Citation
The first synthesis of a transition metal complex of cycloheptatetraene

Material Information

Title:
The first synthesis of a transition metal complex of cycloheptatetraene
Creator:
Winchester, William Randolph, 1957-
Publication Date:
Language:
English
Physical Description:
viii, 185 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Atoms ( jstor )
Carbenes ( jstor )
Carbon ( jstor )
Chemical equilibrium ( jstor )
Geometry ( jstor )
Ligands ( jstor )
Orbitals ( jstor )
Platinum ( jstor )
Potassium ( jstor )
Transition metals ( jstor )
Allene crystals ( lcsh )
Chemistry thesis Ph. D
Cycloheptatetraene ( lcsh )
Dissertations, Academic -- Chemistry -- UF
Transition metal compounds ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1985.
Bibliography:
Includes bibliographical references (leaves 180-184).
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by William Randolph Winchester.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright William Randolph Winchester. Permission granted to the University of Florida to digitize, archive and distribute 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.
Resource Identifier:
14558053 ( OCLC )
029494824 ( ALEPH )

Downloads

This item has the following downloads:


Full Text













THE FIRST SYNTHESIS OF A TRANSITION METAL COMPLEX
OF CYCLOHEPTATETRAENE





















By

WILLIAM RANDOLPH WINCHESTER














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 1985


































TO MY PARENTS.














ACKNOWLEDGEMENTS

By far the most enjoyable aspect of writing a

dissertation is the opportunity to acknowledge assistance received. I would like to thank Professor William M. Jones for his constant encouragement and patience. Professor J. R. Sabin was essential in helping the strange bedfellows of theory and experiment develop together. It is pleasant to acknowledge the assistance of the Quantum Theory Project in all the facets of my education, but especially for the opportunity to attend the Summer Institute in Quantum Chemistry and to attend the annual Sanibel Symposia.

It is my pleasure to acknowledge the capable assistance of and discussions with George Purvis, Marc Radcliffe and Seth Elsheimer.

Finally, I want to thank Joan Raudenbush for her constant assurances I would finish and her capable assistance in making that come true.













iii

























"One takes up fundamental science out of a sense of pure excitement, out of joy at enhancing human culture, out of awe at the heritage handed down by generations of masters and out of a need to publish first and become famous."



Leon M. Lederman
The Value of Fundamental Science Scientific American 251 #5, November, 1984.
























iv












TABLE OF CONTENTS

Pane

ACKNOWLEDGEMENTS ---------------------------------------- iii

ABSTRACT ------------------------------------------------- ii

CHAPTER

I INTRODUCTION ----------------------------------- 1

II RESULTS OF EXTENDED HUCKEL CALCULATIONS -------- 20 III EXPERIMENTAL RESULTS -------------------------- 63

IV SUMMARY ------------------------------------------ 116

V -EXPERIMENTAL ------------------------------------ 118

General ---------------------------------------- 118

7,7-dibromo-2,3-4,5-dibenzo bicyclo[4.1.0]heptane
(122) ----------------------------------- ---- 119

4,5-6,7-10,11-12,13-tetrabenzotricyclo[7.5.0.0 2,8]tetradeca-2,3-4,5-6,7-10,11-12,13-14, 1-hexane
(125) ------------------------------------------ 119

(1,2-i 2-4,5-6,7-dibenzocycloheptatetraene)-bis
(triphenylphosphine)platinum (127) ------------ 120

(1,2-n 2-1,2-cyclononadiene) bis (triphenylphosphine)
platinum (57) ------------------------------- 121

trans-(bromo)(n -cyclohepten-1-yl)bis(triphenylphosphine)platinum (133) ---------------------- 121

trans-(bromo) (l -3,4-5,6-dibenzocycloheptatrien-lyl)bis(triphenylphosphine)platinum (135) 122

(1,2-n 2_4,5-benzocycloheptatetraene) bis(triphenylphosphine)platinum (143) ---------------------- 123







v









trans-(bromo)(4-n -1,2-benzocycloheptatriene-4yl)bis(triphenylphosphine)platinum (146) -------- 124 Tropone (148) ------------------------------------ 125

1-,2-, and 3-bromocycloheptatrienes (149) -------- 125

(1,2-n -cycloheptatetraene)bis(triphenylphosphine)
platinum (150) ---------------------------------- 125

trans-(bromo) ( l-n l-cycloheptatrienyl)bis (triphenylphosphine)platinum (152) ------------------126

7,7'-cycloheptatrienylcycloheptatriene-, 1'-diyl)
bis(triphenylphosphine)platinum (153) ----------- 127

(7'-cycloheptatrienyl-7-cycloheptatriene-1,1'-diyl)
(1,2-bis-diphenylphosphinoethane)platinum (163)-- 128

(7'-3', 4 '-benzocycloheptatrienyl-7-3,4-benzocycloheptatrien-1, 1 '-diyl)bis(triphenylphosphine)platinum (164) ---------------------------------- 129

Reaction of 135 with potassium tert-butoxide ----- 130

The reaction of heptafulvalene with potassium tertbutoxide and tris(triphenylphosphine)platinum --- 130 APPENDIX ---------------------------------------------- 132

REFERENCES -------------------------------------------- 180

BIOGRAPHICAL SKETCH ----------------------------------- 185



















vi














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



THE FIRST SYNTHESIS OF A TRANSITION METAL COMPLEX OF CYCLOHEPTATETRAENE

By

William Randolph Winchester


Chairman: William M. Jones Cochairman: J. R. Sabin
Major Department: Chemistry

It is well established that cycloheptatetraene is lower in energy than cycloheptatrienylidene. To date all transition metal complexes of this ligand have been aromatic carbene complexes, no transition metal cycloheptatetraene complex has been reported.

Extended Huckel molecular orbital calculated relative energies correctly show cycloheptatetraene to be lower in energy than cycloheptatrienylidene and also show that for the dicarbonylcyclopentadienyliron (Fp) complex the aromatic carbene complex is lower in energy than the cycloheptatetraene complex. This result is interpreted as resulting from a strong bonding interaction between the Fp LUMO and the cycloheptatrienylidene HOMO, which is much vii








less strong for the cycloheptatetraene-Fp complex. Further, this result led to the choice of bis(phosphine)platinum as a metal for which the cycloheptatetraene complex would be lower in energy than the cycloheptatrienylidene complex.

The first transition metal cycloheptatetraene complex was synthesized by the generation of cycloheptatetraene in the presence of tris(triphenylphosphine)platinum and exchange to give (n -cycloheptatetraene)bis(triphenylphosphine)platinum. Similarly, the 1,2-n2-4,5-benzocycloheptatetraene and the 1,2- 2-4,5-6,7-dibenzocycloheptatetraene complexes have been synthesized. Interestingly, none of these compounds is fluxional.

A novel reaction between cycloheptatetraene and the cycloheptatetraene complex to give (7'-cycloheptatrienyl7-cycloheptatrien-l, 1 '-diyl)bis(triphenylphosphine)platinum, which has been characterized by 1H, 13C, 31P NMR spectra and an x-ray structure determination, was observed. This is the first (allene)bis(triphenylphosphine)platinum complex which has been observed to undergo this reaction. A mechanism which is consistent with this result is proposed.








viii














CHAPTER I
INTRODUCTION

Allenes, or 1,2-dienes, are linear compounds, with a central sp-carbon bonded to two sp2-hybridized carbons.





C C == C H H H

1



Transition metal complexes of allenes, in which the metal is bonded to one of the double bonds, form an interesting





HM
n

2

group of molecules and have been the subject of several
1-5
reviews.

Crystal structure determinations have been done for a number of allene complexes. The usually linear allene is bent when complexed to a transition metal. The carbon which is not bonded to the transition metal moves away from the metal. This bending is consistent with descriptions of olefin complexes as metallocyclopropanes.


1




2





H H H
C C
II H
C + M C
1 M BENT C C

HH HH
2
LINEAR

1

H

H , C


C

H

METALLACYCLOPROP ANE The metal carbon distances are not symmetrical, the MC1 bond distance is always longer than the MC2 bond distance. This has been attributed to the different hybridizations of C1 and C2, since increasing s-character
4
in an orbital generally shortens the bond length. It has also been attributed to the overlap of an occupied metal




H

H C\ sp2 IM
sp C

HH




3


d-orbital with the orthogonal w*-orbital of the double bond which is not complexed to the metal.6










HH

4




The bonding in transition metal allene complexes is usually discussed within the context of the DewarChatt-Duncanson model.7-8 There is donation from the olefin n-orbital to an unoccupied metal al-orbital and backbonding from an occupied metal bl-orbital to an olefin 7*-orbital. The bending of allene upon complexing to a metal has been attributed to metal backbonding.




H H H H
C \C/
IJ II

Oc<: (>C


HH H H

5 6




4



Rotation about the axis of the olefin transition metal i-bond is rapid in some allene transition metal complexes and in those cases for which rotation is rapid there is also a rapid exchange of the metal between the orthogonal double bonds. For symmetrical allenes the resulting configurations



HH H H
C C 4-M U C --" C
II 4- M
C C
H H H

7 8



are equivalent and this exchange is called fluxional behavior. 9Fluxional processes have been studied by several authors and as will be discussed below the mechanism appears to depend on the metal and its oxidation state. 10-13

The fluxional exchange has been studied for dicarbonylcyclopentadienyliron (Fp) and has been shown to be intramolecular.10 It was shown there is no loss of optical activity during the fluxional exchange of 1,2-cyclononadiene complexed to Fp. This result excludes allyl complex 10 as a possible intermediate in the Fp exchange reaction.




5



H H

\ 1 /.H 2 C Fp - C + 4 Fp+ H2

HH

9 10

The work of Cope et al.11 with platinum(II) complexes of 1,2-cyclononadiene, 11, is in direct contrast to the work with Fp. When optically pure 1,2-cyclononadiene was used to make the complex with platinum, repeated recrystallization led to a decrease in enantiometric purity. When the 1,2-cyclononadiene was released from the platinum it was found to be of lower optical purity than the starting material. Two structures were proposed for the intermediate



Pt(Am)(C1) 2



11





Pt(Am)(Cl)

Pt(Am) (Cl)2 11 CC


12 13




6



in the mutarotation, but 7-allyl complex 13 is the more likely intermediate since other platinum allene complexes have been shown to rearrange to -allyl complexes, vide infra.

Allene complexes of platinum(0) with phosphine ligands have been studied, and show no fluxional behavior. 12 The H-NMR spectrum of 14 shows two signals for H and Hb at all temperatures, with decomposition occurring before coalscence is observed.


Ph H

C /
S PPh

C \pph3
C PPh

Ha Ph

14






As with structure and bonding, the synthesis of allene transition metal complexes has similarities to olefin transition metal complexes, with some interesting differences which can be attributed to the uniqueness of the allene. The majority of allene complexes has been prepared by exchange reactions; the allene is either added to or generated in-situ with the transition metal and displaces a ligand from the metal. For example see Fiaure 1.











Ref

HH HH C C o - -- Pt(PPh3) 4 Pt(PPh3)4 + C 2 1 3) II 1 / C /C\ H H H H
15 1 16









Fp + Fp 10




17 18 1






Me Me Me Me Fe( -*-- Fe (CO) 4 15 II II /C C
Me Me Me Me

20 21 22


Figure i. Synthesis of transition metal allene complexes.












Figure 1. Continued.


Ref


H H
C
Rh(PPh3) Cl + 1 -j-Rh(PPh3)2C1 14
3 3 C
I
c
H H

23 24



HH HH
C C
II ( -*-Pt(Cl) )16 ((C2H4 PtC122 + C C16
II II
C
Me Me Me Me


25 26 27




9



Several allene complexes have resulted from

intramolecular rearrangements. Treatment of 28 with 17
acid leads to the allene complex 9 17 While it is possible the allyl cation 10 is an intermediate, it is most likely there is neighboring group participation of the iron.



OMe


28






CH2

CH 2 OMe

10 29







Fp+42
C

CH2

9


While studying hydride abstraction from 30, it was found that the unusual reagent, 31, gave the w-complex of allene.18 Alpha hydride abstraction gives 32, a highly strained alkylidene, which rearranges to the 7-complex.




10








+
Fp H Fp Fp+






30 31 32



Fp+--II

9

This reaction has been studied starting with resolved 33.19 Treatment of 33 with trimethylsilyltriflate leads to the formation of optically inactive allene complex 36 The racemization was interpreted as occurring via initial formation of carbene 34, followed by opening to achiral Fp OMe Fp +

+ TMSOTf Fp -- + Me


33 34 35






36




11



allyl cation 35 and finally rearrangement to the allene

complex 36.

In what appears to be the reverse process of 35 to 36, platinum(II) allene complexes have been observed to rearrange to r-allyl complexes.2023 Allene displaces

acetone from 37 to give an allene complex, 38, which can be isolated below 0 C. If this is allowed to warm to room temperature in solution, 7-allyl complex 39 is formed.





+
Me Q 2 oc I CH2 RT
Q-Pt-Q + C Me-Pt
o0 CH2 Q CH2


37 1 38 39'



This discussion of allene complexes has led to the introduction of complexes of the general types 40, 41, and 42. The interconversion of metal complexes of this





M
40


M M=
41 42




12



type is of interest, but has been studied very little. However, the interconversion with no metal has been thoroughly studied. When cyclopropylidene 43 is generated from an optically active precursor, the product, 46, is the optically active allene indicating that the achiral, planar
24-27
allene, 45, is not an intermediate in the reaction.





NO Ph Ph



PhPh
Ph
43 44 45



Ph
C= C =C4.Hh
H
46

optically active



Allenes which have been incorporated into a
28-29
ring have been the subject of more recent investigations.





H . I H




47




13



Incorporating an allene into a ring forces it to bend about the central carbon and twists the 7r-bonds from their usual perpendicular geometry. This raises the energy of the allene closer to the bent, planar geometry 48 which experiences little deformation on incorporation in a ring.


H H





48


Cyclic allenes with fewer than eight carbons cannot be isolated and indirect methods have been developed to probe the structure of these molecules.

A cyclic allene has C2 symmetry and therefore is

chiral. Its planar valence isomer, 48, has Cs symmetry






C2 Cs

47 48



and is therefore achiral. Balci and Jones28 have used this difference to probe the structure of the six and seven membered ring compounds. Their results show that the product




14



of dehydrohalogenation is chiral and therefore for the six membered ring the allene, 47, is still lower in energy than its planar form. This is in accord with recent FORS







b + Menthoxide


49







optically active

50


(Full Optimized Reaction Space) calculations which show the allene to be preferred by 13.1kcal/mol for the six membered ring.29



-13.1 H H 48





--.0 H
kcal/mol

47




15



There are two reports of the synthesis of strained cyclic allenes complexed to transition metals in the literature. Visser and Ramakers30 trapped the seven, eight, and nine membered cyclic allenes with 54. The




H' H Pt(PPh3)2
+ Pt(PPh3)2(C2H4) - H H


n
n= 2 51 54 n n= 2 55
3 52
3 56
4 53
4 57


resulting w-complexes were crystalline, air stable complexes that were not fluxional.

The Fp complex of 1,2-cycloheptadiene, 59, has been 31
synthesized by methoxy abstraction from 58. This compound






OMe + TMSOTf



58 59






is fluxional with a 13.9kcal/mol activation barrier for the 1,2-shift.




16



Cycloheptatetraene, 60, is a strained, cyclic allene which has been the subject of much interest.32-36 The planar form of this allene is aromatic and this decreases the difference in energy between the planar form 61 and

the allene 60.











60 61



Calculations are all in agreement that cycloheptatetraene is lower in energy than its planar form. Waali,34 using the MNDO method, has calculated the vibrational force constant matrix and normal modes for the planar form 61 and his results are consistent with the planar form 61 as a transition state connecting the enantiomeric cycloheptatetraene geometries.

Method AE Ref

INDO 13.8 36 MNDO 23 37
AE
EHT 28 38

STO 4-31G 15.8 39





17



Harris and Jones studied the dehydrohalogenation from optically active bromocycloheptatriene 62. The isobenzofuran adduct of cycloheptatetraene is optically active, consistent with the calculated lower energy of cycloheptatetraene relative to its planar form 61.




+ t-BuOK


62 60











63



No transition metal cycloheptatetraene 7-complex

has been synthesized. But many transition metal complexes of its planar valence isomer have been synthesized. In the original work,37 reaction of sigma compound 64 with the hydride abstractor 65 gave the cycloheptatrienylidene complex 66. The compound was sufficiently novel that an



+ Ph3 C + Fp



64 65 66




18



x-ray crystal structure was obtained for it and its benzannelated derivative.3839 The crystal structure shows unequivocally that the ligand is planar and sigma bonded to the metal. Hydride abstraction to give the carbene complex has proven to be a very general reaction and Table 1 gives some of the compounds made and the 13C-NMR shift of the carbene carbon. The allene complex, 67, was not observed in any of the reactions, even as a side product.





M





67





It was the primary objective of this research to

synthesize the first transition metal complex of cycloheptatetraene. First, molecular orbital calculations were used to study the bonding in 40, 41 and 42. Then using the results of these calculations a transition metal system was chosen and the first transition metal cycloheptatetraene complex synthesized.





19




Table 1. 13C NMR chemical shifts (ppm) of the carbene carbon in some cycloheptatrienylidene transition metal complexes.

Structure M=Fe M=Ru




242.3 223.6 68





Mp+ 265.9 244.7


69




Mp + 201 186.6




70



Mpp
278.8 256.2 71





Mp= ,M\ Mpp= M
CO CO CO PR
3















CHAPTER II
RESULTS OF EXTENDED HUCKEL CALCULATIONS

The bonding in the transition metal complexes 40,

41, and 42 was studied using molecular orbital calculations 40,41
of the extended Hickel type. The transition metal complexes







H2C= CCH2

M M M

40 41 42






with M=Fp were studied first to help understand the experiments discussed in the introduction. Then the transition metal complexes with M=Pt(PH3)2 were studied because it was expected a cycloheptatetraene complex of this system could be synthesized and because it was expected that studying the bonding in a different transition metal complex for comparison with Fp complex would be useful.





20





21



The relative energies calculated by the extended HU'ckel molecular orbital method (EHIMO) for 72, 73 and 74 are shown in Figure 2. Experiments have shown that generation of both the cyclopropylidene complex 74 and the allyl complex 73 leads to the formation of the allene complex 72, indicating that the allene complex 72 is the preferred valence isomer.17,18 In agreement with this experimental result the allene complex 72 is calculated to be lowest in energy. Allyl complex 73 is next highest in energy, 24 kcal/mol above 72. The cyclopropylidene complex 74, is calculated to be highest in energy, 33 kcal/mol above the allene complex 72.

A minimum for the energy difference between the allene complex 72 and the allyl complex can be deduced from an earlier reported experiment.17 There is no loss of optical purity on refluxing the optically active allene complex 19 for two hours at 800C. If one assumes a preexponential









Fp Fp19 75




22

















Kcal






Fp 33 74



- 24 Fp

73










II

72 Figure 2. Relative energies of 72, 73 and 74 in kcal/mol.




23


term in the Arrhenius equation,42 A, of 1013 and a T, of two hours then the calculated barrier is 30 kcal/mol. This also assumes that the transition state for racemization



E = RT [in(T ) + 35.73]
a k (T,sec


is the allyl cation and not some intermediate in the conversion of the allene to the allyl cation.

The relative energies of allene 76, planar allene 77,

and cyclopropylidene 78 were calculated for comparison with the Fp complexes. As Figure 3 shows, allene, 76, is lowest in energy, followed by planar allene 77, and highest is cyclopropylidene 78. The ordering by energy is the same for the organic molecules as for their Fp complexes. But, the differences in energy of the organic molecules are much greater than in the Fp complexes.

In order to understand this decrease in relative energy on complexing to Fp and to understand the bonding in the Fp complexes some interpretation of the EHMO calculated molecular orbitals is necessary. The interpretation of calculated molecular orbitals for organometallic complexes has been done by arbitrarily dividing the molecule into an organic fragment and a metal fragment.43-46 The molecular orbitals from the calculation for the molecule are discussed with an interaction diagram in terms of the orbitals calculated for the fragments. This approach was used in discussing the bonding in 72, 73 and 74.




24




Kcal





78










80 C>

77
















-- 33 Fp+ . 74


- 24 Fp 73








12 - 0II
2

76 72 Figure 3. Relative energy of 76-78 as calculated by EHMO.




25



Before the calculated molecular orbitals and energies are discussed in more detail, the method used to obtain these will be described. For all the organic molecules MINDO/3 geometry .optimizations were performed.47 The resulting geometries were used for the EHMO calculations. A similar optimization could not be done with the organometallic complexes because MINDO/3 is not parameterized for transition metals. Initially the EHMO calculations on the organometallic complexes used the geometries found from optimization of the organic fragments combined with the known geometry of the metal fragment. The EHMO calculated relative energies were not compatible with the experimental results, 72 was found to be higher in energy than 73. A study of the effect of bending in the allene ligand on the EHMO total energy is shown in Figure 4. There is an






H H H
C C

C C / bend C
HH H HH


Linear Bent

e = 1800 e < 1800





26



18 kcal/mol drop in total energy on bending the allene ligand from 1800 to 1500, which is enough to make the allene Fp complex lowest in energy, this bend is also consistent with the known geometry of Fp allene complexes.48

Therefore, it was decided that using the geometries of the uncomplexed ligands would not be acceptable for the complexes. Since geometry optimizations using the EHMO method can lead to spurious results, geometries obtained from the MINDO/3 optimization of the model systems 79, 80 and 81 were used for the organic fragments in the organometallic complexes. Using this procedure assumes that the bonding in the oxygen complexes will be similar to the bonding in the Fp complexes, which is reasonable since the oxygen atom is isolobal with Fp. This procedure led to an angle of 161.30 in the geometry used for the complexed allene.







