29

13 but the exact orientation and structure of the Mn complex and its intermediate ligand environment was not clear. Very r ecently another crystal structure22 appeared at 3.5 Figure 1-6. (Left) Crystal stru cture of the WOC at 3.0 (top)26 with the 4 Mn labeled 1-4 in pink, and the 3.5 Mn4Ca cubane-containing crys tal structure (bottom).22 (Right) Possible Mn4CaOx topologies suggested by biophysical studies.27a which assigns most of the amino acids in the protein, identifies 4M n and 1Ca ions and proposes the metal coordination number and geometry. In particular, the authors conclude that the WOC is a [Mn3CaO4] cubane with the fourth Mn (extrinsic Mn) attached to one of the cubane oxygen atom (Figure 1-6 (left (bottom)). This crystal structure by the Barber group, postulating a Mn3CaO4 cubane-like core for the WOC, is not accepted unanimously because it is believed that radiation damage to the crystals took

PAGE 30

14 place during X-ray data collection.23 However a very interesting development in this particular report, when compared to the earlie r structural elucidati ons of the WOC, was the detection of a Ca atom intimately associated with the tetranuclear Mn cluster. It had long been known that the WOC requires Ca2+ for activity (it acts as a cofactor),24 and calcium EXAFS (extended X-ray absorption fi ne structure) studi es of the WOC had revealed a MnCa separation of ~3.4 .25 The most recent and complete crystal structure of the WOC by the Zouni group26 at 3.0 concurs with the fact that the WOC is a pentanuclear heterometallic site and sugge sts that the Mn ions are arranged in a + 1 fashion (Figure 1-6 (left (top)), as proposed by the first two studies. The Mn4Ca structure (Figure 1-6 (left (top)) has Mn1 a nd Mn3 (one is III a nd another IV oxidation state), along with Mn2 (III oxi dation state; ligated by the ca rboxy-terminal carboxylate of Ala 344) and Ca form a trigonal pyramid, to which is attached the extrinsic fourth Mn4 (IV oxidation state; ligated by Asp 170). Vari ous biophysical studies have also suggested similar Mn4Ca cluster topologies (Figure 1-6 (right)).27 At the current resolution, smaller ligands such as C, H, O, H2O, N and Cl cannot be confiden tly located. However, both the studies at 3.0 and 3.5 have identified Ca as being part of the Mn complex using anomalous diffraction data. Thus, although th ere is still an am biguity about the Mn4Ca structure obtained from crystallography due to radiation damage during X-ray data collection,23 there is little doubt that th e WOC is a heterometallic [Mn4CaOx] cluster.22, 26 The availability of inorganic Ca/Mn complexes to act as synthetic models of the WOC would represent an important step forward in understanding the magnetic and spectroscopic properties of the native site and the mechanism of its function. Many groups have in the past applie d the Synthetic Analogue Approach28 to the WOC and a

PAGE 31

15 plethora of Mn4 complexes have been synthesized. However, new bioinorganic modelling approaches to the WOC involve synthesi zing heterometallic Mn/Ca complexes as opposed to homometallic Mn complexes. Theref ore, as part of our ongoing interest in obtaining synthetic models of the WOC and its various modified forms, we wanted to investigate mixed Ca/Mn chemistry extensively. Additionally, Mn clusters have been the primary source of SMMs, and it has become ev er-increasingly important to devise new synthetic routes towards clusters with inte resting magnetic properties and higher blocking temperatures. A totally new approach to wards SMMs is synthesizing mixed-metal complexes. This approach has the benefits of the spin as well as the anisotropy not canceling, thereby increasing the probabil ity of the resulting polynuclear complex displaying SMM behavior. Thus, the primary goal of the research featured in this dissertation is the development of new synthetic routes aimed at the preparation of novel heterometallic complexes incorporating Mn, whose relevance span the magnetic materials as well as the bioinorganic research areas. Chapters IIIV report the syntheses of Mn containing heterometallic complexes of lanthanides and actinides. The magnetic properties of these are described in detail. The subsequent chapters, V and VI are relevant to the bioinorganic modeling of the WOC of PS II, a nd they detail the relvance of mixed Sr/Mn and Ca/Mn complexes. The X-Ray Absorption Spectroscopy of these complexes and the comparisons thereof with the WOC are also reported in these chapters.

PAGE 32

16 CHAPTER 2 SINGLE-MOLECULE MAGNETS: A NOVEL FAMILY OF MnIII / CeIV COMPLEXES WITH A [Mn8Ce8O8]12+ CORE 2.1 Introduction One of the motivating themes in polynucle ar cluster chemistry research is the design of high-nuclearity manganese carboxy late clusters which can function as nanoscale magnetic materials. Since these specie s are molecular in nature, they fall in the nanoscale regime and since some of th em display superparamagnet-like slow magnetization relaxation these are termed as single-molecule magnets (SMMs). Thus, an SMM represents a molecular approach to nanomagnets.9 Such molecules thus behave as magnets below their blocking temperature (TB), exhibiting hysteresis in magnetization versus dc field scans. This behavior results from the combination of a large ground spin state (S) with a large and negative Ising (or easy-axis) type of magnetoanisotropy, as measured by the axial zero-field splitting parameter D. This leads to a significant barrier (U) to magnetization reversal its maximum value given by S2|D| or (S2 )|D| for integer and half-integer spin, respectively. 9, 29 However, in practice, quantum tunneling of the magnetization (QTM) through the barrier via higher lying MS levels of the spin S manifold results in the actual or effective barrier (Ueff) being less than U. The interest in SMMs for scientists in various disciplines is stimulated by not onl y their aesthetically pleasing structures but their ability to disp lay classical magnetic bistability as well as quantum properties.9, 29 The first SMM discovered was [Mn12O12(O2CMe)16(H2O)4],29, 9 (hereafter referred to as Mn12-acetate) which possesses an S = 10 ground state; together

PAGE 33

17 with the continually growing family of [Mn12O12(O2CR)16(H2O)4] (Mn12; R = various) molecules, these clusters are still the best and most thor oughly studied SMMs to date.30 Ever since, several types of SMMs have been discovered, most of them containing primarily MnIII ions.31 However, there have been only a few isostructural families of SMMs studied to date, most notably the Mn4 defect-dicubane,32 Mn4 cubanes,33 and Mn12 wheel complexes.34 As part of our continuing search for ne w synthetic routes to wards novel structural types which can function as SMMs, we have joined on-going efforts in mixed-metal cluster chemistry. Recently, there has been a spurt of research activity in the scientific community towards heterometallic systems which can behave as SMMs.35 We and our coworkers had contributed to this relativ ely nascent field with the successful characterization of the Mn8Ce,36 Mn11Dy4,37 Mn2Dy2,38 and Fe2Dy2 39 SMMs. Indeed, we had earlier reported the temp late synthesis of the Mn8Ce SMM which possessed an S = 16 ground state spin. Although the Ce ion (CeIV) in the complex was diamagnetic, it acted as a template around which 8 ferromagnetically coupled MnIII ions wrapped. Since the ground-state spin was then the largest for any Mn species we carried out a detailed investigation towards obtaining similar st ructural types with interesting magnetic properties. We herein report the successful s ynthesis, structure, magnetic characterization and reactivity of four Mn8Ce complexes and demonstrate th at they are new SMMs with spin-variability within the family. 2.2 Results and Discussion 2.2.1 Syntheses [Mn8CeO8(O2CMe)12(H2O)4]H2O (4) was originally obtained serendipitously from a solution of [Mn6CeO9(O2CMe)9(NO3)(H2O)2]40 in MeCN/Et2O that had been left

PAGE 34

18 undisturbed for some time. However, once the identity of 4 was know it became crucial to develop a rational syntheti c procedure towards obtaining 4 in high yield. One very attractive reaction strategy was to use the linear polymer {[Mn(OH)(O2CMe)2](MeCO2H)(H2O)}n,41 which might wrap around the oxophilic Ce4+ ion as Ce4+OH contacts develop. Thus, with a Ce:Mn ratio of 1:8 and OH deprotonation, this encircleme nt could in principle give 4, since complex 1, whose structure is shown below, provides all the required components needed to obtain 4. Thus, Mn OHMn O C O CH3O C O CH3O O OH C C CH3CH3O O O C O Mn OH CH3O C O CH3OH O O C C O O CH3CH3 the feasibility of the above-state d hypothesis was tested by reacting 1 with (NH4)2Ce(NO3)6 in MeCN as depicted in eq 2-1 below. (8/n)[Mn(OH)(O2CMe)2]n + Ce4+ + 4 H2O [Mn8CeO8(O2CMe)12(H2O)4] + 4 MeCO2H + 4 H+ (2-1) Indeed, the reaction resulted in the formation of 4 in 55% isolated yield. The magnetism studies of 4 revealed that it had an S = 16 ground state spin. However, the complex had considerable intermolecular interactions and a small D value. Thus, we decided to synthe size derivates of 4 with the objectives being two-fold: i) Block the intermolecular intera ctions primarily o ccurring via the bound H2O molecules by replacing the terminal ligation with other chelates. ii) Obtain derivatives of 4 with bulky carboxylates that would, besi des blocking the intermolecula r interactions, help in flattening the Mn8O8 loop to a more planar conf iguration thereby increasing the magnitude of the small D value which complex 4 possessed (primarily because the JT

PAGE 35

19 axes were perpendicular (vide infra)). Among the various MnIII sources which could be e xplored to obtain the 8MnIII compound, [Mn3O(O2CMe)6(py)3] (2MnIII, 1MnII) and(NBun 4)[Mn4O2(O2CPh)9(H2O)] (4MnIII) seemed particularly attractive, as they have been known to give higher nuclearity complexes.42, 43 Hence, the reaction of [Mn3O(O2CMe)6(py)3] with Ce4+ in MeCN in an approximately 3:1 ratio gave [Mn8CeO8(O2CMe)12(py)4]C4H2O2 (5) as depicted in eq 2-2 below. However, it should be noted that varying the Mn3:Ce ratio from 2:1 to 5:1 gave the same product, although the yield was optimized when the ratio of 8:3 was used as stated in eq 2-2. 8 [Mn3O(O2CMe)6(py)3] + 3 Ce4+ + 16 H2O 3 [Mn8CeO8(O2CMe)12(py)4] + 12 py + 12 MeCO2H + 32 H+ + 20 e(2-2) The synthetic strategy shown above pr oved successful in isolating a Mn8Ce complex in which the 4 terminal water molecules had been replaced by pyridine (py) molecules, thereby reducing intermolecular interactions. In order to flatten the Mn8O8 loop, synthesis of the be nzoate version of the Mn8Ce complex was sought. This task was achieved by reacting 3 (a tetranuclear MnIII complex) with (NH4)2Ce(NO3)6 in a 1:1 ratio. Varying the Mn4:Ce ratio to 1:2 also led to the isolation of the Mn8Ce benzoate complex {[Mn8CeO8(O2CPh)12(MeCN)4] [Mn8CeO8(O2CPh)12(dioxane)4]} (6). Complex 6 is isostructural with complexes 4 and 5, the difference being th at the unit cell has 2 Mn8Ce complexes; one having 4 MeCN as terminal ligands and the other with 4 1,4dioxane ligands providing terminal ligation. The mani festation of two complexes co-crystallizing indicates the complexity of this reaction with se veral species likely to be in equilibrium in the reaction solution. Thus, factors such as relative so lubility, lattice energies,

PAGE 36

20 thermodynamics, crystallization kinetics and ot hers undoubtedly determine the identity of the isolated product. The successful isolation of complex 6 encouraged us to seek the syntheses of Mn8Ce complexes with even bulkier carboxylate groups. For this purpose, we decided to employ diphenylacetic acid (Ph2CHCO2H). Such a bulky carboxylate would not only separate the molecules thereby reducing interm olecular interactions, but would also cause strain in the central Mn8O8 core thus flattening the loop. The synthetic technique used to obtain [Mn8CeO8(O2CCHPh2)12(H2O)4] (7) was our standard ligand substitution reaction,11a which has been successful in Mn12 chemistry. The ligand substitution reaction (eq 2-3) is an equilibrium th at must be driven to comple tion by (i) using a carboxylic acid with a much lower pKa than that of acetic acid (4.75) ; and/or (ii) using an excess of RCO2H; and/or (iii) removing the acetic acid as its toluene azeo trope. Hence, 20 equiv. of diphenylacetic acid were reacted with complex 4 and the free acetic acid removed under vacuum as its toluene azeotrope. Although, the reaction proceeded with 12 equiv. of the acid too, as summed up in eq 2-3; the extra aci d is generally needed to ensure complete [Mn8CeO8(O2CMe)12(H2O)4] + 12 Ph2CHCO2H [Mn8CeO8(O2CCHPh2)12(H2O)4] + 12 MeCO2H (2-3) carboxylate substitution.11 The presence of the extra inco ming acid group ensures (by Le Chateliers principle) that th e equilibrium of the reaction in eq 2-3 is broken and pushed forward and a pure product with complete lig and substitution is obtained. Also, the pKa of diphenylacetic acid (3.94) is lower than that of acetic acid (4.75) thereby facilitating the carboxylate substitution reac tion and isolation of complex 7.

PAGE 37

21 2.2.2 Description of Structures 2.2.2.1 X-ray crystal structu res of complexes 4 7 PovRay representations of the labeled crystal structures of 4 (Figure 2-1), 5 (Figure 2-2, left), 6 (Figure2-2, right), and 7 (Figure 2-3) are depicted in the indicated figures. The common [Mn8CeO8] core present in complexes 4-7 is shown in Figure 2-4 (left). The crystallographic data and structure refinement details for complexes 4 7 are collected in Table 2-1. Selected bond distan ces and angles for complexes 4 and 5 are listed in Table A-1. The bond distances and angles for complex 7 are listed in Table A-2. Complex 4H2O crystallizes in the tetragonal space group I 4 with crystallographic S4 symmetry. The cluster contains one CeIV and eight MnIII ions bridged by eight 3-O2and twelve CH3CO2 groups. The structure of 4 can be described as a non-planar saddlelike [MnIII 8( 3-O)8]8+ loop attached to a central CeIV ion via the triply bridging oxides of the loop. Peripheral ligati on around this central [Mn8CeO8]12+ loop is provided by 8 syn, syn, doublyand 4 triply-bridg ing acetate groups. Four H2O molecules (O6 and its symmetry counterparts in Figure 2-1) terminally ligate on four of the MnIII (Mn2 in Fig. 2-1) ions. The central Ce ion is octa-coordi nated, with the CeO bond lengths (2.29-2.37 ) being typical fo r eight-coordinate CeIV.44c All the Mn ions are hexa-coordinate with near-octahedral geometry and display Jahn-Teller (JT) elongation axes (vide infra), with the JT bonds being at least 0.1 0.2 longer than the other MnIII-O bonds, as expected for high-spin MnIII ions. Nevertheless, the metal oxidati on states were also verified by bond-valence sum calculations (BVS; see chap ter III for theory on BVS), and charge considerations.44 The complex contains four unbound water molecules as solvent of crystallization. There are st rong interand intramolecular hydrogen-bonding interactions involving the lattice and bound water molecules as well as O atoms from

PAGE 38

22 oxide and carboxylate ligands. Complex 5C4H8O2 crystallizes in the tetragonal space group P42/n with a crystallographic S4 axis. The cluster is isostructural with complex 4, with the difference being that terminal ligation on four of the MnIII ions is provided by pyridine (py) molecules rather than H2O, as is the case in complex 4. Additionally, there are three 1,4dioxane (C4H8O2) molecules as solvents of crystall ization. Thus inter/intra-molecular interactions which were present in 4 because of bound/unbound water have been nullified in complex 5. This occurs as a consequence of the absence of any H atoms attached to an electronegative atom; the py as well as th e dioxane do not have any hydrogen connected to N / O atoms. Thus, H-bonding does not occur in complex 5. Figure 2-1. PovRay representation at the 50% probability level of the X-ray crystal structure of 4. Color scheme: Mn green, Ce cyan, O red, C grey. Hydrogen atoms have been omitted for clarity. The complex has a four-fold inversion center.

PAGE 39

23 Complex 6C4H8O2MeOH crystallizes in the tetragonal space group P4n2 and the unit cell contains two Mn8Ce clusters; one with four MeCN molecules providing terminal ligation and the othe r with four 1,4-dioxane (C4H8O2) molecules providing terminal ligation. Hence, complex 6 is formulated as {[Mn8CeO8(O2CPh)12(MeCN)4] [Mn8CeO8(O2CPh)12(C4H8O2)4]}. Figure 2-2 (right) depicts the Mn8Ce benzoate subcluster with MeCN providing the terminal ligation. For each of the co-crystallizing Mn8Ce clusters of 6, peripheral ligation is provided by eight doublyand four triplybridging benzoate groups. The central [Mn8CeO8]12+ core found in 4 and 5, is also retained in complex 6. Figure 2-2. PovRay representation at the 50% probability level of the X-ray crystal structure of 5 (left) and 6 (right). Color scheme: Mn green, Ce cyan, N dark blue, O red, C grey. H atoms have been omitted for clarity. The complexes have a four-fold symmetry axis with an inversion center. Complex 7H2O2CH2Cl23MeCN crystallizes in th e lower symmetry monoclinic space group P21/n, and the asymmetric unit contains the entire Mn8Ce cluster. Complex 7 is formulated as [Mn8CeO8(O2CCHPh2)12(H2O)4] and thus is isostructural with complex 4, i.e., contains a central CeIV ion which is bound by 8 3-O2to 8 MnIII ions which

PAGE 40

24 together manifest themselves as a non-plan ar saddle-like loop. Terminal ligation is provided by 4 water molecules, as is the case with complex 4. However, peripheral ligation in 7 is provided by eight doubl yand four triply-bridging diphenylacetate groups (Ph2CHCO2 ) (see Fig. 2-3). Figure 2-3. PovRay representation at the 50% probability level of the X-ray crystal structure of 7. Color scheme: Mn green, Ce cyan, O red, C grey. H atoms have been omitted for clarity. The presence of the big, fat phenyl groups of the bulky carboxylate group in 7 causes a greater degree of separation betw een individual clusters. In fact, the Mn8Ce clusters of 7 are very far apart from each other as was evidenced from the packing diagram of 7. The fact that individual clusters are aloof and separa ted from neighbors might help in improving the magnetic properties of this complex (see later). However,

PAGE 41

25 Figure 2-4. (Top) Co mparison of the [Mn8CeO8]12+ core (left), with the [Mn12O12]16+ core (right). (Bottom) The common core of complexes 4, 5, 6 and 7 depicting the near perpendicular alignmen t of the Jahn-Teller (JT) pairs. Thick black bonds denote JT elongation axes. Color scheme: MnIII green, MnIV purple, Ce cyan, O red, C gray. there are some hydrogen bonding interac tions amongst the water molecules. 2.2.2.2 Structural Comparison of complexes 4 7 The structures of complexes 4-7 are overall very similar, differing only in the nature of one or two peripheral bridging ligands (vide supra). Hence, complex 4 is formulated as [Mn8CeO8(O2CMe)12(H2O)4] and a variation in the terminal ligation from H2O to py gives complex 5. The benzoate version of complex 4 (complex 6) has MeCN/dioxane providing terminal ligation. Finally, complex 7 is the diphenylacetate version of complex 4. One-common structural motif which is conserved through 4-7 is the [Mn8CeO8]12+ core. This core depicted in Figure 2-4 (left) consists of eight [MnO2Ce] rhombs which inter-connect with each other via the eight sh ared triply bridging oxides. Thus, within this description the central core looks like a space-shuttl e with eight flaps and the non-planar arrangement of these wi ngs give the saddle-like loop structure arrangement to the core. The eight CeIVO bonds are undoubtedly crucial in the formation of the [Mn8O8] loop which engulfs the central Ce ion. Thus, the Ce ion acts as a template around which forms the non-plan ar metal-oxo loop. The four Mn2 atoms (Figure 2-1) occupy the corners of an al most perfect tetrah edron (Mn2CeMn2' =

PAGE 42

26 109.43), whereas the four Mn1 ions form a severely distorted (flattened) tetrahedron (Mn1CeMn1' = 92.02). Within this description, the CeIV ion occupies the centre of both the tetrahedra. Interestingly, the [Mn8O8] loop found in these complexes is very similar to that in [Mn12O12(O2CMe)16(H2O)4] (Mn12-ac)11 which has a central [Mn4O4] cubane instead of a smaller Ce atom (Figure 2-4 (center)). Theref ore, both of them have a non-planar ring of 8 MnIII ions which are linked by eight 3-O2to the central unit. Besides the difference in the central unit, Mn12-ac has four more carboxylates pr oviding peripheral ligation than the Mn8Ce complexes. A much clearer view of the cores of both of these complexes is seen in Figure 2-4 whereby the greater folding in the Mn8Ce complex is accentuated. The eight CeO bonds which connect the Ce to the Mn atoms are indisputably crucial to the formation of these complexes and cau se a greater folding of the [Mn8O8] ring in the Mn8Ce complexes, than in Mn12. Thus, a loop is created in the Mn8Ce complexes rather than the non-planar ring found in Mn12 complexes. The relative alignment of the Jahn-Teller (JT) axes is very important with respect to the magnetic properties displayed by the co mplex; it determines the anisotropy in the molecule, or in other words the magnitude of the ZFS term D. As already stated earlier, some of the JT axes are nearly perpendicular. Figure 2-4 (rig ht) depicts the orientation of the JT axes which are shown as thick bl ack bonds. For example the Mn2O4Mn1 angle in Figure 2-4 (right) is 83.9. Indeed, all the eight JTs occur in sets of two with the Mn OMn angle in the range of 80-84 for all the Mn8Ce complexes. Within the four sets of perpendicular JTs, each set shares the doubl y bridging oxygen of the four triply bridging carboxylates as a common vertex. Thus, all th e eight JT axes originate from the doubly

PAGE 43

27Table 2-1. Crystallographic data for 4H2O, 53C4H8O2, 612C4H8O24MeOH, and 7H2O2CH2Cl2MeCN. 4 5 6 7 Formula C24H52O40Mn8Ce C56H80N4O38Mn8Ce C236H264O100Mn16Ce2 C176H161Cl4N3O40Mn8Ce fw, g/mol 1560.30 1996.87 5859.88 3679.52 Space group I 4 P42/n P4n2 P21/n a, 23.947(6) 12.7367(6) 22.5647(6) 19.4171(12) b, 23.947(6) 12.7367(6) 22.5647(6) 31.3568(19) c, 9.953(5) 23.485(2) 26.3631(13) 27.5856(17) 90 90 90 90 90 90 90 95.884(2) 90 90 90 90 V, 3 5708(4) 3809.8(4) 13423.2(8) 16707.2(18) Z 4 2 2 4 T, K 100(2) 173(2) 173(2) 173(2) Radiation, a 0.71073 0.71073 0.71073 0.71073 calc, g/cm3 1.816 1.697 1.784 1.463 mm-1 2.584 1.953 1.168 0.993 R1 b,c 0.0899 0.0497 0.0367 0.0891 wR2 d 0.2163 0.1165 0.1083 0.2136 a Graphite monochromator. b I > 2 (I). c R1 = 100(||Fo| |Fc||)/|Fo|. d wR2 = 100[[w(Fo 2 Fc 2)2]/ [w(Fo 2)2]]1/2, w = 1/[2(Fo 2) + [(ap)2 +bp], where p = [max (Fo 2, O) + 2Fc 2]/3.

PAGE 44

28 bridging oxygens of the four triply bridging carboxylates and terminate in four cases on the singly bridging oxygen of another triply bridging carboxyla te group. In the remaining four cases, the JTs terminate on the atoms (O/N) which provide terminal ligation. Therefore, considerable magnetoanisotropy is expected to get cancelled for these complexes, consequently resulting in a small D value, as was initially reported for complex 4.36 2.2.3 Magnetochemistry of Complexes 4, 5 and 7 2.2.3.1 DC studies Solid-state variable temperature magnetic susceptibility measurements were performed on vacuum-dried microc rystalline samples of complexes 4, 5 and 7, which were suspended in eicosane to prevent to rquing. The dc magnetic susceptibility ( M) data were collected in the 5.0-300 K range in a 0.1 T magnetic field and are plotted as MT vs T in Figure 2-5. For 4, the MT value of 39.36 cm3mol-1K at 300 K remains more or less constant until 70 K. Then, it steadily increases with decreasing temperature to reach 69.28 cm3mol-1K at 5.0 K indicating a la rge ground-state spin for 4. For 5, the MT value of 28.31 cm3mol-1K at 300 K remains stea dy till 70 K. Then, it starts decreasing with decreasing temperature to finally reach a value of 17.08 cm3mol-1K at 5.0 K indicating a small ground-state spin for 5. For 7, MT fractionally increases from 28.00 cm3mol-1K at 300 K to 32.47 cm3mol-1K at 70 K, and then decreases to a minimum value of 19.85 cm3mol-1K at 5.0 K indicating a rela tively small ground state for 7. The spin-only value of eight non-interacting MnIII ions is 24.00 cm3mol-1K assuming g = 2 (CeIV is diamagnetic, f 0). For 4 the MT value of 39.36 cm3mol-1K at 300 K is much higher than the spin-only value and MT increases with decreasing temperature suggesting ferromagnetic interactions w ithin the molecule and a la rge ground-state spin value.

PAGE 45

29 Conversely, for both 5 and 7, the MT value decreases with decreasing temperature, suggesting the presence of overall strong, predominantly antiferromagnetic exchange interactions within these mo lecules and consequently a sm all ground-state spin for them. Temperature (K) 050100150200250300 M T (cm 3 mol -1 K) 0 20 40 60 80 Complex 4 Complex 5 Complex 7 Figure 2-5. Plots of MT vs T for complexes 4, 5 and 7. M is the dc molar magnetic susceptibility measured in a 1.0 kG field. With eight MnIII centers in 4, 5 and 7, total spin values range from 0 to 16. However, due to the size and low symmetry of the molecules, a matrix diagonalization method to evaluate the various Mn pairwise exchange parameters (Jij) within the Mn8Ce molecules is not easy. Similarly, application of the equivalent opera tor approach based on the Kambe vector coupling method45 is not possible. Therefore, we focused only on identifying the ground state S as well as the ZFS term D values, as in any case these would dominate the low temper ature studies we performed (vide infra). Hence, magnetization (M) data were collected in the magnetic field and temperature ranges 0.1-7 T and 1.8-10 K in order to determine the spin ground states of complexes 4, 5 and 7. In

PAGE 46

30 the nomenclature, N is Avogadros number, B is the Bohr magneton, T is temperature and H is the applied magnetic field. The obtained data are plotted as M/N B (reduced magnetization) vs H/T in Figure 2-6 for complexes 4 (left) and 7 (right). For a system occupying only the ground state and experien cing no ZFS, the various isofield lines would be superimposed and M/N B would saturate at a value of gS. The nonsuperposition of the isofield lines in Figure 2-6 is indicative of th e presence of strong ZFS. H/T [kG/K] 101520253035 M/N B 22 24 26 28 30 32 34 3 T 4 T 5 T 6 T 7 T fitting H/T [kG/K] 012345 M/N B 0 2 4 6 8 0.1 T 0.2 T 0.3 T 0.5 T 0.6 T 0.7 T 0.8 T fitting Figure 2-6. Magnetization (M) vs field (H) and temperature (T) data, plotted as reduced magnetization (M/N B) vs H/T, for (left) complex 4 at applied fields of 3, 4, 5, 6 and 7 T and in the 1.8 10 K temper ature range, and (right) for complex 7 at applied fields of 0.10.8 T range and in the 1.8 10 K temperature range. The solid lines are the fit of the data; see the text for the fit parameters. The data were fit, using the program MAGNET,46 by diagonalization of the spin Hamiltonian matrix assuming only the ground state is populated, incorporating axial anisotropy (D z 2) and Zeeman terms, and employing a full powder average. Thus, the complexes are modeled for the magnetization fi t as a giant-spin with Ising-like anisotropy. The corresponding Hamiltonian is gi ven by eq 2-4, where D is the anisotropy = D z 2 + g B0 H (2-4)

PAGE 47

31 constant, B is the axial Bohr magneton, z is the easy-axis spin operator, g is the electronic g factor, 0 is the vacuum permeability, and H is the applied longitudinal field. The last term in eq 2-4 is the Zeeman ener gy associated with an applied magnetic field. The giant-spin model and the same MAGNE T program is used in the magnetization fits for complexes reported in the ensuing chap ters, and for the sake of brevity, the whole process will not be repeated ag ain. Instead, just the spin, D value and the isotropic g values will be mentioned. For 4, the fit (solid lines in Fig. 2-6 (left)) gave S = 16, D = 0.10 cm-1 and g = 1.98. Thus, ferromagnetic couplings within complex 4 aligns the 8 MnIII spins parallel (as has been observed in Mn12ac, vide supra), leading to the second highest spin ground state for a Mnx species yet reported. The largest spin yet observed for a Mn complex is the S = 51/2 reported for a Mn25 complex.47 The D value of -0.1 cm-1 is consistent with the complex having the JT ax es perpendicular as stated earlier, and g is < 2, as expected for Mn. When data collected at fields < 3.0 T were included, a satisfactory fit could not be obtained, which is un derstandable as the complex has weak intermolecular interactions, thus higher fields are needed to overcome those interactions. For complex 5, attempts were made to fit the magnetization data collected in the 0.1 7 T and 1.8-10 K temperature ranges. A satisfactory fit could be obtained only when data collected in the 0.1-2 T applied fiel d range were used. The fit of the data gave a ground state of S = 5, D = -0.30 cm-1 and g = 1.83. For complex 7, attempts were made to fit the magnetization data collected in th e 0.1 0.8 T range and 1.8-10 K temperature ranges. A best fit of the data (Figure 2-6 (right)) yielded a ground state of S = 6, D = 0.34 cm-1 and g = 1.89. The ground states obtained for complexes 5 and 7 are in agreement with the dc magnetic susceptibi lity data. However, there are low-lying

PAGE 48

32 excited-states (relative to kT) which complicate the fitting; this was also confirmed from the sloping nature of the inphase ac susceptibility data (vide infra). We have found that poor quality fits of the magnetization versus H and T plots are a common problem in manganese chemistry when the Mnx species is of high nuclearit y and there is thus a high density of spin states resulting from th e exchange interact ions amongst the many constituent Mn ions. Thus, low-lying excited states are populated, even at these relatively low temperatures, and/or the MS levels from nearby excited states with S greater than that of the ground state are being sufficiently stabiliz ed by the applied dc field that they thus approach or even cross the ground state leve ls; note that the fitting model assumes population of only ground state levels. Populatio n of the excited states will thus be difficult to avoid even at the lowest temper atures normally employed. However, the spin ground states obtained from the fittings we re confirmed by the more reliable ac susceptibility studies48 described later. The large ground state spin value obtained for complex 4 suggested that it may have a barrier to magnetization reversal. Also, the combination of the S and D values obtained for complexes 5 and 7 may enable them to display slow magnetization relaxation characteristic of SMMs. The S and D values obtained for complexes 4, 5, and 7 suggest an upper limit to the energy barrier (U) to magnetization reversal of U = S2|D| = 25.6 cm-1 for 4, 7.5 cm-1 for 5 and 12.2 cm-1 for 7 respectively. However, the effective barrier Ueff, might be a little bit smaller because of quantum tunneling through the barrier. Hence, ac susceptibility measurements were performed to investigate whether these Mn8Ce complexes functioned as single-molecule magnets.

PAGE 49

33 2.2.3.2 AC studies In an ac susceptibility experiment, a weak field (typically 1 5 G) oscillating at a particular frequency ( ) is applied to a sample to probe the dynamics of the magnetization (magnetic moment) relaxation. An out-of -phase ac susceptibility signal ( M ) is observed when the rate at which the magnetization of a molecule relaxes is close to the operating frequency of the ac field, and there is a corresponding decrease in the in-phase (MT) signal. At low enough temperature, where the thermal energy is lower than the barrier for relaxation, the magnetization of the molecule cannot relax fast enough to keep in-phase with the oscillating field. Therefore, the molecule will exhibit a frequency-dependent M signal indicative of slow ma gnetization relaxation. The increase in the frequencydependent, imaginary M signal is accompanied by a conc omitant decrease in the real MT signal. Frequency-dependent M signals are an importa nt indicator of SMMs. Alternating current magnetic susceptibil ity studies were performed on vacuumdried microcrystalline samples of 4, 5 and 7 in the temperature range 1.8-10 K with a zero dc field and a 3.5 G ac field oscillati ng at frequencies between 5-1000 Hz. The inphase (M ) component of the ac sus ceptibility was plotted as M T vs T. Similarly, the out-of-phase M component was plotted as M vs T. Indeed, frequency-dependent signals were seen in the in-phase as well as the out-of-phase ac susceptibility plots for all of the three above-mentioned complexes. Th e strength of the signals varied and the intensity in decreasing order was 4 >7 >5. Although a rise in th e out-of-phase signal was accompanied by a concomitant decrease in the in -phase signal, only tails, of peaks which lie below the operating minimum (1.8 K) of our SQUID magnetometer were seen. Nevertheless, the in-phase M T data proved useful for confirming the spin ground state

PAGE 50

34 obtained from magnetization fits. For example, the M T value of complex 4 is 63.17 cm3mol-1K at 10 K and increases steadily to 94.05 cm3mol-1K at 1.8 K and seems to keep on rising steeply to higher valu es below 1.8 K. Extrapolation of the data to 0 K indicates a M T of ~ 125 which is in ag reement with the ~133 cm3mol-1K value obtained by applying the formula M T = (g2/8)S(S+1) and using the g and S values obtained from the magnetization fits (vide supra). For complexes 5 and 7 the presence of low-lying excited spin states within close separation of the ground state was confirmed by the sloping nature of the MT vs T plots. The in-phase MT value drops sharply and reaches 15.39 cm3mol-1K for 5 and 18.15 cm3mol-1K for 7 at 1.8 K. These values indicate an S of ca. 5 or 6 for these two complexes which is satisf yingly consistent with the dc magnetization fit values. Since all three complexes display freque ncy dependent ac signals which are an indication of the superparamagnet-like slow magnetization relaxation of a SMM we decided to investigate them fu rther. Note however that thes e signals are necessary but not sufficient proof that an SMM has been obtai ned because intermolecu lar interactions and phonon bottlenecks can also lead to such signals.49 Thus, lower temperature (<1.8 K) studies were carried out to expl ore this possibility further. 2.2.3.3 Hysteresis studies below 1.8 K If complexes 4, 5 and 7 were indeed SMMs, they s hould exhibit hysteresis below their blocking temperatures, TB, in a magnetization versus dc field plot. To investigate this, magnetization vs. applied dc field data down to 0.04 K were collected on single crystals of 4H2O, 5C4H2O2 and 7H2O3MeCN2CH2Cl2 using a micro-SQUID apparatus.50 The observation of hysteresis loops in su ch studies represents the diagnostic

PAGE 51

35 property of a magnet, including SMMs and superparamagnets below their blocking temperature (TB). The observed magnetiza tion responses for complex 4 are shown in Figure 2-7 (left) at a fixed sweep rate of 0.004 T/s and at different temperatures. The hysteresis loops of complex 5 at field sweep rates of 0.280 T/s and 0.017 T/s and a constant temperature of 0.3 K are shown in Figures 2-7 (right). Fi nally, the magnetization responses for 7, at different temperatures and a fi xed field sweep rate (0.14 T/s) are shown in Figure 2-8 (left) and at a fixed temp erature (0.04 K) and varying dc field sweep rates in Figure 2-8 (right), respectively. In all cases, hysteresis loops are seen, whose -1 -0.5 0 0.5 1 -0.4-0.200.20.4 0.04 K 0.2 K 0.3 K 0.4 K 0.5 K 0.6 K M/Ms 0H (T) 0.004 T/s -1 -0.5 0 0.5 1 -1.2-0.8-0.400.40.81.2 0.280 T/s 0.017 T/s M/Ms 0H (T) 0.3 K Figure 2-7. Magnetization (M) vs applied magnetic field (H) hysteresis loops: (left) for 4 in the temperature range 0.04.6 K and at a 0.004 T/s sweep rate; (right) for 5 at 0.280 and 0.017 T/s sweep rates and at a fixed temperature of 0.3 K. M is normalized to its saturation value, Ms, for both plots. coercivities increase with increasing sweep rate and with decreasing temperature, as expected for the superparamagnet-like prope rties of a SMM. The data thus confirm complexes 4, 5 and 7 as new additions to the family of SMMs. For 4H2O in Figure 2-7 (left) hys teresis loops are evident below 0.6 K, with the dominating feature in the l oops being the two-step prof ile and its va riation with temperature. This two-step profile and its broadening with decreasing temperatures are characteristic of a weak intermolecular interaction between mo lecules, undoubtedly

PAGE 52

36 mediated by the hydrogen bonds and dipolar inter actions. Similar behavior has been seen in the Fe19 SMMs51 which also display intermolecular interactions. However, the abovementioned interaction in the Mn8Ce complex only perturbs th e SMM behavior; it is too weak to give a classical antiferromagnetically ordered network. An intermolecular exchange parameter (J) of only 0.0025 K and an interaction energy of 0.65 K can be calculated from the loops shown in Figure 2-7 (left). Thus, complex 4 behaves as an SMM and the low temperature at which it s hows magnetization hystere sis is clearly due to the small D value, which is consistent with some of the MnIII JT axes, the primary source of the molecular anisotropy, being nearly perpendicular. For 5C4H2O2 hysteresis loops at 0.3 K are shown in Figure 2-7 (righ t). Clearly the coerci vity increases with increasing scan rate, as expected for an SMM; the loops at 0.280 T/s are thicker than the loops at 0.017 T/s. However, there is very little coer civity at H = 0 and this occurs because of a fast tunnel transition which changes the magnetization direction very rapidly. Thus, the effective barrier to ma gnetization relaxation reduces and hence the hysteresis behavior is seen at lower temp eratures (below 0.3 K). Nevertheless, the hysteresis data are in agreement with the ac signals which were the weakest for complex 5 and the upper limit to magnetization reversal for 5 was also only 7.5 cm-1. For 7H2O3MeCN2CH2Cl2 in Figure 8, the dominating feature in the temperature-dependent (top) as well as th e sweep-rate-dependent (bottom) hysteresis loops is the large step (corre sponding to a large increase in magnetiza tion relaxation rate) at zero field due to quantum tunneling of the magnetization (QTM). Steps at other field positions are only poorly resolved, probably due to broadening effects from low-lying excited states and a distributi on of molecular environments (and thus a distribution of

PAGE 53

37 -1 -0.5 0 0.5 1 -1.2-0.8-0.400.40.81.2 0.04 K 0.1 K 0.2 K 0.3 K 0.4 K 0.5 K M/Ms 0H (T) 0.14 T/s -1 -0.5 0 0.5 1 -1.2-0.8-0.400.40.81.2 0.140 T/s 0.070 T/s 0.035 T/s 0.017 T/s 0.008 T/s M/Ms 0H (T) 0.04 K Figure 2-8. Magnetization (M) vs applied magnetic field (H) hysteresis loops for 7: (left) in the temperature range 0.04.5 K at a 0.14 T/s sweep rate; (right) in the 0.008.140 T/s sweep rate range at 0.04 K. M is normalized to its saturation value, Ms, for both plots. relaxation barriers) caused by disordered latti ce solvent molecules and ligand disorder. In addition, intermolecular interactions (bot h dipolar and exchange ) and population of excited states can result in step broadening.52 Complex 7 displays hysteresis below 0.5 K and the plots in Figure 2-8 show increasing co ercivities with decr easing temperature and increasing sweep rates, as expected for SMMs. There is very little coercivity at H = 0. However, one should remember that that the y-axis for the hysteresis loops of complexes 5 and 7 scan a range of -1.4 till +1.4 T and that of 4 depicts the range of -0.5 to +0.5 T 0H values. Unfortunately, the fast relaxation ra te in zero field that results in the large step at this position also prevents us fr om collecting magnetizati on vs time decay data with which to construct an arrhenius plot and determine the effective barrier to relaxation (Ueff). It must be stated though, that at a qualitative as well as to some extent a quantitative level, the various dc, ac and hysteresis data are all consistent. Thus, for complexes 4, 7 and 5 the upper limit to magnetization reversal was found out to be 25.6 cm-1 >12.2 cm-1 >7.5 cm-1. The strength of the freque ncy-dependent ac signals also followed that same descending order and fi nally the temperature below which they

PAGE 54

38 displayed hysteresis loops (0.6 K >0.5 K >0 .3 K) was also consistent with the earlier mentioned data. All these magnetic parameters are tabulated in Table 2-2 for comparison. Table 2-2. Comparison of the magnetic parameters of complexes 4, 5, and 7. Complex S D (cm-1) g U (cm-1) TB (blocking T)a (K) 4 16-0.10 1.98 25.6 0.6 5 7 5 6 -0.30 -0.34 1.83 1.89 7.5 12.2 0.3 0.5 a TB is the blocking temperature: Temperat ure below which hysteresis loops were observed. 2.3 Conclusions Convenient, high-yield syntheti c routes towards obtaining a family of isostructural Mn8Ce complexes with a common [Mn8CeO8]12+ core, have been developed. Peripheral ligation has been varied within this fam ily and new surrogates obtained with the end objective of improving the magnetic properties; by blocking intermolecular interactions and/or flattening the [Mn8O8] loop to a more planar configuration. Complex 4 with an S = 16 ground state spin posse ss the third largest S value reported for a Mnx species; the highest and second highest being the S = 51/2 for a Mn25 SMM,47 and the S = 22 for a high-symmetry Mn10 complex.97c Complexes 5 and 7 possess an S = 5 and an S = 6 ground state, respectively. Although spin variability is present in this family of complexes, each of them possess a sufficien t combination of spin and anisotropy to display superparamagnet-like slow magneti zation relaxation and thereby function as single-molecule magnets (SMMs). Hysteresis measurements confirm the addition of complexes 4, 5 and 7 to the growing family of SMMs. Thus, in this chapter it has been

PAGE 55

39 aptly demonstrated that a subtle change in the ligand environment has a huge impact on the resulting magnetic properties within this family of complexes. This work lucidly reassert s that synthetic manipulation around the metallic core by organic groups can help in systematically studying the magnetic properties of SMMs, as has been seen for Mn12 complexes.11 It reiterates the advantages of the molecular approach to nanomagnetism wher eby standard chemistry methods give a greater deal of control, when compared to the clas sical top-down nanoparticle approach. 2.4 Experimental 2.4.1 Syntheses All manipulations were performed under aerobic conditions us ing chemicals as received, unless otherwise stated. {[Mn(OH)(O2CMe)2](MeCO2H)(H2O)}n (1)53, [Mn3O(O2CPh)6(py)2(H2O)] (2),54 [Mn3O(O2CMe)6(py)3],54 and (NBun 4)[Mn4O2(O2CPh)9(H2O)] (3),55 were prepared as previously reported. [Mn8CeO8(O2CMe)12(H2O)4]H2O (4). Method A. To a slurry of 1 (2.00g, 7.46mmol) in MeCN (35ml) was added solid (NH4)2Ce(NO3)6 (0.51g, 0.93mmol) and left under magnetic stirrer for 8 h re sulting in the formation of a brownish precipitate and a reddish-brown solution, which were separated by filtration. To the filtrate was added diethylether (40ml) and left under magnetic stirring for 5 more min. This solution was filtered and the filtrate concentrated by evaporation to yield reddish brown crystalline material which was washed with acetone and diethylether. The crystalline material was identified as 4H2O and was obtained in 55% isolated yield. Anal. Calc (Found) for 4H2O: C24H52O40Mn8Ce: C, 18.47 (18.49); H, 3.36 (3.32) %. Selected IR data (KBr, cm-1): 3392(s, br), 1576(s), 1539(s), 1444(s), 1029(w), 680(s), 657(m), 619(m), 589(s, br), 551(m), 496(w), 432(w).

