Citation
Synthesis and Characterization of Colloidal II-VI Semiconductor Nanorods

Material Information

Title:
Synthesis and Characterization of Colloidal II-VI Semiconductor Nanorods
Creator:
LEE, HYEOKJIN ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Alloying ( jstor )
Chemicals ( jstor )
Eggshells ( jstor )
Electrons ( jstor )
Luminescence ( jstor )
Nanocrystals ( jstor )
Nanorods ( jstor )
Photoluminescence ( jstor )
Quantum dots ( jstor )
Semiconductors ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Hyeokjin Lee. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
4/17/2006
Resource Identifier:
77080062 ( OCLC )

Downloads

This item has the following downloads:

lee_h.pdf

lee_h_Page_073.txt

lee_h_Page_058.txt

lee_h_Page_015.txt

lee_h_Page_046.txt

lee_h_Page_005.txt

lee_h_Page_079.txt

lee_h_Page_054.txt

lee_h_Page_096.txt

lee_h_Page_021.txt

lee_h_Page_012.txt

lee_h_Page_070.txt

lee_h_Page_001.txt

lee_h_Page_027.txt

lee_h_Page_033.txt

lee_h_Page_035.txt

lee_h_Page_110.txt

lee_h_Page_067.txt

lee_h_Page_071.txt

lee_h_Page_056.txt

lee_h_Page_050.txt

lee_h_Page_104.txt

lee_h_Page_082.txt

lee_h_Page_092.txt

lee_h_Page_031.txt

lee_h_Page_094.txt

lee_h_Page_085.txt

lee_h_Page_081.txt

lee_h_Page_061.txt

lee_h_Page_119.txt

lee_h_Page_095.txt

lee_h_Page_086.txt

lee_h_Page_025.txt

lee_h_Page_013.txt

lee_h_Page_077.txt

lee_h_Page_040.txt

lee_h_Page_016.txt

lee_h_Page_023.txt

lee_h_Page_121.txt

lee_h_Page_090.txt

lee_h_Page_003.txt

lee_h_Page_026.txt

lee_h_Page_098.txt

lee_h_Page_088.txt

lee_h_Page_049.txt

lee_h_pdf.txt

lee_h_Page_097.txt

lee_h_Page_069.txt

lee_h_Page_063.txt

lee_h_Page_106.txt

lee_h_Page_068.txt

lee_h_Page_037.txt

lee_h_Page_114.txt

lee_h_Page_038.txt

lee_h_Page_105.txt

lee_h_Page_036.txt

lee_h_Page_075.txt

lee_h_Page_099.txt

lee_h_Page_107.txt

lee_h_Page_112.txt

lee_h_Page_078.txt

lee_h_Page_014.txt

lee_h_Page_053.txt

lee_h_Page_062.txt

lee_h_Page_065.txt

lee_h_Page_080.txt

lee_h_Page_120.txt

lee_h_Page_006.txt

lee_h_Page_111.txt

lee_h_Page_018.txt

lee_h_Page_055.txt

lee_h_Page_089.txt

lee_h_Page_059.txt

lee_h_Page_048.txt

lee_h_Page_007.txt

lee_h_Page_030.txt

lee_h_Page_032.txt

lee_h_Page_072.txt

lee_h_Page_041.txt

lee_h_Page_042.txt

lee_h_Page_047.txt

lee_h_Page_087.txt

lee_h_Page_010.txt

lee_h_Page_117.txt

lee_h_Page_115.txt

lee_h_Page_039.txt

lee_h_Page_083.txt

lee_h_Page_108.txt

lee_h_Page_064.txt

lee_h_Page_057.txt

lee_h_Page_044.txt

lee_h_Page_029.txt

lee_h_Page_019.txt

lee_h_Page_103.txt

lee_h_Page_028.txt

lee_h_Page_024.txt

lee_h_Page_066.txt

lee_h_Page_020.txt

lee_h_Page_043.txt

lee_h_Page_017.txt

lee_h_Page_002.txt

lee_h_Page_113.txt

lee_h_Page_022.txt

lee_h_Page_009.txt

lee_h_Page_093.txt

lee_h_Page_101.txt

lee_h_Page_034.txt

lee_h_Page_008.txt

lee_h_Page_045.txt

lee_h_Page_076.txt

lee_h_Page_118.txt

lee_h_Page_004.txt

lee_h_Page_100.txt

lee_h_Page_074.txt

lee_h_Page_109.txt

lee_h_Page_116.txt

lee_h_Page_102.txt

lee_h_Page_051.txt

lee_h_Page_052.txt

lee_h_Page_011.txt

lee_h_Page_060.txt

lee_h_Page_091.txt

lee_h_Page_084.txt


Full Text












SYNTHESIS AND CHARACTERIZATION OF COLLOIDAL II-VI
SEMICONDUCTOR NANORODS















By

HYEOKJIN LEE


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


2005

































Copyright 2005

by

Hyeokjin Lee


































This work is dedicated to my wife, Eunjeong Cho, my son, Keonyoung Lee and my
parents in Korea.















ACKNOWLEDGMENTS

Above all, I am grateful to my adviser and my committee chairman, Dr. Paul H.

Holloway, who has given me sincere guidance and support for four and one-half years.

The four and one half years when I worked for Dr. Holloway is the precious period in my

life. And I have learned the attitude of a scientist from my advisor.

It is also my great honor to have four other professors (Dr. Steve Pearton, Dr.

Wolfgang Sigmund, Dr. Mark Davidson, and Dr. John R. Reynolds) as my committee

members. I am especially grateful to Dr. John R. Reynolds for advice and discussion on

synthesis of conjugate polymer.

For the sample characterization, I am thankful to Lindsay and Dr. Valeria D.

Kleiman from Chemistry Department for time resolved spectroscopy study and

informative discussion. Also I would like to thank Kerry Siebein of the Major Analytical

Instrumentation Center (MAIC) for HRTEM measurement.

I also appreciate valuable help from my colleagues in Dr. Holloway's group. I

especially thank Heesun Yang for useful discussion on luminescent properties of

nanocrystals. Another deep appreciation is expressed to Ludie Harmon for helping out

with every detailed miscellaneous task.

Finally, the continual encouragement and support of my wife, son, and parent are

deeply and sincerely appreciated
















TABLE OF CONTENTS



A C K N O W L E D G M EN T S ......... ......... ....................................................................... iv

L IST O F TA B L E S ............. . ........................... ...... ...................... viii

LIST OF FIGURES ......... ............................... ........ ............ ix

A B S T R A C T .......................................... .................................................. x iii

CHAPTER

1 IN TR O D U C TIO N ..................................................................... .......... .. .. .... 1

2 LITER A TU R E R EV IEW ................................................................. ....................... 4

2.1 Physical Properties of Semiconductor Nanocrystals ...........................................4
2.1.1 Electronic Structure of Semiconductor Nanocrystals..............................4..
2.1.2 Surface Effect on Optical Properties of Semiconductor Nanocrystals.........8
2.1.3 Interdiffusion in ZnSe/CdSe Structure............................................12
2.2 Synthetic Methods for Colloidal II-VI Semiconductor Nanocrystals ..................15
2.2.1 Binary Semiconductor Nanocrystals ................. ............................. 15
2.2.2 Colloidal Ternary Semiconductor Nanocrystals .......................................19
2.3 Anisotropic Crystal Growth of Colloidal Semiconductor Nanocrystals .............21
2.3.1 Reaction Temperature Effects on Anisotropic Nanocrystal Growth..........23
2.3.2 Surfactant Effect on Anisotropic Nanocrystal Growth ...........................25
2.3.3 Concentration Effects on Anisotropic Nanocrystal Growth....................27
2.3.4 Effects of Catalyst on Anisotropic Nanocrystal Growth..........................30
2.4 Application of Semiconductor Nanocrystals................................ .................. 31
2.4.1 Hybrid Electroluminescent (EL) Devices ...............................................31
2.4.2 Hybrid Photovoltaic D evices................................. ........................ 32
2.4.3 M etal-semiconductor Nanoassembly ............................... ............... .34

3 SINGLE STEP GROWTH OF COLLOIDAL ZNCDSE QUANTUM DOTS .............37

3 .1 In tro du ctio n ...................................... ............................ ................ 3 7
3.2 E xperim ental Section ......................................... .................. ............... 38
3.2 .1 M materials .................. ............ ................................. ............ 38
3.2.2 Preparation of ZnCdSe Quantum Dots.................. ........ ............... 38
3.2.3 ZnS Shell Growth on ZnCdSe Nanocrystals............................................39









3.2.4 Characterization of ZnCdSe Quantum Dots................... ............... 39
3.3 R results and D iscu ssion .......................... ................. ...................... ... 39
3.3.1 Single Step Growth of Colloidal Ternary ZnCdSe Quantum Dots............41
3.3.2 Effect of Reactivity of Precursor and Reaction Temperature on Optical
P ro p erties ........................................ ........... ... ...... ........................ 4 3
3.3.3 Effect of Reactivity of Precursor on Particle Growth ..............................46
3.3.4 Shell Growth of ZnS on ZnCdSe Quantum Dots ................ ........... 48
3 .4 C o n clu sio n ............. ..... .......... .............. ............. ................ 4 9

4 SYNTHESIS AND CHARACTERIZATION OF COLLOIDAL TERNAY
ZNCDSE SEMICONDUCTOR NANORODS .................................. ...............51

4.1 Introduction........................................................................ ....... ...... 51
4 .2 E x p erim ental S section ............................................. ......................................... 52
4 .2 .1 M materials .................. ...... ............................................... ....... 52
4.2.2 Preparation of ZnCdSe Nanorods.............................................................53
4.2.3 Characterization...... .......... ............................ ...... ...... ........ 53
4.3 R results and D discussion ................................................ .............. ............... 56
4.3.1 Synthesis of ZnCdSe N anorods............................................................... 56
4.3.2 Structure of ZnCdSe Nanorods .... ........... ....................................... 58
4.3.2 Effect of Alloying on the Phonon Spectra...............................................60
4.3.3 Photoluminescence and Absorption Properties........................................63
4.3.4 Time Resolved Photoluminescence (TRPL) Study ....................................68
4.3.5 Transient Absorption Spectroscopy (TRA) Study ...................................73
4.4 C conclusion ..................................................................... .......... 75

5 SELF ASSEMBLED GROWTH OF GOLD NANOCRYSTALS ON CADMIUM
SULFID E N AN OROD S ............................................. ......... ......................... 77

5.1 Introduction ..................................................................... 77
5 .2 E x p erim ental S section ......................................... .............................................7 8
5.2.1 M materials ...................... .............................. ...... ........ 78
5.2.2 Preparation of CdS Nanorods ...................................... ...............78
5.2.3 Preparation of Au/CdS Nanorods................... ....... .... ... ............ 79
5.2.4 Characterization...... ...................................... ............ ........ 79
5.3 Result and D iscussion........... .................... ................................. .. ...............80
5.3.1 Growth of Au Nanocrystals on CdS Nanorods ............... .....................80
5.3.2 Photoluminescence and Charge Separation....................................82
5.3.4 Photocatalytic A activity ........................................ .......................... 86
5.5 C conclusion ..................................... ................................ .......... 88

6 C O N C L U S IO N ................................................................................................. .. 8 9

6.1 Single Step Growth of Colloidal Ternary ZnCdSe Quantum Dots ......................89
6.2 Synthesis and Characterization of Colloidal Ternary ZnCdSe Semiconductor
N anorods ................ ........ ..... .............. ..... ... ..... ....... ......... ........ 90
6.3 Self Assembled Growth of Au Nanocrystals on CdS Nanorods ........................91









6 .4 F utu re D direction .......... ..... .......................................................... .......... .. ..... .. 92

L IST O F R E F E R E N C E S ...................................... .................................... ....................93

BIOGRAPHICAL SKETCH ............................................................. ..................107
















LIST OF TABLES


Table p

2.1 Diffusion parameters for bulk ternary II-VI compounds...........................................12

2.2 Diffusion parameter of Cd in superlattice structures ..........................................13

2.3 Properties of hybrid OLEDs using various semiconductor nanocrystals. ..................32

3.1 Calculated metal-chalcogen bond energies in metal alkyl chalcogen precursors
with the composition of M(ER)2 (kcal/mol; from reference) ................................45

4.1 Comparison of P and T value of CdSe/ZnSe coreshell and ZnCdSe alloyed
n a n o ro d s .......................................................................... 7 1















LIST OF FIGURES


Figurege

2.1 Schematic diagram of the molecular orbital model for band structure ....................4

2.2 Energy levels in CdSe nanocrystals. (a) The theoretical size dependence of the
electron and hole levels in CdSe nanocrystals. (b) The three lowest transitions as
a function of the energy of the first excited state ...................................................7

2.3 Absorption spectra of TOPO/TOP passivated CdSe nanocrsytals with radii from
1.2 to 4.1 nm. Arrows mark the positions of the four well resolved transition..........8

2.4 Schematic illustration and electronic potential step of valence and conduction
bands, HOMO and LUMO levels of (a) inorganic core and (b) inorganic
core/shell nanocrystals, both with surface attachment of organic molecules..........9

2.5 Radial probability functions for the electron and hole wave functions in bare CdSe,
CdSe/ZnS, and CdSe/CdS nanocrystals. The sketches to the right show the band
offsets between the various components ....... ........ ........................................ 10

2.6 Schematic representation of a core/shell/shell nanocrystal and band gap versus
lattice spacing of the wurzite phase of CdSe, ZnSe, CdS and ZnS.....................11

2.7 ZnSe/CdSe quantum well structure grown on GaAs substrate and PL emission
shift with anneal temperature for (a) ZnS/ZnSe, (b) ZnS/CdS and (c)
ZnSe/CdSe .............................................. ........ ....................... 13

2.8 Classification of solution phase synthetic methods applied to semiconductor
n an o cry stal s .................................................................... ................ 16

2.9 Chemical structures of representative coordinating surfactants widely used for
m olecular precursor m ethods. ............................................................................17

2.10 Illustration of molecular precursor methods for II-VI semiconductor nanocrystals
such as CdSe quantum dots ............................ ................... ............... 18

2.11 Chemical structure of the less reactive metal stearate and oleate precursors. ...........18

2.12 Chemical structure of a single molecular precursor, [CdioSe4(SPh)16]4. .................. 19









2.13 Bandgap versus lattice constant diagram for the common cubic II-VI and III-V
sem iconductors............................................................................................ 20

2.14 Parameters controlling the crystal shape in solution phase synthesis (molecular
precursor m methods) ......... ................... ......... ....... ..........22

2.15 Schematic description of nuclei and resulting shape of CdSe and CdS at (a) high
and (b) low growth temperature. c-f) HR-TEM images of CdS monorod (grown
at 300C), bipod, tripod (grown at 1800C), and tetrapod (grown at 1200C). ..........24

2.16 Illustration of growing CdSe nanorod............................................. .....................26

2.17 Precursor diffusion (arrow) from bulk solution into diffusion sphere for crystal
growth (Color gradient indicate concentration gradient) ......................................28

2.18 The CdSe nanorods grown in different precursor concentration, (a) 0.067 mol/kg
and (b) 0.267 m ol/kg. .......................................... ........................ 29

2.19 Aspect ratio of the CdSe nanorods ((a) 2.5, (b) 6, and (c) 8) as function of initial
Cd/Se ratio of precursors (a) Cd:Se=1:5, (b) 1:2, and (c) 5:1 .............. ...............29

2.20 Schematic of the mechanism of the catalyzed solution-liquid-solid (SLS) growth
process ......................................................... . ........................... 3 1

2.21 External quantum efficiency of hybrid solar cell and TEM image of the CdSe
nanocrystals used in each cell. ........................................................................... 33

2.22 Cadmium-gold nanocomposite. (a) The schematic picture of CdS/Au assembly
bridged by (b) conducting organic spacer, N,N'-bis(2-aminoethyl)-4,4'-
bipyridinium, and (c) electrically insulating spacer, 1,4-trimethyl ammonium
butane. .............................................................................. 34

2.23 TEM images showing growth of Au onto the tips of CdSe nanorods ....................35

3.1 Temporal evolution of photoluminescence spectra at reaction temperature of
320C as a function of reaction time, (a) 2 min (648 nm), (b) 10 min (623 nm),
(c) 20 min (580nm), and (d) 30 min (567 nm). (excitation at 350 nm, doubled
p e a k ) ...................................... .................................................... 4 0

3.2 Picture of luminescence of ZnCdSe nanocrystals dispersed in toluene under UV
irradiation. Nanocrystals were grown at reaction temperature of 320C and
different reaction time, (a) 5 min, (b) 10 min, (c) 20 min, and (d) 30 min..............42

3.3 Dependence of the bandgap energy (Eg) as function of composition of (a) ZnCdSe
nanocrystals grown at reaction temperature of 320C and different reaction time,
5 min, 10 min, and 30 min and (b) bulk alloy on their composition.....................42









3.4 Band gap absorption change as a function of reaction time, (a) reaction
temperature 320 C and (b) reaction temperature 270 C. Inset represents the
m agnified plot of dotted circle in (a)............................................. ............... 44

3.5 X-ray diffraction patterns of (a) bulk CdSe, (b) Cd rich ZnCdSe nanocrystals
obtained grown for 10 min. at 2700C (c) ZnCdSe nanocrystals grown for 230
m in. at 2700 C and (d) bulk ZnSe........................................ .......................... 44

3.6 Schematic reaction of precursors for nanocrystal formation............................. 45

3.7 Comparison of the size of ZnCdSe nanocrystals grown at different temperatures
(a) 270C for 230 min. and (b) 320C for 30 min (20 nm scale bar).
Photoluminescence spectrum of ZnCdSe nanocrystals obtained (c) at 270 C for
230 min. and (d) at 320 C for 30 min. after no further spectral shift....................47

3.8 HR-TEM images of ZnCdSe nanocrystals grown at 3200C as reaction time
increase, (a) 2 min (b) 5 min, and (c) 30 min (20 nm scale bar). Average particle
sizes (diameters) are (a) -5.2 nm, (b) -5.7 nm and (c) -7.6 nm with aspect ratio
1 .5 .............. .................... ........ ................................................4 8

3.9 Photoluminescence spectra (a) before ZnS shell growth and (b) after ZnS shell
growth on ZnCdSe quantum dots........................... .... .................49

4.1 Powder X-ray diffraction patterns of (a) CdSe nanorods,(b) CdSe/ZnSe core/shell
nanorods, and (c) ZnCdSe nanorods. ............................................ ............... 57

4.2 Comparison of powder X-ray diffraction patterns of (a) ZnCdSe nanorods, and
(b) spherical ZnCdSe dots. ..... ........................... .......................................59

4.3 HR-TEM image and histogram of size distribution of ZnCdSe nanorods. Lattice
fringe from a nanorod is shown by the lower right inset. ......................................59

4.4 Raman spectra of LO phonon mode of (a) CdSe nanorods and (b) CdSe/ZnSe core-
shell nanorods .............. ............ ... ............................ .. .. ......... 61

4.5 Raman LO phonon spectra of ZnCdSe nanorods after annealing at 270 C for (a) 1,
(b ) 2 o r (c) 3 h rs.................................................................. 6 2

4.6 Photoluminescence spectra of (a) CdSe-ZnSe core/shell nanorods and (b) CdSe
nanorods. ............................................................................64

4.7 UV-Vis absorption spectra of (a) CdSe nanorods, (b) CdSe/ZnSe core-shell
nanorods, and (c) ZnCdSe nanorods alloyed at 2700C for 3hrs.............................64

4.8 Photoluminescence spectra from (a) CdSe/ZnSe core/shell nanorods and ZnCdSe
nanorods alloyed at 2700C for (a) 1, (b) 2, and (c) 3 hrs ......................................67

4.9 TRPL decay curve of CdSe/ZnSe nanorod and ZnCdSe nanorods ...........................69









4.10 The dot lines are experimental data and the full lines are the fitting data using
equation ln[ln(Io/It)] versus In(time) of (a) CdSe/ZnSe coreshell nanorods, (b)
ZnCdSe alloy nanorods Ihr, (c) ZnCdSe alloy nanorods 2hr, and (d) ZnCdSe
alloy nanorods 3hr. ............................. ........ .............. ...... ...... ...... 70

4.11 Ultrafast carrier relaxation in (a) CdSe/ZnSe core shell nanorods (b) ZnCdSe
alloy Ihr (c) ZnCdSe alloy 2hr (d) ZnCdSe alloy 3hr........................................... 74

4.12 Summary of absoption study. (a) Energy separation between IS and IP transition
(b) 1S bandwidth change (c) Bandgap change of CdSe/ZnSe coreshell and
ZnCdSe nanorods ......................................... .. ...... ...... ........ .. 75

5.1 Hexagonal x-ray diffraction patterns of CdS nanorods prepared at 1200C for (a) 0.5
hrs, (b) 2 hrs, and 10 hrs, respectively. ........................................ ............... 81

5.2 HR-TEM image of CdS nanocrystals obtained at reaction time of (a) 0.5 hrs and
(b) 10 hrs. Well resolved lattice fringe of CdS nanorod is shown at (c)..................81

5.3 HR-TEM image of- 2 nm size Au nanocrystals deposited directly on CdS
nanorods crystals .............. ......................... ........... .... ..... 82

5.4 Photoluminescence spectra of (a) Cd rich and (b) S rich CdS nanorods (excitation
at 3 2 5 n m )....................................................................... 8 3

5.5 Photoluminescence spectra of (a) S rich CdS nanorods and (b) Au deposited CdS
nanorods. ............................................................................85

5.6 Energy diagram of Au deposited CdS nanocrystals: photogenerated charge
separation between CdS and Au nanocrystals (* surface hole trap state); see text
fo r d iscu ssio n ............................. .................................................... ............... 8 5

5.7 Ratio of the concentration C versus initial concentration Co of Procion red mix-5B
(PRB) dye in aqueous solution versus time of exposure to 365 nm UV light
irradiation in the presence of 5mg of (a) CdS nanorods. or (b) Au/CdS nanorods..87

5.8 UV visible absorption spectra of Procion red mix-5B (PRB) dye in aqueous
solution under 365 nm UV light irradiation in the presence of 5mg of (a) CdS
nanorods and (b) Au/CdS nanorods .............................................. ............... 87















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

SYNTHESIS AND CHARACTERIZATION OF COLLOIDAL II-VI
SEMICONDUCTOR NANORODS

By

Hyeokjin Lee

December 2005

Chair: Paul. H. Holloway
Major Department: Materials Science and Engineering

Colloidal ternary alloy ZnCdSe quantum dots (5-8 nm size) were synthesized by a

"single step" reaction using Cd and Zn oleates precursors in a trioctylphosphine oxide

(TOPO) solution. Optical properties and structure of these nanocrystals were

characterized using photoluminescence (PL), UV-Vis absorption spectroscopy, X-ray

diffraction (XRD), and transmission electron microscopy (TEM). Significant blue shifts

from 648 to 567 nm of the PL emission peaks indicated the formation of ternary alloy

nanocrystals during the reaction, and the emission wavelength was dependent on reaction

time and reaction temperature. The properties were discussed in terms of nucleation and

growth processes controlled by the reactivity and diffusion of precursors.

Colloidal ZnCdSe nanorods were synthesized by diffusion of Zn ions into CdSe

nanorods in solution at 270 C. CdSe nanorods were prepared using a mixture of

tetradecylphosphonic acid (TDPA)/TOPO surfactants at 250 C. The PL quantum yield

(QY) of ZnCdSe nanorods was 5-10 %, which is higher than that from pristine CdSe









nanorods (0.6%). XRD and TEM showed that structure of ZnCdSe nanorods was

hexagonal structure with -6 nm in diameter and -13 nm in length. Alloying and

compositional disorder during the reaction were determined by spectral shift and line

broadening of Raman spectroscopy, UV-Vis absorption spectroscopy and PL. The PL

decay was measured using time-resolved photoluminescence (TRPL) and a stretched

exponential function, I(t) = I exp[- (t/r) ], was used to describe PL decay. Comparing

CdSe/ZnSe coreshells to alloy ZnCdSe, we find a significant decrease in the 0 value

(from -0.75 to 0.48 0.58) which is attributed to compositional disorder in nanorod

crystals such as spatial fluctuations of the local Zn concentration. Emission decay time

(z) increases from 173 ns to 270-500 ns. We speculate that the binding energies of

exciton in alloy nanorods increase due to increased localization of exciton by

compositional fluctuation, leading to increase luminescence decay time (z). For

additional insight, femtosecond transient absorption has been utilized to study the

evolution of the absorption bleaching in the nanorods, which support other results

explained by alloying and compositional disorder of ZnCdSe nanorods system.

Gold nanocrystals -2 nm in diameter were grown directly on a S2- rich surface of

CdS nanorods using strong Au-S bonding nature. CdS nanorods grew from Au

nanocrystal nucleation sites prepared by the reaction of nonstoichiometric precursors in

the presence of ethylene diamine. PL was quenched significantly after Au nanocrystals

were deposited on CdS nanorods and this effect was attributed to interfacial charge

separation between Au nanocrystals and CdS nanorods. Efficient charge separation in

Au/CdS nanorods enhanced photocatalytic degradation of the Procion red mix-5B (PRB)

dye in aqueous solution under UV light irradiation.














CHAPTER 1
INTRODUCTION

Since 1990, colloidal semiconductor nanocrystals such as CdSe have been

intensively studied due to their excellent luminescent efficiency, size tunable optical

properties, and their great promise for applications in a variety of optoelectronic devices

such as light emitting diodes (LEDs) [1] and photovoltaic cells [2]. Furthermore,

colloidal nanocrystals are considered by some researchers to be the building blocks for

fabrication of quantum superstructures [3].

Synthesis of colloidal semiconductor nanocrystals with high crystal quality and

luminescent properties have been reported using molecular precursors [4-7] developed by

the research groups of Alivisatos and Bawendi. To improve the emission efficiency, the

surfaces of nanocrystals have been coated with a shell of a larger bandgap material to

confine the charge carriers and minimize the nonradiative decay channels resulting from

electronic surface states. Since 2000, metal oxide precursors with functionalized organic

ligands have been used for synthesis instead of a Cd(CH3)2 precursor for a "greener"

approach [8-10]. Metal oxide precursors are well suited for studies of colloidal

nanocrystal growth due to their slow nucleation and growth rates [8].

Even with metal oxide precursors, nanocrystal less than -2 nm may grow to final

diameters within a few seconds, making control of their size delicate and complicated. In

addition, the photoluminescent efficiency of extremely small nanocrystals is typically

lower than that of larger nanocrystals due to the increased surface-to-volume ratio as the

diameter decreases [11,12].









In order to solve those problems resulting from particle size, ternary semiconductor

quantum dots are of more interest. A few studies of colloidal ternary alloy quantum dots,

such as ZnCdS [13], ZnCdSe [14] and CdSeTe [15,16], have reported excellent

luminescent efficiencies that were comparable to those from binary core/shell structured

nanocrystals. The luminescence wavelengths from UV to IR from the ternary particles

were controlled by composition and size and were shown to be stable.

Development of growth methods for rod shaped CdSe nanocrystals promise new

opportunities to study shape-dependent electronic and optical devices, such as polarized

LED [17,18], lower threshold laser [19], and more efficient photovoltaic cells [2,20]. A

higher conversion efficiency (1.7 %) for plastic solar cell was achieved by controlling the

length of nanorods, which was attributed to better charge transport to cell electrodes [2].

The better charge conductivity of nanorods crystals may be also utilized for composite

organic light emitting diode to solve an injected charge imbalance due to poor conduction

between quantum dot nanocrystals. While they are promising, rod shaped CdSe

nanocrystals still are reported to have low photoluminescent quantum efficiencies [21,22]

and weak confinement along the rod axis has led to inefficient production of blue-green

light.

In this dissertation, studies of the synthesis of colloidal ZnCdSe nanorods will be

reported. Their luminescence in the visible region will be reported. In addition, metal-

semiconductor nanoassemblies (CdS nanorods and gold nanocrystals) were prepared and

charge separation between these two materials was studied.

In chapter 2, a review of the fundamental physics, synthetic methods, anisotropic

nanocrystal growth, and applications of colloidal semiconductor nanocrystals is given. In









chapter 3, the preparation and properties are described of ternary alloy ZnCdSe quantum

dots prepared in a single step process using a mixture of metal carboxylates (Zn-oleate

and Cd-oleate). The effects of the reaction temperature and reactivity of precursors on

optical properties, alloying and crystal growth are reported. In studies described in

chapter 4, ZnCdSe nanorods were synthesized by diffusion of Zn into the CdSe nanorods.

The ZnCdSe nanorods were characterized by absorption spectroscopy,

photoluminescence (PL), X-ray diffraction (XRD), high resolution-TEM (HR-TEM) and

Raman spectroscopy. Furthermore, the lifetimes and dynamics of colloidal ZnCdSe

nanorods were studied using time-resolved photoluminescence (TRPL) and femtosecond

transient absorption measurements (TRA). In chapter 5, CdS nanorods were prepared and

Au nanocrystals were grown directly on CdS nanorods. Structural and optical properties

of CdS nanorods and Au/CdS nanorods assemblies were determined using XRD, HR-

TEM, and PL. Nanoscale charge separation between gold and CdS nanorods was studied

by luminescence quenching and photo-catalytic properties. Finally, chapter 6 draws

conclusions and suggests future directions.
















CHAPTER 2
LITERATURE REVIEW

2.1 Physical Properties of Semiconductor Nanocrystals

2.1.1 Electronic Structure of Semiconductor Nanocrystals

Semiconductor electronic properties can be described by molecular orbitals (MO)

of the solid. In general, semiconductor quantum dots (QD) can be considered as an

"artificial atom" which is built from a small number of individual atoms [23]. When the

atomic orbitals on neighboring atoms are combined pairwise, new energy levels are

formed. Doubly occupied bonding orbital (c) and empty antibonding orbital (o*) are the

result. Each new atom adds one orbital to the bonding orbital set and to the antibonding

orbital set for each bond formed. So the number of levels in a band is equal to the number

of bonding electrons per atom times the number of atoms in the crystals.


SEMICONDUCTOR MOLECULAR ORBITALS




Fiuearam of te m a Cobanducti
UMO ,


p ................. ... ornd Gap Eg


HOMO ,

-^ j urdBnd


sP Loca ..
Atomic orl Moiecular :
Y~ i1 br8 Oss O8b 143h
Or-htoIls Oerin y of Saies

Figure 2.1 Schematic diagram of the molecular orbital model for band structure [24]









A spread of orbital energies develops within each orbital set, and the HOMO-

LUMO separation in the molecule becomes the bandgap of the bulk solid (figure 2.1).

Therefore, the optical properties of semiconductor nanocrystals depend strongly on the

size (number of atoms) of the nanocrystal.

The energy of the electronic state of a quantum dots (QD) can be described by the

Schrodinger equation. The earliest and simplest treatment of the electronic states of a QD

is based on the effective mass approximation (EMA) [25]. The EMA rests on the

assumption that if the QD is larger than the lattice constants of the crystal structure, then

it will retain the lattice properties of the infinite crystal and the same values of the carrier

effective masses. The electronic properties of the QD can then be determined by simply

considering the modification of the energy of the charge carriers produced by the

quantum confinement. Thus, the electronic properties are determined by solving the

Schrodinger equation for a particle in a three dimensional (3D) box. The zeroth order

approximation is a perfectly spherical QD with infinite potential walls at the surface.

Strong confinement is defined as the case in which the QD size is small compared with

the deBroglie wavelength of electrons in the box or compared with the Bohr radius of the

electrons, which is the case for II-VI and III-V semiconductors [26]. Taking into account

the Coulomb interaction between electrons and holes that is enhanced due to confinement

in the QD, the Hamiltonian (the sum of the kinetic and potential energy) can then be

written as

A 2V2 h2V e2
H=--- h V,(r,)+Vh(rh) (2-1)
2me 2mh e rh

and


H Y(r) = EY(r)


(2-2)









where V, and Vh are the confining potential, re and rh are the distances of the

electron and hole from the center of the QD, and c is the dielectric constant of the

semiconductor. Analytical solutions of equations (2-1) and (2-2) are difficult because the

center of mass motion and reduced mass motion cannot be separated as independent

coordinates. Various approaches to solving this problem have been used including

perturbation theory [27,28]), variational calculation [29], etc., but all lead to a solution of

the form

h 2R2 1 1 1.8e2
E 22- me1 1 e 0.25ERyd (2-3)
2m R e m h I R


where E,, is the lowest energy separation between hole and electron states in the QD

(HOMO-LUMO energy gap), ER* is the bulk exciton binding energy in meV, and R

is the QD radius. Emn is often referred to as the band gap of the QD because it

represents the threshold energy for photo absorption, that is blue shifted from the bulk

band gap, Eg, by a value dependent upon the size. The simple EMA treatment was

subsequently improved by incorporating the effective mass k*p approach that has been

used to calculate the electronic structure of bulk semiconductor [26].

The solution of the Schrodinger equation results in a description of the electronic

states in the QD by three quantum numbers plus spin because of the 3D spatial

confinement in QDs. A commonly used notation is for the electron states to be labeled as

nLe and the hole state as nLF, where n is the principal quantum number (1, 2, 3, etc), L is

the orbital angular momentum (S, P, D, etc) and F is the total angular momentum (F=

L+J, and J=L+S) where S is the spin, and the projection ofF along a magnetic axis is mF

= -F to +F.










Iri)A Rf1tiuc (A V




eo a h 0. 30] 2 S is 14
X", .



.-0-




I I Energy of e r st Excited State- (eV)
L' --- l --- f --- a 5?- eo 2 2a.4 2.$ e.B
|/P'I|O*( A ) Energy of 1st Excited State (eV)
(a) (b)
Figure 2.2 Energy levels in CdSe nanocrystals. (a) The theoretical size dependence of the
electron and hole levels in CdSe nanocrystals [30]. (b) The three lowest
transitions as a function of the energy of the first excited state [31]

Figure 2.2 (a) shows the theoretical dependence of the electron and hole levels on the size

of CdSe nanocrystals. Figure 2.2 (b) shows agreement between theory and experiment

obtained in studies of photoluminescence excitation by Norris and Bawendi [31] and the

result for the three lowest transitions as a function of the energy of the first excited state.

The three lowest electron states and hole states in order of increasing energy are 1Se, 1Pe,

1De and 1S3/2, 1P3/2, 2S3/2, respectively [30]. For optical transition in ideal spherical QDs,

the selection rules are An=0, AL=0, 2 and AF=0, 1. Therefore the first three lowest

energy bands in CdSe quantum dots can be assigned to the transitions labeled as 1S

[1Se(e)-iS3/2(h)], 2S [1Se(e)-2S3/2(h)] and IP [1Pe(e)-1P3/2(h)]. Figure 2.3 shows

absorption spectra of five colloidal CdSe nanocrystals samples with radii 1.2, 1.7, 2.3,

2.8, and 4.1 nm and show well resolved features corresponding to interband optical

transitions from coupling to different electron and hole quantized states [32]. These ideal

selection rules can be broken by nonspherical QD. Furthermore,










decreasing the degree of symmetry and the anisotropy due to rod-shaped crystals leads to

changing of the degeneracy and splitting of the excited states of semiconductor

nanostructures [33].

2.1.2 Surface Effect on Optical Properties of Semiconductor Nanocrystals

The surfaces of nanocrystals play a key role in their electronic and optical

properties, in part due to the high surface to volume ratio of semiconductor nanocrystals.

Some important properties of semiconductor nanocrystals for various type of application

are (1) high luminescence quantum efficiency, (2) stability of luminescence properties in

real operational conditions, and (3) dispersion of nanocrystals in a desired solvent for

processing. All of these properties deal with or are influenced by passivation of dangling

bonds present on the nanocrystal surface. Therefore surface modification of





























from 1.2 to 4.1 nm. Arrows mark the positions of the four well resolved
transition [32].
a 7m










S h, I IS< Si I )
I l ISj'aht
2. z.2 2.4 2A6 28 3.0 o .2
Photon energy (eV)


Figure 2.3 Absorption spectra of TOPO/TOP passivated CdSe nanocrystals with radii
from 1.2 to 4.1 nm. Arrows mark the positions of the four well resolved
transition [32].









semiconductor nanocrystals has been the subject to extensive investigation.

Band edge emission from nanocrystals competes with both radiative and

nonradiative decay channel originating from surface states. Organic ligands attached to

the nanocrystal surface can affect the solubility of nanocrystals, as well as result in

improved luminescent efficiency [5, 34]. However, a high luminescence quantum yield

(QY) is difficult to achieve in semiconductor nanocrystals coated by organic ligands due

to imperfect surface passivation. In addition, the organic ligands are labile for exchange

reactions due to weak bonding to the nanocrystal surface atom exposing the surface to

degradation effects such as photooxidation [35, 36]. In some cases, chemical degradation

of the organic ligand molecule itself might lead to degraded luminescent nanocrystals

[36].

organic
molecule










Sand r

offset ---- (

band ... .......




(a) (b)


Figure 2.4 Schematic illustration and electronic potential step of valence and conduction
bands, HOMO and LUMO levels of (a) inorganic core and (b) inorganic
core/shell nanocrystals, both with surface attachment of organic molecules.







10


A proven strategy for increasing both luminescence QY and photostability is to

grow a shell of a higher band gap semiconductor on the core nanocrystals. Examples

include CdSe/ZnS [6,37]}, CdSe/CdS [7], CdSe/ZnSe [38], CdS/ZnS [22] and InAs/ZnSe

[39] (figure 2.4). Not only effective surface passivation, larger band gap shell materials

provide also a potential band offset for electron and holes originating in the core

nanocrystals, reducing the probability for the carriers to be trapped at the surface.

Therefore higher luminescence QY can be achieved by inorganic shell growth over a

luminescent core. When an inorganic shell material such as ZnS surrounds the core

nanocrystal, the absorption and emission energies shift to lower values (ca. 10-20 nm)

than those from bare quantum dots [6,37,40]. Dabbousi et al. explained the influence of

surface passivation on shift in energies using a simple theoretical treatment of charge

carriers in a spherical box [6].


CdSe CIS Ihrlt \"2

(Lor
,0,4 *




D: (CdSnS f enes, zs 1






th iC 1S. Cri ]
c-J L
~ -*

0 LV. -

S S 1V 15 K- 25
Distance from Centi (A)


Figure 2.5 Radial probability functions for the electron and hole wave functions in bare
CdSe, CdSe/ZnS, and CdSe/CdS nanocrystals. The sketches to the right show
the band offsets between the various components. [6]









The wave function of the lighter electron tunnels into the shell, while the hole wave

function has a negligible probability of spreading into the shell layer. The increased

delocalization of the electron lowers its confinement energy and consequently the energy

of the excited state. The probability of penetrating into the shell layer depends on the

barrier height between the core and shell, which means that lower barrier heights lead to

larger energy red shifts [6] (figure 2.5).

An additional effect on the luminescence QY is the lattice mismatch between the core

and shell of the nanocrystals, because dislocations induced by interfacial strain may act as

nonradiative recombination centers. Two methods have been reported for reducing the

defects at the core/shell interface. One method is growth of a compositionally stepped

shell, such as a CdS/ZnS shell/shell structure grown on a CdSe core (figure 2.6). The

intermediate CdS shell results in relaxation of the strain and consequently improved PL

QY and stability of nanocrystals [22,41]. In the other method, which is called

"photoannealing", defects at the highly strained interface can diffuse to the outer surface

upon irradiation of the nanocrystal with a laser [22].