0 0 0


H2C --CCH2 H2C H 2

79 80 81




27



















I o

k














145 155 165 175
ANGLE OF CCC LINK IN 72.

Figure 4. Affect of bending on the energy of 72.





28


The allene valence orbitals are shown in Figure 5.

Because the allene geometry used was not linear, degeneracy of the i orbitals is lifted. While only the lt and the 2n* orbitals are of w-like symmetry the 7 symbol is used for all four orbitals for ease of communication. The 2w and the ir* orbitals have been raised and lowered in energy respectively. Note the large energy difference between the occupied and unoccupied orbitals. Allene is a stable, closed shell molecule and it will not interact with the metal fragments as strongly as will a fragment with non-bonding electrons.

The valence orbitals of the Fp fragment, 82, as calculated by the EHMO method are shown in Figure 6. Several groups have reported calculations on Fp compounds and all have found it useful to discuss their results using the Fp fragment
S49-50
valence orbitals. The LUMO, 3a', is important for the strong a-bonding with incoming ligands. The HOMO, la", is well suited for backbonding to the ligand, both energetically and geometrically.

The calculated two lowest energy conformations of 72

are upright, 72a, and bisecting, 72b, and the bisecting conformation is preferred by 20 kcal/mol. This conformation minimizes the destabilizing steric interaction between the cyclopentadienyl ring and the allene ligand and maximizes the allene LUMO overlap with the Fp HOMO. (cf. Figure 7). The experimental value for the barrier to rotation is




29







el





-7.9 2.-7 4







-9F.2 n1, _ 76



















-12.6 _


-13.0 IT Figure 5. Allene, 76, valence orbitals.





30







eV










-8.3 - 2a"



















-10.9 - 3a' C>O











-12.5 ~ la"
-12.7 2a -12.9 1la'









Figure 6. Fp, 82, valence orbitals.




31







SFe 5 Fe rH oc 4 ' .%H OC I OC H OC C
i1
C
H H

BISECTING UPRIGHT

72b 72a




8.4 kcal/mol, which shows the calculated energy difference for 72a and 72b is probably too high.10 The rotational barrier for the ethylene complex 83 has been calculated to be 21 kcal/mol, again a very high value. The authors state, ". .. in general extended H'ickel energies are not reliable and must be viewed as indicative of trends."49







H
Fe 8
OC
OC ,. C

H

83




32



The interaction diagram for Fp with allene in the bisecting conformation is given in Figure 7. The molecule has no symmetry and the diagram is a simplified one with only the major interactions shown. The allene HOMO is a bonding orbital and much lower than the Fp 3a' LUMO. The bonding orbital resulting from this interaction is only slightly lower in energy than the allene HOMO. Similarly, a large energy difference between Fp la" and the allene lx* results in a small lowering in the resulting bonding orbital.

The bisecting allyl complex 73b is calculated to be 24 kcal/mol above the allene complex 72 and 3 kcal/mol lower than its upright isomer 73a. From the valence orbitals of. the allyl fragment 77, shown in Figure 8, it is apparent the HOMO, al, is a non-bonding orbital. The LUMO, a2, is of the correct symmetry for w-backbonding









Fe Fe

oc OC

BISECTING UPRIGHT

73b 73a





33
eV









-7








-8 - LT*

2a"














-10







C1 0 3a'









-12







la'
-13 Figure . Interaction diagram for 72.










Figure 7. Interaction diagram for 72.




34











-7.9 - bl































-11.0

















-13.7 bl








Figure 8. Allyl fragment valence orbitals.




35



but a node at the central carbon leads to a small overlap with other fragments and little interaction.

Most striking in the interaction diagram for 73,

shown in Figure 9, is the lack of change or interaction in the a2 orbital of allyl fragment 77. Next, there is a bonding orbital formed by the virtual Fp 3a' orbital and the occupied allyl al orbital. Finally, there is a stabilizing interaction of the allyl b1 orbital with the Fp la' orbital. Notice that the bonding orbital resulting from the allyl al orbital and the Fp 3a' orbital is lowered

by almost 2 eV. The allene 27 orbital interaction with the Fp 3a' orbital results in only a .5 eV lowering (cf. Figure 7). The decrease in relative energy of 76 and 77 on complexing to Fp appears to be due to the different

ligand-HOMO, Fp-LUMO interactions.

The Fp-cyclopropylidene, 74, is the highest in energy

of the Fp complexes, 34 kcal/mol above the Fp allene complex 72. The valence orbitals of the cyclopropylidene fragment, shown in Figure 10, are as one would expect. The HOMO, al, is a non-bonding orbital, associated with the lone pair of the carbene. The LUMO, bl, is of the correct symmetry

for backbonding.





36

eV









-7








-8

2a"





-9








-10







-11 0 <<:) 3a'








-12



s Iba"

2a' la'
-13











Figure 9. Interaction diagram for 73.





37
















eV















-10.8 , bl






-11.7















-13.4 . 2 2







Figure 10. Cyclopropylidene valence orbitals.





38



The upright conformer 74a is calculated by EHMO to be lower in energy than the bisecting conformer, 74b, by

5 kcal/mol. The low difference in energy of the two conformations is the result of two opposing affects. In the upright conformation w-backbonding is strongest and in the bisecting conformation the steric repulsion between the cyclopentadienyl ring and the cyclopropylidene ligand is smallest.












+ +
Fe Fe



oc oc
OC OC


UPRIGHT BISECTING

74a 74b




39




From the interaction diagram for 74a, shown in Figure 11, it is apparent the strongest interaction is between the cyclopropylidene al orbital and the Fp 3a' orbital; the a-bond formation. As in the Fp allyl complex this a-interaction is worth almost 2 eV. There is also a stabilizing interaction resulting from backbonding from the Fp la" HOMO to the b1 LUMO of 78. This backbonding accounts for the greater stabilization of the cyclopropylidene 78 relative to the planar allene 77 on complexing to Fp.

Since the acyclic allenes are known to be bent when complexed to Fp, an allene which has been incorporated in a ring should be less strained when complexed to a metal than when free. At the same time EHMO calculations (vide supra) predict the allyl geometry to be closer in energy to the allene when complexed to Fp. These are opposing effects; the first will increase the difference in energy between the complexed allene and the second will decrease the difference. The relative magnitude of the two effects will determine the size of the ring necessary to make the Fp allene complex equal in energy to the Fp allyl complex. The six memebered ring allene 84, its allyl isomer 85 and 86 and their Fp complexes were therefore studied.





40

eV









-71








-8

2a"





-9








-10

















-12





2a' la'
-13











Figure 11. Interaction diagram for 74.




41







e.






84 85 86







The geometries of the uncomplexed valence isomers 84, 85, and 86 were optimized using MINDO/3 and the EHMO calculations were done on the optimized geometries, the resulting calculated relative energies are shown in Figure 12. The cyclic allene 84 is calculated by EHMO to be lower in energy than its allyl isomer by 40 kcal/mol and is 70 kcal/mol lower than bicyclocarbene 86. These relative energies are smaller than those for 76, 77 and 78 following the expected trend. Experiment has shown the allene to be the preferred geometry, but the next accessible geometry and electronic state have not been unambiguously assigned.28,29

The geometries for the EHMO calculations on the Fp complexes were obtained by optimizing the corresponding oxides 87, 88 and 89 with MINDO/3. The EHMO calculated relative energies, shown in Figure 18, place the Fp allyl




42



Kcal


114 78













--80o : 77



-- 70 86











-40 85 4 o 85
















--- O 76 o 84 CH2


Figure 12. Relative energies of 84-86 as calculated by EHMO.




43

Kcal



















-----70 86











-40 "85



Fp+







4 91 84 0 90 Figure 13. Relative energies of 90-92 as calculated by EHMO.




44







0 0- 0






87 88 89











complex 90 lowest in energy, 14 kcal/mol below the Fp allene complex 91. The bicyclo(3.1.0)hexane carbene complex 92 is highest in energy, 23 kcal/mol above the allyl complex 90. While these results corroborate intuitive expectations, they must await experimental verification.

Finally, the cyclic-conjugated seven membered ring

analogues 93, 94, and 95 were studied. Recall that 93 has been shown to be the preferred valence isomer by both experiments and calculations using the MNDO,STO-3G and

STO 4-31G methods.

The geometry of 93, 94, and 95 was optimized

using MINDO/3 and the EHMO calculation was then performed using the MINDO/3 geometry. Consistent with the earlier




45



results, the cycloheptatetraene, 93, was lowest in energy 29 kcal/mol lower than 94 .36,37,39 Highest in energy, 94 was found to be 51 kcal/mol above cycloheptatetraene.

The Fp complexes were studied next. First the geometry of 96, 97, and 98 was optimized using the MINDO/3





0
0 0





96 97 98









program. Then EHMO calculations were done replacing the oxygen with an Fp fragment. The EHMO calculated total energies, shown in Figure.15, place the cycloheptatrienylidene complex 100 lowest in energy with the Fp cycloheptatetraene complex 99 next, 37 kcal/mol higher in energy and 101 highest, 39 kcal/mol higher in energy. NMR experiments and a crystal structure determination show 100 to be the preferred valence isomer, in agreement with the EHMO results.3739 There is no experimental evidence for the relative energies of 99 or 101 to 100.




46


Kcal


-114 ~78














80 . : 77













-51 95








-29 94










CH2
c 93 CH2


76

Figure 14. Relative energies of 93-95 as calculated by EHMO.




47








Kcal










95
--33

-27
94






79= 101
-2
-0
F- o 100


93









--37 FP 99




Figure 15. Relative energies of Fp complexes 99-101 as calculated by EHMO.




48


The explanation for the preference for 99 relies on our earlier results. Recall that allyl fragment 77, was found to be 80 kcal/mol above allene, 76, in energy. But, for the Fp complexes cycloheptatrienylidene, 94, was found to be only 24 kcal/mol higher in energy than cycloheptatetraene 93. The non-bonding allyl ~OMO has a much stronger interaction with the Fp 3a' LUMO than does the bonding allene HOMO. For the cyclic-conjugated systems the non-bonding HOMO of 94 has a stronger interaction with the Fp 3a' LUMO than does the HOMO of 93 and the uncomplexed molecules are much closer in total energy, only 29 kcal/mol different. Therefore, one could expect that complexing to Fp would lower the energy of 94 sufficiently to make it lower than 93.

The energy difference for the upright and bisecting

conformations of 100 was calculated by EHMO as 9.3 kcal/mol, with the bisecting conformation, 100b,preferred. This is








Fe+ Fe+

O C OC


UPRIGHT BISECTING

100a 100b




49



in agreement with the crystal structure determination which found the cycloheptatriene ring parallel to the cyclopentadienyl ring. The rotational barrier for 102 has been determined by D-NMR studies to be 9.7 kcal/mol, fortuitously close to the calculated barrier for 100.51





+
Fe

bu3P


102



With the completion of the study of the Fp complexes, a new metal fragment was chosen for study, the primary criterion for the choice was experimental evidence which might indicate the metal complex of 93 would be lower in energy than the complex with 94. Bis(triphenylphosphine)platinum has been used to form complexes of several reactive
30,52
olefins, so it was decided to study this metal system using the same organic fragments as in the Fp study. Trends in the relative energy as well as qualitative changes in the bonding were studied.

The relative energies of 103, 104, and 105 were calculated using the EHMO method. For the calculations phosphine was substituted for triphenylphosphine and the





50


organic fragment geometries were the same as in the Fp calculations. The substitution of phosphine for triphenylphosphine has been used by several authors49,50 and other than the steric differences it is assumed that the relative energies are not affected by the substitution.

The EHMO calculated total energies show 103 to be

lowest in energy. Complexes 104 and 105 are of similar energy and are higher in energy than 103 by almost 55 kcal/mol. Comparison of these calculated relative energies with those for the uncomplexed molecules 76, 77 and 78 show the relative stabilization to be 25 kcal/mol for the allyl complex 104 and 60 kcal/mol for the cyclopropylidene complex 105. It is also apparent the bis(phosphine)platinum complexes are significantly different from the Fp analogues and an explanation for the difference would be useful. Interaction diagrams for 103, 104 and 105 will be used to study the bonding in these molecules.

The valence orbitals of bis(phosphine)platinum, shown in Figure 17, are very different from those of Fp even though the HOMO and LUMO are of the same symmetry. The HOMO of 106, b2, is high in energy, 0.7 eV higher than the Fp HOMO. The LUMO of 106, 3a', at -6.5 eV, is too high for a strong interaction; it is 4.4 eV higher than the Fp LUMO. There would be no a-interaction with 106, but a combination of the lal and 3al orbitals lead to a a-interaction when complexes are formed.




51





cal

-114 c 78












-80 a- 77











L2Pt = 105
- 55
54

L Pt--4 104











CH2


-o 2 -o L2PS 103 76

Figure 16. Relative energies of platinum complexes 103-105 as calculated by EHMO.





52











eV





-6.5 3a1 C -<



















-11.8 b2 _..-12.3 2al
1 O



laI
2
a2 b Ial




H3P

H pPt 106

Figure 17. Valence orbitals for bis(phosphine)platinum fragment 106.





53


Lowest in energy of the three valence isomers, the allene complex with bis(triphenylphosphine)platinum, is a known compound. NMR experiments are in agreement with a crystal structure determination showing the in-plane geometry, 103i, to be preferred over the perpindicular
14
geometry, 103p. No rotational barrier has been reported; presumably decomposition occurs before rotation. The calculations are consistent with this, the square planar geometry, 103i, is lowest in energy and the calculated difference in energy is 41 kcal/mol.





HP ,PH HP PH Pt Pt cH C= C C H H C C C N-H
HlC H C "t H

IN-PLANE PERPINDICULAR

103i 103p





An interaction diagram for the in-plane geometry, 103i, is shown in Figure 18. The bis(phosphine)platinum b2 orbital interacts with the allene 17 orbital to form a bonding orbital in the complex. It is the lack of a comparable interaction in 103p which leads to a large





54 eV








-7








-8. - " ' '








-9
-9 ,,%.








-10o








-11






b

-12

2a1

a21



-13 . Ir











Figure 18. Interaction diagram for 103i.




55



rotational barrier. There is also a stabilizing .interaction between the la1 and 3al orbitals of bis(phosphine)platinum and the allene 2w orbital.

Next highest in energy is 104, calculated to be 54 kcal/mol above 103. Recall that the LUMO, the a2 orbital, of the allyl fragment has a node at the central carbon, making this a poor ligand for metal backbonding. The calculated lower energy conformation is the upright 104a, with the plane of the allyl group perpindicular to the plane of the bis(phosphine)platinum. The energy difference between 104a and 104i is small, 4 kcal/mol.







HP PH HP PH
Pt- PtIN-PLANE PERPINDICULAR

104i 104a



The interaction diagram for 104a is shown in Figure 19. There is no change in the energy of the allyl a2 orbital in the complex, as expected since this orbital has a node at the carbon bonded to platinum. The only stabilizing interaction is between the la1 and 3al orbitals of bis(phosphine)platinum and a1 of the allyl fragment.





56

eV




a. 0 3a1


-7



L2t b2



-8







-9







-10







-11 a2 al C> S b2
-12

2a, a2


-13









Figure 19. Interaction diagram for 104p.




57



Highest in energy and last to be considered is the cyclopropylidene complex 105. This is a Pt(0) carbene, of which there are no examples in the literature. The perpindicular conformation 105p is calculated to be lower in energy than the bisecting conformation. The energy difference for the two conformers is high, 18 kcal/mol.





PH


PH3 PH


IN-PLANE PERPINDICULAR

105i 105p



The interaction diagram for 105p is shown in Figure 20. As in the allyl system, 104, there is a stabilizing interaction between la1 and 3al of bis(phosphine)platinum and the cyclopropylidene alorbital. There is also backbonding from the platinum b2 orbital to the cyclopropylidene b2LUMO. The decrease in the relative energy fo the allyl fragment and the cyclopropylidene fragment on complexing to platinum is due to this.





58







eV




* 3a



-7

L2P =





-8








-9







-10






b2 :

-11 a1 C>

-12


2a1
a



-13 b









Figure 20. Interaction diagram for 105p.




59


To summarize the results for the bis(phosphine) platinum complexes, the allene complex is lowest in energy. The cyclopropylidene and allyl complexes are of similar energy, 55 kcal/mol above the allene complex. From the interaction diagrams this is attributed to the high energy of the virtual 3al orbital of bis(phosphine)platinum.

As in the Fp series, the bis(phosphine)platinum complexes

of the six membered ring systems 107, 108, and 109 were studied. The EHMO calculated relative energies are shown in Figure 21. Interestingly the allene complex 107 is found to be lowest in energy. This is in contrast with the Fp system in which the allyl complex was lowest in energy. An attempt to make the platinum allene complex 107 has failed, but this is most likely due to the extreme conditions necessary to generate the allene.30

To complete the calculations with bis(phosphine)platinum, the cyclic conjugated systems 110, 111, and 112 were studied. As shown in Figure 22, the EHMO calculations place 110 and 111 at similar energies and 112 36 kcal/mol higher in energy. This is consistent with our results for the acyclic systems 103, 104 and 105. The allyl fragment was stabilized 25 kcal/mol relative to the allene fragment on complexing to platinum. Now, the cycloheptatrienylidene fragment has been stabilized 30 kcal/mol relative to the cycloheptatetraene fragment on complexing to platinum.




60


Kcal



















-70


86






Pt(PH3)2

-4 -39 85 109 Pt(PH) 2


-20 b 108



O Pt(PH3) 2 84 107 Figure 21. Relative energies of platinum complexes 107-109 as calci-Tated by EHMO.




61





Kcal









-% 36
- (PH3) 2Pt <
-33

112
( 27 O









(PH3) 2Pt


111
-3
-0 -0


(PH3 ) 2 Pt 293
110







Figure 22. Relative energies of platinum complexes 110-112.




62



While at this level of theory an energy difference of 3 kcal/mol is not significant, the comparison of this result with that for the Fp complexes clearly shows the shift favoring the cycloheptatetraene complex with bis(phosphine)platinum. This is consistent with the experimental results. In the next chapter the first synthesis of a transition metal cycloheptatetraene complex is reported.













CHAPTER III
EXPERIMENTAL RESULTS

The EHMO calculations placed the cycloheptatetraene complex 110 at approximately the same energy as the cycloheptatrienylidene complex 111. Previous calculations had shown that annelation of a benzene ring at the (4,5) position as in 113 favored the allene form over the carbene form relative to the non-benzannelated valence- isomers.53








113 114



Similarly, annelation of benzene rings at the (4,5) and (6,7) positions had an even greater effect. The bis(triphenylphosphine)platinum complex of 115 was therefore the first cycloheptatetraene complex synthesized.









115 116


63




64



Visser and Ramakers have reported in a communication that the generation of 1,2-cyclophetadiene in the presence of (ethylene)bis(triphenylphosphine)platinum, 54, gave the 1,2-cycloheptadiene complex 55 but they reported no 30
experimental details.



Ph P

Pt-PPh

+ Pt(PPh3)2(C2H 4) 54

51 55






Bennet and Yoshida have reported that generation of cyclohexyne in the presence of tris(triphenylphosphine)platinum, 118, gave the cyclohexyne complex, 119, and reported a detailed experimental procedure.54 Tris(triphenylphosphine)Br

+ Pt(PPh3)3 + Na(Hg)

117 118


Ph P


Ph P
119




65



platinum was chosen as the trapping reagent because of the greater amount of experimental detail available for its use as a trapping agent and because it was available in higher yield than 54 from the common precursor 120.





Eyt, Pt(PPh3)3

K2PtC14 + PPh3 EtOH Pt(PPh 3)4 118


120 15
Pt(02)(PPh3)2

NaBH ,

C2H4

Pt(C2H4) (PPh3)2


54



Dibenzocycloheptatetraene, 115, had never before been generated. Typically, dehydrohalogenation is a high-yield means of generating strained allenes and bromoalkene 124 was chosen as the precursor to the dibenzocycloheptatetraene for this reason. The synthesis of the bromoalkene precursor followed the method of Allison as is outlined below.55 Addition of dibromocarbene to phenanthrene gives norcaradiene derivative 122. Thermolysis of 122 at 1600C for 30 minutes leads to disrotatory cyclopropyl ring opening and formation




66

Br Br


:CBr



121 122

Br Br
r
LAH




124 123


of the dibromocycloheptatriene derivative 123. Reduction of 123 to 124 is the lowest yield step in the synthesis. Lithiumtriethylborohydride, tri(n-butyl)tinhydride, sodiumborohydride, and lithium aluminum hydride were tried, but only lithium aluminum hydride gave 124 which could be readily purified, although low in yield.

Since the dibenzocycloheptatetraene, 115, had never before been generated, the reaction of 124 with potassium tert-butoxide in the absence of a trapping agent was studied.







Br


+ t-BuOK




124 125




67


This reaction gave the allene dimer 125, as the only isolated product. Only one regioisomer is formed and its regiochemistry cannot be assigned from the spectra. However, based on the regiochemistry of the dimerization of other cyclic allenes it is assigned as above.56

The [1,2-2 -4,5-6,7-dibenzocycloheptatetraene]bis(triphenylphosphine)platinum complex, 127, was synthesized in 67% yield by the slow addition at 00C of bromoalkene, 124, to a THF solution of 126 and 118. The assignment of the



Br Ph3 P\
Pt-PPh

+ t-BuOK + Pt(PPh3)3 --O O 126 118


124 127

1 13 31
structure rests on H, C, and P NMR data (cf. Figures 23 to 27) and the elemental analysis. As will be shown, comparison of the NMR shifts of the allene carbons and the phosphorous atoms of 127 with the analogous atoms in the known (1,2-cyclononadiene)bis(triphenylphosphine)platinum complex, 57, provides strong evidence for its structure. Due to the symmetry of the carbene complex, 128,




68



H1 is equivalent to H3, C1 is equivalent to C3, and P1 is equivalent to P2, while for the allene complex 127 all these atoms are non-equivalent. The observed non-equivalence of all these atoms shows that the product cannot be the carbene complex 128, and is consistent with the allene structure.