PAGE 56

40 Method B. In MeCN (20ml) was dissolved [Mn6CeO9(O2CMe)9(NO3)(H2O)2]40 (0.50g, 0.38mmol) and left under magnetic stirri ng for 20 min., filtered off and the filtrate layered with diethylether (40ml). After two weeks the resulting solution was filtered and the filtrate slowly concentrated by evapor ation to yield reddish brown crystals of 4 which were filtered and washed with acetone and di ethylether and dried in vacuo. The product was identified as 4 by IR. The yield was 5%. [Mn8CeO8(O2CMe)12(py)4]C4H2O2 (5). To a slurry of [Mn3O(O2CMe)6(py)3] (1.00g, 1.29mmol) in MeCN (50ml) was added solid (NH4)2Ce(NO3)6 (0.24g, 0.43mmol) and left under magnetic stirring for 30 min. The solution was filtered and the filtrate layered with 50 ml of 1,4-dioxane (C4H2O2). After a week, nice square crystals of 5 were obtained. These were washed with 1,4-dioxane, dried in vacuo and isolated in 60% yield. Anal. Calc (Found) for 5C4H2O2: C56H80N4O38Mn8Ce: C, 33.68 (33.75); H, 4.04 (4.15); N, 2.81 (2.54) %. Selected IR data (KBr, cm-1): 3429(s, br), 1576(s), 1540(s), 1444(s), 1119(w), 871(w), 679(m), 656(m), 619(m), 589(s, br), 551(m), 496(w), 430(m). [Mn8CeO8(O2CPh)12(MeCN)4][Mn8CeO8(O2CPh)12(dioxane)4]12C4H2O2MeO H (6). To a slurry of 3 (1.00g, 0.62mmol) in MeCN/MeOH (25ml/1ml) was added solid (NH4)2Ce(NO3)6 (0.34g, 0.62mmol) and left under magnetic stirring for 20 min. The solution was filtered and the filtrate layered with 25 ml of 1,4-dioxane (C4H2O2). After a week, large dark red crystals of 6 were obtained. These were washed with 1,4-dioxane, dried in vacuo and isolated in 45% yield. Anal. Calc (Found) for 612C4H2O24MeOH: C236H264O100Mn16Ce2: C, 48.37 (48.20); H, 4.54 (4.35) %. Selected IR data (KBr, cm-1): 3430(s, br), 1600(m), 1560(s), 1528(s), 1412(s), 1120(w), 872(w), 718(s), 683(m), 615(m), 573(s, br), 510(w), 422(w).

PAGE 57

41 [Mn8CeO8(O2CCHPh2)12(H2O)4]H2O3MeCN2CH2Cl2 (7). The carboxylate substitution reactions which have been very successful for Mn12 systems30 were employed to synthesize 7. Therefore, diphenylacetic acid (0 .54g, 2.5 mmol) was added to a slurry of 4 (0.20g, 0.12mmol) in MeCN (50ml) and stirred overnight. The resulting solution was concentrated by using a toluen e azeotrope to remove the free acetic acid. The process of removing the free acid under va cuum was repeated thrice and the resulting powder was re-dissolved in MeCN. Concentrat ion of this solution by evaporation gave nice black crystals which unfortunately were not suitable for X-ray diffraction. Hence, recrystallization was achieved by dissolving the crystals in CH2Cl2 and layering with heptanes. After a week, nice black crystals of 7H2O3MeCN2CH2Cl2 were obtained in 60% yield. These were maintained in the mother liquor for X-ray crystallography and other single-crystal studies, or collected by filtr ation, washed with heptanes and dried in vacuo. The synthesis could also be performed by using 5 instead of 4, as the starting material. The dried solid analyzed as 7H2O. Anal. Calc (Found) for 7H2O: C168H148O40Mn8Ce: C, 59.58 (59.45); H, 4.40 (4.25) %. Selected IR data (KBr, cm-1): 3429(s, br), 1590(m), 1550(s), 1527(m), 1494(w), 1404(s), 1032(w), 745(m), 697(s), 650(m), 580(s, br), 433(w). 2.4.2 X-ray Crystallography Data were collected on a Siemens SM ART PLATFORM equipped with a CCD area detector and a graphite monochromator utilizing Mo-K radiation ( = 0.71073 ). Suitable crystals of the complexes were attach ed to glass fibers us ing silicone grease and transferred to a goniostat where they were cooled to 100K for complex 4 and 173 K for complexes 5, 6, and 7 for data collection. An initial search of reciprocal space revealed a tetragonal cell for 4, 5 and 6, and a monoclinic cell for 7; the choice of space groups I 4,

PAGE 58

42 P42/n, P4n2 and P21/n, respectively, were confirmed by the subsequent solution and refinement of the structures. Cell parameters were refined using up to 8192 reflections. A full sphere of data (1850 fram es) was collected using the -scan method (0.3 frame width). The first 50 frames were re-measured at the end of data collection to monitor instrument and crystal stability (maxim um correction on I was < 1 %). Absorption corrections by integration were applied ba sed on measured indexed crystal faces. The structures were solved by direct methods in SHELXTL6,56a and refined on F2 using fullmatrix least squares. The non-H atoms we re treated anisotropically, whereas the hydrogen atoms were placed in calculated, id eal positions and refined as riding on their respective carbon atoms. The asymmetric unit of 4H2O consists of one-fourth of the Mn8Ce cluster lying on an inversion centre and one H2O molecule of crystallization. Both of these are located on a C4 rotation axis. A total of 342 parameters we re included in the structure refinement using 2958 reflections with I > 2 (I) to yield R1 and wR2 of 8.99 % and 21.63 %, respectively. For 5C4H2O2, the asymmetric unit consists of one-fourth of the Mn8Ce cluster (on a 4-fold rotation axis) and a half dioxane molecu le (located on a 2-fold rotation axis), and a dioxane molecule (located on a 4-fold center). A total of 258 parameters were included in the structure refinement on F2 using 23703 reflections with I > 2 (I) to yield R1 and wR2 of 4.97 % and 11.65 %, respectively. The asymmetric unit of 612C4H2O2MeOH consists of a dioxane Mn8Ce cluster, a acetonitrile Mn8Ce cluster, three disordered dioxane molecules and one methanol molecule of crystallization. All of these are on a four-fold rotation axis. The

PAGE 59

43 solvent molecules were disordered and coul d not be modeled properly, thus program SQUEEZE,56b a part of the PLATON56c package of crystallographic software, was used to calculate the solvent disord er area and remove its contri bution to the overall intensity data. A total of 646 parameters were in cluded in the structure refinement on F2 using 6920 reflections with I > 2 (I) to yield R1 and wR2 of 3.67% and 10.83%, respectively. The asymmetric unit of 7H2O3MeCN2CH2Cl2 consists of a Mn8Ce cluster, four water molecules, three acetonitrile molecule s and two dichloromethane molecules. Most of the phenyl rings display considerable di splacement motions but not large enough of a disorder to allow to successfully resolve them. Consequently, large thermal parameters are observed for their carbon atoms. Thus, th ey were refined with isotropic thermal parameters only. The cluster has four coordi nated water molecules; two on each side of its plane. The two water molecules on each side along with two solvent water molecules form a diamond shape as a result of H ydrogen bonding. The solvent water molecules were located inside cavities created by the phe nyl rings. The rest of the solvent molecules were disordered and could not be m odeled properly, thus program SQUEEZE,56b a part of the PLATON package of crystallographic soft ware, was used to calculate the solvent disorder area and remove its contribution to the overall intensity data. A total of 1010 parameters were included in the structure refinement on F2 using 74183 reflections with I > 2 (I) to yield R1 and wR2 of 8. 91% and 21.36%, respectively. Unit cell data and details of the structure refinements for 44H2O, 53C4H8O2, 612C4H8O2MeOH and 74H2O2CH2Cl2MeCN are listed in Table 2-1.

PAGE 60

44 CHAPTER 3 SINGLE-MOLECULE MAGNETS: SYNTHESES AND MAGNETIC CHARACTERIZATION OF A NOVEL FAMILY OF HETEROMETALLIC MANGANESE-LANTHANIDE COMPLEXES 3.1 Introduction The current burgeoning research in nanos ciences and nanotechnology is primarily governed by the ideology of taking materials to the extrem e limit of miniaturization. Once we venture beyond the macro-, mesoand the micro-scale dimensions and reach the nanoscale (and even subnano), interesti ng phenomenon are observed which cannot be described by the classical propert ies of matter. Therefore, ther e has been great interest in the scientific community in the study of nanoscale materials. One such interesting class of nanoscale magnetic material is single-molecule magnets (SMMs). SMMs represent a molecular approach to nanoscale magnetic part icles. Since these molecules are magnets, they show hysteresis like a ny classical magnet. However, their behavior is unlike macroscale magnets because they display quantum tunneling of the magnetization (QTM), a property seen mostly in the mi cro-, nano-scale and beyond. The magnetic behavior in SMMs is due to the intrinsic, in tramolecular properties of these species, and is the result of the combinati on of a large ground state spin (S) value and a significant magnetic anisotropy of the easy-axis (or Ising) type, as reflected in a negative value of the axial zero-field splitting (ZFS) parameter, D As a result, SMMs possess a significant barrier to reversal (relaxatio n) of their magnetization vect or, and the upper limit to this barrier (U) is given by S2|D| and (S2-)|D| for integer and half-integer S values, respectively. However, because of QTM the effective barrier Ueff is generally lower than

PAGE 61

45 U. The slow relaxation results in imaginar y (out-of-phase) magnetic susceptibility (M ) signals in ac studies, and in hysteresis loops in magnetization versus applied dc field sweeps.57 SMMs thus represent a molecular (or bottom-up) approach to nanoscale magnets, and thus differ significantly from classical (or top-down) nanoscale magnets of metals, metal alloys, metal oxides, etc. These differences include monodispersity, crystallinity, true solubility (rather than co lloid formation), and a sh ell of organic groups that prevents close contact of the magnetic cores with those of neighboring molecules, and which can be varied using standard chemi cal methods. Since the initial discovery of the revered [Mn12O12(O2CR)16(H2O)4] family of SMMs,7 a number of other structural types have been discovered, almost all of them being homometallic transition metal clusters and the majority of them being Mn clusters containing at least some MnIII ions.91, 102b, 78 However, recently we and others have been exploring heterometallic clusters as routes to novel SMMs. One, big advantage of heterometallic clusters is that the spins as well as the anisotropy will probably not can cel in the resulting polynuclear complex. Hence, some heterometal SMMs have been reported.58 Again, within this subclass a very attractive route to SMMs with higher blocki ng temperatures is the synthesis of mixed transition metal-lanthanide SMMs. In general, the presence of a lanthanide ions (i) large spin such as the S = 7/2 of Gd3+ and/or (ii) large anisotr opy as reflected in a large D value could serve to generate SMMs with prope rties significantly different from their homometallic predecessors. The field of transition metal / lanthanide SMMs has flourished in the last year or so, with reports of Cu2Tb2, Mn11Dy4, Mn6Dy6, Mn2Dy2, CuTb, Dy2Cu, and Fe2Dy2 SMMs amongst others.59 Additionally, single-molecule magnetism has also been reported in hom onuclear Ho, Dy, and Tb phthalocyanates.60

PAGE 62

46 With the field of single-molecule magnetism now firmly established, new synthetic methodologies and approaches need to be take n to obtain clusters wi th higher blocking temperatures, TB (the temperature below which the mo lecule functions as a magnet). The present work describes one such approach. We had earlier reported our initial br eakthrough in mixed Mn/Ln chemistry with the synthesis of the [Mn11Dy4O8(OH)6(OMe)2(O2CPh)16(NO3)5(H2O)3] SMM which displayed magnetization hyste resis and quantum tunneling.37 Additonally, in Chapter II we also discussed a family of Mn8Ce SMMs, albeit we do not consider them as mixed Mn/Ln SMMs as the Ln ion involved (CeIV) is diamagnetic. We herein describe the extension of the earlier work with the succe ssful syntheses, stru ctures and detailed magnetic characterization of the complete family of this Mn11Ln4 complexes (Ln = Nd, Gd, Dy, Tb, Ho, and Eu). Within this family, all but the Eu complex behave as SMMs; this was confirmed by the observation of fre quency dependent ac susceptibility signals, magnetization hysteresis and QTM for the aforementioned complexes. 3.2 Results and Discussion 3.2.1 Syntheses The synthesis of high-nuclearity Mn clusters incorporating MnIII and/or MnIV ions can be achieved either by oxidizing simple MnII salts61 or by a reductive aggregation of MnVII salts such as MnO4 .62 A yet different approach is the use of preformed metal clusters such as the oxo-centered trinuclear [Mn3O(O2CR)L3]0/+1 (R = Me, Ph, Et etc.; L = py, MeCN, H2O) complexes.54 However, various groups have generally been using flexible alkoxo chelates (tri podal ligands, diols, monols) in reactions employing the above-mentioned triangular unit. This stra tegy helps in imposing some geometric constraints on the resulti ng cluster; the carboxylates im pose little or no geometry

PAGE 63

47 restrictions and bridge multiple metals as oxide bonds are formed. Thus, clusters of various nuclearities have been obtained in which the alkoxide ar ms act as bridging ligands.61 In our research, we have recently b een investigating reactions employing a preformed Mn cluster and a heterometal atom with the end goal of obtaining heterometallic clusters which might be impor tant to diverse research areas such as magnetic materials and bioinorganic modeling. 37, 43 Hence, for the magnetic materials research we have been utilizing the triand tetra-nuclear [Mn3O(O2CPh)6(py)2(H2O)] (2)54 and (NBun 4)[Mn4O2(O2CPh)9(H2O)] (3)55 complexes as starting materials. The reaction typically is carried out in a mixed solvent system of MeOH/MeCN (1:20 v/v) in the presence of a Ln(NO3)3. The solvent mixture is necessary to ensure adequate solubility of the reagen ts, especially the Ln(NO3)3 salts. Indeed, reactions employing 2 or 3 and a Ln(NO3)3 in ratios varying from 1:1 to 1:2 gave the family of [Mn11Ln4]45+ compounds (Ln = Nd, Eu, Gd, Dy, and Ho; complexes = 8, 9, 10, 11, and 12, respectively). For complexes 8-12, the yield was optimized for ratios as stated in the Experimental Section (see later). The reaction proceeds in th e stoichiometric ratio as depicted, for example, in eq 3-1 for th e Mn/Gd reaction and eq 3-2 for the Mn/Dy reaction. In all reactions ex cept the one involving Mn and Dy, the MeOH acts as an inert solvent, inasmuch, it does not end up in the resu lting isolated complex either as MeOH or 11 [Mn3O(O2CPh)6(py)2(H2O)] + 12 Gd(NO3)3 + 35 H2O + 3 H+ + 24 e3 [Mn11Gd4O8(OH)8(O2CPh)16(NO3)5(H2O)3] + 18 PhCO2H + 22 py + 21 NO3 (3-1) as methoxides. However, the M11Dy4 complex has two bridging methoxides. Also, if more methanol was used in the reaction mixture, [Mn2Ln2O2(O2CPh)6(OMe)4(MeOH)4] complexes were obtained by methanolysis; they will be described later.59g These

PAGE 64

48 observations clearly indicate that the reactions are very complicated with an intricate mix of several species likely to be in equilibrium in the reaction solutions; this is perhaps 11 [Mn3O(O2CPh)6(py)2(H2O)] + 12 Dy(NO3)3 + 6 MeOH + 41 H2O + 3 H+ + 24 e3 [Mn11Dy4O8(OH)6(OMe)2(O2CPh)16(NO3)5(H2O)3] + 18 PhCO2H + 22 py + 21 NO3 (3-2) expected given the amount of H2O molecules, their deprotonation and further H removal to form hydroxide/oxide brides etc. Although the reactions are simple, one-pot and straightforward, factors such as relative solubility, lattice energies, crystallization kinetics and others undoubtedly determin e the identity of the isolat ed product. This said, the products isolated were obtained in high yiel ds in the 55-60% range and were definitely thermodynamically most stable. We wanted to further investigate the reaction system and wondered what would happen if we used other alcohols instead of methanol. Additionally, the Tb complex had not been synthesized with the above reacti on system and the magnetic properties of the Dy (which is magnetically similar to Tb) comple x, were most interesting, as stated in our earlier communication.59b Thus, the reaction of 2 with Tb(NO3)3 in a MeCN/PhCH2OH (20ml/5ml) solution was carried out from which complex 13 was successfully isolated. The reaction is depicted in eq 3-3 and the product contains three deprotonated benzyl alcohol (PhCH2OH) molecules, or in othe r words phenyl methoxides. 11 [Mn3O(O2CPh)6(py)2(H2O)] + 12 Tb(NO3)3 + 15 PhCH2OH + 23 H2O 3 [Mn11Tb4O8(OH)6(OCH2Ph)3(O2CPh)20(PhCH2OH)2(H2O)] + 6 PhCO2H + 22 py + 36 NO3 + 47 H+ + 11 e(3-3)

PAGE 65

49 The small volume of methanol / benzyl alcohol in these reactions is very crucial for the attainment of these clusters. In the abse nce of methanol either (i) brown precipitates of manganese oxides/hydroxides we re obtained, or (ii) it did not prove possible to isolate clean products from the reaction solutions. Al so, the relative acidity of methanol (pKa = 15.2) and that of benzyl alcohol (pKa = 15.0) are comparable and hence they provide very similar reaction conditions from which an alogous end products were obtained. When ethanol (pKa = 16.0) was used as the alcohol, no clean products could be isolated. A different strategy was used for the Mn2Ln2 reactions. In this case, an excess of MeOH (10ml) was added to the filtrate, after the reaction of the MnIII 4 complex 3 with with Yb(NO3)3 5H2O (or Y(NO3)3 6H2O) had been performed in a 1:2 molar ratio in MeCN/MeOH (20/5 ml). Thus, the methanolysis of a MnIII species in the presence of a Yb3+ or Y3+ source resulted in the isolation of two isostructural complexes in 10-25 % yields; [Mn2Yb2O2(O2CPh)6(OMe)4(MeOH)4] (14) or [Mn2Y2O2(O2CPh)6(OMe)4(MeOH)4] (15). Note that in both these Mn-containing complexes there are also MeOgroups from the MeOH solvent, in addition to terminal MeOH molecules. Again, there are likely other products from these complicated reactions in the colored filtrates, but we have not pursued a ny further separations. 3.2.2 Description of Structures 3.2.2.1 X-ray crystal structu re of complexes 11 and 13 PovRay representations of the crystal structures and labeled cores of complexes 11 (left) and 13 (right) are depicted in Figures 3-1 an d 3-2, respectively. Crystallographic data for 105MeCN, 11MeCN, and 13PhCH2OH are listed in Table 3-1. Complex 11MeCN crystallizes in the triclinic space group 1 Pwith the Mn11Dy4 molecule lying on an inversion center. Although 11 is heterometallic, it is homovalent as

PAGE 66

50 Figure 3-1. PovRay representation at the 50% probability level of the X-ray crystal structures of 11 (left) and 13 (right). Color scheme: Mn yellow, Dy green, Tb purple, O red, N blue, C grey. H atoms have been omitted for clarity. it contains eleven MnIII and four DyIII ions. The structure (Figure 3-1, left) consists of a [Mn11Dy4]45+ metallic unit held together by six 4-O2-, two 3-O2-, six 3-HO-, and two MeOions. Peripheral ligation is prov ided by twelve -, and four 3-bridging benzoate groups, five chelating NO3 -groups on the Dy ions, two H2O molecules on Mn4 and Mn4', and a water molecule on Dy2. The doubly bridging O atoms of the four 3-PhCO2groups in one case bridge Mn1 / Dy1', and in the other Mn3 / Mn6, and of course the two symmetry related counterparts. The core (Fig ure 3-2, left) consis ts of two distorted [DyMn3O2(OH)2] cubanes (Dy2, O17, Mn3, O16, Mn4, O21, Mn5, O5), each attached via a Mn/Dy pair (one above (Mn6) and anothe r (Dy1) below the plane) to a central, near linear and planar [Mn3O4] unit (Mn1, Mn2, Mn1'). The Mn/Dy pairs are also linked to the Mn3 linear unit via two triply bridging hydroxide s (O12, O12'). Within this description, the remaining four hydroxides (O21, O 16, O21', O16') lie in the two DyMn3 cubanes. Also, two doubly bridging methoxi des (O26, O26') link the DyIII in the cubanes (Dy2, Dy2'), to two MnIII ions (Mn6, Mn6'). The position of the methoxide oxygen (O26) is the crucial distinction between the cores of the remaining members of this family of

PAGE 67

51 complexes (see later). The metal oxidation states and the prot onation levels of O2-, HOand MeOions were established by bond-valence sum calculations,63 charge Figure 3-2. PovRay representation at the 50% probability level of the [Mn11Dy4O8(OH)6(OMe)2]21+ core of 11 (left) and the [Mn11Tb4O8(OH)6(OCH2Ph)2]21+ core of 13 (right). Color scheme: Mn yellow, Dy green, Tb purple, O red, N blue, C grey. H atoms have been omitted for clarity. considerations, inspection of metric pa rameters and the identification of MnIII Jahn-Teller (JT) elongation axes on all Mn atoms. All the Mn ions are hexa-coordinated, and in near octahedral geometry. The Dy ions are nine -coordinate, and the D yO bonds lie in the range 2.303.482 comparable with values reported in the literature.59c,d,f Complex 13PhCH2OH crystallizes in the triclinic space group 1 P and comprises a centrosymmetric MnIII 11TbIII 4 cluster (Figure 3-1, right). The overall structure of 13 is similar to complex 11, the difference being that the Mn11Tb4 complex possesses four extra bridging benzoate groups and two lesser terminal H2O molecules. Instead of two H2O molecules, it now has tw o benzyl alcohols (PhCH2OH), in addition to having three

PAGE 68

Table 3-1. Crystallographic data for 1015MeCN, 1115MeCN and 13PhCH2OH. 10 11 13 Formula C140H135O66N20Mn11Gd4 C142H139O66N20Mn11Dy4 C196H169O63Mn11Tb4 fw, g/mol 4430.07 4436.07 4772.53 Space group P 1 P 1 P 1 a, 16.5278(15) 16.5705(11) 16.6598(19) b, 17.4169(16) 17.2926(11) 18.647(2) c, 18.5303(17) 18.6213(12) 19.859(2) 65.238(2) 64.9700(10) 72.889(2) 67.126(2) 67.2950(10) 66.854(2) 65.125(2) 65.6110(10) 70.608(2) V, 3 4245.6(7) 4256.6(5) 5254.5(2) Z 1 1 1 T, K 173(2) 173(2) 173(2) Radiation, a 0.71073 0.71073 0.71073 calc, g/cm3 1.750 1.731 1.520 mm-1 2.420 2.610 2.042 R1 b,c 0.0656 0.0425 0.0773 wR2 d 0.1622 0.1006 0.1942 a Graphite monochromator. b I > 2 ( I ). c R 1 = 100(|| Fo| | Fc||)/| Fo|. d wR 2 = 100[[ w ( Fo 2 Fc 2)2]/ [ w ( Fo 2)2]]1/2, w = 1/[2( Fo 2) + [( ap )2 + bp ], where p = [max ( Fo 2, O) + 2 Fc 2]/3. 52

PAGE 69

53 bridging phenyl methoxides (PhCH2O-). Thus, 13 is formulated as [Mn11Tb4O8(OH)6(OCH2Ph)3(O2CPh)20(PhCH2OH)2(H2O)], the [Mn11Tb4O8(OH)6(OCH2Ph)2]21+ core of which is shown in Figure 3-2 (right). Amongst the benzoate groups sixteen are bridging as was in complex 11. Two of the remaining four are monodentate on Tb2 and T b2', whereas the remaining two are 2 terminally chelating on Tb1 and Tb1'. Th e two benzyl alcohols provi de terminal ligation on Mn4 and Mn4'. The water molecule and a phenyl me thoxide reside on Tb2. The remaining two phenyl methoxides (O20, O20' in Figure 3-2, right) bridge two Mn/Tb pairs (Mn6, Tb2 and Mn6', Tb2'), as was the case with the methoxides in complex 11. The metaloxo/hydroxo core for the Mn/Tb complex is the same as was for complex 11, with the oxides and hydroxides occurring in the same positions (Figure 3-2). The only difference (Figure 3-2) is that O26 is the oxygen of a methoxide in 11, and the corresponding O20 in 13 comes from a phenyl methoxide. All the elev en Mn ions are hexa-coordinated, and in near octahedral geometry. Tb1 is nine-coordi nate, and Tb2 is eight -coordinate. The TbO bonds lie in the range 2.374.492 compar able with values reported in the literature.59a,e,g 3.2.2.2 Structural Comparis on of Complexes 8-13 The structures of complexes 8-13 are overall very similar to each other, differing slightly in the nature of a few bridging ligands in the periphery. Indeed, complexes 8, 9, and 10 can be formulated as [Mn11Ln4O8(OH)8(O2CPh)16(NO3)5(H2O)3], with Ln = Nd, Eu, and Gd respectively. The common [Mn11Ln4O8(OH)8]21+ core to these complexes is shown for the Ln = Gd complex (10) in Figure 3-3, where O14 is a hydroxide connecting Gd2 and Mn5. The difference between the cores of these complexes and that of 11 and 13 is that the Dy complex contains a methoxide (O26) in the position of O14, and the Tb

PAGE 70

54 complex contains a phenyl methoxide (O20) in the corresponding positions. Peripheral ligation around these cores is the same for the Dy and Gd complexes, whereas the Tb Figure 3-3. PovRay representation at the 50% probability level of the [Mn11Gd4O8(OH)8]21+ core of complex 10. Color scheme: Mn yellow, Gd cyan, O red, C grey. complex differs slightly as already de scribed earlier. The holmium complex, [Mn11Ho4O8(OH)8(O2CPh)18(NO3)3(H2O)7] (12) differs considerably in the peripheral ligands when compared to other members of this family of complexes. The core for 12 is the same as the one depicted for 10 in Figure 3-3. However, peripheral ligation is now provided by eighteen bridging benz oate groups with sixteen of them bridging as was for the Dy complex. The additional two benzoa te groups bridge monodentate on the two Ho ions in the two cubanes. The other two Ho ions have three chelating nitrate groups on them. Additionally, there are five water molecu les on the Ho ions and two on the Mn ions (for example Mn6 in Figure 3-3), providing te rminal ligation. Thus, although the metallic core of all the complexes is very similar, they differ slightly in their periphery. This slight variation in the structures may indeed consequently have an effect on the magnetic properties of these clusters. Hence we decide d to investigate the single-molecule magnet properties of these complexes, which are described later.

PAGE 71

55 3.2.2.3 Structural descriptions of complexes 14 and 15 PovRay representations of th e centrosymmetric crystal structures (top) and labeled cores (bottom), of complexes 14 (left) and 15 (right) are depicted in Figure 3-4. Selected interatomic distances and angles are liste d in Table 3-4. The two complexes both crystallize in the monoclinic space group P 21/c with the molecules lying on an inversion centre; the two complexes are isostructural. The cores of the complexes possess a defectdicubane structure (two fu sed cubanes sharing a face, and each missing an opposite vertex; Fig. 3-4), with two MnIV atoms at the central positions and either two 2 YbIII (complex 14) or two YIII (complex 15) atoms at the end posi tions. The fully-labelled cores of 14 and 15 are provided in Fig. 3-4 (bot tom), which emphasize the near superimposibility of the mixed 3d/4d complex 15 with the mixed 3d/4f complex 14. Peripheral ligation about the cores is provided by four syn, syn bridging benzoate groups bridging each Y/Mn or Yb/Mn edge of the rhombus, two monodentate benzoate groups, one on each of the Y/Yb atoms, and four te rminal MeOH molecules, two on each of the Y/Yb ions. Within this description, two 3-O2atoms (O8, O8a) cap each triangular subunit, and each of the four edges of the M4 rhombus in 14/15 is bridged by a MeOion (O7, O7a, O11, O11a). As a result, the YbIII and YIII atoms of 14 and 15, respectively, are eight-coordinate, and the MnIV atoms are six-coordinate. Th ere are intramolecular OHH hydrogen bonds between the unbound O atom (O1) and the terminal MeOH ligand (O1O10 = 2.577). In addition, ther e are intermolecular OHH hydrogen bonds between the bound MeOH (O9) and intersti tial MeOH (O12) molecules (O9O12 = 2.656 ), and between this interstitial MeOH and the unbound O atom (O1) of a neighboring cluster (O12O1 = 2.687 ). Thus, the hydrogen-bonding through the

PAGE 72

56 interstitial MeOH molecules links adjacent metal clusters in the crystal, which also likely provides a pathway for superexchang e interactions between molecules ( vide infra ). Figure 3-4. (Top) PovRay representations of the crystal structures of complexes 14 (left) and 15 (right). (Bottom) Comparison of the cores of complexes 14 (left) and 15 (right), emphasizing their near superimposibility. Color scheme: Mn blue, Yb orange, Y pink, O red, C grey. It should be noted that mixed Mn2Ln2 complexes with the same kind of defectdicubane core have previously been reported for Ln = Dy, Gd and Tb.59d, 64 However, in all of these previous cases, the Mn atoms were in the MnIII oxidation state, so 14 and 15 are the first to instead contain MnIV. The MnIV oxidation level is suggested by overall charge considerations and inspection of metric parameters; in particular, Mn O bond distances all lie in the range 1.84.97 as expected for MnIV, and thus do not show the Jahn-Teller axial distortion expected for MnIII in near octahedral geometry. The Y O and Yb O bond distances are very similar, lying in the 2.30.42 and 2.27.38 ranges

PAGE 73

57 (Table 3-4), respectively, consis tent with eight-coordinate YIII/YbIII centers. The MnIV oxidation states and the pr otonation levels of the O2, MeO and MeOH were confirmed by bond valence sum (BVS) calculations, shown in Tables 3-2 and 3-3. The BVS values Table 3-2. Bond Valence Sums (BVS)a for the Mn atoms of complexes 15 (Y) and 14 (Yb). Atom MnII MnIII MnIV Assignment Mn1 (15) 4.132 3.779 3.967 MnIV Mn1 (14) 4.124 3.772 3.959 MnIV a The underlined value in bold is the one cl osest to the charge for which it was calculated. The oxidation state of a particular metal is the nearest whole number to that value. Table 3-3. Bond Valence Sums (BVS)a for the O atoms of complexes 15 (Y) and 14 (Yb). Atom BVS (15)BVS (14)Assignment O7 1.972 1.972 MeOO8 1.898 1.859 O2O9 1.210 1.198 MeOH O10 1.339 1.342 MeOH O11 2.038 2.016 MeOa The oxygen atoms is O2-, MeOif BVS 2; MeOH if BVS 1; and H2O if BVS 0. for the Mn atoms are clearly ~ 4, confirming the MnIV oxidation level. Values of ~ 2 are expected for O atoms in the O-II oxidation level and that ha ve no attached atoms that cannot be seen in X-ray crys tallography (i.e. H atoms). This confirms that O7, O8 and O11 are MeO-, O2-, and MeO-, respectively. In contrast, if there is a H atom which is not visible and its contribution to the BVS of that O atom is therefore not included, a lower BVS value is expected, typically 1 1.5 (dep ending on the degree of its participation in hydrogen-bonding); this is clearly the case for O9 and O10, which are therefore confirmed as MeOH groups.

PAGE 74

58 Table 3-4. Selected bond distances () and angles () for complexes 14 and 15. Complex 15 (Mn2Y2) Complex 14 (Mn2Yb2) Y1 O7 2.365(2) Y1 O8 2.356(2) Y1 O9 2.416(2) Y1 O10 2.351(2) Y1 O11 2.299(2) Mn1 O8 1.859(2) Mn1 O11 1.913(2) Y1 Mn1 3.3168(7) Mn1 Mn1a 2.7819(9) Y1-O11-Mn1 102.08(9) Mn1-O8-Mn1a 97.52(9) Y1-O8-Mn1 103.18(9) Yb1 O7 2.336(2) Yb1 O8 2.331(2) Yb1 O9 2.383(3) Yb1 O10 2.323(3) Yb1 O11 2.269(2) Mn1 O8 1.866(2) Mn1 O11 1.912(3) Yb1 Mn1 3.2886(6) Mn1 Mn1a 2.7769(11) Yb1-O11-Mn1 101.79(11) Mn1-O8-Mn1a 97.05(11) Yb1-O8-Mn1 102.61(10) The occurrence of MnIV in 14 and 15 is noteworthy give n that the reaction employed a MnIII 4 starting material and either YIII or YbIII, neither of which are good oxidizing agents. This indicates either the participation of atmospheric O2 gas as the oxidizing agent or the disproportionation of MnIII to MnIV and MnII. We favor the latter possibility given the low yields of thes e complexes, but we have not sought MnII species in the filtrates to confirm the same. 3.2.3 Magnetochemistry of Complexes 8-13, and 15 3.2.3.1 DC studies of complexes 9, 10, 11, and 13 Solid-state variable temperature magnetic susceptibility measurements were performed on vacuum-dried microc rystalline samples of complexes 9, 10, 11, and 13, each suspended in eicosane to prevent to rquing. The dc magnetic susceptibility ( M) data were collected in the 5.0-300 K range in a 0.1 T magnetic field and are plotted as MT vs T in Figure 3-5 for the aforementioned complexes. For 9 the MT value of 26.33 cm3mol1K at 300 K is much lesser than the expected value for 11 Mn3+ ( S = 2, L = 2, 5D0; g = 2) and 4 Eu3+ ( S = 3, L = 3, 7F0) non-interacting ions of 33.00 cm3mol-1K, consistent with antiferromagnetic exchange interactions within 9. The MT value decreases slightly with

PAGE 75

59 Temperature (K) 050100150200250300 M T (cm3mol-1K) 0 20 40 60 80 Complex 9 (Eu) Complex 10 (Gd) Complex 11 (Dy) Complex 13 (Tb) Figure 3-5. MT vs T plots for complexes 9 ( ), 10 ( ), 11 ( ), and 13 ( ). decreasing temperature to reach 13.63 cm3mol-1K at 25 K. Thereafter it falls rapidly to a minimum of 9.46 cm3mol-1K at 5.0 K, indicating a sma ll ground state spin for complex 9. For 10, the MT value of 51.43 cm3mol-1K at 300 K is less than the expected value for 11 Mn3+ ( S = 2, MT = 33 cm3mol-1K) and 4 Gd3+ ( S = 7/2, L = 0, 8S7/2) non-interacting ions of 64.52 cm3mol-1K. The MT value decreases slightly w ith decreasing temperature to reach 43.38 cm3mol-1K at 15 K, consistent with domi nant antiferromagnetic exchange interactions within complex 10. Below 15 K, the MT value decreases steeply to reach 39.03 cm3mol-1K at 5.0 K indicating a ground state spin for 10, which is definitely larger than that of 9. For complexes 11 and 13, the MT values at all temperatures are higher than the corresponding ones for 9 and 10 (Figure 3-5). For 11, the MT value of 74.34 cm3mol-1K at 300 K is less than the expected value for 11 Mn3+ ( S = 2, g = 2) and 4 Dy3+ ( S = 5/2, L = 5, 6H15/2) non-interacting ions of 89.68 cm3mol-1K, and the MT decrease steadily with decreasing temperature to reach 62.17 cm3mol-1K at 40 K; the behavi or being consistent

PAGE 76

60 with antiferromagnetic exchange interactions within 11. After 40 K, the MT drops slightly to 57.78 cm3mol-1K at 10 K, and then remains more or less constant with decreasing temperature to finally reach 57.09 cm3mol-1K at 5.0 K, indicating a relatively large ground-state spin for complex 11. For complex 13, the MT value of 64.35 cm3mol1K at 300 K is less, as anticipated, th an the expected value for eleven Mn3+ ( S = 2) and four Tb3+ ( S = 3, L = 3, 7F6) non-interacting free-ions of 80.28 cm3mol-1K. The magnitude of MT decreases with decreasing temperatures to reach a value of 53.14 cm3mol-1K at 40 K. Below 40 K, the MT drops sharply to reach 46.66 cm3mol-1K at 10 K after which it levels off at a value of 45.04 cm3mol-1K at 5.0 K, indicating a la rge ground-state spin for 13, which might be comparable to that of 11. The molar dc magnetic susceptibility valu es of the above-mentioned complexes are indicative of the contribution of the lanthanide ion to th e net experimental magnetic susceptibility observed and consequently the resulting magnetic properties of these complexes ( vide infra ). This is hypothesized because all of these complexes have eleven MnIII centers, in addition to the Ln ions, and the free-ion MT value (in cm3mol-1K) of individual Ln3+ ions in decreasing order is Dy (14. 17) > Tb (11.82) > Gd (7.88) > Eu (0.00). Accordingly, the magnitude of the MT with decreasing temperatures and even at 300 K, follows the same descending pattern, i.e. 11 > 13 > 10 > 9, as that of the free-ion MT values of the constituent Ln3+ ions of these complexes. Thus, both the data are satisfyingly consistent, and the overall nature of the plots for all th e complexes is very similar as can be seen in Figure 3-5. Howeve r, due to the size and low symmetry of the molecules, a matrix diagonalization method to evaluate the various metal pairwise exchange parameters (Jij) within the Mn11Ln4 molecules is not easy. Similarly,

PAGE 77

61 application of the equivalent operator a pproach based on the Kambe vector coupling method45 is not possible. Additionally, due to the presence of strong spin-orbit coupling effects (arising from the Ln3+ component) in these lantha nide containing mixed-metal complexes, the fitting of the magnetization data to obtain the S and the axial zero-field splitting (ZFS) term D is far from straightforward. The main problem in these procedures is the assumption of an is otropic, single electronic g factor; all the systems are two g (one for Mn3+ and another for Ln3+). However, there is some reprieve provided by the Mn11Gd4 system which has Gd3+ ( S = 7/2, L = 0, g 2) and therefore the resulting absence of spin-orbit coupling effects from the Ln3+ component. Thus, we focused on identifying the ground state S value for complex 10, for which magnetization ( M ) data were collected in the magnetic field and te mperature ranges 0.1-0.5 T and 1.8-10 K; the data are plotted as reduced magnetization ( M/N B) versus H/T in Figure 3-6. The data were fit, using the program MAGNET,46 by diagonalization of the spin Hamiltonian matrix assuming only the ground state is popu lated, incorporating axial anisotropy ( D z 2) and Zeeman terms, and employing a full powder average, as has already been discussed in Chapter II. The fit parameters were S = 9, g = 1.86 ( 0.00) and D = .06 ( 0.00) cm1. The D value of .06 cm-1 is consistent with the complex having four isotropic Gd3+ ions, and therefore an attenuation in the effective magnetoanisotropy component; it has already been stated earlier that the Ln3+ ions play a major role in the magnetic properties displayed by these complexes. Additionally g < 2, as can be expected for a system comprising eleven MnIII ( g < 2) and four GdIII ( g = 2) ions. Also, the expected MT for an S = 9 and g = 1.86 system of 38.92 cm3mol-1K can be compared with the 5 K dc value of 39.03 cm3mol-1K, which are in good agreement.

PAGE 78

62 H/T (kG/K) 0.00.51.01.52.02.53.0 M/N 0 2 4 6 8 10 12 14 0.1 T 0.2 T 0.3 T 0.4 T 0.5 T fitting Figure 3-6. Magnetization (M) vs field (H) and temperature (T ) data, plotted as reduced magnetization (M/N ) vs H/T, for complex 10 at applied fields of 0.1, 0.2, 0.3, 0.4 and 0.5 T and in the 1.8 10 K temperature range. The solid lines are the fit of the data; see the text for the fit parameters. Although 10 has a small D value of .06 cm-1, the large ground st ate spin value of S = 9, which is comparable to the S = 10 of Mn12 SMMs,11 suggested that the barrier to magnetization relaxation might be large enough for the complex to function as an SMM. The S and D values obtained for complex 10 suggest an upper lim it to the potential energy barrier ( U ) to magnetization reversal of U = S2| D | = 4.86 cm-1 = 7.00 K,9 although the actual, effective barrier ( Ueff) was anticipated to be less than this due to quantum tunneling of the magnetization (QTM). Add itionally and more importantly, the ground state spin of the Dy and Tb complexes can be estimated to lie in the S = 10 1 region, based on comparative Dc magnetic susceptibility data ( vide supra ). Thus, ac susceptibility measurements ( see later ) were performed on complexes 9, 10, 11, and 13 to investigate whether they functione d as single-molecule magnets.

PAGE 79

63 3.2.3.2 DC studies of [Mn2Y2O2(O2CPh)6(OMe)4(MeOH)4] (15) Solid-state variable temperature dc magnetic susceptibility measurements were performed on a vacuum-dried microcrystalline sample of complex 15 suspended in eicosane to prevent torquing. Th e dc magnetic susceptibility ( M) data were collected in the 5.0-300 K range in a 0.1 T ma gnetic field and are plotted as MT vs T in Figure 3-7 for 15. Complex 15 has a MT value that steadily decreases almost linearly from 3.13 cm3mol-1K at 300 K to 1.52 cm3mol-1K at 50 K, and then more rapidly decreases to 0.49 cm3mol-1K at 5.0 K (Fig. 3-7). The value at 300 K is less than the expected spin-only ( g = 2) value for a complex consisting of two non-interacting MnIV ions ( S = 3/2; the 2 YIII ions are diamagnetic) of 3.75 cm3mol-1K, indicating an antiferromagnetic exchange 050100150200250300 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Dc data fitting M T (cm3mol-1K)Temperature (K) Figure 3-7. Plot of MT (solid circles, ) vs T for complex 15. The solid line in the MT vs T plot is the fit of the data; see the text for the fit parameters. interaction between the MnIV ions and a resultant S = 0 spin ground state. Attempts to fit the MT vs T data using the isotropic Heisenbe rg-Dirac-van Vleck Hamiltonian described by H = -2 J 1 2 where S1 = S2 = 3/2 and J is the magnetic exchange interaction, gave poor fits that did not reproduce the data over the whole temperature range, particularly the higher T data. However, it did suggest that the J value is relatively weak, in the J = -

PAGE 80

64 10 to -15 cm-1 range, which in fact is consistent with the relatively acute Mn4+-O2--Mn4+ (Mn1-O8-Mn1a) angles of 97.52. The best fit of the data shown as a solid line in Figure 3-7, gave J = -13.5 cm-1, g = 1.88, and paramagnetic impurity, p = 0.10. Although the g value is as expected for MnIV (g < 2), the paramagnetic impurity is very high; approximately 10%. We had extended the work to Y instead of Ln, as it has proven useful for understanding the magnetic prope rties of the isostr uctural Ln-containing species 14. As for complex 14, we believe the overall behavior is similar to 15, albeit fitting of the data becomes really complicated because of the presence of the strong spinorbit coupling effects arising from the highly anisotropic YbIII ions. Additionally, there are intermolecular exchange interactions, via the hydrogenbonding network as described earlier, in complex 14, which are not incorporated in th e model. These points need to be considered further if a more quantitative ma gnetic understanding of this whole family of mixed-metal species, as well as others being prepared, is required. 3.2.3.3 AC studies of complexes 9, 10, 11, and 13 Alternating current magnetic susceptibil ity studies were performed on vacuumdried microcrystalline samples of 9, 10, 11, and 13 in the temperature range 1.8-10 K with a zero dc field and a 3.5 G ac field oscillating at freque ncies between 50-1000 Hz. Typically in an ac susceptibilit y experiment, a weak field (g enerally 1 G) oscillating at a particular frequency ( ) is applied to a sample to probe the dynamics of the magnetization (magnetic moment) relaxation. A decrease in the in-phase ac susceptibility signal and a concomitant increase in the out-o f-phase signal are indica tive of the onset of the slow, superparamagnetlike relaxation of SMMs.7, 9 This occurs because at low enough temperatures, where the thermal energy is lower than the barrier for relaxation, the magnetization of the molecule cannot rela x fast enough to keep in phase with the

PAGE 81

65 oscillating field. The obtained data for complexes 11 and 13, plotted as M T vs. T for the in-phase (M ) component and M (out-of-phase) vs. T, are shown in Figures 3-8 and 39, respectively. For complex 9, the results of the ac susceptibility measurements did not display any frequency-dependent signals or even tails of peaks which might lie below the operating minimum of 1.8 K of our SQUID magnetometer. Additionally for complex 10, at temperatures below 2.5 K, weak tails of M'' signals whose peak maxima lie at temperatures below the operating minimum of our SQUID magnetometer (1.8 K) were observed. These M'' signals are accompanied by a concomitant frequency-dependent decrease in the in-phase ( M'T) signals, albeit at these low temperatures. The results of the ac susceptibility studies on 11 are depicted in Figure 3-8. As can be seen, the Dy complex disp lays strong frequency-dependant signals below 4 K. Indeed, a frequency-dependent decr ease in the in-phase M T component (Figur e 3-8, top) is accompanied by a concomitant increase in the out-of-phase M signals (Figure 3-8, bottom). The latter merely tails of peaks that lie at < 1.8 K, the operating limit of our SQUID. These signals are indica tive of slow magnetization ( M ) relaxation. The M T value for complex 11 at 10 K of 61 cm3mol-1K remains more or less constant till 4 K, after which it drops rapidly from the slow re laxation effect. The ac data are in agreement with the dc data and both in conjunc tion suggest a large ground state spin S for complex 11. The slow-magnetization relaxation effects observed in complex 13 are similar to those seen for 11, as is illustrated in Figure 3-9 for 13. Both in-phase and out-of-phase frequency-dependent signals in dicative of the onset of sl ow-magnetization relaxation are observed at temperatures < 3.4 K. These signa ls are strong, albeit their peaks still lie

PAGE 82

66 below 1.8 K. For 13 the M T value at 10 K of 49.0 cm3mol-1K drops slightly to 47.5 cm3mol-1K, before rising to a maximum of 50.0 cm3mol-1K at 3.4 K. Thereafter, the M' T (cm3mol-1K) 45 50 55 60 65 Temperature (K) 0246810 M'' (cm3mol-1) 0 1 2 3 4 5 6 1000 Hz 500 Hz 250 Hz 50 Hz Figure 3-8. Ac susceptibility of complex 11 in a 3.5 G field oscill ating at the indicated frequencies. (Top) in-phase signal ( M ) plotted as M T vs T ; and (bottom) outof-phase signal M vs T superparamagnet-like slow magnetization re-o rientation initiates and consequently the M T value drops (Figure 3-9, top). The appearance of out-of-phase M'' signals suggests that complexes 10, 11, and 13 may indeed have a significant (vs. kT ) barrier to magnetization relaxation and thus may be SMMs. In fact, dc and ac magnetic su sceptibility studies were also performed on complexes 3 and 7, the results of which have not been discussed for brevitys sake. Nevertheless, both these complexes also s howed frequency depe ndent signals in ac

PAGE 83

67 M' T (cm3mol-1K) 46 48 50 52 54 56 58 60 Temperature (K) 0246810 M'' (cm3mol-1) 0 1 2 3 4 5 6 1000 Hz 500 Hz 250 Hz 50 Hz Figure 3-9. Ac susceptibility of complex 13 in a 3.5 G field oscill ating at the indicated frequencies. (Top) in-phase signal ( M ) plotted as M T vs T ; and (bottom) outof-phase signal M vs T studies with the strength of the signals (tempe rature below which the signals appeared) in descending order being Dy, Tb > Ho, Gd > Nd (complexes 11, 13, 12, 10, and 8, respectively). Note that the Eu complex (9) did not show any frequency-dependent Ac signals. However, frequency-dependent ac signals are necessary but not sufficient evidence that an SMM has been obtained. Conf irmation of this requires magnetization vs applied dc field sweeps to display hysteresis loops, and this was e xplored on with studies at temperatures below 1.8 K on complexes 8-13, as described below.