SZnS Cs

[o ,C"^ CdS


cASeu
62 a6ac ,e B 0
c-lattice spacing A


Figure 2.6 Schematic representation of a core/shell/shell nanocrystal and band gap versus
lattice spacing of the wurzite phase of CdSe, ZnSe, CdS and ZnS. [41]









Despite its great importance, characterization of surfaces and interface of the

core/shell nanocrystals is still limited due to a number of factors.

2.1.3 Interdiffusion in ZnSe/CdSe Structure

Interdiffusion at the core-shell interfaces may occur and change the band gap

energy, alter the electron and hole confinement, and result in large changes in the

subband energies of the conduction and valence bands. Solid state diffusion is thermally

activated and the diffusion coefficient, D, follows the Arrhenius relationship

D(T) = Do exp(-Q/kT) (2-4)

where Do and Q are the experimentally determined infinite temperature diffusivity

and activation energy, respectively. An estimate of the extent of interdiffusion occurring

at temperature T in time t is the root mean square displacement from the random walk

expression

Id = = D(T)t (2-5)

Values of Do and Q for interdiffusion of atomic species in wide band gap II-VI

compounds are not well established, although diffusivities and qualitative results are

available over a limited temperature range. A summary of the experimental diffusion

parameters at the specified temperatures is given in Table 2.1.

Martin [42] determined values of Do = 6.4 x 10-4cm2s-1 and Q=1.87 eV for Cd

diffusion into ZnSe. These values yield diffusion length of 0.13, 0.58, 2.1, 6.7 and 41nm

Table 2.1 Diffusion parameters for bulk ternary II-VI compounds
System Temperature (C) Composition Do (cm2-1) Q (eV)

Zni-xCdxSe 700-950 0 < x <0.36 6.4 x 104 1.87
Zni-xCdxTe 700-1010 0.1< x<0.9 0.29exp(2.32x) 2.14
Ref) [42, 43]







13


S(b) 6 t 5-
4 nm ZnSe
1 nm CdSe I / -
500 nm ZnSe

(a) (C)
(100) GaAs I 0

ZnSe-CdSe system I
28 o L
t W 400 Soo 50tro
Caitnl Arditworlur P ) sunpls Amd. ImpenomlreMCI
swsia


Figure 2.7 ZnSe/CdSe quantum well structure grown on GaAs substrate and PL emission
shift with anneal temperature for (a) ZnS/ZnSe, (b) ZnS/CdS and (c)
ZnSe/CdSe. [44]


in layers annealed for 1 hour at 300, 350, 400, 450, 500, and 550C respectively. Even

one monolayer diffusion distances (-0.3 nm) would be expected to produce observable

effects on optical properties in nanoscale systems, PL shifts could be expected at 350C

according to Martin's data. However, Parbrook et al. reported that ZnSe/CdSe and

ZnS/CdS superlattice structures grown by the MOCVD method show disordering at

annealing temperatures greater than 400 and 450C (figure 2.7) [44].

However, no alteration was found for the samples annealed at temperatures lower

than 400C. Anionic interdiffusion was found not to be significant at temperatures below

550C. They speculated that this discrepancy was due to the high temperature range

(700-950C) of Martin's experiments [44].

Table 2.2 Diffusion parameter of Cd in superlattice structures [47] [45]
System Temperature Do (cm2s-1) Q (eV) d*

ZnSe/CdSe/ZnSe 340-400 1.9 x 10-4 1.8 0.14 nm

ZnCdSe( Onm)/ZnSe(1 Onm) 360-600 1.2 x 10-5 1.7 0.01 nm

Calculated diffusion length after 1 hours at 300C









On the other hand, diffusion coefficient of Cd in quantum well structures in the

range of 340C-400C was calculated by Rosenauer et al. and Strapburg et al. (table 2.2).

Interdiffusion in the ZnSe/CdSe system was negligible at 300C; for example, diffusion

length of Cd after Ihr was only 0.14 nm. However, the diffusion behavior of Cd in

quantum well structures is significantly affected by the growth condition, doping, and

atmosphere during annealing, due to their effects on the density of group II vacancies

which accelerate interdiffusion [45, 46].

Diffusion parameters in colloidal nanocrystals have not been reported. However,

Zhong et al observed significant interdiffusion at 270-290C in colloidal CdSe/ZnSe

core/shell nanocrystals and significant PL blue shift [14]. A temperature of 270C, which

they called the alloyingg point", is much lower than previous reported temperature for

diffusion in bulk or quantum well structures. Diffusion at lower temperatures may be

reasonable in nanoparticles because of the large surface to volume ratio as compared with

bulk or quantum well structure. Another important parameter is imperfections at the

interface between nanocrystal core and shell layer. In a system containing only a few

hundred atoms, a large fraction of these atoms will be located on the surface or the

core/shell interface. As surface atom bonds tend to be unsaturated, there is a large energy

associated with this surface. In the solution, surface atoms move to minimize surface area

and minimize surface energies by reconstructing [48]. Defects at the core/shell interface

might result in higher diffusion rates than for quantum well (QW) structures grown by

MBE. An increased concentration of vacancies or imperfection at the interface between

the core and shell is consistent with "photo-annealing" effects on luminescent efficiency

[22]. When nanocrystals were exposed to UV light irradation, the luminescent efficiency









of CdSe/ZnS core/shell nanocrystals increased by factor of 2-10 due to reduction by

annealing of vacancies [22]. Yet another possible explanation of accelerated diffusion in

nanocrystals is a lower activation energy (Q). The coordination of ions close to the

surfaces is not satisfied in small size nanocrystals, i.e. the coordination number is less

than that for ions in the bulk. In addition, the interaction between distant neighbor atoms

in nanocrystals is much weaker than that in the bulk crystals [49,50]. Weak interactions

between atoms in colloidal nanocrystals may lead to lower activation energy for

interdiffusion. This argument is consistent with weakening of the crystal field strength

with decreasing size [49,50]. For example, in the case of ZnS:Eu2 nanocrystals, the

crystal field strength of ZnS nanocrystals is weaker than for bulk materials. Therefore

emission energy of Eu2+ shifts to higher energy [49,50].

2.2 Synthetic Methods for Colloidal II-VI Semiconductor Nanocrystals

2.2.1 Binary Semiconductor Nanocrystals

The solution phase methods used in the synthesis of semiconductor nanocrystals

are classified into two main categories (figure 2.8): Controlled precipitation, or Molecular

precursors. Controlled precipitation includes those methods based on traditional

precipitation techniques, forming a stable nanocrystal from the mixture of the ionic

components of the semiconductor. In this method, the stability of the nanocrystals results

from using stabilizers in the solution (e.g., polymers [51], or surfactants [52]). Sol-gel

methods [53], reverse micelle [52] and hydrothermal methods [54,55] are included in this

category.

Molecular precursor methods have been developed by the groups of Alivisatos and

Bawendi, as the most promising chemical routes to synthesize quantum dots for the

optoelectronic and biological application. [4-7] In these procedures, the molecular









precursors are decomposed and react in a coordinating solvent at relatively high

temperature, hence promoting the crystallinity and passivating the dots surfaces with

appropriate surfactants. A wide range of II-VI semiconductor nanocrystals have been

prepared using such methods.

Figure 2.9 shows chemical structure of some surfactants that are most widely used

in the molecular precursor methods. In the conventional TOPO method

(trioctylphosphine oxide) described by Bawendi and co-workers, the group VI species

(e.g., TOP-Se) are injected rapidly into hot TOPO solution containing the group II

species (e.g., Cd(CH3)2) and the reaction solution is vigorously stirred to avoid

inhomogeneous particle growth.






Controlled precipitation Molecular precursor

Reverse micelle method Organometallic precursor
Sol-gel method Pyrolysis at high temp.
Hydrothermal method







Figure 2.8 Classification of solution phase synthetic methods applied to semiconductor
nanocrystals

Consequently, a large number of nucleation centers are initially formed, and

continued to grow via Ostwald ripening (i.e. the growth of larger particles at the expense

of smaller particles to minimize the higher surface free energy associated with smaller

particles) [56].

















P P
trioctylphosphineoxide (TOPO) trioctylphosphine (TOP)



OH
O=P-OH

tetradecylphosphonicacid (TDPA) Trioctylamine (TOA)



Figure 2.9 Chemical structures of representative coordinating surfactants widely used for
molecular precursor methods.

The coordinating surfactants (TOPO and TOP) limit or control particle growth.

Size-tuning is achieved during reaction by adjusting the time and temperature of

synthesis. Nanocrystals whose surfaces are passivated with coordinating surfactants are

obtained, allowing further manipulation such as size selective process and shell growth

using other solutions. This method is illustrated in figure 2.10. Several modifications to

this method have been reported for better control over growth dynamics and production

of monodispersed nanocrystals.[8,57] To avoid the use of toxic, highly reactive and

pyrophoric compounds such as Cd(CH3)2 at high temperatures, several modifications to

this method have been reported using alternative precursors which are relatively stable,

inexpensive, and safe [8-10]. The alternative precursors, metal carboxylate complexes,

are prepared by reaction of metal oxide (e.g., CdO or ZnO) with alkyl carboxylic acids,

such as stearic acid (figure 2.11). The reactivity of such precursors toward group IV

species is lower than the traditional organometallic precursors such as Cd(CH3)2 and







18










Se-TOP
injection


N2


A 250-320C


Cd source
In TOPO


Figure 2.10 Illustration of molecular precursor methods for II-VI semiconductor
nanocrystals such as CdSe quantum dots. [4-7]


Zn stearate


Zn stearate


Cd oleate


Figure 2.11 Chemical structures of the less reactive metal stearate and oleate precursors.









and Zn(CH3)2. Therefore, alternative route using metal carboxylate are more suited to

control the growth of high quality semiconductor nanocrystals and for studying growth

mechanisms of colloidal nanocrystals due to their slow nucleation and growth rate [8].

Single molecular precursors, which contain all of the elements for the desired

nanocrystals, have been reported to be alternative precursors (e.g. [CdioSe4(SPh)16]4- or

[MeCdSe2CN(C2H5)2]2). [58-60] The use of single molecular precursors with preformed

metal-chalcogenide bonds provides a convenient reactive intermediate and allows

nanocrystal growth to be initiated at relatively lower temperatures. [58] Incorporation of

dopants, such as Eu3+, into the semiconductor nanocrystals can be achieved effectively by

using the single molecular precursor described above.[61,62] The use of prebonded

complex materials such as Mn2(t-SeCH3)2(CO)s have been applied to doping transition

metal into semiconductor nanoparticles to reduce the extent of dopant segregation on the

particle surface.[63, 64]




leJO



Q I









2.2.2 Colloidal Ternary Semiconductor Nanocrystals

A variety of II-VI ternary and quaternary alloy semiconductors based on ZnSe have

been studied for LEDs and laser diodes (LDs).[65, 66] Compositions of these materials












ZnS


3-


O CdS

U 2- CdTe

CInP

1 -
PbS InAs
PbSe
0 I I I I
5.25 5.50 5.75 6.00 6.25 6.50
Lattice Constant (A)



Figure 2.13 Bandgap versus lattice constant diagram for the common cubic II-VI and III-
V semiconductors. [70]

have been demonstrated whose wavelengths cover the whole visible range. For example,

ZnSSe (blue-green), ZnCdSe (green-yellow) and ZnCdSeTe (yellow-orange) have been

grown as an emission layer on appropriate substrate such as GaAs and InP [67-70].

Modification has been reported to reduce the defects and increase the device life

time. [67-70]. However, very little effort has been reported on growth of colloidal ternary

alloy nanocrystals, although nanocrystalline binary compounds such as CdSe have been

intensively studied for use of optoelectronics and bio-medical imaging. Figure 2.13

shows the energy gap versus the cubic lattice constant of binary II-VI semiconductors.

Ternary alloyed quantum dots, a solid solution of two binary semiconductors, would be

expected to exhibit intermediate energy band gaps between those of the constituent

binaries, and opens new possibilities in band gap engineering. Various types of ternary









alloy quantum dots have been prepared using the molecular precursor methods. In 2003,

Bailey et al. reported composition tuning of colloidal CdSeTe nanocrystals. [15] Light

emission from these CdSeTe nanocrystals could be tuned over the range from 650nm to

800nm with QYs of up to 60% [15,16]. Zhong et al. also reported ternary alloy ZnCdSe

and ZnCdS quantum dots with high PL efficiency (25-70 %) and narrow band

luminescence properties (15-30 nm) in the near UV visible region [13,14].

The luminescence wavelength of ternary systems can be controlled by composition

and particle size from UV to IR. By judicious choice of the composition, the wavelength

desired can be achieved with stable larger particles, while in binary systems the emission

wavelength is only controlled by the nanocrystal size. In general, nanocrystal less than -2

nm may grow to final diameters within a few seconds when molecular precursor method

is used for growth. Therefore controlling particle size below -2 nm is very delicate and

complicated. In addition, the photoluminescent efficiency of smaller nanocrystals is

typically lower than that of larger nanocrystals due to the increased surface-to-volume

ratio as the diameter decreases.[11, 12] Some of the problems resulting from particle size

can be avoided or reduced in ternary quantum dots. High luminescent intensities can also

be achieved from ternary compounds, at times being comparable to core/shell structured

nanocrystals. Zhong et al [14] suggested that high luminescence of ternary quantum dots

may result from spatial composition fluctuation that produce atomically abrupt jumps in

the chemical potential which localizes the free exciton states and leads to high luminance

and stability.

2.3 Anisotropic Crystal Growth of Colloidal Semiconductor Nanocrystals

One-dimensional (ID) nanostructures such as wires, rods, belts, and tubes are of

interest due to their unique applications in mesoscopic physics and fabrication of









nanoscale devices [71]. The understanding of ID nanostructures (nanorods and

nanotubes) has been slow due in part to difficulties in the synthesis and fabrication of

these nanostructures with well controlled dimensions, morphology, phase purity, and

chemical composition. In recent years, shape control of nanocrystals through molecular

precursor method has been successful [72-74]. By controlling growth variables, various

shapes of nanocrystals have been produced. Colloidal nanorods synthesized by these

methods exhibits potential technological advantages over spherical nanocrystals, such as

linearly polarized emission,[17,18,75,76] lasing from quantum rods in the visible range

[19,77], and improved solar cell performance [2, 78].

There are several parameters that can influence the growth pattern of nanocrystals

(figure 2.14).


AG




I \ ,






Surfactants Catalyst





Figure 2.14 Parameters controlling the crystal shape in solution phase synthesis
(molecular precursor methods)

The intrinsic surface energy of the crystallographic face of the seed is important,

since the kinetic energy barrier (AG) is inversely proportional to the surface energy [79].









Therefore, the final shape of nanocrystals is dependent on the crystal structure of the seed

crystals which are influenced by reaction temperature. However, surface properties can

also be tailored by the types and the amounts of adsorbing organic capping molecules.

The concentration of precursors also plays a key role for the determination and evolution

of the shapes of nanocrystals. Rod growth can also be achieved by using catalyst in the

case for materials with zinc blend cubic structures (see 2.3.4). In the following section,

the factors important to the anisotropic growth of II-VI semiconductor nanocrystals are

summarized.

2.3.1 Reaction Temperature Effects on Anisotropic Nanocrystal Growth

One of the critical factors responsible for the shape determination of the

nanocrystals is the crystallographic phase, although the stable phase is highly dependent

on its environment. Initially, the crystalline phase of the nuclei is critical for directing the

intrinsic shapes of nanocrystals due to its characteristic unit cell structure and consequent

variation of surface energy with crystallographic orientation. After the preferred

crystalline phase is nucleated, the subsequent growth stage strongly governs the final

architecture of the nanocrystals through the delicate balance between the kinetics and

thermodynamics of the growth. The crystalline phase of the seeds can be controlled by

adjusting the initial temperature during the nucleation process (figure 2.15).

In the case of CdS semiconductor nanocrystals [80], nuclei with a wurtzite structure

are more stable at high temperature (-300C) and nanorod formation is observed.

Because the surface energy of the (001) face of the wurtzite phase is typicallyhigher than

that of other faces due to its higher atomic density and number of dangling bonds, the

preferred growth is along the c axis. On the other hand, at lower temperatures (120-180

C), zinc blende nuclei are preferred and tetrahedral seeds with four { 111} faces are









formed. The growth of wurtzite pods along the [001] directions from the { 111 } faces and

the formation of CdS multipods can be achieved by low temperature growth


('dSe, ((IS
S () CrI- stallne ,i
I phase of seed



wurtzite Rod
--------------------- -----------------

S(b) -
II !F'

zinc blende n. "
tetrapod



Figure 2.15 Schematic descriptions of nuclei and resulting shape of CdSe and CdS at (a)
high and (b) low growth temperature. c-f) HR-TEM images of CdS nanorod
(grown at 300C), bipod, tripod (grown at 180C), and tetrapod (grown at
120C) [79].

Similarly, in the case of CdSe nanocrystals [73,81 ] control of the temperature

influences the formation of wurtzite or zinc blende nuclei. If conditions that favor the

cubic phase are set, then large tetrahedral nanocrystals are obtained after formation of

zinc blende nuclei. One way of achieving this is keep the system at low temperature (250

C). On the other hand, hexagonal growth is promoted by high temperature or an increase

in the concentration of monomers and result in growth of nanorods. For complicated

anisotropic shape, zinc blende nuclei can be grown, and conditions can be switched from

the thermodynamic growth regime to the kinetic growth regime by increasing the

temperature, or by increasing the precursor concentration, or both. Wurtzite rods are

formed off of the { 111 } faces of the original zinc blende nucleus leading to tetrapod









nanocrystals. However, the rapid switching is difficult to achieve in a reproducible way

in a batch type reactor because the nucleation and growth of nanocrystals take place on a

time scale of seconds to minutes.

In the case of Cd chalcogenide nanocrystals, the thermodynamically stable form

can be controlled with reaction temperature if the precursor concentration is low. Other

types of semiconductor nanocrystals, such as PbS, can be grown with various shape by

changing the reaction temperature. [82] The effects of other parameters on anisotropic

growth must be considered, such as mixture of surfactant, concentration of precursor, or

catalyst.

2.3.2 Surfactant Effect on Anisotropic Nanocrystal Growth

The nanocrystal surface is coated with a layer of organic molecules known as

surfactants during solution growth. At a reaction temperature in the range of 200-400C,

the surfactant molecules are dynamically adsorbed to the surface of the growing crystals.

They are mobile enough to provide access for the addition of precursor units, but stable

enough to prevent the aggregation of nanocrystals. In addition, the surfactant molecules

are chemically stable at the high temperatures required for growth. Surfactants that bind

too strongly to the surface of the crystals would restrict the crystal growth. On the other

hand, a weakly bound molecule would yield large particles, or aggregates. Some

examples for the growth of CdSe include alkyl phosphine, alkyl phosphine oxides, or

amines as shown in figure 2.9. Furthermore, surfactant coating allows for great synthetic

flexibility in organic solution in that it can be exchanged for another coating of organic

molecules having different functional groups or polarity.

The surface energy difference of facets in wurtzite crystals can be changed by

adjusting the type and ratios of surfactants.[72, 79] In other words, the shape of the









nanocrystal can be controlled by using surfactants (or mixture of surfactants) that bind

differently to the different crystallographic faces. Peng et al. pioneered the use of mixed

surfactants to control the shape of CdSe nanocrystals.[72, 81, 83] They found that CdSe

quantum rods with aspect ratios as high as 10 could be obtained in large quantities by

adding hexylphosphonic acid (HPA) to TOPO.

Explanations for nanorod growth with particular orientation are based on inherent

anisotropy in the wurtzite structure and the surface coordination of the surfactants. [73]

The surface energy of the (001) face of the wurtzite phase is typically higher than that of

other faces due to its higher atomic density and number of dangling bonds as mentioned

in a previous section. These facts make the (001) facet chemically more reactive than
















(d:l Se:# Ligand:A 4 AA

Figure 2.16 Illustration of growing CdSe nanorod [83].

other facets. Dangling bonds of surface atoms on the (001) facet cannot be fully

passivated by surfactants due to steric interference between neighboring surfactant

molecules, whereas surface atoms on other faces can be passivated effectively

[73,81,83,84]. In addition, alkyl phosphonic acids (PA) such as HPA are bound more

strongly to the nanocrystal surface than TOPO, as confirmed by calculation by Puzder et









al [84].The strong binding ability of PA is responsible for further raising the surface

energy difference between (001) face and some others of CdSe and enhancing the relative

growth rates of different facets. Figure 2.16 show illustration of growing CdSe nanorods

passivated by a surfactant. The (001) faces are either Cd or Se terminated. If a layer of Se

atoms is exposed on this face, the phosphonic acid molecules will passivate them very

weakly, because Se atoms in CdSe have an anionic character and the surfactants in the

system bind strongly to cationic species. This situation allows for the rapid building up of

a layer of Cd atoms, which would again slow down the growth of this face. When Cd

atoms are attached to Se atoms, the alkyl phosphonic acid cannot passivate all the

dangling bonds from Cd atoms due to steric interference as illustrated. In figure 2.16,

surface area per Cd atom in (001) face and (100) face is about 0.19 nm2/Cd and 0.30

nm2/Cd respectively. However, surface area per surfactants is about 0.25 nm2. Therefore

the steric interference is larger at (001) face that at (100) face. Combined with the fact

that the sides of nanocrystal can be passivated by either phosphonic acid or TOPO, the

(001) face becomes the fastest growing face of the crystal.

2.3.3 Concentration Effects on Anisotropic Nanocrystal Growth

Geometric control has also been demonstrated by varying the concentration of the

precursors in the growth solution. Peng et al. [83,85,86] have systematically studied the

anisotropic growth of CdSe and explained nanocrystal growth in terms of precursor

concentration effects.

Figure 2.17 show schematic diagram of nanocrystal growth at high precursor

concentration. Precursors should diffuse into the diffusion sphere around growing crystal

due to high concentration gradient when precursor concentration in the bulk solution is

high. As a result, the volume of nanocrystals increases noticeably.




























1- Precursor diffusion


Figure 2.17 Precursor diffusion (arrow) from bulk solution into diffusion sphere for
crystal growth (Color gradient indicate concentration gradient)

However, precursor supplied by diffusion is consumed by the quick growth of the

(001) facet. As a result, the diffusion flux goes to the c axis exclusively and makes the

growth reaction rate along the c-axis much faster than that along any other axis leading to

high aspect ratio (length vs. diameter) of nanorods [83].

At low precursor concentration, there is no net diffusive flux between the bulk

solution and the growing nanoparticle. The precursors on the surface of the nanocrystals

adjust their positions to minimize the total surface energy of a given crystals and

eventually the aspect ratio of crystals decreases [83]. A lower precursor concentration

will lead to Ostwald ripening of the particles which is particle growth by consuming

smaller size particle [83, 85, 86]. Consequently, relatively high precursor concentrations

promote the formation of nanorods with a high aspect ratio (figure 2.18). Other

experimental factors can affect nanocrystal growth in the same manner as the initial










[Cd]=0.067 mol/kg [Cd]=0.267 mol/kg





/ f4 -.
... -..-.T.z



100 nm r 100 irnm

(a) (b)
Figure 2.18 The CdSe nanorods grown in different precursor concentration, (a)
0.067mol/kg and (b) 0.267mol/kg. [85]

precursor concentration. For example [85] :

i). Reactivity of precursor: a more stable Cd precursor reduces the nucleation rate,

and less nuclei are formed. Therefore the concentration of the remaining precursor at

initial stage is higher than that of unstable reactive monomer.








.'AM jII : .. :, i I i






(a) (b) (c)
Figure 2.19 Aspect ratio of the CdSe nanorods ((a) 2.5, (b) 6, and (c) 8) as function of
initial Cd/Se ratio of precursors (a) Cd:Se=1:5, (b) 1:2, and (c) 5:1. [85]
t;' :: . '
,-, .... (.c->I i'' ... .. .
Figue 2.19 Apectrati of he C~e nnoros (.) 2.,.(..6, nd o) 8 .............
......... '...e ,,,i o. p e u s r a d S l : (',b' ,,2 :'. ( ..... .. .









ii). Initial Cd/Se ratio: The Se precursor is more active than the Cd precursor. As a

result, higher Se precursors lead to faster nucleation with consequent decreases in the

precursor concentration and a lower aspect ratio of the nanocrystals (figure 2.19).

iii). Multiple injections of precursor: After the primary injection of a large excess

of the Cd precursor, the more reactive Se precursor may be increased gradually as using

multiple injections.

iv). High growth temperature: High temperatures increase the diffusion flux

towards the fast growth facets, leading to a higher aspect ratio.

2.3.4 Effects of Catalyst on Anisotropic Nanocrystal Growth

So far, discussions of the growth of rod-shaped nanocrystals were based on the

presence of chemically dissimilar lattice faces that can be selectively bound by

coordinating surfactants or precursors to achieve different growth rates along different

crystal axes. Therefore nanorod growth has been limited to semiconductors with wurtzite

crystal structures, such as hexagonal CdSe. However rod growth can be prepared

effectively by the solution-liquid-solid (SLS) mechanism for lattice structures that do not

exhibit chemically dissimilar surfaces, e.g. zinc blend cubic structures.[87,88] In one

mechanism for catalyzed SLS growth, precursors diffuse into metal nanocrystals

(catalyst) and precipitate when the catalyst solution is supersaturated, resulting in the

formation of one dimensional nanocrystals (figure 2.20)

Kan et al. reported the formation of InAs nanorods (InP) which were prepared by

the reaction between tris(trimethylsilyl)arsine (TMS)3As and InC13 in TOPO using

dodecanethiol stabilized gold nanocrystals with diameters of 2 nm [87,88]. The Au

nanocrystals act as a catalyst into which reactants dissolve, leading to directed growth

from the supersaturated solid solution. The melting temperature of 1.6 nm Au is 350









C, increasing to 480 C for 2.5 nm Au particles which is significantly lower than that

of bulk gold [87,88]. The smaller nanocrystals, the larger contribution made by surface

energy to overall energy of the system and thus the more dramatic the melting

temperature depression [48]. At temperatures just below this melting but above the

boiling point for TOPO, rapid diffusion can take place leading to nanorod growth. Using

Bi or Au/Bi catalysts, Yu et al. reported growing high quality, narrow diameter CdSe

nanowires. [89, 90]

2.4 Application of Semiconductor Nanocrystals

2.4.1 Hybrid Electroluminescent (EL) Devices

Since the first observation of light emission from organic materials by Tang et al

[91], continuous and rapid improvement in device performance have enabled organic

light emitting devices (OLEDs) to compete with existing technologies. However there are

still many problems to be overcome, such as improving device stability and color purity.

Full width at half maximum (FWHM) of photoluminescence of colloidal semiconductor



Precipitation

Precursor

ij / ** ^*" -k*k.k.k: .\J

Byproducts < .I


Solution Catalyst Nanorod




Figure 2.20 Schematic of the mechanism of the catalyzed solution-liquid-solid (SLS)
growth process [74]









nanocrystals is about 30nm which is narrower than those from organic materials. In

addition these inorganic nanocrystals are much stable and robust than organic molecules.

Therefore hybrid OLEDs using semiconductor nanocrystals as an emission layer have

been to have good stability and efficiency. The first demonstration of a hybrid OLED was

by Colvin et al in 1994 [92]. External quantum efficiencies approaching 1% have been

reported as summarized in table 2.3. In order to enhance the quantum efficiency of hybrid

OLED devices, several problems must be solved including more efficient charge transfer

between the organic layer and nanocrystals, the imbalance of injected charge due to poor

conduction through nanocrystals, a high density of pinhole defects in the nanocrystal

layer, uniformity of nanocrystals, a high density of pinhole defects in the nanocrystal

layer, uniformity of nanocrystals in the deposited layer, and optimization of interlayer

structure of devices [1].

Table 2.3 Properties of hybrid OLEDs using various semiconductor nanocrystals.
Nanocrystals (NC) Device structure X QE (%) Ref
CdSe ITO/PPV/NC/Mg VIS 0.001-0.01 [92]
CdSe/CdS ITO/PPV/NC/Mg/Ag VIS 0.22 [93]
CdSe/ZnS ITO/TPD/NC/Alq3/Mg-Ag/Ag VIS 0.5 [1]
CdS:Mn/ZnS ITO/PEDOT/PVK/NC/A1 VIS [94]
ITO/PEDOT/MEH-PPV-
InAs/ZnSe NIR 0.5 [95]
NC/Ca/A1
PbSe ITO/TPD/NC/Alq3/Mg-Ag/Ag NIR 0.001 [96]
PbS ITO/PPV/NC/Mg/Au NIR [97]
HgTe ITO/PEDOT/NC/Al NIR 0.02 [98]
ITO: Indium tin oxide, PPV: Poly (p-phenylenvinylidene), PVK: Polyvinylcarbazole
PEDOT: Polyethylenedioxythiophene NC: Nanocrystals
2.4.2 Hybrid Photovoltaic Devices

Hybrid photovoltaic devices consist of semiconductor nanocrystals integrated with









conjugated polymers to take advantage of the complementary properties of organic and

inorganic materials [2, 20, 99]. The nanocrystalline semiconductors generally have larger

electron affinities than do conjugated polymer, so interfacial charge separation may occur

more efficiently when electrons and holes are generated by absorption of photons.

Interfacial charge separation reduces the recombination probability of the photoexcited

electrons and holes, and thus increases the probability that the charge will migrate to the

electrodes. As a result, nanocomposites exhibit higher photoconversion efficiencies.

In 1996, N.C. Greenham et al. demonstrated the first conjugated polymer-

semiconductor nanocrystal composite photovoltaic (PV) devices using MEH-PPV and

90wt% CdSe quantum dots.[99] They showed that nanocomposite materials show higher

quantum efficiency than those for pure conjugated polymer device. W.U. Huynh et al.

reported that controlling the nanorod length in poly (3-hexyl-thiophene) (P3HT) resulted

in changes in the distance over which electrons are transported through the thin film. The











S 2 a___. Elm




Figure 2.21 External quantum efficiency of hybrid solar cell and TEM image of the CdSe
nanocrystals used in each cell[2].


external quantum efficiency was improved to 1.7 % which is one of the highest reported

power conversion efficiency for plastic solar cells [2].









The improvement in the external quantum efficiency (EQE) as the nanocrystals

length is increased is shown in Fig 2.21. As the data show, the EQE is larger for longer

nanorods versus spherical nanoparticles. However, the nanorods tend to lie in the plane of

the film, which is not optimum arrangement for charge extraction. B. Sun et. al. reported

a solar cell power efficiency of 1.8% under AM 1.5 illumination using tetrapod CdSe

nanocrystals in poly(p-phenylenevinylene) derivatives [100]. The improvement is

consistent with improved electron transport perpendicular to the substrate because

tetrapod CdSe do not lie flat within the film. Recently, McDonald et al. demonstrated

infrared photovoltaics using PbS quantum dots with MEH-PPV in which absorption can

be tunable from 800 nm to 2000nm. [101]

2.4.3 Metal-semiconductor Nanoassembly

Semiconductor/metal nanoassembled composite particles have improved the

efficiency of photoelectrochemical cells [102] and photocatalysts [103].



H3NH2CH2C-N N+--2C2CH2NH3


(b)
CH3 CH3
L L
H3C-N-CH2CH2CH2CH2-N-CH3
(a) I I
CH3 CH3

(c)

Figure 2.22 Cadmium-gold nanocomposite. (a) The schematic picture of CdS/Au
assembly bridged by (b) conducting organic spacer, N,N'-bis(2-aminoethyl)-
4,4'-bipyridinium, and (c) electrically insulating spacer, 1,4-trimethyl
ammonium butane.

The enhanced photo-conversion efficiency of semiconductor/metal nanoassemblies

is attributed to the interfacial charge separation of photo-generated electrons and holes to









the metal and semiconductor parts of the nanocomposite. Kolny et al. reported that a

nanoassembly of CdS and Au nanoparticles could be achieved by electrostatic interaction

[102]. Positively and negatively charged nanoparticles were prepared by coating their

surface with charged bifunctional ligand (e.g. 2-(diemthylamino) ethanethiol and 3-

mercaptopropionic acid) for self organization. However, the use of functionalized organic

spacers to electrostatistically assemble nanocomposites sometimes reduces the charge

separation between Au and CdS nanocrystals due to the insulating character of the

organic spacers.

Sheeney-Haj-Ichia et al reported that bipyridinium-bridged CdS/Au showed

higher photocurrents than CdS/Au that was crosslinked by 1,4-trimethyl ammonium

butane [104]. The enhanced photocurrent in the bipyridinium crosslinked systems is

attributed to the conducting property of the spacer that enables effective charge

separation. A schematic illustration of a CdS/Au assembly bridged by organic spacer is

shown in Figure 2.22 along with the chemical structure of conducting and insulating

organic spacers.

i .


1* *.





-nm


Figure 2.23 TEM images showing growth of Au onto the tips of CdSe nanorods [105].

Recently Au nanocrystals were grown on the tip of CdSe nanorods by Mokari et al.

(figure 2.23) [105]. Selective tip growth of the Au nanocrystals results from preferential






36


adsorption of the Au complex onto the rod tips. The tips of the CdSe nanorods are more

reactive because of the increased surface energy and imperfect passivation by the

surfactants. This leads to preferential growth along the [001] direction and to CdSe

nanorods. Ultimately Au nucleates on the edge of the CdSe nanorods. Gold tipped

nanostructures could provide natural metal-semiconductor contacts for electrical devices.














CHAPTER 3
SINGLE STEP GROWTH OF COLLOIDAL ZNCDSE QUANTUM DOTS

3.1 Introduction

Colloidal semiconductor nanocrystals can exhibit strong quantum confinement of

excitons and allow continuous tunability of the electronic and optical properties by

changing their size [5,8,32]. However, nanocrystals less than -2 nm may grow to final

diameters within a few seconds, making control of their size delicate and complicated. In

addition, the photoluminescent efficiency of smaller nanocrystals is typically lower than

that of larger nanocrystals due to the increased surface-to-volume ratio as the diameter

decreases [11,12]. While emission wavelength can be controlled with nanocrystal size, it

can also be controlled by composition and a few studies of colloidal ternary alloy

nanocrystals, such as ZnCdS [13], ZnCdSe [14] and CdSeTe [15,16], have been reported.

Luminescence wavelength was controlled by composition from UV to IR with stable,

larger particle size. Excellent luminescent properties were reported, that were comparable

to core/shell structured nanocrystals, which have potential for use in optoelectronic

devices and biomedical imaging.

For synthesis of colloidal ternary semiconductor nanocrystals, better crystal quality

and luminescent properties have been reported using molecular precursors [13-16] versus

controlled precipitation methods such as reverse micelle[106,107]. Recently, metal oxide

precursors have been used with functionalized organic ligands instead of organometallic

precursor for a "greener" approach to synthesis[8-10]. For example, complexes of CdO

with alkyl carboxylic acid and alkyl amine has been reported for synthesis of CdSe









nanocrystals, rather than a classical highly toxic, pyrophoric, expensive organometal

precursor (e.g. Cd(CH3)2). Moreover, metal oxide precursors are well suited for growth

studies of colloidal nanocrystals due to their slow nucleation and growth rates[8].

In this study, ternary alloy ZnCdSe nanocrystals were prepared in a single step

process using a mixture of metal carboxylates (Zn-oleate and Cd-oleate) prepared by

reactions of ZnO and CdO with oleic acid. The effects of reaction temperature on crystal

growth and alloying were determined.

3.2 Experimental Section

3.2.1 Materials

Cadmium oxide (CdO, 99.99+ %), Se powder (100 mesh, 99.999 %), oleic acid (90

%), trioctylphosphine oxide (TOPO, 99 %) and trioctylphosphine (TOP, 90 %) were

purchased from Aldrich chemicals. Zinc oxide (ZnO, 99.999 %) was purchased from

Alfa Aesar. All chemicals were used without further purification.

3.2.2 Preparation of ZnCdSe Quantum Dots

A quantity of 61 mg of ZnO (0.75 mmole) and 45.8 mg of CdO (0.25 mmole) was

dissolved in 2.6 ml of oleic acid at 350 C for 20 min in a reaction flask and cooled to

room temperature. The resultant solid mixture of Zn-oleate and Cd-oleate was loaded into

a reaction vessel containing 3 g of TOPO on a schlenk line under N2 atmosphere and

heated to 320 C with stirring. When the temperature of the reaction mixture was stable,

79 mg (1 mmol) of selenium dissolved in 1.8 ml TOP (Se-TOP) was quickly injected,

and the reaction temperature maintained at 320 C for growth. Aliquots were taken at

selected time intervals, quickly cooled and diluted with toluene to stop further growth,

and nanocrystals collected by precipitation using methanol/toluene co-solvents. These

nanocrystals were characterized using UV-Vis absorption and PL









3.2.3 ZnS Shell Growth on ZnCdSe Nanocrystals

The obtained ZnCdSe nanocrystals were loaded into a reaction vessel containing 10

g of TOPO on a Schlenk line under N2 atmosphere and heated to 180 C with stirring. For

shell growth, Zn oleate was prepared by dissolving a quantity of 41 mg of ZnO (0.5

mmole) in 2.6 ml of oleic acid at 350 C in a flask and diluted with 10ml TOP after

cooling to room temperature. Sulfur precursor (S-TOP) prepared by dissolving 24 mg

(0.75 mmol) of sulfur in 3 ml TOP was mixed with Zn oleate solution at room

temperature. The resultant mixture of Zn-oleate and S-TOP was loaded into syringe and

was injected slowly to reaction vessel containing ZnCdSe nanocrystals over 1.5 hours,

maintaining reaction temperature at 180 C for shell growth. Nanocrystals was collected

by precipitation using methanol/toluene co-solvents and characterized optical properties

using PL spectroscopy.

3.2.4 Characterization of ZnCdSe Quantum Dots

Absorption spectra were collected with a Shimadzu UV-2401PC

spectrophotometer. Photoluminescence (PL) was measured at room temperature from

nanocrystals suspended in toluene using a Fluorolog Tau 3 spectrofluorometer (Jobin

Yvon Spex instruments, S.A. Inc.). High resolution transmission electron microscopy

(HR-TEM) images were obtained using a JEOL 2010F microscope for lattice imaging

and crystal size determination. TEM samples were prepared by dispersing the

nanocrystals in toluene and depositing them onto formvar-coated copper grids. X-ray

diffraction (XRD) patterns to determine crystal structure were obtained using a Philips

APD 3720 X-ray diffractometer.


3.3 Results and Discussion




















(d) (c) (b) (a)


400 500 600 700 800
Wavelength(nm)


Figure 3.1 Temporal evolution of photoluminescence spectra at reaction temperature of
320C as a function of reaction time, (a) 2 min (648 nm), (b) 10 min (623 nm),
(c) 20 min (580nm), and (d) 30 min (567 nm) (excitation at 350 nm, doubled
peak).