PhP\ PPh Ph P
Pt
Pt Pt -- PPh 11 3






128 127





The 1H-NMR data for the dibenzocycloheptatetraene complex 127 and 1,2-cyclononadiene complex 57 are summarized in Table 2 along with the reported data for 1,3-diphenylallene complex 14. The non-equivalence of the hydrogen atoms is consistent with the assignment as an allene 7-complexed to a metal. Notice that for the allene bis(triphenylphosphine)platinum complexes many of the coupling constants to hydrogen




69







Ph P Pt -PPh3 127























3.0 7.s 7.0 6.5 6.0 5.5 5.0 q.5 4.0 3.5 3.0

Figure 23. 1H NMR spectrum (300MHz) of 127.




70





Pt (PPh3) 2 H3




127























5.85 5.80 5.75 5.70 5.65 PPM

Figure 24. Expansion of iH NMR spectrum of 127 showing only H3 .




71




Table 2. IH NMR data for allene complexes 127, 57,
and 14.ab


Pt(PPh3)2 Pt(PPh3)2 CHPh "4-- Pt(PPh3)2
C

CHPh

127 57 14


H1 3.48 2.4 3.75
2PtH 70.8 66 65 3JPH -- -- 5.5

3J -- -- 9.5

JHH (2.8) -- 3.0



H3 5.71 4.95 5.82

3PtH 54.6 74 -PtH
JpH 8.4 11 12.5

JPH 2.8 -- -JHH 2.8 -- (3.0)



aAll spectra were recorded in CDCl3. Chemical shifts are in ppm and coupling constants are in Hz.




72








Pt(PPh3)2 Ph H

Pt(PPh3)2 44- Pt(PPh3)2

C

Ph H
127 57
14






cannot be determined. In fact the H3-platinum coupling constant was not reported for 1,3-diphenylallene complex 14. Otsuka et al. state the resonance was broad and the platinum coupling could not be seen.12 For all the allene complexes studied the hydrogen resonances were broad and
1
of low intensity, making interpretation of the H-NMR spectra difficult.

Before this work was started no 31P NMR data had been reported for allene complexes of bis(triphenylphosphine)platinum. But, the chemical shifts and coupling constants of the phosphorous nuclei for several olefin complexes had been reported and recently White and Stang reported 31p NMR shifts for butatriene complexes 129 and 130.57These are included in Table 3 for comparison with the phosphorous spectra of the allene complexes. The non-equivalence of the




73







Ph Pt(PPh )2 Ph Ph Pt(PPh3)2



Ph Ph Ph
129 130








phosphorous nuclei in the allene complexes is not consistent with a carbene structure, while the chemical shifts and coupling constants are consistent with the assignment as an olefin n-complex.

The 13C-NMR spectra of the allene complexes 127 and 57

are very complicated. With no symmetry and 51 carbons, there is the potential for a spectroscopist's nightmare, especially if one adds long range platinum and phosphorous coupling which can lead to even more signals. Luckily, the chemical shifts of the three carbons of the allene moiety are at unusual regions of the spectrum and do not overlap with any other signals. While the 13C-NMR spectra of several bis(triphenylphosphine)platinum complexes have been reported, one author reported no coupling constants57 and the other reported platinum, but no phosphorous coupling.59 Thus, the first complete spectral data for the 13C-NMR of




74



Table 3. 31P NMR data of some (olefin)bis(triphenylphosphine)platinum complexes.a


Olefin Complex P J P J 2 Ref.
-a P -+ -b- --PP -pp CH2
II 54 32 3660 58
CH2


CH

C

c
U1 131 19.7 3590 58

C
I
CH3



CPh2
11
I 129 29 3416 26 3664 37 57
II
CPh2


CPh
II
S 130 29 2996 25.5 3550 23 57



A




75





Table 3. continued.






57 33.8 3307 29.9 3216 48.8








127 29 3208 27.5 3145 28





chemical shifts are in ppm and coupling constants are in Hz.




76










Ph P
\ PPh Pt






127



















F05 225 520 15 10 PFr Figure 25. 31p NMR spectrum of 127.




77



bis(triphenylphosphine)platinum olefin complexes is reported in Table 4.

Perhaps the most unusual chemical shift in the 13C-NMR spectrum is that of the central allene carbon, which is found at very low field, at about 160 ppm it is consistent with the formulation as an allene complex, since uncomplexed 60
allenes show a resonance for the central carbon at 200 ppm and in all cases there is a shift to higher field on complexing olefins to a metal. The resonance of the central carbon in the iron allene complexes 19 and 59is 150 and 148 ppm, respectively.31





150ppm +
\Fp



146ppm

59 19


In attempting to reproduce Visser and Ramakers' 30work it was found that the insertion product, 133, formed on mixing 1-bromocycloheptene and 118. The insertion of platinum into carbon halogen bonds has been reported previously and is useful for forming carbon-platinum sigma bonds.61,62




78



Table 4. 13C NMR data for allene complexes 57 and 127.a,b Pt(PPh3)2

Pt(PPh3)2




57 127

C1 33.8 29.5

Jpc 6.33 5.43 JPC 38.6 52.2


C2 161.6 164.4

JPC 10.1 9.6 JPC 52.3 62.9

J 490.7 -C3 110.8 109.9

J -- 5.3

JPC 9.5 10.0

2JptC 0 0


a
All spectra were recorded in CDC13. Chemical shifts are in ppm and coupling constants are in Hz.







Ph P 127







FJ













li, ISO 1'O 130 120 110 100 90 80 70 60 so 10 '1 0 Figure 26. 13C NMR spectrum (75MHz) of 127.




80







rllI i\ Pt(PPh )2






166 165 164 163 -- 16 127











C3
12 111 :O 109 i08












1
31 30 29 28

Figure 27. Expansions of 13C NMR spectrum of 127 showing C1-C3.




81









Br
Br
Ph P-Pt-PPh P+ Pt(PPh 3 -3 118 0


132 133








Since the cis insertion product, 134, would show an ABX pattern in its phosphorous NMR spectrum which might be similar to that of the allene complex, the thermal PPh3

Ph P-Pt-Br







134
Br
I
Br Ph3P -P-t -PPh3


+ Pt(PPh 3)4

0 153 124 135




82


insertion reaction of 124 with 15 studied. The product of this reaction is the trans adduct, 135, as shown by the phosphorous NMR in which there are large platinum phosphorous coupling constants of 3189Hz and 3222Hz, which is a simple 1:4:1 triplet, with equivalent phosphorous atoms.

Hydride abstraction from sigma complexes has been useful in the synthesis of transition metal cycloheptatrienylidene complexes.37-39 The reaction of insertion product 135 with trityl cation was studied in methylene chloride at -780C. This gave an insoluble material assigned



Br

Ph -Pt-PPh3 + Ph C Pt (PPh 3)2







135 136




structure 136. A similar observation was made when 137 was treated with silver ion. Since no triphenylmethane was observed and therefore no hydride abstraction had occurred, no further work with this material was attempted.





83





Br Br
Br + Ag+

Pt(PPh3) 2Br Pt (PPh3 )2

137 138



With the successful synthesis of the dibenzocycloheptatetraene complex, 127, the synthesis of the benzocycloheptatetraene complex was undertaken. This allene has been reported previously;63 so it was only necessary to generate the intermediate in the presence of the tris(triphenylphosphine)platinum.trap.

The bromoalkene precursor, 142, was prepared by the method of Waali et al., as outlined below.63 Slow addition of




Br
Br

CBr2 O


139 140


Br Br
Br
-HBr


142 141

142 141




84



this to a THF solution of 118 and 126 gave the allene complex 143. The 1H, 13C, and 31P -NMR spectra (cf Figure 28.) show that is the allene 143 and not the carbene complex. The NMR data for this compound are reported with that for the cycloheptatetraene complex, vide infra.


Ph P
Br Pt- PPh3


+ t-Bu0K + Pt(PPh) --141 126 118 143


For allene complex 143 two regioisomers are possible. From steric considerations one expects the platinum to complex away from the benzannelation, as in 143. This is






Pt(PPh )2 Pt(PPh3)2







144 143




85







Pt (PPh3) 2 H /, H1



"O
143





'1 '
3.3 3.2 3.1- 3.0 2.9 2.8 2.





















'"r, I".j bef) H
5.5 5.9 5.3 5.2 51 5. 3

Figure 28. Expansions of 1H NMR spectrum of 143 showing H1
and H3.




86



consistent with the chemical shifts and coupling constants of the allene hydrogen atoms. The dibenzocycloheptatetraene complex 127 and parent cycloheptatetraene complex were used as model systems. In complex 143 the coordinated hydrogen should be similar to the analagous hydrogen in the cycloheptatetraene complex, while in complex 144 the coordinated hydrogen should be as in the dibenzocycloheptatetraene complex. Similar logic can be used for the uncoordinated

allene hydrogen, and both are consistent with the assignment of structure 143.

Again, there was the possibility of the formation of the cis-insertion product 145. The reaction of bromoalkene 142 with 15 in benzene was studied and the trans-insertion


PPh

Ph P-Pt-Br







145 Br

Br Ph P-Pt - PPh3 + Pt(PPh 3)4 --S15 Q

142 146





87



product was formed as shown by the triplet in the phosphorous spectrum. It was postulated that the insertion product might form during the reaction of bromoalkene 142 with base and 118 and that the insertion product reacted with base to give the allene. However, there was no reaction between the insertion product and base under the reaction conditions.




Pt(PPh3)2Br


+ t-BuOK --- no reaction 126

146




It remained to synthesize the parent cycloheptatetraene complex. Bromocycloheptatriene was prepared as a mixture


0
Se0 2




147 1. BrCCBr 00
2. LAH


Br




149





88



of three isomers according to the procedure of Fohlish and Haug.64 Slow addition of one equivalent of bromocycloheptatriene to a solution of potassium tert-butoxide and tris(triphenylphosphine) platinum gave the cylcoheptatetraene complex 150, as characterized by 1H, 13C, and 31P NMR (cf Figure29,30). The spectroscopic data for the allene complex are summarized along with those for the benzocycloheptatetraene complex and the dibenzocvcloheptatetraene comolev in Tables 5-7.


Ph3 P
Pt- PPh

+ t-BuOK + Pt(PPh3 ) 3 126 118

149 150

Cycloheptatetraene complex 110 was calculated by EHMO to be of comparable energy to cycloheptatrienylidene complex 111. In the carbene complex 151 Ha and Hb are equivalent and should have the same NMR shift. To probe the possibility of interconversion of the cycloheptatetraene complex with its valence isomer, variable temperature NMR studies were performed. On heating to 800C there is no change in the IH-NMR spectrum, this is equivalent to a barrier separating the two valence isomers of at least 17 kcal/mol.65




89





Table 5. 13C NMR data for 150, 143, and 127.a,b









150 143 127 C1 27.7 26.41 29.5

2JpC 5.1 4.9 5.4 2JpC 44.1 43.9 52.2 C2 151.0 156.1 164.4

2JpC 10.4 9.7 9.6 2JpC 61.0 62.3 62.9 C3 114.5 113.0 109.9


3JPC 9.3 -- 10.0 3JpC 5.6 -- 5.3 aAll spectra were recorded in CDC13. Chemical shifts are in ppm and coupling constants are in Hz.




90










Table 6. 31P NMR data for allene complexes 150,
143, and 127.a'b







150 143 127


P1 31.5 30.5 29.0

2pp 33.3 30.1 28.0 1JP 3267 3297 3208



P2 27.3 26.2 27.5

1Jptp 3170 3131 3145 aAll spectra were recorded in CDC13. Chemical shifts are in ppm and coupling constants are in Hz.




91




Table 7. 1H NMR data for H1 and H3 of 150, 143, and 127.ab










150 143 127 H1 2.94 3.02 3.48

2 PtH 70.4 72 70.8
PtH

JpH 11 13.2 -*H2 4.63 5.16 5.71

3JPtH 70.8 66 54.6 4JpH 9 9 8.35


4JpH 7 4.5 2.8 JHH -- -- 2.8



aAll spectra were recorded in CDC1 at 300MHz. Chemical shifts are reported in pam and coupling constants are in Hz.




92
















Pt (PPh 3) 150


















Figure 29. 1H NMR spectrum (100MHz) of 150.




Full Text

PAGE 1

THE FIRST SYNTHESIS OF A TRANSITION METAL COMPLEX OF CYCLOHEPTATETRAENE By WILLIAM RANDOLPH WINCHESTER 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 1985

PAGE 2

TO MY PARENTS

PAGE 3

ACKNOWLEDGEMENTS By far the most enjoyable aspect of writing a dissertation is the opportunity to acknowledge assistance received. I would like to thank Professor William H. Jones for his constant encouragement and patience. Professor J. R. Sabin was essential in helping the strange bedfellows of theory and experiment develop together. It is pleasant to acknowledge the assistance of the Quantum Theory Project in all the facets of my education, but especially for the opportunity to attend the Summer Institute in Quantum Chemistry and to attend the annual Sanibel Symposia. It is my pleasure to acknowledge the capable assistance of and discussions with George Purvis, Marc Radcliffe and Seth Elsheimer. Finally, I want to thank Joan Raudenbush for her constant assurances I would finish and her capable assistance in making that come true. in

PAGE 4

"One takes up fundamental science out of a sense of pure excitement, out of joy at enhancing human culture, out of awe at the heritage handed down by generations of masters and out of a need to publish first and become famous . " Leon M. Lederman The Value of Fundamental Science Scientific American 251 #5, November, 1934 IV

PAGE 5

TABLE OF CONTENTS Paae ACKNOWLEDGEMENTS iii ABSTRACT v ii CHAPTER I INTRODUCTION 1 II RESULTS OF EXTENDED HUCKEL CALCULATIONS 20 III EXPERIMENTAL RESULTS 63 IV SUMMARY 116 V EXPERIMENTAL 118 General ' 118 7 , 7-dibromo-2 , 3-4 , 5-dibenzo bicyclo [4.1.0] heptane ( 122 ) 119 4, 5-6, 7-10,11-12, 13-tetrabenzotricyclo[ 7. 5. 0.0 2 ' 8 ]tetradeca-2 , 3-4 , 5-6 , 7-10 , 11-12 , 13-14 , 11-hexane ( 125 ) 119 2 (1 , 2-n -4,5-6 , 7-dibenzocycloheptatetraene) -bis (triphenylphosphine) platinum ( 127 ) 120 2 (1, 2-n -1, 2-cyclononadiene)bis (triphenylphosphine) platinum (57) 121 trans(bromo) (n -cyclohepten-1-yl) bis (triphenylphosphine) platinum ( 133 ) 121 trans(bromo) (n -3 ,4-5, 6-dibenzocycloheptatrien-lyl) bis (triphenylphosphine) platinum ( 135 ) 122 2 (1, 2-n -4 , 5-benzocycloheptatetraene) bis (triphenylphosphine) platinum (143) 123

PAGE 6

trans(bromo) (4-n -1 , 2-benzocycloheptatriene-4yl) bis (triphenylphosphine) platinum ( 146 ) 124 Tropone ( 148 ) 125 l-,2-, and 3-bromocycloheptatrienes ( 149 ) 125 2 (1 , 2-n -cycloheptatetraene) bis ( triphenylphosphine) platinum ( 150 ) i25 trans(bromo) (1-n -cycloheptatrienyl) bis (triphenylphosphine) platinum ( 152 ) 126 7,1' -cycloheptatrienylcycloheptatriene-1 , 1 ' -diy 1) bis (triphenylphosphine) platinum ( 153 ) 127 (7 '-cycloheptatrienyl-7-cycloheptatriene-l,l '-diyl) (1, 2-bis-diphenylphosphinoethane) platinum ( 163 ) — 128 (7 '-3 ' ,4 '-benzocycloheptatrienyl-7-3,4-benzocycloheptatrien-1,1 '-diyl) bis (triphenylphosphine) platinum ( 164 ) 129 Reaction of 13 5 with potassium tert-butoxide 130 The reaction of heptafulvalene with potassium tertbutoxide and tris (triphenylphosphine) platinum 130 APPENDIX REFERENCES BIOGRAPHICAL SKETCH 132 180 185 VI

PAGE 7

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE FIRST SYNTHESIS OF A TRANSITION METAL COMPLEX OF CYCLOHEPTATETRAENE By William Randolph Winchester Chairman: William M. Jones Cochairman: J. R. Sabin Major Department: Chemistry It is well established that cycloheptatetraene is lower in energy than cycloheptatrienylidene. To date all transition metal complexes of this ligand have been aromatic carbene complexes, no transition metal cycloheptatetraene complex has been reported. Extended Huckel molecular orbital calculated relative energies correctly show cycloheptatetraene to be lower in energy than cycloheptatrienylidene and also show that for the dicarbonylcyclopentadienyliron (Fp) complex the aromatic carbene complex is lower in energy than the cycloheptatetraene complex. This result is interpreted as resulting from a strong bonding interaction between the Fp LUMO and the cycloheptatrienylidene HOMO, which is much vii

PAGE 8

less strong for the cycloheptatetraene-Fp complex. Further, this result led to the choice of bis (phosphine) platinum as a metal for which the cycloheptatetraene complex would be lower in energy than the cycloheptatrienylidene complex. The first transition metal cycloheptatetraene complex was synthesized by the generation of cycloheptatetraene in the presence of tris (triphenylphosphine) platinum and 2 exchange to give (n -cycloheptatetraene) bis (triphenyl2 phosphine) platinum. Similarly, the 1,2-ri -4,5-benzo2 cycloheptatetraene and the 1,2-n -4 , 5-6 ,7-dibenzocycloheptatetraene complexes have been synthesized. Interestingly, none of these compounds is fluxional. A novel reaction between cycloheptatetraene and the cycloheptatetraene complex to give (7 ' -cycloheptatrienyl7-cycloheptatrien-l,l '-diyl)bis (triphenylphosphine) platinum, which has been characterized by H, C, P NMR spectra and an x-ray structure determination, was observed. This is the first (allene) bis (triphenylphosphine) platinum complex which has been observed to undergo this reaction. A mechanism which is consistent with this result is proposed. vm

PAGE 9

CHAPTER I INTRODUCTION Allenes, or 1,2-dienes, are linear compoundswith a 2 central sp-carbon bonded to two sp -hybridized carbons. Transition metal complexes of allenes, in which the metal is bonded to one of the double bonds, form an interesting H. H > i=ctc; ML n group of molecules and have been the subject of several 1-5 reviews . Crystal structure determinations have been done for a number of allene complexes. The usually linear allene is bent when complexed to a transition metal. The carbon which is not bonded to the transition metal moves away from the metal. This bending is consistent with descriptions of olefin complexes as metallocyclopropanes .

PAGE 10

H H H \ C . C II H \\ C (\ H H H H / v A LINEAR 1 H I ^ C H \\ c METALLACYCLOPROPANE The metal carbon distances are not svmmetrical, the MC, bond distance is always lonqer than the MC~ bond distance. This has been attributed to the different hybridizations of CI and C2 , s ince increasing s-character 4 in an orbital generally shortens the bond length. It has also been attributed to the overlap of an occupied metal H \ 2 >3 __c | 31 sp ' /\ H H

PAGE 11

d-orbital with the orthogonal 7r*-orbital of the double bond which is not complexed to the metal. H W H U M O c 4\ H H The bonding in transition metal allene complexes is usually discussed within the context of the Dewar7 — 8 Chatt-Duncanson model. There is donation from the olefin ir-orbital to an unoccupied metal a, -orbital and backbonding from an occupied metal b, -orbital to an olefin Tr*-orbital . The bending of allene upon complexing to a metal has been attributed to metal backbonding. c II <2>co | C>mcO 4 \ H H H H V II c<^ p 1W 1V1 H H

PAGE 12

Rotation about the axis of the olefin transition metal ir-bond is rapid in some allene transition metal complexes and in those cases for which rotation is rapid there is also a rapid exchange of the metal between the orthogonal double bonds. For symmetrical allenes the resulting configurations H H H H \* X / C C H-M II c — C II +M A H H 8 are equivalent and this exchange is called fluxional 9 behavior. Fluxional processes have been studied by several authors and as will be discussed below the mechanism appears to depend on the metal and its oxidation state. 10 ~ 13 The fluxional exchange has been studied for dicarbonylcyclopentadienyliron (Fp) and has been shown to be intramolecular. It was shown there is no loss of optical activity during the fluxional exchange of 1, 2-cyclononadiene complexed to Fp. This result excludes allyl complex 10_ as a possible intermediate in the Fp exchange reaction.

PAGE 13

H H \ s C it c c H H Fp A K 2 Fp — c : + >3H 2 9_ 10 The work of Cope et al. with platinum(II) complexes of 1, 2-cyclononadiene, 11, is in direct contrast to the work with Fp. When optically pure 1, 2-cyclononadiene was used to make the complex with platinum, repeated recrystallization led to a decrease in enantiometric purity. When the 1, 2-cyclononadiene was released from the platinum it was found to be of lower optical purity than the starting material Two structures were proposed for the intermediate 11 Pt(Am) (CI) Pt(Am)(Cl) 2 * 11 Pt(Am)(Cl) CI 12 13

PAGE 14

in the mutarotation, but ir-allyl complex 13 is the more likely intermediate since other platinum allene complexes have been shown to rearrange to ir-allyl complexes, vide infra, Allene complexes of platinum(O) with phosphine ligands 12 have been studied, and show no fluxional behavior. The H-NMR spectrum of 14 shows two sianals for H and H, at — a b all temperatures, with decomposition occurring before coalscence is observed. Ph H, v b / PPh 3 A PPh 3 H Ph a 14 As with structure and bonding, the synthesis of allene transition metal complexes has similarities to olefin transition metal complexes, with some interesting differences which can be attributed to the uniqueness of the allene. The majority of allene complexes has been prepared by exchange reactions; the allene is either added to or generated in-situ with the transition metal and displaces a ligand from the metal. For example see Figure 1.