PAGE 84

68 3.2.3.4 Hysteresis studies below 1.8 K on complexes 8-13 To establish whether complexes 8-13 were SMMs, magnetization vs. applied dc field data down to 0.04 K were collected on si ngle crystals (that had been kept in contact with mother liquor) using a micro-SQUID apparatus.50 The observation of hysteresis loops in such studies represents the diagnos tic property of a magnet, including SMMs and superparamagnets below their blocking temperature ( TB). The observed magnetization responses for complexes 8-10, and 12 at a fixed field sweep ra te and at the indicated variable temperatures ar e shown in Figure 3-10 (8 (top; left), 9 (top; right), 10 (bottom; left), and 12 (bottom; right)). Hysteresis loops were indeed observed for complexes 8, 10, and 12; the coercivities of the loops increase with decreasing temperature as expected for the superparamagnet-like proper ties of a SMM. The results t hus confirm the addition of complexes 8, 10, and 12 as new members to the growing family of SMMs. For complex 9 (Mn11Eu4) there is no coercivity even at temperatures down to 0.04 K, which is satisfyingly consistent with the earlier me ntioned absence of any frequency-dependent signals in ac magnetic suscep tibility studies performed on 9. Additionally, the ground state spin S of 9 was postulated to be very small based on the 9.46 cm3mol-1K dc value at 5.0 K, and comparative analyses with other complexes, hence precluding any possibility of 9 possessing a big enough barrier to ma gnetization relaxation. Thus, complex 9 is not a SMM. For complex 8, the hysteresis loops slightly open up at 0.5 K and a clear coercivity is observed at 0.04 K and at the fixed scan rate of 0.14 T/s. The overall behavior in hysteresis studies of complexes 8 and 12 is very similar as shown at a fixed scan rate of 0.035 T/s in Figure 3-10, bottom (left and ri ght, respectively). In both these cases, hysteresis loops are se en below 0.5 K, although the coercivities of the loops in 12 are broader than those observed for 10. Nonetheless, the observan ce of hysteresis loops with

PAGE 85

69 -1 -0.5 0 0.5 1 -1-0.500.51 0.04 K 0.5 K 1.0 K 2.0 K 4 K 7 K M/Ms 0H (T) 0.14 T/s -1 -0.5 0 0.5 1 -1-0.500.51 0.04 K 1.1 K 7 K M/Ms 0H (T) Mn11Eu4 -1 -0.5 0 0.5 1 -1-0.500.51 0.04 K 1.1 K 7 K M/Ms 0H (T) Mn11Eu4 -1 -0.5 0 0.5 1 -0.8-0.6-0.4-0.200.20.40.60.8 0.06 K 0.2 K 0.3 K 0.4 K 0.5 K M/Ms 0H (T) 0.035 T/s Figure 3-10. Magnetization ( M ) vs dc field ( H ) hysteresis loops at a fixed field sweep rate and at the indicated variable temperatures for single crystals of 8 (top; left), 9 (top; right), 10 (bottom; left), and 12 (bottom; right). The magnetization is normalized to its saturation value, MS. increasing coercivities with decreasing te mperature, unequivocally establish that complexes 10 and 12 are indeed single-molecule magnets. The hysteresis loops for complexes 11 (top) and 13 (bottom) are depicted in Figure 3-11. Shown in Figure 3-11 (top, left) are the magnetization responses for 11 at a dc field sweep rate of 0.14 T/s with the fiel d approximately along the easy axis ( z axis) of the molecule, and at variable temperatures in cluding and below 1.0 K. Figure 3-11 (top, right) depicts the dc field sweep rate depe ndence of the hysteresis loops observed for 11 at 0.04 K. Analogous data collected on complex 13 are presented in Figure 3-11 (bottom), at a sweep rate of 0.14 T/s (left) and at a fixed temperature of 0.04 K (right). -1 -0.5 0 0.5 1 -1-0.500.51 0.04 K 0.2 K 0.3 K 0.4 K 0.5 K M/Ms 0H (T) 0.035 T/s

PAGE 86

70 Hysteresis loops were observed for complexes 11 and 13, and as anticipated, the coercivities (widths) of th ese loops increase with decrea sing temperature (Figure 3-11, left) and increasing sweep rate (Figure 3-11, right), as expected for the superparamagnetlike properties of a SMM. -1 -0.5 0 0.5 1 -0.6-0.4-0.200.20.40.6 0.04 K 0.2 K 0.3 K 0.4 K 0.5 K 0.6 K 0.8 K 1.0 K M/Ms 0H (T) 0.14 T/s -1 -0.5 0 0.5 1 -0.6-0.4-0.200.20.40.6 0.140 T/s 0.070 T/s 0.035 T/s 0.017 T/s 0.008 T/s 0.004 T/s 0.002 T/s 0.001 T/s M/Ms 0H (T) 0.04 K -1 -0.5 0 0.5 1 -0.6-0.4-0.200.20.40.6 0.04 K 0.10 K 0.20 K 0.30 K 0.4 K 0.5 K 0.6 K 0.7 K 0.8 K M/Ms 0H (T) 0.14 T/s -1 -0.5 0 0.5 1 -0.6-0.4-0.200.20.40.6 0.280 T/s 0.140 T/s 0.070 T/s 0.035 T/s 0.017 T/s 0.008 T/s 0.004 T/s 0.002 T/s 0.001 T/s M/Ms 0H (T) 0.04 K Figure 3-11. Magnetization ( M ) vs dc field ( H ) hysteresis loops fo r single crystals of 11 (top), and 13 (bottom) at the indicated temper atures and a fixed field sweep rate of 0.14 T/s (left), and at the indicated field sweep rates and a fixed temperature 0.04 K (right). The magnetiza tion is normalized to its saturation value, MS. Hysteresis in magnetization versus fiel d sweeps is the classical property of a magnet, and such loops are a diagnostic signa ture also of SMMs and superparamagnets. The data thus indicate complexes 11 and 13 to be new additions to the family of mixed Mn/Ln SMMs, and only the second member of this family to display hysteresis, the only

PAGE 87

71 other being the Mn2Dy2 complex reported recently.59d The blocking temperature ( TB) is ~ 1.0 K for 11 and 13, above which there is no hysteresis (Figure 3-11 (left)); i.e. the spin relaxes faster to equilibrium than the time scale of the hysteresis loop measurement. For several other SMMs studied to date, the hyste resis loops have not been smooth but have instead displayed step-like feat ures at periodi c field values.7, 61b,c, 65, 12 These steps correspond to increased magnetization relaxatio n rates and are due to quantum tunneling of the magnetization (QTM) thr ough the anisotropy energy barrier.57 The hysteresis loops in Figure 3-11 do not show such periodic step s because of a distribution of molecular environments arising from the presence of seve rely disordered solvent molecules, as well as chelates such as nitrates. Theref ore, the local environment around the Mn11Ln4 molecules vary, which in combination with low-lying excited spin states (exchange interactions ( J ) of Ln ions are typically weak), lead to a distribution in D values, that is, a distribution in energy barriers to relaxation; the separations between steps is directly proportional to D so a distribution in D would give a distribution in step position as a consequence of which the steps are broadened to the extent that they are smeared out and thus not visible. In addition, intermolecular interactions (both dipol ar and exchange) and population of excited states can result in step broadening. It should be noted that although such disorders in solvents and ligand positions might at first glance appear to be too trivial to cause a noticeable variation in the molecular D value, there is ample precedent that this is not the case. Instead, on seve ral occasions we have observed that such properties of SMMs are acutely sensitive to relatively small changes in the local environment of the molecules.66, 67 The same smearing out of steps has been observed previously, especially in high nuc learity Mn complexes such as Mn16,68 Mn21,69 Mn22,70

PAGE 88

72 Mn25,47 Mn30,52 and Mn84,102b to name but a few, and thes e were similarly assigned to a distribution in D values. Although a justification has been given for the absence of steps in hysteresis loops through the above discussion, the hysteresis loops st ill suggest that QTM is occurring in complexes 11 and 13. This is proposed because at temperatures below 0.1 K the loops become temperature independent, but a scan -rate study (Figure 3-11 (right)) shows that the loops are still time-dependent. Further confirmation of this behavior was sought through time dependent dc magne tization decay data collected in the 0.04 1.0 K range. This study is useful in obtain ing data which can be used to construct an Arrhenius plot based on the Arrhenius relationship of eq 3-4, where 0 is the preexpone ntial factor, Ueff is the mean effective barrier to magnetization relaxation, and k is the Boltzmann constant. = 0 exp( Ueff/k T ) (3-4) Typically in a dc magnetization decay study, th e samples magnetization is first saturated in one direction at ~ 5 K with a large dc field, the temper ature is then lowered to a 0 0.1 0.2 0.3 0.4 0.5 0.6 0.11101001000M/Ms t (s) 0.75 K 0.7 K 0.65 K 0.55 K 0.45 K 0.6 K 0.5 K 0.4 K 0.04 K 0.35 K 0.3 K 0.25 K 0.2 K 0.15 K 0.1 K 10-510-410-310-210-1100101102103012345678 (s) 1/T (1/K) Figure 3-12. (Left) Magnetization vs time decay plots for crystals of complex 13 at the indicated temperatures. (R ight) Arrhenius plot usi ng the resulting relaxation time () versus T data. The dashed line is the fit of the data in the thermally activated region, to the Arrhenius equati on. See the text for the fit parameters.

PAGE 89

73 chosen value, and then the field removed and the magnetization decay monitored with time. This provided magnetiz ation relaxation rate (1/) versus temperature (T) data, which were fit to the Arrh enius relationship of eq 3-4. The corresponding dc magnetization decay plot (left) and the resu lting Arrhenius plot (right) for complex 13 (Mn11Tb4) are depicted in Figure 3-12. The Arrhenius plots for complexes 11 (left) and 10 (right) are depicted in Figure 3-13. For complexes 11 and 13, the dc decay data was collected in the 0.04 1.00 K range, and for complex 10 the relaxation time () was determined for magnetization decay spanning th e 0.5 0.04 K temperature range. Fits of the thermally activated region above ~0.6 K for 13, ~0.5 K for 11, and ~0.3 K for 10 (shown as a dashed line in Figur es 3-12 (right), and 3-13), gave 0 = 5.8 x 10-8 s and Ueff/k = 9.9 K for the Mn11Tb4 complex (13). For the Mn11Dy4 complex the parameters were 0 = 4 x 10-8 s and Ueff/k = 9.3 K, and for the Mn11Gd4 complex 0 = 7 x 10-13 s and Ueff/k = 9.0 K. For the sake of comparison, thes e parameters are listed in Table 3-5 for Figure 3-13. Arrhenius plot of the relaxation time () versus 1/ T constructed from Dc magnetization decay data for complexes 11 (left), and 10 (right). The dashed line in green is the fit of the data in the thermally activated region to the Arrhenius relationship with the indicated parameters. complexes 10, 11 and 13. Below ~ 0.1 K, the relaxati on lifetime becomes essentially temperature-independent for complexes 10, 11 and 13. Although, the magnetization 10-710-510-310-110110310510705101520 (s) 1/T (1/K) = 7e-13 e^(9/T) 10-510-310-110110310510705101520 (s) 1/T (1/K) = (4 x 10-8)exp(9.3/T) 10-510-310-110110310510705101520 10-510-310-110110310510710-510-310-110110310510705101520 05101520 (s) 1/T (1/K) = (4 x 10-8)exp(9.3/T)

PAGE 90

74 relaxation processes in the Tb (13) and Gd (10) complexes are very complicated, the Mn11Dy4 complex's relaxation phenomenon is easi er to comprehend. Thus, a closer Table 3-5. Comparison of the SMM parameters of complexes 10, 11 and 13. Complexes (Mn11Ln4) 0 (s) (Preexponential) Ueff/k (K) (Kinetic barrier) TB (K)a (Blocking temp.) Mn11Gd4 (10) 7.0 x 10-13 9.0 ~0.5 Mn11Dy4 (11) 4.0 x 10-8 9.3 ~1.0 Mn11Tb4 (13) 5.8 x 10-8 9.9 ~1.0 a Approximate value of the temperature be low which hysteresis loops are observed. inspection of the Arrhenius plot for complex 11 (Fig. 3-13, left) reveals that below ~ 0.1 K (1/T = 10), the relaxation lifetime becomes essentially temperature independent at ~106 s, consistent with the purely quantum re gime where QTM through the anisotropy barrier is only via the lowest energy MS levels. In other words, in this ground-state QTM region tunneling is now only betw een the lowest-energy MS = S levels ( S is the ground state spin of the Mn11Dy4 complex), and no longer via a th ermally (phonon) assisted pathway involving higher-energy MS levels. The crossover temperature to this ground-state tunneling from the thermally activated re laxation is between 0.1 and 0.2 K. The results of the dc magnetization decay data study on complex 10 (Mn11Gd4) provided an effective (kinetic) energy barrier Ueff = 9.0 K. This Ueff may be compared to the U = S2| D | = 7.0 K obtained from ma gnetization fits. Although the U is generally slightly bigger than Ueff in most cases, the anomaly here arises due to the reduced magnetization procedure not being so accurate mainly due to the assumption of the population of only the ground state, as well as the application of a very simplified model which neglects rhombic anisotropy ( E ) and higher order anisotropy terms. Nevertheless both the barrier values U and Ueff for complex 10 still fall in the same ballpark. Thus, at a

PAGE 91

75 qualitative level, and to some extent to a qua ntitative level, the various dc, ac, hysteresis, and magnetization relaxation data are all consistent. It should be noted that in ac studies the strength of the signals (temperature belo w which the signals appeared) in descending order was Dy, Tb > Ho, Gd > Nd (complexes 11, 13, 12, 10, and 8, respectively). Correspondingly, the temperature below whic h hysteresis was observed, the kinetic barrier Ueff, as well as the Dc value at 5 K, followed more or less the same decreasing pattern for all of these Mn11Ln4 complexes. 3.3 Conclusions In this and earlier published work,59b we have demonstrated the initial observation of magnetization hysteresis and quantum tunne ling, two confirmatory properties of SMMs, in mixed-metal 3d/4f single-molecule magnets. The observation of hysteresis, especially in heterometallic lanthanide-c ontaining SMMs is of utmost importance because it is a well established fact now that the Ln component causes these complexes to suffer from the disadvantage of fast QTM rates.59d, f-h This represents a diminution of the effective barrier for magnetization relaxation, and thus the temperature below which the relaxation is blocked and the complex will func tion as a SMM. Additionally at zero field, the tunneling is so fast that many of the reported complexes do not show superparamagnet-like behavior, in other word s, a barrier to magnetization relaxation at H = 0.59d However, and very interestingly, the Mn11Ln4 complexes reported here do not suffer from fast QTM. Consequently, important parameters such as the effective barrier to magnetization relaxation, Ueff, were obtained for the Dy and Tb complexes which were 9.3 and 9.9 K, respectively. The preexponential (0) values of 4.0 x 10-8 s (Mn11Dy4) and 4.8 x 10-8 s (Mn11Tb4), are comparable and satisfying c onsistent with those reported for

PAGE 92

76 other SMMs. Both these complexes show fr equency-dependent ac susceptibility signals which in conjunction with the observance of hysteresis loops below 1.0 K confirmed that 11 and 13, irrefutably are SMMs. Magnetization fits for the Mn11Gd4 complex, containing magnetoisotropic GdIII ions, yielded an S = 9 and D = .06 cm-1. The ability to probe both classical (h ysteresis) and quantum (QTM) magnetism behavior as a function of th e lanthanide ion as demonstrated, by the extension of a standard synthetic method to se veral lanthanides, should provi de an invaluable means of improving our understanding of both the chemistry and physics of this area of heterometallic molecular nanomagnetism. In a ddition, the large vari ation in spin and anisotropy within the la nthanide ions mentioned earlier, coupled with th e site-selective replacement of the Ln3+ ions in the various Mn11Ln4 complexes, offers the possibility of spin-injection and/or ani sotropy-injection within th ese SMMs, by the choice of a suitable Ln3+ ion. Thus, it allows greater contro l over a family of related SMMs and investigation of the magnetic properties in a causal manner. Finally, we comment that these complexes along with the ongoing research efforts in this area provide proof-offeasibility of extending the single-molecu le magnetism phenomenon to lanthanidecontaining species. This exciting research area thus offers the tantalizing possibility of being able to raise th e blocking temperature ( TB) of SMMs to above that of the Mn12 family. 3.4 Experimental 3.4.1 Syntheses All chemicals were used as received unless otherwise stated. All manipulations were performed under aerobic conditions. [Mn3O(O2CPh)6(py)2(H2O)] (2)54 and (NBun 4)[Mn4O2(O2CPh)9(H2O)] (3)55 were prepared as described previously.

PAGE 93

77 [Mn11Nd4O8(OH)8(O2CPh)16(NO3)5(H2O)3] 21 MeCN (8). Solid Nd(NO3)3 6 H2O (0.32g, 0.74 mmol) was added to a solution of complex 3 (0.50g, 0.31 mmol) in MeOH/MeCN (1ml/20ml). After stirring for 20 min. the solution was filtered and the brown filtrate slowly concentrated by evapora tion. Within a week nice dark red crystals of 8 were obtained in isolated 50% yield. These were collected by filtration and dried in vacuo. The dried solid analyzed as fu lly desolvated. Anal. Calcd (Found) for C112H94N5O66Mn11Nd4: C, 35.90 (36.13); H, 2.53 (2.52); N, 1.87 (1.68). Selected IR data (KBr, cm-1): 3420 (br), 1599 (m), 1562 (m), 149 2 (w), 1449 (w), 1392 (s), 1178 (w), 1025 (w), 715 (s), 679 (m), 604 (m), 475 (w). [Mn11Eu4O8(OH)8(O2CPh)16(NO3)5(H2O)3] (9). Solid Eu(NO3)3 5 H2O (0.32g, 0.74 mmol) was added to a st irring solution of complex 2 (0.50g, 0.45 mmol) in MeOH/MeCN (1:20 v/v). After stirring for 20 min. the solution was filtered and the brown filtrate slowly concentrated by evapora tion. Within a week, nice orange crystals of 9 were obtained in isolated 50% yield. Thes e were collected by filtration and dried in vacuo. The dried solid analyzed as solvent free. Anal. Calcd (Found) for C112H94N5O66Mn11Eu4: C, 35.61 (35.45); H, 2.51 (2.45); N, 1.85 (1.71). Selected IR data (KBr, cm-1): 3419 (br), 1599 (m), 1560 (m), 149 1 (w), 1449 (w), 1385 (s), 1178 (w), 1025 (w), 816(w), 715 (s), 679 (m), 604 (m), 474 (w). [Mn11Gd4O8(OH)8(O2CPh)16(NO3)5(H2O)3] 15 MeCN (10). Solid Gd(NO3)3 6 H2O (0.41g, 0.91 mmol) was added to a solution of complex 2 (0.50g, 0.45 mmol) in MeOH/MeCN (1:20 v/v). After stirring for 20 min. the solution was filtered and the brown filtrate slowly concentrated by evapor ation. Within a couple of days dark red crystals of 10 were obtained in isolated 55% yield. These were collected by filtration and

PAGE 94

78 dried in vacuo. The dried so lid analyzed as solvent free. Anal. Calcd (Found) for C112H94N5O66Mn11Gd4: C, 35.41 (35.15); H, 2.49 (2.44); N, 1.84 (1.69). Selected IR data (KBr, cm-1): 3421 (br), 1599 (m), 1562 (m), 149 2 (w), 1449 (w), 1392 (s), 1178 (w), 1025 (w), 715 (s), 679 (m), 604 (m), 476 (w). [Mn11Dy4O8(OH)6(OMe)2(O2CPh)16(NO3)5(H2O)3] 15 MeCN (11). Solid Dy(NO3)3 5 H2O (0.40g, 0.90 mmol) was added to a stirring solution of complex 2 (0.50g, 0.45 mmol) in MeOH/MeCN (1:20 v/v) After stirring for 20 min. the solution was filtered and the brown filtr ate slowly concentrated by ev aporation. Within a couple of days nice dark black crystals of 11 were obtained in isolated 55% yield. These were collected by filtration and dried in vacuo. Th e dried solid analyzed as fully desolvated. Anal. Calcd (Found) for C114H98N5O66Mn11Dy4: C, 35.75 (35.91); H, 2.53 (2.56); N, 1.83 (1.74). Selected IR data (KBr, cm-1): 3421 (br), 1599 (m), 1561 (m), 1492 (w), 1449 (w), 1385 (s), 1178 (w), 1025 (w), 716 (s), 680 (m), 606 (m), 567(w), 477 (w). [Mn11Ho4O8(OH)8(O2CPh)18(NO3)3(H2O)7] 21 MeCN (12). Solid Ho(NO3)3 5 H2O (0.40g, 0.90 mmol) was added to a solution of complex 3 (0.50g, 0.31 mmol) in MeOH/MeCN (1:20 v/v). After stirring for 20 min. the solution was filtered and the brown filtrate slowly concentrated by evapora tion. Within a week nice dark red crystals of 12 were obtained in isolated 55% yield. Thes e were collected by filtration and dried in vacuo. The dried solid analyzed as solvent-free. Anal. Calcd (Found) for C126H104N3O64Mn11Ho4: C, 38.33 (38.20); H, 2.66 (2.52); N, 1.06 (1.04). Selected IR data (KBr, cm-1): 3420 (br), 1600 (m), 1560 (m), 1449 (w), 1386 (s), 1178 (w), 1025 (w), 716 (s), 680 (m), 606 (m), 476 (w). [Mn11Tb4O8(OH)6(OCH2Ph)3(O2CPh)20(PhCH2OH)2(H2O)] 3PhCH2OH (13).

PAGE 95

79 Solid Tb(NO3)3 5 H2O (0.19g, 0.45 mmol) was added to a solution of complex 3 (0.5g, 0.31 mmol) in PhCH2OH/MeCN (5ml/20ml). After stirri ng for 20 min. the solution was filtered and the brown filtrate slowly concen trated by evaporation. Within a couple of weeks nice dark red crystals of 13 were obtained in isolated 55% yield. These were collected by filtration and dried in va cuo. The dried solid analyzed as 13 2PhCH2OH. Anal. Calcd (Found) for C189H161O62Mn11Tb4: C, 48.67 (48.51); H, 3.48 (3.35). Selected IR data (KBr, cm-1): 3415 (br), 1601 (m), 1563 (m), 1449 (w), 1385 (s), 1177 (w), 1023 (w), 718 (s), 680 (m), 606 (m), 562 (w), 473 (w). [Mn2Yb2O2(O2CPh)6(OMe)4(MeOH)4] 2MeOH (14). To a solution of complex 3 (0.5g, 0.31 mmol) in MeCN/MeOH (20/5 mL) was added Yb(NO3)35H2O (0.31 g, 0.62 mmol) and stirred for 25 min to give a brow n solution. This was filtered and more MeOH (10 mL) added to the filtrate. The resu lting solution was slowly concentrated by evaporation at room temper ature for 5 days, during which time brown crystals of 14 2MeOH slowly grew. These were collected in 10% yield and dried in vacuum. Anal. Calc. (Found) for 14: C 40.94 (40.87), H 3.99 (3.96). Selected IR data (KBr, cm-1): 3426 (br), 1600 (m), 1540 (m), 1384 (s), 1026 (w), 718 (s), 682 (m), 624 (s), 545 (m), 476 (w). [Mn2Y2O2(O2CPh)6(OMe)4(MeOH)4] 2MeOH (15). To a solution of complex 3 (0.5g, 0.31 mmol) in MeCN/MeOH (20/5 mL) was added Y(NO3)35H2O (0.24 g, 0.62 mmol) and stirred for 25 min to give a brow n solution. This was filtered and more MeOH (10 mL) added to the filtrate. The resulti ng solution was left undisturbed at room temperature for a week to slowly produce orange crystals of 15 2MeOH, which were collected in 25% yield and dried in v acuum. The same product was obtained using YCl3H2O instead of Y(NO3)3H2O. Anal. Calc. (Found) for 15: C 46.24 (46.15), H

PAGE 96

80 4.50 (4.46). Selected IR data (KBr, cm-1): 3412(br), 1594(m), 1545 (s), 1448 (w), 1391 (s), 1025 (w), 718 (s), 681 (m), 623 (s), 540 (m), 474 (w). 3.4.2 X-ray Crystallography Data were collected on a Siemens SM ART PLATFORM equipped with a CCD area detector and a graphite monochromator utilizing MoK radiation ( = 0.71073 ). Suitable crystals of solvated 8, 10, 11, 12, 13, 14, and 15 were attached to glass fibers using silicone grease and transferred to a goni ostat where they were cooled to 173 K for data collection. An initial search of recipr ocal space revealed a triclinic cell with the choice of space group being 1 P for the Mn11Ln4 complexes 8-13, and a monoclinic cell with the choice of space group being P 21/c for the Mn2Ln2 complexes 14-15, respectively. Cell parameters were refined us ing up to 8192 reflections. A full sphere of data (1850 frames) was collected using the -scan method (0.3 fram e width). The first 50 frames were re-measured at the end of data collection to monitor instrument and crystal stability (maximum correction on I was < 1 %). Absorption corrections by integration were applied based on measured indexed crystal faces. The structures were solved by direct methods in SHELXTL6 ,56a and refined on F2 using full-matrix least squares. The non-H atoms were treated anisot ropically, whereas the hydrogen atoms were placed in calculated, ideal positions and refined as ri ding on their respective carbon atoms. The asymmetric unit of 821MeCN consists of half the Mn11Nd4 cluster lying on an inversion centre, along with 10.5 MeCN molecules as solvents of crysta llization. A total of 897 parameters were included in the struct ure refinement using 9582 reflections with I > 2 (I) to yield R1 and wR2 of 8.63% and 12.70%, respectively.

PAGE 97

81 For 10MeCN, the asymmetric unit consists of half the Mn11Gd4 cluster located on an inversion center and 7.5 MeCN molecule s as crystallization solvents. All H atoms of the water ligands could not be located and were not included in the final refinement cycles. A total of 907 parameters were includ ed in the structure refinement using 10499 reflections with I > 2 (I) to yield R1 and wR2 of 6.56% and 16.22%, respectively. The asymmetric unit of 1115MeCN consists of half the Mn11Dy4 cluster lying on an inversion centre, along with 7.5 MeCN molecules as solven ts of crystallization. A total of 1041 parameters were included in the structure refinement using 12052 reflections with I > 2 (I) to yield R1 and wR2 of 4.25% and 10.06%, respectively. The asymmetric unit of 1221MeCN consists of half the Mn11Ho4 complex lying on an inversion centre, along with 10.5 MeCN molecules as solven ts of crystallization. A total of 934 parameters were included in th e structure refinement using 14245 reflections with I > 2 (I) to yield R1 and wR2 of 5.19% and 14.00%, respectively. For all of the above mentioned complexes some of the solvent molecules were disordered and could not be modele d properly, thus the SQUEEZE program,56b a part of the PLATON56c package of crystallographic software was used to calculate the solvent disorder area and remove its contribution to the overall intensity data. All of the complexes 8, 10, 11, and 12 have one MeCN (50% occupanc y) close to a disorder on one of the Ln ions where one coordination site is 50% NO3and another 50% H2O. The acetonitrile exists only with th e water on one of the four Ln ions in each of the complexes (for example Dy2 in the Mn11Dy4 complex). Thus, although all complexes are centrosymmetric, they have odd number of NO3, H2O and MeCN molecules because of

PAGE 98

82 the earlier stated disorder of a NO3 ion related in occupancy with the MeCN/H2O molecules. The asymmetric unit of 13PhCH2OH consists of half the Mn11Tb4 and three halves of benzyl alcohol solvent molecules (all three were disordered and each were refined in two parts). Tb1 has a disorder of a water molecule and benzyl alcohol anion in the same coordination position having O32 co mmon to both. Thus the C and H atoms of this ligand were refined with 50% occupanc ies while the water protons of the water counterpart were not located and not included in the final refinement. The phenyl groups of the solvent molecules, and those of C22 and C82 ligands, were re fined as rigid bodies and constrained to maintain a perfect he xagon. The hydroxyl protons and those of the water molecules were not located and were not included in the final refinement cycles. A total of 744 parameters were refined in the final cycle of refinement using 29065 reflections with I > 2 (I) to yield R1 and wR2 of 7.73% and 19.42%, respectively. The asymmetric unit of 142MeOH consists of half the Mn2Yb2 cluster and a methanol molecule of crystallization. A total of 378 parameters were included in the final cycle of refinement using 4821 reflections with I > 2 (I) to yield R1 and wR2 of 3.11% and 7.67%, respectively. The asymmetric unit of 152MeOH consists of half the Mn2Y2 cluster and a methanol molecule of crystallization. A total of 378 parameters were included in the final cycle of structure refinement using 2803 reflections with I > 2 (I) to yield R1 and wR2 of 3.48% and 7.27%, respectively.

PAGE 99

83 CHAPTER 4 HIGH NUCLEARITY COMPLEXES: HOMOVALENT [Th6Mn10O22(OH)2(O2CPh)16(NO3)2(H2O)8] AND MIXED-VALENT [Mn7O5(OR)2(O2CPh)9(terpy)] (R = Me, CH2Ph) DISPLAYING SLOWMAGNETIZATION RELAXATION 4.1 Introduction We have had a longstanding interest in the development of manganese carboxylate cluster chemistry, mainly due to its relevan ce to a variety of areas including bioinorganic chemistry,18 nanoscale magnetic materials7-11 and catalysis of various oxidation processes.71 For example, Mn carboxylate cluste rs are the primary source of singlemolecule magnets (SMMs), individual molecules that retain their magnetization orientation below a blocking temperature in the absence of an applied field.9 In recent work, we and others have turned our attenti on to high-nuclearity mixe d 3d/4f clusters of Mn as a route to potentially interesting new species, and a number of heterometallic complexes of this type are now available.37, 38, 72 In addition, both mixed 3d/4d and 3d/5d SMMs containing Mn are now known.73 As already illustrated in Chapters II and III, mixed Mn/Ln systems are a rich source of SMMs. An obvious extension of the above -stated efforts in high nuclearity mixedmetal cluster chemistry is to ask whether 3d /5f clusters might be accessible with Mn, and if yes, whether they would function as SMMs We have taken up this challenge using Th. There are relatively few well characterized transition metal/actinide complexes, among which are the dinuclear metal-metal bonded MAn organometallic complexes (M= Fe, Ru and An= Th, U)74a and the family of linear trimetallic M2 IIUIV (M= Co, Ni, Cu, Zn)

PAGE 100

84 complexes containing a he xadentate Schiff base.74b However, only one of these contains Mn, trinuclear [MnU2O2L2(py)4] (L= 1,7-diphenyl-1,3,5,7-heptane-tetronato).75 Although thorium is used in a wide array of pr oducts and processes, the cluster chemistry of Th is poorly developed compared to tr ansition metals: Currently there are metalorganic frameworks76a and organically templated thorium complexes76b known, and the largest molecular Th complex is Th6. 77 We can now report the first mixed Mn/Th molecular cluster, Mn10Th6, which we believe to be the prototype of a potentially large new area of cluster chemistry. Additionally, although mixed-metal systems are a very exciting research area, they still suffer from the fact that in a multicomponent paramagnetic system there is more than one g -tensor value. Hence, magnetic charac terization of these systems is very difficult. Therefore, magnetic parameters such as S D and isotropicg values are difficult to obtain with the current fitting programs which we possess,46 mainly because of secondorder spin-orbit couplin g and higher order anisotropy terms. Thus, in the area of singlemolecule magnetism, research on homometallic Mn containing SMMs is still the primary focus, with the principal objective being th e preparation of a SMM that behaves as a magnet at technologically relevant temperatures, i.e. at least 77 K (that of liquid N2). This goal has been approached primarily by two means: (i) the pr eparation of novel 3d metal carboxylate clusters posse ssing differing topologies that may behave as SMMs and (ii) the addition of new derivatives to already existing families of SMMs so that structural features may be correlated with magnetic properties and ultimately, a rational synthetic method for improved SMMs may be developed. The first SMM discovered was [Mn12O12(O2CMe)16(H2O)4],7 and synthetic manipulation of this complex has provided a

PAGE 101

85 very well studied family of complexes.30 Ever since the discovery of Mn12 complexes, several types of SMMs have been discovere d, most of them containing primarily MnIII ions.78 Thus, an undying theme in the field of SMMs, and polynuclear cluster chemistry for that matter, is achieving synthetic cont rol and investigating properties in a causal manner, i.e. how does one factor (for example stru ctural change) influence all the other resulting properties (like magnetic propertie s). With respect to manganese carboxylate cluster chemistry, there have been thor ough investigations and some structural Mn4 motifs such as the defect-dicubanes,32 cubanes,79 butterflies,80 and adamantane81 complexes which are thermodynamically inhere ntly stable, have been isolated and characterized in great detail. However, one wonders whether ther e exist any trapped intermediates or transiti on-state clusters comprising a combination of the abovementioned motifs which have not yet been isol ated. Indeed, as a re sult of our continuing efforts towards the search of new synthetic ro utes and structural t ypes we herein report two Mn7 clusters which have a very unusua l metallic core resembling a fused butterfly/cubane (via the body oxygen atom of the butterfly) and sharing a common wing Mn atom of the butterfly. The peripherial ligation has been systematically varied to include methoxides and phenyl methoxides. In addition to having a very atypical trapped intermediate core, both these complexes possess a very well isolated ground state spin of S = 6. We describe herein the synthesis, st ructure, and magnetic characterization of both of these complexes, in addition to the Mn10Th6 complex. We also demonstrate that the Mn7 complex with phenyl methoxides displays magnetization hystere sis arising from slow magnetization relaxation.

PAGE 102

86 4.2 Results and Discussion 4.2.1 Syntheses High nuclearity Mn clusters have previ ously been obtained by the reaction of preformed Mnx clusters with bi-, trior tetra-dentate (N/O) chelates.82 Amongst the various MnIII sources which have already been explored, [Mn3O(O2CPh)6(py)2(H2O)](2) (2MnIII, 1MnII) and (NBun 4)[Mn4O2(O2CPh)9(H2O)](3) (4MnIII) are particularly attractive, as they have been known to give higher nuclearity comp lexes. A modification of the above strategy towards developing new s ynthetic routes to heterometallic species is the reaction of preformed Mnx (x = 3 or 4 here) clusters with a heterometal salt / carboxylate. Such reactions are typically carrie d out in a mixed solvent system which not only ensures the solubility of all involved re actants but also provides bridging / terminal ligands. Indeed, reactions employing complexes 2 and 3 in a mixed solvent system like MeCN/MeOH have proven very successful for us in the attainment of heterometallic species, as already discussed in Chapter III.37, 39, 43 Since the use of a mixed solvent system was so successful in reactions including Mn and Ln, we decided to extend the synthetic chemistry to actinides, since actinides have similar reactivity like lanthanides. Thus, the system was explored fu rther with reactions involving ThIV. Caution: Thorium Nitrate is weakly radioactive and a potent ial mutagen and carcinogen; proper safety precautions should be taken. However, Th(NO3)4 was insoluble in MeCN so the reaction was carried out in a mixed MeCN/MeOH (20:1, v/v) solvent system. Thus, the reaction of complex 3 with Th(NO3)4 in a 5:12 molar ratio as shown in eq 4-1 resulted in the isolation of [Mn10Th6O22(OH)2(O2CPh)16(NO3)2(H2O)8] (16) in 20% yield in about a month. The same reaction but with an increased MnIII 4:ThIV ratio of 1:3 or 1:4 also gave the same product. No isolable products were obtained when only MeCN

PAGE 103

87 5 (NBun 4)[Mn4O2(O2CPh)9(H2O)] + 12 Th(NO3)4 + 49 H2O 2 [Mn10Th6O22(OH)2(O2CPh)16(NO3)2(H2O)8] + 5 (NBun 4)+ + 44 NO3 + 13 PhCO2H + 59 H+ + 20 e(4-1) was used. However, in the above-mentioned mixe d-metal strategy ther e are very little geometric constrains imposed on the resulting cluster; the carboxylat es impose little or no geometry restrictions and bridge multiple metals as oxide bonds are formed. Thus, clusters of various nuclearities have been obtained, although most of them are high nuclearity clusters. If one needed to li mit nuclearity growth, for example, for bioinorganic modeling of the WOC of PS II (s ee next two chapters ), a very logical approach is the use of terminal ch elates, such as the bidentate 4,4 -dipyridyl (bpy) or the tridentate 2,2 :6 ,2 -terpyridine (terpy). Indeed, many clusters have been obtained using this technique, notable among which are the tetranuclear bpy butterfly complexes83 and the dimer-of-dimers84 incorporating terpy; both putati vely synthesized as bioinorganic models18c of the water oxidizing complex of photosynthesis.18 Thus using this synthetic route, reactions were carri ed out employing complexes 2 and 3 in the presence of terpy in a mixed-solvent reaction system. The solvent mixture is necessary to ensure adequate solubility of the reagents especially the terpy. [Mn7O5(OMe)2(O2CPh)9(terpy)] (17) was obtained when the trinuclear 2 was reacted with terpy in MeOH/MeCN (1:10 v/v) as shown in eq 4-2 below. 7 [Mn3O(O2CPh)6(py)2(H2O)] + 3 terpy + H2O + 6 MeOH 3 [Mn7O5(OMe)2(O2CPh)9(terpy)] + 14 py + 15 PhCO2H + 7 H+ + 7 e(4-2)

PAGE 104

88 Similarly, [Mn7O5(OCH2Ph)2(O2CPh)9(terpy)] (18) crystallized out of a benzyl alcohol/acetonitrile (1:2 0 v/v) solvent system as depicted in eq 4-3. No isolable 7 [Mn3O(O2CPh)6(py)2(H2O)] + 3 terpy + H2O + 6 PhCH2OH 3 [Mn7O5(OCH2Ph)2(O2CPh)9(terpy)] + 14 py + 15 PhCO2H + 7 H+ + 7 e(4-3) products were obtained when only MeCN wa s used. This is obviously because the alcohols not only act as solvents but also e nd up as bridging ligands (alkoxides) in the end-product. Indeed, ther e are two methoxides in 17 and two phenyl methoxides in complex 18. Thus, the use of an alcohol in the ab ove reactions results in the addition of more variables and finally in the isolati on of unprecedented compounds which were not possible with a homogeneous solvent system. In the Mn7 reactions, we wanted to investigate the effect of the change in th e alcohol group and what resulting outcome it would have on the end-product and/or the resu lting magnetic propertie s of the isolated complex. Hence, we tried reac tions with benzyl alcohol (p Ka = 15.0) which had a phenyl group rather than a methyl, as in the case of methanol (p Ka = 15.2). Not surprisingly, we obtained isostructural complexes because the p Ka of the alcohols is very similar. When ethanol (p Ka = 16.0) was used as the alcohol, only brown precipitates of manganese oxides/hydroxides were obtaine d and/or no clean products c ould be isolated. Also, if more than 0.6 equivalents of terpy was used in the above reacti ons the end-product was obtained but along with some yellow crystals of unreacted terpy as a contaminant. 4.2.2 Description of Structures 4.2.2.1 X-ray crystal structure of [Mn10Th6O22(OH)2(O2CPh)16(NO3)2(H2O)8] (16) A PovRay representation of complex 16 and its labeled [Mn10Th6O22(OH)2]18+ core is provided in Figure 4-1. Selected bond and interatomic distances, and angles for complexes 16, 17, and 18 are listed in Tables A-6, A-7 and A8, respectively.