3.3.1 Single Step Growth of Colloidal Ternary ZnCdSe Quantum Dots

As described in the experimental section, to synthesize ZnCdSe nanocrystals,

appropriate amounts of Zn and Cd oleate were added into the reaction vessel containing

TOPO solution. After Se-TOP was injected, the alloying process was monitored by the

blue shift of the absorption and PL spectra of the extracted nanocrystals. The wavelength

of the band edge emission peak from the ternary nanocrystals decreased from 648 to 567

nm with increasing reaction time at 320C (see Figure 3.1). This significant blue shift is

consistent with the formation of the alloyed ZnCdSe ternary nanocrystals and results

from band gaps intermediate between the wide band gap of ZnSe (2.6 eV or 477 nm) and

the narrow band gap of CdSe (1.74 eV or 713 nm). The PL emission is a single peak,

ruling out separate nucleation and growth of CdSe and ZnSe. However, note that the

single peak at 2 min is broad (FWHM = 52 nm) and asymmetric, but is much more

narrow (FWHM = 32 nm) and symmetric after 30 min. of reaction (Figure 3.1, curve a,

and d, respectively). It is believed that the emission peak is more narrow after long

reaction times because the composition of ZnCdSe is more uniform [13]. Figure 3.2

shows optical photographs of the series of the samples under UV-irradiation

corresponding to growth times and spectra in Figure 3.1. Although the diameter of

ZnCdSe nanocrystals (5.7-7.6 nm, see figure 3.8) is larger than the Bohr exciton radius

of CdSe (-5.5 nm) and ZnSe (2.2 nm), we observed a shift in the absorption spectrum to

the higher energies than bulk ZnCdSe with same composition. Figure 3.3 shows the

comparison between the band gap energy of bulk and nanocrystalline ZnxCdi-xSe alloy as

function of the Zn mole fraction (Figure 3.3). Presumably this shift is explained by two

factors; (1) the size of the particles is in the intermediate confinement regime (ae> a> ah,

where ae and ah are the Bohr radii of electron and holes, respectively) and (2) spatial
































Figure 3.2 Picture of luminescence of ZnCdSe nanocrystals dispersed in toluene under
UV irradiation. Nanocrystals were grown at reaction temperature of 320C
and different reaction time, (a) 5 min, (b) 10 min, (c) 20 min, and (d) 30 min.


0.1 0.2 0.3 0.4
Zn Composition


Figure 3.3 Dependence of the bandgap energy (Eg) as function of composition of (a)
ZnCdSe nanocrystals grown at reaction temperature of 320C and different
reaction time, 5 min, 10 min, and 30 min and (b) bulk alloy on their
compositions.









fluctuation produce atomically abrupt jumps in the chemical potential which localizes the

free exciton states.

3.3.2 Effect of Reactivity of Precursor and Reaction Temperature on Optical
Properties

The bandgap absorption wavelength changed as a function of reaction time for growth at

320C and 270C as shown in Figure 3.4. Two distinct regimes are observed; an initial

red shift regime, and alloying resulting in a blue shift regime. In the initial stage, the

reaction was almost instantaneous and the solution quickly developed to a deep red color

at 320 C under UV irradiation within 1 min. After 1 min. at 320 C, emission from the

solution gradually became yellow green. These data suggest that nucleation and growth

of Cd-rich ZnCdSe is dominant during the first minute of the reaction rather than Zn-rich

nanocrystals, leading to red shift of emission due to increment of size. Red shift in initial

growth stage was clearly seen when nanocrystals were grown at lower growth

temperature (270 C) because the red shift of emission from the solution persisted over

longer (-20 min) reaction time before starting blue shift. The conclusion that the initial

growth of nanocrystals is Cd-rich is consistent with the x-ray diffraction pattern shown in

Figure 3.5. The pattern and 20 values after 10 min. at 2700C (Figure 3.5b) is much closer

to bulk CdSe (Figure 3.5a) than bulk ZnSe (Figure 3.5d). For example, 20 values from

(001) face of bulk CdSe and nanocrystals at initial reaction stage are 25.33 and 25.66

respectively. The XRD data from nanocrystals grown for 230 min. at 2700C (Figure 3.5c)

are shifted towards bulk ZnSe. Note however that the band gap adsorption of the initial

nanocrystals (-650 nm) is blue-shifted from the band gap of bulk CdSe (-713 nm), and

the XRD peaks are at higher 20 values, both suggesting that alloying has already taken

place. The initial red shift for initial growth time may result from larger particles as




















630
E
620
O
0.
610
0 600
.0
: 590
S580
I 570
560

550
540
530


50 100 150 200
Reaction time(min)


Figure 3.4 Band gap absorption change as a function of reaction time, (a) reaction
temperature 320 C and (b) reaction temperature 270 C. Inset represents the
magnified plot of dotted circle in (a).


(d)














(a)
I I I I

20 30 40 50 60
2 Theta


Figure 3.5 X-ray diffraction patterns of (a) bulk CdSe, (b) Cd rich ZnCdSe nanocrystals
obtained grown for 10 min. at 2700C (c) ZnCdSe nanocrystals grown for 230
min. at 2700C and (d) bulk ZnSe


250









longer reaction time, reducing quantum confinement effects. Cd-rich growth in the initial

stage would suggest that the Cd oleate is more reactive than the Zn oleate towards Se-

TOP. Growth of the II-VI semiconductor nanocrystals results from the reaction of metal

precursors with the Se precursor via elimination of chemically bonded organic molecules

(figure 3.6).

(OOCR)n-m
nM(OOCR)2 + nSe-TOP (M-Se)n(OOCR)m
nTOP

R; CH3(CH2)7CH=CH(CH2)6- M; Zn, Cd etc


Figure 3.6 Schematic reaction of precursor for nanocrystal formation

If the bond strength between the metal and organic ligand is sufficiently high, the

rate limiting step is elimination of the organic ligand. The bond strength is therefore

related to the activation energy for the crystal growth. Cundari et al. calculated metal-

ligand bond energies for the study of the reaction of II-VI precursor for organometallic

chemical vapor deposition (MOCVD) (table 1) [108].

Table 3.1 Calculated metal-chalcogen bond energies in metal alkyl chalcogen precursors
with the composition of M(ER)2 (kcal/mol; from reference [108].
M/E C O S Se Te

Zn 37 71 58 50 40

Cd 36 57 50 44 36

Hg 31 44 41 35 28

M : metal, E : chalcogen R: alkyl
From data in table 3.1, metal-ligand bond energies decrease in the order Zn > Cd >

Hg for bonding to the anions reported, including oxygen and selenium. Therefore, the

activation enthalpy for nucleation of CdSe is expected to be lower than that of ZnSe,

consistent with the present data that indicate that the Cd-oleate reacts faster with Se-TOP

than Zn-Oleate. Bailey and Nie have reported, consistent with data in Table 1, that









reaction between Cd and Se and Te precursors resulted in a core/shell structure of

CdTe/CdSe in a single step reaction [15]. In addition, the data in table 1 indicate that the

bond energy of metal-oxygen is -1.5-2 times higher than that of metal-carbon, where

metal is Zn, Cd or Hg. Therefore the reactivity of metal carboxylate, such as Cd oleate, is

lower than that of precursor such as Cd(CH3)2, leading to slower growth and better

development of nanocrystals. In addition, oleic acid acts also as capping agent on the

surface of the nanocrystals, stabilizing and preventing Ostwald ripening [10].

3.3.3 Effect of Reactivity of Precursor on Particle Growth

On the other hand, it was observed that red shift of band gap absorption was faster

and developed to 655nm in initial state of reaction at high reaction temperature, whereas

absorption was slowly changed to 620nm at low reaction temperature. This indicates that

nucleation and growth is substantially faster, followed by the formation of larger crystal

at higher temperature. If almost precursor has been involved in the nucleation process in

case of finite volume of the solution, which is considered in the case of using highly

reactive Cd(CH3)2, it is expected that particle sizes are small at high temperature due to

fast nucleation [10] and depletion of precursor for further growth. Subsequent

nanocrystals grow via the Ostwald ripening reaction. However, in case of less reactive

metal carboxylate precursor such as Cd acetate, it was reported that only -10 % of Cd

precursor was consumed during the nucleation. [9, 10]Therefore, the nanocrystals

continue to grow consuming free precursors from surrounding solution and large amount

of free precursors is still present in the crude reaction solution. In this case, the growth

rate of a single nanoparticle depends strongly on reaction temperature to overcome

activation energy for further growth. The concentration of monomer in solution close to

the nanoparticle's surface significantly affects the growth rate [56]. At higher










temperature, flux of precursors at the surface of nanoparticles increases due to high

diffusivity of unreacted precursors. Growth rate increases also due to fast elimination of

chemically bonded organic ligand and high concentration of precursors, resulting in fast

growth and larger particle size. Furthermore diffusivity of precursors becomes more

important factor in case of ternary system in which two different types of precursors are

involved, because nanocrystals may be shielded by layer of less reactive Zn oleate when




















548 nm (C) 567 nm (d)










500 550 600 650 500 550 600 650
Wavelength (nm) Wavelength (nm)



Figure 3.7 Comparison of the size of ZnCdSe nanocrystals grown at different
temperatures (a) 270C for 230 min. and (b) 320C for 30 min (20 nm scale
bar). Photoluminescence spectrum of ZnCdSe nanocrystals obtained (c) at 270
C for 230 min. and (d) at 320 C for 30 min. after no further spectral shift.









Cd oleates form crystals. From the reasons described above, the size of ZnCdSe

nanocrystals obtained after no further luminescence shift is larger at 320 C (-7.6 nm)

than 270 C (-4 nm), as shown in figure 3.7.

After the initial nucleation and growth of CdSe, nanocrystals continue to grow with

reaction time. HR-TEM image in Figure 3.8 shows the crystal size and prolate shape

evolution of ZnCdSe nanocrystal at 320 C. The increase of particle size presumably is

the result of predominant consumption of Zn and lesser consumption of Cd precursors.

Therefore, it is postulated that Zn-rich layer on the outer of Cd-rich ZnCdSe nanocrystals

led to gradient-induced diffusion, consistent with the blue shift and narrowing of PL

emission and shift of x-ray diffraction peak to larger 2 theta angles with the increase of

Zn in the ZnCdSe nanocrystals








particle sizes (diameters) are (a) 5.2 nm, (b) -5.7 nm and (c) 7.6 nm with







aspect ratio 1.5.
3.3.4 Shell Growth ofZnS on ZnCdSe Quantum Dots









higher band gap semiconductor on the cormages of ZnCdSe nanocrystals suchgrown at 320C as CdSe/ZnS [6,37]. Higher
aspect ratio 1. .






3.3.4 Shell Growth of ZuS on ZnCdSe Quantum Dots
Figher band gap semiconductormages of ZnCdSore nanocrystals sucgrown as CdSe/Zn320C as reaction time






luminescence QY can be achieved by inorganic shell growth due to reduce nonradiative

decay channels that originate from surface-trap electronic states. Not only effective











1.4x107


S6.0x106


4.0x106
-ZnCdSe(#52) QD
2. 6 (a) -ZnCdSe/ZnS core/shell
2.0x106

0.0
450 500 550 600 650 700
Wavelength(nm)



Figure 3.8 Photoluminescence spectra (a) before ZnS shell growth and (b) after ZnS shell
growth on ZnCdSe quantum dots

surface passivation, higher band gap shell materials provide also a potential step for

electron and holes originating in the core nanocrystals, reducing the probability for the

carriers to sample the surface. Figure 3.8 shows that luminescence intensity increase

significantly when ZnS shell was grown on ZnCdSe nanocrystals. In addition, when ZnS

surrounds the ZnCdSe nanocrystal, it is observed -21 nm energy shift in emission to

lower energy than that of ZnCdSe nanocrystals. The wave function of lighter electron

tunnels into shell [6]. The increased delocalization of the electron lowers its confinement

energy and consequently the energy of luminescence [6].

3.4 Conclusion

Colloidal ternary alloy ZnCdSe nanocrystals have been synthesized by a "single









step" reaction in a mixture of Cd and Zn oleates with Se-TOP. After an initial red-shift

due to an increase of size of Cd-rich ZnCdSe nanocrystals, a blue shift of PL was

observed upon diffusion of Zn into the nanocrystals to form a larger band gap ternary

alloy. The initial nucleation of Cd-rich ZnCdSe nanocrystals was attributed to the higher

reactivity of Cd oleate as compared to the Zn oleate. The initial red shift of band gap

absorption was attributed to growth of the nanocrystal size, while the subsequent blue

shift was attributed to diffusion of Zn from the outer layer into the Cd-rich core, to form

the final ZnCdSe particles between 5 and 8 nm in diameter after reacting for 2-30 min. at

320C. Fast growth and larger particle size was observed at higher reaction temperature

due to fast elimination of chemically bonded organic ligand and high concentration of

precursors. Luminescence intensity increase significantly when ZnS shell was grown on

ZnCdSe nanocrystals. In addition, when ZnS surrounds the ZnCdSe nanocrystal, it is

observed -21 nm energy shift in emission to lower energy than that of ZnCdSe

nanocrystals.














CHAPTER 4
SYNTHESIS AND CHARACTERIZATION OF COLLOIDAL TERNAY ZNCDSE
SEMICONDUCTOR NANORODS

4.1 Introduction

Colloidal semiconductor nanocrystals have been intensively studied due to the

excellent luminescent properties, size tunable optical properties, and their high diversity

for a variety of optoelectronic application such as light emitting diodes (LEDs) [1] and

photovoltaics [2,20]. Furthermore, colloidal nanocrystals have been considered to be the

building block of future nanotechnology for the fabrication and investigation of quantum

superstructures [3]. Recently, development of synthetic methods for growth of rod shaped

CdSe nanocrystals promise new opportunities to study shape-dependent electronic and

optical properties such as polarized LED [17, 18], lower threshold laser [19], and more

efficient photovoltaic device [2]. However, rod shaped CdSe nanocrystals still are

reported to have low quantum efficiencies [17, 21, 22] and weak confinement along the

rod axis leading to difficulties in efficient production of blue-green light.

One approach to improve the emission efficiency is to coat the surface of the CdSe

nanorods with a larger bandgap shell material to confine the charge carriers and minimize

the nonradiative decay channels resulting from electronic surface states. The luminescent

quantum yield of as-grown CdSe nanorods is reported to be -1% at room temperature,

and increases to -20 % when a shell of a larger band gap material (CdS or ZnS) was

grown on the surface of the core [21,22]. On the other hand, Zhong et al. reported that

ternary ZnCdSe quantum dots show high photoluminescence (PL) efficiency comparable









with those from CdSe/ZnS core shell quantum dot [13,14].

A solid solution of two binary semiconductors may exhibit intermediate energy

band gaps between those of the constituent binaries, making them attractive materials for

opto-electronic devices. A variety of II-VI ternary alloy semiconductors, such as ZnCdSe,

with a range of compositions and therefore band gaps, have been grown by MBE and

their emitted wavelength shown to cover the whole visible range. They have been studied

for LEDs and laser diodes (LDs) for use in optical DVDs and displays [67-70, 109].

Colloidal ZnCdSe nanorods could also be used in optoelectronic devices since

quantum confinement may allow efficient emission at wavelengths that cover the visible

spectrum. In this study, green-yellow emitting ZnCdSe nanorods were prepared by

diffusion of Zn into the CdSe core. The ZnCdSe nanorods were characterized by

absorption spectroscopy, photoluminescence, X-ray diffraction (XRD), high resolution-

TEM (HR-TEM) and Raman spectroscopy. Raman spectroscopy is an excellent probe of

the nanostructure [110-115]. The Raman data will be shown to complement the XRD and

HR-TEM data on the ZnCdSe nanostructure. In addition, in order to elucidate the energy

relaxation and recombination dynamics in nanoparticles, we present here a

characterization of the luminescence decay and dynamics of CdSe/ZnSe coreshell

nanorods and ZnCdSe nanorods using time-resolved photoluminescence (TRPL) and

femtosecond transient absorption (TRA)

4.2 Experimental Section

4.2.1 Materials

Cadmium oxide (CdO, 99.99+ %), selenium powder (100 mesh, 99.999 %), oleic

acid (90 %), trioctylphosphine oxide (TOPO, 99 %) and trioctylphosphine (TOP, 90 %)

were purchased from Aldrich chemicals. Tetradecylphosphonic acid (TDPA) was









purchased from PolyCarbon Industries, Inc. and zinc oxide (ZnO, 99.999 %) was

purchased from Alfa Aesar. All chemicals were used without further purification.

4.2.2 Preparation of ZnCdSe Nanorods

CdSe nanorods were synthesized using the method described by Peng [85]. In this

method, 0.205g (1.6 mmol) of CdO, 2.903 g of TOPO and 0.8928 g of TDPA were

heated in a three-neck flask on a Schlenk line under a N2 atmosphere to 350 C while

stirring. After the solution became optically clear, it was cooled to room temperatures.

The solid Cd-TDPA complex was used after aging for 24 hr without further purification.

This Cd-TDPA complex was heated in a three-necked flask under a N2 atmosphere to 280

C while stirring, and 0.126 g (1.6 mmol) of selenium dissolved in 2.9 ml TOP was

injected quickly. After injection, the temperature of the mixture was kept at 250 oC for

the 30 min growth of CdSe nanorods, and then cooled to 180 oC.

For shell growth, 0.1302g (1.6mmol) of the ZnO was dissolved in 2.03 ml of oleic

acid (Zn-oleate) at 350 oC and cooled to room temperature, and then lml of TOP was

added to prevent solidification. In addition, 0.126 g (1.6 mmol) of Se was dissolved in 2.9

ml of TOP (Se-TOP). The Zn-oleate and Se-TOP solutions were mixed by stirring for

10min at RT, and this mixture was loaded into a syringe and injected drop-by-drop into

the reaction flask over 1.5 hr. After injection was complete, the solution was stirred at RT

for another 10min. For alloying, the reaction vessel was heated with stirring to 2700C for

up to 3 hrs. After heating for 1, 2 or 3 hrs, a sample was immediately cooled and diluted

with toluene to stop alloying, then was precipitated with methanol/toluene co-solvents.

4.2.3 Characterization

High-resolution transmission electron microscope (HR-TEM) images were

collected using a JEOL 2010F microscope for imaging and direct determination of the









average and distribution of the nanorod dimensions. To prepare TEM samples, the

nanocrystals were dispersed in toluene and deposited onto formvar-coated copper grids.

X-ray diffraction (XRD) patterns were obtained using a Philips APD 3720 X-ray

diffractometer and used for determination of both the crystal structure and size. Raman

spectra were measured at 300K in the backscattering geometry, using the 532 nm line

from a Verdi 8 doubled Nd-YAG solid state laser in a Ramanor U-1000 Jobin-Yvon

Raman spectrometer.

Absorption spectra were collected with a Shimadzu UV-2401PC

spectrophotometer. Photoluminescence (PL) was measured at room temperature using

nanorods suspended in toluene using a Fluorolog Tau 3 spectrofluorometer (Jobin Yvon

Spex instruments, S.A. Inc.). The PL quantum yield (QY) was determined using

Rodamine 6G organic dye standard with a known QY of 95%.

Time resolved spectra were recorded using a spectrograph attached to a charge-

coupled device (CCD) (Shamrock 303i). A commercial Ti-Sapphire (Ti-Sa) laser system

consisting of a Ti-Sa oscillator (Tsunami, Spectra-Physics) and subsequent amplifier

(Spitfire, Spectra-Physics) with a repetition rate of 1 kHz was used to pump and optical

parametric amplifier (OPA) to generate 400 nm excitation pulses.

The relaxation processes of the colloidal nanocrystals were explored using

femtosecond transient absorption (TRA) based on the same laser system described above.

A part of the amplifier output is split off to pump with a 1 mm CaF2 window to generate

a white light continuum probe that ranges from 300 to 900 nm. Prior to white-light

generation, the probe polarization is tilted by 45 degrees with respect to the pump pulse

using a thin-film polarizer. Specifically, the fourth harmonic of the OPA idler output is









used to produce excitation pulses (pump) at 450 nm. This beam is then fed through a

prism compressor, resulting in pulse lengths less than 100 fs (FWHM). The diameter of

the pump was -100 |jm and was set to low energy (- 39 to 45 nJ) to avoid biexciton

formation. Prior to interaction in the sample, a fraction of the probe pulse is split off and

measured as a reference beam. The pump pulses were modulated by an optical chopper at

a frequency of 500 Hz and passed through a computer-controlled optical delay line to

delay the probe relative to the pump. The pump and probe beams were temporally and

spatially overlapped into the sample. The optical chopper blocks every other pump pulse.

The transmitted signal, T, is the probe pulse in the presence of the pump, and To is the

transmission of the probe pulse in absence of the pump. The pump induced absorption

changes to be detected are:

AT/T = T-To/To (4-1)

A Glan-Thompson polarizer splits the transmitted signal, with and without the

pump pulse, into its polarization components, parallel (A) and perpendicular (AJ) with

respect to the pump. TA experiments performed at the Magic angle eliminate directional

and rotational influences on the signal. Therefore, the Magic angle signal (TRA) is

calculated from the parallel and perpendicular components measured simultaneously:


TA = (A 2A)/3 (4-2)

The parallel and perpendicular transmitted signals and reference were focused into

a spectrograph attached to a charge-coupled device (CCD) (Shamrock 303i) for detection.

Sample solutions for TRA measurements were placed in a quartz cuvette with a 2 mm

path length and continuously stirred to guarantee excitation of a new sample volume with

every laser shot. The set up for TRPL is same as TRA except rather than facing the probe









beam on the camera, we collect the fluorescence from the sample using a 2 inch lens and

the focusing this into the camera. We used a 4 nanosecond gate and we excited at 400 nm

which is the second harmonic of the Spitfire output.

4.3 Results and Discussion

4.3.1 Synthesis of ZnCdSe Nanorods

The combination of TDPA/TOPO surfactants has been used to prepare anisotropic

nanocrystals because it was expected to raise the surface energy of some crystal facets

relative to other facets of the nanocrystals because of their strong binding energies with

metal ion and their steric effects [84, 116]. Initial attempts to synthesize ZnCdSe

nanorods using a mixture of Zn-TDPA and Cd-TDPA as metal sources in TOPO solution

were unsuccessful. This failure may have resulted from the fact that these two metal

sources did not lead to crystallite shape control because of their different reactivity with

Se-TOP [38,117]. Furthermore, higher temperatures and much longer reaction times were

required for complexation of ZnO versus CdO with TDPA. To avoid this problem in the

preparation of ternary alloy ZnCdSe nanorods, CdSe nanorods were first synthesized

using TDPA/TOPO surfactants. Following growth of the CdSe nanorod core, the ZnSe

shell was grown from the Zn-Oleate and Se-TOP mixture. Then the temperature was

raised to 270 C to allow diffusion and alloyed ZnCdSe nanorods to form. Steady, slow

addition of the Zn-oleate and Se-TOP mixture was required to avoid homogeneous

nucleation of ZnSe during shell growth. In addition, careful temperature control was

required because alloying led to a blue shift of emission when growth of the shell was

attempted at a temperature >210 C. However at a temperature of <170 C, the Zn-oleate

complex was too slow to react with TOP-Se and grow a shell, resulting in very weak

emission. It was important to slowly increase the temperature to 270 C for diffusion in






57










(002) (110) (103) (112)

(100)
(101)
S 0 (102) (c)



O
(b)




(a)

20 30 40 50 60
2Theta



Figure 4.1 Powder X-ray diffraction patterns of (a) CdSe nanorods,(b) CdSe/ZnSe
core/shell nanorods, and (c) ZnCdSe nanorods.









order to avoid temperature over-shoot.

4.3.2 Structure of ZnCdSe Nanorods

The powder X-ray diffraction patterns from hexagonal CdSe, CdSe/ZnSe, and ZnCdSe

nanorods are shown in Figure 1. For all three materials, the width of the (002) diffraction

peaks is small compared to those from the (100) plane, consistent with less broadening of

the diffraction peak from planes perpendicular to the long axis of the rod-shaped

nanocrystals. The lattice spacing along the c-axis was 7.01 A, 6.94A and 6.77A for CdSe,

CdSe/ZnSe, and ZnCdSe, respectively.

The c-axis lattice parameter reported above for the CdSe/ZnSe core/shell structure is -1%

smaller than that for CdSe alone. It has been reported that the CdSe core is under

compression due to the growth of a ZnS shell with an 11% smaller lattice parameter,

leading to smaller d-spacing and larger 20 values [21]. Similarly, the CdSe core should

be compressed by the ZnSe shell due to its smaller lattice parameter, but in this case the

lattice mismatch is smaller at -7 %. Furthermore, alloying between CdSe and ZnSe is

expected to reduce both the lattice parameter and any strain induced by lattice mismatch.

The smaller lattice parameter reported above for ZnCdSe alloyed nanorods must result

predominantly from the lattice contraction upon interdiffusion, not from strain relaxation

that would increase the lattice parameter. The XRD data in Figure 4.1 show that all

diffraction peaks shifted to a larger 20 (smaller interplanar spacing) upon alloying of

CdSe with Zn, consistent with expectations.

The XRD data from spherical versus rod-like nanocrystals of ZnCdSe are

compared in Figure 4.2. While diffraction peaks from the same planes are observed, the

intensity of diffraction from the (102) and (103) planes is reduced compared to the (110)


































20 30 40 50 60
20 30 40 50 60


2Theta



Figure 4.2 Comparison of powder X-ray diffraction patterns of (a) ZnCdSe nanorods, and
(b) spherical ZnCdSe dots.


0 -
SDiameter

SLength


4Oii


20 nm scale bar


5 nm scale bar


Figure 4.3 HR-TEM image and histogram of size distribution of ZnCdSe nanorods.
Lattice fringe from a nanorod is shown by the lower right inset.


nm









and (112) planes of ZnCdSe nanorods. Attenuation of the (102) and (103) diffraction

peaks probably results from a higher density of stacking faults along the (002) axis of

spherical versus rod nanocrystals [5,12,118]. A HR-TEM image of the ZnCdSe nanorods

is shown in Figure 4.3, along with a histogram of the diameter and lengths of the

nanorods measured from such images. The average diameter is -6 nm and the average

length is -13 nm. The particle sizes of ZnCdSe nanorods calculated from XRD data using

the Debye-Scherrer equation were a diameter of 5.5 nm and a length of 11.8 nm, which

agree well with the HR-TEM data shown in Figure 4.3.

4.3.2 Effect of Alloying on the Phonon Spectra

Raman spectroscopy has been used to study the structure of nanocrystals, including

the core/shell interface, through the dependence of the phonon frequencies upon

compositional homogeneity [113,114].

The Raman peaks detected from CdSe nanorods are shown in Figure 4.4(a). The

peak at -206cm-1 is from the LO phonon, and is -4cm-1 lower in wave number than that

reported for corresponding bulk CdSe (210cm-1) [112,114], presumably due to the

confinement of the optical phonons in the nanorods [112-114]. Surface phonon vibration,

which is observed as the "shoulder" to the left of the main Raman peak (at -180cm-1) is

attributed to the surface phonon for CdSe, that is detected because of the non-spherical

geometry of the CdSe nanorods [110,111].

Similar data from as-grown CdSe/ZnSe core/shell nanorods are shown in Figure

4.4(b). In addition to the 'bulk' LO phonon from the CdSe core (-206 cm-1) and the ZnSe

shell (-247 cm-1), limited formation of interfacial ZnCdSe is indicated by the Raman

peak at -235 cm-1. Unresolved Raman peaks (shoulders) on the both side of the CdSe LO
































Raman shift(cm-1)


175 200 225 250 275


Raman shift(cm-1)




Figure 4.4 Raman spectra of LO phonon mode of (a) CdSe nanorods and (b) CdSe/ZnSe
core-shell nanorods.







62



















I-
(c)






(b)





(a)

I I I I I
180 200 220 240 260 280
Raman shift (cm1)



Figure 4.5 Raman LO phonon spectra of ZnCdSe nanorods after annealing at 270 C for
(a) 1, (b) 2, or (c)3 hrs.









peak (206 cm-) are attributed to the isolated atom-impurity modes of Zn in CdSe (190

cm-1) and Cd in ZnSe (218 cm-1) [119].

The effects of alloying time (1, 2 or 3 hrs at 270 C) on the Raman spectra are shown in

Figure 4.5. After heat treatment, a single phonon mode is detected for the ZnCdSe

nanorods at -223 cm-1, -228 cm-1 and -226 cm-1 (for 1, 2 and 3 hrs, respectively), similar

to the one phonon-mode behavior for bulk ZnCdSe [120,121]. The single phonon mode

suggests that the interface between CdSe and ZnSe has disappeared, the isolated atom-

impurity modes have disappeared, and the reasonably sharp single-mode peak is

consistent with a uniform composition and particle size distribution. Note however that

the Raman peak for ZnCdSe annealed for one hour is considerably broader than those

from the samples annealed for 2 and 3 hrs due to compositional disorder. The differences

between the single mode peak positions at 1 versus 2 hours (-5 cm-) results from a

continuation of the alloying process. The difference in peak position between 2 and 3 hrs

(-2 cm-1) is attributed to compositional disorder[122,123] and stress relaxation by

thermal annealing[115].

4.3.3 Photoluminescence and Absorption Properties

The PL spectra from CdSe core and CdSe/ZnSe core/shell nanorods, with peaks at

642 nm and 638 nm respectively, are shown in Figure 4.6. The PL quantum yield (QY)

for CdSe core versus CdSe/ZnSe core/shell nanorods was 0.6% versus 15%. The

increased PL QY was presumably due to passivation of non-radiative surface states on

the CdSe nanorods by the ZnSe shell. The 4 nm blue shift of PL from CdSe/ZnSe

core/shell versus CdSe core nanorods could result from electron localization or nanorod

size effects in the core/shell structures. Contrary to the current blue shift, Mokari and














6x10 -


5x106-


4x106-


3x106-


2x106-


1x106-


0-


I I I I I I I I
450 500 550 600 650 700 750 800

Wavelength(nm)


Figure 4.6 Photoluminescence spectra of (a) CdSe-ZnSe core/shell nanorods and (b)
CdSe nanorods.


0.25-



0.20-



0.15-


CO
- 0.10-
0


0.05-



0.00-


300 400 500

Wavelength


600 700


Figure 4.7 UV-Vis absorption spectra of (a) CdSe nanorods, (b) CdSe/ZnSe core-shell
nanorods, and (c) ZnCdSe nanorods alloyed at 270C for 3hrs.


m m m m









Banin [21] reported a -10nm red shift for a CdSe/ZnS core/shell structure, and attributed

this shift to tunneling of the electron wave function into the ZnS shell. This tunneling led

to a delocalization of the electron, lowering its confinement energy and consequently the

energy of the exciton levels[6]. Increased localization would be expected for

compressively stressed core/shell particles as suggested above based on XRD data,

resulting in a blue shift. In addition, the Raman data above suggest the formation of

interfacial ZnCdSe in as-grown CdSe/ZnSe core/shell nanorods. This reaction would be

expected to decrease the size of the CdSe core[124] resulting in increased localization

and a blue shift in emission. Finally, as reported below, alloying results in a blue shift of

the PL peak. Presumably the observed blue shift from formation of the core/shell

structure results from a summation of these effects. There were also significant

differences in the absorption spectra for CdSe core, CdSe/ZnSe core/shell, and ZnCdSe

alloyed nanorods, as shown in Figure 4.7. For CdSe core and CdSe/ZnSe core/shell

nanorods, the initial absorption edge was at -650 nm and -645 nm, respectively, in

agreement with the emission peaks in Figure 4.6. This absorption peak is thought to result

from interband optical transitions such as IS [1S(e)-iS1/2(h)], 2S [1S(e)-2S3/2(h)] and 1P

[1P(e)-1P3/2(h)], where e and h represent electron and hole quantized states [32]. The

second absorption structure in Figure 4.7 for CdSe and CdSe/ZnSe nanorods (-520 nm)

is thought to represent IP transition. The energies of the corresponding absorption

features for alloyed ZnCdSe (3 hrs at 270 C) is blue shifted considerably to -555 nm and

-465 nm, consistent with PL emission data reported in Figure 4.8. Presumably these

features originate from transition similar to those in the core and core/shell nanorods,

reflecting the larger band gap resulting from the formation of ZnCdSe. While the









adsorption features from nanorods shown in Figure 4.7 are easily detected, they are not as

sharp and as well resolved as features reported for CdSe quantum dots [32]. Presumably

this results from the fact that a distribution of energy levels will result in the conduction

and valence bands due to the increased degree of freedom for nanorods as compared with

quantum dots [33,125]. In addition, the compositional disorder detected by Raman data

will lead to broader features in the absorption and emission spectra. Absorption feature

described above was studied in detail using time resolved absorption spectroscopy (see

section 4.3.5)

The blue shift in absorption upon alloying to form ZnCdSe is consistent with the

PL emission spectra shown in Figure 4.8.Upon annealing at 270 C, the PL peak from

CdSe/ZnSe core/shell nanorods (638 nm) was shifted to 610, 570 and finally to 565 nm

with the alloying times of 1, 2 or 3 hrs, respectively. Note that the PL peak after 1 hr of

annealing is not only shifted, but is skewed to the short wavelength side, consistent with a

range of compositions as indicated by the breadth of the Raman peak. The peak after 2

hrs of annealing is still broad as compared to the breadth after 3 hrs of annealing. The PL

peak after 3 hrs of annealing is also narrower than the peak from the CdSe/ZnSe

core/shell nanorods, although it is less intense. The PL QYs of the ZnCdSe nanorods (-8,

5 and 10 % for 1, 2 and 3 hrs anneals, respectively) were lower than the QY of 15 %

reported above for CdSe/ZnSe core/shell structures and higher than that for CdSe

nanorods (0.6 %). The increased QY for ZnCdSe versus CdSe probably results from the

compositional disorder discussed above. Composition disorder in ternary alloy nanorod

structures will lead to localization of excitons [126]. Such localization effects apparently

improve the luminescence efficiency by increasing the overlap integral of the electron






67







6x106

(a)
5x106
(d)

4x106


3x106 (
C

2X106
(c)

1x106




400 450 500 550 600 650 700 750 800

Wavelength(nm)



Figure 4.8 Photoluminescence spectra from (a) CdSe/ZnSe core/shell nanorods and
ZnCdSe nanorods alloyed at 2700C for (a) 1, (b) 2, and (c) 3 hrs.









and hole wavefunctions. The decreased QY from ZnCdSe versus the ZnSe/CdSe

core/shell nanorods probably results from the lack of surface passivation on the ZnCdSe

nanorods. This is consistent with the QY decreasing after 1 and further after 2 hrs of

annealing, since diffusion will be reducing the gradient in Zn (i.e. reducing the high

concentration at the surface and increasing the low concentration in the middle of the

nanorods). However, annealing for 3 hrs increased the QY over that from samples

annealed for 1 or 2 hrs. This increased QY is attributed to annealing of crystalline defects

and reduction of stress, consistent with the Raman data reported above. This is also

consistent with the fact that the PL peak from ZnCdSe annealed for 3 hrs at 2700C has the

narrowest FWHM as shown in Figure 4.8. Crystal defects are known to act as

nonradiative recombination centers, reducing the emission efficiency [13,127].

4.3.4 Time Resolved Photoluminescence (TRPL) Study

The TRPL measurements of the nanorods samples were carried out at room

temperature. Figure 4.9 shows PL decay curves of the nanorods samples. At early times

the PL probes particles with fast decaying times while at latter times it probes those with

longer rates. It is found that the decay process is characterized by a non-exponential

function at all nanorods samples, and this non-exponential decay can be well

characterized by a stretched exponential function for disordered low dimensional

semiconductors[128]:


(t)= Io exp- (4-3)


where I(t) is the PL intensity at time t and Io is I(0); 0 is a dispersion factor and T is

emission decay time. Several relaxation phenomena in complex condensed matter

systems have been found to follow the stretched exponential decay law [128].








69
























0
0)



U,)






CdSe/ZnSe Core/Shell nanorods
-ZnCdSe alloy nanorods 1hr
ZnCdSe alloy nanorods 2hr
-ZnCdSe alloy nanorods 3hr



-20 0 20 40 60 80 100 120 140 160 180

Time (ns)




Figure 4.9 TRPL decay curve of CdSe/ZnSe nanorod and ZnCdSe nanorods








70














0- (a) 0- (b)



-1- -1-



-2 -2-
S2 T=300K T=300K
-J,
p= 0.75 p= 0.58
-3. T= 173ns -3. / = 277ns


| CdSeZnSe Coreshell | ZnCd
4 -4
0 1 2 3 4 5 0 1 2 3 4 5
Ln(Time(ns)) Ln(Time(ns))
0 00-

(c) (d)

-1, -10

-1 5-

-2- -20-
T=300K
T=300K
T0 -2O- P= 0.58



-35
L ZnCdSe alloy 2hr L ZnCdSealoy3hr
-4 -40 =
0 1 2 3 4 5 00 05 10 15 20 25 30 3.5 40 45
Ln(Time(ns)) Ln(Time(ns))




Figure 4.10 The dot lines are experimental data and the full lines are the fitting data using
equation ln[ln(Io/It)] versus In(time) of (a) CdSe/ZnSe coreshell nanorods, (b)
ZnCdSe alloy nanorods lhr, (c) ZnCdSe alloy nanorods 2hr, and (d) ZnCdSe
alloy nanorods 3hr.









The parameter P and T depend on the material and the specific phenomenon under

consideration and can be a function of external variables such as temperature. For the

limiting case of 3- 1, we get the single exponential decay with the characteristic life time

T. For single QD, we can expects P=1. It should be mentioned that P<1 result from

superposition of many exponential decays. This decay law is often encountered in the

disordered systems and considered as a consequence of the dispersive diffusion of the

photoexcited carriers[128-131]. In general, carrier dispersive diffusion among different

spatial regions can be due to energetic disorder, or due to the topological disorder. The

alloying process is not uniform and leads to broad PL spectra. Emission of an

inhomogeneous population (different sizes, shapes or composition) leads to the

simultaneous probing of particles with different decaying rates. A stretched exponential

function can be used to describe theses systems. Figure 4.10 shows a replotting of the

data for the nanorod samples in the form of an ln[ln(Io/It)] versus In(time) plot and fitting

Table 4.1 Comparison of P and T value of CdSe/ZnSe coreshell and ZnCdSe alloyed
nanorods
Nanorods Xem(nm) 3 T (ns)

CdSe/ZnSe core
shell 645 0.75 173

ZnCdSe lhr 625 0.58 277

ZnCdSe 2hr 570 0.48 501

ZnCdSe 3hr 566 0.58 276


using linear function. 3 value can be independently determined from T by plotting the

double logarithm of the signal versus the logarithm of the time. Obtained values are

summarized in table 4-1. The fitted 0 of CdSe/ZnSe coreshell nanorod is -0.75, which









reflects higher degree of ordered crystals. Difference of 0 value between 1 and -0.75

might be mainly due to size distribution. Interestingly, this value is matched to reported

value of 0 in quantum well structure of CdSe-ZnSe superlattice (3= -0.75) in quantum

well structure of CdSe-ZnSe superlattice (3= -0.75) [129]. Comparing CdSe/ZnSe

coreshell nanorods to alloy ZnCdSe nanorods, the 3 values are significantly decreased

from 0.75 to 0.48-0.58. If we assume that size distribution is not significantly changed

during alloying process, it is clearly seen that decreasing 0 results from compositional

disorder in nanorod crystals such as spatial fluctuations of the local Zn concentration.