PAGE 15

Ref Pt(PPh 3 )^ ii H H t> C ii c II A H H H H \ + -*— Pt(PPh-)_ C J 2 A H H 16 14 **t CT* 10 12 18 19 Fe 2 (C0) 9 Me Me r ti C ii A Me Me Me Me \ + C c II A Me Me Fe(CO)^ 15 20 21 22 Figure ^. Synthesis of transition metal allene complexes

PAGE 16

Figure 1. Continued. Ref Rh(PPh 3 ) CI H H t# C -fl-Rh(PPh-) ,C1 V C J * i c / \ H H 23 24 ((C 2 H 4 )PtCl 2 ) 2 + H H v C H H c II c — ( -H-Pt(ci) 2 ) 2 II II c / \ Me Me A Me Me 16 25 26 27

PAGE 17

Several allene complexes have resulted from intramolecular rearrangements. Treatment of 28 with acid leads to the allene complex 9 . While it is possible the allyl cation 1_0 is an intermediate, it is most likely there is neighboring group participation of the iron. Fp OMe 28 .CH Fp CH 2 FP'O^Me 10 29 + CH 2 fp— H2 C II CH 2 9 While studying hydride abstraction from 30, it was found that the unusual reagent, 31, gave the Tt-complex of 1 8 allene. Alpha hydride abstraction gives 32, a highly strained alkylidene, which rearranges to the ir-complex.

PAGE 18

Fp 30 10 • Fp 31 / FP + ^ II 32 This reaction has been studied starting with resolved 3_3. 19 Treatment of 3_2 with trimethylsilyltrif late leads to the formation of optically inactive allene complex 3 6 The racemization was interpreted as occurring via initial formation of carbene 3_4, followed by opening to achiral OMe TMSOTf 33 Fp^ II 34 36 FP— ^ Me Me 35

PAGE 19

11 allyl cation 3_5 and finally rearrangement to the allene complex 36 . In what appears to be the reverse process of 3_5 to 36 , platinum(II) allene complexes have been observed to 20-23 rearrange to ir-allyl complexes. Allene displaces acetone from 3J7 to give an allene complex, 3_8, which can be isolated below 0°C. If this is allowed to warm to room temperature in solution, ir-allyl complex 39 is formed. Me Q ™2 „ I 1 1 2 ° c | CH 9 RT Q-Pt-Q + C -^-=» Me-Pt-tf 2 — 4-1 " I C fV4 2 — J Me 37 1 3j8 3? This discussion of allene complexes has led to the introduction of complexes of the general types £0_, 41, and 4_2. The interconversion of metal complexes of this >m # 40 M M ^=1 41 • 42

PAGE 20

12 type is of interest, but has been studied very little. However, the interconversion with no metal has been thoroughly studied. When cyclopropylidene 43^ is generated from an optically active precursor, the product, 46, is the optically active allene indicating that the achiral, planar 24-27 allene, 45, is not an intermediate in the reaction. NO I Ph Es Ph 45 Ph 44 Ph \ /=c=ci;» h H 46 optically active 45 Allenes which have been incorporated into a ring have been the subject of more recent investigations. 28-29 u 47

PAGE 21

13 Incorporating an allene into a ring forces it to bend about the central carbon and twists the Tr-bonds from their usual perpendicular geometry. This raises the energy of the allene closer to the bent, planar geometry 4_8 which experiences little deformation on incorporation in a ring. * V 4 3 Cyclic allenes with fewer than eight carbons cannot be isolated and indirect methods have been develoDed to probe the structure of these molecules. A cyclic allene has C 2 symmetry and therefore is chiral. Its planar valence isomer, 48, has C symmetry s V C 2 -s £7 £8 2 ft and is therefore achiral. Balci and Jones have used this difference to probe the structure of the six and seven membered ring compounds . Their results show that the product

PAGE 22

14 of dehydrohalogenation is chiral and therefore for the six membered ring the allene, 47, is still lower in energy than its planar form. This is in accord with recent FORS 3r 49 + Menthoxide H *.. 0" optically active 50 (Full Optimized Reaction Space) calculations which show the allene to be preferred by 13.1kcal/mol for the six 29 membered ring. 13 '"Cr" 48 kcal/mol -0.0 ty 47

PAGE 23

15 There are two reports of the synthesis of strained cyclic allenes complexed to transition metals in the 30 literature. Visser and Ramakers trapped the seven, eight, and nine membered cyclic allenes with 54 The + p-t(pph 3 ) 2 (c 2 jy 54 Pt(PPh 3 ) 2 resulting ir-complexes were crystalline, air stable complexes that were not fluxional. The Fp complex of 1, 2-cycloheptadiene, 59, has been synthesized by methoxy abstraction from 58_. This compound OMe 58 + TMSOTf Fp + 59 is fluxional with a 13 . 9kcal/mol activation barrier for the 1,2-shift.

PAGE 24

16 Cycloheptatetraene, 60, is a strained, cyclic allene which has been the subject of much interest. 32-36 The planar form of this allene is aromatic and this decreases the difference in energy between the planar form 61 and the allene 60.. 60 61 Calculations are all in agreement that cycloheptatetraene is lower in energy than its planar form. Waali, 34 using the MNDO method, has calculated the vibrational force constant matrix and normal modes for the planar form 6_1 and his results are consistent with the planar form 6_1 as a transition state connecting the enantiomeric cycloheptatetraene geometries. Method AE Ref INDO 13.8 3 6 MNDO 23 3 7 EHT 28 3 8 STO 4-31G 15.3 39 7 6 AE

PAGE 25

17 32 Harris and Jones studied the dehydrohalogenation from optically active bromocycloheptatriene 62. T h e isobenzofuran adduct of cycloheptatetraene is optically active, consistent with the calculated lower energy of cycloheptatetraene relative to its planar form 61. H D 62 3r + t-BuOK o 60 53 No transition metal cycloheptatetraene 7r-complex has been synthesized. But many transition metal complexes of its planar valence isomer have been synthesized. In the original work, 7 reaction of sigma compound 64 with the hydride abstractor 65 gave the cycloheptatrienylidene complex 66. The compound was sufficiently novel that an cr 64 + Ph„C" 65 er 66

PAGE 26

x-ray crystal structure was obtained for it and its benzannelated derivative. 8 ~ 39 The crystal structure shows unequivocally that the ligand is planar and sigma bonded to the metal. Hydride abstraction to give the carbene complex has proven to be a very general reaction and Table 1 gives some of the compounds made and the 13 C-NMR shift of the carbene carbon. The allene complex, 67, was not observed in any of the reactions, even as a side product. 67 It was the primary objective of this research to synthesize the first transition metal complex of cycloheptatetraene. First, molecular orbital calculations were used to study the bonding in 40, 41 and 42. Then using the results of these calculations a transition metal system was chosen and the first transition metal cycloheptatetraene complex synthesized.

PAGE 27

19 13, Table 1. c NMR chemical shifts (ppm) of the carbene carbon in some cycloheptatrienylidene transition metal complexes . Structure M=Fe M=Ru Mp 68 242.3 223.6 265.9 244.7 69 70 \\ Jp= Mpp + 71 201 278.8 186.6 256.2 Mp= CO CO Mpp= / M N CO PR,

PAGE 28

CHAPTER II RESULTS OF EXTENDED HUCKEL CALCULATIONS The bonding in the transition metal complexes 40, 41, and £2 was studied using molecular orbital calculations of the extended Huckel type. 40 ' 41 The transition metal complexes H C=C=tCH i < i n M M M 40 41 42 with M=Fp were studied first to help understand the experiments discussed in the introduction. Then the transition metal complexes with M=Pt(PH 3 ) 2 were studied because it was expected a cycloheptatetraene complex of this system could be synthesized and because it was expected that studying the bonding in a different transition metal complex for comparison with Fp complex would be useful. 20

PAGE 29

21 The relative energies calculated by the extended Hiickel molecular orbital method (EHMO) for 7_2, 73 and 74 are shown in Figure 2. Experiments have shown that generation of both the cyclopropylidene complex 74 and the allyl complex 73 leads to the formation of the allene complex 72, indicating that the allene complex 72 is the preferred valence isomer. 17 ' 18 m agreement with this experimental result the allene complex 72 is calculated to be lowest in energy. Allyl complex 73_ i s next highest in energy, 24 kcal/mol above 72. The cyclopropylidene complex 74, is calculated to be highest in energy, 33 kcal/mol above the allene complex 72 . A minimum for the energy difference between the allene complex J_2 and the allyl complex can be deduced from an earlier reported experiment. 17 There is no loss of optical purity on refluxing the optically active allene complex 19 for two hours at 80°C. If one assumes a preexponential £0 Fp 19 75

PAGE 30

22 Kcal 33 2k 74 Fp— = 73 Fp 72 Figure 2. Relative energies of 72, 73 and 7_4 in kcal/mol

PAGE 31

23 42 13 term in the Arrhemus equation, A, of 10 and a Tj of two hours then the calculated barrier is 30 kcal/mol. This also assumes that the transition state for racemization E = R T° [ln(T. ) + 35.73] a k %,sec is the allyl cation and not some intermediate in the conversion of the allene to the allyl cation. The relative energies of allene 7_6, planar allene 77 , and cyclopropylidene 7_8 were calculated for comparison with the Fp complexes. As Figure 3_ shows, allene, 76, is lowest in energy, followed by planar allene 77, and highest is cyclopropylidene 78. The ordering by energy is the same for the organic molecules as for their Fp complexes. But, the differences in energy of the organic molecules are much greater than in the Fp complexes. In order to understand this decrease in relative energy on complexing to Fp and to understand the bonding in the Fp complexes some interpretation of the EHMO calculated molecular orbitals is necessary. The interpretation of calculated molecular orbitals for organometallic complexes has been done by arbitrarily dividing the molecule into an organic fragment 43-46 and a metal fragment. The molecular orbitals from the calculation for the molecule are discussed with an interaction diagram in terms of the orbitals calculated for the fragments. This approach was used in discussing the bonding in 72, 7 3 and 74.

PAGE 32

24 Kcal 11A 78 -80 77 '2 76 33 Fp + = 2^ Fp — <^+ 74 73 Fp + -||72 Figure 3_. Relative energy of 76-78 as calculated by EHMO,

PAGE 33

25 Before the calculated molecular orbitals and energies are discussed in more detail, the method used to obtain these will be described. For all the organic molecules MINDO/3 geometry optimizations were performed. 47 The resulting geometries were used for the EHMO calculations. A similar optimization could not be done with the organometallic complexes because MINDO/3 is not parameterized for transition metals. Initially the EHMO calculations on the organometallic complexes used the geometries found from optimization of the organic fragments combined with the known geometry of the metal fragment. The EHMO calculated relative energies were not compatible with the experimental results, 7_2 was found to be higher in energy than 73. A study of the effect of bending in the allene ligand on the EHMO total energy is shown in Figure 4. There is an H H \ S c + c J C A H H H I C C \ F P + ^; ccc bend p H H Linear Bent 8 = 180° e < !80°

PAGE 34

18 kcal/mol drop in total energy on bending the allene ligand from 180° to 150°, which is enough to make the allene Fp complex lowest in energy, this bend is also consistent with the known geometry of Fp allene complexes. 48 Therefore, it was decided that using the geometries of the uncomplexed ligands would not be acceptable for the complexes. Since geometry optimizations using the EHMO method can lead to spurious results, geometries obtained from the MINDO/3 optimization of the model systems 79_, 80 and 81 were used for the organic fragments in the organometallic complexes. Using this procedure assumes that the bonding in the oxygen complexes will be similar to the bonding in the Fp complexes, which is reasonable since the oxygen atom is isolobal with Fp. This procedure led to an angle of 161.3° in the geometry used for the complexed allene. /\ i !! H C C CH 2 79 A 81

PAGE 35

27 SI a < o X o Eh < a .x 1*5 155 ANGLE OF CCC LINK IN 72. 165 175 Figure 4_. Affect of bending on the energy of 7 2

PAGE 36

28 The allene valence orbitals are shown in Figure 5. Because the allene geometry used was not linear, degeneracy of the it orbitals is lifted. While only the In and the 2tt* orbitals are of ir-like symmetry the tt symbol is used for all four orbitals for ease of communication. The 2ir and the Itt* orbitals have been raised and lowered in energy respectively. Note the large energy difference between the occupied and unoccupied orbitals. Allene is a stable, closed shell molecule and it will not interact with the metal fragments as strongly as will a fragment with non-bonding electrons. The valence orbitals of the Fp fragment, 82, as calculated by the EHMO method are shown in Figure 6. Several groups have reported calculations on Fp compounds and all have found it useful to discuss their results using the Fp fragment 49-50 valence orbitals. The LUMO, 3a', is important for the strong a-bonding with incoming ligands. The HOMO, la", is well suited for backbonding to the ligand, both energetically and geometrically . The calculated two lowest energy conformations of 7 2 are upright, 72a , and bisecting, 7 2b , and the bisecting conformation is preferred by 20 kcal/mol . This conformation minimizes the destabilizing steric interaction between the cyclopentadienyl ring and the allene ligand and maximizes the allene LUMO overlap with the Fp HOMO, (cf . Figure 7) . The experimental value for the barrier to rotation is

PAGE 37

eV 29 -7.9 Lir""©^. -9.2 Iff* ©0 -12.6 zn 00 -13.0 IT ®=^= Figure 5. Allene, 76, valence orbitals

PAGE 38

eV 30 -8.3 -10.9 -12-5 -12.7 -12.9 2a" '3»' 0§0 -la" •2a* •la' Figure 6. Fp, 82, valence orbitals

PAGE 39

31 H Fe ' H Fe j 1 1C t OC C c / \ H H BISECTING UPRIGHT 72b 72a 8.4 kcal/mol, which shows the calculated energy difference for 72a and 7 2b is probably too high. The rotational barrier for the ethylene complex 83_ has been calculated to be 21 kcal/mol, again a very high value. The authors state, ". ..in general extended Huckel energies are not 49 reliable and must be viewed as indicative of trends." H oc v / ^ oc H ^ H 83

PAGE 40

32 The interaction diagram for Fp with allene in the bisecting conformation is given in Figure 7_. The molecule has no symmetry and the diagram is a simplified' one with only the major interactions shown. The allene HOMO is a bonding orbital and much lower than the Fp 3a' LUMO. The bonding orbital resulting from this interaction is only slightly lower in energy than the allene HOMO. Similarly, a large energy difference between Fp la" and the allene Itt* results in a small lowering in the resulting bonding orbital. The bisecting allyl complex 73b is calculated to be 24 kcal/mol above the allene complex 72 and 3 kcal/mol lower than its upright isomer 73a . From the valence orbitals of. the allyl fragment 77 , shown in Figure £, it is apparent the HOMO, a,, is a non-bonding orbital. The LUMO, a~, is of the correct symmetry for ir-backbonding Fe OC '1 OC BISECTING UPRIGHT 73b 73a

PAGE 41

eV -7 33 -8. ZTT* 2a" -9-10 -11 •. -12 • -13 + -*— lit Figure 7. Interaction diagram for 7 2

PAGE 42

34 -7.9 r -11.0 -11.7 -13.7 1 i^o — a 3 -^ o -i 54* 1 o^o Figure 8_. Allyl fragment valence orbitals

PAGE 43

35 but a node at the central carbon leads to a small overlap with other fragments and little interaction. Most striking in the interaction diagram for 73 , shown in Figure 9_, is the lack of change or interaction in the a„ orbital of allyl fragment 77_. Next, there is a bonding orbital formed by the virtual Fp 3a' orbital and the occupied allyl a orbital. Finally, there is a stabilizing interaction of the allyl b orbital with the Fp la' orbital. Notice that the bonding orbital resulting from the allyl a, orbital and the Fp 3a' orbital is lowered by almost 2 eV. The allene 2tt orbital interaction with the Fp 3a' orbital results in only a . 5 eV lowering (cf. Figure 7) . The decrease in relative energy of 7_6 and 7_7 on complexing to Fp appears to be due to the different ligand-HOMO, Fp-LUMO interactions. The Fp-cyclopropylidene, 7_4, is the highest in energy of the Fp complexes, 3 4 kcal/mol above the Fp allene complex 72 . The valence orbitals of the cyclopropylidene fragment, shown in Figure 10_, are as one would expect. The HOMO, a., is a non-bonding orbital, associated with the lone pair of the carbene. The LUMO, b , is of the correct symmetry for backbonding.

PAGE 44

eV -7 • 36 -8. 2a" -9--10 • Figure 9_. Interaction diagram for 73

PAGE 45

eV 37 -10.8 11.7 a, O-13. ^ », <&^± Figure 10. Cyclopropylidene valence orbital s .

PAGE 46

38 The upright conformer 74a is calculated by EHMO to be lower in energy than the bisecting conformer, 74b, by 5 kcal/mol. The low difference in energy of the two conformations is the result of two opposing affects. In the upright conformation Tr-backbonding is strongest and in the bisecting conformation the steric repulsion between the cyclopentadienyl ring and the cyclopropylidene ligand is smallest. OC _ + + Fe Fe^. / V7 oc J ^^ oc v oc UPRIGHT BISECTING 74a 74b

PAGE 47

39 From the interaction diagram for 74a , shown in Figure 11 , it is apparent the strongest interaction is between the cyclopropylidene a, orbital and the Fp 3a 1 orbital; the a -bond formation. As in the Fp allyl complex this a-interaction is worth almost 2 eV. There is also a stabilizing interaction resulting from backbonding from the Fp la" HOMO to the b, LUMO of 7_8. This backbonding accounts for the greater stabilization of the cyclopropylidene 7_8 relative to the planar allene 77 on complexing to Fp . Since the acyclic allenes are known to be bent when complexed to Fp, an allene which has been incorporated in a ring should be less strained when complexed to a metal than when free. At the same time EHMO calculations (vide supra) predict the allyl geometry to be closer in energy to the allene when complexed to Fp . These are opposing effects; the first will increase the difference in energy between the complexed allene and the second will decrease the difference. The relative magnitude of the two effects will determine the size of the ring necessary to make the Fp allene complex equal in energy to the Fp allyl complex, The six memebered ring allene 84_, its allyl isomer 8 5 and 8 6 and their Fp complexes were therefore studied.

PAGE 48

eV -7. 40 -8.. 2a" -9-Ppi -10 •• -11 .. OflO 3a' -12 © <5> la -13 a ! O
PAGE 49

41 • « ^ 84 85 86 The geometries of the uncomplexed valence isomers 84 , 11' and 86_ were optimized using MINDO/3 and the EHMO calculations were done on the optimized geometries, the resulting calculated relative energies are shown in Figure 12, The cyclic allene 84_ i s calculated by EHMO to be lower in energy than its allyl isomer by 40 kcal/mol and is 70 kcal/mol lower than bicyclocarbene 8_6_. These relative energies are smaller than those for 7_6, 7_7 and 7_8 following the expected trend. Experiment has shown the allene to be the preferred geometry, but the next accessible geometry and electronic state have not been unambiguously , 28,29 assigned. ' The geometries for the EHMO calculations on the Fp complexes were obtained by optimizing the corresponding oxides 87, 88 and 89_ with MINDO/3 . The EHMO calculated relative energies, shown in Figure 1_8, place the Fp allyl

PAGE 50

42 Kcal •11* 78 -30 77 70 86 — 4 ° Cj ii CH, ii ' C ii CH, 76 84 Figure 12. Relative energies of 84-86 as calculated by EHMO.

PAGE 51

43 Kcal 70 86 — 6 85 Fp 1 -23 -11 ,*P _ 21 Fp 92 91 90 Figure r3. Relative energies of 90-9 2 as calculated by EHMO,

PAGE 52

87 4 4 88 89 complex ££ lowest in energy, 14 kcal/mol below the Fp allene complex 9_1. The bicyclo (3 .1 . 0) hexane carbene complex 9_2 is highest in energy, 23 kcal/mol above the allyl complex 9_0. While these results corroborate intuitive expectations, they must await experimental verification. Finally, the cyclic-conjugated seven membered ring analogues 93_, 94_, and 9_5 were studied. Recall that 9_3 has been shown to be the preferred valence isomer by both experiments and calculations using the MNDO,STO-3G and STO 4-3 1G methods. The geometry of 93, 9_4, and 9_5 was optimized using MINDO/3 and the EHMO calculation was then performed using the MINDO/3 geometry. Consistent with the earlier

PAGE 53

45 results, the cycloheptatetraene, 9_3, was lowest in energy 29 kcal/mol lower than 9_4 . ' 37 ' 39 Highest in energy, 94 was found to be 51 kcal/mol above cycloheptatetraene. The Fp complexes were studied next. First the geometry °f iL£' 22.' and 2JL was optimized using the MINDO/3 96 97 98 program. Then EHMO calculations were clone replacing the oxygen with an Fp fragment. The EHMO calculated total energies, shown in Figure. 15_, place the cycloheptatrienylidene complex 100 lowest in energy with the Fp cycloheptatetraene complex 9_9 next, 37 kcal/mol higher in energy and 101 highest, 39 kcal/mol higher in energy. NMR experiments and a crystal structure determination show 100 to be the preferred valence isomer, in agreement with the EHMO results. There is no experimental evidence for the relative energies of 99 or 101 to 100. 37-39

PAGE 54

4 6 Kcal llif aa CHII 2 C II CH 2 78 77 51 •29 95 94 93 76 Figure 14. Relative energies of 93-95 as calculated by EHMO,

PAGE 55

47 Kcal -33 •27 / 93 95 94 Fp = Fp -37 Fp 101 100 o 99 Figure 1J5. Relative energies of Fp complexes 99101 as calculated by EHMO.