PAGE 105

89 Complex 16MeCN crystallizes in the triclinic space group P with the cluster lying on an inversion centre. Although complex 16 is heterometallic, it is homovalent with 10 MnIV and 6 ThIV ions. The structure (Figure 41 top (left)) consists of a Figure 4-1. (Top) PovRay representation at th e 50% probability level of the X-ray crystal structure (left) of 16 (with the benzoate rings omitted for clarity except for the ipso C atoms) and (right) the 2:3:6:3: 2 (Mn:Th:Mn:Th:Mn) layer structure of 16. Bottom depicts the labeled [Mn10Th6O22(OH)2]18+ core of complex 16. The complex resides on a two-fold rotation ax is with an inversion center. Color scheme: MnIV cyan; ThIV light green; N dark blue ; O red; C gray. H atoms have been omitted for clarity. [Mn10Th6O22(OH)2]18+ core comprising 4 4-O2-, 16 3-O2-, 2 -O2and 2 -HObridging ions (Figure 4-1 (bo ttom)). Peripheral ligation aroun d the core is provided by 16

PAGE 106

90 -PhCO2 groups in their familiar syn, syn binding mode, an 2 chelating nitrate on Th3 and Th3', and eight terminal H2O molecules on the other Th atoms, three each on Th1 and Th1', and one each on Th2 and Th2'. Th e doubly bridging benzoate groups bridge either Mn/Th pairs or Mn2 pairs, with no benzoate groups bridging Th2 pairs. The latter are instead bridged by only oxide ions, and this is consistent with the known high oxophilicity of actinides. Indeed, Th2, Th2', Th3 and Th3' are each bound to six oxide ions, and Th1 and Th1' are bound to five. In f act, it is noteworthy th at there are so many oxide ions in the complex; a total of 22 oxide s bridging 16 metal atoms. This is clearly a manifestation of the high oxidation stat e of +4 of all the metal atoms in 16, and the resulting increase in the prefer ence for hard oxide ions as ligands. A closer examination of the centrosymmetric [Th6Mn10O22(OH)2]18+ core of 16, depicted in Figure 4-1 (bottom), reveals that each half can be described as two distorted Mn2Th2O4 cubane units sharing a common vertex (Th2). The two cubane s comprise Mn1, O1, Mn2, O2, Th1, O7, O3, Th2 and Th2, O14, Mn5, O10, Mn4, O15, Th3, O11, and are connected by -HOoxygen atom O8, which is the only hydroxide ion in each half of the cluster. Each of the two dicubane units in the molecule is linke d by three oxides (O12, O13, O16, and their symmetry partners) to central Mn atoms Mn3 and Mn3'. All the Mn atoms are sixcoordinate with near-octahedral geometry. In contrast, 4 Th ions (Th1, Th1', Th2, Th2') are nine-coordinate, and the remaining two (Th3, Th3') are ten-coordinate. The metal oxidation states and the protonation level of O2-, OHand H2O groups were determined by bond valence sum (BVS) calculations,85 inspection of metric parameters, and charge balance considerations. A bond valence sum (BVS) is an empirical value based on crystallographically determined metal-ligand dist ances that is routinely used to determine

PAGE 107

91 the oxidation level of an atom in a molecule. The valence of the i th atom, Vi, can be defined in terms of the sum of the individual bond valences, sij, of the atom i with those atoms, j in its coordination sphere (eq 4-4). 44b, 85 jj b R R ij iije s V] / ) [(0 (4-4) The valences of the individual bonds, sij, can be calculated from the observed bond length in the crystal structure of a molecule using eq 4-4, where Rij is the observed bond length, R0 is the length expected for a bond of unit valence and b is an experimentally determined constant equal to 0.37. Values of R0 for Mn-O bonds for Mnn+ (n = 2, 3, 4, and 7) are available, 44b, 85 allowing the application of this ca lculation to determine the oxidation states of the Mn centers in our clusters. The calculations can also be extended to include inorganic oxygen atoms and are a useful m eans of assessing the protonation levels of such atoms in a complex. In addition to tran sition/lanthanide/actinid e metal clusters, bond valence sum analysis has been used to determ ine the oxidation states of metal centers in metalloenzymes86 and superconductors.87 BVS values for O2-, OHand H2O groups are typically ~ 2.0, 1.0-1.2 and 0.2-0.4, respectively, reflecting the noninclusion of contributions fr om strong O-H bonds that are not observed with accuracy in some high-nuc learity X-ray studies. Participation in hydrogen-bonds will also affect these values. Oxygen BVS calculations are invaluable in high-nuclearity cluster chemistry, and for 16 they have unequivocally identified the protonation levels, with one exception: O13 had a BVS of 1.04, suggestive of an OHion and thus inconsistent with its expected nature as an O2ion. However, the low BVS value was explained by the observation that O 13 is involved in strong OH-O hydrogenbonding with two terminal water molecules, O6 and O9', on Th1 and Th2', respectively,

PAGE 108

92 Table 4-1. Bond valence suma calculations for the Mn ions in complex 160MeCN. 16 Atom MnII MnIII MnIV Mn(1) 4.50 4.12 4.33 Mn(2) 4.33 3.96 4.15 Mn(3) 4.37 4.00 4.20 Mn(4) 4.38 4.00 4.20 Mn(5) 4.32 3.95 4.15 a The underlined value in bold is the one closest to the actu al charge for which it was calculated. The oxidation state of a particul ar atom can be taken as the nearest whole number to the underlined value depicted in bold. (O13O6 = 2.796 and O13O9' = 2.834 ) This lengthens the corresponding MnO13 and Th-O13 bonds and leads to a lower BV S for O13. Thus, O13 is indeed an O2ion, consistent with charge balance for the complete molecule. The BVS values for the Mn ions of 16 are presented in Table 4-1. The Mn-O bond distances in 16 are in the range 1.802-2.027 consistent with the MnIV oxidation level. Similarly, the Th-O bond lengths are in the range 2.354-2.655 typical of ThIV-O bonds in the literature.88 Interestingly, a layer arrangement can be ascribed to the metal ions of complex 16, as is generally observed in minerals and ores of Mn and Th. Indeed 16 can be described as consisting of alternating layers of Th and Mn atoms, as emphasized by the layers in cyan and green in Figure 4-1 (top (right)), with a 2:3:6:3:2 pattern of metal atoms. Within this description, the Th atoms occur as two Th3 isosceles triangles (depicted in green in Figure 4-1 (top (righ t)), the Th1Th2 a nd Th2Th3 distances (3.829 ) being significantly shorter than Th1Th3 (4.808 ). The Th6Mn10 complex is the largest mixed transition metal/actinide complex to date, and the first Th/Mn cluster. Additionally, it is the largest molecular Thorium complex. Another fascinating

PAGE 109

93 observation is that 16 contains heterometallic cuba nes with high-oxidation state MnIV ions. These mixed-metal Mn containing c ubane units are potentially relevant to bioinorganic modelling (struc tural as well as functional) of the higher S states19 of the Mn4Ca site of the water oxidi zing complex of photosynthesis.22 Such Mn-containing heterometallic cubanes were unknown until seen in mixed Mn/lanthanide (Chapter III) and Mn/Ca (Chapter VI) complexes reported very recently.37, 43 4.2.2.2 X-ray crystal st ructures of [Mn7O5(OR)2(O2CPh)9(terpy)] complexes (17, 18) PovRay representation of the crystal structure of 17 (top) and its labeled core (bottom) are presented in Figure 4-2. A PovRay representation of the complete Mn7 molecule of complex 18 along with a stereopair is provi ded in Figure 4-3. Unit cell data and details of the structure refinements for 175MeCN and 18MeCN are listed in Table 4-3. Selected bond and interatomic di stances, and angles for complexes 17 and 18 are listed in Tables A-7 and A-8, respectively. Complex 17MeCN crystallizes in the triclinic space group P 1 with a crystallographic C1 symmetry. The cluster is mixed valent and comprises 1 MnIV, 5 MnIII and 1 MnII ions. The Mn ions are held together by 9 syn, syn doubly-bridging benzoates, 2 doubly bridging methoxides, 4 3-O2-, 1 5-O2and a terminal terpy. Figure 4-2 (bottom) illustrates the [Mn7O5(OMe)2(terpy)]9+ core of complex 17 which can be described as a MnIVMnIII 3O4 cubane (Mn4, Mn5, Mn6, Mn7) linked to a MnIII 3MnIIO2 butterfly (Mn2, Mn3 body Mn atoms; Mn1, Mn6 wing-tip Mn atoms) through a common 5-O2(O17) and also sharing a MnIII ion (Mn6 in this description). Within this portrayal, the cubane and butterfly motif s are also held together by two doubly bridging methoxides (O12, O18) which link the 2 MnIII body atoms of the butterfly (Mn2, Mn3) to the

PAGE 110

94 Figure 4-2. (Top) PovRay representation at th e 50% probability level of the X-ray crystal structure of 17 (with the benzoate rings omitted for clarity except for the ipso C atoms) and (bottom) its labeled [Mn7O5(OMe)2(terpy)]9+ core. Color scheme: MnIV purple; MnIII cyan; MnII yellow; N dark blue; O red; C grey. H atoms have been omitted for clarity. unshared MnIII ions of the cubane (Mn4, Mn5). Thus, the 4 MnIII ions along with the two oxos of the methoxides form two distorted Mn2O2 rhombs which share and are linked by the 5-O2(O17). Three benzoates bridge the 3 MnIII ions of the cubane with the MnIV ion of the cubane. The 3 MnIII ions of the cubane ar e further bridged by 4 PhCO2 2to the

PAGE 111

95 two body Mn ions of the butterfly. Finally, the remaining two benzoates link the two body Mn ions to the only MnII ion (Mn1), which has 3 of its coordination sites blocked by the 3 chelating N atoms of the termin al terpy. The terpy ligates to the MnII ion in a merfashion with the three N atoms lying on the meridian of a sphere. The tridentate terpy causes strain in the bonding of it's three nitrogen atom s (N1-Mn1-N3 = 143) to the Mn ion and distorts the geometry from ideal octahedral. Thus, the coordination sphere of the ligands around Mn1 is a distor ted-octahedral with the two sets of 3 N and 3 O ligands merto each other. The terpy is nearly orthogonal to the rest of the core (Figure 4-2) with the O3-Mn1-N3 angle ~116. The fused cubane / butterfly are also a bit distorted (O17O3-Mn1 = 161, O3-O17-Mn6 = 128 for the butte rfly) which is clearly a manifestation of the strain caused by the pr esence of the unusual, shared 5-O2-(O17). Complex 18MeCN crystallizes in the orthorhombic space group P bca with the asymmetric unit consisting of the whole Mn7 cluster. The structure of 18 (Figure 4-3) is very similar to 17, the only difference being the benz yl vs methyl difference in the alkoxide groups in the distorte d rhombs mentioned earlier ( vide supra ). In essence, both complexes can be described as [Mn7O5(OR)2(O2CPh)9(terpy)] (R = Me for 17 and R = CH2Ph for 18). However, the ligand induced co re distortion is less profound for 18, inasmuch the butterfly portion of 18 encompassing Mn1 is nearly linear (O12-O23-Mn7 = 179, O23-O12-Mn4 = 129 for the butterfly motif in Figure 4-3). The terpy is also more orthogonal to the rest of the core for 18 with the corresponding O23-Mn7-N3 angle in Figure 4-3 being ~104, which is smaller than the corresponding 116 value mentioned earlier for complex 17. This slight distortion in the co re angles is undoubtedly caused by

PAGE 112

96 the bulk of the phenyl groups (when compared to the methyl groups of 17) of the bridging phenyl methoxides. Figure 4-3. PovRay representation at the 50% probability level of the X-ray crystal structure of 18 (top) and its stereopair (bottom). Color scheme: MnIV purple; MnIII cyan; MnII yellow; N dark blue; O red; C grey. Solid black thicker bonds denote Jahn-Teller elongation axes. H atoms have been omitted for clarity. All the Mn ions of 17 and 18 are hexa-coordinate with near-octahedral geometry except Mn1 (the only MnII ion), which possess a severe ly distorted-octahedral coordination sphere as already stated earli er. The assignments of the metal oxidation

PAGE 113

97 states were established by bond-valence sum calculations89 (see Table 4-2), charge considerations and the observance of evid ent Jahn-Teller (JT) elongation axes for the MnIII ions. The 5 MnIII ions display JT elongation axes with the JT bonds (depicted in bold in Figure 4-3) being at least 0.1 0.3 longer than the other MnIII-O bonds, as expected for high-spin MnIII ions. Table 4-2. Bond valence suma calculations for the Mn ions in complexes 17 and 18. 17 18 Atom MnII MnIII MnIV MnII MnIII MnIV Mn(1) 2.13 2.00 2.01 4.14 3.79 3.98 Mn(2) 3.20 2.93 3.08 3.23 2.95 3.09 Mn(3) 3.20 2.92 3.07 3.27 2.99 3.14 Mn(4) 3.22 2.95 3.09 3.21 2.94 3.08 Mn(5) 3.20 2.93 3.08 3.10 2.84 2.98 Mn(6) 3.12 2.85 2.99 3.18 2.91 3.05 Mn(7) 4.24 3.88 4.07 2.12 1.99 2.00 a The underlined value in bold is the one closest to the actu al charge for which it was calculated. The oxidation state of a particul ar atom can be taken as the nearest whole number to the underlined value depicted in bold. The overall structures of 17 and 18 are very unusual and can be described as a fused butterfly/cubane or a linked triangle/cu bane. We have in the past observed the transformation of a butterfly core to a cubane core by controll ed potential electrolysis, or disproportionation triggered by carboxylate abstraction,79e, 80g,h so it is interesting to find a complex that is an amalgamation of both stru ctural types. Additionally, the fact that the complex comprises a triumvirate of Mn oxidation states is al so very rare, with previous examples including Mn3,90a Mn4,90b,c Mn13,90d Mn25,47 Mn30,91 and the reduced Mn12 complexes,92 to name a few.

PAGE 114

98 Table 4-3. Crystallographic data for 175MeCN and 18MeCN. 17 18 Formula C90H77N8O25Mn7 C102H85N8O25Mn7 fw, g/mol 2055.18 2207.34 Space group P 1 P bca a, 15.9564(13) 27.695(2) b, 16.5066(14) 22.430(2) c, 20.5637(17) 30.293(3) 110.824(2) 90 110.666(2) 90 95.156(2) 90 V, 3 4590.8(7) 18818(3) Z 2 8 T, K 173(2) 173(2) Radiation, a 0.71073 0.71073 calc, g/cm3 1.487 1.481 mm-1 1.010 1.244 R1 b,c 0.0423 0.0574 wR2 d 0.0871 0.1221 a Graphite monochromator. b I > 2 ( I ). c R 1 = 100(|| Fo| | Fc||)/| Fo|. d wR 2 = 100[[ w ( Fo 2 Fc 2)2]/ [ w ( Fo 2)2]]1/2, w = 1/[2( Fo 2) + [( ap )2 + bp ], where p = [max ( Fo 2, O) + 2 Fc 2]/3. A few other Mn7 complexes have been reported in the literature. Amongst these, are the high-spin Mn7 complexes (4MnII, 3MnIII) consisting of a Mn6 hexagon of alternating MnII and MnIII ions, surrounding a central MnII ion.93 Also, a Mn7 complex with a similar core consisting of a Mn6 hexagon, but instead now comprising 3MnII and 4MnIII ions has been reported.94 However, structurally pertinent to the Mn7 complexes reported in this chapter are: i) a Mn7 complex which can be desc ribed as two butterflies linked together by, and sharing a common Mn atom95; and ii) two Mn4 cubanes fused together via a shared Mn ion.96

PAGE 115

99 4.2.3 Magnetochemistry of Complexes 16, 17 and 18 4.2.3.1 DC studies Solid-state variable temperature magnetic susceptibility measurements were performed on vacuum-dried microc rystalline samples of complexes 16, 17 and 18, which were suspended in eicosane to prevent to rquing. The dc magnetic susceptibility ( M) data were collected in the 5.0-300 K range in a 0.1 T magnetic field and are plotted as MT vs T for complex 16 in Figure 4-4 (left), and complexes 17 and 18 in Figure 4-5. The MT value at 300 K of complex 16 is 17.54 cm3 mol-1 K, slightly lower than the 18.75 cm3 mol1 K value expected for a clus ter of ten non-interacting MnIV ions (ThIV is diamagnetic), indicating the presence of do minant antiferromagnetic exchange interactions within 16. MT decreases only slightly with decreasing temperature until ~ 25 K and then decreases more rapidly with decreasing temperature to 7.41 cm3 mol-1 K at 5.0 K. The MT value at 5 K indicates a relatively small ground state spin for the Mn10Th6 complex. To determine the spin ground state, magnetization ( M ) data were collected in the magnetic field 0.1-0.5 T and in the 1.8-10 K temperature range. The data are plotted as reduced magnetization ( M/N B) versus H/T in Figure 4-4 (right) for complex 16. The data were fit, using the program MAGNET,46 by diagonalization of the spin Hamiltonian matrix assuming only the ground state is populated, inco rporating axial anisotropy ( D z 2) and Zeeman terms, and employing a full powder average as alr eady discussed in detail in Chapter II. Although 16 is heterometallic, magnetically it ca n be treated as homometallic; the ThIV ions are diamagnetic. However, there exist low lying (relative to kT ) excited spin states with S greater than the S of the ground state. This was ev ident from the sloping nature of the dc plot at lower temperatures and the in-phase ac susceptibility data (see later)

PAGE 116

100 corroborates the same. Thus, population of the excited states will be difficult to avoid even at the lowest temperatures normally empl oyed. In addition, in the presence of a big enough applied dc field, MS components of the excited state(s) can approach in energy the lowest lying MS of the ground state and even cross below it. We thus employed the weaker fields, and the best fit (although stil l poor) of the data is shown in Figure 4-4 (right). The fit of the data represented as solid lines yielded a ground state of S = 4, D = -0.81 (16) cm-1 and g = 1.85 (2). Attempts to fit the data with an S = 3 gave unreasonable ( g > 2.5) g values. It is evident that the fit is not really satisfactory and thus an error in the D as well as S is generated. Complex 16 has isotropic MnIV ions and it is very unusual that the magnetoanisotropy term should be so high. However, the error in the value is also high. Additionally, the in-phase ac sus ceptibility measurements which do not employ a dc field are a more reliable technique48 to estimate the S in such situations and hence they were performed on 16 (see later). Temperature (K) 050100150200250300 M T (cm 3 mol -1 K) 0 5 10 15 20 H/T (kG/K) 0.00.51.01.52.02.53.0 M/N B 0 1 2 3 4 0.1 T 0.2 T 0.3 T 0.4 T 0.5 T Fitting Figure 4-4. (Left) Plot of MT vs T for complex 16. (Right) Magnetization (M) vs field (H) and temperature (T) data, pl otted as reduced magnetization ( M/N B) vs H/T for 16 at applied fields of 0.10. 5 T range and in the 1.8 10 K temperature range. The solid lines are the f it of the data; see the text for the fit parameters. The results of the dc magnetic su sceptibility studies on complexes 17 and 18 are depicted in Figure 4-5. For 17, the MT value of 16.91 cm3mol-1K at 300 K decreases

PAGE 117

101 slightly to 15.21 at 50 K and then it st eadily increases marginally with decreasing temperature to reach 17.83 cm3mol-1K at 5.0 K indicating an isolated large ground-state spin for 17. The MT value of 18 follows a similar pattern as can be seen in Figure 4-5 and has a value of 16.66 cm3mol-1K at 300 K which decreases slightly to 15.46 at 50 K. Then the MT value increases fractionally with d ecreasing temperature to reach 17.48 cm3mol-1K at 5.0 K indicating a ground-state spin of 18 which is very similar to that of 17. The spin-only value of 1 MnIV, 5 MnIII, and 1 MnII non-interacting ions is 21.25 cm3mol-1K, assuming g = 2. Thus, for both 17 and 18 the MT value at 300 K is lesser than the non-interacting sp in-only value of 21.25 cm3mol-1K and remains more or less constant with decreasing te mperature indicating dominant antiferromagnetic exchange interactions within the Mn7 complexes, and signifying a very well isolated ground state spin for both of them since the MT value does not change much with temperature. Temperature (K) 050100150200250300 M T (cm 3 mol -1 K) 10 12 14 16 18 20 22 24 Complex 17 Complex 18 Figure 4-5. Plots of MT vs T for complexes 17 and 18. The nuclearity, low symmetry and large num ber of exchange interactions of the molecules (eight under CS virtual symmetry) makes a ma trix diagonalization method to evaluate the various Mn2 pairwise exchange parameters ( Jij) onerous, and also rules out

PAGE 118

102 application of the equivalent operator a pproach based on the Kambe vector coupling method.45 However, a qualitative rationalization of the individual spin alignments and signs of the exchange interactions ( Jij) would be possible on ce the ground state was identified ( vide infra ). Therefore, we focused on identifying the ground state S the ZFS term D values, and the electronic g factors, for complexes 17 and 18. Hence, magnetization ( M ) data were collected in the magnetic field and temperature ranges 0.1-4 T and 1.8-10 K in order to determin e the spin ground states of the Mn7 complexes. The data are plotted as reduced magnetization ( M/N B) versus H/T in Figure 4-6 (top) for complexes 17 (left) and 18 (right). The data were fit usi ng the same procedure applied to complex 16. For 17, the fit (solid lines in Fi g. 4-6 (top (left))) gave S = 6, D = -0.22 (0) cm-1 and g = 1.88 (0). When fields higher than 4 T were employed in the fitting the same S value was obtained; it can be clearly seen in Fig. 4-6 (top (left )) that the magnetization saturates at the applied field of 4 T. Thus dominant antiferromagnetic couplings within complex 1 leads to a spin ground state of S = 6 which is consistent with the dc and inphase ac magnetic susceptibility data (see later). For complex 18, attempts were made to fit the magnetization data collected in the 0.1 3 T and 1.8-10 K temperature ranges. Beyond 3 T, the magnetization saturates and portions of the isofield lines at fi elds > 3 T start superimposing with M/N B values of lower fields. Thus, a satisfactory fit could be obtained only when data collected in the 0.1-3 T applied field range was used. The fit of the data, represented as solid lines in Fig. 4-6 (top (right)) yiel ded a ground state of S = 6, D = -0.18 (0) cm-1 and g = 1.86 (1). The ground states obtained for complexes 17 and 18 are in agreement with the dc magnetic susceptibility data. Also, the g values obtained were smaller than 2.00, as

PAGE 119

103 expected for Mn. In general for these magne tization fits, for a system occupying only the ground state and experiencing no zero-field sp litting (ZFS), the various isofield lines would be superimposed and M/N B would saturate at a value of gS .97 The nonH/T (kG/K) 05101520 M/N B 0 2 4 6 8 10 12 0.1 T 0.5 T 1 T 2 T 3 T 4 T Fitting H/T [kG/K] 051015 M/N B 0 2 4 6 8 10 12 0.1 T 0.5 T 1 T 2 T 3 T fitting g 1.801.821.841.861.881.901.921.941.961.982.00 D (cm-1) -0.30 -0.28 -0.26 -0.24 -0.22 -0.20 -0.18 -0.16 g 1.821.841.861.881.901.92 D (cm -1 ) -0.24 -0.22 -0.20 -0.18 -0.16 -0.14 Figure 4-6. (Top) Magnetization (M) vs field (H) and temperature (T) data, plotted as reduced magnetization ( M/N B) vs H/T for (left) complex 17 at applied fields of 0.1 4 T range and in the 1.8 10 K temperature range, and (right) for complex 18 at applied fields of 0.13 T range and in the 1.8 10 K temperature range. The solid lines are the f it of the data; see the text for the fit parameters. (Bottom) Two-dimensional contour plot of the error surface for the D vs g fit for complexes 17 (left), and 18 (right). superimposition of the isofield lines in Figur e 4-6 (top) clearly indi cates the presence of ZFS, albeit, small values of D were obtained for both the complexes. The small D values

PAGE 120

104 are nevertheless consistent w ith the fact that both complexes have some JT axes perpendicular (especially in the cubane) as can be seen in Figure 4-3. However, we wanted to confirm that the obtained parameters were the true global rather than local error minima. Thus for this reason and also to assess the uncert ainty in the obtained g and D values, root-mean square D vs g error surfaces for the fits were generated using the program GRID.98 The error surface for 17 is shown in Figure 4-6 (bottom (left)) as a twodimensional contour plot for the D = -0.15 to -0.30 cm-1 and g = 1.8 to 2.0 ranges. Only one minimum is observed in this range, a nd this is a fairly well-defined minimum, indicating a fairly small level of uncertainty in the best-fit parameters. The two lowest error contour lines cover a range of D ~ -0.20 to -0.23 cm-1 and g ~ 1.87 to 1.89, giving fit values and their uncertainties of D = -0.22 (1) cm-1 and g = 1.88 (1). Similarly for complex 18, although in this case the minimum is somewhat shallower and there is a correspondingly slightly gr eater level of uncertainly giving D = -0.18 (2) cm-1 and g = 1.86 (2). Note, however, that the error surf ace indicates the precision of the fit minima, not the accuracy of the obtained D and g pa rameters; bulk magnetization data are not in general the most accurate way to obtain these, more sensitive techniques such as EPR being superior for this purpose. Although 17 and 18 possess a complicated, low symmetry Mn7 core, their S = 6 ground states can nevertheless be rationalized in a very sa tisfying manner (Figure 4-7). This is achieved using the known spin coupling pattern in the discrete Mn4 complexes mentioned earlier that possess the same MnIV, 3MnIII cubane core found within 17 and 18. These Mn4 complexes of C3V symmetry exhibit ferromagnetic and antiferromagnetic exchange interactions within the MnIIIMnIII and MnIVMnIII pairs, respectively.79 As a

PAGE 121

105 result the ground state is S = 9/2, due to the three MnIII ( S = 2) spins being parallel to each other and antiparallel to the MnIV ( S = 3/2) spin. Assuming (very reasonably) that the same situation exists for this cubane unit within the larger Mn7 core of 17 and 18, the question then becomes how this S = 9/2 unit couples to the ot her Mn atoms. Inspection of Figures 4-2 shows that the primary exch ange pathways between the cubane MnIII atoms and the butterfly body MnIII atoms (Mn2 and Mn3) are via the 5-O2ion (O17) and the -MeOgroups (O12 and O18). Remembering that (i) the MnIII JT axes define the local Mn z axes, and (ii) the cubane MnIII JT axes intersect at O17 whereas those for body MnIII atoms include neither O17 nor O12/O18 (s ee Figure 4-3 for JT axes), then Mn-OMn d-p-d exchange pathways through O17 and O12/O18 atoms will involve a singly occupied dz 2 orbital on cubane MnIII atoms and an empty dx 2 -y 2 on body MnIII atoms. These pathways would thus be expected to give ferromagnetic contributions to the total exchange between these metals Notwithstanding Mn-O-Mn d-p-d exchange pathways involving singly-occupied Mn d orbitals that would be expected to give antiferromagnetic contributions, it is concl uded that there are overall ferromagnetic exchange interactions between the cubane and butterfly body MnIII atoms in 17 and 18. The same was observed for the ferromagnetically-coupled [Mn6O4X4(dbm)6] (X = Cl, Br) complexes with an S = 12 ground state, whose ferromagnetic coupling between MnIII atoms was rationalized in the same fashion as for 17 and 18.97a,b Finally, it is expected that the exchange interacti ons between the butterfly body MnIII atoms and the MnII atom will be antiferromagnetic, as they are in the discrete Mn4 butterfly complexes,80 and so the ground states of 17 and 18 are concluded to comprise all the MnIII spins being parallel to each other and antiparallel to the MnII and MnIV spins, giving a predicted ground state

PAGE 122

106 Figure 4-7. Schematic representation of the Mn spin alignments giving the S = 6 ground state for the Mn7 complexes, based on a fused cubane/butterfly arrangement. of S = 10 5/2 3/2 = 6, as observed experiment ally. This is summarized in Figure 4-7. Hence, it was very satisfying that the spin obtained from the above-stated rationalization was in perfect agreement with the S value obtained from magne tization fits for both the complexes. Of course, the rationale given a bove is just a hypothesi s and there are other ways of justifying the ground state spin. A mo re quantitative justif ication of the spinalignments by DFT (Density Functional Theory) calculations34a, 99 is needed for confirmation of the ground state spin. Although the D values of both the complexes are small, the combination of this D with the S = 6 may provide a big enough barrier for magnetization reversal for these complexes. Thus, they might display slow magnetization relaxation characteristic of

PAGE 123

107 single-molecule magnets (SMMs) resulting from a barrier, the upper limit of which is given by U = S2| D | for integer spins.9 Indeed, complexes 17 and 18 possess an energy barrier ( U ) to magnetization reversal, which computes to 7.92 cm-1 for 17 and 6.48 cm-1 for 18, respectively. However, the effective barrier Ueff, might be a little bit smaller because of quantum tunneling through the barrie r. Hence, ac susceptibility measurements were performed to investigate whether these Mn7 complexes functioned as SMMs. 4.2.3.2 AC studies In an ac magnetic susceptibility experiment, a weak field (typically in the 1 5 G range) oscillating at a particular ac frequency ( ) is applied to a sample to investigate the dynamics of its magnetization relaxation. If th e magnetization vector of the molecule can reorient at the frequency of the oscillating ac field, then there is no out-of-phase ( M ) susceptibility signal, an d the in-phase signal ( M) is equal to the dc magnetic susceptibility. However, if the magnetization vector of the molecule cannot keep up with the oscillating field, then an out -of-phase signal is observed. The M signal is dependent on the frequency of the oscillating ac field, i. e., faster the oscillati on of the ac field, the higher the temperatur es at which the M signal is observed, and is accompanied by a frequency-dependent decrease in the in-phase signal (as M T ). Such signals are a characteristic signature of the superpar amagnet-like properties of a SMM, however, should not be taken as sufficient proof that a molecule behaves as a SMM; intermolecular interactions and phonon bottlenecks have been s hown to also lead to such signals. Thus, ac magnetic susceptibility data were collected on dried, microcrystalline samples of 16, 17 and 18 in the temperature range 1.8-10 K in a zero dc field and a 3.5 G ac field oscillating at several frequencies in the 25997 Hz range. The in-phase data, plotted as M T versus T is depicted in Figure 4-8 for complex 16 (left), and complexes 17 and 18

PAGE 124

108 (right) at an applied frequency of 500 Hz. There was no out-of-phase signal observed for any of the complexes 16-18, at least till 1.8 K, the ope rating minimum of our MPMS SQUID instrument. As already stated, for Mn10Th6 the S obtained from magne tization fits is inconsistent with the S derived from the in-phase da ta. This anomaly is undoubtedly because the quality of the ma gnetization fit is really poor Additionally, the 5.0 K dc magnetic susceptibility value for complex 16 of 7.41 cm3 mol-1 K may be compared to the spin-only (g = 2) values of 3.0 and 6.0 cm3mol-1K expected for S = 2 and 3 states, respectively. The M T value at 0 K where only the ground state would be populated, for an S = 4 and g = 1.85 (values obtained from magn etization fits) system is 8.56 cm3mol1K. The in-phase M T (Figure 4-8 (left )) value of complex 16 of 14.30 cm3mol-1K at 10 K decreases sharply to reach 6.80 cm3mol-1K at 1.8 K. The drastic sloping nature of the plot indicates the presence of low lying (re lative to kT) excited spin states with S greater than the S of the ground state. Thus, with lowering of temperature the M T value decreases steeply as depopulation of excited states with greater S value than the ground Temperature (K) 024681012 M T (cm 3 mol -1 K) 0 5 10 15 20 Temperature (K) 0246810 M T (cm 3 mol -1 K) 10 12 14 16 18 20 22 24 Complex 17 Complex 18 Figure 4-8. Plots of the in-phase (as M T ) ac susceptibility signals vs T for complex 16 (left), and complexes 17 and 18 (right). The measurement is in a 3.5 G ac field oscillating at a 500 Hz frequency.

PAGE 125

109 state is occurring. Unfortunate ly, the latter are commonly en countered in high nuclearity clusters; weak exchange inter actions between the constituent ions and/or spin frustration effects within the molecule give rise to a hi gh density of spin states Extrapolation to 0 K of the in-phase signal of 16 (Figure 4-8 (left)), where only the ground state would be populated, gives a value of ~ 5.2 cm3mol-1K indicative of an S = 3 ground state and g < 2, as expected for Mn. We are more confident w ith the in-phase ac susceptibility data as it avoids the complications resulting from the a pplication of a dc field and/or the population of low-lying excited spin-states other than the ground state, and thus conclude that complex 16 has a ground state spin S = 3. In principle, in a zero dc field the in-phase ac data shoul d be in agreement with the dc data in the same temperature range if there is population of only the ground state and the complex has no significant intermolecular interactions. Thus, the M T value extrapolated to zero Kelvin where only the ground state would be populated should be in agreement with the formula M T = ( g2/8) S ( S +1). This provides a reliable approach Table 4-4. Comparison of the magnetic susceptibility parameters of complexes 17 and 18. Complex S D (cm-1) g U (cm-1) M T valuea (cm3mol-1K) M T valueb (cm3mol-1K) 17 6 -0.22 1.88 7.92 18.56 18.64 18 6 -0.18 1.86 6.48 18.16 19.45 a The M T value computed at 0 K, using S and g values obtained from magnetization fits and applying the formula M T = ( g2/8) S ( S +1). b The experimental M T value at 1.8 K, obtained from the in-phase ac susceptibility plots. towards determining the ground state spin fo r polynuclear homometallic complexes as it avoids the complications resulting from the a pplication of a dc field and/or the population

PAGE 126

110 of low-lying excited spin-states other than the ground state. We have in the past used this technique to confirm and/or determine the ground-state spin of our complexes.48 The M T value of complex 17 is 17.79 cm3mol-1K at 10 K and increases steadily to 18.64 cm3mol-1K at 1.8 K (Fig. 4-8 (right)) and seem s to be dropping slightly to a lower value below 1.8 K. For 18 the M T value (Fig. 4-8 (right)) of 17.32 cm3mol-1K at 10 K increases steadily to 19.71 cm3mol-1K at 2.2 K. Then it drops fractionally to reach 19.45 cm3mol-1K at 1.8 K and seems to be falling to a lower value at 0 K. The expected M T at 0 K for an S = 6 and g = 1.88 system (see magnetizati on fits earlier) like complex 17 is 18.56 cm3mol-1K. Similarly, for complex 18 one expects a value of 18.16 cm3mol-1K at 0 K for the M T component. Thus, the M T values for complexes 17 and 18 are satisfyingly consistent with the S and g values obtained from the magnetization fits ( vide supra ). The non-sloping nature of the dc and ac in-phase magnetic su sceptibility data clearly indicate that both the complexes have a very well-isolated S = 6 ground state spin and there is minimal population of othe r excited spin-states, if any. The S D g U and M T values (expected at 0 K from S and g values from magnetizati on fits and obtained at 1.8 K from in-phase ac data) are di splayed in Table 4-4 for complexes 17 and 18 for comparison. The plots of the out-of-phase signal ( M ) versus T did not display any frequencydependant signals for complexes 17 or 18. However, no tails of peaks were seen until the operating minimum of 1.8 K of our SQUID in strument. Since, both complexes possess a barrier to magnetization relaxation of ~ 7 cm-1 there was always a possibility that the frequency-dependant ac signals could perhaps be present at temperatures below 1.8 K.

PAGE 127

111 Thus, to confirm this and to investigate th e system further we decided to perform low temperature (< 1 K) hysteresis studies on complex 18, as well as the Mn10Th6 complex. 4.2.3.3 Hysteresis studies below 1.8 K Hysteresis loops are the definitive property of a magnet and can provide unquestionable proof that a molecule behaves as a single-molecule magnet. The absence of an out-of-phase signal, at least ab ove 1.8K (the operating minimum of our magnetometer) suggests that the Mn10Th6 complex is not a SMM, but this is to be expected for a complex with a small molecular spin and containing only MnIV ions; the latter is a fairly isotropic ion and the cluster will therefore have at best only a very small magnetoanisotropy ( i.e. zero-field splitting parameter, D ); there are currently no SMMs containing only MnIV ions. Thus, to confirm the conclusion from the ac data, magnetization vs dc field sweeps on a single crystal of 16MeCN were carried out at temperatures down to 0.04 K, and the results are shown in Fig. 4-9(left). Even at the lowest temperature, no hysteresis was obser ved, confirming the compound not to be a SMM. Since the M T showed no (significant) decrease down to 1.8 K, the complexes 17 and 18 clearly did not exhibit th e slow magnetization relaxation that is suggestive of a SMM. This was supported by the absence of an out-of-phase ac susceptibility signal ( M ) for the frequency range examined. To explore whether slow relaxation might nevertheless be manifested at even lower te mperatures, magnetization vs dc field scans were carried out on a single crystal of 18MeCN using a micro-SQUID apparatus.50 The observation of hysteresis loops in such studi es represents the diagnostic property of a magnet, including SMMs and superparamagne ts below their blocking temperature ( TB).

PAGE 128

112 In particular, the hysteresis loops of SMMs e xhibit increasing coercivities with increasing sweep-rates and with decreasing temperatures. The magnetization vs dc field sweeps at a fixed sweep-rate of 0.14 T/s and at temperatures of 0.7 and 0.1 K are presented in Figure 4-9(right). The 0.7 K scan displays an apparently greater coercivity than the 0.1 K scan, but that must be due to the former being a result of a phonon bottleneck rather than a significant barrier. The 0.1 K does show a very small amount of hysteresis, evid ent at non-zero field positions, but this is truly very small. We conclude from this that in spite of a predicted barrier U of ~7 cm-1 -1 -0.5 0 0.5 1 -1.2-0.8-0.400.40.81.2 0.04 K 0.5 K 1 K 2 K 4 K 7 K M/Ms 0H (T) 0.035 T/s -1 -0.5 0 0.5 1 -1-0.500.51 M/Ms 0H (T) 0.14 T/s 0.7 K 0.1 K Figure 4-9. (Left) Magnetization ( M ) vs applied magnetic field ( H ) hysteresis loops for single-crystals of 1610MeCN at a fixed sweep rate of 0.035 T/s and at the indicated temperatures. (Ri ght) Hysteresis loops at te mperatures of 0.7 and 0.1 K and at a 0.14 T/s sweep rate for single crystals of 185MeCN. In both plots M is normalized to its saturation value, Ms. calculated from S2| D |, it is clear that the true barrier Ueff is much smaller due to fast QTM. The latter is responsible fo r the large step (magnetizati on change) at zero field that effectively relaxes almost all the magnetization, and allows only a very small hysteresis at non-zero field values. Large QTM rates woul d be consistent with the low symmetry of the molecule, which would result in a signifi cant transverse anis otropy (rhombic ZFS parameter, E ). The latter will result in significant mi xing of states on either side of the

PAGE 129

113 anisotropy barrier, and thus significant QT M rates. We conclude that complexes 17 and 18 are not SMMs. 4.3 Conclusions In summary, the first polymetallic cluster containing manganese and thorium atoms has been synthesized by the reaction of a preformed MnIII 4 cluster with ThIV nitrate. The fact that the resulting ThIV 6MnIV 10 product contains only MnIV is noteworthy, and is reminiscent of the CeIV xMnIV y products reported recently.100 The hard ThIV ion is known to be very oxophilic and favor the incorporation of similarly hard O2ions, and the latter in turn will favour the formation of hard MnIV ions from MnIII, presumably facilitated by aerial oxidation. As for the CeIV xMnIV y complexes, which were obtained from the oxidation of MnII sources with CeIV, we believe the hi gh oxidation state MnIV ions are stabilized by the rich oxide environment, which in turn is facilitated by the CeIV ions. The Th6Mn10 complex is the largest mixed transition metal/actinide complex to date and augurs well that it is merely the prototype of a rich new area of mixed 3d/5f cluster chemistry. The next largest member of this family is a Cu2Mo12U polyoxometallate.101 Indeed, polyoxometallates are relatively common in actinide chemistry, but in contrast there are only very few structurally char acterized mixed-metal carboxylate complexes containing actinides, and these are small nucl earity species. Finally, the high oxidative strength of the Mn10Th6 complex, complimented by its solubility in most organic solvents, makes it a very attractive candidate for catalytic and/or noncatalytic oxidation processes. In this chapter, we have also detail ed the synthesis, st ructure and magnetic properties of two novel Mn7 complexes. Thus, the use of terpy in reactions with certain Mn3 and Mn4 species has led to an interesting ne w structural type in Mn carboxylate

PAGE 130

114 cluster chemistry that can be described as a hybr id of two previously observed cores. It is also a rare example of three different Mn oxi dation states in the same unit. The obtained complexes possess S = 6 ground states, and it is satisfyi ng that this can be rationalized by qualitative considerations of the expected Mn2 pairwise couplings. The low anisotropy, however, as reflected in the small D values, as well as the low C1 (virtual CS) symmetry, means that these complexes are not new additions to the family of SMMs. Nevertheless, they are interesting new additions to the family of Mnx clusters. The preparation of 17 and 18 are obviously dependent on the presence of terpy in the reaction, but only one terpy is incorporated into the structure, and then only on the MnII atom on the periphery of the molecule and not apparently of much importance to the topology of the remaining Mn6 portion of the structure. In fact, with hindsight, there seems no reason why this type of structure could not have b een previously encountered with other chelates such as bpy and a monodentate ligand (H2O, Cl, etc) in place of te rpy. Of course, this merely emphasizes how complicated and unpredic table are the precise nuclearities and topologies of products of such labile a nd complicated multi-component reactions under thermodynamic control. 4.4 Experimental 4.4.1 Syntheses All manipulations were performed under aerobic conditions us ing chemicals as received, unless otherwise stated. [Mn3O(O2CPh)6(py)2(H2O)](2),54 and (NBun 4)[Mn4O2(O2CPh)9(H2O)] (3),55 were prepared as previously reported. [Mn10Th6O22(OH)2(O2CPh)16(NO3)2(H2O)8]0MeCN (16). To a slurry of 3 (0.25g, 0.16mmol) in MeCN/MeOH (20ml/1ml) was added Th(NO3)4H2O (0.17g, 0.31mmol) and left under magnetic stirrer for 30 min. The resulting brown solution was

PAGE 131

115 filtered and the filtrate concentrated by slow evaporation to yield reddish crystals of 16 in about a month. These crystals were washed with acetone and dried overnight in vacuo. The dried crystalline material identified as fully desolvated and was obtained in 20% isolated yield. Anal. Calc (Found) for 16: C112H98O70N2Mn10Th6: C, 29.67 (29.52); H, 2.18 (2.15); N, 0.62 (0.48) %. Selected IR data (KBr, cm-1): 3392(s, br), 1598(m), 1522(s, br), 1448(w), 1384(s, br), 1025(w), 718(s) 685(m), 615(s, br), 571(w), 482(w). [Mn7O5(OMe)2(O2CPh)9(terpy)]MeCN (17). Method A. To a slurry of 2 (0.50g, 0.45mmol) in MeCN/MeOH (20ml/2ml) was added solid 2,2 :6 ,2 -terpyridine (in short, terpy) (0.05g, 0.23mmol) and left under magnetic stirrer for 20 min. The resulting reddish-brown solution was filtere d and the filtrate concentrated by slow evaporation to yield brown crys tals in a couple of days. These were washed with acetone and dried in vacuo. The dried crys talline material identified as 175MeCN and was obtained in 60% isolated yiel d. Anal. Calc (Found) for 17MeCN: C90H77O25N8Mn7: C, 52.60 (52.55); H, 3.78 (3.72); N, 5.45 (5.40) %. Selected IR data (KBr, cm-1): 3432(s, br), 1599(s), 1563(s), 1449(w), 1385(s), 1025(w), 772(w), 717(s), 681(m), 621(m), 542(w), 473(w). Method B. In MeCN/MeOH (20ml/2ml) was dissolved complex 3 (0.50g, 0.31mmol) and terpy (0.03g, 0.12mmol) and left stirring for 20 min. The resulting solution was filtered and the filtrate slowly concentrated by evaporation to yield brown crystals which were washed with acetone a nd dried in vacuo. The product was identified as 17 by IR and elemental analysis. Anal. Calc (Found) for 17MeCN: C90H77O25N8Mn7: C, 52.60 (52.62); H, 3.78 (3.82); N, 5.45 (5 .49) %. The yield was 45% based on Mn.

PAGE 132

116 [Mn7O5(OCH2Ph)2(O2CPh)9(terpy)]MeCN (18). To a slurry of 2 (0.25g, 0.23mmol) in MeCN/PhCH2OH (20ml/1ml) was added terpy (0.02g, 0.07mmol) and left under magnetic stirrer for 30 min. The resu lting brown solution was filtered and the filtrate concentrated by slow evaporation to yield black crystals of 2 in about a week. These crystals were washed with acetone and dried overnight in vacuo. The dried crystalline material identified as fully desolv ated and was obtained in 50% isolated yield. The synthesis could also be repeated by employing complex 3 as the starting material and reacting it with terpy in a 23:7 molar ratio. The product from this reaction was identified as 18 by IR and elemental analysis. Anal. Calc (Found) for 18: C92H70O25N3Mn7: C, 55.19 (54.95); H, 3.52 (3.24); N, 2.10 (2.27) %. Selected IR data (KBr, cm-1): 3440(s, br), 1599(s), 1562(s), 1450(w), 1370(s), 1024(w), 771(w), 717(s), 683(m), 628(m), 541(w), 472(w). 4.4.2 X-ray Crystallography Data were collected on a Siemens SM ART PLATFORM equipped with a CCD area detector and a graphite monochromator utilizing Mo-K radiation ( = 0.71073 ). Suitable crystals of the complexes were attach ed to glass fibers us ing silicone grease and transferred to a goniostat where they were cool ed to 173 K for data collection. An initial search of reciprocal space revealed a triclinic cell for 16 and 17, and an orthorhombic cell for 18; the choice of space groups P P and P bca respectively, were confirmed by the subsequent solution and refinement of the st ructures. Cell parameters were refined using up to 8192 reflections. A full sphere of data (1850 frames) was collected using the -scan method (0.3 frame width). The first 50 frames were re-measured at the end of data collection to monitor instrument and crysta l stability (maximum correction on I was < 1 %). Absorption corrections by integration were applied based on measured indexed

PAGE 133

117 crystal faces. The structures were solved by direct methods in SHELXTL6,56a and refined on F2 using full-matrix least squares. The non-H atoms were treated anisotropically, whereas the hydrogen atoms were placed in ca lculated, ideal positions and refined as riding on their respective carbon atoms. For 16MeCN, the asymmetric unit consists of half the cluster and five MeCN molecules as crystallization solvents. The so lvent molecules were disordered and could not be modeled properly, thus program SQUEEZE,56b a part of the PLATON56c package of crystallographic software, was used to calcu late the solvent disorder area and remove its contribution to the overall intensity data. A total of 791 parameters were included in the structure refinement on F2 using 16874 reflections with I > 2 (I) to yield R1 and wR2 of 7.07% and 16.47%, respectively. For 17MeCN, the asymmetric unit consists of two Mn7 clusters located on an inversion center, and five acetonitrile molecules as solvents of crysta llization. A total of 1176 parameters were included in the structure refinement on F2 using 29290 reflections with I > 2 (I) to yield R1 and wR2 of 4.23% and 8.71%, respectively. For 18MeCN, the asymmetric unit consists of the Mn7 cluster and five molecules of acetonitrile. The solvent molecules were disordered and could not be modeled properly, thus program SQUEEZE,56b a part of the PLATON56c package of crystallographic software, was used to calcula te the solvent disorder area and remove its contribution to the overall intensity data. A to tal of 1144 parameters were included in the structure refinement on F2 using 21072 reflections with I > 2 (I) to yield R1 and wR2 of 5.74% and 12.21%, respectively.