However, annealing for 3 hrs increased the 0 value over that from samples annealed for 1

or 2 hrs. This increased 0 is attributed to compositional homogeneity of the sample which

is consistent with photoluminescence and Raman data reported in previous section (see

4.3.2 and 4.3.3 section). In addition, T of the samples increases with alloying time from

173 ns to 276-501ns (table 4-1). This behavior is consistent with previous theoretical

arguments predicting that the radiative lifetime of bound excitons increases with binding

energy [127]. Exciton binding energy is increased by exciton confinement which is

obtained by size reduction or localization of carrier wave function by composition

variation [132]. Therefore the binding energies of exciton in alloy nanorods presented

here can be expected to increase due to increased localization of exciton by

compositional fluctuation, leading to increase luminescence decay time (z). Kim et al

observed also increment of the PL decay lifetime with Al content in case of AlGaN alloys

due to the exciton localization [127].









4.3.5 Transient Absorption Spectroscopy (TRA) Study

Additional insight can be gained by investigating the ultrafast changes observed in

the absorption spectra[32,33]. Femtosecond TRA was utilized to study the evolution of

the absorption bleaching in the nanorods. Changes in population density of different

energy levels was examined using femtosecond time resolved pump probe spectroscopy.

A pump pulse excites the sample, which causes a depopulation of the ground state to an

excited state and a probe pulse arrives at different delay times to monitor the populations

of various energy states. The signals collected are changes in transmission (eq. 4-1). We

have excited all samples with low energy excitation (35 to 49 nJ) at 450 nm and

monitored the bleach signal at wavelengths between 450 and 725 nm. Although we

cannot make any definitive statements about the dynamics at this point, we are able to

complement the steady state absorption data and to show the general trends that occur

during alloying. Figure 4.11 compares the bleach spectrum of the core/shell and ZnCdSe

alloyed samples using the same delay times of 0, 50, 100, 200, 400, 800 and 2470 fs.

Two bleach transitions are observed, denoted as IS and IP. The bleach of IP transition

grows instantaneously (<150 fs) while the bleach of the IS transition grows on a 2 ps

timescale due to relaxation of hot electron from IP to IS level.

It is observed that band width of IS transition increases during alloying process due

to compositional disorder of ZnCdSe nanorods. In addition in early time bleach spectra of

ZnCdSe nanorods (200-400fs), it is clear that ZnCdSe nanorods have more absorption

band than CdSe/ZnSe coreshell nanorods. It is speculated that energy level may split by

intrinsic electric field induced by compositional fluctuation within single ZnCdSe

nanorods. In figure 4.12, we summarized band structures of the nanorod samples. As

shown in figure 4.12(a), the band gap energies rise rapidly from 1.95 eV in the core/shell









nanorods but plateau to 2.2 eV as shown steady state PL and UV-Vis absorption

data(figure 4.7 and figure 4.8). Figure 4.12(b) shows that IS bandwidth increases from

the core/shell to ZnCdSe alloyed for lhrs, and then narrows to approximately the original


I(nm) 74nm)


Figure 4.11 Ultrafast carrier relaxations in (a) CdSe/ZnSe core shell nanorods (b)
ZnCdSe alloy Ihr (c) ZnCdSe alloy 2hr (d) ZnCdSe alloy 3hr












26, ,. Q30 a5,-
(a) 0 ;3 (b) 048 (c)
-21 24- 0 o46-.
23. 022



.f (14-12
010-
o B-I 036.

OB& AlhylTr Al2llTr lA3GT Qh l AB Vlf T AlV21T AIV31T Q1 AICa1 R- AlVo/2R Ally3H



Figure 4.12 Summary of absoption study. (a) Energy separations between IS and 1P
transition, (b) 1S bandwidth change and (c) bandgap change of CdSe/ZnSe
coreshell and ZnCdSe nanorods

width after 3hrs. This trend is consistent with the luminescence inhomogeneous

broadening shown in figure 4.8 due to compositional disorder and annealing effect.

Finally, we measure the energy separation between the IS and IP (Figure 4.12(c)). These

values do not change significantly and the average spacing is 0.4075+0.012 eV. As

shown in figure 2.2, energy separation between IS and IP is significantly dependent on

size of nanocrystals [31]. So we can expect that size change may be negligible during

alloying process. Overall, data indicate the occurrence of a transformation of the band

gap, band structure as function of alloying time in each of the rods studied. This is

evident in both the steady state and time-resolved data presented. Further work is

necessary to elucidate the dynamics of the processes discussed above.

4.4 Conclusions

Green-yellow emitting ZnCdSe ternary alloy nanorods (6 nm x 13 nm) with

relatively high quantum yields (QY = 10 %) were synthesized by alloying ZnSe/CdSe

core/shell nanorods at 270C for times up to 3 hrs. The nanorods were characterized by









X-ray diffraction (XRD), transmission electron microscopy (TEM), and Raman

spectroscopy, as well as optical absorption and emission. The XRD and TEM were used

to quantify the size and shape of the nanorods, as reported above. The Raman data were

used to detect limited alloying in as-synthesized core/shell nanorods, and composition

disorder in the alloyed material. The QY of ZnCdSe nanorods was a function of

annealing time, was greater than pristine CdSe nanorods (QY = 0.6 %), but was lower

than that from the core/shell nanorods (15 %). The luminescent efficiency of these

materials was discussed in terms of compositional disorder, defects induced by the

alloying process, and surface passivation by larger band gap surface layers resulting from

higher Zn concentrations near the surface.

Time resolved emission provided information regarding the role of diffused Zn. A

stretched exponential function was used to describe these systems, where 3 < 1

corresponds to disperse populations. Comparing CdSe/ZnSe coreshell nanorods to alloy

ZnCdSe nanorods we found a significant decrease in the 3 value (from -0.75 to

0.48-0.58). Luminescence decay life time c of the samples increased with alloying time

from 173 ns to 276-501ns. This was explained by compositional disorder and exciton

localization. Additional insight was gained by the evolution of the absorption bleaching

in the nanorods using femtosecond TRA. After excitation at 450 nm, band structures of

the nanorod samples were determined.














CHAPTER 5
SELF ASSEMBLED GROWTH OF GOLD NANOCRYSTALS ON CADMIUM
SULFIDE NANORODS

5.1 Introduction

Nanoscale charge transfer [133] is an important issue for both the fundamental

properties and applications of molecular electronics in displays, photovoltaics, sensors,

and photocatalyst. For photovoltaics and photocatalyst, efficient charge separation to

reduce the recombination of the photogenerated electron-hole pairs is essential.

Assemblies of noble metals with semiconductor nanoparticles, such as Au/TiO2, have

been studied because of their ability to separate charge [134]. The basic role of the noble

metal nanoparticle is to shuttle photogenerated electrons from semiconductor

nanocrystals.

Simple self organized mixtures of Au nanoparticles and spherical CdS nanocrystals

have been reported previously to exhibit increased photocurrent generation and

photoelectrochemical properties [102,104,135,136]. In these mixtures, appropriate

organic spacers have been used for electrostatic self assembly and surface modification of

nanocrystals. However, such organic spacers may bridge nanoparticles by electrostatic

attraction without promoting charge separation due to their insulating properties [104]. It

seems reasonable that direct contact of Au nanocrystals with CdS nanocrystals would be

more effective for charge separation and improved device performance.

In addition to nanospheres, nanorods of CdSe have been studied for photovoltaic

devices because of their potential for better charge transfer to the solar cell electrodes









[2,78], and for laser [137] and thin film transistor [138]. Various methods have been

shown to result in one-dimensional (1-D) nanocrystals such as nanorods [71]. Solution

phase growth using capping reagents may be used to produce large quantities at a low

cost, and involve simple procedures. Moreover, solution phase synthesis is appropriate

for in situ preparation of nanocomposite from two different materials.

In this study, CdS nanorods were prepared using solution phase growth and Au

nanocrystal could also be grown directly on S2- rich surfaces due to strong Au-S

bonding. Photoluminescence (PL) and photo-catalytic properties of Au/CdS nanorod

composites are reported, which are attributed to more effective charge separation at the

interface between Au and CdS.

5.2 Experimental section

5.2.1 Materials

Cadmium nitrate tetrahydrate (Cd(NO3)2, 99.999 %), thiourea (NH2CSNH2, 99.0

%), hydrogen tetrachloroaurate(III) hydrate (99.999 %), sodium borohydride (NaBH4,

powder, 98%) and ehthylendiamine (redistilled,99.5 %) were purchased from Aldrich

chemicals and used without further purification.

5.2.2 Preparation of CdS Nanorods

123.2 mg (0.4 mmol) of Cd(NO3)2 and 60.8 mg (0.8 mmol) of thiourea were

dissolved in 10mL of ethylenediamine. Nanorods of S- rich CdS were grown in this

solution by heating to 120 C for 10hrs, after which the reaction vessel was cooled to

room temperature to stop crystal growth. Nanorods were collected by precipitation and

washing with distilled water. Cd+ rich CdS nanorods were prepared using the same

procedure except the stoichiometry of the Cd:S precursors was 2:1 rather than 1:2.

Obtained CdS nanorods were used for deposition of Au nanocrystals.









5.2.3 Preparation of Au/CdS Nanorods

A solution of 7.87mg (0.02mmol) HAuCl4 in 10ml of ethylenediamine was added

to the solution containing the S- rich CdS nanorods, and stirred for 15 min at room

temperature. The resulting solution contained an Au-S complex on the surface of the CdS

nanocrystals. This solution was added to 20 ml of distilled water to which 0.9 mg of

NaBH4 had been dissolved, and the mixture turned grey indicating Au nanocrystals had

formed. Nanocrystals was collected by centrifugation and washed with distilled water.

5.2.4 Characterization

High resolution transmission electron microscopy (HR-TEM) analysis of the

crystalline lattice and rod shape was conducted using a JEOL 2010F microscope. X-ray

diffraction (XRD) patterns were obtained using a Philips APD 3720 ray diffractometer

for crystal structure and size. For photoluminescence data, a He-Cd laser emitting at 325

nm was used for excitation. The light emitted was collected using a HR-320

monochromator (Instruments SA, Inc.) with a Hamamatsu R943-02 GaAs

photomultiplier detector. To evaluate the photocatalytic activity of the Au/CdSe

composites, the rate of destruction of a red dye, procion red mix-5B (PRB) from Aldrich

Chemicals, was measured in a 100ml Pyrex glass vessel with magnetic stirring. The

illumination source for photocatalyst was four 12 inch 8 W UV bulbs (UVP inc.)

dominated by emission at 365nm. Aqueous solution of PRB (50 ml, 10 mg/L) and 5 mg

of Au-CdS nanorods composite were placed in the reaction vessel and sonicated for 20

min in the dark. After selected time intervals of UV-illumination, a specimen of the

suspension was collected and analyzed by UV-Vis absorption spectroscopy using a

Perkin-Elmer Lambda 800 spectrophotometer.









5.3 Result and Discussion

5.3.1 Growth of Au Nanocrystals on CdS Nanorods

As described above, nanorods of CdS were synthesized by solution phase method

using ethylene diamine [139]. In general, symmetry breaking is required in the nucleation

step for anisotropic nanocrystals [71]. In solution phase growth, a variety of chemical

techniques to lower the symmetry of a nucleus have been reported by using appropriate

capping reagents. For example, micelles methods [140], kinetically controlled growth

rates of various facets [72], polyol process and solvo-thermal method [139,141].

Bidentate ligand such as ethylene diamine can serve as a "molecular template", leading to

rod shape CdS nanocrystals [139,142]. The XRD pattern from a typical hexagonal

structure of CdS nanorods is shown in Figure 5.1.

For long reaction times, the (002) diffraction peak became more intense and

narrower than peaks from other planes, consistent with growth of rod shape crystals along

the c-axis of the hexagonal structure. The (002) interplanar spacing of the rods increased

from 6.60 to 6.66 A as reaction time increased, presumably because of relaxation of

compressive stress on the larger particle size. After reaction for 10 hr at 120 C, Debye-

Scherrer analysis based on the XRD FHM data indicated a diameter of-15 nm and a

length of -80 nm. The HR-TEM image in figure 5.2 (b) from the same batch of

nanocrystals are consistent with these dimensions, and also shows well resolved

crystalline lattice planes (figure 5.2 (c)). Figure 2 (a) shows that the particles produced in

ethylene diamine after 0.5hr at 120C were anisotropic.

Nanorods of CdS that were S2- rich [139,143] were prepared with an excess of

thiourea, and served as nucleation sites for Au nanocrystals. The photoluminescent

spectra from S2- rich versus Cd2+ rich CdS nanorods were different, as will be discussed
































I I I
20 30 40 50
2 theta


Figure 5.1 X-ray diffraction patterns from hexagonal CdS nanorods prepared at 120C for
(a) 0.5 hrs, (b) 2 hrs, and 10 hrs, respectively.





















Figure 5.2 High resolution transmission electron micrograph of CdS nanorods obtained
after reaction at 120C for (a) 0.5 hrs and (b) 10 hrs. Lattice fringes from CdS
nanorods are shown in (c)









later (Figure 5.5). Ions of Au3+ on the surface of CdS nanorods were reduced to Auo by

NaBH4 at room temperature, resulting in -2 nm particle size [144]. A HR-TEM image of

-2 nm Au nanocrystals deposited directly on CdS nanorods crystals is shown in Figure

5.3. The chemisorption energy of thiolate to Au has been reported to be 28-44 kcal/mol

[145], due to the dominant covalent and minor ionic contributions to the surface bond

[146,147]. This strong chemisorption enables nucleation of Au nanocrystals directly on

the CdS nanorods. The size of the Au nanocrystals is presumably restricted to -2nm size

by the ethylene diamine capping agent [144].




















Figure 5.3 High resolution transmission electron micrographs of CdS nanorods with -2
nm Au nanocrystals.

5.3.2 Photoluminescence and Charge Separation

Luminescence from all nanocrystals, including CdS, may be strongly influenced by

trapped electron/holes at surface defects (traps) that quench radiative band gap

recombination. Therefore the optical properties of CdS nanorods depend strongly upon

their preparation conditions and aspect ratio [143,148-152]. The dependence of the PL

spectra upon the enriched species on the surface is shown in Figure 5.4. The broad PL





















(a)


V /-0.7 eV
I hv -


- i hv
~710 nm
S surface
-0.8 eV/


I I I I
350 400 450 500


550 600 650 700 750 800 850


VWavelength(nm)



Figure 5.4 Photoluminescence spectra from (a) Cd rich and (b) S rich CdS nanorods
(excitation at 325 nm). See text for discussion of defect levels illustrated in the
diagrams.


II









peak from S2- rich CdS nanorods occurred at -710 nm which was 40 nm longer than that

from Cd2+ rich surfaces (-670 nm). Weak excitonic emission with a peak at -489 nm

(2.54 eV) was observed from S2- rich CdS nanorods. In case of S2- rich nanorods, a

photo-generated hole could migrate to the surface, forming S- traps before radiative

electron-hole pair recombination. Such traps might be located -0.8 eV above the valence

band. Recombination of photo-generated electrons from the conduction band with holes

at S- surface traps lead to red emission at -710 nm [152]. In the case of CdS nanorods

with Cd2+ rich surface, S- surface traps are not expected. Instead, surface electron traps,

such as Cd2+ or sulfur vacancies (Vs), are expected to influence luminescence [143].

These photo-generated trapped electrons recombine with holes in the valence band,

leading to emission at -670nm. Based on this PL data, the sulfur vacancy is located about

-0.7 eV below the conduction band in CdS nanorods, which is comparable with the

values reported for bulk CdS [153,154].

Deposition of Au nanocrystals on S2- rich CdS severely quenched the PL intensity

from CdS nanorods (Figure 5.5). Presumably electron-transfer from CdS nanorods to the

Au metal nanoparticles reduced the probability of radiative electron-hole pair

recombination [135]. Luminescence quenching by charge separation was also reported by

Million et al. in type II heterostructure CdSe/CdTe nanocrystals [155]. It has also been

recently reported that the Fermi level of the TiO2 particles covered with Au nanoparticles

shifted closer to the conduction band after UV-irradiation. This was attributed to a large

accumulation of electrons on the Au nanocrystals, indicating a high efficiency for

interfacial charge-transfer between the metal and semiconductor nanocrystals [134].

Furthermore, smaller Au particles induce greater shifts in the Fermi level than did larger












35000-

30000 -

25000-
(a) CdS nanorods
20000-

15000-

10000-

(b) Au/CdS nanorods
5000-

0-

400 500 600 700 800
Wavelength(nm)



Figure 5.5 Photoluminescence spectra from (a) S rich CdS nanorods and (b) Au deposited
CdS nanorods. Note the severe quenching when Au nanocrystals are present
on the CdS nanorods.


------- Vac


5 eV


CdS Au
nanorods nanocrystals


Figure 5.6 Energy level diagrams for Au deposited on CdS nanocrystals illustrating
separation of photogenerated charge.










particles, due to the discrete energy levels for the smaller Au nanocrystals [134,156].

Based on the TEM images and quenching of PL in Figures 3 and 4, respectively,

the present CdS nanorods with 2 nm Au nanocrystals show a high efficiency for charge

separation of photogenerated electron-hole pairs. A schematic energy level diagram of

Au deposited on CdS nanorods that lead to charge separation is shown in figure 6. The

work function of Au is -5ev which causes the top of the valence band to align below the

conduction band minimum of CdS. Therefore any electron photoexcited from the valence

band to the conduction band of CdS will have a strong driving force to transfer to the

lower lying states in the Au nanocrystal. The resulting hole may be trapped on the S-

surface state shown at in Figure 5.6.

5.3.4 Photocatalytic Activity

To further test whether charge separation was enhanced in Au/CdS

nanocomposites, the rate of photocatalytic degradation of the Procion red mix-5B (PRB)

red dye was measured as described above. PRB is easily oxidized in aqueous solution

containing a photocatalyst under UV light irradiation [157]. The structure of PRB and the

normalized concentration versus time of exposure to UV light is shown in Figure 5.7. The

normalized PRB concentration was determined from absorbance at 539 nm (figure 5.8).

Under UV irradiation, the Au/CdS nanorods degrade the PRB at a rate that is 9 % faster

than the rate for only CdS nanorods. Enhanced charge separation by Au/CdS (versus CdS

only) with a subsequently larger rate of production of oxygen radicals (e.g. 02*, *OOH,

*OH [158] are thought to explain the rate increase. These oxygen radicals degrade PRB,

resulting in bleaching of the red color of the solution. The yield of oxygen radicals is

lower with pure CdS nanorods due to the more rapid electron-hole recombination and




Full Text

PAGE 1

SYNTHESIS AND CHARACTERIZ ATION OF COLLOIDAL II-VI SEMICONDUCTOR NANORODS By HYEOKJIN LEE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Hyeokjin Lee

PAGE 3

This work is dedicated to my wife Eunjeong Cho, my son, Keonyoung Lee and my parents in Korea.

PAGE 4

ACKNOWLEDGMENTS Above all, I am grateful to my adviser and my committee chairman, Dr. Paul H. Holloway, who has given me sincere guidance and support for four and one-half years. The four and one half years when I worked for Dr. Holloway is the precious period in my life. And I have learned the attitude of a scientist from my advisor. It is also my great honor to have four other professors (Dr. Steve Pearton, Dr. Wolfgang Sigmund, Dr. Mark Davidson, and Dr. John R. Reynolds) as my committee members. I am especially grateful to Dr. John R. Reynolds for advice and discussion on synthesis of conjugate polymer. For the sample characterization, I am thankful to Lindsay and Dr. Valeria D. Kleiman from Chemistry Department for time resolved spectroscopy study and informative discussion. Also I would like to thank Kerry Siebein of the Major Analytical Instrumentation Center (MAIC) for HRTEM measurement. I also appreciate valuable help from my colleagues in Dr. Holloways group. I especially thank Heesun Yang for useful discussion on luminescent properties of nanocrystals. Another deep appreciation is expressed to Ludie Harmon for helping out with every detailed miscellaneous task. Finally, the continual encouragement and support of my wife, son, and parent are deeply and sincerely appreciated iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................iv LIST OF TABLES ...........................................................................................................viii LIST OF FIGURES ...........................................................................................................ix ABSTRACT .....................................................................................................................xiii CHAPTER 1 INTRODUCTION............................................................................................................1 2 LITERATURE REVIEW.................................................................................................4 2.1 Physical Properties of Semiconductor Nanocrystals..............................................4 2.1.1 Electronic Structure of Semiconductor Nanocrystals...................................4 2.1.2 Surface Effect on Optical Properties of Semiconductor Nanocrystals.........8 2.1.3 Interdiffusion in ZnSe/CdSe Structure.......................................................12 2.2 Synthetic Methods for Colloidal II-VI Semiconductor Nanocrystals..................15 2.2.1 Binary Semiconductor Nanocrystals..........................................................15 2.2.2 Colloidal Ternary Semiconductor Nanocrystals........................................19 2.3 Anisotropic Crystal Growth of Colloidal Semiconductor Nanocrystals..............21 2.3.1 Reaction Temperature Effects on Anisotropic Nanocrystal Growth..........23 2.3.2 Surfactant Effect on Anisotropic Nanocrystal Growth..............................25 2.3.3 Concentration Effects on Anisotropic Nanocrystal Growth.......................27 2.3.4 Effects of Catalyst on Anisotropic Nanocrystal Growth............................30 2.4 Application of Semiconductor Nanocrystals........................................................31 2.4.1 Hybrid Electroluminescent (EL) Devices..................................................31 2.4.2 Hybrid Photovoltaic Devices......................................................................32 2.4.3 Metal-semiconductor Nanoassembly.........................................................34 3 SINGLE STEP GROWTH OF COLLOIDAL ZNCDSE QUANTUM DOTS..............37 3.1 Introduction...........................................................................................................37 3.2 Experimental Section............................................................................................38 3.2.1 Materials.....................................................................................................38 3.2.2 Preparation of ZnCdSe Quantum Dots.......................................................38 3.2.3 ZnS Shell Growth on ZnCdSe Nanocrystals..............................................39 v

PAGE 6

3.2.4 Characterization of ZnCdSe Quantum Dots...............................................39 3.3 Results and Discussion.........................................................................................39 3.3.1 Single Step Growth of Colloidal Ternary ZnCdSe Quantum Dots............41 3.3.2 Effect of Reactivity of Precursor and Reaction Temperature on Optical Properties.........................................................................................................43 3.3.3 Effect of Reactivity of Precursor on Particle Growth................................46 3.3.4 Shell Growth of ZnS on ZnCdSe Quantum Dots.......................................48 3.4 Conclusion............................................................................................................49 4 SYNTHESIS AND CHARACTERIZATION OF COLLOIDAL TERNAY ZNCDSE SEMICONDUCTOR NANORODS..........................................................51 4.1 Introduction...........................................................................................................51 4.2 Experimental Section............................................................................................52 4.2.1 Materials.....................................................................................................52 4.2.2 Preparation of ZnCdSe Nanorods...............................................................53 4.2.3 Characterization..........................................................................................53 4.3 Results and Discussion.........................................................................................56 4.3.1 Synthesis of ZnCdSe Nanorods..................................................................56 4.3.2 Structure of ZnCdSe Nanorods..................................................................58 4.3.2 Effect of Alloying on the Phonon Spectra..................................................60 4.3.3 Photoluminescence and Absorption Properties..........................................63 4.3.4 Time Resolved Photoluminescence (TRPL) Study....................................68 4.3.5 Transient Absorption Spectroscopy (TRA) Study.....................................73 4.4 Conclusion............................................................................................................75 5 SELF ASSEMBLED GROWTH OF GOLD NANOCRYSTALS ON CADMIUM SULFIDE NANORODS.............................................................................................77 5.1 Introduction...........................................................................................................77 5.2 Experimental Section............................................................................................78 5.2.1 Materials.....................................................................................................78 5.2.2 Preparation of CdS Nanorods.....................................................................78 5.2.3 Preparation of Au/CdS Nanorods...............................................................79 5.2.4 Characterization..........................................................................................79 5.3 Result and Discussion...........................................................................................80 5.3.1 Growth of Au Nanocrystals on CdS Nanorods..........................................80 5.3.2 Photoluminescence and Charge Separation................................................82 5.3.4 Photocatalytic Activity...............................................................................86 5.5 Conclusion............................................................................................................88 6 CONCLUSION...............................................................................................................89 6.1 Single Step Growth of Colloidal Ternary ZnCdSe Quantum Dots......................89 6.2 Synthesis and Characterization of Colloidal Ternary ZnCdSe Semiconductor Nanorods................................................................................................................90 6.3 Self Assembled Growth of Au Nanocrystals on CdS Nanorods..........................91 vi

PAGE 7

6.4 Future Direction....................................................................................................92 LIST OF REFERENCES...................................................................................................93 BIOGRAPHICAL SKETCH...........................................................................................107 vii

PAGE 8

LIST OF TABLES Table page 2.1 Diffusion parameters for bulk ternary II-VI compounds.............................................12 2.2 Diffusion parameter of Cd in superlattice structures .................................................13 2.3 Properties of hybrid OLEDs using various semiconductor nanocrystals....................32 3.1 Calculated metal-chalcogen bond energies in metal alkyl chalcogen precursors with the composition of M(ER) 2 (kcal/mol; from reference) ..................................45 4.1 Comparison of and value of CdSe/ZnSe coreshell and ZnCdSe alloyed nanorods...................................................................................................................71 viii

PAGE 9

LIST OF FIGURES Figure page 2.1 Schematic diagram of the molecular orbital model for band structure ........................4 2.2 Energy levels in CdSe nanocrystals. (a) The theoretical size dependence of the electron and hole levels in CdSe nanocrystals. (b) The three lowest transitions as a function of the energy of the first excited state.......................................................7 2.3 Absorption spectra of TOPO/TOP passivated CdSe nanocrsytals with radii from 1.2 to 4.1 nm. Arrows mark the positions of the four well resolved transition..........8 2.4 Schematic illustration and electronic potential step of valence and conduction bands, HOMO and LUMO levels of (a) inorganic core and (b) inorganic core/shell nanocrystals, both with surface attachment of organic molecules.............9 2.5 Radial probability functions for the electron and hole wave functions in bare CdSe, CdSe/ZnS, and CdSe/CdS nanocrystals. The sketches to the right show the band offsets between the various components..................................................................10 2.6 Schematic representation of a core/shell/shell nanocrystal and band gap versus lattice spacing of the wurzite phase of CdSe, ZnSe, CdS and ZnS..........................11 2.7 ZnSe/CdSe quantum well structure grown on GaAs substrate and PL emission shift with anneal temperature for (a) ZnS/ZnSe, (b) ZnS/CdS and (c) ZnSe/CdSe................................................................................................................13 2.8 Classification of solution phase synthetic methods applied to semiconductor nanocrystals..............................................................................................................16 2.9 Chemical structures of representative coordinating surfactants widely used for molecular precursor methods...................................................................................17 2.10 Illustration of molecular precursor methods for II-VI semiconductor nanocrystals such as CdSe quantum dots......................................................................................18 2.11 Chemical structure of the less reactive metal stearate and oleate precursors............18 2.12 Chemical structure of a single molecular precursor, [Cd 10 Se 4 (SPh) 16 ] 4...................19 ix

PAGE 10

2.13 Bandgap versus lattice constant diagram for the common cubic II-VI and III-V semiconductors.........................................................................................................20 2.14 Parameters controlling the crystal shape in solution phase synthesis (molecular precursor methods)...................................................................................................22 2.15 Schematic description of nuclei and resulting shape of CdSe and CdS at (a) high and (b) low growth temperature. c-f) HR-TEM images of CdS monorod (grown at 300 o C), bipod, tripod (grown at 180 o C), and tetrapod (grown at 120 o C). ..........24 2.16 Illustration of growing CdSe nanorod........................................................................26 2.17 Precursor diffusion (arrow) from bulk solution into diffusion sphere for crystal growth (Color gradient indicate concentration gradient).........................................28 2.18 The CdSe nanorods grown in different precursor concentration, (a) 0.067 mol/kg and (b) 0.267 mol/kg................................................................................................29 2.19 Aspect ratio of the CdSe nanorods ((a) 2.5, (b) 6, and (c) 8) as function of initial Cd/Se ratio of precursors (a) Cd:Se=1:5, (b) 1:2, and (c) 5:1..................................29 2.20 Schematic of the mechanism of the catalyzed solution-liquid-solid (SLS) growth process......................................................................................................................31 2.21 External quantum efficiency of hybrid solar cell and TEM image of the CdSe nanocrystals used in each cell..................................................................................33 2.22 Cadmium-gold nanocomposite. (a) The schematic picture of CdS/Au assembly bridged by (b) conducting organic spacer, N,N`-bis(2-aminoethyl)-4,4`-bipyridinium, and (c) electrically insulating spacer, 1,4-trimethyl ammonium butane.......................................................................................................................34 2.23 TEM images showing growth of Au onto the tips of CdSe nanorods ......................35 3.1 Temporal evolution of photoluminescence spectra at reaction temperature of 320 o C as a function of reaction time, (a) 2 min (648 nm), (b) 10 min (623 nm), (c) 20 min (580nm), and (d) 30 min (567 nm). (excitation at 350 nm, doubled peak).........................................................................................................................40 3.2 Picture of luminescence of ZnCdSe nanocrystals dispersed in toluene under UV irradiation. Nanocrystals were grown at reaction temperature of 320 o C and different reaction time, (a) 5 min, (b) 10 min, (c) 20 min, and (d) 30 min..............42 3.3 Dependence of the bandgap energy (E g ) as function of composition of (a) ZnCdSe nanocrystals grown at reaction temperature of 320 o C and different reaction time, 5 min, 10 min, and 30 min and (b) bulk alloy on their composition........................42 x

PAGE 11

3.4 Band gap absorption change as a function of reaction time, (a) reaction temperature 320 o C and (b) reaction temperature 270 o C. Inset represents the magnified plot of dotted circle in (a)........................................................................44 3.5 X-ray diffraction patterns of (a) bulk CdSe, (b) Cd rich ZnCdSe nanocrystals obtained grown for 10 min. at 270 o C (c) ZnCdSe nanocrystals grown for 230 min. at 270 o C and (d) bulk ZnSe..............................................................................44 3.6 Schematic reaction of precursors for nanocrystal formation.......................................45 3.7 Comparison of the size of ZnCdSe nanocrystals grown at different temperatures (a) 270 o C for 230 min. and (b) 320 o C for 30 min (20 nm scale bar). Photoluminescence spectrum of ZnCdSe nanocrystals obtained (c) at 270 o C for 230 min. and (d) at 320 o C for 30 min. after no further spectral shift......................47 3.8 HR-TEM images of ZnCdSe nanocrystals grown at 320 o C as reaction time increase, (a) 2 min (b) 5 min, and (c) 30 min (20 nm scale bar). Average particle sizes (diameters) are (a) ~5.2 nm, (b) ~5.7 nm and (c) ~7.6 nm with aspect ratio 1.5.............................................................................................................................48 3.9 Photoluminescence spectra (a) before ZnS shell growth and (b) after ZnS shell growth on ZnCdSe quantum dots.............................................................................49 4.1 Powder X-ray diffraction patterns of (a) CdSe nanorods,(b) CdSe/ZnSe core/shell nanorods, and (c) ZnCdSe nanorods........................................................................57 4.2 Comparison of powder X-ray diffraction patterns of (a) ZnCdSe nanorods, and (b) spherical ZnCdSe dots........................................................................................59 4.3 HR-TEM image and histogram of size distribution of ZnCdSe nanorods. Lattice fringe from a nanorod is shown by the lower right inset.........................................59 4.4 Raman spectra of LO phonon mode of (a) CdSe nanorods and (b) CdSe/ZnSe core-shell nanorods...........................................................................................................61 4.5 Raman LO phonon spectra of ZnCdSe nanorods after annealing at 270 o C for (a) 1, (b) 2, or (c) 3 hrs.......................................................................................................62 4.6 Photoluminescence spectra of (a) CdSe-ZnSe core/shell nanorods and (b) CdSe nanorods...................................................................................................................64 4.7 UV-Vis absorption spectra of (a) CdSe nanorods, (b) CdSe/ZnSe core-shell nanorods, and (c) ZnCdSe nanorods alloyed at 270 o C for 3hrs...............................64 4.8 Photoluminescence spectra from (a) CdSe/ZnSe core/shell nanorods and ZnCdSe nanorods alloyed at 270 o C for (a) 1, (b) 2, and (c) 3 hrs..........................................67 4.9 TRPL decay curve of CdSe/ZnSe nanorod and ZnCdSe nanorods.............................69 xi

PAGE 12

4.10 The dot lines are experimental data and the full lines are the fitting data using equation ln[ln(Io/It)] versus ln(time) of (a) CdSe/ZnSe coreshell nanorods, (b) ZnCdSe alloy nanorods 1hr, (c) ZnCdSe alloy nanorods 2hr, and (d) ZnCdSe alloy nanorods 3hr....................................................................................................70 4.11 Ultrafast carrier relaxation in (a) CdSe/ZnSe core shell nanorods (b) ZnCdSe alloy 1hr (c) ZnCdSe alloy 2hr (d) ZnCdSe alloy 3hr..............................................74 4.12 Summary of absoption study. (a) Energy separation between 1S and 1P transition (b) 1S bandwidth change (c) Bandgap change of CdSe/ZnSe coreshell and ZnCdSe nanorods.....................................................................................................75 5.1 Hexagonal x-ray diffraction patterns of CdS nanorods prepared at 120 o C for (a) 0.5 hrs, (b) 2 hrs, and 10 hrs, respectively.....................................................................81 5.2 HR-TEM image of CdS nanocrystals obtained at reaction time of (a) 0.5 hrs and (b) 10 hrs. Well resolved lattice fringe of CdS nanorod is shown at (c)..................81 5.3 HR-TEM image of ~ 2 nm size Au nanocrystals deposited directly on CdS nanorods crystals......................................................................................................82 5.4 Photoluminescence spectra of (a) Cd rich and (b) S rich CdS nanorods (excitation at 325 nm).................................................................................................................83 5.5 Photoluminescence spectra of (a) S rich CdS nanorods and (b) Au deposited CdS nanorods...................................................................................................................85 5.6 Energy diagram of Au deposited CdS nanocrystals: photogenerated charge separation between CdS and Au nanocrystals (* surface hole trap state); see text for discussion............................................................................................................85 5.7 Ratio of the concentration C versus initial concentration C o of Procion red mix-5B (PRB) dye in aqueous solution versus time of exposure to 365 nm UV light irradiation in the presence of 5mg of (a) CdS nanorods. or (b) Au/CdS nanorods..87 5.8 UV visible absorption spectra of Procion red mix-5B (PRB) dye in aqueous solution under 365 nm UV light irradiation in the presence of 5mg of (a) CdS nanorods and (b) Au/CdS nanorods.........................................................................87 xii

PAGE 13

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SYNTHESIS AND CHARACTERIZATION OF COLLOIDAL II-VI SEMICONDUCTOR NANORODS By Hyeokjin Lee December 2005 Chair: Paul. H. Holloway Major Department: Materials Science and Engineering Colloidal ternary alloy ZnCdSe quantum dots (5~8 nm size) were synthesized by a single step reaction using Cd and Zn oleates precursors in a trioctylphosphine oxide (TOPO) solution. Optical properties and structure of these nanocrystals were characterized using photoluminescence (PL), UV-Vis absorption spectroscopy, X-ray diffraction (XRD), and transmission electron microscopy (TEM). Significant blue shifts from 648 to 567 nm of the PL emission peaks indicated the formation of ternary alloy nanocrystals during the reaction, and the emission wavelength was dependent on reaction time and reaction temperature. The properties were discussed in terms of nucleation and growth processes controlled by the reactivity and diffusion of precursors. Colloidal ZnCdSe nanorods were synthesized by diffusion of Zn ions into CdSe nanorods in solution at 270 C. CdSe nanorods were prepared using a mixture of tetradecylphosphonic acid (TDPA)/TOPO surfactants at 250 C. The PL quantum yield (QY) of ZnCdSe nanorods was 5~10 %, which is higher than that from pristine CdSe o o xiii

PAGE 14

nanorods (0.6%). XRD and TEM showed that structure of ZnCdSe nanorods was hexagonal structure with ~6 nm in diameter and ~13 nm in length. Alloying and compositional disorder during the reaction were determined by spectral shift and line broadening of Raman spectroscopy, UV-Vis absorption spectroscopy and PL. The PL decay was measured using time-resolved photoluminescence (TRPL) and a stretched exponential function, /exp)(tItIo was used to describe PL decay. Comparing CdSe/ZnSe coreshells to alloy ZnCdSe, we find a significant decrease in the value (from ~0.75 to 0.48 ~ 0.58) which is attributed to compositional disorder in nanorod crystals such as spatial fluctuations of the local Zn concentration. Emission decay time () increases from 173 ns to 270~500 ns. We speculate that the binding energies of exciton in alloy nanorods increase due to increased localization of exciton by compositional fluctuation, leading to increase luminescence decay time () For additional insight, femtosecond transient absorption has been utilized to study the evolution of the absorption bleaching in the nanorods, which support other results explained by alloying and compositional disorder of ZnCdSe nanorods system. Gold nanocrystals ~2 nm in diameter were grown directly on a S 2rich surface of CdS nanorods using strong Au-S bonding nature. CdS nanorods grew from Au nanocrystal nucleation sites prepared by the reaction of nonstoichiometric precursors in the presence of ethylene diamine. PL was quenched significantly after Au nanocrystals were deposited on CdS nanorods and this effect was attributed to interfacial charge separation between Au nanocrystals and CdS nanorods. Efficient charge separation in Au/CdS nanorods enhanced photocatalytic degradation of the Procion red mix-5B (PRB) dye in aqueous solution under UV light irradiation. xiv

PAGE 15

CHAPTER 1 INTRODUCTION Since 1990, colloidal semiconductor nanocrystals such as CdSe have been intensively studied due to their excellent luminescent efficiency, size tunable optical properties, and their great promise for applications in a variety of optoelectronic devices such as light emitting diodes (LEDs) [1] and photovoltaic cells [2]. Furthermore, colloidal nanocrystals are considered by some researchers to be the building blocks for fabrication of quantum superstructures [3]. Synthesis of colloidal semiconductor nanocrystals with high crystal quality and luminescent properties have been reported using molecular precursors [4-7] developed by the research groups of Alivisatos and Bawendi. To improve the emission efficiency, the surfaces of nanocrystals have been coated with a shell of a larger bandgap material to confine the charge carriers and minimize the nonradiative decay channels resulting from electronic surface states. Since 2000, metal oxide precursors with functionalized organic ligands have been used for synthesis instead of a Cd(CH 3 ) 2 precursor for a greener approach [8-10]. Metal oxide precursors are well suited for studies of colloidal nanocrystal growth due to their slow nucleation and growth rates [8]. Even with metal oxide precursors, nanocrystal less than ~2 nm may grow to final diameters within a few seconds, making control of their size delicate and complicated. In addition, the photoluminescent efficiency of extremely small nanocrystals is typically lower than that of larger nanocrystals due to the increased surface-to-volume ratio as the diameter decreases [11,12]. 1

PAGE 16

2 In order to solve those problems resulting from particle size, ternary semiconductor quantum dots are of more interest. A few studies of colloidal ternary alloy quantum dots, such as ZnCdS [13], ZnCdSe [14] and CdSeTe [15,16], have reported excellent luminescent efficiencies that were comparable to those from binary core/shell structured nanocrystals. The luminescence wavelengths from UV to IR from the ternary particles were controlled by composition and size and were shown to be stable. Development of growth methods for rod shaped CdSe nanocrystals promise new opportunities to study shape-dependent electronic and optical devices, such as polarized LED [17,18], lower threshold laser [19], and more efficient photovoltaic cells [2,20]. A higher conversion efficiency (1.7 %) for plastic solar cell was achieved by controlling the length of nanorods, which was attributed to better charge transport to cell electrodes [2]. The better charge conductivity of nanorods crystals may be also utilized for composite organic light emitting diode to solve an injected charge imbalance due to poor conduction between quantum dot nanocrystals. While they are promising, rod shaped CdSe nanocrystals still are reported to have low photoluminescent quantum efficiencies [21,22] and weak confinement along the rod axis has led to inefficient production of blue-green light. In this dissertation, studies of the synthesis of colloidal ZnCdSe nanorods will be reported. Their luminescence in the visible region will be reported. In addition, metal-semiconductor nanoassemblies (CdS nanorods and gold nanocrystals) were prepared and charge separation between these two materials was studied. In chapter 2, a review of the fundamental physics, synthetic methods, anisotropic nanocrystal growth, and applications of colloidal semiconductor nanocrystals is given. In

PAGE 17

3 chapter 3, the preparation and properties are described of ternary alloy ZnCdSe quantum dots prepared in a single step process using a mixture of metal carboxylates (Zn-oleate and Cd-oleate). The effects of the reaction temperature and reactivity of precursors on optical properties, alloying and crystal growth are reported. In studies described in chapter 4, ZnCdSe nanorods were synthesized by diffusion of Zn into the CdSe nanorods. The ZnCdSe nanorods were characterized by absorption spectroscopy, photoluminescence (PL), X-ray diffraction (XRD), high resolution-TEM (HR-TEM) and Raman spectroscopy. Furthermore, the lifetimes and dynamics of colloidal ZnCdSe nanorods were studied using time-resolved photoluminescence (TRPL) and femtosecond transient absorption measurements (TRA). In chapter 5, CdS nanorods were prepared and Au nanocrystals were grown directly on CdS nanorods. Structural and optical properties of CdS nanorods and Au/CdS nanorods assemblies were determined using XRD, HR-TEM, and PL. Nanoscale charge separation between gold and CdS nanorods was studied by luminescence quenching and photo-catalytic properties. Finally, chapter 6 draws conclusions and suggests future directions.