PAGE 56

48 The explanation for the preference for 99 relies on our earlier results. Recall that allyl fragment 77, was found to be 80 kcal/mol above allene, 76, in energy. But, for the Fp complexes cycloheptatrienylidene, 94, was found to be only 24 kcal/mol higher in energy than cycloheptatetraene 93. The non-bonding allyl HOMO has a much stronger interaction with the Fp 3a 1 LUMO than does the bonding allene HOMO. For the cyclic-conjugated systems the non-bonding HOMO of 94 has a stronger interaction with the Fp 3a* LUMO than does the HOMO of 93 and the uncomplexed molecules are much closer in total energy, only 29 kcal/mol different. Therefore, one could expect that complexing to Fp would lower the energy of 9_4 sufficiently to make it lower than 93_. The energy difference for the upriaht and bisecting conformations of 100 was calculated by EHMO as 9.3 kcal/mol, with the bisecting conformation, 100b, preferred . This is »Fe \^<5^ Fe + ° c oc V/ ° c ° c ^w> UPRIGHT BISECTING 100a 100b

PAGE 57

49 in agreement with the crystal structure determination which found the cycloheptatriene ring parallel to the cyclopentadienyl ring. The rotational barrier for 10 2 has been determined by D-NMR studies to be 9.7 kcal/mol, fortuitously close to the calculated barrier for 100. 102 With the completion of the study of the Fp complexes, a new metal fragment was chosen for study, the primary criterion for the choice was experimental evidence which might indicate the metal complex of 9_3 would be lower in energy than the complex with 94. Bis (triphenylphosphine) platinum has been used to form complexes of several reactive olefins, ' so it was decided to study this metal system using the same organic fragments as in the Fp study. Trends in the relative energy as well as qualitative changes in the bonding were studied. The relative energies of 103 , 104 , and 105 were calculated using the EHMO method. For the calculations phosphine was substituted for triphenylphosphine and the

PAGE 58

50 organic fragment geometries were the same as in the Fp calculations. The substitution of phosphine for triphenyl49 50 phosphine has been used by several authors ' and other than the steric differences it is assumed that the relative energies are not affected by the substitution. The EHMO calculated total energies show 103 to be lowest in energy. Complexes 10 4 and 105 are of similar energy and are higher in energy than 103 by almost 55 kcal/mol Comparison of these calculated relative energies with those for the uncomplexed molecules 7_6, 77_ and 7 8 show the relative stabilization to be 25 kcal/mol for the allyl complex 104 and 60 kcal/mol for the cyclopropylidene complex 105 . It is also apparent the bis (phosphine) platinum complexes are significantly different from the Fp analogues and an explanation for the difference would be useful. Interaction diagrams for 103 , 104 and 105 will be used to study the bonding in these molecules. The valence orbitals of bis (phosphine) platinum, shown in Figure 17 , are very different from those of Fp even though the HOMO and LUMO are of the same symmetry. The HOMO of 106 , b 2 , is high in energy, 0.7 eV higher than the Fp HOMO. The LUMO of 106, 3a', at -6.5 eV, is too high for a strong interaction; it is 4.4 eV higher than the Fp LUMO. There would be no a-interaction with 10 6, but a combination of the la 1 and 3a-, orbitals lead to a a-interaction when complexes are formed.

PAGE 59

51 Kcal 114 o78 80 o77 J _ *2 76 55 5^ L 2 Pt L 2 Pt II .o V* -H105 104 103 Figure 16. Relative energies of platinum complexes 103 -105 as calculated by EHMO.

PAGE 60

eV 52 -6.5 3»i O -11.8 -12.3 -12.6 2a, la, 1 ^ @ ® Q4P a 2 b l la l H 3 P v H 3 r Pt 106 Figure 17. Valence orbitals for bis (phosphine) platinum fragment 10 6.

PAGE 61

53 Lowest in energy of the three valence isomers, the allene complex with bis (triphenylphosphine) platinum, is a known compound. NMR experiments are in agreement with a crystal structure determination showing the in-plane geometry, 103i , to be preferred over the perpindicular 14 geometry, 103p . No rotational barrier has been reported; presumably decomposition occurs before rotation. The calculations are consistent with this, the square planar geometry, 103i , is lowest in energy and the calculated difference in energy is 41 kcal/mol. h 3 p n y m 3 h 3 p^ph 3 Pt Pt K ir H h^ ^ H IN-PLANE PERPINDICULAR 103i 103p An interaction diagram for the in-plane geometry, 103i , is shown in Figure 1_8. The bis (phosphine) platinum b 2 * orbital interacts with the allene Itt orbital to form a bonding orbital in the complex. It is the lack of a comparable interaction in 103p which leads to a large

PAGE 62

eV 54 O 3a -7--8-. -9t -10--11 . © ^ b -12 • •DO a -13 -• 'igure 18_. Interaction diagram for 103i ,

PAGE 63

55 rotational barrier. There is also a stabilizing .interaction between the la, and 3a orbitals of bis (phosphine) platinum and the allene 2tt orbital. Next highest in energy is 104 , calculated to be 54 kcal/mol above 103 . Recall that the LUMO, the a orbital, of the allyl fragment has a node at the central carbon, making this a poor ligand for metal backbonding. The calculated lower energy conformation is the upright 104a , with the plane of the allyl group perpindicular to the plane of the bis (phosphine) platinum. The energy difference between 104a and 104i is small, 4 kcal/mol, H 3 P N / H 3 H 3 P »/ H 3 PtPtIN-PLANE PERPINDICULAR 104i 104a The interaction diagram for 104a is shown in Figure 19 . There is no change in the energy of the allyl a_ orbital in the complex, as expected since this orbital has a node at the carbon bonded to platinum. The only stabilizing interaction is between the la, and 3a, orbitals of bis (phosphine) platinum and a, of the allyl fragment.

PAGE 64

eV 56 <=». O 3a -7--8.. -9 • -10 -11 o @ -12 • =>g<3 -13 • a i O (o Figure 19. Interaction diagram for 104p,

PAGE 65

57 Highest in energy and last to be considered is the cyclopropylidene complex 105 . This is a Pt(0) carbene, of which there are no examples in the literature. The perpindicular conformation 105p is calculated to be lower in energy than the bisecting conformation. The energy difference for the two conformers is high, 18 kcal/mol. PH >= < PH 3 >-»*
PAGE 66

58 O 3a -? -8.. -9 • -10 • -11 • -12 .. >8o -13 •CI 4 <><] 9. ^ *3 o Figure 20. Interaction diagram for 105p,

PAGE 67

59 To summarize the results for the bis (phosphine) platinum complexes, the allene complex is lowest in energy. The cyclopropylidene and allyl complexes are of similar energy, 55 kcal/mol above the allene complex. From the interaction diagrams this is attributed to the high energy of the virtual 3a x orbital of bis (phosphine) platinum. As in the Fp series, the bis (phosphine) platinum complexes of the six membered ring systems 107 , 108 , and 109 were studied The EHMO calculated relative energies are shown in Figure 21. Interestingly the allene complex 107 is found to be lowest in energy. This is in contrast with the Fp system in which the allyl complex was lowest in energy. An attempt to make the platinum allene complex 107 has failed, but this is most likely due to the extreme conditions necessary to generate the allene. To complete the calculations with bis (phosphine) platinum, the cyclic conjugated systems 110 , 111 , and 112 were studied. As shown in Figure 22, the EHMO calculations place 110 and 111 at similar energies and 112 3 6 kcal/mol higher in energy. This is consistent with our results for the acyclic systems 103 , 104 and 105 . The allyl fragment was stabilized 25 kcal/mol relative to the allene fragment on complexing to platinum. Now, the cycloheptatrienylidene fragment has been stabilized 30 kcal/mol relative to the cycloheptatetraene fragment on complexing to platinum.

PAGE 68

60 Kcal 70 86 — 6 39 85 -20 84 P.t(PH 3 ) 2 109 Pt(PH 3 ) 2 108 Pt(PH 3 ) 2 107 Figure 21. Relative enerqies of platinum complexes 107-109 as calculated by EHMO.

PAGE 69

61 Kcal «• 33 25 27 «• 9 i+ 22 36 (PH 3 ) 2 Pt 112 (PH 3 ) 2 Pt-^^ 3 -o in (PH 3 ) £ Pt = 110 Figure 2_2. Relative energies of platinum complexes 110-112

PAGE 70

62 While at this level of theory an energy difference of 3 kcal/mol is not significant, the comparison of this result with that for the Fp complexes clearly shows the shift favoring the cycloheptatetraene complex with bis(phosphine) platinum. This is consistent with the experimental results. In the next chapter the first synthesis of a transition metal cycloheptatetraene complex is reported.

PAGE 71

CHAPTER III EXPERIMENTAL RESULTS The EHMO calculations placed the cycloheptatetraene complex 110 at approximately the same energy as the cycloheptatrienylidene complex 111 . Previous calculations had shown that annelation of a benzene ring at the (4,5) position as in 113 favored the allene form over the carbene form relative to the non-benzannelateri valence isomers . 53 113 114 Similarly, annelation of benzene rings at the (4,5) and (6,7) positions had an even greater effect. The bis (triphenylphosphine) platinum complex of 115 was therefore the first cycloheptatetraene complex synthesized, 115 116 63

PAGE 72

64 Visser and Ramakers have reported in a communication that the generation of 1, 2-cyclophetadiene in the presence of (ethylene) bis (triphenylphosphine) platinum, 5£, gave the 1, 2-cycloheptadiene complex 55 but they reported no experimental details. + Pt(PPh 3 ) 2 (C 2 H /4> ) 54 51 Pt— PPh, 55 Bennet and Yoshida have reported that generation of cyclohexyne in the presence of tris (triphenylphosphine) platinum, 118 , gave the cyclohexyne complex, 119 , and reported 54 a detailed experimental procedure, Tris (triphenylphosphine) 117 .Br 00 'Br Pt(PPhJ 3 + Na(Hg) 118 Ph 3 P N Ph 3 P / Pt 119

PAGE 73

65 platinum was chosen as the trapping reagent because of the greater amount of experimental detail available for its use as a trapping agent and because it was available in higher yield than 54 from the common precursor 120, ^ Pt(PPh )„ EtOH^-^ 3 3 K 2 PtCl 4 + PPh„ Et0H . Pt(PPh 3 )^" 120 15 Pt(0 2 )(PPh 3 ) 2 NaBH^ • C 2 K k Pt(C 2 H^)(PPh 3 ) 2 54 Dibenzocycloheptatetraene, 115 , had never before been generated. Typically, dehydrohalocrenation is a high-yield means of generating strained allenes and bromoalkene 124 was chosen as the precursor to the dibenzocycloheptatetraene for this reason. The synthesis of the bromoalkene precursor followed the method of Allison as is outlined below. Addition of dibromocarbene to phenarthrene gives norcaradiene derivative 122 . Thermolysis of 122 at 160°C for 30 minutes leads to disrotatory cyclopropyl ring opening and formation

PAGE 74

66 Br. Br 124 123 of the dibromocycloheptatriene derivative 123 . Reduction of 123 to 124 is the lowest yield step in the synthesis. Lithiumtriethylborohydride, tri (n-butyl) tinhydride, sodiumborohydride, and lithium aluminum hydride were tried, but only lithium aluminum hydride gave 124 which could be readily purified, although low in yield. Since the dibenzocycloheptatetraene, 115 , had never before been generated, the reaction of 124 with potassium tert-butoxide in the absence of a trapping agent was studied 124 + t-BuOK 125

PAGE 75

67 This reaction crave the allene dimer 125 , as the only isolated product. Only one regioisomer is formed and its regiochemistry cannot be assigned from the spectra. However, based on the regiochemistry of the dimerization of other cyclic allenes it is assigned as above. 2 The [1,2-n -4, 5-6,7-dibenzocycloheptatetraene] bis (triphenylphosphine) platinum complex, 127 , was synthesized in 67% yield by the slow addition at 0°C of bromoalkene, 124, to a THF solution of 126 and 118 . The assignment of the Ph 3 P Pt — PPh, + t-BuOK + Pt(PPh 3 ) 3 126 118 124 127 structure rests on H, C, and 31 p NMR data (cf. Figures 23 to 27_) and the elemental analysis. As will be shown, comparison of the NMR shifts of the allene carbons and the phosphorous atoms of 127 with the analogous atoms in the known (1, 2-cyclononadiene) bis (triphenylphosphine) platinum complex, 5_7, provides strong evidence for its structure. Due to the symmetry of the carbene complex, 128,

PAGE 76

68 H.^ is equivalent to H 3 , C. is equivalent to C 3 , and P. is equivalent to P~, while for the allene complex 127 all these atoms are non-equivalent. The observed non-equivalence of all these atoms shows that the product cannot be the carbene complex 128 , and is consistent with the allene structure. — PPh, 128 127 The H-NMR data for the dibenzocycloheptatetraene complex 127 and 1, 2-cyclononadiene complex 57_ are summarized in Table 2 along with the reported data for 1 , 3-diphenylallene complex 14_. The non-equivalence of the hydrogen atoms is consistent with the assignment as an allene iT-complexed to a metal. Notice that for the allene bis (triphenylphosphine) platinum complexes many of the coupling constants to hydrogen

PAGE 77

69 Pt— PPh, a.o 7.3 7.0 B s e.o 5.5 5.0 4 5 1-0 3.5 3.0 Figure 23_. X H N MR spectrum (300MHz) of 127.

PAGE 78

70 Pt(PPh 3 ) 2 127 Figure 24. Expansion of 1 H 5.70 NMR spectrum of 127 showing only H

PAGE 79

71 Table 2. H NMR data for allene complexes 127, 57, and 14. a ' b Pt(PPh,) 2 Pt(PPh 3 ) 2 @^> CHPh C .11 CHPh Pt(PPh 3 ) 2 127 5 7 14 PtH PH PH HH 3.48 70.8 (2.8) 2.4 66 3.75 65 5.5 9.5 3.0 PtH PH HH 5.71 54.6 8.4 2.8 2.8 4.95 74 11 5.82 12.5 (3.0) .All spectra were recorded in CDC1,. Chemical shifts are in ppm and coupling constants are in Hz.

PAGE 80

72 Pt(PPh 3 ) 2 PhH f ^^S^5 ^P " t(PPh 3 ) 2 44-Pt(PPh 3 ) 2 k^ 127 57 C II c / \ Ph H 14 cannot be determined. In fact the H, -platinum coupling constant was not reported for 1, 3-diphenylallene complex 14 . Otsuka et al . state the resonance was broad and the 1 7 platinum coupling could not be seen. " For all the allene complexes studied the hydrogen resonances were broad and of low intensity, making interpretation of the H-NMR spectra difficult. Before this work was started no P MMR data had been reported for allene complexes of bis (triphenylphosphine) platinum. But, the chemical shifts and coupling constants of the phosphorous nuclei for several olefin complexes had been reported and recently White and Stang reported p NMR shifts for butatriene complexes 129 and 130. 57 These are included in Table 3_ for comparison with the phosphorous spectra of the allene complexes. The non-equivalence of the

PAGE 81

73 Ph Pt(PPh ) Pt(PPh-,) 9 rn r-c^rrn 3 ; 2 Ph Ph 3 2 Ph Ph Ph 129 130 phosphorous nuclei in the allene complexes is not consistent with a carbene structure, while the chemical shifts and coupling constants are consistent with the assignment as an olefin Tr-complex. mu 13 The C-NMR spectra of the allene complexes 127_ and 57 are very complicated. With no symmetry and 51 carbons, there is the potential for a spectroscopist ' s nightmare, especiallv if one adds long range platinum and phosphorous coupling which can lead to even more signals. Luckily, the chemical shifts of the three carbons of the allene moiety are at unusual regions of the spectrum and do not overlap with any other signals. While the 13 C-NMR spectra of several bis(triphenylphosphine) platinum complexes have been reported, one author reported no coupling constants 57 and the other reported platinum, but no phosphorous coupling. 59 Thus, the first complete spectral data for the 13 C-NMR of

PAGE 82

74 31 Table 3. P NMR data of some (olefin) bis (triphenylphosphine) platinum complexes. 3 Olefin Complex P^ _^^ _p^_ \j^ fj^ Ref, CH 2 H 54 32 3660 58 CH 2 i 3 c 111 131 19.7 3590 58 C I CH, CPh 2 II II 129 29 3416 26 3664 37 57 11 CPh 2 CPh ? 1 1 1 130 29 2996 25.5 3550 23 57 A

PAGE 83

75 Table 3. continued. 57 33.8 3307 29.9 3216 48.8 127 29 3208 27.5 3145 28 Chemical shifts are in ppm and coupling constants are in Hz.

PAGE 84

76 JLJ 1 — i — p 40 "I i f i 1 — r — r 35 ^~ | i I i — r — | — i r 30 25 Ph P Pt 127 20 UL_ Figure 25. 31 p NMR spe ctrum of 127 I i i i — t — | — i — i — 15 io FFr

PAGE 85

77 bis (triphenylphosphine) platinum olefin complexes is reported in Table 4_. Perhaps the most unusual chemical shift in the 13 C-NMR spectrum is that of the central allene carbon, which is found at very low field, at about 160 ppm it is consistent with the formulation as an allene complex, since uncomplexed allenes show a resonance for the central carbon at 200 ppm 60 and in all cases there is a shift to higher field on complexing olefins to a metal. The resonance of the central carbon in the iron allene complexes 19 and 59_ is 150 an ^ 14 8 ppm, respectively. 150ppm >» Fp + ex FP + l46ppm 51 19 In attempting to reproduce Visser and Ramakers' 3 °work it was found that the insertion product, 133 , formed on mixing 1-bromocycloheptene and 118 . The insertion of platinum into carbon halogen bonds has been reported previously and is useful for forming carbon-platinum sigma bonds. 61 ' 62

PAGE 86

78 13 Table 4. C NMR data for allene complexes 57 and 127. a ' b PC PC 57 33.8 6.33 38.6 Pt(PPh 3 ) 2 Pt(PPh 3 ) 2 127 29.5 5.43 52.2 PC PC PtC PC PC PtC 161.6 10.1 52.3 490.7 110.8 9.5 164.4 9.6 62.9 109.9 5.3 10.0 b All spectra were recorded in CDC1.,. Chemical shifts are in ppm and coupling constants are in Hz.