PAGE 134

118 CHAPTER 5 THE FIRST STRONTIUM-MANGANE SE CLUSTER: SINGLE-MOLECULE MAGNETISM AND Sr-EXAFS COMPARISON WITH THE WATER OXIDIZING COMPLEX OF PHOTOSYSTEM II 5.1 Introduction Manganese carboxylate cluster chemistry ha s assumed much importance in recent years owing mainly to its relevance to bi oinorganic chemistry and nanoscale magnetic materials. The latter area primarily involve s the synthesis of single-molecule magnets (SMMs), which are molecules that reta in their magnetization below a blocking temperature in the absence of an applied field.7, 9 For a molecule to function as an SMM, it should possess a large ground state spin ( S ) and a negative (easy axis) magnetoanisotropy ( D ). Several classes of SMMs are now known,102 most of them containing MnIII, but there is a continuing need to develop methods to modify known structural types and synthesize new ones in order to improve our understanding of this interesting nanomagnetic phenomenon. In th e bioinorganic field, the water oxidizing complex (WOC) in Photosystem II (PS II) catalyses the oxidation of H2O to O2 gas in green plants and cyanobacteria.18 This four-electron process involves various oxidation levels of the WOC (the so-called Sn states, n = 0 to 4),19, 20 and is the source of essentially all the O2 on this planet. Amongst these various states, S1 is the most stable, well-studied and understood one; various crystallographic st udies have also been performed on this state. The WOC has been studied for many y ears by a variety of spectroscopic and physicochemical techniques, and has long been known to comprise a tetranuclear, oxide-

PAGE 135

119 bridged Mn4 cluster containing primarily MnIII and MnIV ions and mainly carboxylate peripheral ligation. It has also b een known that the WOC requires Ca2+ for activity,103 and calcium EXAFS (extended X-ray absorption fi ne structure) studies of the WOC have revealed a MnCa separation of ~3.4 .104 Many groups have studied the Ca2+ binding site of the WOC by substitution with Sr2+, motivated by the genera l lack of spectroscopic features of Ca2+.105, 106 Additionally, there are experimental complications involved in Ca EXAFS studies mainly arising due its abundance in nature and the re sulting difficulty to exclude extraneous Ca, from PS II samples. Strontium is the only metal that can substitute for the Ca2+ in active PS II samples with major retention of activity (~ 40%),105 and a corresponding MnSr separation of ~ 3.5 has been detected in Sr EXAFS studies of Sr-substituted preparations.106 In addition, recent crystallographic studies on the PS II reaction center of the cyanobacterium Thermosynechococcus elongatus at 3.5 22 and 3.0 26 resolutions have identified Ca as being an integral part of the Mn complex using anomalous diffraction data. Alth ough there is still an uncertainty about the CaMn4 structure obtained from crystallogra phy due to radiation damage during X-ray data collection,23, 26 there is little doubt that the WOC is indeed a heterometallic [Mn4CaOx] cluster.22, 26 The availability of synthetic Ca/Mn (and Sr/Mn) complexes to act as synthetic models of the WOC would re present an important step forward in understanding the magnetic and spectroscopic pr operties of the native site and the mechanism of its function. However, in contrast to the gr owing knowledge of the WOC, as described above, there is essentially nothing in the in organic chemistry literature of mixed-metal Ca/Mn or Sr/Mn compounds. In fact, only rece ntly was the first example of a molecular

PAGE 136

120 Ca/Mn cluster reported.43 Similarly, there are no molecular Sr/Mn clusters currently known, only some polymeric SrII/MnII species.107 Therefore, as part of our continuing interest in obtaining synthetic models of the WOC and its various modified forms, we have been investigating mixed Sr/Mn chemistr y and have successfully obtained the first such molecular species, a SrMn14 complex. Although the nuclea rity is much higher than that in the WOC, it is an important step fo rward and has also allowed Sr EXAFS data to be obtained for the first time on a structurally well-characte rized synthetic complex that can act as a benchmark for comparison w ith similar data on the WOC. We also demonstrate that this new Sr/Mn complex is a SMM with a high blocking temperature arising primarily from a hi gh magnetoanisotropy component. 5.2 Results and Discussion 5.2.1 Syntheses We recently began seeking synthetic rout es to Ca/Mn, Sr/Mn and Ln/Mn (Ln = lanthanide) heterometallic complexes, and initia l results from this effort have been very satisfying.37, 43 Our synthetic strategy has been to employ preformed manganese clusters such as [Mn3O(O2CR)6(py)3] (R = Me, But, Ph) and (NBun 4)[Mn4O2(O2CPh)9(H2O)] and treat them with a heterometal carboxylate or other salt. Particularly attractive as a starting material is (NBun 4)[Mn4O2(O2CPh)9(H2O)], which contains 4 MnIII ions and only carboxylate and water ligands around the Mn/O core.55 It is thus an attractive (and easily prepared) reagent that can potentially yield MnIII(and even MnIV-) containing products. We have thus been carrying out a thorough i nvestigation of the reactivity of this Mn4 complex with various metal(II) salts, seeking to access new mixed-metal species. As part of the above inves tigation, the following procedur e with Sr was developed, which is similar to those used prev iously for Ca/Mn and Ln/Mn products. 37, 39, 43

PAGE 137

121 Treatment of (NBun 4)[MnIII 4O2(O2CPh)9(H2O)] with 1 equiv. of Sr(ClO4)2xH2O in MeCN/MeOH (10:1 v/v) resulted in the subsequent isolation of [SrMn14O11(OMe)3(O2CPh)18(MeCN)2]12MeCN (1912MeCN) in 55% yield (based on Mn). The Mn4:Sr ratio of 1:1 was obviously directed at a Mn4Sr product as would be present in the WOC, and the small amount of MeOH was necessary to ensure solubility of Sr(ClO4)2. However, the absence of any chela ting agents in the reaction solution allowed any smaller nuclearity intermediate s to aggregate further and yield the Mn14Sr product 19; reactions with added chelates such as bipyridine, terpyridine and 8hydroxyquinoline are under investiga tion but have yet to provid e pure, isolable products. The reaction mixture is almost certainly a complicated mixture involving fragmentation, aggregation and redox processes, with severa l species likely in equilibrium, and the attainment of complex 19 in good yield (55%) and purity is undoubtedly facilitated by its crystallization directly from the reaction mixture, driving an y equilibria to this product. The presence of MeOH as a solvent or co-sol vent is a recurring theme in much of our recent work,48, 108 and not only ensures solubi lity of all reactants, as already stated, but also serves to (potentially) provide terminal and/or br idging ligands. Indeed, complex 19 contains three bridging methoxide groups ( vide infra ). The present work is thus the most recent member of a growing class of methoxide-bridged mixed-metal species, 37, 39, 43 although it should be added that 19 is the first mixed Sr/Mn cluster of any type and with any ligation set. 5.2.2 Description of Structures A PovRay representation of 19 and its stereopair are pres ented in Figure 5-1. The labeled core is shown in Fi gure 5-2. Crystallographic data and structure refinement

PAGE 138

122 details are collected in Table 5-1, and selected interatomic distances are listed in Table 52. Complex 19MeCN crystallizes in the triclinic space group P1 with the SrMn14 molecule in a general position. The structure consists of a [SrMn14O11(OMe)3]18+ core whose Mn ions are mixed-valent (MnIII 13, MnII) and bridged by six 4-O2-, five 3-O2-, Figure 5-1. (Top) PovRay representation at th e 50% probability level of the X-ray crystal structure of 19 (with the benzoate rings omitted for clarity, except for the ipso C atoms) and (bottom) its stereopair. Color scheme: MnIII light blue, MnII yellow, Sr green, O red, N dark blue, C grey. H atoms have been omitted for clarity. two 3-MeO-, and one 2-MeOions (Figure 5-2). The meta l oxidation states and the protonation levels of O2and MeOions were established by bond-valence sum

PAGE 139

123 calculations calculations,109, 89 charge considerations, inspecti on of metric parameters and the identification of MnIII Jahn-Teller (JT) elongation axes The core consists of two [Mn4O3(OMe)] cubanes attached, via oxide bridges, on either side of a central, near linear and planar [Mn3O4] unit (Mn6, Mn7, Mn9). To the la tter are also attached the MnII (Mn8) and SrII atoms above the central unit, and a [Mn2O(OMe)] rhombus (Mn5, Mn10) below it. Within this descri ption, two triply -bridging methoxides (O43, O46) lie within the two cubanes whereas the doubly-bridging methoxide (O41) is within the [Mn2O(OMe)] rhombus. Peripheral ligation is provided by fourteen doublyand four triply-bridging Figure 5-2. The labeled [Mn14SrO11(OMe)3]18+ core of complex 19. Color scheme: MnIII light blue, MnII yellow, Sr green, O red, N da rk blue, C grey. H atoms have been omitted for clarity. benzoates, as well as two term inal MeCN molecules on the Sr2+ ion. Three of the 3PhCO2 groups have one O at om doubly-bridging a SrII/MnIII pair, whereas the fourth benzoate has it bridging a MnII/MnIII pair. Three of the MnIII atoms (Mn5, Mn9, Mn10) are five-coordinate, whereas all the remaini ng Mn are six-coordinate. The Sr is eightcoordinate with six Sr-O bonds in the range 2.501 2.769 and two Sr-N bonds to the

PAGE 140

124 Table 5-1. Crystallographic data for [SrMn14O11(OMe)3(O2CPh)18(MeCN)2]12MeCN Parameter 1912MeCN formulaa C157H141Mn14Sr1N14O50 fw, g mol-1 3880.62 space group P 1 a 18.4417(16) b 19.0430(16) c 25.447(2) deg 69.789(2) deg 78.549(2) deg 73.792(2) V 3 8001.2(12) Z 2 T K 173(2) radiation, b 0.71073 calc, g cm-3 1.611 cm-1 8.825 R 1 ( wR 2), %c,d 8.99 (20.69) a Including solvent molecules. b Graphite monochromator. c R 1 = || Fo| | Fc|| / | Fo|. d wR 2 = [ [ w ( Fo2 Fc2)2] / [ w Fo2)2]]1/2 where S = [[ w ( Fo2 Fc2)2] / ( n p )]1/2, w = 1/[ 2( Fo2) + ( mp )2 + np ], p = [max( Fo2, 0) + 2 Fc2]/3, and m and n are constants. two MeCN molecules of 2.711 and 2.763 Th e closest SrMn separation is 3.3 to Mn3. 5.2.3 Magnetochemistry of Complex 19 5.2.3.1 DC studies of 19 Solid-state variable-temperature magnetic susceptibility measurements were performed on vacuum-dried, micr ocrystalline samples of complex 19, restrained in eicosane to prevent torquing. Th e dc magnetic susceptibility ( M) data were collected in the 5.00 300 K range in a 1 kG (0.1 T) field, and they are plotted as MT vs. T in Figure 5-3 (left). The MT at 300 K is 30.67 cm3mol-1K, much lower than the 43.38 spin-only value (g = 2) expected for a cluster comprising thirteen MnIII and one MnII noninteracting ions, indicating extensive intramolecular antiferromagnetic interactions. The

PAGE 141

125 MT steadily decreases further with decreasi ng temperature and finally reaches 10.51 cm3mol-1K at 5.00 K, suggesting a small (but non-zero) ground state spin, S. Owing to the size and low symmetry of the molecule, a matrix diagonalization method to determine the various exchange parameters ( Jij) between MniMnj pairs was clearly unfeasible. We thus concentrated instead on identifying the ground state spin of th e molecule, and this was accomplished by collecting variable-tempe rature and variable -field magnetization (M) data in the 1.8 10 K and 0.1 4.0 T ranges; the data are plotted as reduced magnetization (M/N ) versus H/T in Figure 5-3 (right) The data were fit, using the program MAGNET,46 by diagonalization of the spin Hamiltonian matrix assuming only the ground state is populated, inco rporating axial anisotropy ( D z 2) and Zeeman terms, 050100150200250300 5 10 15 20 25 30 35 MT (cm 3 mol -1 K)Temperature (K) 0510152025 0 2 4 6 8 10 12 14 0.1 T 1 T 2 T 3 T 4 T Fitting H/T (kG/K)M/N Figure 5-3. (Left) Plot of MT vs T for complex 19. (Right) Magnetization (M) vs field (H) and temperature (T) data, plo tted as reduced magnetization (M/N B) vs H/T, for 19 at applied fields of 0.1, 1.0, 2.0, 3.0 and 4.0 T and in the 1.8 10 K temperature range. The solid lines are th e fit of the data; see the text for the fit parameters. and employing a full powder average. The co rresponding Hamiltonian is given by eq 5-1, where D is the axial anisotropy constant, B is the Bohr magneton, z is the easy-axis spin = D z 2 + g B0Hz (5-1) operator, g is the electronic g factor, 0 is the vacuum permeability, and Hz is the applied

PAGE 142

126 longitudinal field. The last term in eq 5-1 is the Zeeman term associated with an applied magnetic field. The fit parameters were S = 9/2, g = 1.88 ( 0.01) and D = -0.50 ( 0.01) cm-1. The D value of -0.5 cm-1 is consistent with the complex having predominantly MnIII ions and g < 2, as expected for Mn. When data collected at fields > 4.0 T were included, a satisfactory fit could not be obtained. We have found that poor quality fits of the magnetization versus H and T plots are a common problem in manganese chemistry when (i) the Mnx species is of high nuclearity and there is thus a high density of spin states resulting from the many excha nge interactions present amongst the constituent Mn ions; and/or (ii) one or more MnII ions are present, which typica lly give very weak (and usually antiferromagnetic) exchange inte ractions and thus small energy separations. As a result, there are many excited states th at are low-lying (relative to kT ), and some of these will have S greater than the S of the ground state. Population of the excited states will thus be difficult to avoid even at the lowest temperat ures normally employed. In addition, in the presence of a big enough applied dc field, Ms components of the ex cited state(s) can approach in energy the lowest-lying Ms of the ground state and even cross below it. The fitting procedure assumes only a single state is occupied, and thus an S value greater than the true ground-state S is given by the best fit of the data because it is affected by the contributions from the populated excited state( s). As a result, we t ypically observe that the best fit using all of the data collected over many field values will thus overestimate the M value at low fields and underestimate the M value at large fields. For this reason, we used for the fit of Figure 5-3 (right) only the M data collected at small fields and as a result a satisfactory fit was obtained, with the fit parameters given above.

PAGE 143

127 In order to confirm the global minimum had been obtained, and to assess the precision of the fit parameters, the root -mean-square error surface for the fit was generated as a function of g and D using the program GRID.98 The obtained error surface is depicted in Figure 5-4 as a 2D contour plot (left) and as a 3-D mesh plot (right), in the ranges of D = -0.35 to -0.68 cm-1 and g = 1.79 to 2.00 for the contour plot. A single minimum is observed, with the best fit at g = 1.88 and D = -0.50 cm-1 (the star in Figure 5-4). The error surface near this minimum is rather soft with an elongated contour describing the region from D = -0.49 to -0.51 cm-1 and g = 1.87 to 1.89, and we thus estimate uncertainties in the fit parameters of D = -0.50 0.01 cm-1 and g = 1.88 0.01, 1.801.851.901.952.00 -0.65 -0.60 -0.55 -0.50 -0.45 -0.40 -0.35 gD (cm-1) 1.801.851.901.952.00 -0.65 -0.60 -0.55 -0.50 -0.45 -0.40 -0.35 gD (cm-1) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 1.80 1.85 1.90 1.95 2.00 2.05 -1.0 -0.5 0.0 0.5E r r o rgD ( c m1) Figure 5-4. (Left) Two-Dimens ional contour plot of the r oot-mean-square error surface for the reduced magnetization (M/N B) vs H/T fit for complex 19 as a function of g and D The star marks the point of minimum error. (Right) Three-Dimensional mesh plot of the error vs D vs g for the same fit for 19. as already stated. Although 19 has a small ground state spin value of S = 9/2, the large D value of 0.5 cm-1, which is comparable to Mn12 SMMs,30 suggested that the barrier to magnetization relaxation might be large enough for the complex to function as an SMM. The S and D values obtained for complex 19 suggest an upper lim it to the potential

PAGE 144

128 energy barrier ( U ) to magnetization reversal of U = ( S2 )| D | = 10 cm-1 = 14 K,9 although the actual, effective barrier ( Ueff) was anticipated to be less than this due to quantum tunneling of the magnetization (QTM ). Thus, ac susceptib ility measurements were performed to investigate whether 19 was an SMM or not. 5.2.3.2 AC studies of 19 Ac studies were performed in the 1.8 10 K range in a zero dc field and a 3.5 G AC field oscillating at frequencies in the 10 1000 Hz range to determine if 19 might be an SMM. Typically in an ac susceptibility e xperiment, a weak field (generally 1 G) oscillating at a particular frequency ( ) is applied to a sample to probe the dynamics of the magnetization (magnetic moment) relaxation. A decrease in the in-phase ac susceptibility signal and a concomitant increas e in the out-of-phase signal are indicative of the onset of the slow, superp aramagnet-like relaxation of SMMs.7, 9 This occurs because at low enough temperatures, where the thermal energy is lower than the barrier for relaxation, the magnetization of the mo lecule cannot relax fast enough to keep in phase with the oscillating field. The obtained data for complex 19 are shown in Figure 55, and it can be seen that the molecule at temperatures below 3 K exhibits the frequencydependent tails of M'' signals whose peak maxima lie at temperatures below the operating minimum of our SQUID magnetometer (1.8 K). These M'' signals are accompanied by a concomitant frequency-dependent decrease in the in-phase ( M'T) signals. In the absence of slow relaxa tion, the ac in-phase signal ( M'T) should be equal to the dc MT, except that the former employs no dc fi eld, of course. In fact, this is very useful as an extra assessme nt of the ground state S value, since ac studies preclude the complicating possibility of low-lying excite d states, as discussed for the magnetization

PAGE 145

129 fits above. Thus, examination of the low temperature M'T value is a good indicator of the ground state S. The plot in Figure 5-5 shows that M'T is decreasing at temperatures < 10 K, consistent with population of excited states with S greater than that of the ground state. Below ~ 6 K, the line begins to slope more steeply and then drops rapidly below 3 K, 8 10 12 14 0246810 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 1000 Hz 250 Hz 50 Hz 10 Hz Temperature (K)M' T (cm3mol-1K) M'' (cm3mol-1) Figure 5-5. Plot of the in-phase (as M T ) and out-of-phase ( M ) AC susceptibility signals vs temperature for complex 19 in a 3.5 G field oscill ating at the indicated frequencies. from the slow relaxation effect. Some of the d ecrease in the 3 6 K region is likely due to weak intermolecular inte ractions, so extrapolating the plot from above 6 K to 0 K gives a M'T 10.5 cm3mol-1K, in satisfying agreement with the S = 9/2 and g = 1.88 values obtained earlier from the magne tization fits, which predict M'T to be 10.9 cm3mol-1K. Note that an S = 7/2 or 11/2 ground state would give M'T = 6.96 or 15.8 cm3mol-1K,

PAGE 146

130 respectively, for g = 1.88 (or M'T = 7.9 or 17.9 cm3mol-1K, respectively, for g = 2.0), both much more in disagreement with the expe rimental data in Figure 5-5. The ac data thus provide an independent confirmation of an S = 9/2 ground state for complex 19. The appearance of out-of-phase M'' signals suggests that complex 19 may indeed have a significant (vs kT ) barrier to magnetization rela xation and thus may be a SMM. However, confirmation of this requires ma gnetization vs applied dc field sweeps to display hysteresis loops, and this was explored on with studies at temperatures below 1.8 K, as described below. 5.2.3.3 Hysteresis studies below 1.8 K To establish whether 19 is a SMM, magnetization versus applied dc field data were collected on single crystals of 1912MeCN (that had been kept in contact with the -1 -0.5 0 0.5 1 -1-0.500.51 0.04 K 0.2 K 0.3 K 0.4 K 0.5 K 0.6 K 0.7 K 0.8 K 0.9 K 1.0 K 1.1 K M/Ms 0H (T) 0.14 T/s -1 -0.5 0 0.5 1 -1.2-0.8-0.400.40.81.2 0.140 T/s 0.070 T/s 0.035 T/s 0.017 T/s 0.008 T/s 0.004 T/s 0.002 T/s 0.001 T/s M/Ms 0H (T) 0.04 K Figure 5-6. (Left) Magnetization ( M ) vs applied magnetic field ( H ) hysteresis loops for single-crystals of 1912MeCN at a fixed sweep rate of 0.14 T/s and at the indicated temperatures, and (right) at 0.04 K and at th e indicated sweep rates. In both plots M is normalized to its saturation value, Ms. mother liquor) down to 0.04 K using a mi cro-SQUID instrument, with the field approximately along the easy axis ( z axis) of the molecule.50 Magnetization vs field hysteresis, the diagnostic property of a magne t, was indeed observed below 1.3 K (Figure 5-6). The hysteresis loops exhi bit increasing coercivity with increasing field sweep rate at

PAGE 147

131 a constant temperature (Figur e 5-6, right), and increasing coercivity with decreasing temperature at a constant sweep rate (Figure 5-6, left). This is as expected for the superparamagnet-like properties of a SMM. Th ese hysteresis loops thus confirm complex 19 to be a new addition to the family of SMMs. The blocking temperature ( TB) is ~1.4 K, above which there is no hysteresis, i.e., the sp in relaxes faster to equilibrium than the timescale of the hysteresis loop measurements. The most dominating feature of the hysteres is loops in Figure 5-6 is the large step at zero field, which is due to quantum tunne ling of the magnetizati on (QTM) through the Figure 5-7. Plot of the magneti zation versus time decay data of 19. Because of fast QTM, decay time, and hence relaxation rate vs temperature kinetic data cannot be obtained from this. anisotropy barrier, with a second step at ~ 1 T. The large zero-field step (which corresponds to a surge in the magnetization relaxation ra te) indicates that QTM in zero field is fast, and this in turn is consistent with the low symmetry of the molecule, which introduces a non-zero rhombic (transverse) anisotropy into the spin Hamiltonian, i.e. E ( x2y2), where E is the rhombic zero-fi eld splitting parameter. The greater is the transverse anisotropy, the grea ter will be the mixing of leve ls on either of the anisotropy barrier, leading to increased rates of QTM. Unfo rtunately, the fast relaxation rate in zero 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.11101001000M/Ms t (s) from 1 K to 0.05 K in steps of 0.05 K

PAGE 148

132 field that results in the large step at this position also prevents us from collecting magnetization vs time decay data with which to construct an arrhenius plot and determine the effective barrier to relaxation ( Ueff). Typically, in such a dc magnetization decay experiment, the magnetization is saturated wi th an applied dc field at a particular temperature, then the temperature is lowered to a chosen value, the field switched off, and the magnetization monitored with time (Fig. 5-7). The magnetization relaxation of a SMM obeys the Arrhenius relationship = 0 exp ( Ueff/kT), the characteristic behavior of a thermally-activated Orbach process,110 where Ueff is the effective (kinetic) anisotropy energy barrier, k is the Boltzmann constant, and 1/ 0 is the pre-exponential term. Thus, if we can obtain the Temperatures (T) and the corresponding decay times ( ), we can determine the Ueff and 0. However, in cases where there is a fast QTM at H = 0 (resulting in a sudden increase in the magnetization relaxation rate ( d M/ d t), which corresponds to a very small time ( d t)), it is difficult to obtain the decay time ( d t) from, for example Fig. 5-7, and thus to construct th e Arrhenius plot. Thus, important parameters such as the Ueff and the preexponential, 0, could not be obtained for 19. Hence, efforts to compare these values with the U of complex 19 and similar parameters of other SMMs proved futile. 5.2.4 X-ray Absorption Spectroscopy of Complex 19 Complex 19 is the first molecular Sr/Mn complex and provides a crystallographically characterized benchmark w hose data can be compared with those of Sr-reactivated PS II preparations. Previous Sr EXAFS studies at the Sr K-edge on isotropic and oriented Sr-reactivated PS II samples, and at the Ca K-edge on native PS II samples, confirmed that Sr or Ca is located at ~3.5 or ~3.4 respectively, from the Mn cluster.106, 25 Analogous data have now been collected on the SrMn14 compound 19. Note

PAGE 149

133 that although the Mn nuclearity in 19 is higher than that in the WOC, the Sr nuclearity is identical, i.e. one. Note also that even after the X-ray crystallographic21, 22, 26 on one Sn state (dark-adapted S1 state) becomes available at a re solution sufficient to define the precise structure of the WOC, the exact position of Ca and the nature of CaMn interactions rely particular ly on the anomalous diffraction data which contains severe radiation damage effects. Therefore, biophys ical techniques whic h employ soft X-rays, such as XAS and EPR on different Sn states will be imperativ e to gain an understanding of the structural, topological and mechanistic changes o ccurring during the catalytic cycle at the WOC, especially as it sh uttles through its other, less stable Sn (n = 0,2,3,4) states.19, 20 For this reason, model compounds such as 19 with precisely known structures are and will continue to be inva luable to assist the interpretation of the data obtained from XAS analysis of the native site at various Sn states. 5.2.4.1 Sr EXAFS of 19 The Sr K-edge EXAFS (k3-weighted) spectra of complex 19 and Sr-reactivated PS II preparations in the S1 state106a are shown in Figure 5-8 (top). The Fourier transforms of the Sr EXAFS of 19 and PS II S1 state are shown in Figure 5-8 (bottom). Each peak in the Fourier-transformed (FT) spect rum indicates a radial dist ribution of atoms surrounding the Sr atoms in 19 and appears at distances which are shorter than the actual Srbackscatterer distances due to an average phase shift induced by the potentials of the given absorber-scatterer pair on the photoelectron. The first F ourier peak is from the first shell of ligands about Sr (O and/or N backsca tters), and it can be simulated by 7 2 light atoms (O and N) at 2.53 0.02 and 2.66 0.02 which is in good agreement with the X-ray structure of the complex (Table 5-3) The second Fourier peak is from SrMn separations. Although there are three such separations (3.3, 3.7 and 3.9 ) in the Mn14Sr

PAGE 150

134 structure, only the shorter vectors (3.3 and 3.7 ) are the dominant feature in the FT peak. A detailed list of intera tomic distances can be found in Table 5-2. The second peak Figure 5-8. (Top) k3-weighted Sr K-edge EXAFS spectra of the Mn14Sr compound (red) and Sr reactivated PS II samples in the S1 state (black). (Bottom) Fourier transforms of the averaged Sr K-edge EXAFS spectra for the Mn14Sr complex 19 (red) and Sr-reactivated PS II samples in the S1 state (black). of the PS II spectrum (black) in Figure 5-8 (bo ttom) is clearly shifte d to a longer distance compared to the Mn14Sr model, which is in good agreemen t with the fact that the SrMn distance in Sr reactivated PS II wa s earlier reported to be 3.5 106a and thus longer than the shortest SrMn vector in complex 19 (3.3 ). The higher inte nsity of the second peak

PAGE 151

135 in Sr-reactivated PS II compared to the shoulder seen for complex 19 is likely due to presence of more than one SrMn interaction in the WOC of PS II. This hypothesis is in Table 5-2. Selected bond and inte ratomic distances () for complex 19. Mn3-O50 1.863(5) Mn3-O48 1.905(5) Mn3-O37 1.909(6) Mn3-O21 1.935(6) Mn3-O47 2.180(6) Mn3-O6 2.265(7) Mn3Sr1 3.289(2) Mn6-O50 1.883(5) Mn6-O39 1.902(5) Mn6-O36 1.968(6) Mn6-O22 1.974(5) Mn6-O15 2.200(7) Mn6-O33 2.236(7) Mn6-Mn7 2.8920(18) Mn6Sr1 3.839(2) Mn7-O49 1.896(5) Mn7-O45 1.908(5) Mn7-O39 1.920(5) Mn7-O50 1.933(5) Mn7-O38 2.149(6) Mn7-O30 2.486(10) Mn7Mn9 2.8474(19) Mn7Sr1 3.7458(19) Mn8-O29 1.986(10) Mn8-O16 2.150(8) Mn8-O23 2.151(8) Mn8-O44 2.271(6) Mn8-O39 2.498(6) Mn8-O35 2.505(6) Mn9-O49 1.846(5) Mn9-O45 1.861(5) Mn9-O19 1.904(6) Mn9-O12 1.953(6) Mn9-O17 2.256(8) Sr1-O30 2.501(6) Sr1-O37 2.583(6) Sr1-O6 2.607(7) Sr1-O4 2.626(11) Sr1-O15 2.685(7) Sr1-N1 2.711(16) Sr1-N2 2.763(17) Sr1-O50 2.769(6) agreement with the recent crystallographic data on PS II at 3.0 wherein 3 CaMn separations were postulated.26 The Fourier peaks I and II were isolated separately and simulated with Srligand (2.5-2.7 ) and Sr Mn (> 3.3 ) distances (Table 5-3). Fit I-1 attempts to fit peak I with only SrO inter actions, while Fit I-2 in cludes both SrO and SrN interactions. A noticeable improvement in fit quality was observed when peak I was fit using the coordination number ( N ) from the crystal structure data. The coordination number of Sr deviates from the true N (6 SrO and 2 Sr N bonds) by less than 10 % when unconstrained. Peak II, which clearly c ontains at least two co mponents, was fit to multiple SrMn distances; as stated above, th ere are three SrMn interactions (3.3, 3.7

PAGE 152

136 and 3.9 ) in the crystal structure. If complex 19 was treated as an unknown structure with two-shell SrMn intera ctions (Fit II-1), the total N value ( N = 1.4 for 3.34 and N = 1.2 for 3.72 ) is between 2 to 3, which indicates the presence of at least two SrMn Table 5-3. Least-squares fits of Fourier-f iltered peaks I and II of Sr EXAFS data compared with structural parameters from the X-ray crystal structure on Complex 19 and the Sr-substituted WOC of Photosystem II in the S1 state.[a] Fit No. Shell R () N 2 /103 2 /103 2 /105 XRD R / XRD N Mn14Sr I-1 Sr-O 2.59 8.8 12.0 0.27 0.10 I-2 Sr-O Sr-N 2.56 2.73 5.2 2.3 7.1 6.0 0.22 0.084 Pea k I I-3 Sr-O Sr-O Sr-N 2.54 2.62 2.74 4.0[b]2.0[b]2.0[b]6.7 [c] 0.22 0.084 Sr-O ~2.5 ~2.6 Sr-N ~2.7 4 2 2 II-1 Sr-Mn Sr-Mn 3.34 3.72 1.4 1.2 6.0 [c] 1.10 0.76 Pea k II II-2 Sr-Mn Sr-Mn Sr-Mn 3.35 3.69 4.00 1.0[b]1.0[b]1.0[b]4.0 7.3 4.5 0.57 0.22 Sr-Mn 3.3 3.7 3.9 1 1 1 Sr-Photosystem II S1 [d] Peak I Sr-O 2.57 9.0 Peak II Sr-Mn 3.54 2.0 7.8 0.402 0.30 [a] N the number of interactions; 2, Debye-Waller parameter; and 2, fit-quality parameters. S0 2 value was fixed to 1.0. For details, see Appendix C. [b] N is fixed to known values from XRD (crystal stru cture) data. [c] Single value of 2 is used for all the interactions. [d] Data was taken from reference 106a. interactions. Increasing the number of shells from two to three (Fit II-2) with fixed N values obtained from the crystal structure data significantly improved the fit quality. In comparison to the Mn14Sr compound, the second peak of the Sr-subs tituted PS II S1 state shows a single FT peak at 3.5 .106a This was best fit to 2 SrMn interactions ( N = 2) with similar fitting parameters ( 2 and S0 2).

PAGE 153

137 Figure 5-9 (top) shows two SrMn4 sub-units within 19 with an open-cubane topology. Both of these subunits are important with respect to the Sr EXAFS results 3.75 3.84 2.89 2.84 3.65 3.72 3.36 3.36 3.29 3.75 3.84 2.89 2.84 3.65 3.72 3.75 3.84 2.89 2.84 3.65 3.72 3.36 3.36 3.29 3.36 3.36 3.29 Figure 5-9. (Top) PovRay representation at th e 50% probability level of two open-cubane containing sub-units within 19. On the left, the extrinsic Mn (Mn9) is linked by two 2 bridging oxides, whereas on the right the extrinsic Mn (Mn3) is attached to an oxide of the open-cuba ne. (Bottom) Left depicts a sub-unit which resembles the 3.0 cr ystal structure of the WOC26 of PS II. The relevant metal-metal separations with in the above-mentioned sub-units are depicted in center and right. Color scheme: MnIII light blue, MnII yellow, Sr green, O red, C grey. because they contain the 3 MnIII ions (Mn3, Mn6 and Mn7) which are most proximal to the Sr ion in the SrMn14 complex. In particular, the Sr separation from Mn3 and Mn7 is 3.3 and 3.7 respectively. A more detail ed list of the bond lengths and interatomic distances can be found in Table 5-2. Additi onally, Figure 5-9 (bottom) gives the exact

PAGE 154

138 metal-metal separations with in the earlier mentioned s ub-units. Although the study by Ferreira et al. at 3.5 suggests22 that the WOC might have a CaMn3O4 cubane linked to a fourth Mn atom, the earlier X-ray studi es at 3.8 and 3.7 resolution postulated21 that the four Mn ions are arranged in a + 1 (i.e. open-cubane ) fashion. Thus, the sub-units of 19 in Figure 5-9 have some intriguing simila rity to the structural models proposed from the crystallographic studies of the WOC S1 state. Further, to make this discussion more interesting, another view of one of thes e sub-units (Figure 5-9 t op (left)) is shown in Figure 5-9 (bottom (left)), and this spatial orientation highlights the metals (3 Mn, Sr) arranged in a pyramid (depicted by green line s); Mn6, Mn7 and Sr1 form a plane and at the apex of the pyramid is Mn8. Extrinsically connected to this pyramidal unit is Mn9. This description of the geometry and arrangement of metal atoms is comparable to the most recent crystallographic study of the S1 state of the WOC of PS II at 3.0 in which the aforementioned model was proposed.26 Thus, it is perhaps not surprising that the FT peak positions and the intensity observed in Sr EXAFS spectrum for 19 in Figure 5-8 are similar to those of the S1 state of PS II, because it is possible that the local ligand and structural environment that the Sr atom of complex 19 finds itself in, might well be very similar to that of the Sr atom in Sr-reactivat ed PS II preparations, even though, of course, the overall Mn nuclearity of the SrMn14 complex is much higher than that of the native site. 5.2.4.2 Mn EXAFS of 19 We also carried out Mn EXAFS studies on complex 19. Comparison of Mn K-edge EXAFS (k3-weighted) spectra with those of the WOC in the S1 state is shown in Figure 510 (top), and the Fourier transf orms are shown in Figure 5-10 (bottom). The three main peaks, labeled I III, have been assigned in the WOC to the Mnligand (peak I), di-

PAGE 155

139 oxide-bridged MnMn (peak II), and Mn Mn and MnCa (peak III) separations, respectively. Figure 5-10. (Top) k3-weighted Mn K-edge EXAFS spectra of the Mn14Sr compound (red) and PS II samples in the S1 state (black). (Bottom) Fourier transforms of the averaged Mn K-edge EXAFS spectra for the Mn14Sr complex 19 (red) and PS II samples in the S1 state (black). Inspection of the FT spectra in Figur e 5-10 (bottom) indi cates significant differences in the peak positions a nd relative intensities between complex 19 and the WOC. MnO distances are longer in 19 than in the WOC, which is indicated by the shift of peak I to longer distances. This is consistent with the lower average Mn oxidation level

PAGE 156

140 in 19 (13 MnIII, MnII) compared with the WOC S1 state (2MnIV, 2MnIII),18 which will give longer average Mn-O bonds in the former. In general, MnIV ions (2 MnIV in WOC S1 state) are symmetric and isotropic, and hen ce have shorter MnO bond lengths (generally below 2 ) when compared to the anisotropic JT distorted MnIII ions of complex 19. In the Mn14Sr compound, only weak peaks are observe d in the region for peaks II and III. This is clearly due to the ma ny different MnMn separations in the molecule spanning distances in a wide range of 2.9 4.0 In such a case, the contribution of each individual vector is less pr onounced in the spectrum normalized to one Mn atom, and the EXAFS oscillations are damped by the high distance distribution. 5.3 Conclusions The first molecular Sr/Mn complex has been synthesized. Its magnetochemical study has established that it possesses an S = 9/2 ground state and a significant D value arising from the projection of the large number of MnIII single-ion anisotropies onto the molecular anisotropy axis. As a result of this combination of significant S and D values, the complex has a sufficient barrier to magne tization relaxation to function as a singlemolecule magnet (SMM). It also exhibits quantum effects, and undergoes fast QTM in zero field, as reflected in a very large QTM st ep at zero field in the hysteresis loops. In fact, this fast QTM has precluded detailed magnetization vs time decay studies to obtain relaxation rate vs T kinetic da ta; we cannot therefore construct an Arrhenius plot from which could be determined the effectiv e barrier to magnetization relaxation, Ueff. Nevertheless, the appearance of frequency de pendent signals along with hysteresis loops with increasing coercivity with decreasi ng temperatures firmly establish complex 19 to be a new addition to the family of SMMs.

PAGE 157

141 Although the overall nucl earity of complex 19 is too large for it to act as a true model of the WOC of PS II, it nevertheless has only one Sr atom and thus its Sr EXAFS data can be informatively compared with thos e for Sr-reconstituted PS II samples. This Sr EXAFS comparison has revealed significant similarities between the local Sr environment in the model complex and that in the Sr-substituted WOC. Complex 19 is thus both an important step forward in obt aining model complexes for this mixed-metal biological site, and a useful benchmark for a ssisting the interpretation of data arising from EXAFS and related studies on the biol ogical site. The obvious next step is to synthetically access smaller nuclearity Sr/Mn species, ideally at the SrMn4 pentanuclearity and with a structure analogous or similar to those shown for the sub-units in Figure 5-9. It would then be possible to carry out more extensive Sr and Mn EXAFS comparisons between model complexes a nd the Sr-reactivated WOC, as well as comparative studies by other techniques such as EPR. Attempts to obtain such smaller nuclearity mixed Sr/Mn (and Ca/Mn) complexes are in progress. 5.4 Experimental 5.4.1 Synthesis All manipulations were performed under aerobic conditions us ing chemicals as received, unless otherwise stated. (NBun 4)[Mn4O2(O2CPh)9(H2O)] (3) was prepared as described elsewhere.55 [SrMn14O11(OMe)3(O2CPh)18(MeCN)2] (1912MeCN). To a stirred solution of 3 (0.50 g, 0.31 mmol) in MeCN/MeOH (20/2 mL) was slowly added Sr(ClO4)2xH2O (0.09 g, 0.31 mmol), and this produced a dark brow n solution. This was stirred for a further 20 minutes, filtered, and the filtrate slowly concentrated by evaporation at ambient temperature, which slowly pr oduced dark red crystals of 1912 MeCN over two days. The

PAGE 158

142 yield was 55%. The crystals of 19 were maintained in the mother liquor for X-ray crystallography and other single-crystal studies or collected by filtration, washed with MeCN, and dried in vacuo The synthesis can also be carried out with Sr(NO3)2. Vacuum-dried solid analyzed as solvent-free. Elemental analysis (%) calcd. for 19 (C133 H105Mn14Sr1N2O50): C, 47.15; H, 3.12; N, 0.83; found: C, 46.92; H, 2.92; N, 0.63. Selected IR data (KBr, cm-1): 3430(br), 2968(w), 1600(m), 1559(m), 1405(s), 1177(w), 1069(w), 1025(w), 841(w), 715(s) 676(m), 615(m), 503(w). 5.4.2 X-ray Crystallography Diffraction Data were collected on a Si emens SMART PLATFORM equipped with a CCD area detector and a graphite monochromator utilizing Mo K radiation ( = 0.71073 ). A suitable crystal of 192MeCN was attached to a gla ss fiber using silicone grease and transferred to a goniostat where it wa s cooled to 173 K for data collection. An initial search of reciprocal sp ace revealed a triclinic cell for 19, the choice of space group P 1 was confirmed by the subsequent solution and refinement of the structure. Cell parameters were refined usi ng up to 8192 reflections. A full s phere of data (1850 frames) was collected using the -scan method (0.3 frame width). The first 50 frames were remeasured at the end of data collection to monitor instrument and crystal stability (maximum correction on I was < 1 %). Abso rption corrections by integration were applied based on measured indexed crystal f aces. The structures were solved by direct methods in SHELXTL6 ,56a and refined on F2 using full-matrix least squares. The non-H atoms were treated anisotropically, wher eas the hydrogen atoms were placed in calculated, ideal positions and refined as riding on their respective carbon atoms. The asymmetric unit contains the complete SrMn14 cluster and 12 MeCN molecules of crystallization. The latter were disorder ed and could not be m odeled properly, thus

PAGE 159

143 program SQUEEZE,56b a part of the PLATON56c package of crystallographic software, was used to calculate the solvent disorder area and remove its contribution to the overall intensity data. A total of 1750 pa rameters were refined in the final cycle of refinement using 35042 reflections with I > 2 (I) to yield R1 and wR2 of 8.96 and 20.69%, respectively. 5.4.3 XAS studies X-ray absorption spectroscopy (XAS) was pe rformed at the Stanford Synchrotron Radiation Laboratory (SSRL) on beamline 9-3 at an electron energy of 3.0 GeV with an average current of 70-90 mA. The experi ments were conducted by Dr. Junko Yano and Dr Yulia Pushkar, members of the Yachandr a group in the Physical Biosciences Division at Lawrence Berkeley Laboratory, Berkeley CA. The radiation wa s monochromatized by a Si(220) double-crystal monochr omator. The intensity of th e incident X-ray beam was monitored by a N2-filled ion chamber (I0) in front of the sample. XAS samples were made by carefully grinding 5-10 mg of compoun d and diluting it with a 10-fold excess of boron nitride. The mixture was packed into 0. 5-mm-thick sample holders and sealed with Mylar windows. For complex 19, the data were collected as fluorescence excitation spectra with a Lytle detector.111 Energy was calibrated by th e pre-edge peak of KMnO4 (6543.3 eV) for Mn XAS, and by the edge peak of strontium acetate (16120 eV) for Sr XAS. The standards were placed between two N2-filled ionization chambers (I1 and I2) after the sample. The X-ray flux at 16-17 keV was between 2 and 5 x 109 photons s-1mm-2 of the sample. The monochromator was det uned at 6600 eV to 50% of maximal flux to attenuate the X-ray 2nd harmonic. Samples were kept at a temperature of 10 K in a liquid helium flow cryostat to minimize radiation damage. Data reduction has been detailed previously112 and will be only briefly summarized here: After conversion of background-

PAGE 160

144 corrected spectra from energy space to phot oelectron wave vect or (k) space, and weighted by k3, a four-domain spline was subtracted for a final background removal. See also Appendix C (Physical Measurements) fo r details regarding fitting of the data.