PAGE 18

CHAPTER 2 LITERATURE REVIEW 2.1 Physical Properties of Semiconductor Nanocrystals 2.1.1 Electronic Structure of Semiconductor Nanocrystals Semiconductor electronic properties can be described by molecular orbitals (MO) of the solid. In general, semiconductor quantum dots (QD) can be considered as an artificial atom which is built from a small number of individual atoms [23]. When the atomic orbitals on neighboring atoms are combined pairwise, new energy levels are formed. Doubly occupied bonding orbital () and empty antibonding orbital (*) are the result. Each new atom adds one orbital to the bonding orbital set and to the antibonding orbital set for each bond formed. So the number of levels in a band is equal to the number of bonding electrons per atom times the number of atoms in the crystals. Figure 2.1 Schematic diagram of the molecular orbital model for band structure [24] 4

PAGE 19

5 A spread of orbital energies develops within each orbital set, and the HOMO-LUMO separation in the molecule becomes the bandgap of the bulk solid (figure 2.1). Therefore, the optical properties of semiconductor nanocrystals depend strongly on the size (number of atoms) of the nanocrystal. The energy of the electronic state of a quantum dots (QD) can be described by the Schrdinger equation. The earliest and simplest treatment of the electronic states of a QD is based on the effective mass approximation (EMA) [25]. The EMA rests on the assumption that if the QD is larger than the lattice constants of the crystal structure, then it will retain the lattice properties of the infinite crystal and the same values of the carrier effective masses. The electronic properties of the QD can then be determined by simply considering the modification of the energy of the charge carriers produced by the quantum confinement. Thus, the electronic properties are determined by solving the Schrdinger equation for a particle in a three dimensional (3D) box. The zero th order approximation is a perfectly spherical QD with infinite potential walls at the surface. Strong confinement is defined as the case in which the QD size is small compared with the deBroglie wavelength of electrons in the box or compared with the Bohr radius of the electrons, which is the case for II-VI and III-V semiconductors [26]. Taking into account the Coulomb interaction between electrons and holes that is enhanced due to confinement in the QD, the Hamiltonian (the sum of the kinetic and potential energy) can then be written as )()(222*2*22hheehehheerVrVrremmH (2-1) and (2-2) )()(rErH

PAGE 20

6 where and are the confining potential, and are the distances of the electron and hole from the center of the QD, and eV hV er hr is the dielectric constant of the semiconductor. Analytical solutions of equations (2-1) and (2-2) are difficult because the center of mass motion and reduced mass motion cannot be separated as independent coordinates. Various approaches to solving this problem have been used including perturbation theory [27,28]), variational calculation [29], etc., but all lead to a solution of the form *2**222min25.08.1112RydheERemmRE (2-3) where is the lowest energy separation between hole and electron states in the QD (HOMO-LUMO energy gap), is the bulk exciton binding energy in meV, and minE *RydE R is the QD radius. is often referred to as the band gap of the QD because it represents the threshold energy for photo absorption, that is blue shifted from the bulk band gap, E minE g by a value dependent upon the size. The simple EMA treatment was subsequently improved by incorporating the effective mass kp approach that has been used to calculate the electronic structure of bulk semiconductor [26]. The solution of the Schrdinger equation results in a description of the electronic states in the QD by three quantum numbers plus spin because of the 3D spatial confinement in QDs. A commonly used notation is for the electron states to be labeled as nL e and the hole state as nL F where n is the principal quantum number (1, 2, 3, etc), L is the orbital angular momentum (S, P, D, etc) and F is the total angular momentum (F= L+J, and J=L+S) where S is the spin, and the projection of F along a magnetic axis is m F = -F to +F.

PAGE 21

7 (a) (b) Figure 2.2 Energy levels in CdSe nanocrystals. (a) The theoretical size dependence of the electron and hole levels in CdSe nanocrystals [30]. (b) The three lowest transitions as a function of the energy of the first excited state [31] Figure 2.2 (a) shows the theoretical dependence of the electron and hole levels on the size of CdSe nanocrystals. Figure 2.2 (b) shows agreement between theory and experiment obtained in studies of photoluminescence excitation by Norris and Bawendi [31] and the result for the three lowest transitions as a function of the energy of the first excited state. The three lowest electron states and hole states in order of increasing energy are 1S e 1P e 1D e and 1S 3/2 1P 3/2 2S 3/2 respectively [30]. For optical transition in ideal spherical QDs, the selection rules are n=0, L=0, 2 and F=0, Therefore the first three lowest energy bands in CdSe quantum dots can be assigned to the transitions labeled as 1S [1S e (e)-1S 3/2 (h)], 2S [1S e (e)-2S 3/2 (h)] and 1P [1P e (e)-1P 3/2 (h)]. Figure 2.3 shows absorption spectra of five colloidal CdSe nanocrystals samples with radii 1.2, 1.7, 2.3, 2.8, and 4.1 nm and show well resolved features corresponding to interband optical transitions from coupling to different electron and hole quantized states [32]. These ideal selection rules can be broken by nonspherical QD. Furthermore,

PAGE 22

8 decreasing the degree of symmetry and the anisotropy due to rod-shaped crystals leads to changing of the degeneracy and splitting of the excited states of semiconductor nanostructures [33]. 2.1.2 Surface Effect on Optical Properties of Semiconductor Nanocrystals The surfaces of nanocrystals play a key role in their electronic and optical properties, in part due to the high surface to volume ratio of semiconductor nanocrystals. Some important properties of semiconductor nanocrystals for various type of application are (1) high luminescence quantum efficiency, (2) stability of luminescence properties in real operational conditions, and (3) dispersion of nanocrystals in a desired solvent for processing. All of these properties deal with or are influenced by passivation of dangling bonds present on the nanocrystal surface. Therefore surface modification of Figure 2.3 Absorption spectra of TOPO/TOP passivated CdSe nanocrystals with radii from 1.2 to 4.1 nm. Arrows mark the positions of the four well resolved transition [32].

PAGE 23

9 semiconductor nanocrystals has been the subject to extensive investigation. Band edge emission from nanocrystals competes with both radiative and nonradiative decay channel originating from surface states. Organic ligands attached to the nanocrystal surface can affect the solubility of nanocrystals, as well as result in improved luminescent efficiency [5, 34]. However, a high luminescence quantum yield (QY) is difficult to achieve in semiconductor nanocrystals coated by organic ligands due to imperfect surface passivation. In addition, the organic ligands are labile for exchange reactions due to weak bonding to the nanocrystal surface atom exposing the surface to degradation effects such as photooxidation [35, 36]. In some cases, chemical degradation of the organic ligand molecule itself might lead to degraded luminescent nanocrystals [36]. organicmolecule Eg(core)Eg(shell) band offset band offset Organic molecule organicmolecule Eg(core)Eg(shell) band offset band offset organicmolecule Eg(core)Eg(shell) band offset band offset Organic molecule (a) (b) Figure 2.4 Schematic illustration and electronic potential step of valence and conduction bands, HOMO and LUMO levels of (a) inorganic core and (b) inorganic core/shell nanocrystals, both with surface attachment of organic molecules.

PAGE 24

10 A proven strategy for increasing both luminescence QY and photostability is to ll of a higher band gap semiconductor on the core nanocrystals. Examples grow a sheincludSe s er a of e sketches to the right show e CdSe/ZnS [6,37]}, CdSe/CdS [7], CdSe/ZnSe [38], CdS/ZnS [22] and InAs/Zn[39] (figure 2.4). Not only effective surface passivation, larger band gap shell materialprovide also a potential band offset for electron and holes originating in the core nanocrystals, reducing the probability for the carriers to be trapped at the surface. Therefore higher luminescence QY can be achieved by inorganic shell growth ovluminescent core. When an inorganic shell material such as ZnS surrounds the corenanocrystal, the absorption and emission energies shift to lower values (ca.10~20 nm) than those from bare quantum dots [6,37,40]. Dabbousi et al. explained the influencesurface passivation on shift in energies using a simple theoretical treatment of charge carriers in a spherical box [6]. Figure 2.5 Radial probability functions for the electron and hole wave functions in bare CdSe, CdSe/ZnS, and CdSe/CdS nanocrystals. Ththe band offsets between the various components. [6]

PAGE 25

11 The wave function of the lighter electron tunnels into the shell, while the hole wave function hadelocalization of the electron lowers its confinement energy and consequently the energy of the excited state. The probability of penetrating into the shell layer depends on the barrier height between the core and shell, which means that lower barrier heights lead to larger energy red shifts [6] (figure 2.5). An additional effect on the luminescence QY is the lattice mismatch between the core and shell of the nanocrystals, because dislocations induced by interfacial strain may act as nonradiative recombination centers. Two methods have been reported for reducing the defects at the core/shell interface. One method is growth of a compositionally stepped shell, such as a CdS/ZnS shell/shell structure grown on a CdSe core (figure 2.6). The intermediate CdS shell results in relaxation of the strain and consequently improved PL QY and stability of nanocrystals [22,41]. In the other method, which is called photoannealing, defects at the highly strained interface can diffuse to the outer surface upon irradiation of the nanocrystal with a laser [22]. Figure 2.6 Schematic representation of a core/shell/shell nanocrystal and band gap versus lattice spacing of the wurzite phase of CdSe, ZnSe, CdS and ZnS. [41] s a negligible probability of spreading into the shell layer. The increased c-lattice spacing Bandgapof the bulk (V e) c-lattice spacing Bandgapof the bulk (V e)

PAGE 26

12 Despite its great importance, characterization of surfaces and interface of the core/shell nanocrystals is still limited due to a number of factors. 2.1.3 Interdiffusion in ZnSe/CdSe Structure Interdiffusion at the core-shell interfaces may occur and change the band gap energy, alter the electron and hole confinement, and result in large changes in the subband energies of the conduction and valence bands. Solid state diffusion is thermally activated and the diffusion coefficient, D, follows the Arrhenius relationship )exp()(kTQDTDo (2-4) where ity and aring the random walk and Q are the experimentally determined infinite temperature diffusiv 0D ctivation energy, respectively. An estimate of the extent of interdiffusion occurat temperature T in time t is the root mean square displacement from expression tTDxld)(2 2 (2-0compounds are not well established, although diffusivities and qualitative results are 5) Values of and for interdiffusion of atomic species in wide band gap II-VI available over a limited temperature range. A summary of the experimental diffusion terssMartin [42] determined values of = 6.4 x 10-4 cm2s-1 and =1.87 eV for Cd iffusion parameters for bulk ternary II-VI compounds Temperature (oC) Composition (cm2s-1) (eV) D Q parame at the pecified temperatures is given in Table 2.1. 0diffusion into ZnSe. These values yield diffusion length of 0.13, 0.58, 2.1, 6.7 and 41nmTable 2.1 DSystem D Q 0D Q Zn1 6.4 x 101.87 -4 -x Cd x Se 700~950 36.00 x Zn1-xCdxTe 700~1010 9.01.0 x )32.2exp(29.0x 2.14 Ref) [42, 43]

PAGE 27

13 Figure 2.7 Zgrown on GaAs substrate and PL et with aerature for (a) ZnSnSe,CdS and (c)e/CdSoone moC accordn 400oC. Anionic interdiffusion was found not to be significant at temperatures below (700~950oC) of Martins experiments [44]. System Temp (a)(b) (c) (100) GaAs1 nmCdSeZnSe-CdSesystem 500 nm4 nmZnSeZnSe (a)(b) (c) (a)(b) (c) (100) GaAs1 nmCdSeZnSe-CdSesystem 500 nm4 nmZnSeZnSe (100) GaAs1 nmCdSeZnSe-CdSesystem 500 nm4 nmZnSeZnSe nSe/CdSe quantumshif well structure nneal temp mission /Z (b) ZnS/ ZnS e. [44] in layers annealed for 1 hour at 300, 350, 400, 450, 500, and 550C respectively. Even onolayer diffusion distances (~0.3 nm) would be expected to produce observable effects on optical properties in nanoscale systems, PL shifts could be expected at 350ing to Martins data. However, Parbrook et al. reported that ZnSe/CdSe and ZnS/CdS superlattice structures grown by the MOCVD method show disordering at annealing temperatures greater than 400 and 450oC (figure 2.7) [44]. However, no alteration was found for the samples annealed at temperatures lower tha 550 o C. They speculated that this discrepancy was due to the high temperature range Table 2.2 Diffusion parameter of Cd in superlattice structures [47] [45] D (cm2s-1) Q (eV) l* erature 0 d ZnSe/CdSe/ZnSe 340~400 1.9 x 10 -4 1.8 0.14 nm ZnCdSe(10nm)/ZnSe(10nm) 360~600 1.2 x 10 -5 1.7 0.01 nm Calculated diffusion length after 1 hours at 300 o C

PAGE 28

14 On the other hand, diffusion coefficient of Cd in quantum well structures in therange of 340 g et al. (table 2.2). Interd cts on the density of group II vacancies Diffusion parameters in colloidal nanocrystals have not been reported. However, Zt int at 2 in coll CdS]. A temperature of 270oC, which they c s tend to be unsaturated, there is a large energy assoca e y o C~400 o C was calculated by Rosenauer et al. and Strabur iffusion in the ZnSe/CdSe system was negligible at 300 o C; for example, diffusion length of Cd after 1hr was only 0.14 nm. However, the diffusion behavior of Cd in quantum well structures is significantly affected by the growth condition, doping, andatmosphere during annealing, due to their effe which accelerate interdiffusion [45, 46]. hong et al observed significan erdiffusion 70~290 o C loida e/ZnSe core/shell nanocrystals and significant PL blue shift [14 alled the alloying point, is much lower than previous reported temperature for diffusion in bulk or quantum well structures. Diffusion at lower temperatures may be reasonable in nanoparticles because of the large surface to volume ratio as compared withbulk or quantum well structure. Another important parameter is imperfections at theinterface between nanocrystal core and shell layer. In a system containing only a few hundred atoms, a large fraction of these atoms will be located on the surface or the core/shell interface. As surface atom bond iated with this surface. In the solution, surface atoms move to minimize surface areand minimize surface energies by reconstructing [48]. Defects at the core/shell interfacmight result in higher diffusion rates than for quantum well (QW) structures grown by MBE. An increased concentration of vacancies or imperfection at the interface between the core and shell is consistent with photo-annealing effects on luminescent efficienc[22]. When nanocrystals were exposed to UV light irradation, the luminescent efficiency

PAGE 29

15 of CdSe/ZnS core/shell nanocrystals increased by factor of 2~10 due to reduction by annealing of vacancies [22]. Yet another possible explanation of accelerated diffusionanocrystals is a lower activation energy (Q). The coordination of ions close to the surfaces is not satisfied in small size nanocrystals, i.e. the coordination number is less than that for ions in the bulk. In addition, the interaction between distant neighbor atoms in nanocrystals is much weaker than that in the bulk crystals [49,50]. Weak interactionsbetween atoms in colloidal nanocrystals may lead to lower activation energy for interdiffusion. This argument is consistent with weakening of the crystal field strength with decreasing size [49,50]. For example, in the case of ZnS:Eu n in r ults ods [54,55] are included in this categor developed by the groups of Alivisatos and Bawe 2+ nanocrystals, the crystal field strength of ZnS nanocrystals is weaker than for bulk materials. Thereforeemission energy of Eu 2+ shifts to higher energy [49,50]. 2.2 Synthetic Methods for Colloidal II-VI Semiconductor Nanocrystals 2.2.1 Binary Semiconductor Nanocrystals The solution phase methods used in the synthesis of semiconductor nanocrystals are classified into two main categories (figure 2.8): Controlled precipitation, or Moleculaprecursors. Controlled precipitation includes those methods based on traditional precipitation techniques, forming a stable nanocrystal from the mixture of the ionic components of the semiconductor. In this method, the stability of the nanocrystals resfrom using stabilizers in the solution (e.g., polymers [51], or surfactants [52]). Sol-gel methods [53], reverse micelle [52] and hydrothermal meth y. Molecular precursor methods have been ndi, as the most promising chemical routes to synthesize quantum dots for the optoelectronic and biological application. [4-7] In these procedures, the molecular

PAGE 30

16 precursors are decomposed and react in a coordinating solvent at relatively high temperature, hence promoting the crystallinity and passivating the dots surfaces withappropriate surfactants. A wide range of II-VI semiconductor nanocrystals have been prepared using such methods. Figure 2.9 shows chemical structure of some surfactants that are most widely used in the mole cular precursor methods. In the conventional TOPO method (trioc on phase synthetic methods applied to semiconductor nanocrystals Consequently, a large number of nucleation centers are initially formed, and continued to grow via Ostwald ripening (i.e. the growth of larger particles at the expense of smaller particles to minimize the higher surface free energy associated with smaller particles) [56]. tylphosphine oxide) described by Bawendi and co-workers, the group VI species (e.g., TOP-Se) are injected rapidly into hot TOPO solution containing the group II species (e.g., Cd(CH 3 ) 2 ) and the reaction solution is vigorously stirred to avoid inhomogeneous particle growth. Controlled prec ipitationHydrothermal methodMolecular precursorOrganometallicprecursormp. Figure 2.8 Classification of soluti Reverse micelle methodSol-gel method-Pyrolysisat high te Controlled prec ipitationHydrothermal methodMolecular precursorOrganometallicprecursormp. Reverse micelle methodSol-gel method-Pyrolysisat high te

PAGE 31

17 Figure 2.9 Chemical structures of representative coordinating surfactants widely used for crystals whose surfaces are passivated with coordinating surfactants are btained, allowing further manipulation such as size selective process and shell growth using other sodifications to this mecs and production of mpyrophoric comthis meplexes, are prepared by reactisuch as stearic acid (figure 2.11). The reactivity of such precursors toward group IV PO trioctylphosphineoxide (TOPO)P trioctylphosphine (TOP)N Trioctylamine (TOA)POOHOH tetradecylphosphonicacid (TDPA) molecular precursor methods. The coordinating surfactants (TOPO and TOP) limit or control particle growth. Size-tuning is achieved during reaction by adjusting the time and temperature of synthesis. Nano o olutions. This method is illustrated in figure 2.10. Several mthod have been reported for better control over growth dynamionodispersed nanocrystals.[8,57] To avoid the use of toxic, highly reactive and pounds such as Cd(CH3)2 at high temperatures, several modifications to thod have been reported using alternative precursors which are relatively stable, inexpensive, and safe [8-10]. The alternative precursors, metal carboxylate comon of metal oxide (e.g., CdO or ZnO) with alkyl carboxylic acids, species is lower than the traditional organometallic precursors such as Cd(CH 3 ) 2 and

PAGE 32

18 igure 2.11 Chemical structures of the less reactive metal stearate and oleate precursors. OO Figure 2.10 Illustration of molecular precursor methods for II-VI semiconductor nanocrystals such as CdSe quantum dots. [4-7] OO Zn 2Cd 2Zn stearateCd oleate Se-TOPCdsource N F injectionIn TOPO250~320oC N2 2 Se-TOPCdsource N injectionIn TOPO250~320oC N2 2

PAGE 33

19 and Zn(CH 3 ) 2 Therefore, alternative route using metal carboxylate are more suite d to control the growth of high quality semiconductor nanocrystals and for studying growth echanisms of colloidal nanocrystals due to their slow nucleation and growth rate [8]. Single molecular precursors, which contain all of the elements for the desired nanocrystals, have been reported to be alternative precursors (e.g. [Cd10Se4(SPh)16]4or [MeCdSe2CN(C2H5)2]2). [58-60] The use of single molecular precursors with preformed metal-chalcogenide bonds provides a convenient reactive intermediate and allows nanocrystal growth to be initiated at relatively lower temperatures.[58] Incorporation of dopants, such as Eu3+, into the semiconductor nanocrystals can be achieved effectively by using the single molecular precursor described above.[61,62] The use of prebonded complex materials such as Mn2(-SeCH3)2(CO)8 have been applied to doping transition etal into semiconductor nanoparticles to reduce the extent of dopant segregation on the particle surface.[63, 64] m m Figure 2.12 Chemical structure of a single molecular precursor, [Cd 10 Se 4 (SPh) 16 ] [58] 2.2.2 Colloidal Ternary Semiconductor Nanocrystals A variety of II-VI ternary and quaternary alloy semiconductors based on ZnSe been studied for LEDs and laser diodes (LDs).[65, 66] Compositions of these materials 4have

PAGE 34

20 4 V semiconductors.[70] Figure 2.13 Bandgap versus lattice constant diagram for the common cubic II-VI and III-have been demonstrated whose wavelengths cover the whole visible range. For example, ZnSSe (blue-green), ZnCdSe (green-yellow) and ZnCdSeTe (yellow-orange) have been grown as an emission layer on appropriate substrate such as GaAs and InP [67-70]. Modification has been reported to reduce the defects and increase the device life time. [67-70]. However, very little effort has been reported on growth of colloidal ternary alloy nanocrystals, although nanocrystalline binary compounds such as CdSe have been intensively studied for use of optoelectronics and bio-medical imaging. Figure 2.13 shows the energy gap versus the cubic lattice constant of binary II-VI semiconductors. Ternary alloyed quantum dots, a solid solution of two binary semiconductors, would be expected to exhibit intermediate energy band gaps between those of the constituent binaries, and opens new possibilities in band gap engineering. Various types of ternary 5.255.505.756.006.256.50123 0 ZnS InPInAsPbSePbSCdTe CdSeCdS ZnSeEnergy Gap (eV)5.255.505.756.006.256.50123 4 ZnS InPInAsPbSePbSCdTe CdSeCdS ZnSeEnergy Gap (eV) 0 Lattice Constant () Lattice Constant ()

PAGE 35

21 alloy quantum dots have been prepared using the molecular precursor methods. In 2003, Bailey et al. reported composition tuning of colloidal CdSeTe nanocrystals. [15] Light emission from these CdSeTe nanocrystals could be tuned over the range from 650nm to 800nm with QYs of up to 60% [15,16]. Zhong et al. also reported ternary alloy ZnCdSe and ZnCdS quantum dots with high PL efficiency (25~70 %) and narrow band luminescence properties (15~30 nm) in the near UV visible region [13,14]. The luminescence wavelength of ternary systems can be controlled by composition and particle size from UV to IR. By judicious choice of the composition, the wavelength desired can be achieved with stable larger particles, while in binary systems the emission wavelength is only controlled by the nanocrystal size. In general, nanocrystal less than ~2 nm may grow to final diameters within a few seconds when molecular precursor method is used for growth. Therefore controlling particle size below ~2 nm is very delicate and complicated. In addition, the photoluminescent efficiency of smaller nanocrystals is lso be ac ance of typically lower than that of larger nanocrystals due to the increased surface-to-volume ratio as the diameter decreases.[11, 12] Some of the problems resulting from particle sizecan be avoided or reduced in ternary quantum dots. High luminescent intensities can a hieved from ternary compounds, at times being comparable to core/shell structurednanocrystals. Zhong et al [14] suggested that high luminescence of ternary quantum dots may result from spatial composition fluctuation that produce atomically abrupt jumps in the chemical potential which localizes the free exciton states and leads to high luminand stability. 2.3 Anisotropic Crystal Growth of Colloidal Semiconductor Nanocrystals One-dimensional (1D) nanostructures such as wires, rods, belts, and tubes are interest due to their unique applications in mesoscopic physics and fabrication of

PAGE 36

22 nanoscale devices [71]. The understanding of 1D nanostructures (nanorods and nanotubes) has been slow due in part to difficulties in the synthesis and fabrication of these nanostructures with well controlled dimensions, morphology, phase purity, and chemical composition. In recent years, shape control of nanocrystals through molecular precursor method has been successful [72-74]. By controlling growth variables,shapes of nanocrystals have been produced. Colloidal nanorods synthesized by various these methoFigure 2.14 Parameters controlling the crystal shape in solution phase synthesis y [79]. ds exhibits potential technological advantages over spherical nanocrystals, such as linearly polarized emission,[17,18,75,76] lasing from quantum rods in the visible range [19,77], and improved solar cell performance [2, 78]. There are several parameters that can influence the growth pattern of nanocrystals (figure 2.14). G (molecular precursor methods) The intrinsic surface energy of the crystallographic face of the seed is important, since the kinetic energy barrier (G) is inversely proportional to the surface energ Concentration Temperature Surfactants Catalyst Concentration G Temperature Surfactants Catalyst

PAGE 37

23 Therefore, the final shape of nanocrystals is dependent on the crystal structure ocrystals which are influenced by reaction temperature. However, surface properties can also be tailored by the types and the amounts of adsorbing organic capping molecules.The concentration of precursors also plays a key role for the determination and evolutionof the shapes of nanocrystals. Rod growth can also be achieved by using catalyst in the case for materials with zinc blend cubic structures (see 2.3.4). In the following sectionthe factors important to the anisotropic growth of II-VI semiconductor nanocrystals are summarized. 2.3.1 Reaction Temperature Effects on Anisotropic N f the seed anocrystal Growth the crystallographic phase, although the stable phase is highly dependent on its environmical for directing the ructure and consequent variation of surface energy warchitecture of the nanocrystthermodynamican be controlled by adjusting thare more stable at hecause the surface energy of the (001) face of the wurtzite phase is typicallyhigher than that of otheensity and number of dangling bonds, the prefer One of the critical factors responsible for the shape determination of the nanocrystals is ent. Initially, the crystalline phase of the nuclei is critintrinsic shapes of nanocrystals due to its characteristic unit cell stith crystallographic orientation. After the preferred crystalline phase is nucleated, the subsequent growth stage strongly governs the final als through the delicate balance between the kinetics and cs of the growth. The crystalline phase of the seeds e initial temperature during the nucleation process (figure 2.15). In the case of CdS semiconductor nanocrystals [80], nuclei with a wurtzite structure igh temperature (~300oC) and nanorod formation is observed. B r faces due to its higher atomic d red growth is along the c axis. On the other hand, at lower temperatures (120~180 o C), zinc blende nuclei are preferred and tetrahedral seeds with four {111} faces are

PAGE 38

24 formed. The growth of wurtzite pods along the [001] directions from the {111} faces and the formation of CdS multipods can be achieved by low temperature growth Figure 2.15 Schematic descriptions of nuclei and resulting shape of CdSe and CdS at (a) high and (b) low growth temperature. c-f) HR-TEM images of CdS nanorod (grown at 300 erature (250 oC). Od o C), bipod, tripod (grown at 180 o C), and tetrapod (grown at 120 o C) [79]. Similarly, in the case of CdSe nanocrystals [73,81] control of the temperature influences the formation of wurtzite or zinc blende nuclei. If conditions that favor the cubic phase are set, then large tetrahedral nanocrystals are obtained after formation of zinc blende nuclei. One way of achieving this is keep the system at low temp CdSe, CdSCrystalline phase of seedw TemperatureHigh wurtziteRod (a) n the other hand, hexagonal growth is promoted by high temperature or an increase in the concentration of monomers and result in growth of nanorods. For complicateanisotropic shape, zinc blende nuclei can be grown, and conditions can be switched from the thermodynamic growth regime to the kinetic growth regime by increasing the temperature, or by increasing the precursor concentration, or both. Wurtzite rods are formed off of the {111} faces of the original zinc blende nucleus leading to tetrapod Lo zinc blende tetrapod(b) CdSe, CdSCrystalline phase of seedw TemperatureHigh Lo wurtziteRod zinc blende tetrapod CdSe, CdSCrystalline phase of seedw TemperatureHigh wurtziteRod (a) zinc blende tetrapod(b) Lo

PAGE 39

25 nanocrystals. However, the rapid switching is difficult to achieve in a reproducible way in a batch type reactor because the nucleation and growth of nanocrystals take p lace on a time scale of seconds to minutes. In the case of Cd chalcogenide nanocrystals, the thermodynamically stable form can be controlled with reaction temperature if the precursor concentration is low. Other types of semiconductor nanocrystals, such as PbS, can be grown with various shape by changing the reaction temperature.[82] The effects of other parameters on anisotropic growth must be considered, such as mixture of surfactant, concentration of precursor, or catalyst. 2.3.2 Surfactant Effect on Anisotropic Nanocrystal Growth The nanocrystal surface is coated with a layer of organic molecules known as othe surfacta They are mto provide access for the addition of precursor units, but stable enougles d r etic surfactants during solution growth. At a reaction temperature in the range of 200~400C, nt molecules are dynamically adsorbed to the surface of the growing crystals.obile enough h to prevent the aggregation of nanocrystals. In addition, the surfactant molecuare chemically stable at the high temperatures required for growth. Surfactants that bintoo strongly to the surface of the crystals would restrict the crystal growth. On the othehand, a weakly bound molecule would yield large particles, or aggregates. Some xamples for the growth of CdSe include alkyl phosphine, alkyl phosphine oxides, or amines as shown in figure 2.9. Furthermore, surfactant coating allows for great syntheflexibility in organic solution in that it can be exchanged for another coating of organic molecules having different functional groups or polarity. The surface energy difference of facets in wurtzite crystals can be changed by adjusting the type and ratios of surfactants.[72, 79] In other words, the shape of the

PAGE 40

26 nanocrystal can be controlled by using surfactants (or mixture of surfactants) that bind differently to the different crystallographic faces. Peng et al. pioneered the use of mixed surfactants to control the shape of CdSe nanocrystals.[72, 81, 83] They found that CdSe quant due to its higher atomic density and number of dangling bonds as mentioned ly more reactive than um rods with aspect ratios as high as 10 could be obtained in large quantities by adding hexylphosphonic acid (HPA) to TOPO. Explanations for nanorod growth with particular orientation are based on inherentanisotropy in the wurtzite structure and the surface coordination of the surfactants. [73]The surface energy of the (001) face of the wurtzite phase is typically higher than that of other faces in a previous section. These facts make the (001) facet chemical Figure 2.16 Illustration of growing CdSe nanorod [83]. other facets. Dangling bonds of surface atoms on the (001) facet cannot be fully passivated by surfactants due to steric interference between neighboring surfactant molecules, whereas surface atoms on other faces can be pas sivated effectively [73,8 er et 1,83,84]. In addition, alkyl phosphonic acids (PA) such as HPA are bound morestrongly to the nanocrystal surface than TOPO, as confirmed by calculation by Puzd

PAGE 41

27 al [84].The strong binding ability of PA is responsible for further raising the surface energy difference between (001) face and some others of CdSe and enhancing the relativegrowth rates of different facets. Figure 2.16 show illustration of growing CdSe nanorods passivated by a surfactant. The (001) faces are either Cd or Se terminated. If a layer of atoms is exposed on this face, the phosphonic aci Se d molecules will passivate them very weakf surface area per Cd atom in (001) face and (100) face is about 0.19 nm2/Cd and 0.30 nm2/Cd respectively. However, surface area per surfactants is about 0.25 nm2. Therefore the steric interference is larger at (001) face that at (100) face. Combined with the fact that the sides of nanocrystal can be passivated by either phosphonic acid or TOPO, the (001) face becomes the fastest growing face of the crystal. 2.3.3 Concentration Effects on Anisotropic Nanocrystal Growth Geometric control has also been demonstrated by varying the concentration of the precursors in the growth solution. Peng et al. [83,85,86] have systematically studied the rowth in terms of precursor ng crystal ly, because Se atoms in CdSe have an anionic character and the surfactants in the system bind strongly to cationic species. This situation allows for the rapid building up oa layer of Cd atoms, which would again slow down the growth of this face. When Cd atoms are attached to Se atoms, the alkyl phosphonic acid cannot passivate all the dangling bonds from Cd atoms due to steric interference as illustrated. In figure 2.16, anisotropic growth of CdSe and explained nanocrystal g concentration effects. Figure 2.17 show schematic diagram of nanocrystal growth at high precursor concentration. Precursors should diffuse into the diffusion sphere around growidue to high concentration gradient when precursor concentration in the bulk solution ishigh. As a result, the volume of nanocrystals increases noticeably.

PAGE 42

28 Bulk solutionDiffusion sphere Precursor diffusion Bulk solutionDiffusion sphere Precursor diffusion Bulk solutionDiffusion sphere Precursor diffusion crystal growth (Color gradient indicate concentration gradient) However, precursor supplied by diffusion is consumed by the quick growth of the(001) facet. As a result, the diffusion flux goes to the c axis exclusively and makes the growth reaction rate along the c-axis much faster than that a Figure 2.17 Precursor diffusion (arrow) from bulk solution into diffusion sphere for long any other axis leading to io of crystals decreases [83]. A lower precursor concentration will l ner as the initial high aspect ratio (length vs. diameter) of nanorods [83]. At low precursor concentration, there is no net diffusive flux between the bulk solution and the growing nanoparticle. The precursors on the surface of the nanocrystals adjust their positions to minimize the total surface energy of a given crystals and eventually the aspect rat ead to Ostwald ripening of the particles which is particle growth by consumingsmaller size particle [83, 85, 86]. Consequently, relatively high precursor concentrations promote the formation of nanorods with a high aspect ratio (figure 2.18). Other experimental factors can affect nanocrystal growth in the same man

PAGE 43

29 (a) (b) Figure 2.18 The CdSe nanorods grown in different precursor concentration, (a) 0.067mol/kg and (b) 0.267mol/kg. [85] precursor concentration. For example [85] : i). Rcleation rate, and le (a) (b) (c) eactivity of precursor: a more stable Cd precursor reduces the nu ss nuclei are formed. Therefore the concentration of the remaining precursor at initial stage is higher than that of unstable reactive monomer. Figure 2.19 Aspect ratio of the CdSe nanorods ((a) 2.5, (b) 6, and (c) 8) as function of initial Cd/Se ratio of precursors (a) Cd:Se=1:5, (b) 1:2, and (c) 5:1. [85]

PAGE 44

30 ii). Initial Cd/Se ratio: The Se precursor is more active than the Cd precursor. As a iii). Multipof the Cd precursor, the more multiple injeSo fa nanocrystals were based on the s that can be selectively bound by coordite th can be prepared ffectively by the solution-liquid-solid (SLS) mechanism for lattice structures that do not mforfrom the su0 result, higher Se precursors lead to faster nucleation with consequent decreases in the precursor concentration and a lower aspect ratio of the nanocrystals (figure 2.19). le injections of precursor: After the primary injection of a large excess reactive Se precursor may be increased gradually as using ctions. iv). High growth temperature: High temperatures increase the diffusion flux towards the fast growth facets, leading to a higher aspect ratio. 2.3.4 Effects of Catalyst on Anisotropic Nanocrystal Growth r, discussions of the growth of rod-shaped presence of chemically dissimilar lattice face inating surfactants or precursors to achieve different growth rates along different crystal axes. Therefore nanorod growth has been limited to semiconductors with wurtzcrystal structures, such as hexagonal CdSe. However rod grow e exhibit chemically dissimilar surfaces, e.g. zinc blend cubic structures.[87,88] In one echanism for catalyzed SLS growth, precursors diffuse into metal nanocrystals (catalyst) and precipitate when the catalyst solution is supersaturated, resulting in the mation of one dimensional nanocrystals (figure 2.20) Kan et al. reported the formation of InAs nanorods (InP) which were prepared by the reaction between tris(trimethylsilyl)arsine (TMS)3As and InCl3 in TOPO using dodecanethiol stabilized gold nanocrystals with diameters of ~2 nm [87,88]. The Au nanocrystals act as a catalyst into which reactants dissolve, leading to directed growth persaturated solid solution. The melting temperature of 1.6 nm Au is ~ 35

PAGE 45

31 oC, in tempe, Yu et al. reported growing high quality, narrow diameter CdSe nanowystals l anic re uctor creasing to ~ 480 o C for 2.5 nm Au particles which is significantly lower than that of bulk gold [87,88]. The smaller nanocrystals, the larger contribution made by surfaceenergy to overall energy of the system and thus the more dramatic the melting rature depression [48]. At temperatures just below this melting but above the boiling point for TOPO, rapid diffusion can take place leading to nanorod growth. Using Bi or Au/Bi catalysts ires. [89, 90] 2.4 Application of Semiconductor Nanocr 2.4.1 Hybrid Electroluminescent (EL) Devices Since the first observation of light emission from organic materials by Tang et a[91], continuous and rapid improvement in device performance have enabled orglight emitting devices (OLEDs) to compete with existing technologies. However there astill many problems to be overcome, such as improving device stability and color purity. Full width at half maximum (FWHM) of photoluminescence of colloidal semicond SolutionCatalystNanorod PrecursorByproducts Precipitation Figure 2.20 Schematic of the mechanism of the catalyzed solution-liquid-solid (SLS) growth process [74] SolutionCatalystNanorod PrecursorByproducts Precipitation

PAGE 46

32 nanocrystals is about 30nm which is narrower than those from organic materials. In addition these inorganic nanocrystals are much stable and robust than organic moleculesTherefore hybrid OLEDs using semiconductor nanocrystals as an emission layebeen to have good stability and efficiency. The first demonstration of a hybrid OLEDby Colvin et al in 1994 [92]. External quantum efficiencies approaching 1% have been reported as summarized in table 2.3. In order to enhance the quantum efficiency of hybOLED devices, seve r have was rid ral problems must be solved including more efficient charge transfer between the organicharge due to poor pinhole defects in the nanocrystal layer, layer and nanocrystals, the imbalance of injected c conduction through nanocrystals, a high density of uniformity of nanocrystals, a high density of pinhole defects in the nanocrystal layer, uniformity of nanocrystals in the deposited layer, and optimization of interlayer structure of devices [1]. Table 2.3 Properties of hybrid OLEDs using various semiconductor nanocrystals. Nanocrystals (NC) Device structure QE (%) Ref CdSe ITO/PPV/NC/Mg VIS 0.001~0.01 [92] CdSe/CdS ITO/PPV/NC/Mg/Ag VIS 0.22 [93] CdSe/ZnS ITO/TPD/NC/Alq3/Mg-Ag/Ag VIS 0.5 [1] CdS:Mn/ZnS ITO/PEDOT/PVK/NC/Al VIS [94] InAs/ZnSe ITO/PEDOT/MEH-PPV-NC/Ca/Al NIR 0.5 [95] PbSe ITO/TPD/NC/Alq3/Mg-Ag/Ag NIR 0.001 [96] PbS ITO/PPV/NC/Mg/Au NIR [97] HgTe ITO/PEDOT/NC/Al NIR 0.02 [98] ITO: Indium tin oxide, PPV: Poly (p-phenylenvinylidene), PVK: Polyvinylcarbazole PEDOT: Polyethylenedioxythiophene NC: Nanocrystals .2 Hybrid Photovoltaic Devices 2.4 Hybrid photovoltaic devices consist of semiconductor nanocrystals integrated with

PAGE 47

33 conjugated polymers to take advantage of the complementary properties of organic ainorganic materials [2, 20, 99]. The nanocrystalline semiconductors generally have larger electron affinities than do conjugated polymer, so interfacial charge separation may occmore efficiently when electrons and holes are generated by absorption of photons. Interfacial charge separation reduces the recombination probability of the photoexcited electrons and holes, and thus increases the probability that the charge will migrate to the electrodes. As a result, nanocomposites exhibit higher photoconversion efficiencies. In 1996, N.C. Greenham et al. demonstrated the first conjugated polymer-semiconductor nanocrystal composite photovoltaic (PV) devices using MEH-PPV and90wt% CdSe quantum dots.[99] They showed that nanocomposite materials show highquantum efficiency than those for pure conjugated polymer device. W.U. Huynh et al. reported that controlling nd ur er the nanorod length in poly (3-hexyl-thiophene) (P3HT) resulted The in changes in the distance over which electrons are transported through the thin film roved to 1.7 % which is one of the highest reported powe Figure 2.21 External quantum efficiency of hybrid solar cell and TEM image of the CdSe nanocrystals used in each cell[2]. external quantum efficiency was imp r conversion efficiency for plastic solar cells [2].