PAGE 87

79

PAGE 88

80 ffW^fl fi fkmWW'i ' i ' ' ' ' i ' ' ' ] i ' ' ' ' i ' ' i ' i 166 165 164 163 162 Pt(PPh 3 ) 2 127 vMfrtofafffi ^W^v>4 112 in IU 109 108 #fVffl 0l 1 I ' ' ' ' | ' i . r31 30 1 ' ' ' ' | -i 29 28 Figure 27. Expansions of 13 C NMR spectrum of 127 showing C -C

PAGE 89

81 Br Br 1 1 Ph„P-Pt-PPh~ 3 1 3 J + Pt(PPh 3 ) 3 118 -o 132 133 Since the cis insertion product, 134 , would show an ABX pattern in its phosphorous NMR spectrum which might be similar to that of the allene complex, the thermal PPh~ I 5 Ph 3 P-Pt-Br 134 Br I Ph 3 P-Pt— PPh„ 124 135

PAGE 90

82 insertion reaction of 124 with 15 studied. The product of this reaction is the trans adduct, 135, as shown by the phosphorous NMR in which there are large platinum phosphorous coupling constants of 3189Hz and 3 222Hz, which is a simple 1:4:1 triplet, with equivalent phosphorous atoms. Hydride abstraction from sigma complexes has been useful in the synthesis of transition metal cycloheptatrienylidene complexes. 37 " 39 The reaction of insertion product 135 with trityl cation was studied in methylene chloride at -78 C. This gave an insoluble material assigned Br I Ph^B-Pt— PPh g 135 + Ph 3 C + y Pt'(PPh 3 ) 2 136 structure 136. A similar observation was made when 137 was treated with silver ion. Since no triphenylmethane was observed and therefore no hydride abstraction had occurred, no further work with this material was attempted

PAGE 91

83 137 + Ag Pt(PPh 3 ) 2 Br + Br Pt + (PPh 3 ) 2 138 With the successful synthesis of the dibenzocycloheptatetraene complex, 127, the synthesis of the benzocycloheptatetraene complex was undertaken. This allene has been reported previously; 63 so it was only necessary to generate the intermediate in the presence of the tris (triphenylphosphine) platinum trap. The bromoalkene precursor , 142, was prepared by the method of Waali et al., as outlined below. 63 Slow addition of 139 Br \^Br CBr, -zr -HBr 142 141

PAGE 92

84 this to a THF solution of 118 and 126 gave the allene complex 143. The H, C, and 31 p -NMR spectra (cf Figure 28.) show that is the allene 14J_ and not the carbene complex. The NMR data for this compound are reported with that for the cycloheptatetraene complex, vide infra . + t-BuOK + Pt(PPhJ 3 Ph 3 P N Pt-PPh, 141 126 118 143 For allene complex _143_ two regioisomers are possible. From steric considerations one expects the platinum to complex away from the benzannelation, as in 143_. This is Pt(PPh 3 ) 2 Pt(PPh 3 ) 2 144 143

PAGE 93

85 K, Pt(PPh 3 ) 2 143 yptittfhW'b h i 3.1 ' 3.0 2 3 2.8 2.7 '^''/WW/^A H, 5.0 4.9 Figure 28, Expansions of X h and H . NMR spectrum of 143 showing H

PAGE 94

86 consistent with the chemical shifts and coupling constants of the allene hydrogen atoms. The dibenzocycloheptatetraene complex 127 and parent cycloheptatetraene complex were used as model systems. In complex 143 the coordinated hydrogen should he similar to the analagous hydrogen in the cycloheptatetraene complex, while in complex 144 the coordinated hydrogen should be as in the dibenzocycloheptatetraene complex. Similar logic can be used for the uncoordinated allene hydrogen, and both are consistent with the assignment of structure -illAgain, there was the possibility of the formation of the cis-insertion product L4_5. The reaction of bromoalkene 142 with 15 in benzene was studied and the trans -insert ion PPh, I -> Ph^P-Pt-Br 145 142 + Pt(PPh 3 )^ 15 Br I Ph-P-Pt-PPh146

PAGE 95

87 product was formed as shown by the triplet in the phosphorous spectrum. It was postulated that the insertion product might form during the reaction of bromoalkene 142 with base and 118 an 3 that the insertion product reacted with base to give the allene. However, there was no reaction between the insertion product and base under the reaction conditions. Pt(PPh 3 ) 2 Br It remained to synthesize the parent cycloheptatetraene complex. Bromocycloheptatriene was prepared as a mixture I \ 147 _SeH e-»» 148 1. BrCCBr II H 00 2. LAH I I Br 149

PAGE 96

of three isomers according to the procedure of Fohlish 64 and Haug. Slow addition of one equivalent of bromocycloheptatriene to a solution of potassium tert-butoxide and tris (triphenylphosphine) platinum gave the cylcoheptatetraene complex 150 , as characterized by 1 H, 13 C, and 31 P NMR (cf Figure 29,30) . The spectroscopic data for the allene complex are summarized along with those for the benzocycloheptatetraene complex and the dibenzocycloheptatetraene coirole in Tables 5-7 . Mi .150 Cycloheptatetraene complex 110 was calculated by EHMO to be of comparable energy to cycloheptatrienylidene complex 111. In the carbene complex 151 H a and H are equivalent and should have the same NMR shift. To probe the possibility of interconversion of the cycloheptatetraene complex with its valence isomer, variable temperature NMR studies were performed. On heating to 80°C there is no change in the H-NMR spectrum, this is equivalent to a barrier separating the two valence isomers of at least 17 kcal/mol. 65

PAGE 97

89 13 Table 5. C NMR data for 150, 143, and 127. a,b PC PC 150 27.7 5.1 44.1 143 26.41 4.9 43.9 5.4 52.2 PC J PC PC PC 151.0 10.4 61.0 114.5 9.3 5.6 156.1 9.7 62.3 113.0 164 .4 9 .6 62 .9 109, ,9 10. 5.3 jjAll spectra were recorded in CDC1,. Chemical shifts are in ppm and coupling constants are in Hz.

PAGE 98

90 31 Table 6_. P NMR data for allene complexes 150 , 143, and 127. a,b PP PtP PtP 150 143 31.5 30.5 33.3 30.1 3267 3297 27.3 26.2 3170 3131 127 29.0 28.0 3208 27.5 3145 .All spectra were recorded in CDC1,. Chemical shifts are in ppm and coupling constants are in Hz.

PAGE 99

91 Table 7. X H NMR data for a and H, of 150, 143, and 127 a,b PtH PH PtH PH 150 143 2.94 3.02 70.4 72 11 13.2 4.63 5.16 70.8 66 9 9 7 4.5 HH 127 3.48 70.8 5.71 54.6 8.35 2.8 2.8 b All spectra were recorded in CDC1, at 300MHz. Chemical shifts are reported in ppm and coupling constants are in Hz.

PAGE 100

92 Figure 29. X H NMR spectrum (100MHz) of 1 50 .

PAGE 101

93 Pt(PPh 3 ) 2 Figure 30. Expansion Qf l fl NMR ^^ ^ ^ ^

PAGE 102

94 Ph-P 3 \ Pt — PPh_ K XT 150 Ph P PPh 3 P1T 3 151 RTjln Tc/fiv + 22.96) Eq (2) T c = coalescence temperature Sv = peak separation in Hz The insertion reaction of bromocycloheptatriene with tetrakis (triphenylphosphine) platinum was studied. The trans-isomer was formed and when excess bromocycloheptatriene was used, the 1-isomer was the major product. This material was inert to potassium tert-butoxide in THF at room temperature. Br (I j + "(I 5 a Pt(PPh 3 ) 2 Br 149 152

PAGE 103

95 Pt(PPh 3 ) 2 Br 152 Figure 31. lg NMR spectrum (10QMHz) Qf

PAGE 104

96 Initially there was difficulty in isolating the cyoloheptatetraene complex and the reaction was attempted using two equivalents of bromocycloheptatriene and excess base. The product of the reaction was the novel bis-adduct 153 which has been characterized by 1 H, ^C, -^P-NMR (cf. Figures 3_2 to 36), comparison of thesedata with that for model compounds, and an x-ray crystal structure determination. Br u 2 eq. + t-BuOK + Pt(PPh,)_126 118 Ph,P^ ^PPh Q 149 153 Integration of the Tt-NMR, shownin Figure 32, shows there to be a 1:1 ratio of triphenylphosphine to cyoloheptatetraene. Notice there is a high field proton, 1.68 ppm which is coupled to platinum.

PAGE 105

97 153 _A_ Figure 32. H NMR spectrum (3 00MHz) of 153.

PAGE 106

98 Ph 3 P x /P Ph 3 5-8 5.7 5.6 5.5 5 -y 5.3 pph ^4^0•m {/ \mhmM^ 2.0 "" 1 1.8 1.6 1.4 -| i 1 1 l 2 1.0 PPM hf?h r fiifd r !^r° n ° f ^ mR *^™ °f i" showinc

PAGE 107

99 Consideration of the 13 C-NMR spectrum, shown in Figure 34, and comparison with the 13 C NMR spectra of known platinacyclopentanes shows the compound to be a platinacyclopentane with the regiochemistry shown. Most important for the assignment is a resonance at 157.6 ppm with a small coupling constant of 8.5 Hz and a large coupling constant of 105.16 Hz, consistent with two bond cisand transphosphorous coupling. One would expect to see a platinum coupling constant of about 900 Hz, unfortunately signal to noise limitations prevented observation of the peaks. However, the platinum coupling was observed in a derivative, vide infra . The phosphorous coupling constants show the carbon to be bonded to platinum and the chemical shift and multiplicity are consistent with a vinylic assignment. This eliminates 154 as a possible structure. At 57.8 ppm is a peak for carbon bonded to Ph_P N ^PPh ? 154

PAGE 108

100 Table 8^ NMR data for some platinum metallacyclopentanes . COD Me,P PMe3 \ / j Et 3 P x/ PEt 3 Pt Ph,P PPh 7 3 \/ 3 155 156 157 155 C a 33.5 48.1 36.0 157.6 J PtC 725 584 600 .. 2 J J PC na 7.3 9 8.5 J PC na 93 9 2 105.1 C b 162 148.9 157.5 57.8 2 J J PtC 60 37 119 3 J J PC na 4 li c c 101.2 102.7 132.7 na 3 J J PtC 82 55 86 na J PC na 7.3 3 na Ref . 66 66 67 a r-T,„™i ,-„i ,-U -J X4„„. Chemical shifts are reported in ppm and coupling constants are in Hz.

PAGE 109

101 1H a, x e 3 U *> o o a s z

PAGE 110

102 hydrogen and showing a platinum coupling constant of 119 Hz, both these facts are consistent with structure 153 . Finally, the phosphorous NMR spectrum, shown in Figure 3J5, yields a simple 1:4:1 triplet at 25.7 ppm with a platinum coupling constant of 1960.5 Hz. All the phosphorous nuclei must be equivalent. Table 9 shows some typical phosphorous shifts and coupling constants for known cis-platinum(II) phosphorous compounds. The similarity of the coupling constant for 153 with those of the compounds in Table 9_ is consistent with the assignment of a cisplatinum(II) phosphorous compound. For further confirmation of the structure a derivative was synthesized. Bennet and Yoshida had reported that heating 119 with DIPHOS 161 led to exchange of the phosphorous 54 ligands and formation of 162 . r"V-Pt ( PPh 3 ) 2 ___ r^\ \^ Ph 2 p pph 2 \x^ 119 161 162 DIPHOS

PAGE 111

103 Tatle 231 P NMR data for some cis-phosphorous platinum compounds. Coumpound Shift(ppm) 1 J ptp (Hz) Pt(PEt 3 ) 2 Me 2 -9.7 185 6 158 Pt(PKt 5 ) 2 Ph 2 -3.5 1705 159 Pt(PMe 2 Ph) 2 Me 2 u. 2 1819 160 Ph 3 P N /PP^ _£t^ 25-7 1961 8 121

PAGE 112

104 Ph,P. PPh, 153 25.7 PPM J Pt,P= 1960.5Hz 31 Figure 35. p NMR spectrum (121MHz) of 153.

PAGE 113

105 Heating 153 with excess DIPHOS also led to exchange. This compound was characterized by Y 13 C, 31 P-NMR and elemental analysis and is similar to the parent in all respects. The one bond platinum carbon coupling constant was observed for 163, 904.7Hz. PhjP^ PPh 3 ., + DIPHOS 153 161 i 6 3 Finally, the x-ray crystal structure proved the assignment for 153. The ortep drawing of 153 is in Figure 11. There was a large standard deviation in the determined C-C bond distances because the crystal was slightly deformed, the data was rapidly collected and the platinum and phosphorous atoms dominate the scattering pattern. Once the structure was assigned, the generality of the bis-adduct formation was studied. An excess of benzocycloheptatetraene was reacted with tris (triphenylphosphine) platinum and the bis-adduct 164 was isolated.

PAGE 114

106 ItSSff: $2C$ **t J7, ."J® ,* ».,Figure R. Stereo drawing of 153 produced by Ortep computer program.

PAGE 115

107 // \ + t-BuOK + Pt(PPh 3 ) 3 (Cj) M£ lis 142 Ph 3 P N /P Ph 3 1£ 164 Table 10 summarizes the NMR data for 153 , 163 , and 164 . These are the only compounds which have been observed to form the novel bis-adduct. For example, when an excess of the dibenzocycloheptatetraene 115 was generated in the presence of 116 only the allene complex 127 was formed, and in 70% yield. Similarly, no bis-adduct is formed when an excess of the saturated cyclic allene 5_1 is generated in the presence of the tris (triphenylphosphine) platinum. Finally, no acyclic allenes have been reported to form bis-adducts with tris (triphenylphosphine)platinum.

PAGE 116

108 Table 10. 31 P and 13 C NMR data for 153, 163 , and 164. a Ph,P PPh, 3 X/ 3 31 155 P 25.7 1 J ptp 1960.5 C y 57.8 J PtC 11£ C 1 157.6 PtC J pc 8.5 J pc 105.1 Ph.P PPh, 44.22 1865.8 58.1 114 161.8 904.7 7.3 108.4 25.9 1950.8 57.3 98.6 168.6 9.8 103.8 All chemical shifts are reported in ppm and all coupling constants are in Hz.

PAGE 117

109 CH, II 2 C II CH., 51 115 However, an excess of allene does react with "ligand free platinum", 165, to give the metallacyclopentane 155. 66 CH 9 II 2 C II CH, Pt(COD). 165 155 COD = 1,5-cyclooctadiene Unfortunately, Barker et al. did not study the reaction of 165 with any other allenes. While this is the only example of allene dimerization by platinum, numerous other metals are known to react with two equivalents of allene to form bis-adducts. 69 ' 70

PAGE 118

110 For example , Tf Rh(AcAc)(C 2 H 4 ) 3 + C 3 H 4 » -£-». L 2 (AcAc)Rh^^ 166 167 r^ Ni(C0D) 2 + 1 + R 2 P PR 2 Oct 168 169 170 A mechanism which is consistent with all the above is outlined below. There is initial formation of the allene ir-complex, and if there is only one equivalent of allene this is the major product. If there is excess allene Cil H H II 2 C ML n * Vl M -#„ H H 1 171 ,, 172 «cc / C 3 H 4 L n-1 M N C 3 H 4 174 173

PAGE 119

Ill present, the mono allene adduct 172 , can react with a second equivalent of allene to give a metal complex which has two allene ligands TT-bonded, 173_, This complex can either go on to the metallacyclopentane or lose one of the allenes. T he transition state for metallacyclopentane formation is presumed to be very crowded. Thus, increasing substitution on the allene inhibits the metallacycle formation by steric blocking, as is observed in experiments. Consistent with the proposed mechanism, reaction of cycloheptatetraene with 150 gives the metallocyclopentane 153. ,Pt(PPh 3 ) 2 Ph 3\ / PPh 3 t-BuOK 149 150 t 153 Interestingly, an attempt to make the unsymmetrical metallocyclopentane, 175 , failed, reaction of cycloheptatetraene with 57 gave only 153 . Apparently, loss of cyclononadiene to give 150 occurs more rapidly than metallocyclopentane formation.

PAGE 120

112 t-BuOK 149 Pt(PPh 3 ), 57 Ph 3 P x /P Ph 3 153 Ph 3 P N /PP h 3 "it 175 not formed Metallacyclopentane formation from two allenes is a special case of metallacyclopentane formation from bis-olefin complexes. McKinney et al . have studied the formation of metallacyclopentanes from bis-olefin complexes and have also studied the affect of substituents on the regiochemistry of the reaction. At the level of one electron theory the dominant interaction in this rearrangement is backbonding from the metal to the LUMO of the olefins. If the olefin is unsymmetrical the carbon-carbon bond formation should occur at the position which has the largest LUMO coefficient, as illustrated below.

PAGE 121

113 X X K 176 For 1,2-propadiene extended Huckel calculations show the largest coefficient in the LUMO to be at the central carbon, thus C-C bond formation should occur at the central carbon, leading to the observed regiochemistry . For cycloheptatetraene the largest coefficient is again at the central carbon, leading to the prediction of the same regiochemistry C •83 §>C0 .78 OC3S H H as is observed for 1,2-propadiene, which is not what is observed.

PAGE 122

114 not formed 60 177 Thus, the simple model proposed by McKinney et al. is apparently not applicable to the allene reactions. Another explanation, which is consistent with the observed results is that there is much radical character in the transition state. For allene, a radical on the sp carbon 2 xs more stable than on the terminal sp carbon and carbon bond formation occurs at the central carbon. For cycloheptatetraene the tropyl radical is more stable, leading to a reverse in the observed regiochemistry .

PAGE 123

115 & M(C 3 H 4 ) 2 1/ 173 M 179 M 180 M(C 7 H 6 ) 2 181 182

PAGE 124

CHAPTER IV SUMMARY It has been reported that methoxy abstraction from 28 gives the allene complex 9, but that hydride abstraction from 64 gives the carbene complex 6J5. One objective of this research was a qualitative explanation for this result. EHMO calculations revealed that the Fp LUMO-ligand HOMO interaction is much greater with a sigma orbital than with a pi orbital. Thus, the relative energy of allene, 76, compared with planar allene, 77, decreases on complexation with Fp. But, the difference in energy of 76 and 77 is so great there is not any change in energy ordering. The energy difference between cycloheptatetraene, 93, and cycloheptatrienylidene, 94, is much smaller and on complexation with Fp there is a reversal in order. Because no cycloheptatetraene transition metal complexes had been synthesized, a second objective of this research was the synthesis of a cycloheptatetraene transition metal complex. To this end, EHMO calculations were carried out which showed a significant difference in the bonding of Fp and Pt(PPh 3 > 2 complexes, and which suggested the synthesis of a cycloheptatetraene bis (phosphine) platinum complex. 116

PAGE 125

117 The generation of cycloheptatetraene in the presence of tris (triphenylphosphine) platinum led to rapid exchange and formation of a cycloheptatetraene transition metal complex. Similar reactions gave the benzocycloheptatetraene and dibenzocycloheptatetraene-bis (triphenylphosphine) platinum complexes. These compounds were characterized by H, C, and P NMR spectroscopy and by elemental analysis. These are the first examples of transition metal cycloheptatetraene complexes and demonstrate the significant difference in the bonding of Fp and Pt(PPh,)_. A novel bis-adduct has been isolated from the reaction of excess cycloheptatetraene with tris (triphenylphosphine) platinum. While several transition metal complexes react with allene to form metallacyclopentanes, this is the first example of metallacycle formation with tris (triphenylphosphine) platinum. This bis-adduct, 153 , 1 13 3 1 has been characterized by 33, C, and P NMR spectroscopy and by an x-ray structure determination. The generality of this reaction was studied and benzocycloheptatetraene was the only other allene to show this novel reactivity.

PAGE 126

CHAPTER V EXPERIMENTAL Benzene, diethyl ether, hexane and tetrahydrofuran (THF) were distilled from sodium benzophenone ketyl. Methylene chloride was distilled from P 2 5 under nitrogen. Silica gel was M.C.B. 230-400 mesh. Alumina was Brockman 80-200 mesh, activity I, which was deactivated to activity II by the addition of 3% (w/w) water. 1 H and 13 C NMR were taken on a JEOL FX-100 (100MHz) or a Nicolet NC-300 (300MHz) . 31 p NMR were recorded on a Nicolet NC-300 (121.4MHz) with trimethylphosphite used as an external reference. Elemental analysis were performed by University of Florida Microanalysis Service or by Atlantic Microlab, Inc. Melting points (uncorrected) were obtained using a Thomas Hoover apparatus. All solutions containing transition metals were manipulated under nitrogen either using a glove box or schlenk techniques. 118

PAGE 127

119 Preparation of 7 , 7-dibromo-2, 3-4, 5-dibenzobicyclo [4 .1 .0] heptane ( 122 ) . This compound was prepared by the method of Allison. Recrystallization from chloroform/hexane yielded white needles of 122: mp 131.0°-132.0°C; IR (nujol) 1580m, 780s cm" 1 ; X H NMR (100MHz, CDClj) 3.46(s,2H), 7.34(m,6H), 7.94(m,2H); C MR (25MHz, CDC1 3 > 30.8(C7), 37.4(d), 123.0, 128.1, 128.2, 129.5, 130.9, 131.2; Anal. Calcd. for C 15 H 10 Br 2 : C,51.46; H,2.86. Found: C, 51.37; H,2.68. Preparation of 4,5-6, 7-10 , 11-12 , 13-tetrabenzotricyclo [7.5.0.0 2,8 ] tetradeca-2, 3-4, 5-6, 7-10, 11-12, 13-14, 11-hexaene (125) . Potassium tert-butoxide (lOOmg, .9mmol) was weighed into a schlenk tube and 5 mL THF added. The bromoalkene , 124 , (300mg,l . lmmol) was dissolved in 1 mL of THF and added to the solution in the schlenk tube. This solution was stirred overnight. The reaction was quenched by filtering through alumina (neutral, grade II) and the solvent removed in vacuo. The resulting solid was purified by column chromatography. Elution on a 1" by 6" low pressure, silica gel column (230-400 mesh) with 7% methylene chloride/pentane and collection of the first band yielded 125 (92mg, 54%): mp. 153-155°C; """H NMR (100MHz, CDC1 3 ) 54.42 (S,1H), 6.64(s,lH), 7.2-7.6(m, aromatic); 13 CNMR (25MHz, CDC1 3 )«47.2, 119.6, 124.5, 126.6, 127.8, 129.5, 131.8, 132.0, 135.8, 139.3, 139.9, 143.8, 147.1; mass spectrum, m/e 380 (parent ion); calcd for C 3Q H 20 : 380.1557, found 380.1565.