PAGE 161

145 CHAPTER 6 THE FIRST FAMILY OF HETERO METALLIC CALCIUM-MANGANESE COMPLEXES: Ca-EXAFS AND -XANES COMPARISON WITH THE WATER OXIDIZING COMPLEX OF PHOTOSYSTEM II 6.1 Introduction The water oxidizing complex (WOC) at Photosystem II (PS II) catalyses the oxidation of H2O to O2 and H2 gas in green plants and cyanobacteria.18 This four-electron photo-induced process involves various oxidation levels of the WOC (the so-called Sn states, n = 0 to 4),19, 20 and is the source of essentially all the O2 on this planet. Given the role of PS II in maintaining life on the bios phere and the future vi sions of a renewable energy economy, it is vital to elucidate the st ructure and mechanism of the WOC which is sometimes referred to as the heart of th e water oxidation process. The WOC, also known as the Oxygen-Evolving Complex (OEC), has been studied for many years by a variety of spectroscopic a nd biochemical techniques.113, 114 The WOC has long been known to comprise a tetranuclear, oxide-bridged Mn4 cluster containing primarily MnIII and MnIV ions and carboxylate periph eral ligation. It has also been known that the WOC requires Ca2+ for activity,24 and calcium EXAFS (exten ded X-ray absorption fine structure) studies of the WOC have re vealed a MnCa separation of ~3.4 .25 Amongst the various S states, S1 is the most stable, well-studi ed and understood one; various crystallographic studies have also been performed on this state. The first two crystal structures of PS II at 3.8 and 3.7 resoluti on indicated that the WO C is a tetranuclear Mn cluster with the four Mn ions organized in a + 1 fashion.21 The most recent crystallographic studies on the PS II reaction center of the cyanobacterium

PAGE 162

146 Thermosynechococcus elongatus at 3.5 22 and 3.0 26 resolutions have identified Ca as being part of the Mn complex using anomalous diffraction data. At the current resolutions smaller ligands such as C, H, O, H2O, N and Cl cannot be conf idently located. Although there is still ambiguity about the Mn4Ca structure obtained from crystallography due to radiation damage during X-ray data collection,23 there is little doubt that the WOC is a heterometallic [Mn4CaOx] cluster on the basis of both XAS,25, 27, 106, 129 and XRD22, 26 studies. The availability of inorganic Ca/Mn heterometallic complexes to act as synthetic models of the WOC would re present an important step forward in understanding the magnetic and spectroscopic pr operties of the native site and the mechanism of its function. Many groups have in the past applied the Syntheti c Analogue Approach115 to the WOC and a plethora of Mn4 complexes have been synthesized18c which can be broadly categorized as i) butterfly,79 ii) cubane,80 iii) adamantane,81 and iv) linear,116 to name a few. However, in contrast to the growing knowledge of the Mn4Ca complex in PS II, there is essentially nothi ng in the inorganic chemistry lit erature of mixed-metal Ca/Mn compounds, compared to the copious multinuc lear homometallic Mn complexes known. In fact, only recently was the first exampl e of a molecular Ca/Mn cluster reported.43 Therefore, as part of our ongoing interest in obtaining synt hetic models of the WOC and its various modified forms, we have b een extensively investigating mixed Ca/Mn chemistry and have successfully obtained the first family of such molecular species; Mn13Ca2, Mn11Ca4 and Mn8Ca complexes. Although the Mn nuclearity in these complexes is much higher than that in the WOC, the Ca nuclearity is comparable. Thus a comparison of Ca X-ray absorption spectra fr om these complexes to Ca X-ray spectra

PAGE 163

147 (EXAFS and XANES (X-ray absorption near edge st ructure)) from PS II is of interest. In this chapter, we report the synthesis, st ructure, magnetic properties and XAS (X-ray absorption spectroscopy) studies, on the first family of heterometallic Ca/Mn complexes. The work herein represents an important advancement in obtaining bioinorganic model complexes for the mixed-metal WOC site of PS II and serves as a benchmark for comparison of similar data from the WOC. It also generates important parameters for calibration of biophysical t echniques such as XAS. 6.2 Results and Discussion 6.2.1 Syntheses With the recent crystal structures of PS II becoming available,22, 26 we began our quest for synthetic routes to Ca/Mn, Sr/M n and Ln/Mn (Ln = lanthanide) heterometallic complexes, and initial results from these effort have been very fulfilling.37, 39, 43, 117 Our synthetic strategy has been to employ pr eformed manganese clusters such as [Mn3O(O2CR)6(py)3] (R = Me, But, Ph)118 and (NBun 4)[Mn4O2(O2CPh)9(H2O)], and treat them with a heterometal carboxylate or other salt. Particularly at tractive as a starting material is the MnIII 4 complex (NBun 4)[Mn4O2(O2CPh)9(H2O)], which contains only carboxylate and water ligands around the Mn/O core.55 It is thus an attractive (and easily prepared) starting material th at can potentially yield MnIII(and even MnIV-) containing products with primarily carboxylate and oxide ligation; complexes of obvious relevance to the WOC of PS II. We have thus been carrying out a thorough investigation of the reactivity of this Mn4 complex under a variety of conditi ons, with various divalent metal salts, seeking to access new mixed-metal species. As part of the above inves tigation, the following procedur e with Ca was developed, which has been communicated recently.43 Treatment of (NBun 4)[MnIII 4O2(O2CPh)9(H2O)]

PAGE 164

148 with 0.25 equiv. of Ca(NO3)2H2O in MeCN/MeOH (20:1 v/v) resulted in the subsequent isolation of [Mn13Ca2O10(OH)2(OMe)2(O2CPh)18(H2O)4]0MeCN (20MeCN) in 40% yield (based on Mn). The Mn4:Ca ratio of 1:1 was initially explored, obviously directed at a Mn4Ca product as would be present in the WOC. However, this yielded crystals of 20 along with white, insoluble precipitate, which was identified by IR as un-reacted Ca(NO3)2. Thus, lowering the equiv. of Ca(NO3)2 to 0.25 resulted in isolation of pure, uncontaminated product. The small amount of MeOH was necessary to ensure solubility of Ca(NO3)2. As can be seen in eq 6-1 the ideal ratio of Mn4:Ca for the formation of 20 is 13:8. Thus, it was not surprising that 20 was obtained when the ratio of Mn4:Ca was varied from 1:0.25 till 1: 1, with the yield maximized for the ratio of 1:0.25 as mentione d in the experimental section (see later). The synthesis can also be repeated by employing Ca(ClO4)2 or Ca(O2CPh)2 as Ca sources. 13 [Mn4O2(O2CPh)9(H2O)]+ + 8 Ca(NO3)2 + 25 H2O + 8 MeOH + 1 H+ + 2 e4 [Mn13Ca2O10(OH)2(OMe)2(O2CPh)18(H2O)4] + 45 PhCO2H + 16 NO3 (6-1) The absence of any chelating agents in the above-mentioned reaction solution allowed any smaller nuclearity intermediate s to aggregate further and yield the Mn13Ca2 product 20. An obvious modification of the above reaction is the addition of chelates such as bipyridine (bpy), which would po ssibly facilitate the trapping of smaller nuclearity clusters. Thus, the treatment of (NBun 4)[Mn4O2(O2CPh)9(H2O)] and 1 equiv. of Ca(O2CPh)2H2O in MeCN/MeOH (20:1 v/v) with bpy (1 equiv.) was explored and this led to the isolation of 21 as shown in eq 6-2. The reaction solution is almost certainly a complicated mi xture involving fragmentati on, aggregation and redox processes, with several species likely in equilibrium, and the attainment of complex 21

PAGE 165

149 containing a [Mn11Ca4]2anion, co-crystallizing with 2 Mn4 bpy butterfly cations is a testimony to the fact that fact ors such as solubility, lat tice energies, crystallization 19 [Mn4O2(O2CPh)9(H2O)]+ + 16 Ca(O2CPh)2 + 15 H2O + 8 MeOH + 16 bpy + 15 H+ + 34 e4 [Mn4O2(O2CPh)7(bpy)2]2[Mn11Ca4O10(OH)2(OMe)2(O2CPh)20(H2O)2] + 67 PhCO2H (6-2) kinetics, and others undoubtedly determine th e identity of the is olated product. Since the strategy of employing bidentat e smaller chelates did not help in obtaining a relatively smaller nuc learity product, we decided to turn our attention to the use of bulky carboxylates/chelates. One inte resting candidate in this respect was diphenylphosphinic acid, as it has been known pr eviously to give high oxidation state Mn complexes.119 This ligand has been explored extensively by Dismukes and coworkers119 towards obtaining homometallic cubane complexes as putative models of the WOC of PS II. Thus the reaction of (NBun 4)[Mn4O2(O2CPh)9(H2O)] with 0.5 equiv. of Ca(O2CPh)2 in MeCN/MeOH (10:1) in the presence of 9 equiv. of diphenylphosphinic acid gave 22. The stoichiometric ratio for the attainment of 22 is shown in eq 6-3, and the excess of acid is used to push the equilibrium forward, resulting in the isolation of pure 22. Additionally, the 9 equiv. of diphenylphosphinic acid we re added to ensure substitution of the 9 benzoates in the Mn4 starting material; the 2 benzoates in 22 can be imagined to have come from the calcium benzoate reagent. This reaction is relatively clean and 22 is obtained in 55% yield. 2 [Mn4O2(O2CPh)9(H2O)]+ + Ca(O2CPh)2 + 12 (Ph)2PO2H + 4 MeOH + 2 e[Mn8CaO6(O2P(Ph)2)12(O2CPh)2(OMe)2(MeOH)2] + 18 PhCO2H (6-3)

PAGE 166

150 The presence of MeOH as a solvent or co -solvent, not only ensu res solubility of all reactants, as already stated, but also serv es to (potentially) pr ovide terminal and/or bridging ligands. Indeed, complexes 20-22 contain two bridging me thoxide groups in two cubanes and additionally 22 contains 2 terminal MeOH on the Ca ( vide infra ). The above-mentioned compounds are thus the most recent members of a growing class of methoxide-bridged mixed-metal species,37, 39, 43, 117 although it should be added that 20-22 represent the first family of mixed Ca/Mn cl usters of any type, containing high-oxidation state Mn ions and thus are of obvious relevance to the WOC of PS II. 6.2.2 Description of the Stru ctures of complexes 20-22 PovRay representations of the crystal structures and labeled cores of complexes 20-22 are depicted in Figures 6-1, 6-3, and 6-4, respectively. Un it cell and structure refinement data of complexes 200MeCN, 214MeCNH2O, and 22 are listed in Table 6-2. Selected interatomic distances are tabulated in Table 6-3. Complex 20MeCN crystallizes in the triclinic space group P 1 with the Mn13Ca2 molecule lying on an inversion cent er. The structure consists of a [Mn13Ca2O10(OH)2(OMe)2]18+ core whose Mn ions are mixed-valent (MnIV, MnIII 10, MnII 2) and bridged by six 4-O2-, four 3-O2-, two 3-HO-, and two 3-MeOions (Figure 6-1). The metal oxidation states and the protonation levels of O2-, HOand MeOions were established by bond-valence su m calculations (see Table 6-1),120, 89 charge considerations, inspection of metric pa rameters and the identification of MnIII Jahn-Teller (JT) elongation axes on all Mn atoms except Mn2 (MnIV), and Mn3, Mn3' (MnII). The core (Figure 6-1, bottom) consists of two [MnIII 4O3(OMe)] cubanes attached, via oxide bridges, on either side of a centr al, near linear and planar [Mn3O4] unit (Mn1, Mn2,

PAGE 167

151 Figure 6-1. (Top) PovRay representation at th e 50% probability level of the X-ray crystal structure of 20 (with the benzoate rings omitted for clarity, except for the ipso C atoms) and (bottom) its labeled [Mn13Ca2O10(OH)2(OMe)2]18+ core. Color scheme: MnIV dark blue, MnIII cyan, MnII purple, Ca yellow, O red, C grey. H atoms have been omitted for clarity. O C Ph O Mn Mns y n s y n O C Ph O Mn Mns y n s y n a n t i 3O C Ph O Mn Mn3,1, 4CaMn Ca Figure 6-2. Schematic representation of the three bridging modes found in complex 20. Mn1'). To the latter are also attached two MnIICaII pairs, one above and one below the plane, via the triply bridging hydroxides (O3, O3'). In this description, the two triply-

PAGE 168

152 bridging methoxides (O19, O19') lie within the two cubanes. The MnIV ion (Mn2) lies in the center, whereas the two MnII ions (Mn3, Mn3') are adjacent to the Ca2+ ions as already stated. Peripheral ligation is provided by fourteen -, two 3and two unusual 3, 1, 4-bridging benzoate groups, as well as four terminal H2O molecules, one each on Mn6, Mn6', Mn7, a nd Mn7'. The two 3-PhCO2groups have one O atom doublyTable 6-1. Bond valence sum calculations for the Mna and Ob atoms of complex 20. a The underlined value in bold is the one clos est to the actual charge for which it was calculated. The oxidation state of a particul ar atom can be taken as the nearest whole number to the underlined value. b The oxygen atoms is O2if Vi 2, MeOif Vi 2, OHif Vi 1, and H2O if Vi 0. bridging a CaII/MnII pair, whereas the two 3 O atoms of the 4-PhCO2groups bridge a MnIII ion (Mn1) in addition to bridging the CaII/MnII pair. These three prevalent carboxylate bridging modes found in 20 are illustrated in Figure 6-2. All Mn and the Ca2+ ions are sixand eight-coordina te, respectively; seven of th e eight CaO bonds are in the range 2.297.770 but the eighth is longer (Ca1O2 = 3.039 ). The geometry and symmetry of the eight oxygen atoms around the Ca is very unusual and asymmetric. The closest CaMn separation is 3.50 to Mn5. Complex 2114MeCNH2O crystallizes in the triclinic space group P 1 and comprises a Mn11Ca4 cluster (-2 charge) and two Mn4 cluster cations (+1 charge each). 2010MeCN Atom Mn2+ Mn3+ Mn4+ Mn(1) 3.27 3.00 3.14 Mn(2) 4.36 3.99 4.19 Mn(3) 1.59 1.45 1.53 Mn(4) 3.17 2.91 3.05 Mn(5) 3.14 2.88 3.02 Mn(6) Mn(7) 3.11 3.17 2.85 2.90 2.99 3.04 2010MeCN Atom Vi Assignment O(19) 1.80 MeOO(16) 1.89 O2O(32) 1.77 O2O(43) 1.88 O2O(5) 1.92 O2O(3) 1.15 HOO(13) O(17) 0.25 0.24 H2O H2O

PAGE 169

153 The two [Mn4O2(O2CPh)7(bpy)2]+ cations are the normal charged bpy butterfly complexes80 consisting of 4 MnIII ions, and will not be discussed henceforth. Instead, Figure 6-3. (Top) PovRay representation at th e 50% probability level of the X-ray crystal structure of the doubly charged anion of 21 (with the benzoate rings omitted for clarity, except for the ipso C atoms) and (bottom) its labeled [Mn11Ca4O10(OH)2(OMe)2]18+ core. Color scheme: MnIV dark blue, MnIII cyan, Ca yellow, O red, C grey. H atoms have been omitted for clarity. complex 21 will be referred to as the Mn11Ca4 complex hereafter. This Mn11Ca4 doubly charged motif is formulated as [Mn11Ca4O10(OH)2(OMe)2(O2CPh)20(H2O)2]2and is mixed-valent in the Mn component (MnIV, MnIII 10). The overall structure of this anion (Figure 6-3, top) is very similar to complex 20, the difference being that the Mn11Ca4 complex possesses two extra bridging benz oate groups and two fewer terminal H2O

PAGE 170

154 molecules. The most striking difference as seen in the core in Figure 6-3 (bottom), is the incorporation of two additional Ca2+ ions which replace the Mn2+ ions of 20. Among the benzoate groups, 18 are doubly bridging and the remaining two triply bridging. O8 in complex 21 is the hydroxide. All the Mn ions are hexa-coordinated, with the ligation in near octahedral geometry. Ca1 is sevencoordinate with the O atoms occurring in approximately pentagonal bipyramidal geometry whereas Ca2 is six-coordinated in near octahedral geometry. The closest CaMn se paration of 3.46 is between Ca2 and Mn4. Figure 6-4. (Top) PovRay representation at th e 50% probability level of the X-ray crystal structure of 22 (with the benzoate rings omitted for clarity, except for the ipso C atoms) and (bottom) its labeled [Mn8CaO6(O2P(Ph)2)4(OMe)2(MeOH)2]10+ core. The cluster is located on a two -fold rotation axis. Color scheme: MnIV dark blue, MnIII cyan, Ca yellow, O red, P green, C grey. H atoms have been omitted for clarity. Complex 22 crystallizes in th e monoclinic space group C 2/c with the Mn8Ca complex lying on a two-fold rotation axis. The complex

PAGE 171

155Table 6-2. Crystallographic data of 2010MeCN, 2114MeCNH2O, and 22. 20 21 22 Formula C148H136O54N10Mn13Ca2 C308H258O89N22Mn19Ca4 C162H144O38P12Mn8Ca fw, g/mol 3713.05 6895.73 3550.00 Space group P 1 P 1 C 2/c a, 15.0839(17) 18.179(2) 27.134(2) b, 16.3794(19) 19.160(2) 13.3809(11) c, 17.959(2) 23.003(3) 44.872(4) 112.343(2) 90.529(2) 90 103.301(2) 106.824(2) 94.4730(10) 92.272(2) 95.526(2) 90 V, 3 3953.9(8) 7627.9(14) 16242(2) Z 1 1 4 T, K 173(2) 173(2) 173(2) Radiation, a 0.71073 0.71073 0.71073 calc, g/cm3 1.559 1.501 1.451 mm-1 1.153 0.912 0.826 R1 b,c 0.0906 0.0879 0.0748 wR2 d 0.2168 0.1974 0.1633 a Graphite monochromator. b I > 2 ( I ). c R 1 = 100(|| Fo| | Fc||)/| Fo|. d wR 2 = 100[[ w ( Fo 2 Fc 2)2]/ [ w ( Fo 2)2]]1/2, w = 1/[2( Fo 2) + [( ap )2 + bp ], where p = [max ( Fo 2, O) + 2 Fc 2]/3.

PAGE 172

156 [Mn8CaO6(O2P(Ph)2)12(O2CPh)2(OMe)2(MeOH)2] is mixed-valent (2MnIV, 6MnIII) and centrosymmetric. The Mn ions form two cuba nes within which they are bridged by six 3-O2-, and two 3-MeOions (Figure 6-4, top). Periphera l ligation is provided by twelve doubly bridging diphenylphosphinates, and tw o doubly bridging benzoates (each bridges a MnIII and a MnIV ion). Terminal ligation is provided by two methanol molecules on the calcium ion. The structure can be described as two [Mn4O3(O2P(Ph)2)4(O2CPh)(OMe)] (MnIV, 3MnIII) cubes on either side of a central CaII ion. Each cube is linked to the Ca ion by two doubly bridging diphenylphosphinates (Fi gure 6-4, bottom). Thus, the Ca ion is six-coordinated and in near octahedral geom etry with Ca-O bond di stances in the range 2.288.396 All the Mn ions are also hexa-coo rdinated with the ox ide ligation in near octahedral geometry. The closest Ca Mn separation is 5.23 to Mn4. 6.2.3 Magnetochemistry of Complexes 20-22 Solid-state variable-temperature magnetic susceptibility measurements were performed on vacuum-dried, microc rystalline samples of complexes 20-22, restrained in eicosane to prevent torquing. Th e dc magnetic susceptibility ( M) data were collected in the 5.00 300 K range in a 1 kG (0.1 T) field, and they are plotted as MT vs. T in Figure 6-5 for 20 and the doubly charged anion of 21. The MT value for complex 20 at 300 K of 30.96 cm3 mol-1 K, is much lower than the 40.63 spin-only value (g = 2) expected for a cluster of MnIV, 10 MnIII, 2 MnII non-interacting ions, indicat ing extensive intramolecular antiferromagnetic interactions within 20. The MT steadily decreases with decreasing temperature to reach 19.06 at 100 K. Thereafte r, it falls sharply to a value of 3.47 cm3 mol-1 K at 5.00 K, suggesting a small ground state spin ( S ) in the range S = 5/2 1 for complex 20. For the anion of complex 21 (Mn11Ca4) the MT value is 70.30 cm3mol-1K at

PAGE 173

157 Temperature (K) 050100150200250300 MT (cm3mol-1K) 0 10 20 30 40 50 60 70 80 Mn 13 Ca 2 Mn 11 Ca 4 Figure 6-5. Plots of MT vs T for Mn13Ca2 ( ) and Mn11Ca4 ( ) complexes. 300 K and decreases linearly and uniformly (Figure 6-5) to re ach a value of 10.97 cm3mol-1K at 5.00 K, suggesting a sp in ground state in the range S = 7/2 1 for the Mn11Ca4 cluster anion. Owing to the size and low symmetry of the above-mentioned molecules, a matrix diagonalization met hod to determine the various exchange parameters ( Jij) between MniMnj pairs was clearly unfeasible We thus concentrated instead on identifying the ground state spin of the molecule, and this was accomplished for complex 20 by collecting variable-t emperature and variable -field magnetization ( M ) data in the 1.8 10 K and 0.5 5.0 T ranges; the data are plotted as reduced magnetization ( M/N B) versus H/T in Figure 6-6 for complex 20. The data were fit, using the program MAGNET,46 by diagonalization of the spin Hamiltonian matrix assuming only the ground state is populated, incorporating axial anisotropy ( D z 2) and Zeeman terms, and employing a full powder average. The corresponding Hamiltonian is given by eq 6-4, where D is the axial anisotropy constant, B is the Bohr magneton, z is the easy= D z 2 + g B0 Hz (6-4) axis spin operator, g is the electronic g factor, 0 is the vacuum permeability, and Hz is the applied longitudinal field. The last term in eq 6-4 is the Zeeman term associated with

PAGE 174

158 an applied magnetic field. The fit parameters were S = 5/2, g = 1.86, and D = -1.63 cm-1 for complex 20 and the fitting is shown in Figure 6-6. The D value of -1.63 cm-1 is consistent with the complex having predominantly MnIII ions and g < 2, as expected for Mn. When data collected at fields > 5.0 T we re included, a satisfact ory fit could not be obtained. We have found that poor qual ity fits of the magnetization versus H and T plots are a common problem in manganese chemistry when (i) the Mnx species is of high nuclearity and there is thus a high density of spin states resulting from the many exchange interactions present amongst the cons tituent Mn ions; and/or (ii) one or more MnII ions are present, which typically give very weak (and usua lly antiferromagnetic) exchange interactions and thus small ener gy separations. As a result, there are many H/T (kG/K) 051015202530 M/N 0 2 4 6 8 0.5 T 1 T 3 T 4 T 5 T fitting Figure 6-6. Magnetization (M) vs field (H) and temperature (T ) data, plotted as reduced magnetization (M/N B) vs H/T, for complex 20 at applied fields of 0.5, 1.0, 3.0, 4.0 and 5.0 T and in the 1.8 10 K temperature range. The solid lines are the fit of the data; see the text for the fit parameters. excited states that are low-lying (relative to kT ), and some of these will have S greater than the S of the ground state. Population of the ex cited states will thus be difficult to avoid even at the lowest temperatures normally employed. In addition, in the presence of a big enough applied DC field, Ms components of the excited state(s) can approach in

PAGE 175

159 energy the lowest-lying Ms of the ground state and even cross below it. The fitting procedure assumes only a single st ate is occupied, and thus an S value greater than the true ground-state S is given by the best fit of the da ta because it is affected by the contributions from the populated excited state( s). As a result, we t ypically observe that the best fit using all of the data collected over many field values will thus overestimate the M value at low fields and underestimate the M value at large fields. For this reason, we used for the fit of Figure 6-6 only the M data collected at small fields and as a result a satisfactory fit was obtained, with the fit parameters mentioned earlier. The results of the dc magnetic susceptibility studies on complex 22 are depicted in Figures 6-7 and 6-8. The MT value for complex 22 (Figure 6-8, top) at 300 K of 19.13 cm3mol-1K, is lower than the 21.75 spin-only va lue (g = 2) expected for a cluster comprising 2 MnIV, and 6 MnIII non-interacting ions, i ndicating the presence of intramolecular antiferromagnetic interactions. The MT value decreases marginally with decreasing temperature to 14.32 cm3 mol-1 K at 50 K and then falls steeply to finally reach a value of 8.83 cm3mol-1K at 5.00 K, suggesting a ground state spin ( S ) of ~ 9/2 1 for complex 22. In order to obtain the ground state spin for 22, variable-temperature and variable-field magnetization ( M ) data in the 1.8 10 K and 0.1 3.0 T ranges were collected; the data are plotted as reduced magnetization ( M/N B) versus H/T in Figure 6-7 for complex 22. The data were fit, by diagonalizati on of the spin Hamiltonian matrix assuming only the ground state is popul ated, incorporating axial anisotropy ( D z 2) and Zeeman terms, and employing a full powder average, as was done for complex 20. The fit parameters for 22 were S = 9/2, g = 1.94, and D = -0.22 cm-1 and the fitting is shown in Figure 6-7.

PAGE 176

160 H/T (kG/K) 051015 M/N 0 2 4 6 8 10 12 0.1 T 0.5 T 1 T 2 T 3 T Fitting Figure 6-7. Magnetization (M) vs field (H) and temperature (T ) data, plotted as reduced magnetization (M/N B) vs H/T, for complex 22 at applied fields of 0.1, 0.5, 1.0, 2.0 and 3.0 T and in the 1.8 10 K temperature range. The solid lines are the fit of the data; see the text for the fit parameters. Although the size of complex 22 is large, magnetically it can be simplified as consisting of two non-interacting C3v symmetric MnIVMnIII 3 cubanes. The closest MnMn separation between the two cubanes is 10.07 hence effectively they are magnetically non-interacting. Thus, the M value for each cubane is essentially half of the overall molar magnetization observed for the Mn8Ca complex. Under this assumption each cubane can be treated as a simplified model shown in Figure 6-8 (bottom) whereby there are only two ex change interactions; J33 (between 2 MnIII ions) and J34 (between a MnIII and a MnIV ion). This simplification of th e model has been discussed earlier121 for other [Mn4O3X]6+ complexes that consist of a Mn4 trigonal pyramid with a C3v symmetry. In order to determine the Mn2 pairwise exchange interactions within each of the cubanes, half (in magnitude) of the MT vs T data for complex 22 was fit to the appropriate theoretical expression. The virtual C3v symmetry of each cubane in the solid state of these compounds requires two exchange parameters ( J ). The isotropic Heisenberg spin Hamiltonian for the model shown in Figure 68 (bottom) is given in eq. 6-5, where J34

PAGE 177

161 and J33 refer to the exchange interactions for MnIIIMnIV and MnIIIMnIII pairs, respectively. The expression in eq 6-5 was further modifi ed by applying the Kambe vector coupling H = 2 J33( 2 3 + 2 4 + 3 4) 2 J34( 1 2 + 1 3 + 1 4) (6-5) Temperature (K) 050100150200250300 MT (cm 3 mol -1 K) 0 5 10 15 20 Each cubane Fitting Mn8Ca Mn1 IVMn2 IIIMn4 IIIMn3 IIIJ33J33J33J34J34J34 Mn1 IVMn2 IIIMn4 IIIMn3 IIIJ33J33J33J34J34J34 Figure 6-8. (Top) MT vs T plot for 22 and for each of the two cubanes of the Mn8Ca complex. The solid lines in the MT vs T plot are the fit of the data; see the text for the fit parameters. (Bottom) Model used to fit the exchange interactions within the cubes. method,45 as described elsewhere.122 This leads to the eigenvalue expression of eq 6-6, which gives the energy, E ( ST), of each of the possi ble total spin states, ST, of the complex, E ( ST) = J33[ SA( SA + 1)] J34[ ST( ST + 1) SA( SA + 1)] (6-6)

PAGE 178

162 where SA = S2 + S3 + S4 and ST = S1 + SA. In the above eqs. S2 = S3 = S4 = 2 for the MnIII centers, and S1 = 3/2 for the MnIV ion; the overall multiplicity of the spin system is 500, made up of individual spin states ranging from ST = 1/2 to 15/2. Thus, a theoretical MT vs T expression was derived using the ST values, their energies E ( ST), and the Van Vleck equation,123 and this expression was used to fit the experimental dc data (Figure 6-8, top). The best fit, shown as a solid line in Fi gure 6-8 (top), of the experimental points ( ) to the theoretical model was obtained with J33 = 24.6 cm-1, J34 = -21.6 cm-1, and g = 1.98, with temperature-independent paramagnetism (TIP) parameter held constant at 400 x 10-6 cm3mol-1. This leads to an ST = 9/2 ground state for this complex, since MnIIIMnIII pairwise interactions are ferromagnetic, and MnIIIMnIV exchange interactions antiferromagnetic. This is indeed satisfyingly consistent with the ST = 9/2 obtained from magnetization fits ( vide supra ), albeit the g = 1.94 differs slightly from the 1.98 value obtained here. The exchange interaction values obtained for 22 are similar to the ones reported in the literature where J33 and J34 fall in the ranges 5-11 and -21 to -34 cm-1, respectively.79 6.2.4 Calcium XAS studies of complexes 20-22 Complex 20-22 are the first molecular Ca/Mn complexes synthesized and provide a crystallographically characterized benchm ark whose data can be compared and contrasted with those of PS II preparations. Previously, the Mn cluster has been investigated extensively by XAS, EPR, etc. studies on many inorganic model complexes, as well as PS II.18 However, the Ca cofactor has been comparatively lesser researched; this can be justified by the lack of spect roscopic handles for Ca compared to Mn. Nevertheless, CaEXAFS25 as well as XANES124 have been perfor med on the WOC of PS II. However, such a study has never been extended to inorganic model complexes

PAGE 179

163 containing Ca/Mn for the simple reason that until now there were none such examples. XAS studies on biological Ca-containing samp les entail several daunting experimental challenges. First, compared with Mn or Sr, the X-ray fluorescence yield of Ca is lower making signal detection difficult. Second, the X-ray energies involve d (3.6-4.5 keV) are attenuated significantly by air and regular cryostat windows and othe r materials used in conventional hard X-ray EXAFS studies. Thir d, the low concentration of Ca in PS II ( 60 ppm by weight) means that for biochemical sample preparation, adventitious Ca contamination is a major problem. Extreme care must be taken to remove such extraneous Ca from glassware, solutions, and PS II samples, because nonspecific Ca would contribute to the EXAFS signal, dilu ting the desired signal from the WOC. Finally, biochemical studies have shown that hi gher plants have two types of Ca bound to PS II membrane fragments.125 One of these Ca seems to be bound tightly to the lightharvesting complex II (LHC II)126 and does not have any metals around it, while the other, more loosely bound Ca embedded in the WOC is important for oxygen evolution.26 Any Ca EXAFS experiment will measure the averaged signal from these two types. Despite these obstacles, Ca EXAFS had been conducted on PS II, the results of which indicated a CaMn separation of 3.4 .25 Analogous Ca-XAS data have now been collected on the Mn13Ca2, Mn11Ca4 and Mn8Ca complexes. Note that although the Mn nuclearity in these complexes is higher than the WOC, the range that Ca provides is comparable and amenable for analysis. Note also that even after X-ray crystallographic data21, 22, 26 on one Sn state (dark-adapted S1 state) becomes available at a resolution sufficient to define the precise structure of the WOC, studies on different Sn states using biophysical techniques such as EXAFS and XANES will be imperative to gain an

PAGE 180

164 understanding of the structural topological, and mechanis tic changes occurring during the catalytic cycle at the WOC as it shu ttles through its other, lesser stable Sn (n = 0,2,3,4) states.20 For this and other reasons, m odel compounds such as complexes 20-22 with precisely known structures are and will contin ue to be invaluable to assist in the interpretation of data obtained from XAS anal ysis, and to serve as reference species for physical characterization of the WOC in the various Sn states. 6.2.4.1 Calcium EXAFS of complexes 20-22 Figure 6-9 shows the Ca E XAFS spectra ((left) the k3-weighted spectra and (right) the Fourier transforms (FTs)), of complexes 20-22 and PS II in the S1 state (which contains two intact Ca per PS II sample). Fi gure 6-10 shows the Ca EXAFS spectra (the k3-weighted spectra in (A) and the FTs in (B)) of the Mn13Ca2 and Mn11Ca4 compounds and PS II in the S1 state. Note that in the Fourier Transforms, the Apparent Distance R is less than the actual distance by about 0.5 The spectra of only Mn13Ca2 and Mn11Ca4 model compounds are shown separately in Figure 6-10 since these were fit ( vide infra ), and also since the Mn8Ca compound does not contain any MnCa distances relevant to the WOC in PS II; CaMn distances in Mn8Ca are longer than 5.0 as already stated earlier. In the Fourier transfor ms of EXAFS, each peak indica tes a radial distribution of atoms surrounding the Ca atoms. The apparent distances are shorter than the actual Cabackscatterer distances due to an average phase shift induced by the potentials of the given absorber-backscatterer pair on the photoe lectron. The peak I region corresponds to the first shell of Ca-ligands (oxygen backscatterer for th e model compounds, and most likely oxygen and nitrogen backscatterer for PS II), and the peak II region is generally due to the CaMn interaction and contributions from second and third shell C or O atoms from the ligands (Figures 6-9 and 6-10). The actual distances ( R ) and the coordination

PAGE 181

165 numbers ( N ) are obtained from the EXAFS curve fitting summarized in Table 6-4 for Mn13Ca2 and Mn11Ca4 complexes, and compared with the XRD data. The EXAFS curve fitting quality was evaluated by the fit parameter, Ca-O Ca Mn Ca P I II I II Ca-O Ca Mn Ca P I II I II Figure 6-9. (Left) k3-weighted Ca K-edge EXAFS spect ra of the Ca/Mn complexes and PS II S1 state, and (right) Fourier tr ansforms of the EXAFS spectra. Although the Mn8Ca complex does not contain a ny CaMn distance below 5 we still see a broad shoulder in the peak II re gion of the FT spectrum (Figure 6-9) in the 3.5 to 4.0 range. This arises because of th e presence of CaP interactions. The P atoms are the second shell for the Ca ion, and ther e are four CaP separa tions, 2 each at 3.6 and 3.7 in complex 22. A list of the relevant bond and interatomic distances for complexes 20-22 is depicted in Table 6-3. For both Mn11Ca4 and Mn13Ca2, the fit quality

PAGE 182

166Table 6-3. Selected interatomic and bond distance s () for the Ca and Mn atoms of complexes 20-22. 20 21 22 Ca1-O10' 2.296(8) Ca1-O43' 2.339(7) Ca1-O42 2.346(8) Ca1-O1 2.364(9) Ca1-O22' 2.396(9) Ca1-O3 2.459(8) Ca1-O32' 2.771(8) Ca1-O2 3.039(7) Ca1Mn5 3.505(3) Mn3Ca1 3.634(3) Mn4Ca1' 3.796(3) Mn2Ca1 3.534(3) Mn1Mn2 2.8961(2) Mn5Mn7 2.822(2) Mn5Mn6 3.007(2) Mn4Mn6 3.016(3) Mn4Mn5 3.067(2) Mn4Mn7 3.199(3) Ca1-O14' 2.279(7) Ca1-O40' 2.299(6) Ca1-O10' 2.343(7) Ca1-O19 2.4077) Ca1-O37 2.410(6) Ca1-O2 2.478(10) Ca1-O1 2.598(10) Ca1Mn2' 3.541(2) Ca1Mn3' 3.602(2) Ca1Ca2 3.703(3) Ca2-O1 2.313(10) Ca2-O6 2.335(6) Ca2-O38 2.358(6) Ca2-O20 2.420(7) Ca2-O4 2.437(8) Ca2-O37 2.438(6) Ca2-O42 2.695(6) Ca2-O19 3.049(6) Ca-O12 2.288(6) Ca-O13 2.313(6) Ca-O3 2.397(6) Mn1-O18 1.822(5) Mn1-O19 1.860(5) Mn1-O17 1.883(5) Mn1-O2 1.937(6) Mn1-O7 1.943(6) Mn1-O5 1.956(5) CaP4 3.711(2) CaP5 3.593(2) CaMn3 5.540(2) CaMn4 5.238(2) Mn1Mn4 2.7673(19) Mn1Mn2 2.8580(18) Mn1Mn3 2.9693(19) Mn2Mn3 3.0530(19) Mn2Mn4 3.0860(19)

PAGE 183

167 Table 6-4. Least-squares fits of Fourier-f iltered peaks I and II of Ca EXAFS data compared with structural parameters from the X-ray crystal structures on complexes 20 and 21, and the WOC of Photosystem II in the S1 state. Fit No. Shell R (b) () N (a) 2 (2) /103 XRD (b) R / XRD N Mn13Ca2 I-1 Ca-O2.37 7.0 0.012 1.3 Peak I I-2 Ca-O Ca-O 2.35 2.83 6.0 1.0 0.010 0.001 0.84 Ca-O 2.3-2.5 ~2.8 6 1 II-1 Ca-Mn Ca-Mn 3.43 3.79 3.0 2.0 0.016 0.019 1.57 Peak II II-2 Ca-O Ca-Mn Ca-Mn 3.06 3.45 3.86 1.0 3.0 2.0 0.003 0.017 0.025 0.26 Ca-O ~3.0 Ca-Mn 3.5-3.6 3.8-3.9 1 3 2 Mn11Ca4 I-1 Ca-O 2.39 7.0 0.013 1.36 Peak I I-2 Ca-O Ca-O 2.37 2.85 6.0 1.0 0.011 0.002 0.89 Ca-O 2.3-2.5 ~2.8 6 1 II-1 Ca-Mn Ca-Mn 3.45 3.69 2.0 1.5 0.014 0.011 3.7 II-2 Ca-O Ca-MnCa-Mn 3.13 3.46 3.72 1.0 2.0 1.5 0.001 0.022 0.025 0.39 Peak II II-2 Ca-O Ca-MnCa-Ca Ca-Mn 3.07 3.42 3.68 3.76 1.0 2.0 1.0 1.5 0.004 0.012 0.005 0.015 0.47 Ca-O ~3.0 Ca-Mn 3.46-3.61 3.72-3.86 Ca-Ca ~3.7 1 2 1.5 1 Photosystem II S1 (c) Peak I Ca-O 2.4 5 6 Peak II Ca-Mn3.40 1.0 0.009 0.70 (a) Bold letter indicates fixed N from XRD data. (b) CaMn distances longer than 4.0 were not included for the fitting for both Mn13Ca2 and Mn11Ca4. (c) The data was taken from reference 25. Single value of 2 is used for all the interactions. 2, Debye-Waller parameter; fitquality parameter. For details see Appendix C, Physical Measurements. improves significantly when peak I wa s fit with two Ca-O distances ( R = ~2.4 and ~2.8 ) in accord with the XRD data. For Mn13Ca2, one broad FT peak is observed at R = 3.0

PAGE 184

168 in the peak II region, which is primarily th e mixture of two CaMn distances at ~3.5 and ~ 3.8 (see Table 6-4). On the other hand, there are no distinctive FT peaks observed in the spectrum from Mn11Ca4. This demonstrates a difficulty with EXAFS analysis when dealing with very disordered shells and there is no guiding crystal structure or any other supporting data (f or example, polarized EXA FS data to orientationally remove the distance heterogeneity). In the current case, the decreased intensity and asymmetry of peak II in Mn11Ca4 is explained by the large distance distribution of the Figure 6-10. (A) k3-weighted Ca K-edge EXAFS spectra of the Mn13Ca2 and Mn11Ca4 complexes and PS II S1 state, and (B) Fourier transform of the EXAFS spectra. CaMn vectors across the range of 3.5 to 4. 0 as observed in the XRD data. In fact, peak II of Mn11Ca4 fits well with the XRD data with a high Debye-Waller factor ( 2 value) of ~ 0.015 2 for the CaMn interaction. Additi on of the Ca-O co ntribution to the peak II region based on the XRD data significa ntly improved the curv e fitting quality for both Mn11Ca4 and Mn13Ca2 (see Table 6-4).

PAGE 185

169 Compared to the model compounds EXAFS spectra (Fig. 6-10), the presence of a clear single peak in PS II S1 state suggests a single CaMn distance. In fact, peak II in the PS II S1 state has been assigned to Ca Mn distances of 3.4 with N = 2 (after correcting for the stoichiometry of Ca in PS II; two Ca per PS II, one in LHC II and another in the WOC) for the Ca atoms in the WOC cluster. 6.2.4.2 Calcium XANES of complexes 20-22 The Ca K-edge XANES spectra of the Ca/Mn compounds and PS II in the S1 state (2Ca / PS II)124 are shown in Figure 6-11. The XANES spectra of Mn13Ca2 and Mn11Ca4 showed main edge features similar to that of PS II which has a maximum at 4050 eV, while more varied features (s houlder) are observed for the Mn8Ca complex. The differences in the edge shape are most likely due to differences in the symmetry, coordination numbers, and ligand environments around the Ca atoms in each of the three complexes. In all three Ca/Mn compounds, the first shell around Ca contains oxygen atoms. The Ca atoms in Mn8Ca are in a nearly pe rfect octahedral envir onment, while those of Mn13Ca2 are in asymmetric eight-coordinate geometry. Mn11Ca4 has a mixed environment, where one Ca (Ca1) is sevencoordinate and the O atoms occur in an approximately pentagonal bipyramidal geometry whereas Ca2 is hexa-coordinated and in near octahedral geometry. Since all thr ee compounds have similar first coordination shells, the difference in the main edge region (1s to 4p transition), especially the unique feature (shoulder) in the edge of Mn8Ca that is not present in Mn13Ca2 and Mn11Ca4, is most likely due to differences in the second shell. The second shell of Ca is Mn in Mn13Ca2 and Mn11Ca4, while it is mainly phosphorus atoms in Mn8Ca. The origin of the

PAGE 186

170 Figure 6-11. (Top) Ca K-edge XANE S spectra of Ca/Mn complexes 20-22 and PS II in the S1 state, and (bottom) their second de rivative spectra. (Inset) The pre-edge like peaks. The second derivative sp ectra were derived by analytical differentiation of a third-order polyno mial fit over a 3.0 eV range around each point. characteristic shoulder feature in the Ca XANES, similar to the one observed in Mn8Ca,

PAGE 187

171 was explained in the literature as a contributi on from a transition from Ca 1s to O 2p Ca 4p hybridized state.127 Surprisingly and quite unexpectedly, in the Ca K-edge XANES spectra, there is a pre-edge feature (Figure 611 (inset)) similar to those found in other open-shell 3d transition-metal species. The spectrum is unique for each compound, suggesting the strong influence of the different oxygen ligand geometries. Since Ca2+ has an empty 3d shell and the 1s to 3d transition is forbi dden unless there is some d-p mixing, the observed pre-edge features in the Ca/Mn comp ounds can be explaine d by Ca 3d orbitals gaining charge from O valence 2p orbitals.127, 128 Thus, the Ca XANES pre-edge position and intensity must be related to th e first shell oxygen configuration. In case of higher plant PS II, there are two types of Ca; one Ca which is part of the Mn4Ca cluster and a second Ca which is bound to the light harvesting complex (LHC II). The observed XANES spectrum of PS II is theref ore a superposition of spectral features from these two Ca atoms (hence 2 Ca per PS II). However, in spite of this, it is possible to observe that the PS II spectrum is dissi milar to the Ca/Mn model compounds. This suggests that Ca in the WOC in PS II might have a unique ligand environment, possibly from an oxo bridge as proposed. It would be interesting to collect Ca XANES from cyanobacterial PS II, which contains only one Ca per protein and ma y serve as a better comparison to the model compounds. Small energy shifts (Figure 6-11 (inset)) were also observed in the pre-edge region ; the energy at the top of the peak is ~ 4041 eV for the model compounds, while ~ 4040 eV for PS II. Fu rther investigation is necessary for understanding the relation between the Ca pr e-edge energy and the geometry of the neighboring oxygen conf iguration around Ca.

PAGE 188

172 6.3 Conclusions The successful entry and subsequent development of mixed-metal Ca/Mn chemistry pertinent to bioinorganic m odeling of the WOC of PS II has been demonstrated. Complexes 20-22 represent the first family of heterometallic Mn/Ca complexes. Complexes 20 and 22 possess a ground state spin of S = 3/2 and 9/2, respectively. Magneto-structural correlation provided the exchange-interaction between individual MnIIIMnIII and MnIIIMnIV pairs as 24.6 cm-1 and -21.6 cm-1 in complex 22. Although the overall nuclea rity of complexes 20-22 is larger than the Mn4Ca site of the WOC, the Ca provides an anal ogous range. Hence, the first application of Ca XAS to inorganic model complexes was achieved. Ca EXAFS revealed a close similarity between the local Ca environment of 20 and the WOC of PS II. In fact, the Ca-EXAFS results on 20, in conjunction with the Sr-EXAFS on the Mn14Sr complex 19, offer compelling evidence that there exists a single-atom O bridge between Mn and Ca/Sr in the WOC; these were the features proposed on basis of EXAFS studies of PS II. Of utmost importance is the observation that motifs in these complexes are also present in the most recently proposed structure of the Mn4Ca site of the WOC on th e basis of single-crystal EXAFS studies of PS II.129 Figure 6-12 lucidly illustrates the results of the Ca/Sr EXAFS, along with the germane Mn4Ca and Mn4Sr sub-units and the relevant Ca/SrMn interactions present in complexes 20 and 19, respectively. A very interesting and curious result of the Ca XANES of the model comp lexes was the observance of pre-edge like features. A re-examination of the XANES of th e WOC revealed that it too had pre-edge peaks. There is very clear dichroism of the main-edge and pre-edge peaks in the Mn13Ca2 complex and thus we are currently studying th is in greater detail with orientationdependent single-crystal XAS st udies. Additionally, efforts ar e underway to achieve the

PAGE 189

173 Figure 6-12. (Top) (A) Ca K-edge EXAFS spectra of the Mn13Ca2 complex (red) and PS II S1 state (black), and (B) Sr Kedge EXAFS spectra of the Mn14Sr complex (red) and Srreactivated PS II samples in the S1 state (black). Note that there are two Ca per PS II, including one extr a Ca which is bound to the harvesting complex. Also, note that the apparent distances ( R ) of the FT peaks are shorter by ~0.5 compared to the actu al distances because of a phase shift induced by the interaction of the given absorber-backscatterer pair with the photoelectron. (Bottom) Mn4Ca and Mn4Sr interactions present in complexes 20 and 19, containing the relevant Ca/SrMn distances to PS II. Color scheme: MnIV dark blue, MnIII cyan, MnII purple, Ca yellow, Sr green, O red. smaller nuclearity holy grail Mn4Ca complex in discrete form. Once the ideal pentanuclearity is achieved the research can be further fortified from speculative modeling to a true structural and functional bioinorganic approach, by application of other techniques such as EPR, catalytic oxyge n evolution activity testing, DFT, FTIR, and ENDOR to asses the similarities.