PAGE 48

34 The improvement in the external quantum efficiency (EQE) as the nanocrystals length is increased is shown in Fig 2.21. As the data show, the EQE is larger for longer nanorods versus spherical nanoparticles. However, the nanorods tend to lie in the plane of the film, which is not optimum arrangement for charge extraction. B. Sun et. al. repa solar cell power efficiency of 1.8% under AM 1.5 illumination using tetrapod CdSe nanocrystals in poly(p-phenylenevinylene) derivatives [100]. The improvement is consistent with improved electron transport perpendicular to the substrate because orted tetrapated lly insulating spacer, 1,4-trimethyl ammonium butane. generated electrons and holes to od CdSe do not lie flat within the film. Recently, McDonald et al. demonstrinfrared photovoltaics using PbS quantum dots with MEH-PPV in which absorption canbe tunable from 800 nm to 2000nm. [101] 2.4.3 Metal-semiconductor Nanoassembly Semiconductor/metal nanoassembled composite particles have improved the efficiency of photoelectrochemical cells [102] and photocatalysts [103]. N+N+H3NH2CH2CCH2CH2NH3 CH2CH2CH2CH2N N CH3 CH3 CH 3 CH3 CH 3 H3C (a)(b)(c)CdSAu N+N+H3NH2CH2CCH2CH2NH3 CH2CH2CH2CH2N N CH3 CH3 CH 3 CH3 CH 3 H3C (a)(b)(c) N+N+H3NH2CH2CCH2CH2NH3 CH2CH2CH2CH2N N CH3 CH3 CH 3 CH3 CH 3 H3C (a)(b)(c)CdSAu F igure 2.22 Cadmium-gold nanocomposite. (a) The schematic picture of CdS/Au assembly bridged by (b) conducting organic spacer, N,N`-bis(2-aminoethyl)-4,4`-bipyridinium, and (c) electrica The enhanced photo-conversion efficiency of semiconductor/metal nanoassembliesis attributed to the interfacial charge separation of photo

PAGE 49

35 the mn nic ge osslinked by 1,4-trimethyl ammonium e bipyridinium crosslinked systems is attribnic spacer is shown in Figure 2.22 along with the chemicaltructure of conducting and insulating organic spacers. Figure 2.23ng growth of Au onto the tips of CdSe nanorods [105]. etal and semiconductor parts of the nanocomposite. Kolny et al. reported that a nanoassembly of CdS and Au nanoparticles could be achieved by electrostatic interactio[102]. Positively and negatively charged nanoparticles were prepared by coating their surface with charged bifunctional ligand (e.g. 2-(diemthylamino) ethanethiol and 3-mercaptopropionic acid) for self organization. However, the use of functionalized orgaspacers to electrostatistically assemble nanocomposites sometimes reduces the charseparation between Au and CdS nanocrystals due to the insulating character of the organic spacers. Sheeney-Haj-Ichia et al reported that bipyridinium-bridged CdS/Au showed higher photocurrents than CdS/Au that was cr butane [104]. The enhanced photocurrent in th uted to the conducting property of the spacer that enables effective charge separation. A schematic illustration of a CdS/Au assembly bridged by orga s TEM images showi Recently Au nanocrystals were grown on the tip of CdSe nanorods by Mokari et al. (figure 2.23) [105]. Selective tip growth of the Au nanocrystals results from preferential

PAGE 50

36 adsorption of the Au complex onto the rod tips. The tips of the CdSe nanorods are moreactive because of the increased surface energy and imperfect passivation by the surfactants. This leads to preferential growth along the [001] direction and to CdSe nanorods. Ultimately Au nucleates on the edge of the CdSe nanorods. Gold tipped nanostructures could provide natural metal-semiconductor contacts for electrical devices. re

PAGE 51

CHAPTER 3 SINGLE STEP GROWTH OF COLLOIDAL ZNCDSE QUANTUM DOTS3.1 Introduction Colloidal semiconductor nanocrystals can exhibit strong quantum confinement of excitons and allow continuous tunability of the electronic and optical properties by changing their size [5,8,32]. However, nanocrystals less than ~2 nm may grow to final diameters within a few seconds, making control of their size delicate and complicated. In addition, the photoluminescent efficiency of smaller nanocrystals is typically lower than that of larger nanocrystals due to the increased surface-to-volume ratio as the diameter decreases [11,12]. While emission wavelength can be controlled with nanocrystal size, it can also be controlled by composition and a few studies of colloidal ternary alloy nanocrystals, such as ZnCdS [13], ZnCdSe [14] and CdSeTe [15,16], have been reported. Luminescence wavelength was controlled by composition from UV to IR with stable, larger particle size. Excellent luminescent properties were reported, that were comparable to core/shell structured nanocrystals, which have potential for use in optoelectronic devices and biomedical imaging. For synthesis of colloidal ternary semiconductor nanocrystals, better crystal quality and luminescent properties have been reported using molecular precursors [13-16] versus controlled precipitation methods such as reverse micelle[106,107]. Recently, metal oxide precursors have been used with functionalized organic ligands instead of organometallic precursor for a greener approach to synthesis[8-10]. For example, complexes of CdO with alkyl carboxylic acid and alkyl amine has been reported for synthesis of CdSe 37

PAGE 52

38 nanocrystals, rather than a classical highly toxic, pyrophoric, expensive organometal precursor (e.g. Cd(CH3)2). Moreover, metal oxide precursors are well suited for growth studies of colloidal nanocrystals due to their slow nucleation and growth rates[8]. I process using a mixture of metal carte and Cd-oleate) prepared by reacti s d into 2 atmosphere and heate n this study, ternary alloy ZnCdSe nanocrystals were prepared in a single step boxylates (Zn-olea ons of ZnO and CdO with oleic acid. The effects of reaction temperature on crystalgrowth and alloying were determined. 3.2 Experimental Section 3.2.1 Materials Cadmium oxide (CdO, 99.99+ %), Se powder (100 mesh, 99.999 %), oleic acid (90%), trioctylphosphine oxide (TOPO, 99 %) and trioctylphosphine (TOP, 90 %) were purchased from Aldrich chemicals. Zinc oxide (ZnO, 99.999 %) was purchased from Alfa Aesar. All chemicals were used without further purification. 3.2.2 Preparation of ZnCdSe Quantum Dots A quantity of 61 mg of ZnO (0.75 mmole) and 45.8 mg of CdO (0.25 mmole) wadissolved in 2.6 ml of oleic acid at 350 o C for 20 min in a reaction flask and cooled to room temperature. The resultant solid mixture of Zn-oleate and Cd-oleate was loadea reaction vessel containing 3 g of TOPO on a schlenk line under N d to 320 o C with stirring. When the temperature of the reaction mixture was stable, 79 mg (1 mmol) of selenium dissolved in 1.8 ml TOP (Se-TOP) was quickly injected, and the reaction temperature maintained at 320 o C for growth. Aliquots were taken at selected time intervals, quickly cooled and diluted with toluene to stop further growth, and nanocrystals collected by precipitation using methanol/toluene co-solvents. These nanocrystals were characterized using UV-Vis absorption and PL

PAGE 53

39 3.2.3 ZnS Shell Growth on ZnCdSe Nanocrystals The obtained ZnCdSe nanocrystals were loaded into a reaction vessel containing 1g of TOPO on a Schlenk line under N 0 g. For shell xed with Zn oleate solution at room temperature. The resultant mixt was loaded into syringe and ly to reaction vessel containing ZnCdSe nanocrystals over 1.5 hours, maintes himadzu UV-2401PC spectr 2 atmosphere and heated to 180 o C with stirrin growth, Zn oleate was prepared by dissolving a quantity of 41 mg of ZnO (0.5 mmole) in 2.6 ml of oleic acid at 350 o C in a flask and diluted with 10ml TOP after cooling to room temperature. Sulfur precursor (S-TOP) prepared by dissolving 24 mg (0.75 mmol) of sulfur in 3 ml TOP was mi ure of Zn-oleate and S-TOP was injected slow aining reaction temperature at 180 o C for shell growth. Nanocrystals was collected by precipitation using methanol/toluene co-solvents and characterized optical propertiusing PL spectroscopy. 3.2.4 Characterization of ZnCdSe Quantum Dots Absorption spectra were collected with a S ophotometer. Photoluminescence (PL) was measured at room temperature from nanocrystals suspended in toluene using a Fluorolog Tau 3 spectrofluorometer (Jobin Yvon Spex instruments, S.A. Inc.). High resolution transmission electron microscopy (HR-TEM) images were obtained using a JEOL 2010F microscope for lattice imagingand crystal size determination. TEM samples were prepared by dispersing the nanocrystals in toluene and depositing them onto formvar-coated copper grids. X-ray diffraction (XRD) patterns to determine crystal structure were obtained using a Philips APD 3720 X-ray diffractometer. 3.3 Results and Discussion

PAGE 54

40 Figure 3.1 Temporal evolution of photoluminescence spectra at reaction tempe320 rature of ), led o C as a function of reaction time, (a) 2 min (648 nm), (b) 10 min (623 nm(c) 20 min (580nm), and (d) 30 min (567 nm) (excitation at 350 nm, doubpeak). 400500600700800 400500600700800 400500600700800 400500600700800 (c)(b)(a) (d) Wavelength(nm) 400500600700800 400500600700800 400500600700800 400500600700800 (c)(b)(a) (d) Wavelength(nm)

PAGE 55

41 3.3.1 Single Step Growth of Colloidal Ternary ZnCdSe Quantum Dots s described in the experimental section, to synthesize ZnCdSe nanocrystals, appropriate amounts of Zn and Cd oleate were added into the reaction vessel containing TOPO solution. After Se-TOP was injected, the alloying process was monitored by the blue shift of the absorption and PL spectra of the extracted nanocrystals. The wavelength of the band edge emission peak from the ternary nanocrystals decreased from 648 to 567 nm with increasing reaction time at 320oC (see Figure 3.1). This significant blue shift is consistent with the formation of the alloyed ZnCdSe ternary nanocrystals and results from band gaps intermediate between the wide band gap of ZnSe (2.6 eV or 477 nm) and the narrow band gap of CdSe (1.74 eV or 713 nm). The PL emission is a single peak, ruling out separate nucleation and growth of CdSe and ZnSe. However, note that the single peak at 2 min is broad (FWHM = 52 nm) and asymmetric, but is much more narrow (FWHM = 32 nm) and symmetric after 30 min. of reaction (Figure 3.1, curve a, and d, respectively). It is believed that the emission peak is more narrow after long reaction times because the composition of ZnCdSe is more uniform [13]. Figure 3.2 shows optical photographs of the series of the samples under UV-irradiation corresponding to growth timSe nanocrystals (5.7~7.6 nm, see figure 3.8) is larger than the Bohr exciton radius the higher compison between the band gap energy of bulk and nanocrystalline ZnxCd1-xSe alloy as function of the Zn mole fraction (Figure 3.3). Presumably this shift is explained by two factors; (1) the size of the particles is in the intermediate confinement regime (ae> a> ah, where ae and ah are the Bohr radii of electron and holes, respectively) and (2) spatial A es and spectra in Figure 3.1. Although the diameter of ZnCd of CdSe (~5.5 nm) and ZnSe (2.2 nm), we observed a shift in the absorption spectrum to energies than bulk ZnCdSe with same composition. Figure 3.3 shows the ar

PAGE 56

42 UV irradiation. Nanocrystals were grown at reaction temperature of 320 Figure 3.2 Picture of luminescence of ZnCdSe nanocrystals dispersed in toluene under and different reaction time, (a) 5 min, (b) 10 min, (c) 20 min, and (d) 30 min. o C0.20.41.82.02.22.4 0.10.20.30.40.51.82.02.22.4 Eg(Bulk)= 1.74 + (0.89-0.75)x + 0.75x2Energy (eV)Zn Composition ZnCdSe quantum dots Conect(a)ZnCdSe nanocrystals grown at reaction temperature of 320C and different compositions. finement eff(b) Figure 3.3 Dependence of the bandgap energy (E g ) as function of composition of (a) oreaction time, 5 min, 10 min, and 30 min and (b) bulk alloy on their (a)(b)(c)(d) (a)(b)(c)(d)

PAGE 57

43 fluctuation produce atomically abrupt jumps in the chemical potential which localizes the free exciton states. 3.3.2 Effect of Reactivity of Precursor and Reaction Temperature on Optical Properties The bandgap absorption wavelength changed as a function of reaction time for growth at 320oC and 270oC as shown in Figure 3.4. Two distinct regimes are observed; an initial red shift regime, and alloying resulting in a blue shift regime. In the initial stage, the reaction was almost instantaneous and the solution quickly developed to a deep red color at 320 oC under UV irradiation within 1 min. After 1 min. at 320 oC, emission from the solution gradually became yellow green. These data suggest that nucleation and growth h nanocrystagrowth stage was clearly seen when nanocrystals were grown at lower growth temperature (270 oC) because the red shift of emission from the solution persisted over longer (20 min) reaction time before starting blue shift. The conclusion that the initial growth of nanocrystals is Cd-rich is consistent with the x-ray diffraction pattern shown in Figure 3.5. The pattern and 2 values after 10 min. at 270oC (Figure 3.5b) is much closer to bulk CdSe (Figure 3.5a) than bulk ZnSe (Figure 3.5d). For example, 2 values from (001) face of bulk CdSe and nanocrystals at initial reaction stage are 25.33 and 25.66 respectively. The XRD data from nanocrystals grown for 230 min. at 270oC (Figure 3.5c) are shifted towards bulk ZnSe. Note however that the band gap adsorption of the initial d the XRD pplace. The initial red shift for initial growth time may result from larger particles as of Cd-rich ZnCdSe is dominant during the first minute of the reaction rather than Zn-ricls, leading to red shift of emission due to increment of size. Red shift in initial nanocrystals (650 nm) is blue-shifted from the band gap of bulk CdSe (713 nm), aneaks are at higher 2 values, both suggesting that alloying has already taken

PAGE 58

44 Figure 3.4 Band gap absorption change as a function of reaction time, (a) reaction temperature 320 C and (b) reaction temperature 270 C. Inset represenmagnified plot of dotted circle in (a). Figure 3.5 X-ray diffraction patterns of (a) bulk CdSe, (b) Cd rich ZnCdSe nanocrystalsobtained grown for 10 min. at 270C (c) ZnCdSe nanocrystals grown for 230 min. at 270 oots the o o C and (d) bulk ZnSe 2030405060 (a) Cots (arbit)2 Theta (c) un. un(b) (d) 050100150200250530540560580600610620 550570590670 630640650 660 050100150200250540560580600620 640 660 Badgapbsortion(n(b)0246810600610 n ApmReaction time(min)620650660670 )(a)630640 0246810600 620660 640 050100150200250530540560580600610620 550570590670 630640650 660 050100150200250540560580600620 640 660 Badgapbsortion(n(b)0246810600610 n ApmReaction time(min)620650660670 )(a)630640 620660 0246810600 640

PAGE 59

45 longer reaction time, reducing quantum confinement effects. Cd-rich growth in the initial stage would suggest that the Cd oleate is more reactive than the Zn oleate towards Se-TOP. Growth of the II-VI semiconductor nanocrystals results from the reaction of metal precursors with the Se precursor via elimination of chemically bonded organic molecules (figure 3.6). R; CH3(CH2)7CH=CH(CH2)6M; Zn, Cd etc nM(OOCR)2 + nSe-TOP (M-Se)n(OOCR)m -(OOCR)n-mnTOP Figure 3.6 Schematic reaction of precursor for nanocrystal formation If the bond strength between the metal and organic ligand is sufficiently high, the rate limiting step is elimination of the organic ligand. The bond strength is therefore ligand bonof II-VI precursor for organometallic hemical vapor deposition (MOCVD) (table 1) [108]. Table 3.1 Calculated metal-chalcogen bond energies in metal alkyl chalcogen precursors with the composition of M(ER)2 (kcal/mol; from reference [108]. M/E C O S Se Te related to the activation energy for the crystal growth. Cundari et al. calculated metal-d energies for the study of the reaction c Zn 37 71 58 50 40 Cd 36 57 50 44 36 Hg 31 44 41 35 28 M : metal, E : chalcogen R: alkyl From data in table 3.1, metal-ligand bond energies decrease in the order Zn > Cd > Hg for bonding to the anions reported, including oxygen and selenium. Therefore, the activation enthalpy for nucleation of CdSe is expected to be lower than that of ZnSe, consistent with the present data that indicate that the Cd-oleate reacts faster with Se-TOP than Zn-Oleate. Bailey and Nie have reported, consistent with data in Table 1, that

PAGE 60

46 reaction between Cd and Se and Te precursors resulted in a core/shell structure of in a single step reaction [15]. In addition, the data in table 1 indicate that the CdTe/CdSe, is nocrystals, stabilizing and preventing Ostwald ripening [10]. 3.3.3 Effect of Reactivity of Precursor on Particle Growth On the other hand, it was observed that red shift of band gap absorption was faster and developed to 655nm in initial state of reaction at high reaction temperature, whereas This indicates that nucle in er growth. Subsequent mexylate ursor such Cd acetate was repo that only ~10 % of Cd precursor was consumed during the nucleation. [9, 10]Therefore, the nanocrystals continue to grow consuming free precursors from surrounding solution and large amount the crude reaction solution. In this case, the growth rate o bond energy of metal-oxygen is ~1.5-2 times higher than that of metal-carbon, where metal is Zn, Cd or Hg. Therefore the reactivity of metal carboxylate, such as Cd oleatelower than that of precursor such as Cd(CH 3 ) 2 leading to slower growth and better development of nanocrystals. In addition, oleic acid acts also as capping agent on the surface of the na absorption was slowly changed to 620nm at low reaction temperature. ation and growth is substantially faster, followed by the formation of larger crystal at higher temperature. If almost precursor has been involved in the nucleation processcase of finite volume of the solution, which is considered in the case of using highly reactive Cd(CH 3 ) 2 it is expected that particle sizes are small at high temperature due to fast nucleation [10] and depletion of precursor for furth nanocrystals grow via the Ostwald ripening reaction. However, in case of less reactive tal carbo prec as it rted of free precursors is still present in f a single nanoparticle depends strongly on reaction temperature to overcome activation energy for further growth. The concentration of monomer in solution close tothe nanoparticles surface significantly affects the growth rate [56]. At higher

PAGE 61

47 temperature, flux of precursors at the surface of nanoparticles increases due to high diffusivity of unreacted precursors. Growth rate increases also due to fast eliminatiochemically bonded organic ligand and high concentration of precursors, resulting ingrowth and larger particle size. Furthermore diffusivity of precursors becomes more important factor in case of ternary system in which two different types of precursors arinvolved, because nanocrystals may be shielded by layer of less reactive Zn oleate when n of fast e (a)(b) 500550600650 548 nm 500550600650 Wavelength (nm) 567 nm (c)(d) Wavelength (nm) (a)(b) 500550600650 548 nm 500550600650 Wavelength (nm) 567 nm (c)(d) temperatures (a) 270 Wavelength (nm) Figure 3.7 Comparison of the size of ZnCdSe nanocrystals grown at different bar). Photoluminescence spectrum of ZnCdSe nanocrystals obtained (c) at 270 ft. o C for 230 min. and (b) 320 o C for 30 min (20 nm scale o C for 230 min. and (d) at 320 o C for 30 min. after no further spectral shi

PAGE 62

48 Cd oleates form crystals. From the reasons described above, the size of ZnCdSe nanocrystals obtained after no further luminescence shift is larger at 320 o C (~7than 270 .6 nm) is d to gradient-induced diffusion, consistent with the blue shift and narrowing of PL emission and shift of x-ray diffraction peak to larger 2 theta angles with the increase of Zn in the ZnCdSe nanocrystals Figure 3.8 HR-TEM images of ZnCdSe nanocrystals grown at 320oC as reaction time increase, (a) 2 min (b) 5 min, and (c) 30 min (20 nm scale bar). Average particle sizes (diameters) are (a) ~5.2 nm, (b) ~5.7 nm and (c) ~7.6 nm with aspect ratio 1.5. 3.3.4 Shell Growth of ZnS on ZnCdSe Quantum Dots A proven strategy for increasing both luminescence QY is to grow a shell of a higher band gap semiconductor on the core nanocrystals such as CdSe/ZnS [6,37]. Higher luminescence QY can be achieved by inorganic shell growth due to reduce nonradiative decay channels that originate from surface-trap electronic states. Not only effective o C (~4 nm), as shown in figure 3.7. After the initial nucleation and growth of CdSe, nanocrystals continue to grow withreaction time. HR-TEM image in Figure 3.8 shows the crystal size and prolate shape evolution of ZnCdSe nanocrystal at 320 o C. The increase of particle size presumably the result of predominant consumption of Zn and lesser consumption of Cd precursors. Therefore, it is postulated that Zn-rich layer on the outer of Cd-rich ZnCdSe nanocrystals le (a)(b)(c) (a)(b)(c)

PAGE 63

49 4505005506006507000.066.0x106 2.0x1064.0x10 8.0x1061.0x10 71.2x10 71.4x107 450500550600650700 450500550600650700 450500550600650700 21nm Shift (b) (a) ZnCdSe(#52) QDenWavelength(nm) Intsi ty ZnCdSe/ZnS core/shellgrowth on ZnCdSe q Figure 3.8 Photoluminescence spectra (a) before ZnS shell growth and (b) after ZnS shell uantum dots surface passivation, higher band gap shell materials provide also a potential step for electron and holes originating in the core nanocrystals, reducing the probability for the carriers to sample the surface. Figure 3.8 shows that luminescence intensity increase significantly when ZnS shell was grown on ZnCdSe nanocrystals. In addition, when ZnS surrounds the ZnCdSe nanocrystal, it is observed ~21 nm energy shift in emission to tunnels intot energy and consequently the energy of luminescence [6]. 3.4 Conclusion Colloidal ternary alloy ZnCdSe nanocrystals have been synthesized by a single lower energy than that of ZnCdSe nanocrystals. The wave function of lighter electron shell [6]. The increased delocalization of the electron lowers its confinemen

PAGE 64

50 step reaction in a mixture of Cd and Zn oleates with Se-T OP. After an initial red-shift due toabsorption was attributed to growth of the nanocrystal size, while the subsequent blue shift was attributed to diffusion of Zn from the outer layer into the Cd-rich core, to form the final ZnCdSe particles between 5 and 8 nm in diameter after reacting for 2-30 min. at 320oC. Fast growth and larger particle size was observed at higher reaction temperature due to fast elimination of chemically bonded organic ligand and high concentration of precursors. Luminescence intensity increase significantly when ZnS shell was grown on ZnCdSe nanocrystals. In addition, when ZnS surrounds the ZnCdSe nanocrystal, it is observed ~21 nm energy shift in emission to lower energy than that of ZnCdSe nanocrystals. an increase of size of Cd-rich ZnCdSe nanocrystals, a blue shift of PL was observed upon diffusion of Zn into the nanocrystals to form a larger band gap ternary alloy. The initial nucleation of Cd-rich ZnCdSe nanocrystals was attributed to the higher reactivity of Cd oleate as compared to the Zn oleate. The initial red shift of band gap

PAGE 65

CHAPTER 4 SYNTHESIS AND CHARACTEROLLOIDAL TERNAY ZNCDSE SEMICONDUCT ersity um ed Se minimize decay channels resulting from electronic surface states. The luminescent quantum yield of as-grown CdSe nanorods is reported to be 1% at room temperature, nd increases to 20 % when a shell of a larger band gap material (CdS or ZnS) was grown on the surface of the core [21,22]. On the other hand, Zhong et al. reported that ternary ZnCdSe quantum dots show high photoluminescence (PL) efficiency comparable IZATION OF COR NANORODS 4.1 Introduction Colloidal semiconductor nanocrystals have been intensively studied due to the excellent luminescent properties, size tunable optical properties, and their high divfor a variety of optoelectronic application such as light emitting diodes (LEDs) [1] andphotovoltaics [2,20]. Furthermore, colloidal nanocrystals have been considered to be the building block of future nanotechnology for the fabrication and investigation of quantsuperstructures [3]. Recently, development of synthetic methods for growth of rod shapCdSe nanocrystals promise new opportunities to study shape-dependent electronic and optical properties such as polarized LED [17, 18], lower threshold laser [19], and more efficient photovoltaic device [2]. However, rod shaped CdSe nanocrystals still are reported to have low quantum efficiencies [17, 21, 22] and weak confinement along therod axis leading to difficulties in efficient production of blue-green light. One approach to improve the emission efficiency is to coat the surface of the Cdnanorods with a larger bandgap shell material to confine the charge carriers and the nonradiative a 51

PAGE 66

52 with those from CdSe/ZnS core shell quantum dot [13,14]. A solid solution of two binary semiconductors may exhibit intermediate energy band gaps between those of the constituaking them attractive materials for op, with a range of compositions and thhave been grown by MBE and their died D and chara 4.2.1 Materials Cadmium oxide (CdO, 99.99+ %), selenium powder (100 mesh, 99.999 %), oleic acid (90 %), trioctylphosphine oxide (TOPO, 99 %) and trioctylphosphine (TOP, 90 %) were purchased from Aldrich chemicals. Tetradecylphosphonic acid (TDPA) was ent binaries, m to-electronic devices. A variety of II-VI ternary alloy semiconductors, such as ZnCdSe erefore band gaps, emitted wavelength shown to cover the whole visible range. They have been stufor LEDs and laser diodes (LDs) for use in optical DVDs and displays [67-70, 109]. Colloidal ZnCdSe nanorods could also be used in optoelectronic devices since quantum confinement may allow efficient emission at wavelengths that cover the visible spectrum. In this study, green-yellow emitting ZnCdSe nanorods were prepared by diffusion of Zn into the CdSe core. The ZnCdSe nanorods were characterized by absorption spectroscopy, photoluminescence, X-ray diffraction (XRD), high resolution-TEM (HR-TEM) and Raman spectroscopy. Raman spectroscopy is an excellent probe ofthe nanostructure [110-115]. The Raman data will be shown to complement the XRHR-TEM data on the ZnCdSe nanostructure. In addition, in order to elucidate the energyrelaxation and recombination dynamics in nanoparticles, we present here a cterization of the luminescence decay and dynamics of CdSe/ZnSe coreshell nanorods and ZnCdSe nanorods using time-resolved photoluminescence (TRPL) and femtosecond transient absorption (TRA) 4.2 Experimental Section

PAGE 67

53 purchased from PolyCarbon Industries, Inc. and zinc oxide ( ZnO, 99.999 %) was purch. The son. r tirring for into n vessel was heated with stirring to 270oC for up to 3 hrs. After heating foriately cooled and diluted alloying, then was precipitated with methanol/toluene co-solvents. 4.2.3 he ased from Alfa Aesar. All chemicals were used without further purification. 4.2.2 Preparation of ZnCdSe Nanorods CdSe nanorods were synthesized using the method described by Peng [85]. In this method, 0.205g (1.6 mmol) of CdO, 2.903 g of TOPO and 0.8928 g of TDPA were heated in a three-neck flask on a Schlenk line under a N 2 atmosphere to 350 o C while stirring. After the solution became optically clear, it was cooled to room temperatures olid Cd-TDPA complex was used after aging for 24 hr without further purificatiThis Cd-TDPA complex was heated in a three-necked flask under a N 2 atmosphere to 280 o C while stirring, and 0.126 g (1.6 mmol) of selenium dissolved in 2.9 ml TOP was injected quickly. After injection, the temperature of the mixture was kept at 250 o C fothe 30 min growth of CdSe nanorods, and then cooled to 180 o C. For shell growth, 0.1302g (1.6mmol) of the ZnO was dissolved in 2.03 ml of oleic acid (Zn-oleate) at 350 o C and cooled to room temperature, and then 1ml of TOP was added to prevent solidification. In addition, 0.126 g (1.6 mmol) of Se was dissolved in 2.9ml of TOP (Se-TOP). The Zn-oleate and Se-TOP solutions were mixed by s10min at RT, and this mixture was loaded into a syringe and injected drop-by-dropthe reaction flask over 1.5 hr. After injection was complete, the solution was stirred at RT for another 10min. For alloying, the reactio 1, 2 or 3 hrs, a sample was immed with toluene to stop Characterization High-resolution transmission electron microscope (HR-TEM) images were collected using a JEOL 2010F microscope for imaging and direct determination of t

PAGE 68

54 average and distribution of the nanorod dimensions. To prepare TEM samples, thenanocrystals were dispersed in toluene and deposited onto formvar-coated copper g rids. diffrae von m l arization is tilted by 45 degrees with respect to the pump pulse usingut is X-ray diffraction (XRD) patterns were obtained using a Philips APD 3720 X-ray ctometer and used for determination of both the crystal structure and size. Raman spectra were measured at 300K in the backscattering geometry, using the 532 nm linfrom a Verdi 8 doubled Nd-YAG solid state laser in a Ramanor U-1000 Jobin-Yvon Raman spectrometer. Absorption spectra were collected with a Shimadzu UV-2401PC spectrophotometer. Photoluminescence (PL) was measured at room temperature using nanorods suspended in toluene using a Fluorolog Tau 3 spectrofluorometer (Jobin YSpex instruments, S.A. Inc.). The PL quantum yield (QY) was determined using Rodamine 6G organic dye standard with a known QY of 95%. Time resolved spectra were recorded using a spectrograph attached to a charge-coupled device (CCD) (Shamrock 303i). A commercial Ti-Sapphire (Ti-Sa) laser systeconsisting of a Ti-Sa oscillator (Tsunami, Spectra-Physics) and subsequent amplifier (Spitfire, Spectra-Physics) with a repetition rate of 1 kHz was used to pump and opticaparametric amplifier (OPA) to generate 400 nm excitation pulses. The relaxation processes of the colloidal nanocrystals were explored using femtosecond transient absorption (TRA) based on the same laser system described above. A part of the amplifier output is split off to pump with a 1 mm CaF 2 window to generate a white light continuum probe that ranges from 300 to 900 nm. Prior to white-light generation, the probe pol a thin-film polarizer. Specifically, the fourth harmonic of the OPA idler outp

PAGE 69

55 used to produce excitation pulses (pump) at 450 nm. This beam is then fed through prism compressor, resulting in pulse lengths less than 100 fs (FWHM). The diameter of the pump was ~100 m and was set to low energy (~ 39 to 45 nJ) to avoid biexcitoformation. Prior to interaction in the sample, a fraction of the probe pulse is split off andmeasured as a reference beam. The pump pulses were modulated by an optical choppera frequency of 500 Hz and passed through a computer-controlled optical delay line todelay the probe relative a n at to the pump. The pump and probe beams were temporally and spatiather pump pulse. (4-1) the al simultaneously: TA = (A 2A)/3 (4-2) The parallel and perpendicular transmitted signals and reference were focused into a spectrograph attached to a charge-coupled device (CCD) (Shamrock 303i) for detection. Sample solutions for TRA measurements were placed in a quartz cuvette with a 2 mm path length and continuously stirred to guarantee excitation of a new sample volume with every laser shot. The set up for TRPL is same as TRA except rather than focing the probe lly overlapped into the sample. The optical chopper blocks every oThe transmitted signal, T, is the probe pulse in the presence of the pump, and T 0 is the transmission of the probe pulse in absence of the pump. The pump induced absorption changes to be detected are: T/T = T-T 0 /T 0 A Glan-Thompson polarizer splits the transmitted signal, with and without pump pulse, into its polarization components, parallel (A || ) and perpendicular (A ) with respect to the pump. TA experiments performed at the Magic angle eliminate directionand rotational influences on the signal. Therefore, the Magic angle signal (TRA) is calculated from the parallel and perpendicular components measured ||

PAGE 70

56 beam on the camera, we collect the fluorescence from the sample using a 2 inch lens athe focusing this into the camera. We used a 4 nanosecond gate and we excited at 400 nmwhich is the second harmonic of the Spitfire output. 4.3 Results and Discussion 4.3.1 Synthesis of ZnCdSe Nanorods The combination of TDPA/TOPO surfactants has been used to prepare anisotropnanocrystals because it was expected to raise the surface energy of some crystal facets relative to other facets of the nanocrystals because of their strong binding energies with metal ion and their steric effects [84, 116]. Initial attempts to synthesize ZnCdSe nanorods using a mixture of Zn-TDPA and Cd-TDPA as metal sources in TOPO solutiowere unsuccessful. This failu nd ic n re may have resulted from the fact that these two metal sourcbecause of their dSe-TOSe ised to 270 oC to allow diffusion and alloyed ZnCdSe nanorods to form. Steady, slow additinucle es did not lead to crystallite shape control ifferent reactivity with P [38,117]. Furthermore, higher temperatures and much longer reaction times were required for complexation of ZnO versus CdO with TDPA. To avoid this problem in the preparation of ternary alloy ZnCdSe nanorods, CdSe nanorods were first synthesized using TDPA/TOPO surfactants. Following growth of the CdSe nanorod core, the Znshell was grown from the Zn-Oleate and Se-TOP mixture. Then the temperature was ra on of the Zn-oleate and Se-TOP mixture was required to avoid homogeneous ation of ZnSe during shell growth. In addition, careful temperature control was required because alloying led to a blue shift of emission when growth of the shell was attempted at a temperature > 210 o C. However at a temperature of <170 o C, the Zn-oleatcomplex was too slow to react with TOP-Se and grow a shell, resulting in very weak emission. It was important to slowly increase the temperature to 270 e o C for diffusion in

PAGE 71

57 2030405060 2030405060 2030405060 Count s (112)(110)(102)(103) core/shell nanorods, and (c) ZnCdSe nanorods. (002) (101)(100)(c)(b)(a)2ThetaFigure 4.1 Powder X-ray diffraction patterns of (a) CdSe nanorods,(b) CdSe/ZnSe

PAGE 72

58 order to avoid temperature over-shoot. 4.3.2 Structure of ZnCdSe Nanorods The powder X-ray diffraction patterns from hexagonal CdSe, CdSe/ZnSe, and ZnCdSe nanorods are shown in Figure 1. For all three materials, the width of the (002) diffraction peaks is small compared to those from the (100) plane, consistent with less broadening of the diffraction peak from planes perpendicular to the long axis of the rod-shaped nanocrystals. The lattice spacing along the c-axis was 7.01 6.94 and 6.77 for CdSe, CdSe/ZnSe, and ZnCdSe, respectively. The c-axis lattice parameter reported above for the CdSe/ZnSe core/shell structure is 1% smaller than that for CdSe alone. It has been reported that the CdSe core is under compression due to the growth of a ZnS shell with an 11% smaller lattice parameter, leading to smaller d-spacing and larger 2 values [21]. Similarly, the CdSe core should be compressed by the ZnSe shell due to its smaller lattice parameter, but in this case the lattice mismatch is smaller at ~7 %. Furthermore, alloying between CdSe and ZnSe is expected to reduce both the lattice parameter and any strain induced by lattice mismatch. The smaller lattice parameter reported above for ZnCdSe alloyed nanorods must result ation at would increase the lattice parameter. The XRD data in Figure 4.1 show that all iffraction peaks shifted to a larger 2 (smaller interplanar spacing) upon alloying of dSe with Zn, consistent with expectations. The XRD data from spherical versus rod-like nanocrystals of ZnCdSe are ompared in Figure 4.2. While diffraction peaks from the same planes are observed, the tensity of diffraction from the (102) and (103) planes is reduced compared to the (110) predominantly from the lattice contraction upon interdiffusion, not from strain relax th d C c in

PAGE 73

59 2030405060 2030405060 (b)d Conts u (a) 2Theta Figure 4.2 Comparison of powder X-ray diffraction patterns of (a) ZnCdSe nanorods, an(b) spherical ZnCdSe dots. 01234567891011121314151617181920 010304050608090 2070100 LengthDiameter 01234567891011121314151617181920010304050608090 Count2070100 nm 01234567891011121314151617181920010304050608090 2070100 LengthDiameter 01234567891011121314151617181920010304050608090 Count2070100 nm Lattice fringe from a nanorod is shown by the lower right inset. 20 nm scale bar 5 nm scale bar Figure 4.3 HR-TEM image and histogram of size distribution of ZnCdSe nanorods.

PAGE 74

60 and (112) planes of ZnCdSe nanorods. Attenuation of the (102) and (103) diffraction peaks probably results from a higher density of stacking faults along the (002) axis of spherical versus rod nanocrystals [5,12,118]. A HR-TEM image of the ZnCdSe nanorods is shown in Figure 4.3, along with a histogram of the diameter and lengths of the nanorods measured from such images. The average diameter is 6 nm and the average length is 13 nm. The particle sizes of ZnCdSe nanorods calculated from XRD data using the Debye-Scherrer equation were a diameter of 5.5 nm and a length of 11.8 nm, which agree well with the HR-TEM data shown in Figure 4.3. 4.3.2 Effect of Alloying on the Phonon Spectra Raman spectroscopy has been used to study the structure of nanocrystals, including the core/shell interface, through the dependence of the phonon frequencies upon he Raman peaks detected from CdSe nanorods are shown in Figure 4.4(a). The peak at ~206cm-1 is from the LO phonon, and is 4cm-1 lower in wave number than that reported for corresponding bulk CdSe (210cm-1) [112,114], presumably due to the confinement of the optical phonons in the nanorods [112-114]. Surface phonon vibration, which is observed as the shoulder to the left of the main Raman peak (at ~180cm-1) is attributed to the surface phonon for CdSe, that is detected because of the non-spherical geometry of the CdSe nanorods [110,111]. Similar data from as-grown CdSe/ZnSe core/shell nanorods are shown in Figure 4.4(b). In addition to the bulk LO phonon from the CdSe core (~206 cm-1) and the ZnSe the Raman -1e LO compositional homogeneity [113,114]. T shell (~247 cm -1 ), limited formation of interfacial ZnCdSe is indicated by peak at ~235 cm. Unresolved Raman peaks (shoulders) on the both side of the CdS

PAGE 75

61 150175200225250275 iZnSeCdSe(b) Raman Inensity ZnCdSei t Raman shift(cm-1) 150200250 (a) tensy Init-1 Raman shift(cm) 150175200225250275 iZnSeCdSe(b) Raman Inensity ZnCdSei t Raman shift(cm-1) 150200250 (a) tensy Init-1 Raman shift(cm) Figure 4.4 Raman spectra of LO phonon mode of (a) CdSe nanorods and (b) CdSe/ZnSe core-shell nanorods.

PAGE 76

62 Figure 4.5 Raman LO phonon spectra of ZnCdSe nanorods after annealing at 270 oC for (a) 1, (b) 2, or (c) 3 hrs. 180200220240260280 180200220240260280 180200220240260280 Raman shift (cm-1)Intensity (arb. unit)(c)(b)(a)

PAGE 77

63 p eak (206 cm-1) are attributed to the isolated atom-impurity modes of Zn in CdSe (190 m-1) and Cd in ZnSe (218 cm-1) [119]. he effects of alloying time (1, 2 or 3 hrs at 270 oC) on the Raman spectra are shown in Figure 4.5. After heat treatment, a single phonon mode is detected for the ZnCdSe nanorods at 223 cm-1, 228 cm-1 and 226 cm-1 (for 1, 2 and 3 hrs, respectively), similar to the one phonon-mode behavior for bulk ZnCdSe [120,121]. The single phonon mode suggests that the interface between CdSe and ZnSe has disappeared, the isolated atom-impurity modes have disappeared, and the reasonably sharp single-mode peak is consistent with a uniform composition and particle size distribution. Note however that the Raman peak for ZnCdSe annealed for one hour is considerably broader than those from the samples annealed for 2 and 3 hrs due to compositional disorder. The differences between the single mode peak positions at 1 versus 2 hours (5 cm-1) results from a continuation of the alloying process. The difference in peak position between 2 and 3 hrs (2 cm-1) ial disorder[122,123] and stress relaxation by ermal annealing[115]. The PL spectra from CdSe core and CdSe/ZnSe core/shell nanorods, with peaks at 642 nm and 638 nm respectively, are shown in Figure 4.6. The PL quantum yield (QY) for CdSe core versus CdSe/ZnSe core/shell nanorods was 0.6% versus 15%. The increased PL QY was presumably due to passivation of non-radiative surface states on the CdSe nanorods by the ZnSe shell. The 4 nm blue shift of PL from CdSe/ZnSe core/shell versus CdSe core nanorods could result from electron localization or nanorod size effects in the core/shell structures. Contrary to the current blue shift, Mokari and c T s attributed to composition th 4.3.3 Photoluminescence and Absorption Properties

PAGE 78

64 Figure 4.6 Photoluminescence spectra of (a) CdSe-ZnSe core/shell nanorods and (b) CdSe nanorods. 0.200.25 Figure 4.7 UV-Vis absorption spectra of (a) CdSe nanorods, (b) CdSe/ZnSe core-shell 3004005006007000.000.15 nanorods, and (c) ZnCdSe nanorods alloyed at 270 o C for 3hrs. 0.050.10 3004005006007000.000.15 0.050.10 0.200.25 3004005006007000.000.15 0.050.10 0.200.25 Wavelength(c)(b)(a) sorb Aba66665x1066x106 45050055060065070075080001x102x103x104x10 45050055060065070075080001x102x103x104x10 66665x1066x106 IntensWavelength(nm)(b) ity(a) nce

PAGE 79

65 Banin [21] reported a ~10nm red shift for a CdSe/ZnS core/shell structure, and attributed this shift to tunneling of the electron wave function into the ZnS shell. This tunneling led to a delocalization of the electron, lowering its confinement energy and consequently the energy of the exciton levels[6]. Increased localization would be expected for compressively stressed core/shell particles as suggested above based on XRD data, resulting in a blue shift. In addition, the Raman data above suggest the formation of interfacial ZnCdSe in as-grown CdSe/ZnSe core/shell nanorods. This reaction would be expected to decrease the size of the CdSe core[124] resulting in increased localization and a blue shift in emission. Finally, as reported below, alloying results in a blue shift of structure reation of these effects. There were also significant ifferences in the absorption spectra for CdSe core, CdSe/ZnSe core/shell, and ZnCdSe alloyed nanorods, as shown in Figure 4.7. For CdSe core and CdSe/ZnSe core/shell nanorods, the initial absorption edge was at 650 nm and 645 nm, respectively, in agreement with the emission peaks in Figure 4.6. This absorption peak is thought to result from interband optical transitions such as 1S [1S(e)-1S1/2(h)], 2S [1S(e)-2S3/2(h)] and 1P [1P(e)-1P3/2(h)], where e and h represent electron and hole quantized states [32]. The second absorption structure in Figure 4.7 for CdSe and CdSe/ZnSe nanorods (520 nm) is thought to represent 1P transition. The energies of the corresponding absorption features for alloyed ZnCdSe (3 hrs at 270 oC) is blue shifted considerably to 555 nm and 465 nm, consistent with PL emission data reported in Figure 4.8. Presumably these reflecting te the the PL peak. Presumably the observed blue shift from formation of the core/shell sults from a summ d features originate from transition similar to those in the core and core/shell nanorods, he larger band gap resulting from the formation of ZnCdSe. Whil

PAGE 80

66 adsorption features from nanorods shown in Figure 4.7 are easily detected, they are not as sharp and as well resolved as features reported for CdSe quantum dots [32]. Presumably this results from the fact that a distribution of energy levels will result in the conduction and valence bands due to the increased degree of freedom for nanorods as comquantum dots [33,125]. In addition, the compositional disorder detected by Raman dwill lead to broader features in the absorption and emission spectra. Absorption featudescribed above was studied in detail using time resolved absorption spectroscopy (see section 4.3.5) The blue shift in absorption upon alloying to form ZnCdSe is consistent with the PL emission spectra shown in Figure 4.8.Upon annealing at 270 pared with ata re om 65 nm ith a 2 ~8, d ly tron o C, the PL peak frCdSe/ZnSe core/shell nanorods (638 nm) was shifted to 610, 570 and finally to 5with the alloying times of 1, 2 or 3 hrs, respectively. Note that the PL peak after 1 hr of annealing is not only shifted, but is skewed to the short wavelength side, consistent wrange of compositions as indicated by the breadth of the Raman peak. The peak afterhrs of annealing is still broad as compared to the breadth after 3 hrs of annealing. The PL peak after 3 hrs of annealing is also narrower than the peak from the CdSe/ZnSe core/shell nanorods, although it is less intense. The PL QYs of the ZnCdSe nanorods (5 and 10 % for 1, 2 and 3 hrs anneals, respectively) were lower than the QY of 15 % reported above for CdSe/ZnSe core/shell structures and higher than that for CdSe nanorods (0.6 %). The increased QY for ZnCdSe versus CdSe probably results from the compositional disorder discussed above. Composition disorder in ternary alloy nanorostructures will lead to localization of excitons [126]. Such localization effects apparentimprove the luminescence efficiency by increasing the overlap integral of the elec

PAGE 81

67 d Figure 4.8 Photoluminescence spectra from (a) CdSe/ZnSe core/shell nanorods anZnCdSe nanorods alloyed at 270 o C for (a) 1, (b) 2, and (c) 3 hrs. 400450500550600650700750800 4004505005506006507007508001x106 02x1064x106x10 3x10665x1066 400450500550600650700750800 400450500550600650700750800 400450500550600650700750800 Inte(d)(b)(a) (c) nsity Wavelength(nm)

PAGE 82

68 a nd hole wavefunctions. The decreased QY from ZnCdSe versus the ZnSe/CdSe core/shell nanorods probably results from the lack of surface passivation on the ZnCdSe nanorods. This is consistent with the QY decreasing after 1 and further after 2 hrs of annealing, since diffusion will be reducing the gradient in Zn (i.e. reducing the high concentration at the surface and increasing the low concentration in the middle of the nanorods). However, annealing for 3 hrs increased the QY over that from samples annealed for 1 or 2 hrs. This increased QY is attributed to annealing of crystalline defects and reduction of stress, consistent with the Raman data reported above. This is also consistent with the fact that the PL peak from ZnCdSe annealed for 3 hrs at 270oC has the narrowest FWHM as shown in Figure 4.8. Crystal defects are known to act as nonradiative recombination centers, reducing the emission efficiency [13,127]. 4.3.4 Time Resolved Photoluminescence (TRPL) Study The TRPL measurements of the nanorods samples were carried out at room temperature. Figure 4.9 shows PL decay curves of the nanorods samples. At early times the PL probes particles with fast decaying times while at latter times it probes those with ples, and this non-exponential decay can be well d exponential function for disordered low dimensional iconductors[128]: longer rates. It is found that the decay process is characterized by a non-exponential function at all nanorods sam characterized by a stretche sem tItIoexp)( (4-3) where I(t) is the PL intensity at time t and Io is I(0); is a dispersion factor and is ission decay time. Several relaxation phenomena in complex condensed matter systems have been found to follow the stretched exponential decay law [128]. em

PAGE 83

69 -20020406080100120140160180 Intenity (lg I/I soo) CdSe/ZnSe Core/Shell nanorods ZnCdSe alloy nanorods 1hr ZnCdSe alloy nanorods 2hr ZnCdSe alloy nanorods 3hr Time (ns) Figure 4.9 TRPL decay curve of CdSe/ZnSe nanorod and ZnCdSe nanorods

PAGE 84

70 012345-4-3-2-10 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr-4.0 0.00.51.01.52.02.53.03.54.04.5-3.5-3.0-2.5-2.0-1.5-1.0-0.50.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345-4-3-2-10 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345-4-3-2-10 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K= 0.75= 173nsT=300K= 0.58= 277nsT=300K= 0.48= 501nsT=300K= 0.58= 276ns(a)(b)(c)(d)012345-3-2-10 -4 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr-4.0 0.00.51.01.52.02.53.03.54.04.5-3.5-3.0-2.5-2.0-1.5-1.0-0.50.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345-4-3-2-10 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345-4-3-2-10 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hr012345-3-2-10 -4 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr-4.0 0.00.51.01.52.02.53.03.54.04.5-3.5-3.0-2.5-2.0-1.5-1.0-0.50.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345-4-3-2-10 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345-4-3-2-10 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K= 0.75= 173nsT=300K= 0.58= 277nsT=300K= 0.48= 501nsT=300K= 0.58= 276ns(a)(b)(c)(d) igure 4.10 The dot lines are experimental data and the full lines are the fitting data using equation ln[ln(Io/It)] versus ln(time) of (a) CdSe/ZnSe coreshell nanorods, (b) ZnCdSe alloy nanorods 1hr, (c) ZnCdSe alloy nanorods 2hr, and (d) ZnCdSe alloy nanorods 3hr. F

PAGE 85

71 T he parameter and depend on the material and the specific phenomenon under onsideration and can be a function of external variables such as temperature. For the miting case of 1, we get the single exponential decay with the characteristic life time For single QD, we can expects =1. It should be mentioned that <1 result from superposition of many exponential decays. This decay law is often encountered in the disordered systems and considered as a consequence of the dispersive diffusion of the photoexcited carriers[128-131]. In general, carrier dispersive diffusion among different spatial regions can be due to energetic disorder, or due to the topological disorder. The alloying process is not uniform and leads to broad PL spectra. Emission of an inhomogeneous population (different sizes, shapes or composition) leads to the simultaneous probing of particles with different decaying rates. A stretched exponential function can be used to describe theses systems. Figure 4.10 shows a replotting of the data for the nanorod samples in the form of an ln[ln(Io/It)] versus ln(time) plot and fitting Table 4.1 Comparison of and value of CdSe/ZnSe coreshell and ZnCdSe alloyed nanorods Nanorods em(nm) (ns) c li CdSe/ZnSe core shell 645 0.75 173 ZnCdSe 1hr 625 0.58 277 ZnCdSe 2hr 570 0.48 501 ZnC dSe 3hr 566 0.58 276 sing linear function. value can be independently determined from by plotting the ouble logarithm of the signal versus the logarithm of the time. Obtained values are ummarized in table 4-1. The fitted of CdSe/ZnSe coreshell nanorod is ~0.75, which u d s

PAGE 86

72 reflects higher degree of ordered crystals. Difference of value between 1 and ~0.7might be mainly due to size distribution. Interestingly, this value is matched to reportevalue of in quantum well structure of CdSe-ZnSe superlattice (= ~0.75) in quantum well structure of CdSe-ZnSe superlattice (= ~0.75) [129]. Comparing CdSe/ZnSe coreshell nanorods to alloy ZnCdSe nanorods, the values are significantly decreased from 0.75 to 0.48~0.58. If we assume that size distribution is not significantly changedduring alloying process, it is clearly seen that decreasing results from compositional disorder in nanorod crystals such as spatial fluctuations of the local Zn concentration. However, annealing for 3 hrs increased the value over that from samples annor 2 hrs. This increased is attributed to compositional homogeneity of the sampis consistent with photoluminescence and Raman data reported in previous section (see 4.3.2 and 4.3.3 section). In addition, of the samples increases with alloying time from173 ns to 276~501ns (table 4-1). This behavior is consistent with previous theoretical 5 d ealed for 1 le which ing energy [127]. Exciton binding energy is increased by exciton confinement whiobtained by size reduction or localization of carrier wave function by composition variation [132]. Therefore the binding energies of exciton in alloy nanorods presented here can be expected to increase due to increased localization of exciton by compositional fluctuation, leading to increase luminescence decay time (). Kim et al observed also increment of the PL decay lifetime with Al content in case of AlGaN alloys arguments predicting that the radiative lifetime of bound excitons increases with bind ch is due to the exciton localization [127].

PAGE 87

73 4.3.5 Transient Absorption Spectroscopy (TRA) Study Additional insight can be gained by investigating the ultrafast changes observed inthe absorption spectra[32,33]. Femtosecond TRA was utilized to study the evolution of the absorption bleaching in the nanorods. Changes in population density of differentenergy levels was examined using femtosecond time resolved pump probe spectroscopyA pump pulse excites the sample, which causes a depopulation of the ground state to anexcited state and a probe pulse arrives at different delay times to monitor the populationof various energy states. The signals collected are changes in transmission (eq. 4-1). Whave excited all samples with low energy excitation (35 to 49 nJ) at 450 nm and monitored the bleach signal at wavelengths between 450 and 725 nm. Although we cannot make any definitive statements about the dynamics at this point, we are able to complement the steady state absorption data and to show the general trends that occur during alloying. Figure 4.11 compares the bleach spectrum of the core/shell and ZnCdSalloyed samples using the same delay times of 0, 50, 100, 200, 400, 800 and 2470 fs. Two bleach transitions are observed, denoted as 1S and 1P. The bleach of 1P transitigrows instantaneously (<150 fs) while the bleach of the 1S transition grows on a 2 ptimescale due to relaxation of hot electron from 1P to 1S level. It is observed that band width of 1S transition increases during alloyingto compositional disorder of ZnCdSe nanorods. In addition in early time bleach spectraZnCdSe nanorods (200~400fs), it is clear that ZnCdSe nanorods have more absorption band than CdSe/ZnSe coreshell nanorods. It is specula s e e on s process due of ted that energy level may split by intrinsic electric field induced by compositional fluctuation within single ZnCdSe nanorods. In figure 4.12, we summarized band structures of the nanorod samples. As shown in figure 4.12(a), the band gap energies rise rapidly from 1.95 eV in the core/shell

PAGE 88

74 nanorods but plateau to 2.2 eV as shown steady state PL and UV-Vis absorption data(fl 0.020.040.060.08 igure 4.7 and figure 4.8). Figure 4.12(b) shows that 1S bandwidth increases from the core/shell to ZnCdSe alloyed for 1hrs, and then narrows to approximately the origina 450525600675-0.010.010.030.050.07 0.00 T/ T (nm) 1S1P450525600675-0.010.000.010.02 0.030.04 T T/(nm) 1S1P450525600675-0.0050.0050.0100.0150.0200.030 0.0000.0250.035 T/ 0.07 T(nm) 1S1P450525600675-0.010.010.020.030.040.06 0.000.05 T/T (nm) 1S1P 0 fs0.1000.2000.8002.47 0.050.400 0 fs0.1000.2000.8002.47 0.050.400 0 fs0.1000.2000.8002.47(a)(b)(c)(d)450525600675-0.010.010.030.050.07 0.050.4000.000.020.040.060.08 T/ T (nm) 1S1P450525600675-0.010.000.010.02 0.030.04 T T/(nm) 1S1P450525600675-0.0050.0050.0100.0150.0200.030 0.0000.0250.035 T/ 0.07 T(nm) 1S1P450525600675-0.010.010.020.030.040.06 0.000.05 T/T (nm) 1S1P 0 fs0.1000.2000.8002.47 0.050.400 0 fs0.1000.2000.8002.47 0.050.400 0 fs0.1000.2000.8002.47(a)(b)(c)(d) Figure 4.11 Ultrafast carrier relaxations in (a) CdSe/ZnSe core sZnCdSe alloy 1hr (c) ZnCdSe alloy 2hr (d) ZnCdSe 0.050.400hell nanorods (b) alloy 3hr

PAGE 89

75 0.240.260.280.30 CdSe/ZnSeAlloy 1 hrAlloy 2 hrAlloy 3 hr0.060.080.100.120.140.160.180.200.22 Bandwidth (eVCdSe/ZnSeAlloy 1 hrAlloy 2 hrAlloy 3 hr1.71.81.92.02.12.22.3 )2.42.52.6 Band Gap (eCdSe/ZnSeAlloy 1 HrAlloy 2 HrAlloy 3 Hr0.360.380.400.420.440.48 V)0.460.50 Peak Separation EnerCdSe/ZnSeAlloy 1 hrAlloy 2 hrAlloy 3 hr0.060.080.100.120.140.160.180.200.22 gy(a)(b)(c)0.240.260.280.30 Bandwidth (eVCdSe/ZnSeAlloy 1 hrAlloy 2 hrAlloy 3 hr1.71.81.92.02.12.22.3 )2.42.52.6 Band Gap (eCdSe/ZnSeAlloy 1 HrAlloy 2 HrAlloy 3 Hr0.360.380.400.420.440.48 V)0.460.50 Peak Separation Ener Figure 4.12 Summary of absoption study. (a) Energy separations between 1S and 1P transition, (b) 1S bandwidth change and (c) bandgap change of CdSe/ZnSe coreshell and ZnCdSe nanorods width after 3hrs. This trend is consistent with the luminescence inhomogeneous broadening shown in figure 4.8 due to compositional disorder and annealing effect. Finally, we measure the energy separation between the 1S and 1P (Figure 4.12(c)). These values do not change significantly and the average spacing is 0.4075.012 eV. As shown in figure 2.2, energy separation between 1S and 1P is significantly dependent on size of nanocrystals [31]. So we can expect that size change may be negligible during gap, band s. This is vident in both the steady state and time-resolved data presented. Further work is necessary to elucidate the dynamics of the processes discussed above. 4.4 Conclusions Green-yellow emitting ZnCdSe ternary alloy nanorods (6 nm x 13 nm) with relatively high quantum yields (QY = 10 %) were synthesized by alloying ZnSe/CdSe core/shell nanorods at 270oC for times up to 3 hrs. The nanorods were characterized by gy(a)(b)(c) alloying process. Overall, data indicate the occurrence of a transformation of the band tructure as function of alloying time in each of the rods studied e

PAGE 90

76 X -ray diffraction (XRD), transmission electron microscopy (TEM), and Raman spectroscopy, as well as optical absorption and emission. The XRD and TEM were used to quantify the size and shape of the nanorods, as reported above. The Raman data were used to detect limited alloying in as-synthesized core/shell nanorods, and composition disorder in the alloyed material. The QY of ZnCdSe nanorods was a function of annealing time, was greater than pristine CdSe nanorods (QY = 0.6 %), but was lower than that from the core/shell nanorods (15 %). The luminescent efficiency of these alloying prm lloy e hing es of m aterials was discussed in terms of compositional disorder, defects induced by the ocess, and surface passivation by larger band gap surface layers resulting fro higher Zn concentrations near the surface. Time resolved emission provided information regarding the role of diffused Zn. A stretched exponential function was used to describe these systems, where < 1 corresponds to disperse populations. Comparing CdSe/ZnSe coreshell nanorods to aZnCdSe nanorods we found a significant decrease in the value (from ~0.75 to 0.48~0.58). Luminescence decay life time of the samples increased with alloying timfrom 173 ns to 276~501ns. This was explained by compositional disorder and exciton localization. Additional insight was gained by the evolution of the absorption bleacin the nanorods using femtosecond TRA. After excitation at 450 nm, band structurthe nanorod samples were determined.

PAGE 91

CHAPTER 5 SELF ASSEMBLED GROWTH OF GOLD NANOCRYSTALS ON CADMIUM SULFIDE NANORODS 5.1 Introduction Nanoscale charge transfer [133] is an important issue for both the fundamental properties and applications of molecular electronics in displays, photovoltaics, sensand photocatalyst. For photovoltaics and photocatalyst, efficient charge separation treduce the recombination of the photogenerated electron-hole pairs is essential. Assemblies of noble metals with semicondu ors, o ctor nanoparticles, such as Au/TiO2, have been crystals f aration due to their insulating properties [104]. It seems reasonable that direct contact of Au nanocrystals with CdS nanocrystals would be more effective for charge separation and improved device performance. In addition to nanospheres, nanorods of CdSe have been studied for photovoltaic devices because of their potential for better charge transfer to the solar cell electrodes studied because of their ability to separate charge [134]. The basic role of the noble metal nanoparticle is to shuttle photogenerated electrons from semiconductor nanocrystals. Simple self organized mixtures of Au nanoparticles and spherical CdS nanohave been reported previously to exhibit increased photocurrent generation and photoelectrochemical properties [102,104,135,136]. In these mixtures, appropriate organic spacers have been used for electrostatic self assembly and surface modification onanocrystals. However, such organic spacers may bridge nanoparticles by electrostatic attraction without promoting charge sep 77

PAGE 92

78 [2,78], and for laser [137] and thin film transistor [138]. Various methods have been shown to result in one-dimensional (1-D) nanocrystals such as nanorods [71]. Solution phase growth using capping reagents mquantities at a low cosfor in situ preparation of nanocomporent materials. at the nitrate tetrahydrate (Cd(NO3)2, 99.999 %), thiourea (NH2CSNH2, 99.0 %), hydrogen tetrachloroaura123.2 mg (0.4 mmol) of Cd(NO) and 60.8 mg (0.8 mmol) of thiourea were dissolved in 10mL of ethylenediamine. Nanorods of Srich CdS were grown in this solution by heating to 120 oC for 10hrs, after which the reaction vessel was cooled to room temperature to stop crystal growth. Nanorods were collected by precipitation and washing with distilled water. Cd+ rich CdS nanorods were prepared using the same procedure except the stoichiometry of the Cd:S precursors was 2:1 rather than 1:2. Obtained CdS nanorods were used for deposition of Au nanocrystals. ay be used to produce large t, and involve simple procedures. Moreover, solution phase synthesis is appropriate site from two diffe In this study, CdS nanorods were prepared using solution phase growth and Au nanocrystal could also be grown directly on S2rich surfaces due to strong Au-S bonding. Photoluminescence (PL) and photo-catalytic properties of Au/CdS nanorod composites are reported, which are attributed to more effective charge separationinterface between Au and CdS. 5.2 Experimental section 5.2.1 Materials Cadmium te(III) hydrate (99.999 %), sodium borohydride (NaBH 4 powder, 98%) and ehthylendiamine (redistilled,99.5 %) were purchased from Aldrich chemicals and used without further purification. 5.2.2 Preparation of CdS Nanorods 3 2

PAGE 93

79 5.2.3 Preparation of Au/CdS Nanorods A solution of 7.87mg (0.02mmol) HAuCl 4 in 10ml of ethylenediamine was addedto the solution containing the Srich CdS nanorods, and stirred for 15 min at room temperature. The resulting solution contained an Au-S complex on the surface of the Cdnanocrystals. This solution was added to 20 ml of distilled water to wh S ich 0.9 mg of NaBHd ater. as conducted using a JEOL 2010F microscope. X-ray diffraction (XRD) patterns werD 3720 ray diffractometer re and size. For photoluminescence data, a He-Cd laser emitting at 325 nm wrocion red mix-5B (PRB) from Aldrich yrex glass vessel with magnetic stirring. The illummg 4 had been dissolved, and the mixture turned grey indicating Au nanocrystals haformed. Nanocrystals was collected by centrifugation and washed with distilled w5.2.4 Characterization High resolution transmission electron microscopy (HR-TEM) analysis of the crystalline lattice and rod shape w e obtained using a Philips AP for crystal structu as used for excitation. The light emitted was collected using a HR-320 monochromator (Instruments SA, Inc.) with a Hamamatsu R943-02 GaAs photomultiplier detector. To evaluate the photocatalytic activity of the Au/CdSe composites, the rate of destruction of a red dye, p Chemicals, was measured in a 100ml P ination source for photocatalyst was four 12 inch 8 W UV bulbs (UVP inc.) dominated by emission at 365nm. Aqueous solution of PRB (50 ml, 10 mg/L) and 5 of Au-CdS nanorods composite were placed in the reaction vessel and sonicated for 20min in the dark. After selected time intervals of UV-illumination, a specimen of the suspension was collected and analyzed by UV-Vis absorption spectroscopy using a Perkin-Elmer Lambda 800 spectrophotometer.

PAGE 94

80 5.3 Result a nd Discussion 5.3.1 ], polyol process and solvo-thermal method [139,141]. Bidening to rystals along ods increased of n anocrystals. The photoluminescent ectra from S2rich versus Cd2+ rich CdS nanorods were different, as will be discussed Growth of Au Nanocrystals on CdS Nanorods As described above, nanorods of CdS were synthesized by solution phase method using ethylene diamine [139]. In general, symmetry breaking is required in the nucleation step for anisotropic nanocrystals [71]. In solution phase growth, a variety of chemicaltechniques to lower the symmetry of a nucleus have been reported by using appropriate capping reagents. For example, micelles methods [140], kinetically controlled growth rates of various facets [72 tate ligand such as ethylene diamine can serve as a molecular template, leadrod shape CdS nanocrystals [139,142]. The XRD pattern from a typical hexagonal structure of CdS nanorods is shown in Figure 5.1. For long reaction times, the (002) diffraction peak became more intense and narrower than peaks from other planes, consistent with growth of rod shape cthe c-axis of the hexagonal structure. The (002) interplanar spacing of the rfrom 6.60 to 6.66 as reaction time increased, presumably because of relaxationcompressive stress on the larger particle size. After reaction for 10 hr at 120 o C, Debye-Scherrer analysis based on the XRD FHM data indicated a diameter of ~15 nm and alength of ~80 nm. The HR-TEM image in figure 5.2 (b) from the same batch of nanocrystals are consistent with these dimensions, and also shows well resolved crystalline lattice planes (figure 5.2 (c)). Figure 2 (a) shows that the particles produced iethylene diamine after 0.5hr at 120 o C were anisotropic. Nanorods of CdS that were S 2rich [139,143] were prepared with an excess ofthiourea, and served as nucleation sites for Au n sp

PAGE 95

81 2030405060 2030405060 2030405060 Intnsitya. u.)2 theta(c) e ((b)(a) Figure 5.1 X-ray diffraction patterns from hexagonal CdS nanorods prepared at 120C for (a) 0.5 hrs, (b) 2 hrs, and 10 hrs, respectively. o Figurd after reaction at 120C for (a) 0.5 hrs and (b) 10 hrs. Lattice fringes from CdS 20nm 5nm 5nm(a) 5nm(a)(b)(c) e 5.2 High resolution transmission electron micrograph of CdS nanorods obtaineonanorods are shown in (c)

PAGE 96

82 later (Figure 5.5). Ions of Au3+ on the surface of CdS nanorods were reduced to Au0 by NaBH4 at room temperature, resulting in ~2 nm particle size [144]. A HR-TEM image of ~2 nm Au nanocrystals deposited directly on CdS nanorods crystals is shown in Figure 5.3. The chemisorption energy of thiolate to Au has been reported to be 28~44 kcal/mol [145], due to the dominant covalent and minor ionic contributions to the surface bond [146,147]. This strong chemisorption enables nucleation of Au nanocrystals directly on the CdS nanorods. The size of the Au nanocrystals is presumably restricted to ~2nm size by the ethylene diamine capping agent [144]. Figure 5.3 High resolution transmission electron micrographs of CdS nanorods with ~2 nm Au nanocrystals. 5.3.2 Photoluminescence and Charge Separation Luminescence from all nanocrystals, including CdS, may be strongly influenced by apped electron/holes at surface defects (traps) that quench radiative band gap their prepace of the PL ectra upon the enriched species on the surface is shown in Figure 5.4. The broad PL (a)(b)20nm10nm (a)(b)20nm10nm tr recombination. Therefore the optical properties of CdS nanorods depend strongly upon ration conditions and aspect ratio [143,148-152]. The dependen sp

PAGE 97

83 Figure 5.4 Photoluminescence spectra from (a) Cd rich and (b) S rich CdS nanorods e 350400450500550600650700750800850 ~0.8 eV~710 nm~670 nm~0.7 eV(b) (a) Vs (excitation at 325 nm). See text for discussion of defect levels illustrated in thdiagrams. Ssurface EVECh EV EC h Intensity (AniWavelength(nm) rb.ut)

PAGE 98

84 p eak from S2rich CdS nanorods occurred at ~710 nm which was 40 nm longer than that om Cd2+ rich surfaces (~670 nm). Weak excitonic emission with a peak at ~489 nm .54 eV) was observed from S2rich CdS nanorods. In case of S2rich nanorods, a photo-generated hole could migrate to the surface, forming Straps before radiative electron-hole pair recombination. Such traps might be located ~0.8 eV above the valence band. Recombination of photo-generated electrons from the conduction band with holes at Ssurface traps lead to red emission at ~710 nm [152]. In the case of CdS nanorods with Cd2+ rich surface, Ssurface traps are not expected. Instead, surface electron traps, such as Cd2+ or sulfur vacancies (V+s), are expected to influence luminescence [143]. These photo-generated trapped electrons recombine with holes in the valence band, leading to emission at ~670nm. Based on this PL data, the sulfur vacancy is located about ~0.7 eV below the conduction band in CdS nanorods, which is comparable with the values reported for bulk CdS [153,154]. Deposition of Au nanocrystals on S2rich CdS severely quenched the PL intensity from CdS nanorods (Figure 5.5). Presumably electron-transfer from CdS nanorods to the u metal nanoparticles reduced the probability of radiative electron-hole pair d by Milliron et al. in type II heterostructure CdSe/CdTe nanocrystals [155]. It has also been ently reported that the Fermi level of the TiO2 particles covered with Au nanoparticles on band after UV-irradiation. This was attributed to a large accumulation of electrons on the Au nanocrystals, indicating a high efficiency for interfacial charge-transfer between the metal and semiconductor nanocrystals [134]. Furthermore, smaller Au particles induce greater shifts in the Fermi level than did larger fr (2 A recombination [135]. Luminescence quenching by charge separation was also reporte rec shifted closer to the conducti

PAGE 99

85 Figure 5.5 Photoluminescence spectra from (a) S rich CdS nanorods and (b) Au deposited on the CdS nanorods. CdS nanorods. Note the severe quenching when Au nanocrystals are present separation of photogenerated charge. Figure 5.6 Energy level diagrams for Au deposited on CdS nanocrystals illustrating e-h + CdSAu nanorodsnanocrystalse----+ + + Vac e-h + CdSAu nanorodsnanocrystalse----+ + + Vac1500020000 4005006007008000500010000250003000035000 (b)(a) Au/CdS nanorodsCdS nanorods Intnsity eWavelength(nm)

PAGE 100

86 particles, due to the discrete energy levels for the smaller Au nanocrystals [134,156]. Based on the TEM images and quenching of PL in Figures 3 and 4, respectively, the present CdS nanorods with 2 nm Au nanocrystals show a high efficiency for charge separation of photogenerated electron-hole pairs. A schematic energy level diagram of Au deposited on CdS nanorods that lead to charge separation is shown in figure 6. The work function of Au is ~5ev which causes the top of the valence band to align below the conduction band minimum of CdS. Therefore any electron photoexcited from the valence band to the conduction band of CdS will have a strong driving force to transfer to the lower lying states in the Au nanocrystal. The resulting hole may be trapped on the Ssurface state shown at in Figure 5.6. 5.3.4 PhotTo further test whether charge separation was enhanced in Au/CdS nanocomposites, the rate of photocatalytic degradation of the Procion red mix-5B (PRB) red dye was measured as described above. PRB is easily oxidized in aqueous solution containing a photocatalyst under UV light irradiation [157]. The structure of PRB and the normalized concentration versus time of exposure to UV light is shown in Figure 5.7. The normalized PRB concentration was determined from absorbance at 539 nm (figure 5.8). Under UV irradiation, the Au/CdS nanorods degrade the PRB at a rate that is 9 % faster than the rate for only CdS nanorods. Enhanced charge separation by Au/CdS (versus CdS only) with a subsequently larger rate of production of oxygen radicals (e.g. O2, OOH, OH [158] are thought to explain the rate increase. These oxygen radicals degrade PRB, resulting in bleaching of the red color of the solution. The yield of oxygen radicals is lower withpid electron-hole recombination and ocatalytic Activity pure CdS nanorods due to the more ra

PAGE 101

87 Figure 5.7 Ratio of the concentration C versus initial concentration C of Procion red V light irradiation, (a) without catalyst and in the presence of 5mg of (b) CdS o mix-5B (PRB) dye in aqueous solution versus time of exposure to 365 nm Unanorods or (c) Au/CdS nanorods. 01020304050600.00.20.40.60.8 Figure 5.8 UV visible absorption spectra of Procion red mix-5B (PRB) dye in aqueous solution under 365 nm UV light irradiation in the presence of 5mg of (a) CdSnanorods and (b) Au/CdS nanorods. 1.0 01020304050600.00.20.40.60.8 1.0 01020304050600.00.20.40.60.8 1.0 (a)(b)Co (c) Time (min)C/ 01020304050600.00.20.40.60.8 1.0 01020304050600.00.20.40.60.8 1.0 01020304050600.00.20.40.60.8 1.0 (a)(b)Co (c) Time (min)C/ 3004005006007000.000.100.200.250.300.35 0.050.150.50 0.400.45 0 min(a) bance0.100.200.45 10 min20 min30 min40 min60 min AbsorWavelength3004005006007000.000.050.150.250.300.350.400.50 0 min10 min40 min60 minAbsrbance(b) Wavelength(nm) 20 min30 mino

PAGE 102

88 relatively inefficient charge transfer to O2. Furthermore, it is expected that Au/CdS nanorods may be used as photocatalyst under visible light because the band gap of CdS nanorods is only 2.54 eV (489 nm), whereas the widely used TiO2 photocatalyst has a higher bandgap ( Eg > 3.2 eV or 388 nm) [159]. 5.5 Conclusion Nanorods of S2rich CdS were used to further grow Au nanocrystals at surface nucleation sites by the reaction of nonstoichiometric precursors in the presence of the ak from CdS nfrom a Cd2le trap state on S2and Cd2+rich surfaces, respectively. Au nanocrystals ~2 nm in diameter were grown on S2rich CdS nanorods in the presence of NaBH4 due to strong Au-S bonding. The deposition of Au nanocrystals on CdS nanorods resulted in a dramatic quenching of luminescence. This was postulated to be a result of separation of electrons to the Au and holes to the CdS nanoparticles. The postulate of charge separation between Au and CdS nanocrystals was tested by measuring the relative rates of photocatalytic degradation of an aqueous solution of Procion red mix-5B (PRB) dye under UV light irradiation. The 9 % increase in photocatalytic rate was consistent with the postulate of enhanced charge separation leading to accelerated rates of production of oxygen radicals. bidentate molecular template, ethylene diamine. The wavelength of the PL emission peanorods with a S 2rich surface was lowered by 42 nm relative to emission + rich surface due to the existence of electron or ho

PAGE 103

CHAPTER 6 CONCLUSI ONS The following conclusions werre works were suggested in this chaptingle nto o higher reaction temperature due to rapid elimination of chemically bonded organic ligand and high diffusivity of the precursors. The luminescence intensity increased significantly when a ZnS shell was grown on ZnCdSe nanocrystals. In addition, when ZnS surrounded the ZnCdSe nanocrystal, a e drawn and futu er. 6.1 Single Step Growth of Colloidal Ternary ZnCdSe Quantum Dots Colloidal ternary alloy ZnCdSe quantum dots have been synthesized and their structure and optical properties was studied. 1. Colloidal ternary alloy ZnCdSe quantum dots have been synthesized by a sstep reaction in a mixture of Cd and Zn oleates with Se-TOP. 2. The band gap absorption and luminescence peak initially red shifted and this was attributed to growth of the Cd-rich ZnCdSe nanocrystals size. After the red shift, there was a subsequent blue shift that was attributed to diffusion of Zn from the outer layer ithe Cd-rich core, to form larger band gap ternary alloy ZnCdSe particles between 5 and 8 nm in diameter after reacting for 2-30 min. at 320C. 3. The initial nucleation of Cd-rich ZnCdSe nanocrystals was attributed to the higher reactivity of Cd oleate as compared to the lower reactivity of Zn oleate. 4. Faster growth and larger particle size was observed at 5 89

PAGE 104

90 ~21 nm shift to lower emission energy versus that from ZnCdSe nanocrystals was observed due to electron tunneling into the ZnS shell. 6.2 Synthesis and Characterization of nCdSe Semiconductor operties was studied. 1. Green-yequanto 0 othe alloying process. the e data, Comprease in order nanorods can be expected to increase due to increased localization of exciton due to Colloidal Ternary ZNanorods Colloidal ternary alloy ZnCdSe quantum rods have been synthesized and their structure and optical pr llow emitting ZnCdSe ternary alloy nanorods (6 nm x 13 nm) with um yields (QY) of 10% were synthesized by alloying ZnSe/CdSe core/shell nanorods at 270C for times up to 3 hrs. 2. Raman data were used to detect the limited alloying in as-synthesized core/shell nanorods, and the compositional disorder in the alloyed material. 3. The QY of alloyed ZnCdSe nanorods increased from 0.6% for as-synthesized nanorods, and increased with increased annealing time. However, even after 3 hrs at 27C, the QY of alloyed ZnCdSe nanorods was lower than that from CdSe/ZnSe core/shell nanorods (QY=15%). The luminescent efficiency of alloyed ZnCdSe was discussed in terms of compositional disorder and defects induced by 4. Time resolved photoluminescence (TRPL) data were collected to investigaterole of diffused Zn. A stretched exponential function was found to describe thes aring CdSe/ZnSe coreshell to alloyed ZnCdSe nanorods, a significant decthe value (from ~0.75 to 0.48 -0.58) was found and attributed to compositional disof alloyed ZnCdSe nanorods. 5. The decay lifetime of the PL emission () of ZnCdSe nanoparticles increased with alloying time from 173 to 276~501 ns. The binding energies of excitons in the alloy

PAGE 105

91 compositional fluctuations, leading to increase luminescence decay time. 6. Femtosecond transient absorption spectroscop y has been utilized to study the evtransitions are observed and were designated 1P and 1S. The 1P transition appeared instantaneously, while the 1S transition reached its maximum after 2 ps. The energy spacing between 1S and 1P was ~ 0.40 eV and remains unchanged during alloying. The origins of these transitions were discussed 6.3 Self Assembled Growth of Au Nanocrystals on CdS Nanorods Metal-semiconductor nanoassemblies (CdS nanorods and gold nanocrystals) were prepared and charge separation between these two materials was studied. 1. Nanorods of S rich CdS for Au nanocrystals nucleation site were synthesized by the reaction of nonstoichiometric precursors in the presence of ethylene diamine. 2. The PL peak from S rich CdS nanorods was broad with a peak at ~ 710 nm, which was 40 nm lower in energy than the PL peak from Cd rich CdS (~670 nm). The influence of surface electron or hole trap states on the luminescent pathway of CdS nanorods was discussed. 3. Nanocrystals of Au ~2 nm in size were grown on S rich surface of CdS nanorods due to strong Au-S bonding. 4. Significant luminescence quenching was observed from Au nanocrystals on CdS nanorods due to interfacial charge separation. Charge separation by Au nanocrystals on CdS resulted in enhanced photocatalytic degradation of Procion red mix-5B (PRB) dye in aqueous solution under UV light irradiation. olution of absorption bleaching in the nanorods. After excitation at 450 nm, two main 22-2+ 2

PAGE 106

92 6.4 Future Direction The work in this dissertation introduces new functional nanorods. These functionnanorods can be further used as important building blocks to construct optoelectronic devices. The ultimate goal for developing novel functional nanorods is to use them inapplications such as composite organic LED and photovoltaics. For future direction, research in three major areas listed below is suggested. (1) For OLED application, optimizati al real on of synthesis of functional nanorods presented h high lumin the na nanostructure can be achieved by controlling the functionality of orgge Therefore development of compatible organic cappi in this dissertation is needed, especially controlling nanorod length wit escence efficiency. Moreover, it is desired to study device efficiency as function of aspect ratio of nanorods. (2) For solar cell application, perpendicular alignment of nanorods between the electrodes is the key to the future success of high efficient plastic solar cell. Currently, nocrystals do not align in solution. This could be overcome by co-dissolving nanorods in conventional liquid crystals or liquid crystalline conducting polymer in orderto cooperatively align them in electric fields. (3) Self assembly of anic capping agents. Capping molecule could be engineered to facilitate charseparation or charge conduction as well ng molecule and conjugated polymer should be accompanied with developments of nanocrystals.

PAGE 107

monolayers of nanocrystals in molecular organic devices. Nature 420 (2002) 800(2) Science 295 (2002) 2425-27. (3) M Brust, CJ Kiely: Some recent advances in nanostructure preparation from gold Physicochemical and Engineering Aspects 202 (2002) 175-86. (4) AR Kortan, R Hull, RL Opila, MG Bawendi, ML Steigerwald, PJ Carroll, LE Vice Versa, in Inverse Micelle Media. Journal of the American Chemical Society rly Monodisperse CdE (E = S, Se, Te) Semiconductor Nanocrystallites. Journal of the ikulec, JR Heine, H Mattoussi, R Ober, KF Jensen, MG Bawendi: (CdSe)ZnS core-shell quantum dots: Synthesis and Physical Chemistry B 101 (1997) 9463-75. (7) XG Peng, MC Schlamp, AV Kadavanich, AP Alivisatos: Epitaxial growth of electronic accessibility. Journal of the American Chemical Society 119 (1997) XG Peng: Formation of high-quality CdTe, CdSe, and CdS nanocrystals using CdO as precursor. Journal of the American Chemical Society 123 (2001) 183-84. (9) I Mekis, DV Talapin, A Kornowski, M Haase, H Weller: One-pot synthesis of highly luminescent CdSe/CdS core-shell nanocrystals via organometallic and "greener" chemical approaches. Journal of Physical Chemistry B 107 (2003) 7454-62. LIST OF REFERENCES (1) S Coe, WK Woo, M Bawendi, V Bulovic: Electroluminescence from single 03. WU Huynh, JJ Dittmer, AP Alivisatos: Hybrid nanorod-polymer solar cells and silver particles: a short topical review. Colloids and Surfaces ABrus: Nucleation and Growth of CdSe on ZnS Quantum Crystallite Seeds, and 112 (1990) 1327-32. (5) CB Murray, DJ Norris, MG Bawendi: Synthesis and Characterization of NeaAmerican Chemical Society 115 (1993) 8706-15. (6) BO Dabbousi, J RodriguezViejo, FV M characterization of a size series of highly luminescent nanocrystallites. Journal of highly luminescent CdSe/CdS core/shell nanocrystals with photostability and 7019-29. (8) ZA Peng 93

PAGE 108

94 (10) CR Bullen, P Mulvaney: Nucleation and growth kinetics of CdSe nanocrystals in octadecene. Nano Letters 4 (2004) 2303-07. (11) LH Qu, XG Peng: Control of photoluminescence properties of CdSe nanocrystals in growth. Journal of thety 124 (2002) 2049-55. 2) JY Zhang, XY Wang, M Xiao, L Qu, X Peng: Lattice contraction in free-standing CdSe nanocrystals. Applied Physics Letters 81 (2002) 2076-78. (13) ly narrow luminescence spectral width. Journal of the American Chemical Society 125 (2003) 13559-63. (14) g, TJ White, W Knoll: Composition-tunable ZnxCd1-xSe nanocrystals with high luminescence and stability. Journal of the (15) ing the optical properties without changing the particle size. Journal of the American Chemical (16) d on alloyed semiconductor quantum dots. Journal of Nanoscience and Nanotechnology 4 (2004) 569-74. (17) ds. Science 292 (2001) 2060-63. (18) 9) M Kazes, DY Lewis, Y Ebenstein, T Mokari, U Banin: Lasing from 14 (20) h, JJ Dittmer, N Teclemariam, DJ Milliron, AP Alivisatos, KWJ Barnham: Charge transport in hybrid nanorod-polymer composite photovoltaic (21) i, U Banin: Synthesis and properties of CdSe/ZnS core/shell nanorods. Chemistry of Materials 15 (2003) 3955-60. (22) l e Chemical Society 124 (2002) 7136-45. American Chemical Socie (1 XH Zhong, YY Feng, W Knoll, MY Han: Alloyed ZnxCd1-xS nanocrystals with high XH Zhong, MY Han, ZL Don American Chemical Society 125 (2003) 8589-94. RE Bailey, SM Nie: Alloyed semiconductor quantum dots: Tun Society 125 (2003) 7100-06. RE Bailey, JB Strausburg, SM Nie: A new class of far-red and near-infrared biological labels base JT Hu, LS Li, WD Yang, L Manna, LW Wang, AP Alivisatos: Linearly polarized emission from colloidal semiconductor quantum ro X Chen, A Nazzal, D Goorskey, M Xiao, ZA Peng, XG Peng: Polarization spectroscopy of single CdSe quantum rods. Physical Review B 64 (2001) -. (1 semiconductor quantum rods in a cylindrical microcavity. Advanced Materials(2002) 317-+. WU Huyn cells. Physical Review B 67 (2003). T Mokar L Manna, EC Scher, LS Li, AP Alivisatos: Epitaxial growth and photochemicaannealing of graded CdS/ZnS shells on colloidal CdSe nanorods. Journal of thAmerican

PAGE 109

95 MA Kastner: Artificial Atoms. Physics Today (23) 46 (1993) 24-31. 7) LE Brus: Electron and Electron-Hole Interactions in Small Semiconductor ournal 8) N Chestnoy, R Hull, LE Brus: Higher Excited Electronic States in Clusters of al 9) Y Kayanuma: Quantum-Size Effects of Interacting Electrons and Holes in 988) 0) AI Ekimov, F Hache, MC Schanneklein, D Ricard, C Flytzanis, IA Kudryavtsev, lectronic-Transitions. Journal of the Optical Society of America B-Optical Physics 10 (1993) 100-07. (31) t optical spectrum in CdSe quantum dots. Physical Review B 53 (1996) 16338-46. (32) urnal of Physical Chemistry B 104 (2000) 6112-23. n Dots. Physical Review Letters 72 (1994) 2612-15. (24) ML Steigerwald, LE Brus: Semiconductor Crystallites a Class of Large Molecules. Accounts of Chemical Research 23 (1990) 183-88. (25) AL Efros: Interband Absorption of Light in a Semiconductor Sphere. Soviet Physics Semiconductors-Ussr 16 (1982) 772-75. (26) AL Efros, M Rosen: The electronic structure of semiconductor nanocrystals. Annual Review of Materials Science 30 (2000) 475-521. (2 Crystallites the Size Dependence of the Lowest Excited Electronic State. Jof Chemical Physics 80 (1984) 4403-09. (2 ZnSe, CdSe, and ZnS Spin-Orbit, Vibronic, and Relaxation Phenomena. Journof Chemical Physics 85 (1986) 2237-42. (2 Semiconductor Microcrystals with Spherical Shape. Physical Review B 38 (19797-805. (3 TV Yazeva, AV Rodina, AL Efros: Absorption and Intensity-Dependent Photoluminescence Measurements on Cdse Quantum Dots Assignment of the 1st E DJ Norris, MG Bawendi: Measurement and assignment of the size-dependen VI Klimov: Optical nonlinearities and ultrafast carrier dynamics in semiconductor nanocrystals. Jo (33) MB Mohamed, C Burda, MA El-Sayed: Shape dependent ultrafast relaxatiodynamics of CdSe nanocrystals: Nanorods vs. nanodots. Nano Letters 1 (2001) 589-93. (34) DJ Norris, A Sacra, CB Murray, MG Bawendi: Measurement of the Size-Dependent Hole Spectrum in Cdse Quantum

PAGE 110

96 (3 5) JEB Katari, VL Colvin, AP Alivisatos: X-Ray Photoelectron-Spectroscopy of al Surface. Journal of Physical Chemistry 98 (1994) 4109-17. (36) Se nanocrystals coated by hydrophilic thiols. Journal of the American Chemical Society 123 7) MA Hines, P Guyot-Sionnest: Synthesis and characterization of strongly 00 rystals 9) YW Cao, U Banin: Growth and properties of semiconductor core/shell 0) JJ Li, YA Wang, WZ Guo, JC Keay, TD Mishima, MB Johnson, XG Peng: stals ican Chemical Society 125 (2003) 12567-75. of (42) ence Determinations of Cd Diffusion in ZnSe. Journal of Applied Physics 44 (1973) 5639-41. (43) (1988) 778-96. (44) terdiffusion in Wide-Bandgap Zn(Cd)S(Se) Strained Layer Superlattices. Semiconductor (45) burg, M Kuttler, UW Pohl, D Bimberg: Diffusion of Cd, Mg and S in ZnSe-based quantum well structures. Thin Solid Films 336 (1998) 208-12. (46) mical analysis of CdSe/ZnSe nanostructures by transmission electron microscopy. Physical Review B 61 (2000) 16015-24. Cdse Nanocrystals with Applications to Studies of the Nanocryst J Aldana, YA Wang, XG Peng: Photochemical instability of Cd (2001) 8844-50. (3 luminescent ZnS-Capped CdSe nanocrystals. Journal of Physical Chemistry 1(1996) 468-71. (38) P Reiss, J Bleuse, A Pron: Highly luminescent CdSe/ZnSe core/shell nanocof low size dispersion. Nano Letters 2 (2002) 781-84. (3 nanocrystals with InAs cores. Journal of the American Chemical Society 122 (2000) 9692-702. (4 Large-scale synthesis of nearly monodisperse CdSe/CdS core/shell nanocryusing air-stable reagents via successive ion layer adsorption and reaction. Journal of the Amer (41) DV Talapin, I Mekis, S Gotzinger, A Kornowski, O Benson, H Weller: CdSe/CdS/ZnS and CdSe/ZnSe/ZnS core-shell-shell nanocrystals. JournalPhysical Chemistry B 108 (2004) 18826-31. WE Martin: Photoluminesc D Shaw: Diffusion Mechanisms in Ii-Vi Materials. Journal of Crystal Growth 86 PJ Parbrook, B Henderson, KP Odonnell, PJ Wright, B Cockayne: In Science and Technology 6 (1991) 818-21. M Strass N Peranio, A Rosenauer, D Gerthsen, SV Sorokin, IV Sedova, SV Ivanov: Structural and che

PAGE 111

97 (47) A Rosenauer, T Reisinger, E Steinkirchner, J Zweck, W Gebhardt: High-Resolution Transmission Electron-Microscopy Determination of Cd DiffusioCdse/ZnSe Single-Quantum-Well Structures. Journ n in al of Crystal Growth 152 (1995) 42-50. (48) nal of Physical Chemistry 100 (1996) 13226-39. (49) ce in ZnS : Eu2+ nanoparticles. Journal of Applied Physics 89 (2001) 2671-75. (50) allenberg, JO Bovin, V Zwiller, NA Kotov: Crystal field, phonon coupling and emission shift of Mn2+ in (51) M Antonietti: Synthesis and characterization of CdS nanoparticles stabilized by double-hydrophilic block copolymers. Nano Letters 1 (52) n in Colloid & Interface Science 9 (2004) 264-78. (53) ical Crystal Growth 253 (2003) 467-71. anocrystals. Journal of Physics and Chemistry of Solids 62 (2001) 1957-60. (56) etical study. Journal of Physical Chemistry B 105 (2001) 12278-85. (57) scent monodisperse CdSe and CdSe/ZnS nanocrystals synthesized in a hexadecylamine-7-11. (58) SM Woessner, CS Yun: Inorganic clusters as single-source precursors for preparation of CdSe, ZnSe, and CdSe/ZnS nanomaterials. Chemistry of Materials 14 (2002) 1576-84. AP Alivisatos: Perspectives on the physical chemistry of semiconductor nanocrystals. Jour W Chen, JO Malm, V Zwiller, R Wallenberg, JO Bovin: Size dependence of Eu2+ fluorescen W Chen, R Sammynaiken, YN Huang, JO Malm, R W ZnS : Mn nanoparticles. Journal of Applied Physics 89 (2001) 1120-29.LM Qi, H Colfen, (2001) 61-65. MA Lopez-Quintela, C Tojo, MC Blanco, LG Rio, JR Leis: Microemulsion dynamics and reactions in microemulsions. Current Opinio EA Meulenkamp: Synthesis and growth of ZnO nanoparticles. Journal of PhysChemistry B 102 (1998) 5566-72. (54) CH An, KB Tang, Y Jin, JW Liu, H Lv, YT Qian: A simple method to synthesize PbE (E = S, SE) nanocrystals. Journal of (55) C Wang, G Zhang, S Fan, Y Li: Hydrothermal synthesis of PbSe, PbTe semiconductor n DV Talapin, AL Rogach, M Haase, H Weller: Evolution of an ensemble of nanoparticles in a colloidal solution: Theor DV Talapin, AL Rogach, A Kornowski, M Haase, H Weller: Highly lumine trioctylphosphine oxide-trioctylphospine mixture. Nano Letters 1 (2001) 20SL Cumberland, KM Hanif, A Javier, GA Khitrov, GF Strouse

PAGE 112

98 T Trindade, P Obrien, XM Zhang: Synthesis of CdS and CdSe nanocrystallites using a novel single-molecule precursors approach. Chemistry of Materials 9 (1997) 523-30. (59) thesis, -58. 1) KM Hanif, RW Meulenberg, GF Strouse: Magnetic ordering in doped Cd12) OE Raola, GF Strouse: Synthesis and characterization of Eu-doped cadmium (63) Organometallic synthesis and spectroscopic characterization of manganese-doped 0) 25324) TH Ji, WB Jian, JY Fang: The first synthesis of Pb1-xMnxSe nanocrystals. (65) rmikko, W Xie, DC Grillo, M Kobayashi, RL Gunshor: Blue-Green Injection-Laser Diodes in (Zn,Cd)Se/ZnSe Quantum6) MA Haase, J Qiu, JM Depuydt, H Cheng: Blue-Green Laser-Diodes. Applied 7) WR Chen, CJ Huang: ZnSe-based mixed-color LEDs. Ieee Photonics Technology (68) ng, YK Su, JF Chen, WH Lan, WJ Lin, YT Cherng, CH Liu, UH Liaw: ZnCdSeTe-based orange light-emitting diode. Ieee Photonics (69) ductors on InP for green-yellow emitters. Ieee Journal of Selected Topics in Quantum Electronics 8 (2002) 773-86. (70) (60) T Trindade, P O'Brien, NL Pickett: Nanocrystalline semiconductors: Synproperties, and perspectives. Chemistry of Materials 13 (2001) 3843 (6 xCoxSe diluted magnetic quantum dots. Journal of the American Chemical Society 124 (2002) 11495-502. (6 selenide nanocrystals. Nano Letters 2 (2002) 1443-47. FV Mikulec, M Kuno, M Bennati, DA Hall, RG Griffin, MG Bawendi: CdSe nanocrystals. Journal of the American Chemical Society 122 (20040. (6 Journal of the American Chemical Society 125 (2003) 8448-49. H Jeon, J Ding, W Patterson, AV Nu Wells. Applied Physics Letters 59 (1991) 3619-21. (6 Physics Letters 59 (1991) 1272-74. (6 Letters 16 (2004) 1259-61. WR Chen, SJ Cha Technology Letters 14 (2002) 1061-63. K Kishino, I Nomura: II-VI semicon A Nurmikko, RL Gunshor: Blue and green semiconductor lasers: a status report. Semiconductor Science and Technology 12 (1997) 1337-47. (71) YN Xia, PD Yang, YG Sun, YY Wu, B Mayers, B Gates, YD Yin, F Kim, YQ Yan: One-dimensional nanostructures: Synthesis, characterization, and applications. Advanced Materials 15 (2003) 353-89.

PAGE 113

99 (72) XG Peng, L Manna, WD Yang, J Wickham, E Scher, A Kadavanich, AP Alivisatos: Shape control of CdSe nanocrystals. Nature 404 (2000) 59-61. 3) L Manna, EC Scher, AP Alivisatos: Shape control of colloidal semiconductor 4) YW Jun, JW Seo, SJ Oh, J Cheon: Recent advances in the shape control of ) 5) E Rothenberg, Y Ebenstein, M Kazes, U Banin: Two-photon fluorescence ly al Chemistry B 108 (2004) 2797-800. (76) 7) H Htoon, JA Hollingworth, AV Malko, R Dickerson, VI Klimov: Light dots. 9) SM Lee, SN Cho, J Cheon: Anisotropic shape control of colloidal inorganic 0) YW Jun, SM Lee, NJ Kang, J Cheon: Controlled synthesis of multi-armed CdS ) 5150-51. 12700-06. s ournal of the American Chemical Society 124 (2002) 11244-45. (83) Se nanocrystals. Journal of the American Chemical Society 123 (2001) 1389-95. (7 nanocrystals. Journal of Cluster Science 13 (2002) 521-32. (7 inorganic nano-building blocks. Coordination Chemistry Reviews 249 (20051766-75. (7 microscopy of single semiconductor quantum rods: Direct observation of highpolarized nonlinear absorption dipole. Journal of Physic LS Li, AP Alivisatos: Origin and scaling of the permanent dipole moment in CdSe nanorods. Physical Review Letters 90 (2003). (7 amplification in semiconductor nanocrystals: Quantum rods versus quantumApplied Physics Letters 82 (2003) 4776-78. (78) AP Alivisatos: Hybrid-nanorod polymer solar cells. Abstracts of Papers of the American Chemical Society 227 (2004) U1240-U40. (7 nanocrystals. Advanced Materials 15 (2003) 441-44. (8 nanorod architectures using monosurfactant system. Journal of the American Chemical Society 123 (2001 (81) L Manna, EC Scher, AP Alivisatos: Synthesis of soluble and processable rod-, arrow-, teardrop-, and tetrapod-shaped CdSe nanocrystals. Journal of the American Chemical Society 122 (2000) (82) SM Lee, YW Jun, SN Cho, J Cheon: Single-crystalline star-shaped nanocrystaland their evolution: Programming the geometry of nano-building blocks. J ZA Peng, XG Peng: Mechanisms of the shape evolution of Cd

PAGE 114

100 (8 4) A Puzder, AJ Williamson, N Zaitseva, G Galli, L Manna, AP Alivisatos: The ed by ournal of the American Chemical Society 124 (2002) 3343-53. (86) ctor nanocrystals. Advanced Materials 15 (2003) 459-63. 8) S Kan, T Mokari, E Rothenberg, U Banin: Synthesis and size-dependent 03) selenide 8-69. recursor toward II-VI nanowires. Journal of Physical Chemistry B 108 (2004) 9745-51. (91) Diodes. Applied Physics Letters 51 (1987) 913-15. (92) iconducting Polymer. Nature 370 (1994) 354-57. (93) ystals and a semiconducting polymer. Journal of Applied Physics 82 (1997) 5837-42. (94) Physical Chemistry B 107 (2003) 9705-10. (95) ear-infrared polymer nanocrystal light-emitting diodes. Science 295 (2002) 1506-08. effect of organic ligand binding on the growth of CdSe nanoparticles probAb initio calculations. Nano Letters 4 (2004) 2361-65. (85) ZA Peng, XG Peng: Nearly monodisperse and shape-controlled CdSe nanocrystals via alternative routes: Nucleation and growth. J XG Peng: Mechanisms for the shape-control and shape-evolution of colloidal semicondu (87) S Kan, T Mokari, E Rothenberg, U Banin: Synthesis and size-dependent properties of zinc-blende semiconductor quantum rods (vol 2, pg 155, 2003). Nature Materials 3 (2004) 72-72. (8 properties of zinc-blende semiconductor quantum rods. Nature Materials 2 (20155-58. (89) H Yu, JB Li, RA Loomis, PC Gibbons, LW Wang, WE Buhro: Cadmiumquantum wires and the transition from 3D to 2D confinement. Journal of the American Chemical Society 125 (2003) 1616 (90) JW Grebinski, KL Richter, J Zhang, TH Kosel, M Kuno: Synthesis and characterization of Au/Bi Core/Shell nanocrystals: A p CW Tang, SA Vanslyke: Organic Electroluminescent VL Colvin, MC Schlamp, AP Alivisatos: Light-Emitting-Diodes Made from Cadmium Selenide Nanocrystals and a Sem MC Schlamp, XG Peng, AP Alivisatos: Improved efficiencies in light emitting diodes made with CdSe(CdS) core/shell type nanocr HS Yang, PH Holloway: Electroluminescence from hybrid conjugated polymer CdS : Mn/ZnS core/shell nanocrystals devices. Journal of N Tessler, V Medvedev, M Kazes, SH Kan, U Banin: Efficient n

PAGE 115

101 (96) JS Steckel, S Coe-Sullivan, V Bulovic, MG Bawendi: 1.3 m to 1.55 m tunabelectroluminescence from PbSe quantum dots embedded within an organic device. Advanced Materials 15 (2003) 1862-66. le -22. (98) A Cuddihy, G Redmond: Near-infrared electroluminescent devices based on colloidal HgTe quantum dot arrays. 9) NC Greenham, XG Peng, AP Alivisatos: Charge separation and transport in nched 01) SA McDonald, G Konstantatos, SG Zhang, PW Cyr, EJD Klem, L Levina, EH 02) J Kolny, A Kornowski, H Weller: Self-organization of cadmium sulfide and gold (103) ite nanostructures. To what extent do metal nanoparticles improve the photocatalytic 04) L Sheeney-Haj-Ichia, S Pogorelova, Y Gofer, I Willner: Enhanced 05) T Mokari, E Rothenberg, I Popov, R Costi, U Banin: Selective growth of metal 06) BA Korgel, HG Monbouquette: Controlled synthesis of mixed core and layered 07) WZ Wang, I Germanenko, MS El-Shall: Room-temperature synthesis and (97) EH Sargent: Infrared quantum dots. Advanced Materials 17 (2005) 515E O'Connor, A O'Riordan, H Doyle, S Moynihan, Applied Physics Letters 86 (2005) -. (9 conjugated-polymer/semiconductor-nanocrystal composites studied by photoluminescence quenching and photoconductivity. Physical Review B 54 (1996) 17628-37. (100)BQ Sun, E Marx, NC Greenham: Photovoltaic devices using blends of braCdSe nanoparticles and conjugated polymers. Nano Letters 3 (2003) 961-63. (1 Sargent: Solution-processed PbS quantum dot infrared photodetectors and photovoltaics. Nature Materials 4 (2005) 138-U14. (1 nanoparticles by electrostatic interaction. Nano Letters 2 (2002) 361-64. V Subramanian, E Wolf, PV Kamat: Semiconductor-metal compos activity of TiO2 films? Journal of Physical Chemistry B 105 (2001) 11439-46. (1 photoelectrochemistry in CdS/Au nanoparticle bilayers. Advanced FunctionalMaterials 14 (2004) 416-24. (1 tips onto semiconductor quantum rods and tetrapods. Science 304 (2004) 1787-90. (1 (Zn,Cd)S and (Hg,Cd)S nanocrystals within phosphatidylcholine vesicles. Langmuir 16 (2000) 3588-94. (1 characterization of nanocrystalline CdS, ZnS, and CdxZn1-xS. Chemistry of Materials 14 (2002) 3028-33.

PAGE 116

102 (108)TR Cundari, WT Fu: Reaction pathways for model II-VI precursors: A computational study. International Journal of Quantum Chemistry 71 (1999)56. 47: Patterned three-color ZnCdSe/ZnCdMgSe quantum-well structures for integrated (110) GE Marques: Interface optical phonons in spheroidal dots: Raman selection rules. Physical Review B 65 (2002) -. (111) kimov: sical ) 4664-69. s matrices. Physical Review B 54 (1996) 15120-24. (113) al Review B 70 (2004) -. non spectra in CdSe quantum dots. Physical Review B 68 (2003) -. (115) etched CdxZn1-xSe/ZnSe quantum wires. Physical Review B 57 (1998) 15439(116) isatos: Shape control and applications of nanocrystals. Philosophical Transactions of the Royal Society of London Series a(117) Hines, P Guyot-Sionnest: Bright UV-blue luminescent colloidal ZnSe nanocrystals. Journal of Physical Chemistry B 102 (1998) 3655-57. (118) structural transformations of CdSe nanocrystals. Physical Review Letters 84 (2000) 923-26. (119) conductors and Related Superlattices Cubic Cd1-Xmnxse and Cd1-Xznxse. Physical Review B 40 (1989) 3720-28. (109)Y Luo, SP Guo, O Maksimov, MC Tamargo, V Asnin, FH Pollak, YC Chen full-color and white light emitters. Applied Physics Letters 77 (2000) 4259-61. F Comas, C Trallero-Giner, N Studart C Trallero-Giner, A Debernardi, M Cardona, E Menendez-Proupin, AI EOptical vibrons in CdSe dots and dispersion relation of the bulk material. PhyReview B 57 (1998 (112) YN Hwang, SH Shin, HL Park, SH Park, U Kim, HS Jeong, EJ Shin, D Kim: Effect of lattice contraction on the Raman shifts of CdSe quantum dots in glas RW Meulenberg, T Jennings, GF Strouse: Compressive and tensile stress incolloidal CdSe semiconductor quantum dots. Physic (114) AV Baranov, YP Rakovich, JF Donegan, TS Perova, RA Moore, DV Talapin, AL Rogach, Y Masumoto, I Nabiev: Effect of ZnS shell thickness on the pho T Kummell, G Bacher, A Forchel, G Lermann, W Kiefer, B Jobst, D Hommel, G Landwehr: Size dependence of strain relaxation and lateral quantization in deep 47. EC Scher, L Manna, AP Aliv Mathematical Physical and Engineering Sciences 361 (2003) 241-55. MA JN Wickham, AB Herhold, AP Alivisatos: Shape change as an indicator of mechanism in the high-pressure RG Alonso, EK Suh, AK Ramdas, N Samarth, H Luo, JK Furdyna: Raman-Spectroscopy of 2 Novel Semi

PAGE 117

103 (120) VV Travnikov, VK Kaibyshev: Resonance exciton-phonon spectra in opZnCdSe/ZnSe nanowires: Raman scattering and hot luminescence, extended and loca en lized exciton states. Physics of the Solid State 45 (2003) 1379-87. l, Growth 159 (1996) 103-07. (122) loying effect in GaAs1-xPx ternary alloy semiconductors. Physical Review B 54 (1996) (123) i: Raman-scattering probe of anharmonic effects due to temperature and compositional disorder in III-V binary and ternary alloy (124) to: Alloying of CdSe/ZnSe quantum dot grown by an alternate molecular beam supply. Journal of Crystal 25) PR Yu, JM Nedeljkovic, PA Ahrenkiel, RJ Ellingson, AJ Nozik: Size dependent 26) Y Kawakami, K Omae, A Kaneta, K Okamoto, Y Narukawa, T Mukai, S Fujita: (127) im, RA Mair, J Li, JY Lin, HX Jiang: Time-resolved photoluminescence studies of AlxGa1-xN alloys. Applied Physics Letters 76 (2000) 1252-54. (128) ow-2-74. (121) W Meredith, G Horsburgh, GD Brownlie, KA Prior, BC Cavenett, W RothwelAJ Dann: Microprobe Raman study of the variation of LO phonon frequency with the Cd concentration in the ternary compound Zn1-xCdxSe. Journal of Crystal C Ramkumar, KP Jain, SC Abbi: Resonant Raman scattering probe of al 7921-28. C Ramkumar, KP Jain, SC Abb semiconductors. Physical Review B 53 (1996) 13672-81. M Ohishi, K Tanaka, T Fujimoto, M Yoneta, H Sai Growth 237 (2002) 1320-25. (1 femtosecond electron cooling dynamics in CdSe quantum rods. Nano Letters 4 (2004) 1089-92. (1 In inhomogeneity and emission characteristics of InGaN. Journal of Physics-Condensed Matter 13 (2001) 6993-7010. HS K R Chen: Apparent stretched-exponential luminescence decay in crystalline solids. Journal of Luminescence 102 (2003) 510-18. (129) X Chen, B Henderson, KP Odonnell: Luminescence Decay in Disordered LDimensional Semiconductors. Applied Physics Letters 60 (1992) 267 (130) M Pophristic, FH Long, C Tran, IT Ferguson, RF Karlicek: Time-resolved photoluminescence measurements of InGaN light-emitting diodes. Applied Physics Letters 73 (1998) 3550-52. (131) JS Huang, X Dong, XD Luo, XL Liu, ZY Xu, WK Ge: Localized exciton dynamics in AlInGaN alloy. Solid State Communications 126 (2003) 473-77.

PAGE 118

104 (132) R Cingolani, P Prete, D Greco, PV Giugno, M Lomascolo, R Rinaldi, L Calcagnile, L Vanzetti, L Sorba, A Franciosi: Exciton Spectroscopy in Zn1-Xcdxse/Znse Quantum-Wells. Physical Review B 51 (1995) 5176-83. (133) DM Adams, L Brus, CED Chidsey, S Creager, C Creutz, CR Kagan, PV Kamat, M Lieberman, S Lindsay, RA Marcus, RM Metzger, ME Michel-Beyerle, JR Miller, MD Newton, DR Rolison, O Sankey, KS Schanze, J Yardley, XY Zhu: Charge transfer on the nanos cale: Current status. Journal of Physical Chemistry B 107 (2003) 6668-97. (134) etal particle size on the Fermi level equilibration. Journal of the American Chemical Society 126 (2004) 4943-50. (135) article arrays by surface plasmon resonance: Application for the detection of acetylcholine esterase (136) i, A Javier, GF Strouse: Synthesis and characterization of a 1 : 6 Au-CdSe nanocomposite. Chemistry of Materials 15 (137) g, R Agarwal, CM Lieber: Single-nanowire electrically driven lasers. Nature 421 (2003) 241-45. (138) igh-iconductor nanowires and nanoribbons. Nature 425 (2003) 274-78. (139) f CdS nanorods via solvothermal route. Chemistry of Materials 12 (2000) 3259-63. (140) nanorods: Electrochemical synthesis and optical properties. Journal of Physical Chemistry B 101 (1997) 41) JR Heath, FK Legoues: A Liquid solution synthesis of single-crystal germanium (142) L Yang, GE Zhou: Nonaqueous synthesis of CdS nanorod semiconductor. Chemistry of Materials 10 (1998) 2301-+. (143) CdS clusters by ZnTPP. Journal of Physical Chemistry 96 (1992) 2868-73. V Subramanian, EE Wolf, PV Kamat: Catalysis with TiO2/gold nanocomposites. Effect of m M Zayats, AB Kharitonov, SP Pogorelova, O Lioubashevski, E Katz, I Willner: Probing photoelectrochemical processes in Au-CdS nanop inhibitors. Journal of the American Chemical Society 125 (2003) 16006-14. SL Cumberland, MG Berrettin (2003) 1047-56. XF Duan, Y Huan XF Duan, CM Niu, V Sahi, J Chen, JW Parce, S Empedocles, JL Goldman: Hperformance thin-film transistors using sem J Yang, JH Zeng, SH Yu, L Yang, GE Zhou, YT Qian: Formation process o YY Yu, SS Chang, CL Lee, CRC Wang: Gold 6661-64. (1 quantum wires. Chemical Physics Letters 208 (1993) 263-68. YD Li, HW Liao, Y Ding, YT Qian, J Chrysochoos: Recombination luminescence quenching of nonstoichiometric

PAGE 119

105 (144) MC Daniel, D Astruc: Gold nanoparticles: Assembly, supramolecular chquantum-size-related properties, and applications toward biology, catalysis, ananotechnology. Chemical Reviews 104 (2004) 293-346. emistry, nd (146) ld and Silver Surfaces Implications for Self-Assembled Monolayers. Journal of the American Chemical Society 115 (1993) 9389-401. 48) AV Isarov, J Chrysochoos: Optical and photochemical properties of (149) hexagonal nanocrystals. Journal of Materials Chemistry 13 (2003) 2641-48. (150) rived CdS quantum dots. Applied Physics Letters 69 (1996) 3224-26. (151) CH Liang, J Zhang: Catalytic growth of large-scale single-crystal CdS nanowires by physical evaporation and their (152) iangular CdS nanocrystals: Structural and optical studies. Advanced Materials 13 (2001) 261-+. (153) 13 (1959) 40-42. (154) tor cles the CdS/PbS Systems Photoluminescence Properties. Chemistry of Materials 6 (1994) 1534-41. (155) P Alivisatos: Colloidal nanocrystal heterostructures with linear and branched topology. Nature 56) SW Chen, RS Ingram, MJ Hostetler, JJ Pietron, RW Murray, TG Schaaff, JT (145) LH Dubois, RG Nuzzo: Synthesis, structure, and properties of model organic-surfaces. Annual Review of Physical Chemistry 43 (1992) 437-63. H Sellers, A Ulman, Y Shnidman, JE Eilers: Structure and Binding of Alkanethiolates on Go (147) A Johansson, S Stafstrom: Interactions between molecular wires and a gold surface. Chemical Physics Letters 322 (2000) 301-06. (1 nonstoichiometric cadmium sulfide nanoparticles: Surface modification with copper(II) ions. Langmuir 13 (1997) 3142-49. YC Li, XH Li, CH Yang, YF Li: Controlled synthesis of CdS nanorods and J Butty, N Peyghambarian, YH Kao, JD Mackenzie: Room temperature optical gain in sol-gel de YW Wang, GW Meng, LD Zhang, photoluminescence. Chemistry of Materials 14 (2002) 1773-77. N Pinna, K Weiss, J Urban, MP Pileni: Tr Vuyesteke, Y. T.: Sulfur vacancy mechanism in pure CdS. Physical Review 1 HS Zhou, H Sasahara, I Honma, H Komiyama, JW Haus: Coated SemiconducNanoparti DJ Milliron, SM Hughes, Y Cui, L Manna, JB Li, LW Wang, A 430 (2004) 190-95. (1 Khoury, MM Alvarez, RL Whetten: Gold nanoelectrodes of varied size: Transition to molecule-like charging. Science 280 (1998) 2098-101.

PAGE 120

106 (157) SW Lee, J Drwiega, CY Wu, D Mazyck, WM Sigmund: Anatase TiO2 nanoparticle coating on barium ferrite using titanium bis-ammonium lactato dihydroxide and its use as a magnetic photocatalyst. Chem istry of Materials 16 (2004) 1160-64. (158) radation of dye pollutants. V. Self-photosensitized oxidative transformation of Rhodamine B hysical r Journal of Solar Energy Engineering-Transactions of the ASME 119 (1997) 101-07. TX Wu, GM Liu, JC Zhao, H Hidaka, N Serpone: Photoassisted deg under visible light irradiation in aqueous TiO2 dispersions. Journal of PChemistry B 102 (1998) 5845-51. (159) DY Goswami: A review of engineering developments of aqueous phase solaphotocatalytic detoxification and disinfection processes

PAGE 121

BIOGRAPHICAL SKETCH yeokjin Lee was born in Jeonju, Republic of Korea. He grew up and was H Univer degree in 1992. He continued his study on organic chemistry and obtained his masters degree in 1995. After graduation, he was hired by LG Cable Inc. as associate senior research scientist. He was in charge of the development of photoalignment polymer for liquid crystal display (LCD) for the purpose of widening viewing angle. In 2001, he was admitted to Department of Materials Science and Engineering at the University of Florida to pursue his Ph.D. degree with a specialty of electronic materials. His research area in Dr. Holloways group was luminescent II-VI semiconductor nanocrystals for optoelectronic devices. Furthermore, his research interests are luminescent materials, nanocrystals, and organic materials for displays such as organic light emitting diode (OLED) and LCD. He obtained Ph.D. degree in December of 2005. educated in his hometown. In 1988, He entered the Department of Chemistry at Koreasity in Seoul, where he studied basic science. He obtained his Bachelor of Science 107