PAGE 128

120 2 Preparation of (1,2-n -4, 5-6 ,7-dibenzocycloheptatetraene) bis (triphenylphosphine) platinum ( 127 ) . Potassium tertbutoxide (.35g, .31 mmol) and 118 (.75g, .76mmol) were weighed into an oven dried schlenk tube along with a magnetic stirrer. These were dissolved in 2 5 mL of THF, the schlenk tube was sealed with a septum, connected to a nitrogen line and cooled to 0°c. Bromoalkene 124 (.5g, 1.43mmol) was weighed into a vial and transferred to a syringe with 1 mL of THF. This solution was added dropwise to the schlenk solution over fifteen minutes. After stirring overnight the solution was filtered through celite and the celite washed with THF. The solvent was removed in vacuo and the resulting oil triturated with 2 mL of ethanol . This was filtered and the precipitate washed with pentane. Drying in vacuo yielded 127 (.474g, 67%): mp. 144-150d; IR (KBr) 3050w, 2900w, 16602, 1585m, 1570w, 1475s, 1430s, 1180m, 1090s, 995m, 810m, 740s, 690s, 500s, 4 20m cm" 1 ; 1 E NMR (300MHz, CD 2 C1.) S3.48(t, 2 J = 70.8Hz), 5.71(tdt, 3 J ptH =54.6Hz, 4 Jp H =8 .3 5Hz , 4 4 J P(cis)H = j hh =2 8Hz) ' 6 -4Kd,J=7.6Hz, 1H) , 6.67(td, J=7.3Hz), 1.1Hz), 6.7-7.5(Ph 3 P) ; 13 CNMR (75MHz, CDC1 ) 629.5(dd, 2j P(cis)C =5 43Hz ' 2j P(trans)C =52 2Hz ' C5 > ' "9.9(da, Jp C =5.3Hz, 3 J pc =10Hz, C7), 124.11, 124.37, 125.96, 126.08, 130.68, 131.4, 134.66, 134.8, 140.03, 140 . 71 (d, J=9 . 1Hz) , 143.97(d,J=10.3Hz), 146.17, 164,36 (dd, 2 J„ , . .„-9.6HZ, P (cis) C

PAGE 129

121 J P(trans)C =62 9Hz ' C2) ' 127 • 8 < vt > 3j PC = 9 8Hz < ph 3 P ' meta), 129.24(s,Ph 3 P, para), 129.22(s, Ph P, para), 134. l(m, Ph,P, meta), 135.385, 135.94, 136.10, 136.64(135-136, d, l J =2.5Hz, 31 Ph P, ipso); P NMR (121.5MHz, CDC1,) «28.97(td, 1 J 2 3 PtP 3208Hz, J pp =28Hz), 26.46 (td, 1 J ptp =3145Hz , 2 J pp =28Hz) ; Anal, calcd for CgjH^PjPt^: C,67.33; H,4.75; Found: C, 66.82; H,4.48. 2 Pr eparation of (1,2-n -1, 2-cyclononadiene)bis-(triphenyl phosphine) platinum (57) . This compound was prepared according to the procedure of Visser and Ramakers 30 : mp 149-150°C; IR (KBr) 3060w, 2920w, 1590w, 1480m, 1440s, 1390m, 1180w, 1100s, 745m, 700s, 700s, 510s cm" 1 ; 1 H NMR (300 MHz, CDC1 ) 61.0-2.3 (m,12H), 2.4(td, 2 J ptH =66Hz, 3 J pH =7Hz, HI), 4.95(td, 3 J p1 . H =74Hz, 4 JpH ,^, aJ .,, »., a i TO , j ptH = 13. J pH =10Hz, H3), 7.1-7.5(18, Ph 3 P, 30H) ; 13 C NMR (75MHz, CDC1 ) 622.84, 24.52, 26.20 (C5-C7), 29.80(d, J-ll.OHz), 31.64, 32.78(d, J-12.2H,,, 33.78(dd, % (trans) c =38 . 6Hz , 2 Jp(cis)c = 6. 33Hz, 1), U0.88(d, 3 J pc =9.5Hz, C3 ) , 161.62(tdd, ^=490 . 7Hz , J .f t ,„=52.26Hz, 2 C_, J P (trans) C =52 26Hz ' J P (cis) C =1 ° 05Hz) < 12 7-6(d, 3 J pc =9.5Hz, Ph 3 P, meta), 128.85(s, Ph 3 P, para), 133. 9 (m, Ph 3 P, ortho) , 136.75(t, J ptc =22.9Hz, Ph 3 P, ipso), 137.3 (t, 2 J ptc =23 .0Hz, Ph 3 P, ipso) . Preparat ion of trans(bromo) (n 1 -cyclohepten-l-yl) bis (triphenyl phosp hine) platinum (133) . Tetrabis (triphenylphosphine) platinum (200mg, .16mmol) and bromocycloheptene 73 (.05g, .28mmol) were weighed into a round bottom flask. A magnetic

PAGE 130

122 stirrer and 25 mL benzene were added and the solution refluxed under nitrogen for two hours . The flask was cooled to room temperature and the solvent removed in vacuo . The solid was triturated with ether and filtered yielding 13_3 (116mg, 81%): mp 235-235. 5d; IR (KBr) 3040w, 3020w, 1570m, 1500m, 1480s, 1430s, 1180m, 1090s, 740s, 690s, 520s cm" ; H NMR (100MHz, CDC1 ) 6 . 2-1 . 8 ( 10H) , 5.5 (H, 3 J ptH =74.7Hz, J HH =5.37Hz, Hz), 7.4 (s, PhjP, 18H) , 7.8(s, Ph 3 Pm 12H) ; 13 C NMR (25MHz, CDC1 3 > 626.4, 26.8, 29.6, 32.5, 41.2, 139.9(C1), 127.7(vt, 3 J pc =5.2, Ph 3 P, meta) , 129.9(s, Ph 3 P, para), 131.9(vt, J=27.5Hz, PhjP, ipso), 135. 4(m, Ph 3 P, ortho) ; Anal, calcd. for C 4 ,H Br..P Pt.: C,57.7; H,4.62; Found: C,57.42; H,4.66. P reparation of trans(bromo) (n -3 , 4-5, 6-dibenzocyclo h eptatrien-l-yl)bis (triphenylphosphine) platinum ( 135 ) . Tetrahis (triphenylphosphine) platinum (350mg, .3mmol) was weighed into a round bottom flask equipped with a magnetic stirrer and nitrogen inlet bromoalkene 124 (180mg, .6mmol) in 25 mL toluene was added and the mixture heated overnight. In the course of the reaction there is a color change from bright orange to pale yellow. The solvent was removed in vacuo and the solid triturated with hexane and filtered. The white precipitate was 13 5 (256mg, 86%): mp 281-283°d; IR (KBr) 3060w, 1590m, 1485s, 1440s, 1390s, 1195w, 1105s, 1000m, 840m, 755s, 695s, 520s cm -1 ;

PAGE 131

123 1 H NME (100MHz, CDC1 3 ) S5.50(t, 3 J ptH =68 . 4Hz , 2H , H7) , 6.6(m), 7.2(bs, Ph 3 P) ; 13 C NMR (75MHz, CDC1 3 ) S45.6(C7), 123.4(CH), 124.9(CH), 125.6(CH), 126.5(CH), 127.0(CH), 127.3(CH), 127.7 (Ph 3 P, meta) , 128.4, 129.1, 130.0 (PhjP, para), 132.1 (t, 3 2 J pc =7.2Hz, C2) , 135.1 (vt, J =6Hz, Ph P , ortho) , 136.9, 139.2, 139.6, 139.8, 147.1 (t, 2 J =17.9Hz, CI); 31 PC P NMR (121.5MHz, CDC1 3 ) 622.7 (t, l J ptp =3190Hz); Anal, calcd for C 51 H 41 Br 1 P 2 Pt 1 : C,61.82; H,4.4; Found: C, 62.93, H,4.25. Pr eparation of (1, 2-n 2 -4, 5-benzocycloheptatetraene) bi s (triphenylphosphine) platinum ( 143 ) . Potassium tert-butoxide (57mg, .51mmol) and 118 72 (500mg, .51mmol) were weighed into an oven dried schlenk tube along with a magnetic stirrer. These were dissolved in 15 mL of THF, and the schlenk tube was sealed. The schlenk tube was connected to a nitrogen line and cooled Bromoalkene 142 55 (113mg, .51mmol) was weighed into a vial and transferred to a syringe with 1 mL THF . This solution was added dropwise to the schlenk solution over fifteen minutes. After stirring overnight the solution was filtered through celite and the celite washed with 25mL ether. The solvent was removed in vacuo and the resulting oil triturated with lOmL hexane. This was filtered and the precipitate dried in vacuo yileding a white powder, 142 (360mg, 82%) : mp 140-144d: IR: 3045w, 1480s, 1435s, 1385s, 1090s, 740s, 690s, 515s cm -1 ; X H NMR (300 MHz, CDC1 3 ) S3.02(td, 2 J ptH =72Hz, 3 J pH =13.2Hz, HI),

PAGE 132

124 5.16(tdd, 3 J ptH =66Hz, 4 J pH =9Ha, 4 J pH =4.5Hz, H3 ) , 6.07(d, J=12Hz) , 6.13(d, J=12Hz) , 6.32, 6.7(m), 7.0-7.7(m, Ph P) ; 13 C NMR (25MHz, CDC1,) S26.41(dd, 2 J„, < _ . =43.9Hz, -3 J? \ THT3I1S I C 2 J p(cis)c =4.9Hz, CI), 113.0(rac, C2) , 124.1, 125.8, 127.5 (bs^ Ph 3 P, raeta) , 128.6, 129. 2(s, Ph 3 P, para), 129.9, 130.4, 131.7, 132.1, 134.3 (m, PhjP, ortho) , 136.8 (Ph 3 P, ipso, 2 coupling not observed), 156.1(dd, J„,,. . =62.3Hz, P( trans )C 2 J p(cis)c =9.79Hz) ; 31 PNMR (121.4MHz, CDClj) 626.2(td, 1 ^ tp = 3134Hz, 2 J pp =30.lHz) , 32.2(td, 1 J ptp =3297Hz J pp =30.1Hz); Anal, calcd for C^H^P^t^ C, 65.67; H,4.46; Found: C,65.25; H,4.46. Preparation of trans(bromo) (4-n -1 , 2-benzocycloheptatriene 4-yl) bis (triphenylphosphine) platinum ( 146 ) . Tetrakis (triphenylphosphine) platinum (l.Og, .8mmol) was weighed into a round bottom flask equipped with a magnetic stirrer and a 55 nitrogen inlet. Bromoalkene, 142 , (200mg, .9mmol) in 25 mL benzene was added to this flask and the mixture refluxed 72 hours . After cooling the flask to room temperature the solvent was removed in vacuo and the product triturated with ether. The precipitate was filtered and dried in vacuo to yield 146 (.46g, 61%): mp 218-219d; IR (KBr) 3065, 3020w, 1610w, 1485s, 1440s, 1390s, 1100s, 750m, 700s, 540s cm -1 ; X H NMR (100MHz, CDC1 3 )
PAGE 133

125 124.7, 126.1, 126.2, 126.5, 128. l(d, 3 J pc =4.9Hz, Ph 3 P, meta) , 130.4, 130.5, (s, Ph 3 P, para), 131.3 (td, 2 J ptc =28Hz, 1 J pc =15.9Hz, Ph 3 P, ipso), 134.4, 135. 5(d, 2 J pc =6.1Hz, PhjP, ortho) , 139.1, 142. 8(d, 2 J pc =17.1Hz, C4) ; 31 P NMR (121.5MHz, CDC1 3 ) (522. 0(t, 1 J ptp =3189.4Hz) ; Anal, calcd for H^C^Br^jPt^ C,60.0; H,4.18; Found: C,59.1, H,4.02. Preparation of Tropone ( 148 ) This compound was prepared by a modified procedure of Radlich from cycloheptatriene. Instead of allowing the reaction mixture to stand overnight, it was stirred overnight with a mechanical stirrer in a Morton flask. This improved the yield to 37% (lit 25%) . Preparation of 1-,2-and 3-bromocycloheptatrienes ( 149 ) . This isomeric mixture was prepared from bromotropylium bromide according to the method of Fohlisch and Haug. 75 A slight modification greatly simplified the work-up. Excess LAH was quenched with wet ether followed by addition of a saturated sodium potassium tartrate solution. 2 Preparation of (1,2-n -cycloheptatetraene) bis (triphenylphosphine) platinum ( 150 ) . Potassium tert-butoxide (4 5mg, .401mmol) and tris (triphenylphosphine) platinum (400mg., .41mmol) were weighed into an oven dried schlenk tube and a magnetic stirrer added. After adding 10 mL of THF, the schlehk tube was sealed with a rubber septum, removed from the drybox', and connected to a nitrogen line.

PAGE 134

126 Bromocycloheptatriene (69.4mg, .4iranol) was weighed into a vial and 1 mL of THF added. This solution was drawn into a syringe and added dropwise to the solution in the schlenk tube over ten minutes. After stirring overnight the solution was filtered through celite and the celite washed with 25 mL ether. The solvent was removed in vacuo and this was triturated with 25 mL hexane. Filtration of this mixture yielded the light-yellow 150 as a powder (170mg, 51%) : mp 140-144°C d; IR (KBr) 3045w, 3000w, 1585m, 1570s, 1480s, 1430s, 1090s, 740s, 700s, 520s cm" ; H NMR (100MHz, CDC1,) 62.94(td, J ptH =70.4Hz, 3 J pH =llHz, HI), 4.63 (tdd, 3 J ptH =70 . 8Hz , J=9Hz, J=7Hz), 5.64(m, 2H) , 6.0(t, J=4Hz, 1H) , 6.24(m, 1H) , 7.2(Ph P, 30H); 13 C NMR (25MHz, CDC1,) S27.7(dd, 2 J 2 3 P(cis)C 5-09HZ, J p(trans)c =44.10Hz), 114.5(dd, 3 J pc =5.6Hz, 3 J pc =9.3Hz, C3), 118.5, 125.7, 127. 7(d, J=8.6Hz, Ph 3 P, meta) , 129. 2(s, Ph 3 P, para), 133. 5 (m, Ph 3 P, ortho) , 136.0 (vd, 1 J =41.7, Ph 3 P, ipso), 151.0(dd, 2 J p(ois)c =10.4Hz, 2 J p(trans)c =60.98Hz, C2) 31 P NMR (121.5MHz, CDC1 3 > «27.3(td, l J =3170. 1Hz, J pp =33.3Hz), 31.5(td, 1 J ptp =3266.7Hz, 2 J pp =33 .3Hz) ; Anal, calcd for C 43 H 36 P 2 P tl : C,63.74; H,4.45; Found: C,63.45; H,4.46. P reparation of bromo (ln 1 -cycloheptatrienyl) bis (triphenyl phosphine) platinum (152). Tetrahis (triphenylphosphine) platinum (1.07g, .86mmol) was weighed into a round bottom flask equipped with a magnetic stirrer and a nitrogen inlet. Bromocycloheptatriene (l.Og, 5.85mmol) in 50 mL benzene was added and the mixture stirred overnight.

PAGE 135

127 The solution was removed in vacuo and the product triturated with 50 mL of ether. The precipitate was filtered and dried in vacuo to give 152 (.680g, 89%) : mp 216.0-216.5 d; IR (KBr) 3050w, 3010w, 1571w, 1505m, 1480m, 1430s, 1385m, 1180m, 1100s, 1030m, 1000m, 750m, 700s, 500s cm" 1 ; 1 H NMR (100MHZ, CDC1„) 62.23(td, 3 CT =39Hz, J„ =7.1Hz, H7), J rt.n tin 4.30(dt, J=6.84Hz, J=6.0Hz, 1H) , 5.6(t, J ptH =32Hz, 1H) , 5.64(s, 2H) , 6.0(s, 3H) , 7.2(Ph 3 P, 18H) , 7.7 (Ph.jP, 12H); 13 C NMR (75MHz, CDC1 3 > <580.5(C7), 115.4, 122.9, 124.58, 127.70 (t, J pc =5.27Hz, Ph 3 P, meta) , 130. Kb, Ph g P, para), 131.0 (t, 2 J ptc =46Hz, Ph 3 P, ipso), 14.3, 134.45, 135. 2(m, Ph 3 P, ortho) , 139.4(d); 31 PNMR (121.5MHz, CDC1 3 ) S21.34(t, J Hp =3215Hz) ; Anal, calcd for C^H^Br^Pt^ C.57.94; H,4.15; Found: C,57.76, H,4.23. P reparation of (7 7-cycloheptatrienyl-7-cycloheptatriene1 ,1' -diyl) bis (triphenylphosphine) platinum ( 153 ) . Tris (triphenylphosphine) platinum (1.75g, 1.78mmol) and potassium tert-butoxide (.60g, 5.33mmol) were weighed into a dry schlenk tube equipped with a magnetic stirrer. This was dissolved in 15 mL of THF and a septum used to seal the schlenk tube. After removing from the dry box this was connected to a nitrogen line and cooled to 0°C. Bromocycloheptatriene (910mg, 5.33mmol) was weighed in a dry vial and then .5 mL of THF added. This solution was drawn into a syringe and added dropwise over ten minutes to the solution in the schlenk tube. The mixture was stirred overnight.

PAGE 136

128 The brown solution was filtered through celite and the solvent removed from the filtrate in vacuo . Trituration of this solid overnight with hexane yields pure 153 (1.48g, 93%): mp 154-5°C d; IR (KBr) 3050w, 3010w, 1480m, 1440s, 1390m, 110-m, 740m, 700s, 520s cm -1 ; 1 H NMR (300MHz, CDC1 ) S1.68(t, 3 J ptH =43.5Hz, H7 , H7 1 ), 5.65(t, J ptH =58.5Hz, 2H) , 5.75(m, 2H) , 6.15(m, 6H) , 6. 66-7. 4 (m, 30H, Ph P) ; 13 CNMR (75MHz, CDC1 3 ) S57.76(t, 2 J ptc =118 . 9 5 , C7), 118.54, 123.64, 125.06, (C 3 -C 5 ), 127.44(vt, 3 J pc =4.4Hz, Ph 3 P, meta) , 129.23 (s, Ph 3 P, para), 132.39(t, 2 J ptc =75Hz, C 2 ) , 133.58(vd, 1 PC =45Hz, Ph 3 P, ipso), 133.62(C 6 ), 134.79(vt, Z J =5.7Hz, PC Ph 3 P,^ortho), 157.62(dd, 2 J p (cis) c =8 . 5Hz, % (trans) c =l 5.12Hz, CI); P NMR (41MHz, CDClj) 525. 7(t, """Jp^ig 60 . 5Hz) . Preparation of (7 '-cycloheptatrienyl-7-cycloheptatriene-l,l ' diyl) (1,2-bisdiphenylphosphinoethane) platinum ( 163 ) . The bisadduct, 153_, (510mg, .57mmol) was weighed into a round bottom flask and toluene added. Diphos, 161, 250mg, .63mmol) was added to the flask along with a reflux condenser and the solution refluxed overnight. After cooling, the solution was concentrated to 15 mL and 20 mL of hexanes added. This was concentrated to 15 mL and filtered, yielding orange 163 (270mg, 61%) : mp 170-75°C d; IR (KBr): 3040w, 3000w, 1575w, 1482m, 1435s, 1385m, 1188m, 1100m, 690s, 725m, 740m, 530m; "4 NMR (100MHz, a,D ) 7 8 S1.4-2.0(m, 4H, diphos), 2.52(t, 3 J ptH =34Hz), 6.2-6.8(m, C ? H 6 ) , 7.0 (s, Ph 3 P) , 7. 4-7. 8m (PhjP) ; 13 C NMR (75MHz, CDC1 )

PAGE 137

129 530.21-30. 95(m, diphos) , 58.07(t, 2 J tc =114.2Hz, C), 119.45, 123.39, 126.08, 128.14(d, J J pc =5.1Hz, diphos, meta) , 128.21(d, J pc +4.7Hz, diphos, meta), 128.657(d, 3 J pc =4.8Hz, diphos, meta), 128.72(d, 3 J pc =5.0Hz, diphos, meta), 130. 4(s, diphos, para), 130. 59 (s, diphos, para), 131 .79 (diphos , ipso), 13 2.7 5 (diphos, ipso), 153 .45-134 . 26 (m, diphos, ortho) , 161.76(tdd, 1 J D .„=904.8Hz, 2 J . , =7.29Hz, 2 J ptc P(cis)C '•™' J P(trans)C 108.38Hz); 31 p NMR (121.4MHz, CDC1 3 ) 644. 2(t, 1 J p tp =186 5 . 8Hz) ; Anal, calcd for C^H^P^^ : C,61.95, H4.68; Found: C,63.29; H,5.09. Pr eparation of (7 '-3 ' , 4 ' -benzocycloheptatrienyl-7-3 , 4 benzocycloheptatrien-1, 1 '-diyl) bis (triphenylphosphine) platinum (1£4) . Tris (triphenylphosphine) platinum (500mg, .51mmol) and potassium tert-butoxide (150mg, 1.34mmol) were weighed into a dry schlenk tube equipped with a magnetic stirrer. These were dissolved in 5 mL of THF and the schlenk tube sealed with a septum. The schlenk tube was connected to a nitrogen line and cooled to 0°c. Bromoalkene 142^ (250mg, 1.14mmol) was weighed into a dry vial and then .5 mL of THF added. This solution was drawn into a syringe, removed from the dry box and added dropwise over ten minutes to the solution in the schlenk tube. After adding the solution from the syringe, the mixture was allowed to stir overnight.

PAGE 138

13 The resulting brown solution was filtered through celite and the celite washed with 25 mL ether. The solvent was removed in vacuo and the product triturated with hexane overnight. Filtration yielded the off white powder, 164. (.22mmol, 44%): mp 140-146°C d; IR (KBr) 3050w, 3000w, 1585m, 1475s, 1435s, 1383w, 1310w, 1180w, 1090m, 1000m, 840w, 740s, 690s cm" ; H NMR (100MHz, CDC1 ) 63.38(t, 3 J =28Hz, H7), 5.4-6.4(m, 6H) , 7.2-7.7(m, Ph 3 P, 38H) ; 13 C NMR (25MHz, CDC1 3 ) 657. 3(t, 2 J ptc =98.6), 123.15, 123.6, 124.6, 125.1, 125.9, 127. 6(d, 3 J pc =5.2Hz, Ph 3 P, meta) , 129. 4(s, Ph 3 P, para), 129.8, 130.0, 131.9, 132.6(Ph 3 P, ipso); 134. 9 (m, PhjP, ortho) , 137.6, 140.0, 168.6(dd, 2 Jp (cis) c =9 . 8Hz , 2 J p(trana) c =103 .8Hz) ; 31 P NMR (121.4MHz, CDC1 3 ) 625.9 ( 1 J ptp =19 50 , 8Hz) . Reaction of 135 with potassium tert-butoxide. Potassium tert-butoxide (43mg, .38mmol) and insertion product 13 5 (53mg, 50mmol) were weighed into a round bottom flask and 20 mL THF added. The mixture was stirred overnight and then filtered through celite. The solvent was removed in vacuo and the product triturated with ethanol, yielding 13 5 (39mg, 39mmol) . Similar experiments were performed with the insertion products 146 and 152 with like results. The reaction of heptaful v a lene with potassium tert-butoxide and tris (triphenylphosphine) platinum . Heptafulvalene (53mg, .29mmol), potassium tert-butoxide (lOOmg, .89mmol), and tris (triphenylphosphine) platinum were weighed into a schlenk

PAGE 139

131 tube and 15 mL THF added. The schlenk tube was sealed with a septum and the mixture stirred overnight under nitrogen. After this it was filtered through alumina (20g, neutral, activity I, 80-200 mesh) and the alumina washed in vacuo . H NMR of the crude mixture showed heptafulvalene and tris (triphenylphosphine) platinum identical with the starting material.