PAGE 190

174 All the complexes mentioned in this ch apter provide a diversity of Mn/Ca/O arrangements and topologies. A selected few geometries are illust rated in Figure 6-13. Many of the topologies are reminiscen t of the Mn topologies found in MnO2 minerals and Figure 6-13. PovRay representations at the 50% probability level of units / sub-units present in the Mn/Ca complexes 20-22, and in the Mn14Sr complex 19. Subunits of complex 20 in the first row resemble the 3.0 26 (left) and 3.5 22 (center) crystal structures of the WOC. Color scheme: MnIV dark blue, MnIII cyan, MnII purple, Ca yellow, Sr green, O red, C grey. which were proposed as plausi ble evolutionary origin of the WOC of PS II by Sauer and Yachandra.130 Obviously, in the absence of a conclusive crystal structure21, 22, 23, 26 or one particular preferred topology based on biophysical studies,131 researchers will have to

PAGE 191

175 keep an open-mind towards this bioinorganic research of the active site of PS II. In retrospect, what Mother Natu re has achieved through billio ns of years of evolution cannot be replicated in the lab within a coupl e of years. Neverthele ss, the WOC of PS II is an intriguing and interesting research area, especially with the recent crystal structure reports becoming available. Thus with rene wed research vigour in this field, PS II definitely promises much more new science as it further matures. 6.4 Experimental 6.4.1 Syntheses All manipulations were performed under aerobic conditions us ing chemicals as received, unless otherwise stated. (NBun 4)[Mn4O2(O2CPh)9(H2O)] was prepared as described elsewhere.55 [Mn13Ca2O10(OH)2(OMe)2(O2CPh)18(H2O)4] (2010MeCN). To a stirred solution of (NBun 4)[Mn4O2(O2CPh)9(H2O)] (0.25 g, 0.16 mmol) in MeCN/MeOH (20/1 mL) was slowly added Ca(NO3)2H2O (0.01 g, 0.04 mmol), and this produced a brown solution. This was stirred for a further 20 minutes, filtere d, and the filtrate slowly concentrated by evaporation at ambient temperature, which slowly produced dark orange crystals of 2010MeCN over a couple of days. The yield was 40%. The crystals of 20 were maintained in the mother liquor for X-ra y crystallography and other single-crystal studies, or collected by filtration, washed with MeCN, and dried in vacuo The synthesis can also be carried out with Ca(ClO4)2 or Ca(O2CPh)2. Vacuum-dried solid analyzed as solvent-free. Elemental analysis (%) calcd. for 20 (C128H106O54Mn13Ca2): C, 46.55; H, 3.24; found: C, 46.43; H, 3.15. Selected IR data (KBr, cm-1): 3475(br), 1600(m), 1548(m), 1408(s), 1384(s), 1177( w), 1069(w), 1025(w), 716(s), 677(w), 613(m, br), 508(w).

PAGE 192

176 [Mn4O2(O2CPh)7(bpy)2]2[Mn11Ca4O10(OH)2(OMe)2(O2CPh)20(H2O)2] (214MeCNH2O). To a stirred solution of (NBun 4)[Mn4O2(O2CPh)9(H2O)] (0.25 g, 0.16 mmol) and Ca(O2CPh)2H2O (0.06 g, 0.18 mmol) in MeCN/MeOH (20/1 mL) was slowly added 2, 2'-bipyridyl (in short bpy) (0.03g, 0.19 mmol). The resulting brown solution was stirred for 20 minutes, filtered, and the filtrate concentrated by evaporation at ambient temperature to yield orange crystals of 2114MeCNH2O over a week. The yield was 35%. The crystals of 21 were washed with MeCN, and dried in vacuo Vacuum-dried solid analyzed as fully desolvated. Elemental analysis (%) calcd. for 21 (C280H214N8O88Mn19Ca4): C, 53.36; H, 3.42; N, 1.78; f ound: C, 53.35; H, 3.35; N, 1.91. Selected IR data (KBr, cm-1): 3375(br), 1602(m), 1558(m), 1400(s, br), 1175(w), 1069(w), 1025(w), 772(w), 716(s), 688( w), 676(w), 612(m, br), 473(w). [Mn8CaO6(O2P(Ph)2)12(O2CPh)2(OMe)2(MeOH)2] (22). To a stirred solution of (NBun 4)[Mn4O2(O2CPh)9(H2O)] (0.25 g, 0.16 mmol) and Ca(O2CPh)2H2O (0.03 g, 0.08 mmol) in MeCN/MeOH (20/2 mL) was slow ly added diphenylphosphinic acid (0.31g, 1.42 mmol) which turned the solution to a light orange color. The resulting solution was stirred for 20 minutes, filtered, and the filtrate concentrated by evaporation at ambient temperature to yield orange crystals of 22 over a couple of days. The yield was 55%. The crystals of 22 were washed with MeCN, and dried in vacuo Vacuum-dried solid analyzed as solvent-free. Elem ental analysis (%) calcd. for 22 (C162H144O38P12Mn8Ca): C, 54.81; H, 4.09; found: C, 54.73; H, 4.07. Selected IR data (KBr, cm-1): 3408(w), 1592(w), 1548(w), 1438(m), 1368(m), 1130(s, br ), 1010(s), 991(s), 753(m), 726(s), 693(s), 556(s, br), 489(w).

PAGE 193

177 6.4.2 X-ray Crystallography Data were collected on a Siemens SM ART PLATFORM equipped with a CCD area detector and a graphite monochromator utilizing Mo K radiation ( = 0.71073 ). Suitable crystal of 200MeCN, 214MeCNH2O and 22 were attached to a glass fiber using silicone grease and transferred to a goni ostat where they were cooled to 173 K for data collection. An initial search of r eciprocal space revealed a triclinic cell for 20 and 21, and a monoclinic cell for 22; the choice of space groups P 1, P 1 and C 2/c, respectively, were confirmed by the subsequent solution and refinement of the structures. Cell parameters were refined using up to 8192 reflections. A full sphere of data (1850 frames) was collected using the -scan method (0.3 frame widt h). The first 50 frames were re-measured at the end of data collecti on to monitor instrument and crystal stability (maximum correction on I was < 1 %). Abso rption corrections by integration were applied based on measured indexed crystal f aces. The structures were solved by direct methods in SHELXTL6,56a and refined on F2 using full-matrix least squares. The non-H atoms were treated anisotropically, wher eas the hydrogen atoms were placed in calculated, ideal positions and refined as riding on their respective carbon atoms. The asymmetric unit of 2010MeCN contains half the cluster and 5 MeCN molecules of crystallization. The latter were disordered and could not be modeled properly, thus program SQUEEZE,56b a part of the PLATON56c package of crystallographic software, was used to calcula te the solvent disorder area and remove its contribution to the overall intensity data. A to tal of 884 parameters were refined in the final cycle of refinement us ing 4146 reflections with I > 2 (I) to yield R1 and wR2 of 9.06 and 21.68%, respectively.

PAGE 194

178 The asymmetric unit of 2114MeCNH2O consists of a half Mn11Ca4 cluster (-1 charge), a Mn4 cluster cation (+1 charge), and seven MeCN and a half H2O molecules of crystallization. The solvent molecules we re disordered and c ould not be modeled properly, thus program SQUEEZE,56b a part of the PLATON56c package of crystallographic software, was used to calcula te the solvent disorder area and remove its contribution to the overall intensity data. A to tal of 1751 parameters were refined in the final cycle of refinement usi ng 42241 reflections with I > 2 (I) to yield R1 and wR2 of 8.79 and 19.74%, respectively. The asymmetric unit of 22 consists of a half Mn8Ca cluster. There are no solvent molecules. The structure is best described as two Mn4 moieties joined by a Ca center. There are four disordered phenyl rings, one on P4, one on P5 and both of the phenyl rings on P5'. A total of 960 parameters were refine d in the final cycle of refinement using 10624 reflections with I > 2 (I) to yield R1 and wR2 of 7.48 and 16.33%, respectively. 6.4.3 XAS Studies X-ray absorption spectra were measured on beamline 10.3.2 at the Advanced Light Source (ALS) operating at an electron energy of 1.9 GeV with an average current of 300 mA. The experiments were conducted by Dr. Junko Yano and Dr Yulia Pushkar, members of the Yachandra group in the P hysical Biosciences Division at Lawrence Berkeley Laboratory, Berkeley, CA. The ra diation was monochro matized by a Si(111) double-crystal monochromator. Intensity of the incident X-ray was monitored by an N2filled ion chamber (I0) in front of the sample. XAS samples were made by gently grinding ~1 mg of compound and sealing in a sample holder with Mylar tape. Fluorescence spectra were recorded by us ing a seven-element Ge detector array.132 Energy was calibrated by the pre-edge peak of KMnO4 (6543.3 eV) for Mn XAS, and by

PAGE 195

179 the edge peak of calcium acetate (4050 eV) fo r Ca XAS. All spectra were collected at room temperature. Each XANES and EXA FS scan required 17 and 30 minutes to complete. From a radiation damage study on each sample prior to data collection, it was considered safe to collect three scans pe r spot for XANES and two scans per spot for EXAFS with a spot si ze of 100 (H) 10 (V) m2 using a defocused beam. Spectra were background-subtracted, normalized, and analyzed as previously described in the literature.112 After conversion of backgr ound-corrected spectra from energy space to photoel ectron wave vector ( k ) space, and weighted by k3, a four-domain spline was subtracted for a final b ackground removal. Appendix C (Physical Measurements) contains details regard ing fitting procedures employed.

PAGE 196

180 APPENDIX A BOND DISTANCES AND ANGLES Table A-1. Selected interatomic distances () and angles () for [Mn8CeO8(O2CMe)12(H2O)4] (4) and [Mn8CeO8(O2CMe)12(py)4] (5) Complex 4 Complex 5 Mn1-O3 1.85(2) Mn1-O1 1.92(2) Mn1-O18 1.94(3) Mn1-O5 1.94(3) Mn1-O2 2.18(3) Mn1-O4 2.28(3) Mn2-O3 1.86(2) Mn2-O1 1.87(2) Mn2-O11 1.93(3) Mn2-O17 1.98(3) Mn2-O6 2.18(2) Mn2-O4 2.26(2) Ce1-O3 2.33(19) Ce1-O1 2.37(2) Mn2-O4-Mn1 83.9(8) Mn4-O8-Mn3 83.9(8) Mn4-O9-Mn3 117.8(14) Mn1-O2 1.867(3) Mn1-O1 1.874(3) Mn1-O6 1.963(4) Mn1-O7 1.973(3) Mn1-O3 2.150(4) Mn1-O4 2.343(4) Mn2-O2 1.847(3) Mn2-O1 1.872(3) Mn2-O8 1.977(3) Mn2-O5 1.980(3) Mn2-O4 2.245(3) Mn2-N1 2.326(4) Ce1-O2 2.322(3) Ce1-O1 2.369(3) Mn2-O4-Mn1 82.42(11) Mn2-O2-Mn1 121.70(18) Mn2-O2-Ce1 105.45(14)

PAGE 197

181 Table A-2. Selected interatomic distances () and angles () for [Mn8CeO8(O2CCHPh2)12(H2O)4] (7). Ce(1)-O(24) 2.319(7) Ce(1)-O(5) 2.321(7) Ce(1)-O(14) 2.337(8) Ce(1)-O(33) 2.342(8) Ce(1)-O(9) 2.395(7) Ce(1)-O(28) 2.398(7) Ce(1)-O(19) 2.401(8) Ce(1)-O(3) 2.413(8) Ce(1)-Mn(5) 3.344(2) Ce(1)-Mn(2) 3.347(2) Ce(1)-Mn(6) 3.351(2) Ce(1)-Mn(4) 3.352(2) Mn(1)-O(5) 1.859(8) Mn(1)-O(3) 1.888(7) Mn(1)-O(6) 1.964(9) Mn(1)-O(2) 1.984(9) Mn(1)-O(4) 2.221(9) Mn(1)-O(1) 2.243(9) Mn(1)-Mn(8) 3.007(3) Mn(1)-Mn(2) 3.225(3) Mn(2)-O(5) 1.857(7) Mn(2)-O(9) 1.882(8) Mn(2)-O(8) 1.964(8) Mn(2)-O(10) 1.982(8) Mn(2)-O(7) 2.143(9) Mn(2)-O(11) 2.298(8) Mn(2)-Mn(3) 3.032(3) Mn(3)-O(14) 1.847(8) Mn(3)-O(9) 1.893(8) Mn(3)-O(15) 1.975(9) Mn(3)-O(12) 1.979(8) Mn(3)-O(13) 2.232(8) Mn(3)-O(11) 2.283(8) Mn(3)-Mn(4) 3.196(3) Mn(4)-O(14) 1.858(8) Mn(4)-O(19) 1.891(8) Mn(4)-O(17) 1.958(9) Mn(4)-O(18) 1.987(9) Mn(4)-O(16) 2.167(8) Mn(4)-O(21) 2.292(8) Mn(4)-Mn(5) 3.048(3) Mn(6)-O(24) 1.852(7) Mn(6)-O(28) 1.891(8) Mn(6)-O(27) 1.957(8) Mn(6)-O(26) 1.960(8) Mn(6)-O(25) 2.134(9) Mn(6)-O(29) 2.262(9) Mn(6)-Mn(7) 3.034(3) Mn(7)-O(33) 1.834(7) Mn(7)-O(28) 1.896(8) Mn(7)-O(30) 1.951(8) Mn(7)-O(32) 1.969(9) Mn(7)-O(31) 2.230(8) Mn(7)-O(29) 2.270(8) Mn(7)-Mn(8) 3.199(3) Mn(8)-O(33) 1.872(8) Mn(8)-O(3) 1.876(8) Mn(8)-O(36) 1.958(9) Mn(8)-O(35) 1.979(8) Mn(8)-O(34) 2.146(8) Mn(8)-O(1) 2.273(9) O(5)-Ce(1)-O(3) 65.1(2) O(14)-Ce(1)-O(3) 159.8(2) O(33)-Ce(1)-O(3) 65.6(3) O(9)-Ce(1)-O(3) 115.6(3) O(28)-Ce(1)-O(3) 114.0(2) O(19)-Ce(1)-O(3) 98.8(3) O(24)-Ce(1)-Mn(5) 32.04(19) O(5)-Ce(1)-Mn(5) 112.94(19) O(14)-Ce(1)-Mn(5) 79.8(2) O(33)-Ce(1)-Mn(5) 131.45(18) Mn(5)-Ce(1)-Mn(6) 57.21(5) Mn(2)-Ce(1)-Mn(6) 164.21(5) Mn(5)-Ce(1)-Mn(4) 54.16(5) Mn(2)-Ce(1)-Mn(4) 91.67(5) Mn(6)-Ce(1)-Mn(4) 92.66(5) Mn(8)-Mn(1)-Ce(1) 63.33(5) Mn(2)-Mn(1)-Ce(1) 61.03(5) Mn(1)-Mn(2)-Ce(1) 61.51(5)

PAGE 198

182 Table A-3. Selected interatomic distances () and angles () for [Mn11Gd4O8(OH)8(O2CPh)16(NO3)5(H2O)3] (10). Gd1-O20 2.361(6) Gd1-O21 2.364(5) Gd1-O22 2.410(6) Gd1-O10 2.417(6) Gd1-O24' 2.438(6) Gd1-O28 2.501(6) Gd1-O26 2.524(5) Gd1-O5 2.524(5) Gd1-O4 2.551(5) Gd1-N2 2.909(7) Gd1-Mn3 3.5085(13) Gd1-Mn2 3.5498(5) Gd2-O14 2.341(6) Gd2-O25 2.415(5) Gd2-O12 2.431(6) Gd2-O9 2.436(5) Gd2-O29 2.447(8) Gd2-O8 2.456(5) Gd2-O32 2.462(8) Gd2-O33 2.483(8) Gd2-O31 2.492(8) Gd2-N3 2.777(17) Gd2-N1 2.874(10) Gd2-Mn5 3.4401(13) Mn1-O3 1.887(5) Mn1-O4 1.897(5) Mn1-O23' 1.951(6) Mn1-O2 1.968(5) Mn1-O1 2.238(6) Mn1-O24' 2.298(6) Mn1-Mn2 2.9324(13) Mn2-O3' 1.900(5) Mn2-O3 1.900(5) Mn2-O4 1.983(5) Mn2-O4' 1.983(5) Mn2-O5 2.173(5) Mn2-O5' 2.173(5) Mn2-Mn1' 2.9324(13) Mn2-Gd1' 3.5498(5) Mn3-O10 1.848(6) Mn3-O4 1.923(5) Mn3-O6 1.982(6) Mn3-O9 1.989(5) Mn3-O7 2.120(6) Mn3-O8 2.201(5) Mn3-Mn6 2.8975(17) Mn3-Mn4 3.0991(18) Mn4-O3' 1.875(5) Mn4-O8 1.878(5) Mn4-O11 1.919(6) Mn4-O12 1.950(5) Mn4-O13 2.164(6) Mn4-O10 2.468(5) Mn6-O10 1.854(5) Mn6-O16 1.930(6) Mn6-O17 1.945(6) Mn6-O9 1.985(6) Mn6-O18 2.200(6) Mn6-O12 2.260(5) O20-Gd1-O21 90.84(19) O20-Gd1-O22 76.3(2) O21-Gd1-O22 73.86(18) O20-Gd1-O10 152.1(2) O21-Gd1-O10 72.66(18) O22-Gd1-O10 77.4(2) O28-Gd1-N2 26.36(19) O26-Gd1-N2 25.1(2) O5-Gd1-N2 145.1(2) O4-Gd1-N2 108.8(2) O20-Gd1-Mn3 168.36(12) O4-Mn1-Mn2 42.03(15) O23'-Mn1-Mn2 136.88(17) O2-Mn1-Mn2 139.39(18) O1-Mn1-Mn2 94.47(16) O24'-Mn1-Mn2 88.21(15) O3'-Mn2-O3 180.0(4) O3'-Mn2-O4 101.1(2)

PAGE 199

183 Table A-4. Selected interatomic distances () and angles () for [Mn11Dy4O8(OH)6(OMe)2(O2CPh)16(NO3)5(H2O)3] (11). Dy1-O2' 2.336(4) Dy1-O4 2.349(4) Dy1-O6 2.373(3) Dy1-O5 2.385(3) Dy1-O8' 2.408(3) Dy1-O3 2.466(4) Dy1-O1 2.483(4) Dy1-O12' 2.504(3) Dy1-O7 2.556(3) Dy1-N1 2.906(5) Dy1-Mn3 3.4908(8) Dy1-Mn2 3.5413(3) Dy2-O26 2.303(4) Dy2-O30 2.34(2) Dy2-O27 2.354(8) Dy2-O34 2.357(4) Dy2-O16 2.383(4) Dy2-O17 2.406(3) Dy2-O21 2.407(3) Dy2-O33 2.43(3) Dy2-O28 2.453(8) Dy2-O29 2.464(9) Dy2-O32 2.51(2) Dy2-O31 2.558(17) Mn1-O13' 1.891(3) Mn1-O7 1.895(3) Mn1-O10 1.943(4) Mn1-O9 1.944(4) Mn1-O11 2.225(4) Mn1-O8' 2.290(4) Mn1-Mn2 2.9411(8) Mn2-O13 1.893(3) Mn2-O13' 1.893(3) Mn2-O7 2.003(3) Mn2-O7' 2.003(3) Mn2-O12' 2.158(3) Mn2-O12 2.158(3) Mn2-Mn1' 2.9411(8) Mn2-Dy1' 3.5413(3) Mn5-O13 1.879(3) Mn5-O17 1.883(3) Mn5-O22 1.939(4) Mn5-O21 1.947(4) Mn5-O23' 2.177(4) Mn5-O5 2.466(4) Mn3-Mn4 2.8798(11) Mn3-Mn5 3.1128(11) Mn4-O5 1.848(4) Mn4-O20 1.931(4) Mn4-O19 1.944(4) Mn4-O16 1.994(4) Mn4-O18 2.204(4) Mn4-O21 2.251(4) O4-Dy1-O1 72.25(13) O6-Dy1-O1 129.61(13) O5-Dy1-O1 124.74(13) O8'-Dy1-O1 73.62(13) O3-Dy1-O1 51.37(12) O2'-Dy1-O12' 74.14(12) O4-Dy1-O12' 140.82(12) O6-Dy1-O12' 68.22(12) O2'-Dy1-Mn3 169.03(9) O4-Dy1-Mn3 99.05(9) O6-Dy1-Mn3 99.29(9) O5-Dy1-Mn3 30.48(8) O8'-Dy1-Mn3 90.75(9) O3-Dy1-Mn3 70.48(9) O7-Mn1-O8' 79.36(14) O10-Mn1-O8' 95.04(15) O9-Mn1-O8' 85.90(15) O11-Mn1-O8' 167.76(14) O13'-Mn1-Mn2 39.01(10) O12'-Mn2-Dy1 44.36(9) O12-Mn2-Dy1 135.64(9) Mn1-Mn2-Dy1 68.594(17 O15-Mn3-Mn4 141.35(12) O14-Mn3-Mn4 97.09(11) O17-Mn3-Mn4 86.77(10) O5-Mn3-Mn5 52.40(11)

PAGE 200

184 Table A-5. Selected interatomic distances () and angles () for [Mn11Tb4O8(OH)6(OCH2Ph)3(O2CPh)20(PhCH2OH)2(H2O)] (13) Tb1-O3' 2.374(8) Tb1-O29 2.375(7) Tb1-O27 2.392(7) Tb1-O31 2.406(8) Tb1-O1 2.445(8) Tb1-O5' 2.447(8) Tb1-O11 2.490(7) Tb1-O2 2.491(8) Tb1-O10' 2.632(8) Tb1-Mn2 3.5235(19) Tb1-Mn3 3.5573(6) Tb2-O20 2.324(8) Tb2-O17 2.340(9) Tb2-O16 2.361(8) Tb2-O24 2.396(10) Tb2-O19 2.411(7) Tb2-O26 2.412(8) Tb2-O14 2.413(8) Tb2-O32 2.426(12) Tb2-Mn6 3.4147(19) Tb2-Mn5 3.4172(18) Tb2-Mn4 3.487(2) Tb2-Mn2 3.5321(19) Mn1-O10 1.891(7) Mn1-O6 1.893(7) Mn1-O4 1.936(7) Mn1-O9 1.937(7) Mn1-O12 2.246(9) Mn1-O5 2.300(9) Mn1-Mn3 2.9088(16) Mn3-Tb1' 3.5573(6) Mn4-O27 1.875(7) Mn4-O14 1.931(8) Mn4-O18 1.939(8) Mn4-O30 1.945(8) Mn4-O7 2.215(9) Mn4-O26 2.275(8) Mn4-Mn5 3.239(2) Mn4-Mn5 3.239(2) Mn5-O19 1.868(8) Mn5-O6 1.870(7) Mn5-O26 1.933(7) Mn5-O28 1.947(8) Mn5-O25 2.161(8) Mn5-O27 2.396(7) Mn6-O19 1.911(8) Mn6-O20 1.915(8) Mn6-O21 1.934(8) Mn6-O11' 1.963(8) Mn6-O22 2.149(8) Mn6-O13' 2.469(8) Mn4-O21 2.251(4) O3' -Tb1-O29 73.8(3) O3' -Tb1-O27 148.7(3) O29-Tb1-O27 76.5(3) O3' -Tb1-O31 88.6(3) O27-Tb1-Mn2 30.06(18) O31-Tb1-Mn2 98.7(2) O1-Tb1-Mn2 123.2(2) O5' -Tb1-Mn2 89.05(19) O11-Tb1-Mn2 95.36(17) Mn6-Tb2-Mn5 62.72(4) O20-Tb2-Mn4 130.4(2) O17-Tb2-Mn4 63.4(2) Mn6-Tb2-Mn2 57.30(4) Mn5-Tb2-Mn2 52.44(4) Mn4-Tb2-Mn2 48.05(4) O10-Mn1-Mn3 42.2(2) O6-Mn1-Mn3 39.8(2) O4-Mn1-Mn3 136.5(2) O9-Mn1-Mn3 139.5(2) Mn4-Mn2-Mn5 66.10(6) O27-Mn2-Tb1 39.5(2) O10'-Mn2-Tb1 47.2(2) Tb1-Mn2-Tb2 130.13(5)

PAGE 201

185 Table A-6. Selected interatomic distances () and angles () for [Mn10Th6O22(OH)2(O2CPh)16(NO3)2(H2O)8] (16). Th1-O3 2.376(14) Th1-O7 2.393(12) Th1-O25 2.425(17) Th1-O11 2.455(14) Th1-O8 2.479(13) Th1-O6 2.501(19) Th1-O2 2.514(14) Th1-O5 2.538(17) Th1-O4 2.54(2) Th3-O14 2.354(14) Th3-O32 2.384(16) Th3-O16 2.397(13) Th3-O15 2.424(11) Th3-O29 2.484(16) Th3-O13 2.491(13) Th3-O11 2.547(13) Th3-O17 2.584(15) Th3-O18 2.59(2) Th3-O12 2.655(12) Mn1-O3 1.802(13) Mn1-O2 1.807(15) Mn1-O1 1.820(16) Mn1-O24 1.909(16) Mn1-O21 1.945(17) Mn1-O22 1.984(14) Mn3-O7 1.828(14) Mn3-O16' 1.831(13) Mn3-O12 1.856(14) Mn3-O12' 1.900(13) Mn3-O13 1.928(13) Mn3-O27 1.970(16) Mn4-O10 1.800(15) Mn4-O15 1.855(15) Mn4-O11 1.869(11) Mn4-O28 1.899(18) Mn4-O8 1.929(17) Mn4-O30 1.960(12) Mn5-O15 1.809(15) Mn5-O14 1.831(11) Mn5-O10 1.862(15) Mn5-O34 1.889(16) Mn5-O33 1.943(16) Mn5-O31 2.027(14) Mn4Mn5 2.702(6) Mn1Mn2 2.695(4) Mn3Mn3' 2.821(7) Mn2Mn3 2.763(5) Mn4Mn5 2.702(6) Th2Mn5 3.362(3) Th2Mn1 3.420(4) Th2Mn4 3.489(3) Th3Mn5 3.348(3) Th3Mn4 3.400(3) Mn3Th3' 3.460(3) O2-Th1-Mn4 133.4(4) Mn1-Th1-Mn4 104.58(10) O3-Th1-Mn2 66.4(3) O11-Th1-Mn2 88.0(3) O8-Th1-Mn2 134.6(4) Mn1-Th1-Mn2 46.35(8) Mn4-Th1-Mn2 115.92(9) O10-Th2-O1 140.0(5) Mn5-Th2-Mn4 46.40(10) Mn1-Th2-Mn4 100.92(9) Mn5-Th3-Mn4 47.19(10) Mn2-Mn1-Th1 70.16(11) Mn2-Mn1-Th2 73.42(12) Th1-Mn1-Th2 69.06(7) Mn1-Mn2-Mn3 127.94(15) Mn2-Mn3-Th3 125.43(11) Mn3'-Mn3-Th2 77.49(12) Th3'-Mn3-Th2 106.87(9) Th3-Mn3-Th2 63.24(5) Mn3-O13-Th3 107.9(6) O11-Th1-Mn1 97.7(3) O8-Th1-Mn1 98.9(3) O6-Th1-Mn1 140.7(4)

PAGE 202

186 Table A-7. Selected interatomic distances () and angles () for [Mn7O5(OMe)2(O2CPh)9(terpy)] (17). Mn1-O3 2.048(2) Mn1-O2 2.163(2) Mn1-O4 2.167(2) Mn1-N2 2.243(3) Mn1-N1 2.268(3) Mn1-N3 2.273(3) Mn2-O3 1.842(2) Mn2-O17 1.925(1) Mn2-O5 1.954(2) Mn2-O18 1.984(2) Mn2-O8 2.205(2) Mn2-O13 2.222(2) Mn3-O3 1.840(2) Mn3-O17 1.936(1) Mn3-O1 1.942(2) Mn3-O12 1.987(2) Mn3-O6 2.212(2) Mn3-O10 2.231(2) Mn4-O12 1.892(2) Mn4-O19 1.910(2) Mn4-O11 1.928(2) Mn4-O25 1.946(2) Mn4-O24 2.138(2) Mn4-O17 2.334(2) Mn5-O18 1.894(2) Mn5-O14 1.912(2) Mn5-O20 1.920(2) Mn5-O19 1.938(2) Mn5-O15 2.123(2) Mn5-O17 2.439(2) Mn6-O25 1.913(2) Mn6-O7 1.931(2) Mn6-O9 1.933(2) Mn6-O20 1.939(2) Mn6-O22 2.193(2) Mn6-O17 2.300(2) Mn7-O19 1.840(2) Mn7-O25 1.847(2) Mn7-O20 1.860(2) Mn7-O16 1.931(2) Mn7-O21 1.941(2) Mn7-O23 1.959(2) Mn4Mn7 2.7908(7) Mn4Mn5 3.1124(7) Mn4Mn6 3.1740(7) Mn5Mn7 2.7894(7) Mn5Mn6 3.1997(7) Mn6Mn7 2.8127(7) Mn2Mn3 2.7898(7) Mn2Mn5 3.0361(7) Mn3Mn4 2.9958(7) O3-Mn1-O2 89.96(8) O3-Mn1-O4 92.21(8) O2-Mn1-O4 177.83(8) O3-Mn1-N2 168.47(9) O2-Mn1-N2 81.30(9) O4-Mn1-N2 96.54(9) O3-Mn1-N1 100.48(9) O2-Mn1-N1 87.96(10) O4-Mn1-N1 91.63(9) N2-Mn1-N1 71.85(10) O3-Mn1-N3 116.08(9) O2-Mn1-N3 89.93(10) N2-Mn1-N3 71.69(10) N1-Mn1-N3 143.38(10) O3-Mn2-O17 84.57(8) O3-Mn2-O5 98.74(9) O17-Mn2-O5 176.68(9) O9-Mn6-Mn4 151.07(7) O20-Mn6-Mn4 78.69(6) O22-Mn6-Mn4 121.82(6) O17-Mn6-Mn4 47.22(5) Mn7-Mn6-Mn4 55.172(16) Mn7-Mn6-Mn5 54.827(15) Mn4-Mn6-Mn5 58.458(15)

PAGE 203

187 Table A-8. Selected interatomic distances () and angles () for [Mn7O5(OCH2Ph)2(O2CPh)9(terpy)] (18). Mn7-O23 2.035(2) Mn7-O25 2.170(3) Mn7-O22 2.192(3) Mn7-N2 2.249(4) Mn7-N3 2.259(4) Mn7-N1 2.276(5) Mn3-O13 1.887(3) Mn3-O4 1.912(3) Mn3-O10 1.923(3) Mn3-O7 1.926(3) Mn3-O6 2.105(3) Mn3-O12 2.407(3) Mn4-O14 1.896(3) Mn4-O3 1.906(3) Mn4-O7 1.924(3) Mn4-O16 1.934(3) Mn4-O9 2.161(3) Mn4-O12 2.366(3) Mn2-O18 1.910(3) Mn2-O4 1.914(3) Mn2-O20 1.915(3) Mn2-O3 1.940(3) Mn2-O2 2.140(3) Mn2-O12 2.307(2) Mn1-O4 1.856(3) Mn1-O7 1.860(3) Mn1-O3 1.867(3) Mn1-O8 1.926(3) Mn1-O1 1.960(3) Mn1-O5 1.960(3) Mn5-O23 1.846(3) Mn5-O12 1.931(3) Mn5-O21 1.964(3) Mn5-O20 1.994(3) Mn5-O15 2.223(3) Mn5-O19 2.283(3) Mn5-Mn6 2.8013(9) Mn6-O23 1.849(3) Mn6-O12 1.938(2) Mn6-O24 1.949(3) Mn6-O13 1.994(3) Mn6-O17 2.191(3) Mn6-O11 2.222(3) Mn1Mn3 2.7787(8) Mn1Mn2 2.8028(9) Mn1Mn4 2.8090(8) Mn2Mn5 3.0088(8) Mn2Mn3 3.0624(9) Mn2Mn4 3.1528(9) Mn3Mn6 3.0020(9) Mn3Mn4 3.2031(9) Mn5Mn6 2.8013(9) O23-Mn7-O22 90.90(11) O25-Mn7-O22 177.29(13) O23-Mn7-N2 172.24(13) O25-Mn7-N2 82.71(12) O22-Mn7-N2 95.62(12) O23-Mn7-N3 104.08(13) O25-Mn7-N3 90.89(14) O22-Mn7-N3 90.61(13) N2-Mn7-N3 71.73(16) Mn1-O3-Mn2 94.81(11) Mn4-O3-Mn2 110.10(11) Mn1-O4-Mn3 95.04(12) Mn1-O4-Mn2 96.04(12) Mn3-O4-Mn2 106.35(12) O21-Mn5-O19 84.19(11) O20-Mn5-O19 82.96(11) O15-Mn5-O19 167.31(11) O23-Mn5-Mn6 40.73(8) O2-Mn2-Mn5 145.42(8) O12-Mn2-Mn5 39.91(6) Mn1-Mn2-Mn5 126.85(3) O18-Mn2-Mn3 93.07(10) O4-Mn2-Mn3 36.80(8) Mn1-Mn2-Mn4 55.91(2) Mn5-Mn2-Mn4 73.08(2) Mn3-Mn2-Mn4 62.02(2) Mn1-Mn3-Mn2 57.10(2) Mn6-Mn3-Mn2 86.27(2) O13-Mn3-Mn4 85.71(9)

PAGE 204

188 APPENDIX B LIST OF COMPOUNDS {[Mn(OH)(O2CMe)2]MeCO2HH2O}n (1) [Mn3O(O2CPh)6(py)2(H2O)] (2) (NBun 4)[Mn4O2(O2CPh)9(H2O)] (3) [Mn8CeO8(O2CMe)12(H2O)4] (4) [Mn8CeO8(O2CMe)12(py)4] (5) [Mn8CeO8(O2CPh)12(MeCN)4][Mn8CeO8(O2CPh)12(dioxane)4] (6) [Mn8CeO8(O2CCHPh2)12(H2O)4] (7) [Mn11Nd4O8(OH)8(O2CPh)16(NO3)5(H2O)3] (8) [Mn11Eu4O8(OH)8(O2CPh)16(NO3)5(H2O)3] (9) [Mn11Gd4O8(OH)8(O2CPh)16(NO3)5(H2O)3] (10) [Mn11Dy4O8(OH)6(OMe)2(O2CPh)16(NO3)5(H2O)3] (11) [Mn11Ho4O8(OH)8(O2CPh)18(NO3)3(H2O)7] (12) [Mn11Tb4O8(OH)6(OCH2Ph)3(O2CPh)20(PhCH2OH)2(H2O)] (13) [Mn2Yb2O2(O2CPh)6(OMe)4(MeOH)4] (14) [Mn2Y2O2(O2CPh)6(OMe)4(MeOH)4] (15) [Mn10Th6O22(OH)2(O2CPh)16(NO3)2(H2O)8] (16) [Mn7O5(OMe)2(O2CPh)9(terpy)] (17) [Mn7O5(OCH2Ph)2(O2CPh)9(terpy)] (18) [Mn14SrO11(OMe)3(O2CPh)18(MeCN)2] (19) [Mn13Ca2O10(OH)2(OMe)2(O2CPh)18(H2O)4] (20)

PAGE 205

189 [Mn4O2(O2CPh)7(bpy)2]2[Mn11Ca4O10(OH)2(OMe)2(O2CPh)20(H2O)2] (21) [Mn8CaO6(O2P(Ph)2)12(O2CPh)2(OMe)2(MeOH)2] (22)

PAGE 206

190 APPENDIX C PHYSICAL MEASUREMENTS Infrared spectra were recorded in the so lid state (KBr pellets) on a Nicolet Nexus 670 FTIR spectrophotome ter in the 400-4000 cm-1 range. Elemental analyses (C, H, and N) were performed at the in-house facilities of the University of Florida Chemistry Department. Variable-temperature DC magnetic susceptibility data down to 1.80 K were collected on a Quantum Design MPMS-XL SQUID magnetometer equipped with a 70 kG (7 T) DC magnet at the University of Florida. Pascals constants were used to estimate the diamagnetic corrections, which were subtracted from the experimental susceptibility to give the mola r magnetic susceptibility ( M). AC magnetic susceptibility data were collected on the same instrume nt employing a 3.5 G fi eld oscillating at frequencies up to 1500 Hz. Samples were embe dded in solid eicosane, unless otherwise stated, to prevent torquing. Magnetization vs fi eld and temperature data were fit using the program MAGNET, and contour plots were obtained using the program GRID, both written at Indiana University by Dr. E. R. Davidson. Low temp erature (< 1.8 K) hysteresis loop and DC relaxa tion measurements were perf ormed at Grenoble using an array of micro-SQUIDS.50 The high sensitivity of this magnetometer allows the study of single crystals of SMMs on the order of 10-500 m. The field can be applied in any direction by separately dr iving three orthogonal coils. X-ray absorption spectroscopy (XAS) was pe rformed at the Stanford Synchrotron Radiation Laboratory (SSRL) on beamline 9-3 at an electron energy of 3.0 GeV with an average current of 70-90 mA. Synchrotron radiation facilities were provided by the

PAGE 207

191 Advanced Light Source (ALS), Berkeley, whic h is operated by the De partment of Energy, Office of Basic Energy Sciences. Data reduction of the EXAFS spectra was performed as described earlier.25, 106, 112 Curve fitting was performed using ab initio -calculated phases and amplitudes from the program FEFF 8.133 These ab initio phases and amplitudes were used in the EXAFS equation: ( k )S0 2 N jkRj 2 jfeffj(, k Rj) e2j 2k2e2 Rj/j( k )sin(2 kRjij( k )) The neighboring atoms to the centr al atom(s) are divided into j shells, with all atoms with the same atomic number and dist ance from the central atom grouped into a single shell. Within each she ll, the coordination number Nj denotes the number of neighboring atoms in shell j at a distance of Rj from the central atom. feffj( k Rj) is the ab initio amplitude function for shell j and the Debye-Waller term e 2j 2k2 accounts for damping due to static and thermal disorder in absorber-backscatterer distances. The mean free path term e2 R j / j( k ) reflects losses due to inelastic scattering, where j(k) is the electron mean free path. The oscillations in the EXAFS spectrum are reflected in the sinusoidal term sin(2 kRj ij( k )) where ij( k ) is the ab initio phase function for shell j This sinusoidal term shows the direct re lation between the frequency of the EXAFS oscillations in k space and the absorber-backscatterer distance. The EXAFS equation was used to f it the experimental data using N R and 2 as variable parameters. The coordination numbers, N are evaluated on a per Sr/Mn/Ca basis. Fit quality was evaluated using two different fit parameters, and 2. is a normalized sum of residuals between the data and the simulations. The 2 error takes into account the

PAGE 208

192 number of variable parameters. S0 2 is an amplitude reduction factor due to shakeup/shake-off processes at the central atom(s). Fit-quality is: 1 si 2[expt( ki) calc( ki)]2 1 NT where NT is the total number of data points collected, expt( ki) is the experimental EXAFS amplitude at point i and calc( ki) is the theoretical EXAFS amplitude at point i The normalization factor si is given by: 1 si ki 3kj 3expt( kj) j N The 2 error takes into account the nu mber of variable parameters p in the fit and the number of independent data points Nind, as shown in: 2 NindNind p N1 N is the total number of data points colle cted, and the number of independent data points Nind is estimated from the Nyquist sampling theorem, as shown in Nind 2 k R k is the k -range of the data and R is the width of the Fourier-filtered peak in Therefore, 2 provides a gauge of whet her the addition of another shell to the fit is justified.

PAGE 209

193 LIST OF REFERENCES (1) (a) Boca, R. Theoretical Foundations of Molecular Magnetism; Elsevier Science: Amsterdam, 1999; pp 360. (b) Kahn, O. Molecular Magnetism; VCH Publishers: New York, 1993; pp 287-328. (2) Smart, L.; Moore, E. Solid State Chemistry: An Introduction ; Chapman and Hall: New York, 1992; pp 294-320. (3) Epstein, A. J. MRS Bulletin 2000, 25 33. (4) Awaga, K.; Coronado, E.; Drillon, M. MRS Bulletin 2000, 25 52. (5) Miller, J. S.; Epstein, A. J. MRS Bulletin 2000, 25 21 (6) Hosk; B. F.; Martin, R.L. Nature 1966, 627. (7) (a) Sessoli, R.; Tsai, H.L.; Schake, A.R.; Wang, S.; Vincent, J.B.; Folting, K.; Gatteschi, D.; Christou, G.; Hendrickson, D.N. J. Am. Chem. Soc. 1993, 115 1804. (b) Sessoli, R.; Gatteschi, D. ; Novak, M.A.; Caneschi, A. Nature 1993, 356 141. (c) Thomas, L.; Lionti, L.; Ballou, R.; Gatteschi, D.; Sessoli, R.; Barbara, B. Nature 1996, 383, 145. (8) (a) Sessoli, R. Mol. Cryst. Liq. Cryst. 1995, 274 145. (b) Cheesman, M. R.; Oganesyan, V. S.; Sessoli, R.; Gatteschi, D.; Thomson, A. J. Chem. Commun. 1997, 17 1677. (c) Ruiz, D.; Aubin, S. M. J.; Rumberger, E.; Incarvito, C. D.; Folting, K.; Rheingold, A. L.; Ch ristou, G.; Hendrickson, D. N. Mol. Cryst. Liq. Cryst. 1999, 335 413. (9) (a) Christou, G.; Gatteschi, D.; Hendrickson, D.N.; Sessoli, R. MRS Bulletin 2000, 25 66. (b) Christou, G. Polyhedron 2005, 24 2065. (c) Aromi, G.; Aubin, S. M. J.; Bolcar, M. A.; Christou, G.; Eppley, H. J.; Folting, K.; Hendrickson, D. N.; Huffman, J. C.; Squire, R. C.; Tsai, H. L.; Wang, S.; Wemple, M. W. Polyhedron 1998, 17 3005. (10) Gatteschi, D.; Sessoli, R. Angew. Chem. Intl. Ed. 2003, 42 268. (11) Eppley, H.J.; Tsai, H.L.; de Vries, N. ; Folting, K.; Hendrickson, D.N.; Christou, G. J. Am. Chem. Soc. 1995, 120 301. (12) Wernsdorfer, W.; Aliaga-Alcalde N.; Hendrickson, D.N.; Christou, G. Nature 2002, 416 406.

PAGE 210

194 (13) Hill, S.; Edwards, R. S.; Aliaga-Alcalde, N.; Christou, G. Science 2003, 302 1015. (14) Jones, J. A. Science 1998, 280 229. (15) Wernsdorfer, W., Sessoli, R. Science 1999, 284 133. (16) Pecoraro, V. L. Manganese Redox Enzymes ; Ed.; Verlag Chemie: Weinheim, 1992; and references cited therein. (17) Wu, A.J.; Penner-Hahn, J.E.; Pecoraro, V.L. Chem. Rev. 2004, 104 903. (18) (a) Christou, G., Acc. Chem. Res. 1989, 22 328. (b) Yachandra, V. K.; Sauer, K.; Klein, M. P., Chem. Rev. 1996, 96 2927. (c) Mukhopadhyay, S.; Mandal, S. K.; Bhaduri, S.; Armstrong, W. H., Chem. Rev. 2004, 104 3981. (d) Yachandra, V. K.; Sauer, K. Biochim. Biophys. Acta. 2004, 1655 140 (19) Kok, B.; Forbrush, B.; McGloin, M. Photochem. Photobiol. 1970, 11 457. (20) Haumann, M.; Liebisch, P.; Muller C.; Barra, M.; Grabolle, M.; Dau, H. Science 2005, 310 1019. (21) (a) Zouni, A.; Witt, H. T.; Kern, J.; Fr omme, P.; Krauss, N.; Saenger, W.; Orth, P. Nature 2001, 409 739. (b) Kamiya, N.; Shen, J. R., Proc. Natl. Acad. Sci. USA 2003, 100 98. (22) (a) Ferreira, K. N.; Iverson, T. M.; Maghlaoui, K.; Barber, J.; Iwata, S. Science 2004, 303 1831 (b) Iwata, S.; Barber, J. Curr. Opin. Struct. Biol. 2004, 14 447. (23) (a) Yano, J.; Kern, J.; Irrgang, K.; Latimer, M. J.; Bergmann, U.; Glatzel, P.; Pushkar, Y.; Biesiadka, J.; Loll, B.; Sauer, K.; Messinger, J.; Zouni, A.; Yachandra, V. K. Proc. Natl. Acad. Sci. USA 2005, 102 12047. (b) Dau, H.; Liebisch, P.; Haumann, M. Phys. Chem. Chem. Phys. 2004, 6 4781. (24) (a) Debus, R. J. Biochim. Biophys. Acta. 1992, 1102 269. (b) Ruttinger, W.; Dismukes, G. C. Chem. Rev. 1997, 97 1. (25) Cinco, R. M.; Holman, K. L. M.; Robblee, J. H.; Yano, J.; Pizarro, S. A.; Bellacchio, E.; Sauer, K.; Yachandra, V. K. Biochemistry 2002, 41 12928. (26) Loll, B.; Kern, J.; Saenger, W.; Zouni, A.; Biesiadka, J. Nature 2005, 438 1040. (27) (a) Sauer, K.; Yano, J.; Yachandra, V. K., Photosynthesis Research 2005, 85 73. (b)Yano, J.; Pushkar, Y.; Glatzel, P.; Lewis, A.; Sauer, K.; Messinger, J.; Bergmann, U.; Yachandra, V. K. J. Amer. Chem. Soc. 2005, 127 14974. (28) Ibers, J. A.; Holm, R. H. Science 1980, 209 223.

PAGE 211

195 (29) (a) Sessoli, R.; Gatteschi, D.; Caneschi, A.; Novak, M. A. Nature 1993, 365 141. (b) Sessoli R.; Ysai, H. -L.; Schake, A. R. ; Wang, S.; Vincent, J. B.; Folting, K.; Gatteschi, D.; Christou, G.; Hendrickson, D. N. J Am. Chem. Soc. 1993, 115 1804. (c) Thomas, L.; Lionti, L.; Ballou, R.; Gatteschi, D.; Sessoli, R.; Barbara, B. Nature 1996, 383 145. (30) (a) Eppley, H. J.; Tsai, H.-L.; de Vrie s, N.; Folting, K.; Christou, G.; Hendrickson, D. N. J. Am. Chem. Soc 1995, 117 301. (b) Aubin, S. M. J.; Spagna, S.; Eppley, H. J.; Sager, R. E.; Christou, G.; Hendrickson, D. N. Chem. Commun. 1998, 803. (c) Aubin, S. M. J.; Sun, Z.; Pardi, L.; Krzystek, J.; Folting, K.; Brunel, L. -C.; Rheingold, A. L.; Christou, G.; Hendrickson, D. N. Inorg. Chem. 1999, 38 5329. (d) Soler, M.; Chandra, S. K.; Ruiz, D. ; Davidson, E. R.; Hendrickson, D. N.; Christou, G. Chem. Commun. 2000, 2417. (e) Boskovic, C.; Pink, M.; Huffman, J. C.; Hendrickson, D. N.; Christou, G. J. Am. Chem. Soc. 2001, 123 9914. (31) (a) Miyasaka, H.; Clrac, R.; Wern sdorfer, W.; Lecren, L.; Bonhomme, C.; Sugiura, K.; Yamashita, M. A. Angew. Chem. Int. Ed 2004, 43 2801. (b) Milios, C. J.; Raptopoulou, C. P.; Terzis, A.; Lloret F.; Vicente, R.; Perlepes, S. P.; Escuer, A. Angew. Chem. Int. Ed 2003, 43 210. (c) Brechin, E. K.; Soler, M.; Davidson, J.; Hendrickson, D. N.; Parsons, S.; Christou, G. Chem. Commun, 2002, 2252. (d) Price, D. J.; Batten, S. R.; Moubaraki B.; Murray, K. S. Chem. Commun 2002, 762. (e) Brechin, E. K.; Boskovic, C.; Wernsdorfer, W.; Yoo, J.; Yamaguchi, A.; Sanado, E. C.; Concolino, T. R.; Rheingol d, A. L.; Ishimoto, H.; Hendrickson, D. N.; Christou, G. C. J. Am. Chem. Soc. 2002, 124 9710. (f) Sanudo, E. C.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. Inorg. Chem 2004, 43 4137. (g) Tasiopoulos, A. J.; Vinslava, A.; Wernsdor fer, W.; Abboud, K. A.; Christou, G. Angew. Chem. Int. Ed 2004, 43 2117. (h) Tasiopoulos, A. J.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. Angew. Chem. Int. Ed 2004, 43 6338. (i) Zaleski, C. M.; Depperman, E. C.; Dendrinou-Samara C.; Alexiou, M.; Kampf, J. W.; Kessissoglou, D. P.; Kirk, M. L.; Pecoraro, V. L. J. Am.Chem.Soc. 2005, 127 12862. (32) (a) Brechin, E. K.; Yoo, J.; Huffman, J. C.; Hendrickson D. N.; Christou, G. Chem Commun. 1999, 783. (b) Yoo, J.; Brechin, E. K.; Yamaguchi, A.; Nakano, M.; Huffman, J. C.; Maniero, A. L.; Brunel, L.-C.; Awaga, K.; Ishimoto, H.; Christou; G.; Hendrickson, D. N. Inorg. Chem. 2000, 39 3615. (c) Yoo, J.; Yamaguchi, A.; Nakano, M.; Krzystek, J.; Streib, W. E.; Brunel, L. -C.; Ishimoto, H.; Christou G.; Hendrickson, D. N. Inorg. Chem. 2001, 40 4604. (d) Lecren, L.; Wernsdorfer, W.; Li, Y.-G.; Roubeau, O.; Miyasaka, H.; Clerac, R. J. Am.Chem. Soc. 2005, 127 11311. (e) Lecren, L.; Roubeau, O.; Coulon, C.; Li, Y.-G.; Le Goff, X. F.; Wernsdorfer, W.; Miyasaka, H.; Clerac, R. J. Am. Chem. Soc. 2005, 127 17353. (33) (a) Aubin, S. M. J.; Wemple, M. W.; Adams, D. M.; Tsai, H. -L.; Christou, G.; Hendrickson, D. N. J. Am. Chem. Soc. 1996, 118 7746. (b) Aubin, S. M.; Gilley, N. R.; Pardi, L.; Krzystek, J.; Wemple, M. W.; Brunel, L. C.; Marple, M. B.; Christou, G.; Hendrickson, D. N. J. Am. Chem. Soc. 1998, 120 4991.

PAGE 212

196 (34) (a) Foguet-Albiol, D.; O'Brien, T. A.; Wernsdorfer, W.; Moulton, B.; Zaworotko, M. J.; Abboud, K. A.; Christou, G. Angew. Chem. Int. Ed., 2005, 44 897. (b) Rumberger, E. M.; Shah, S. J.; Beedle, C. C.; Zakharov, L. N.; Rheingold, A. L.; Hendrickson, D. N. Inorg. Chem. 2005, 44 2742. (35) (a) Sokol, J. J.; Hee, A. G.; Long, J. R. J. Am. Chem. Soc. 2002, 124 7656. (b) Schelter, E. J.; Prosvirin, A. V.; Dunbar, K. R. J. Am. Chem. Soc. 2004, 126 15004. (c) Osa, S.; Kido, T.; Matsumoto, N.; Re, N.; Pochaba, A.; Mrozinski, J. J. Am. Chem. Soc. 2004, 126 420. (d) Zaleski, C. M.; Depperman, E. C.; Kampf, J. W.; Kirk, M. L.; Pecoraro, V. L. Angew. Chem. Int. Ed. 2004, 43 3912. (e) Oshio, H.; Nihei, M.; Koizumi, S.; Shiga, T.; Nojiri, H.; Nakano, M.; Shirakawa, N.; Akatsu, M.; J. Am. Chem. Soc. 2005, 127 4568. (f) Costes, J.-P.; Dahan, F.; Wernsdorfer, W. Inorg. Chem. 2006, 45 5. (g) Mori, F.; Nyui, T.; Ishida, T.; Nogami,T.; Choi,K.; Nojiri; H. J. Am. Chem. Soc. 2006, 128 1440. (36) Tasiopoulos, A. J.; Wernsdorfer, W.; Mou lton, B.; Zaworotko, M. J.; Christou, G. J. Am. Chem. Soc. 2003, 125 15274. (37) Mishra, A.; Wernsdorfer, W. ; Abboud, K. A.; Christou, G. J. Am. Chem. Soc. 2004, 126 15648. (38) Mishra, A.; Wernsdorfer, W.; Parsons, S.; Christou, G.; Brechin, E. K. Chem. Commun 2005, 2086. (39) Murugesu, M.; Mishra, A.; Wernsdor fer, W.; Abboud, K. A.; Christou, G. Polyhedron 2006, 25 613. (40) (a) Tasiopoulos, A. J.; OBrien, T. A.; Abboud, K. A.; Christou, G. unpublished results. (b) Tasiopoulos, A. J.; OBrien, T. A.; Abboud, K. A.; Christou, G. Angew. Chem. Int. Ed 2004, 43 345. (41) Tasiopoulos, A. J.; Harden, N. C.; Abboud, K. A.; Christou, G. Polyhedron 2003, 22 133. (42) (a) Brechin, E. K. Chem Commun. 2005, 5141. (b) Piligkos, S.; Rajaraman, G.; Soler, M.; Kirchner, N.; Slageren, J.; Bircher, R.; Parsons, S.; Gudel, H.-U.; Kortus, J.; Wernsdorfer, W.; Christou, G.; Brechin, E. K. J. Am. Chem. Soc. 2005, 127 5572.(c) Boskovic, C.; Wernsdorfer, W.; Folting, K.; Huffman, J. C.; Hendrickson, D. N.; Christou, G. Inorg. Chem. 2002, 41 5107.(d) Brechin, E. K.; Boskovic, C.; Wernsdorfer, W.; Yoo, J.; Yamaguchi, A.; Sanudo, E. C.; Concolino, T. R.; Rheingold, A. L.; Ishimoto, H.; Hendrickson, D. N.; Christou, G. J. Am. Chem. Soc. 2002, 124 9710. (43) Mishra, A.; Wernsdorfer, W. ; Abboud, K. A.; Christou, G. Chem. Commun 2005, 54.

PAGE 213

197 (44) (a) Bond valence sum calculations for the crystallographically independent Mn,44b and Ce44c ions of 4 gave oxidation state values of 2.84 3.03 and 3.73 3.96 respectively. (b) Liu, W.; Thorp, H. H. Inorg. Chem. 1993, 32 4102. (c) Roulhac, P. L.; Palenik, G. J. Inorg. Chem. 2003, 42 118. (45) Kambe, K. J. Phys. Soc. Jpn 1950, 48 15. (46) Davidson, E. R. MAGNET Indiana University. (47) Murugesu, M.; Habrych, M.; Werns dorfer, W.; Abboud, K. A.; Christou, G. J. Am. Chem. Soc. 2004, 126 4766. (48) (a) King, P.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. Inorg. Chem 2005, 44 8659. (b) Foguet-Albiol, D.; Abboud, K. A.; Christou, G. Chem. Commun. 2005, 4282. (c) Mishra, A.; Abboud, K. A.; Christou, G. Inorg. Chem 2006, 45 2364. (49) Chakov, N. E.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. Inorg. Chem 2004, 43 5919. (50) Wernsdorfer, W. Adv. Chem. Phys 2001, 118 99 (51) Goodwin, J. C.; Sessoli, R.; Gatteschi, D. ; Wernsdorfer, W.; Powell, A. K.; Heath, S. L. J. Chem. Soc., Dalton Trans. 2000, 1835. (52) Soler, M.; Wernsdorfer, W.; Folting, K.; Pink, M.; Christou, G. J. Am. Chem. Soc. 2004, 126 2156. (53) Tasiopoulos, A. J.; Harden, N. C.; Abboud, K. A.; Christou, G. Polyhedron 2003, 22 133. (54) Vincent, J. B.; Chang, H. R.; Folting, K.; Huffman, J. C.; Christou, G.; Hendrickson, D. N. J. Am. Chem. Soc. 1987, 109 5703. (55) Wemple, M. W.; Tsai, H.-L.; Wang, S.; Cl aude, J.-P.; Streib, W. E. ; Huffman, J. C.; Hendrickson, D. N.; Christou, G. Inorg. Chem. 1996, 35 6450. (56) (a) Sheldrick, G. M. SHELXTL6 ; Bruker-AXS: Madison, WI, 2000. (b) van der Sluis, P.; Spek, A. L. Acta Crystallogr. 1990, A46 194. (c) Spek, A. L. Acta Crystallogr., Sect. A 1990, 46, C34.

PAGE 214

198 (57) (a) Christou, G.; Gatteschi, D. ; Hendrickson, D. N.; Sessoli, R. MRS Bulletin 2000, 25 66 and references therein. (b) Piligkos, S.; Rajaraman, G.; Soler, M.; Kirchner, N.; van Slageren, J.; Bircher, R.; Pa rsons, S.; Gudel, H.-U.; Kortus, J.; Wernsdorfer, W.; Christou, G.; Brechin, E. K. J. Am. Chem. Soc. 2005, 127 5572. (c) Stamatatos, T. C.; Foguet-Albiol, D. ; Stoumpos, C. C.; Raptopoulou, C. P.; Terzis, A.; Wernsdorfer, W.; Perlepes, S. P.; Christou, G. J. Am. Chem. Soc. 2005, 127 15380. (d) King, P.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. Inorg. Chem. 2005, 44 8659. (e) Zaleski, C. M.; Depperman, E. C.; Dendrinou-Samara, C.; Alexiou, M.; Kampf, J. W.; Kessissoglou, D. P.; Kirk, M. L.; Pecoraro, V. L. J. Am. Chem. Soc. 2005, 127 12862. (f) Oshio, H.; Hoshino, N.; Ito, T.; Nakano, M. J. Am. Chem. Soc. 2004, 126 8805. (g) Bell, A.; Arom, G.; Teat, S. J.; Wernsdorfer, W.; Winpenny, R. E. P. Chem. Commun. 2005, 2808. (h) Brechin, E. K.; Soler, M.; Christou, G.; Helliwell, M.; Teat, S.J.; Wernsdorfer, W. Chem. Commun 2003, 1276 (58) (a) Schelter, E. J.; Pros virin, A. V.; Dunbar, K. R. J. Am. Chem. Soc. 2004, 126 15004. (b) Oshio, H.; Nihei, M.; Koizumi, S.; Shiga, T.; Nojiri, H.; Nakano, M.; Shirakawa, N.; Akatsu, M.; J. Am. Chem. Soc. 2005, 127 4568. (c) Sokol, J. J.; Hee, A. G.; Long, J. R. J. Am. Chem. Soc. 2002, 124 7656. (59) (a) Osa, S.; Kido, T.; Matsumoto, N.; Re, N.; Pochaba, A.; Mrozinski, J. J. Am. Chem. Soc. 2004, 126 420. (b) Mishra, A.; Werns dorfer, W.; Abboud, K. A.; Christou, G. J. Am. Chem. Soc. 2004, 126 15648. (c) Zaleski, C. M.; Depperman, E. C.; Kampf, J. W.; Kirk, M. L.; Pecoraro, V. L. Angew. Chem. Int. Ed. 2004, 43 3912. (d) Mishra, A.; Wernsdorfer, W.; Parsons, S.; Christou, G.; Brechin, E. K. Chem. Commun 2005, 2086. (e) Costes, J.-P.; Auchel M.; Dahan, F.; Peyrou, V.; Shova, S.; Wernsdorfer, W. Inorg. Chem. 2006, 45 1924. (f) Mori, F.; Nyui, T.; Ishida, T.; Nogami,T.; Choi,K.; Nojiri; H. J. Am. Chem. Soc. 2006, 128 1440. (g) Murugesu, M.; Mishra, A.; Wernsdorfe r, W.; Abboud, K. A.; Christou, G. Polyhedron 2006, 25 613. (h) Costes, J.-P.; Dahan, F.; Wernsdorfer, W. Inorg. Chem. 2006, 45 5. (60) (a) Ishikawa, N.; Sugita, M.; Wernsdorfer, W. J. Am. Chem. Soc. 2005, 127 3650. (b) Ishikawa, N.; Sugita, M.; Wernsdorfer, W. Angew. Chem. Int. Ed. 2005, 44 2931. (61) (a) Murugesu, M.; Habrych, M.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. J. Am. Chem. Soc. 2004, 126 4766. (b) Foguet-Albiol, D.; O'Brien, T. A.; Wernsdorfer, W.; Moulton, B.; Zaworotko, M. J.; Abboud, K. A.; Christou, G. Angew. Chem. Int. Ed., 2005, 44 897. (c) Brechin, E. K. Chem Commun. 2005, 5141. (62) (a) Tasiopoulos, A. J.; OBrien, T. A.; Abboud, K. A.; Christou, G. Angew.Chem. Int.Ed 2004, 43 345. (b) Tasiopoulos, A. J.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. Inorg. Chem 2005, 44 6324. (c) Price, D. J.; Batten, S. R.; Moubaraki, B.; Murray, K. S. Chem. Commun. 2002, 762. (d) King, P.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. Inorg. Chem 2005, 44 8659.

PAGE 215

199 (63) Bond valence sum (BVS) calculations for the MnIII and DyIII ions of 6 gave values in the range 2.88 2.92 and 2.89 2.90, re spectively. Also, for the six hydroxides the range was 1.05-1.29. For the methoxide (O26) the value was 1.76. (a) Brown, I. D.; Altermatt, D. Acta. Crystallogr. Sect. B 1985, 41 244. (b) Liu, W.; Thorp, H. H. Inorg. Chem. 1993, 32 4102. (c) Palenik, G. J. Inorg. Chem. 1997, 36 4888. (64) Benelli, C.; Murrie, M.; Parsons, S.; Winpenny, R. E. P. Dalton Trans. 1999, 4125. (65) Piligkos, S.; Rajaraman, G.; Soler, M.; Kirchner, N.; van Slageren, J.; Bircher, R.; Parsons, S.; Gudel, H.-U.; Kortus, J.; Wern sdorfer, W.; Christou, G.; Brechin, E. K. J. Am. Chem. Soc. 2005, 127 5572 (66) Soler, M.; Wernsdorfer, W.; Sun, Z.; Hu ffman, J. C.; Hendrickson, D. N.; Christou, G. Chem. Commun. 2003, 2672. (67) Soler, M.; Wernsdorfer, W.; Abboud, K. A.; Huffman, J. C.; Davidson, E. R.; Hendrickson, D. N.; Christou, G. J. Am. Chem. Soc. 2003, 125 3576. (68) King, P.; Wernsdorfer, W. ; Abboud, K. A.; Christou, G. Inorg. Chem. 2004, 43 7315. (69) Saudo, E. C.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. Inorg. Chem. 2004, 43 4137. (70) Murugesu, M.; Raftery, J.; Wernsdor fer, W.; Christou, G.; Brechin, E. K. Inorg. Chem. 2004, 43 4203. (71) Arndt, D. Manganese Compounds as Oxidizing Agents in Organic Chemistry ; Open Court Publishing Co.: La Salle, Illinois, USA, 1981. (72) Zaleski, C. M.; Depperman, E. C.; Kamp f, J. W.; Kirk; M. L.; Pecoraro, V. L. Angew. Chem. Int. Ed. 2004, 43 3912. (73) (a) Sokol, J. J.; Hee, A. G.; Long, J. R. J. Am. Chem. Soc. 2002, 124 7656. (b) Schelter, E. J.; Prosvirin, A. V.; Dunbar, K. R. J.Am. Chem. Soc. 2004, 126 15004. (74) (a) Sternal, R. S.; Marks, T. J. Organometallics 1987, 6 2621. (b) Borgne, T. L.; Rivire, E.; Marrot, J. ; Thury, P. ; Girerd, J. J.; Ephritikhine, M. Chem. Eur. J. 2002, 8 74. (75) Lintvedt, R. L.; Schoenfelner, B. A.; Ceccarelli, C.; Glick, M. D. Inorg. Chem. 1984, 23 2867. (76) (a) Kim, J.; Norquist, A. J.; OHare, D. J. Am. Chem. Soc. 2003, 125 12688. (b) Kim, J.; Norquist, A. J.; OHare, D. Chem. Commun. 2002, 2198.

PAGE 216

200 (77) Barnhart, D. M.; Butcher, R. J.; Clark, D. L.; Gordon, J. C.; Watkin J. G; Zwick, B. D. New J. Chem. 1995, 19 503. (78) (a) Miyasaka, H.; Clrac, R.; Wern sdorfer, W.; Lecren, L.; Bonhomme, C.; Sugiura, K.; Yamashita, M. A. Angew. Chem. Int. Ed 2004, 43 2801. (b) Milios, C. J.; Raptopoulou, C. P.; Terzis, A.; Lloret F.; Vicente, R.; Perlepes, S. P.; Escuer, A. Angew. Chem. Int. Ed 2003, 43 210. (c) Brechin, E. K.; Soler, M.; Davidson, J.; Hendrickson, D. N.; Parsons, S.; Christou, G. Chem. Commun, 2002, 2252. (d) Price, D. J.; Batten, S. R.; Moubaraki B.; Murray, K. S. Chem. Commun 2002, 762. (e) Sanudo, E. C.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. Inorg. Chem 2004, 43 4137. (f) Tasiopoulos, A. J.; Vinslava, A.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. Angew. Chem. Int. Ed 2004, 43 2117. (g) Tasiopoulos, A. J.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. Angew. Chem. Int. Ed 2004, 43 6338. (79) (a) Aubin, S.M.J.; Wemple, M.W.; Adam s, D.M.; Tsai, H.-L.; Christou, G. and Hendrickson, D.N. J. Am. Chem. Soc. 1996, 118, 7746. (b) Wang, S.; Wemple, M. S.; Tsai, H.-L.; Folting, K.; Huffman, J. C. ; Hagen, K. S.; Hendrickson, D. N.; Christou, G. Inorg. Chem. 2000, 39 1501. (c) Aubin, S. M.; Gilley, N. R.; Pardi, L.; Krzystek, J.; Wemple, M. W.; Brunel, L. C.; Marple, M. B.; Christou, G.; Hendrickson, D. N. J. Am. Chem. Soc. 1998, 120 4991. (d) Aliaga-Alcalde, N.; Edwards, R. S.; Hill, S. O.; Wernsdorfer, W.; Folting, K.; Christou, G. J. Am. Chem. Soc 2004, 126 12503. (e) Wu,J.; De Angelis, F.; Carrell, T. G.; Yap, G. P. A.; Sheats, J.; Car, R.; Dismukes, G. C. Inorg. Chem. 2000, 39 1501. (80) (a) Wemple, M. W.; Tsai, H.-L.; Wang, S.; Claude, J.-P.; Streib, W. E.; Huffman, J. C.; Hendrickson, D. N.; Christou, G. Inorg. Chem. 1996, 35 6450. (b) Wang, S. Y.; Huffman, J. C.; Folting, K.; Streib, W. E.; Lobkovsky, E. B.; Christou, G. Angew.Chem. Int.Ed. 1991, 30 1672. (c) Kulawiec, R. J.; Crabtree, R. H.; Brudvig, G. W.; Gayle, K.; Schulte, G. K. Inorg.Chem. 1988, 27 1309. (d) Wang, S.; Tsai, H.-L.; Folting, K.; Martin, J. D.; Hendrickson, D. N.; Christou, G. J. Chem. Soc., Chem. Commun. 1994, 671. (e) Libby, E.; McCusker, J. K.; Schmitt, E. A.; Folting, K.; Hendrickson, D. N.; Christou, G. Inorg.Chem. 1991, 30 3486. (f) Albela, B.; El Fallah, M. S.; Ribas, J.; Folting, K.; Christou, G.; Hendrickson, D.N. Inorg. Chem. 2001, 40 1037. (g) Wang, S.; Tsai, H.-L.; Folting, K.; Streib, W. E.; Hendrickson, D. N.; Christou, G. Inorg. Chem. 1996, 35 7578. (h) Arom, G.; Bhaduri, S.; Arts, P.; Folting, K.; Christou, G., Inorg. Chem. 2002, 41 805. (81) (a) Wieghardt, K.; Bossek, U.; Nuber, B.; Weiss, J.; Bonvoisin, J.; Corbella, M.; Vitols, S. E.; Girerd, J. J. J. Am. Chem. Soc. 1988, 110 7398. (b) Hagen, K. S.; Westmoreland, T. D.; Scott, M. J.; Armstrong, W. H. J. Am. Chem. Soc. 1989, 111 1907. (c) Dube, C. E.; Wright, D. W.; Pal, S.; Bonitatebus, P. J.; Armstrong, W. H. J. Am. Chem. Soc. 1998, 120 3704.

PAGE 217

201 (82) (a) Brechin, E. K. Chem Commun. 2005, 5141. (b) Piligkos, S.; Rajaraman, G.; Soler, M.; Kirchner, N.; Slageren, J.; Bircher, R.; Parsons, S.; Gudel, H.-U.; Kortus, J.; Wernsdorfer, W.; Christou, G.; Brechin, E. K. J. Am. Chem. Soc. 2005, 127 5572.(c) Boskovic, C.; Wernsdorfer, W.; Folting, K.; Huffman, J. C.; Hendrickson, D. N.; Christou, G. Inorg. Chem. 2002, 41 5107. (83) Vincent, J. B.; Christmas, C.; Chang, H. R.; Li, Q.; Boyd, P. D. W.; Huffman, J. C.; Hendrickson, D. N.; Christou, G. J. Am.Chem. Soc. 1989, 111 2086. (84) (a) Chen, H.; Faller, J. W.; Crabtree, R. H.; Brudvig, G. W. J. Am.Chem. Soc. 2004, 126 7345. (b) Limburg, J.; Vrettos, J. S.; Chen, H.; de Paula, J. C.; Crabtree, R. H.; Brudvig, G. W. J. Am.Chem. Soc. 2001, 123 423. (c) Chen, H. Y.; Collomb, M.N.; Duboc, C.; Blondin, G.; Riviere, E.; Falle r, J. W.; Crabtree, R. H.; Brudvig, G. W. Inorg. Chem. 2005, 44 9567. (85) BVS value44b for the hydroxide O8 gave a value of 1.05. The r0 = 2.167 and B0 = 0.37 for a ThIV-O bond. (b) Brese, N. E.; OKeeffe, M. Acta. Crystallogr. Sect. B 1991, 47 192. BVS values gave the range 4.23 4.31 for the Th ions of 16. (86) (a) Thorp, H. H. Inorg. Chem. 1998, 37 5690. (b) Thorp, H. H. Inorg. Chem. 1992, 31 1585. (87) Ramesh, S.; Wedge, M. S. Physica C: Superconductivity 1994, 230 135. (88) (a) Kim, J.; Norquist, A. J.; OHare, D. J. Am. Chem. Soc. 2003, 125, 12688. (b) Sullens, T. A.; Albrecht-Schmitt, T. E. Inorg. Chem. 2005, 44 2282. (89) (a) Brown, I. D.; Altermatt, D. Acta. Crystallogr. Sect. B 1985, 41 244. (b) Palenik, G. J. Inorg. Chem. 1997, 36 4888. (90) (a) Mukherjee, C.; Weyhermller, T.; Wieghardt, K.; Chaudhuri, P. Dalton Trans. 2006, 2169. (b) Chan, M. K.; Armstrong, W. H. J. Am. Chem. Soc. 1990, 112 4985. (c) Suzuki, M.; Senda, H.; Suenag a, M.; Sugisawa, T.; Uehara, A. Chem. Lett. 1990, 923. (d) Sun, Z.; Gantzel, P. K.; Hendrickson, D. N. Inorg.Chem. 1996, 35 6640. (91) Soler, M.; Wernsdorfer, W.; Folting, K; Pink, M.; Christou, G. J. Am. Chem. Soc. 2004, 126 2156. (92) (a) Eppley, H. J.; Tsai, H.-L.; Vries, N.; Folting, K.; Christou, G.; Hendrickson, D. N. J. Am. Chem. Soc 1995, 117 301. (b) Soler, M.; Wernsdorfer, W.; Abboud, K.A.; Huffman, J.C.; Davidson, E.R.; Hendrickson, D.N.; Christou, G. J. Am. Chem. Soc 2003, 135 3576. (93) (a) Bolcar, M.A.; Aubin, S.M.J.; Fo lting, K.; Hendrickson, D.N.; Christou, G. Chem. Commun. 1997, 1485. (b) Harden, N.C.; Bolcar, M.A.; Wernsdorfer, W.; Abboud, K.A.; Streib, W.E.; Christou, G. Inorg. Chem. 2003, 42 7067.

PAGE 218

202 (94) Abbati, G.L.; Cornia, A.; Fabretti, A.C.; Caneschi, A.; Gatteschi, D. Inorg. Chem. 1998, 37 3759. (95) Bhula, R.; Weatherburn; D.C. Angew. Chem. Int. Ed. 1991, 30 688. (96) Clerk, M.D.; Zaworotko, M.J. Chem. Commun 1991, 1607. (97) (a) Aromi, G.; Wemple, M. W.; Aubin, S. M. J.; Folting, K.; Hendrickson, D. N.; Christou, G. J. Am. Chem. Soc. 1998, 120 2977. (b) Aromi, G.; Knapp, M. J.; Claude, J. -P.; Huffman, J. C.; Hendrickson, D. N.; Christou, G. J. Am. Chem. Soc. 1999, 121 5489. (c) Stamatatos, T. C.; Abboud, K. A.; Wernsdorfer, W.; Christou, G. Angew.Chem. Int. Ed. 2006, 45 4134. (98) Davidson, E. R. GRID Indiana University. (99) (a) Cano, J.; Alemany, P.; Alvarez, S.; Verdaguer, M.; Ruiz, E. Chem. Eur. J. 1998, 4 476. (b) Ruiz, E; Cano, J.; Alvarez, S.; Alemany, P. J. Am. Chem. Soc. 1998, 120 11122. (c) Ruiz, E; Cano, J.; Alvarez, S.; Alemany, P.; J.Comput.Chem. 1999, 20 1391. (d) Carrasco, R.; Cano, J.; Mallah, T.; Jones, L. F.; Collison, D.; Brechin, E. K. Inorg. Chem. 2004, 43 5410. (100) Tasiopoulos, A. J.; O'Brien, T. A.; Abboud, K. A.; Christou, G. Angew. Chem. Int. Ed. 2004, 43 345. (101) Petrukhina, M. A.; Sergienko, V. S. ; Sadikov, G. G.; Tat'yanina, I. V.; Torchenkova, E. A. Koord. Khim. (Russ.)(Coord.Chem.) 1990, 16 354. (102) (a) Selected examples of SMMs: Mn,b Fe,c Ni,d Co,e mixed-metal,f-k and lanthanidel. (b) Tasiopoulos, A. J.; Vinslava, A.; Wern sdorfer, W.; Abboud, K. A.; Christou, G. Angew. Chem. Int. Ed. 2004, 43 2117. (c) Jones, L. F.; Low, D. M.; Helliwell, M.; Raftery, J.; Collison, D.; Aromi, G. ; Cano, J.; Mallah, T.; Wernsdorfer, W.; Brechin, E. K.; McInnes, E. J. L. Polyhedron 2006, 25 325. (d) Bell, A.; Aromi, G.; Teat, S. J.; Wernsdorfer, W.; Winpenny, R. E. P. Chem. Commun. 2005, 2808. (e) Yang, E. C.; Hendrickson, D. N.; Wern sdorfer, W.; Nakano, M.; Zakharov, L. N.; Sommer, R. D.; Rheingold, A. L. ; Ledezma-Gairaud, M.; Christou, G. J. Appl. Phys. 1991, 91 7382. (f) Mishra, A.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. J. Am. Chem. Soc. 2004, 126 15648. (g) Zaleski, C. M.; Depperman, E. C.; Kampf, J. W.; Kirk, M. L.; Pecoraro, V. L. Angew. Chem. Int. Ed. 2004, 43 3912. (h) Schelter, E. J.; Prosvirin, A. V.; Dunbar, K. R. J. Am. Chem. Soc. 2004, 126 15004. (i) Oshio, H.; Nihei, M.; Koizumi, S.; Shiga, T.; Nojiri, H.; Nakano, M.; Shirakawa, N.; Akatsu, M.; J. Am. Chem. Soc. 2005, 127 4568. (j) Sokol, J. J.; Hee, A. G.; Long, J. R. J. Am. Chem. Soc. 2002, 124 7656. (k) Murugesu, M.; Mishra, A.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. Polyhedron 2006, 25 613. (l) Ishikawa, N.; Sugita, M.; Wernsdorfer, W. Angew. Chem. Int. Ed. 2005, 44 2931. (103) (a) Debus, R. J. Biochim. Biophys. Acta. 1992, 1102 269. (b) Ruttinger, W.; Dismukes, G. C. Chem. Rev. 1997, 97 1.

PAGE 219

203 (104) Cinco, R. M.; Holman, K. L. M.; Robblee, J. H.; Yano, J.; Pizarro, S. A.; Bellacchio, E.; Sauer, K.; Yachandra, V. K. Biochemistry 2002, 41 12928. (105) (a) Ghanotakis, D. F.; Babcock, G. T.; Yocum, C. F. FEBS Lett. 1984, 167 127. (b) Boussac, A.; Rutherford, A. W. Biochemistry 1988, 27 3476. (106) (a) Cinco, R. M.; Robblee, J. H.; Rompel A.; Fernandez, C.; Yachandra, V. K.; Sauer, K.; Klein, M. P. J. Phys. Chem. B 1998, 102, 8248. (b) Cinco, R. M.; Robblee, J. H.; Messinger, J.; Fernandez, C.; Holman, K. L. M.; Sauer, K.; Yachandra, V. K. Biochemistry 2004, 43, 13271. (c) Kim, S. H.; Gregor, W.; Peloquin, J. M.; Brynda, M.; Britt, R. D. J. Am. Chem. Soc 2004, 126 7228. (107) (a) Gil de Muro, I.; Insaus ti, M.; Lezama, L.; Pizarro, J. L.; Arriortua, M. I.; Rojo, T. Eur. J. Inorg. Chem. 1999, 935. (b) Schauer, C. K.; Anderson, O. P. Acta Crystallogr., Sect C: Cryst. Struct. Commun. 1988, 44 981. (108) (a) Stamatatos, T. C.; Foguet-Albiol D.; Stoumpos, C. C.; Raptopoulou, C. P.; Terzis, A.; Wernsdorfer, W.; Perlepes, S. P.; Christou, G. J. Am. Chem. Soc. 2005, 127 15380. (b) Tasiopoulos, A. J.; Wernsdorfe r, W.; Abboud, K. A.; Christou, G. Inorg. Chem. 2005, 44 6324. (109) Bond valence sum (BVS) calcu lations for the Mn ions of 19 gave values in the range 2.83 3.06 for the 13 MnIII ions. For the MnII ion (Mn8) the value was 1.78. Also, for the three methoxides O41, O43 and O46 the values were 2.18, 1.88 and 1.93, respectively. (110) Abragam, A.; Bleaney, B. Electron Paramagnetic Res onance of Transition Ions ; Dover Press: Minneaola, NY, 1986. (111) Cinco, R. M.; Robblee, J. H.; Rompel, A. ; Fernandez, C.; Sauer, K.; Yachandra, V. K.; Mlein, M. P. J. Synchrotron Rad. 1999, 6 419. (112) Robblee, J.H.; Messinger, J.; Cinco, R. M.; McFarlane, K. L.; Fernandez, C.; Pizarro, S. A.; Sauer, K.; Yachandra, V. K. J. Am. Chem. Soc. 2002, 124 7459. (113) (a) Strickler, M. A.; Walker L. M.; Hillier, W.; Debus, R. J. Biochemistry 2005, 44 8571. (b) Yano, J.; Pushkar, Y.; Glatzel, P.; Lewis, A.; Sauer, K.; Messinger, J.; Bergmann, U.; Yachandra, V. K. J. Am. Chem. Soc. 2005, 127 14974. (c) Yano, J.; Sauer, K.; Girerd, J.-J.; Yachandra, V. K. J. Am. Chem. Soc. 2004, 126 7486. (d) Visser, H.; Dube, C. E.; Armstrong, W. H.; Sauer, K.; Yachandra, V. K. J. Am. Chem. Soc. 2002, 124 11008. (e) Vrettos, J. S.; Stone, D. A.; Brudvig, G. W. Biochemistry 2001, 40 7937. (f ) Ruettinger, W.; Yagi, M.; Wolf, K.; Bernasek, S.; Dismukes, G. C. J. Am. Chem. Soc. 2000, 122 10353. (114) (a) Peloquin, J. M.; Campbell, K. A.; Randall, D. W.; Evanchik, M. A.; Pecoraro, V. L.; Armstrong, W. H.; Britt, R. D. J. Am. Chem. Soc. 2000, 122 10926. (b) Kulik, L. V.; Epel, B.; Lubitz, W.; Messinger, J. J. Am. Chem. Soc. 2005, 127 2392.

PAGE 220

204 (115) Ibers, J. A.; Holm, R. H. Science 1980, 209 223. (116) (a) Philouze, C.; Blondin, G.; Girerd, J. J.; Guilhem, J.; Pascard, C.; Lexa, D. J. Am. Chem. Soc. 1994, 116 8557. (b) Philouze, C.; Blondi n, G.; Menage, S.; Auger, N.; Girerd, J. J.; Vigner, D.; Lance, M.; Nierlich, M. Angew. Chem., Int. Ed. 1992, 31 1629. (c) Chen, H. Y.; Faller, J. W. ; Crabtree, R. H.; Brudvig, G. W. J. Am.Chem. Soc. 2004, 126 7345. (d) Chen, H. Y.; Collomb, M.-N.; Duboc, C.; Blondin, G.; Riviere, E.; Faller, J. W.; Crabtree, R. H.; Brudvi g, G. W. Inorg. Chem. 2005, 44, 9567. (117) Mishra, A.; Pushkar, Y.; Yano, J.; Yach andra, V. K.; Wernsdorfer, W.; Abboud, K. A.; Christou, G. submitted for publication. (118) Vincent, J. B.; Chang, H. R.; Folting, K.; Huffman, J. C.; Christou, G.; Hendrickson, D. N. J. Am. Chem. Soc. 1987, 109 5703. (119) (a) Ruettinger, W. F.; Campana, C.; Di smukes, G. C. J. Am. Chem. Soc. 1997, 119, 6670. (b) Ruettinger, W. F.; Ho D. M.; Dismukes, G. C. Inorg. Chem. 1999, 38 1036. (c) Ruettinger, W.; Yagi, M.; Wolf K.; Bernasek, S.; Dismukes, G. C. J. Am. Chem. Soc. 2000, 122 10353. (d) Ruettinger, W. F.; Dismukes, G. C. Inorg. Chem. 2000, 39 1021. (e) Carrell, T. G.; Bourles, E.; Lin, M.; Dismukes, G. C. Inorg. Chem. 2003, 42 2849. (f) Chakov, N. E.; Abboud, K. A.; Zakharov, L. N.; Reingold, A. L.; Hendrickson, D. N.; Christou, G. Polyhedron 2003, 22 1759. (g) Wu, J.-Z.; Sellitto, E.; Yap, G. P. A.; Sheats, J.; Dismukes, G. C. Inorg. Chem. 2004, 43 5795. (h) Jian-Zhong Wu, Eric Sellitto Glenn P. A. Yap, John Sheats, and G. Charles Dismukes Inorg. Chem. 2005, 44 8659. (i) Wu, J.-Z.; De Angelis, F.; Carrell, T. G.; Yap, G. P. A.; Sheats, J.; Car, R.; Dismukes, G. C. Inorg. Chem. 2006, 45 189. (120) Bond valence sum (BVS) calcu lations for the Mn ions of 20 gave values in the range 2.85 3.00 for the 10 MnIII ions. For the MnII ion (Mn3) the value was 1.59, and for the MnIV ion (Mn2) the value was 4.19. Also, for the hydroxide O3, the value was 1.15. (121) Wang, S.; Tsai, H.-L.; Libby, E.; Folti ng, K.; Streib, W. E.; Hendrickson, D. N.; Christou, G. Inorg. Chem. 1996, 35 7578 (122) Hendrickson, D. N.; Christou, G.; Schmitt, E. A.; Libby, E.; Bashkin, J. S.; Wang, S.; Tsai, H.-L.; Vincent, J. B.; Boyd, P. D. W.; Huffman, J. C.; Folting, K.; Li, Q.; Streib, W. E. J. Am. Chem. Soc. 1992, 114 2455. (123) Van Vleck, J. H. The Theory of Electric and Magnetic Susceptibilities ; Ed.; Oxford Univesity Press: London, 1932. (124) Cinco, M.R.; Robblee, J.H.; Rompel A.; Fernandez C.; Sauer, K.; Yachandra, V.K.; Klein M.P. J. Synchrotron Rad. 1999, 6 419. (125) Ono, T.-A.; Inoue, Y. FEBS Lett. 1988, 227 147.

PAGE 221

205 (126) Han, K.-C.; Katoh, S. Plant Cell Physiol. 1993, 34 585. (127) Asokan, K; Jan, J. C.; Chiou, J. W.; P ong, W. F.; Tsai, M-H.; Shih, H. L.; Chen, H. Y.; Hsueh, H. C.; Chuang, C. C.; Chang, Y. K.; Chen, Y. Y.; Lin, I. N. J. Phys. Condens. Matter 2001, 13 11087. (128) Barkyoumb, J. H.; Mansour, A. N. Phys. Rev. B 1992, 46 8768.. (129) Yano, J.; Kern, J.; Sauer, K.; Latimer, M. J.; Pushkar, Y.; Biesiadka, J.; Loll, B.; Saenger, W.; Messinger, J.; Zouni, A.; Yachandra, V. K. Science 2006, In Press. (130) Sauer, K.; Yachandra, V. K. Proc. Natl. Acad. Sci. USA 2002, 99 8631. (131) (a) Yano, J.; Pushkar, Y.; Glatzel, P.; Lewis, A.; Sauer, K.; Messinger, J.; Bergmann, U.; Yachandra, V. K. J. Amer. Chem. Soc. 2005, 127 14974. (b) Yano, J.; Sauer, K.; Yachandra, V. K. Photosynthesis Research 2005, 85 73. (132) Stern, E. A.; Heald, S. M. Rev. Sci. Instrum. 1979, 50 1579. (133) Rehr, J. J.; Albers, R. C. Rev. Mod. Phys. 2000, 72 621.

PAGE 222

206 BIOGRAPHICAL SKETCH Abhudaya Mishra was born in Jamshedpur India, on November 19, 1978. He performed his high school studi es at Little Flower School in Jamshedpur, after which he successfully qualified in the joint entrance exam (JEE) of the prestigi ous Indian Institute of Technology (IIT), India, and took admi ssion at IIT, Kharagpur. He received a Bachelor of Science degree w ith honors in industrial chem istry from IIT in July, 2001. During his undergraduate studies he performed research in the group of Professor P. K. Chattaraj, primarily on quantum and theoretical chemistry. After the completion of his undergraduate degree, Abhudaya joined the re search group of Professor George Christou at the University of Florida in August, 2001. He decided to pursue his doctoral studies under the guidance of Dr. Christou. Abhudaya's doctoral research prim arily involves the preparation and the physical and magnetic ch aracterization of polynuc lear, oxide bridged, Mn-containing, mixed-metal complexes that f unction as single-molecule nanomagnets. In addition, these heterometallic complexes are also relevant to the bioinorganic modeling of the water oxidizing comp lex of Photosystem II.