PAGE 140

APPENDIX EXTENDED HUCKEL CALCULATIONS All calculations were performed using the extended Hiickel method. The Hii's are taken from the literature along with the orbital exponents used. The values for the Hii's and orbital exponents are listed in Table 11 . The modified Wolf sberg-Helmholz formula was used in all calculations. The geometry for the Fp fragment is idealized and taken from the literature. The structure for bis (phosphine) platinum is based on the known structure of (ethylene) bis (triphenylphosphine) platinum. All organic structures were optimized using MINDO/3. Metal ligand distances were fixed at 2.0 8. 132

PAGE 141

13 3 Table 11. Parameters used in extended Huckel calculations. C l C 2 0.5366 0.6678 Orbital Hii,eV 6 1 6 2 Fe 3d -12.70 5.35 1. 4s -9.17 1.90 4p -5.37 1.90 Pt 5d -12.59 6.01 2. 6s -9.08 2.55 6p -5.48 2.55 P 3s -18.60 1.60 3p -14.00 1.60 C 2s -21.40 1.625 2p -11.40 1.625 H Is -13.60 1.30 0.6334 0.5513

PAGE 142

13 4 This page left blank intentionally.

PAGE 143

135 a o •H SS
PAGE 144

136 c o •H 4J n c m « P a 01 m H k a O CO § o o < a; e s o

PAGE 145

13 7 C o •H +J a) to in +> B O OJ 05 tn U n < V $ B col e o H (0 +J z < 0) c o T3 u -H CQ ^H 3 >-. V s u o, 44 H o O >i ffl g u e o 10 +> 2 < •a a «< O -H H o <; a) g § o 10 -P •z <

PAGE 146

138 c (1 •H +J 91 HI m +J c n u as in a u n 3 a) e fc e o G Z < ffl A tl +J <1> u +J 03 n <: +J a a ^H 14-1 u u >. i ) a) E e o rd 4-> 2 < x hi (ii c I' c pg m +J 4 to S H a a CN ,-) r-1 o cq cu E e o Z < O < a) e B O 10 4-1 SB
PAGE 147

13 9 c. () •H 4-1 m Kl (0 4-> c n I!) a w p u ^r 1H ^ Q „ m e (1) § o C 10 4-1 (1) 2 < Tt •H rH >l c U III e -H a 14 -P 01 4-1 a i) M-l A u <: H B (J e o & <0 4-1 z a; C U O C m ro 4J XI m O (D e o 10 4-1 z < a) e e o

PAGE 148

140 in a c c k 0) •H I o >1 CN M H

PAGE 149

141 a o •H P (D nj en -P c O 0) Q U a < «*
PAGE 150

142 U'MUUUUUHHHBBHHH C U CD 5S O 0) (6 tn oooooooooo o o o Q u * d) B a) S c IT) -P (U 2 «C H TJ at X U 5 § H >1 CN U-l * u rH 1 m g u e o « (0 -P g z < IS W r-t rH c a m +j q CO m H a a o m s o (0 -P ss < s o 3 -p 2 «!

PAGE 151

143 a 0) H H I gj J3 o Q IN rH I E •P <

PAGE 152

o < V B B O 2 < 144 w x: x fcXXUUOUUaCKKKXUO c o •-H +> < tH c Q a) -H (1) B T1 B m (0 JJ +J B < C (11 :; (! H O >1 § u V H >i c 11 X! m M O CJ « il 01 E -H 6 o T) IB +> Z < (11 C ai rH T> 0) m C U 1 c 4 m ib C EQ a) E E O n) +J

PAGE 153

145 UOOUOffiKKK oooooooo c a) a H >i Q H & M

PAGE 154

146 B -H 4-> (i) IB in +J c n CD a M c a 1) u u to H < H v >, C III H -n l*H CI o a V C « e (l) E i<0 4J () Z < r-i >1 I) rH >. CQ c 3 V a M a) 11
PAGE 155

147 uooobbsbis o o o o o o o

PAGE 156

148 C H j-i m fO en P c BS to c n U •H « Q C 9 (1) V T) -H ^H >• u, •u (J P n tt g E () S ^H 10 -P (1 2 < >1 1 4 IT ^^ pa H a T) 01 IH +J u q id 0) E A E tO 4J r-f IS < t) >1 H >1 •a i) c C <) c A n rd U P ffl a 2 < K-l O «! a § E O us +> 2 <

PAGE 157

149 OOXOUUWffiffiffi l ooooooooo 1 9 c H -P CN CN ro frj ro o CO CO CO CO C CN .—1 rH rH rH a a o O o O C\ 0\ CO CO CO CO u UD o o C CTl CI M2 io ^o KD . — . t*» O o O CO CO H r-i H rH t! (N CO *^ •^ CN CN CN CN c 01 0 rH H H H rH r-H H H •H H fc a CO CO G\ O a> c^ u a m H CN H H H rH rH rH rH rH >i i H o o c o a O o O o O — m m o o CO CO CO CO CO CO rH -H r*« rH o o ^r ^T o o o o <-* H o c T-) H rH rH rH rH •0 £ -p a a) a H rH 10 H CN Ol CXi O O rH rH U H H H CN CN CN CN >i D H >i C .a M id CO r* CO CTi O H CN n *3in u H H rH rH CN CN CN CN CN CN -H •0

PAGE 158

150 C H JJ d) rH tO 0) 0* JJ C 1 cc c < 4) V 11 o a u i S3 «C o 0) c UJ o H CQ Tl 5j !T1 V X (!) a n ^H IM t) U >l n g S i g CM « JJ •• z m w O (1) H C >1 c c n « (> 4J JJ ja oi u rd -H m « a O CO J! 6 g O rd 4J Z < O ft 111 g g o 10 -u Z Mi

PAGE 159

151 g •H 10 X ID a o rH u >• D I

PAGE 160

152 c rH o -P 1 u C a) a H V Ti m •p c 11 U-l a o a u H a) E II g :> ^H £> i (0 -H c OS n -'j u o m (U g g o a -u o < 0) g g o 10 -P 2 < En

PAGE 161

153 x
PAGE 162

154 CN c en H ~ 4J 0) c id in II P c u 0) -H OS en H >i e 01 a H u 13 10 s 4J C CI a rH IH o o a >1 CO g f-i E >1 id 4-) s z < (1 fl m U o H PQ TD < 111 c a) m H M-l H CJ >i 1 a) e to i o 1 Id -P c Z < m 1 — 1 •a a a c u • o c H m q • +) ro a in — i id -H i) K a ^H >1 H <« XI o pa H 9> S C" g o 1 rd -P l£> z <: O -H ^H IW o «e J) g g o id jj

PAGE 163

155 X IN H r-l

PAGE 164

156 o < « | E O <0 -P s < m fa X 01 c 4-> (1) (13 W 4J C 0) K en Q U a 0) g e o H <0 +J a < >i u 0) c y 0) c s (0 V >J 1 u M T3 0) c c o ^^ o c m j3 c J3 to 1TJ tH A! OS Q u a t> -H TJ IM o n tu E E 10 +> Z ft

PAGE 165

157 OOXUUCJUUUOKffiEffiES: ooooooooooooooo >1 c (D •H •a c ID ft H O >i O IS e 0) e in 01 V ID 4J a 5 s. o H O a u HI o

PAGE 166

158 jj a) ^ (0 w c 4-> c () o • c a 01 o H •n § m V *J c 0) ft i) in rH o a I) > a) E 6 -^ 10 4-1 01 c a) z < T1 H rH u >1 D c < (1) V H M 4J m 4J 1H ft u in J5 *" c o -H c >t to i« a 4-> o ja m id -H K Q nl o « J) E E O id -P IB rt i« o «c 0) g E O id 4J Z <

PAGE 167

159 0) C i C 0) H H -P i H o >l o u O

PAGE 168

160 •. c c -rH u 4J a. to c 0> H ^ tt u 4J y c (1) 9 a V .-) u >1 11 n-i iH o a >B -H •0 ~— y k c 3 (1) V 0 1 c o> g e o H CO 4-1 >0 z < ^ OJ 1 •a v 01 c o +J c u. 10 10 (!) 4-1 -C XI w 1 — , ro -H o « a o i g O nj 4J H 2 < c < a g g o id 4J 2 <

PAGE 169

161

PAGE 170

162 T1 a) C n n c n tn -p Q n m •H a. a O n fl) g e o
PAGE 171

163 as a ta o o o CN O O CF» O O if) O O H CO C -H 3 o ^

PAGE 172

164 O u < o a a e e o z <; o x; (I) c (1) 14-1 rH u rtl ai g M g It! 10 +J c 2 < 10 a •a a) c c u i o c N tn ib +j .0 "1 IT] -H « a o 0) g g o 10 4-> z <; o <
PAGE 173

165 IN t ri H r-i

PAGE 174

Q » o g p s c V H -P m H ft M-l Q ill C gi E -H E n 10 -P ft 2 < in () a a n u -H n fl < a. a) E E U (0 +J ft z < o n & o •O
PAGE 175

167

PAGE 176

168 c o •H 4J «o o ^H j-i 1 A 0! t (0 -H fZ « Q o m 2 e o z < o < 3 o 10 V z <

PAGE 177

169

PAGE 178

170 a •H 4-> ID rc en +J C o o « en 0) c 44 H o a £ Si CD g 91 g n 113 J-l 4 S < ^ n H A ^-* CJ (U cq c < X in 4-1 A n H a> g !> g o >, a -M u z « (-4 01 c a •a g g o m 4J

PAGE 179

171 O ft T-i 0) B H •a 01 £ H a c ai H fl (0 | a) o

PAGE 180

172 c H 4J l A 01 O (0 -H H K Q e g g rO JJ Z rf in < £ 1 g G 10 J-> z <

PAGE 181

173 s.

PAGE 182

174 X c o 4J 0) 10 01 P c o S < (1) V c (1) III u 4J 14-1 (1) u 4J M i O HI !> C O j C 4 n « C" 4J ' — ' XI W Id -H K a O (0
PAGE 183

175 UUUKKKffiKffi ooooooooo

PAGE 184

!<8 176 X X Q u a o a a g E O „ CJ CI a c < a) V T3 n < c u m a « Q H i-H iH o n g e E O ro +J z ml o < a) E E o (0 +> 2 bC

PAGE 185

177 UUSBKBBSBBGiC oooooooo 0) 3 c •H -U c o u g 1 •H re H 0) c H £ a a o S & 0) H M 4J rO 4J a a a o H o >1 o i

PAGE 186

178 fN r^ rH H P (1) 13 in P c g a) a « m c H P IS rH a 0, u 0) § c V •H -C a n 4-1 r o a a n (D g •H e o J3 10 P 2 < "a C 8 t) •H u m a >i < 1 rc 0) rH

. U-l H o m 1
PAGE 187

179 i

PAGE 188

REFERENCES !. Shaw, B. L.; Stringer A. J. Inorq. Chim. Acta Rev., 1973, 7, 1. 2. Baker, R. Chem. Rev. , 1973, 73, 523. 3. Otsuka, S.; Nakamura, A. Adv. Organomet. Chem., 1976, 14, 245. 4. Bowden, F. L.; Giles, R. Coordina tion Chem. Rev., 1976, 20, 81. 5. Jacobs, T. L., "Transition metal complexes of Allenes", Ch. 4 in Vol. II of The Chemistry of Allenes (ed. S.R. Landor) Academic Press, New York, 1982, 277. 6. Hewitt, T. G.; Deboer, J. J. J. Chem. Soc. A , 1971, 817. 7. Dewar, M. J. S. Bull. Soc. Chim. Fr. , 1951, 18, c79. 8. Chatt, J.; Duncanson, L. A. J. Chem. Soc , 1953, 2339. 9. Cotton, F. A. Accts. Chem. Res. , 1968, 1, 257. 10. 11. 12. 13. Foxman, B.; Marten, D.; Rosan, A.; Raghus, S., Rosenblum, M. J. Am. Chem. Soc , 1977, 99, 2160. Cope, A. C.j Moore, W. R. ; Back, R. D.; Winkler, H. J. S. J. Am. Chem. Soc , 1979, 92, 1234. Otsuka, S.; Nakamura, A.; Tani, K. J. Organometal . Chem., 1968, 14, c30. Vrieze, K.; Volger, H. C.J Pratt, A. P. J. Organometal. Chem. , 1970, 21, 467. 14. Osborn, J. A. J. C. S. Chem. Comm. , 1968, 1231. 15. Ben-Shoshan, R. ; Petit, R. J. Am. C hem. Soc, 1967, 89, 2231. 16. Vrieze, K.; Volger, H. C.J Gronert, M. ; Praat, A. P. J. Organometal. Chem . 1969, 16, cl9. 180

PAGE 189

181 17. Lennon, P.; Madharro, M. ; Rasan, A.; Rosenblum, M. J. Orqanometal . Chem. , 1976, 108 , 93. 18. Cohen, L.; Kennedy, D.; Magatti, C. V.; Sanders, A. ; Giering, W. P. J. Orqanometal. Chem. , 1974, 65_, c57. 19. Lisko, J., Dissertation, University of Florida, 1984. 20. Chisholm, M. H.; Clark, H. C. Inorg. Chem. , 1973, 12, 991. 21. Chisholm, M. H.j Johns, W. S. Inorg. Chem. , 1975, 14, 1189. 22. Chisholm, M. H. ; Clark, H. C; Hunter, D. H. Chem. Comm. , 1971, W, 809. 23. Chisholm, H. H.; Clark, H. C; Hunter, D. H. Inorg. Chem. , 1971, 10, 2557. 24. Jones, W. M. ; Wilson, J. W. , Jr.; Tutweiler, F. B. J. Am. Chem. Soc. , 1963, 8J5, 3309. 25. Jones, W. M. ; Wilson, J. W. , Jr. Tetrahedron Lett. , 1965, 1587. 26. Jones, W. M. ; Walbrick , J. M. Tetrahedron Lett. , 1968, 5229. 27. Walbrick, J. M.; Wilson, Jr. J. W.; Jones, W. M. J. Am. Chem. Soc , 1968, 9J3, 2895. 28. Balci, M.J Jones, W. M. J. Am. Chem. Soc , 1980, 102 , 7607. 29. Schmidt, M. W. ; Angus, R. 0., Jr.; Johnson, R. P. J. Am. Chem. Soc , 1982, 10_4, 6838. 30. Visser, J. P.; Ramakers, J. E. J. C. S. Chem. Comm. , 1972, 178. 31. Manganiello, F., Dissertation, University of Florida, 1983. 32. Harris, J.; Jones, W. M. J. Am. Chem. Soc , 1982, 104 , 7329. 33. Tyner, R. L.J Jones, W. M. ; Ohrn, Y.; Sabin, J. R. J. Am. Chem. Soc , 1974, 96, 3765. 34. Waali, E. E. J. Am. Chem. Soc , 1981, 103 , 3604. 35. This work.

PAGE 190

182 36. Radom, L.; Schaeffer, H. F. Nou. J. Chem. , 1980, i, 411. 37. Allison, N. T. ; Kawada, Y.; Jones, W. M. J. Am. Chem. Soc , 1978, 100 , 5224. 38. Riley, P. E.; Davis, R. E.; Allison, N. T.; Jones, W. M. J. Am. Chem. Soc , 1980, 102 , 2458. 39. Riley, P. E.; Davis, R. E.; Allison, N. T.j Jones, W. M. Inorq. Chem. , 1982, 2JL, 1321. 40. Hoffmann, R. J. Chem. Phys . , 1963, 59_, 1397. 41. Hoffmann, R.; Thibeault, J. C; Burgi, H. A.; Ammeter, J. H. J. Am. Chem. Soc , 1978, 100 , 3686. 42. Frost, A. A.; Pearson, R. G. "Kinetics and Mechanism", 1961, J. Wiley and Sons, New York, p. 75. 43. Mingos, D. M. P. Adv. Orqanomet . Chem. , 1977, 15_, 1. 44. Hoffmann, R. Science , 1981, 211 , 995. 45. Burdett, J. K. , "Molecular Shapes", 1980, J. Wiley and Sons, New York. 46. Albright, T. A. Tetrahedron , 1982, 3£, 1339. 47. Dewar, M. J. S.; Thiel, W. J. Am. Chem. S oc, 1976, 99, 4907. 48. Foxmann, B. M. J. Chem. Soc, Chem. Comm. , 1975, 221. 49. Schilling, B.; Hoffmann, R. ; Lichtenberger , D. J. Am. Chem. Soc , 1979, 101, 585. 50. Fenske, R. F.; Kostic, N. M. Qrqanometallics , 1982, 1, 974. 51. Manganiello, R. ; Radcliffe, M. D.; Jones, W. M. J. Orqanometal. Chem. , 1982, 228. 52. Jason, M. W.; McGinnety, J. A.; Wiberg, K. B. J. Am. Chem. Soc , 1974, £6, 6531. 53. Balci, M.; Winchester, W. R. ; Jones, W. M. J. Org. Chem. , 1982, 47, 5180. 54. Bennet, M. A.; Yoshida, T. J. Am. Chem. Soc , 1978, 100 , 1750. 55. Allison, N. T., Dissertation, University of Florida, 1978.

PAGE 191

183 56. Hopf, H., "[2+2] Cycloadditions of Allenes", The Chemistry of Allenes, Landor, S. R. Ed., 1982, Academic Press, New York, p525. 57. White, M. R. ; Stang, P. J. Organometallics , 1983, 2, 1655. 58. Gassmann, P. G.; Cesa, I. G. Organometallics , 1984, 3, 119. 59. Chisholm, M. H.; Clark, H. C; Manzer, L.; Strothers , J. B. J. Am. Chem. Soc , 1972, 94, 5087. 60. Silverstein, R. M. ; Bassler, G. C; Morrill, T. C, "Spectrometeric Identification of Organic Compounds", 1981, 4th Ed., J. Wiley and Sons, New York. 61. Mann, B. E.; Shaw, B. L.; Tucker, N. I. J. Chem. Soc , 1971, 2667. 62. Rajaram, J.; Pearson, R. G.; Ibers , J. A. J. Am. Chem. Soc. , 1974, 96, 2103. 63. Waali, E. E. , ; Lewis, J. M. ; Lee, D. E.; Allen, E. W. ; Chappel, A. K. J. Org. Chem. , 1977, 42, 3460. 64. Fohlisch, B.; Haug, E. Chem. Ber. , 1971, 104 , 2324. 65. Abraham, R. J.; Loftus, P.," C and H NMR Spectroscopy", Heyden and Sons Ltd., Philadelphia, Pa., 1979. 66. Barker, G. K.; Green, M.; Howard, J. A. K.; Spencer, J. L. Stone, F. G. A. J. C. S. Dalton Trans. , 1978, 1839. 67. Chappel, S. D.; Cole-Hamilton, D. J. J. C. S. Dalton Trans . , 1983, 1051. 68. Beluco, U. "Organometallic and Coordination Chemistry of Platinum", 1974, Academic Press, London, 250. 69. Ingrosso, G.; Immirizi, A.; Parri, L. J. Organomet. Chem. , 1973, 60, c35. 70. Jolly, P. W.; Kruger, C.j Salz, R. ; Skeutowski, J. C. J. Organomet. Chem. , 1979, 165 , c39. 71. McKinney, R. J.; Thorn, D. L.; Hoffmann, R. ; Stockis, A. J. Am. Chem. Soc , 1981, 103, 2595. 72. Ugo, R. ; Cariati , F.; La Monica, G. Inorganic Synthesis , 1968, 11, 105.

PAGE 192

184 73. Wittig, G.; Meske-Schuler , K. J. Liebigs Ann. Chimie, 1968, 711 , 76. 74. Radlich, T. J. Org. Chem. , 1964, 29_, 960. 75. Fohlisch, B.; Haug, E. Chem. Ber. , 1971, 104 , 2324. 76. Elsheimer, S., personal communication.

PAGE 193

BIOGRAPHICAL SKETCH William Randolph Winchester was born July 13, 1957, in Roswell, New Mexico. As his father was a pilot in the United States Air Force, his formative years were spent moving almost annually and living in nine different states. In spring, 1979, he received a B.S. in chemistry /physics from New College in Sarasota, Florida. This had followed brief stints as a Robert A. Welch Fellow at Texas Christian University and as an Undergraduate Research Fellow at Argonne National Laboratory in Chicago, Illinois. In fall, 1979, Randy entered graduate study at the University of Florida. As a graduate student he has held a University of Florida Fellowship from 1979 to 1981; won the Dupont Teaching award; held a Proctor and Gamble Fellowship; and worried often about ever finishing. Mr. Winchester has accepted an NSF-NATO Fellowship to study with Professor Dr. P. v. R. Schleyer at the University of Erlangen-Nuremberg in February 1985. 185

PAGE 194

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. William M. Jones Chairman ""~ Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. J. 9^ Sabin Professor of Physics I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. M. A. Battiste Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. occqu W. R. Dolbier, Jr. Professor of Chemistry c/

PAGE 195

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. N. Y.||Ohrn Professor of Chemistry This dissertation was submitted to the Graduate Faculty of the Department of Chemistry in the College of Liberal Arts and Sciences and to the Graduate School, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. May 1985 Dean of Graduate Studies and Research

PAGE 196

Slip


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EFW0EPOHG_36ITYW INGEST_TIME 2015-03-27T19:02:49Z PACKAGE AA00029776_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES