Citation
Ultrafast Spectroscopy of Novel Materials

Material Information

Title:
Ultrafast Spectroscopy of Novel Materials
Creator:
Hardison, Lindsay Michelle
Place of Publication:
[Gainesville, Fla.]
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (150 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Chemistry
Committee Chair:
Kleiman, Valeria D.
Committee Members:
Brucat, Philip J.
Fanucci, Gail E.
Omenetto, Nicolo
Hagen, Stephen J.
Graduation Date:
12/14/2007

Subjects

Subjects / Keywords:
Alloys ( jstor )
Electrons ( jstor )
Energy transfer ( jstor )
Excitons ( jstor )
Fluorescence ( jstor )
Nanocrystals ( jstor )
Nanorods ( jstor )
Photoluminescence ( jstor )
Polymers ( jstor )
Wavelengths ( jstor )
Chemistry -- Dissertations, Academic -- UF
cpes, nanorods, spectroscopy, ultrafast
Genre:
Electronic Thesis or Dissertation
born-digital ( sobekcm )
Chemistry thesis, Ph.D.

Notes

Abstract:
My research focused on steady state and time-resolved photophysical characterization of a series of semiconductor nanoparticles and water soluble conjugated polyelectrolytes. Several studies have shown that the electronic structure and relaxation dynamics in CdSe nanocrystals are not only size but are also shape and passivation dependent; however, there is no detailed comparison of the photophysical properties of ZnCdSe particles with different relative amounts of Zn. This dissertation presents data collected for colloidal CdSe, CdSe/ZnSe and ZnCdSe nanoparticles with rod-like architectures synthesized and investigated in our labs to determine how size, shape, passivation and composition affect the quantum confinement and dynamics. In addition, a series of different polymer repeat unit lengths of a linear conjugated polyelectrolyte (CPE) with a carboxylate ionic side chain have been synthesized and their photophysical properties have been explored. Spectral shifts and line broadening exhibited within the Raman spectroscopy, UV-Vis spectroscopy and photoluminescence aided in determining the extent of alloying and compositional disorder created during the alloying process. The photoluminescence quantum yield of ZnCdSe nanorods is higher than that from pristine CdSe nanorods indicating a higher binding energy of the exciton. This effect is speculated to be due to increased localization of the exciton as a result of fluctuations in the composition, ultimately resulting in increases in luminescence efficiencies. Moreover, time-resolved photoluminescence characterized lifetimes of nanoparticles with similar shape but different composition. Emission of an inhomogeneous population distribution (different sizes, shapes or composition) leads to the simultaneous probing of particles with different decaying rates. A stretched exponential function, I(t)= A*exp[-(t/?)^beta], can be used to describe these systems, where beta < 1 corresponds to disperse populations. In the experiments presented here, the photoluminescence data yields small beta values, independent of the emitted photon energy. Photoluminescence decay lifetime, ?, of the samples increased with alloying time due to compositional disorder leading to exciton localization. The dynamics of each nanorod was studied by absorption changes using ultrafast pump-probe spectroscopy. An excitation wavelength dependence study has been conducted to gain insight into the intraband/interband relaxation in core/shell nanorods with small valence band offsets. Determination of the dynamics and mechanisms of these systems will be useful for the study of fundamental physics and light emitting applications such as LEDs, photovoltaic devices, lasing and fluorescence tagging. CPEs are soluble in polar solvents and their conformational properties can be tuned to enhance their emissive behavior for sensing and device applications. It was found that polymer concentration, solvent, aggregation inducer and chain length, all affect the quenching efficiency; therefore, this dissertation examines energy transfer mechanism responsible for this behavior using ultrafast upconversion. Upon excitation of the aggregates from energetically higher lying isolated chains, the fluorescence lifetimes result in multi-exponential behavior due to the competition between the radiative and non-radiatve decay. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2007.
Local:
Adviser: Kleiman, Valeria D.
Statement of Responsibility:
by Lindsay Michelle Hardison.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Hardison, Lindsay Michelle. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
663880943 ( OCLC )
Classification:
LD1780 2007 ( lcc )

Downloads

This item has the following downloads:

hardison_l.pdf

hardison_l_Page_134.txt

hardison_l_Page_122.txt

hardison_l_Page_032.txt

hardison_l_Page_071.txt

hardison_l_Page_146.txt

hardison_l_Page_093.txt

hardison_l_Page_114.txt

hardison_l_Page_002.txt

hardison_l_Page_132.txt

hardison_l_Page_062.txt

hardison_l_Page_013.txt

hardison_l_Page_139.txt

hardison_l_Page_037.txt

hardison_l_Page_027.txt

hardison_l_Page_012.txt

hardison_l_Page_045.txt

hardison_l_Page_106.txt

hardison_l_Page_058.txt

hardison_l_pdf.txt

hardison_l_Page_035.txt

hardison_l_Page_019.txt

hardison_l_Page_144.txt

hardison_l_Page_016.txt

hardison_l_Page_113.txt

hardison_l_Page_131.txt

hardison_l_Page_021.txt

hardison_l_Page_030.txt

hardison_l_Page_083.txt

hardison_l_Page_073.txt

hardison_l_Page_115.txt

hardison_l_Page_138.txt

hardison_l_Page_070.txt

hardison_l_Page_109.txt

hardison_l_Page_140.txt

hardison_l_Page_049.txt

hardison_l_Page_133.txt

hardison_l_Page_145.txt

hardison_l_Page_090.txt

hardison_l_Page_057.txt

hardison_l_Page_003.txt

hardison_l_Page_136.txt

hardison_l_Page_066.txt

hardison_l_Page_068.txt

hardison_l_Page_020.txt

hardison_l_Page_072.txt

hardison_l_Page_105.txt

hardison_l_Page_040.txt

hardison_l_Page_082.txt

hardison_l_Page_094.txt

hardison_l_Page_024.txt

hardison_l_Page_018.txt

hardison_l_Page_038.txt

hardison_l_Page_141.txt

hardison_l_Page_011.txt

hardison_l_Page_059.txt

hardison_l_Page_112.txt

hardison_l_Page_031.txt

hardison_l_Page_081.txt

hardison_l_Page_104.txt

hardison_l_Page_060.txt

hardison_l_Page_148.txt

hardison_l_Page_026.txt

hardison_l_Page_054.txt

hardison_l_Page_130.txt

hardison_l_Page_028.txt

hardison_l_Page_076.txt

hardison_l_Page_149.txt

hardison_l_Page_023.txt

hardison_l_Page_080.txt

hardison_l_Page_053.txt

hardison_l_Page_077.txt

hardison_l_Page_001.txt

hardison_l_Page_075.txt

hardison_l_Page_095.txt

hardison_l_Page_061.txt

hardison_l_Page_069.txt

hardison_l_Page_142.txt

hardison_l_Page_055.txt

hardison_l_Page_121.txt

hardison_l_Page_033.txt

hardison_l_Page_048.txt

hardison_l_Page_085.txt

hardison_l_Page_116.txt

hardison_l_Page_124.txt

hardison_l_Page_014.txt

hardison_l_Page_042.txt

hardison_l_Page_029.txt

hardison_l_Page_110.txt

hardison_l_Page_086.txt

hardison_l_Page_078.txt

hardison_l_Page_044.txt

hardison_l_Page_092.txt

hardison_l_Page_126.txt

hardison_l_Page_056.txt

hardison_l_Page_067.txt

hardison_l_Page_041.txt

hardison_l_Page_096.txt

hardison_l_Page_111.txt

hardison_l_Page_120.txt

hardison_l_Page_084.txt

hardison_l_Page_127.txt

hardison_l_Page_099.txt

hardison_l_Page_088.txt

hardison_l_Page_034.txt

hardison_l_Page_007.txt

hardison_l_Page_009.txt

hardison_l_Page_091.txt

hardison_l_Page_047.txt

hardison_l_Page_063.txt

hardison_l_Page_123.txt

hardison_l_Page_097.txt

hardison_l_Page_103.txt

hardison_l_Page_147.txt

hardison_l_Page_137.txt

hardison_l_Page_025.txt

hardison_l_Page_051.txt

hardison_l_Page_065.txt

hardison_l_Page_100.txt

hardison_l_Page_102.txt

hardison_l_Page_050.txt

hardison_l_Page_128.txt

hardison_l_Page_107.txt

hardison_l_Page_015.txt

hardison_l_Page_101.txt

hardison_l_Page_052.txt

hardison_l_Page_017.txt

hardison_l_Page_108.txt

hardison_l_Page_005.txt

hardison_l_Page_118.txt

hardison_l_Page_036.txt

hardison_l_Page_143.txt

hardison_l_Page_098.txt

hardison_l_Page_039.txt

hardison_l_Page_150.txt

hardison_l_Page_079.txt

hardison_l_Page_046.txt

hardison_l_Page_117.txt

hardison_l_Page_089.txt

hardison_l_Page_087.txt

hardison_l_Page_064.txt

hardison_l_Page_008.txt

hardison_l_Page_004.txt

hardison_l_Page_119.txt

hardison_l_Page_074.txt

hardison_l_Page_006.txt

hardison_l_Page_125.txt

hardison_l_Page_129.txt

hardison_l_Page_010.txt

hardison_l_Page_135.txt

hardison_l_Page_022.txt

hardison_l_Page_043.txt


Full Text






Time-Resolved Fluorescence


Isotropic Upconversion

The fluorescence dynamics of PPE-CO2- with repeat unit lengths equal to 8, 35, 108 in

MeOH were excited at 375 nm and detected at magic angle at several different wavelengths.

Data was fit with a sum of exponentials using the following equation:


ft (-
I(t) = C A exp (-5


where A, represents the weight of each rate constant and r, is the associated time constant. Data

from these fits are summarized in Table 4-2.

1.2 g. .


1.0


0.8 --



S 0.6 --


0.4 --


0.2 --


0.0 ls llll
325 350 375 400 425 450 475 500

h (nm)

Figure 4-15. Excitation spectra 10 pLM 35 PRU PPE-CO2- in methanol with ~ 6 pLM Ca2+
Detection at 430 (-), 475 ( ), 510 (-), and 590 (-) nm

Figure 4-16 shows the time-resolved fluorescence decay of 8 PRU PPE-CO2- in methanol

at three different detection wavelengths (430, 436 and 450 nm) excited at 375 nm. Detection at

430 nm (blue line) shows a multi-exponential behavior which disappears as the detection









Core/Shell versus Alloys

Prior to our work, limited information appeared in the literature involving the synthesis of

colloidal ZnCdSe nanorods for use in optoelectronic devices. Green-yellow emitting ZnCdSe

nanorods were prepared by diffusion of Zn into the CdSe core. For alloying, the reaction vessel

containing CdSe/ZnSe nanorods was heated and stirred for up to 3 hrs. An aliquot was removed

after heating for 1, 2 and 3 hrs, immediately cooled and diluted with toluene to terminate the

alloying process and then precipitated with methanol/toluene co-solvents. The Raman data

presented in Chapter 2 indicate the enhancement of the ZnCdSe phonon mode but disappearance

of the CdSe and ZnSe modes, confirming the alloyed nanorod composition.(42)

Figure 3-6 compares the bleach spectra of the core/shell and ZnCdSe alloyed samples

using the delay times of 0.400, 2.47, 150 and 575 ps. Note that this data is presented differently

than in Figure 3-3. Each plot includes the four samples at one particular delay time. For example,

Plot 1 corresponds to the transients for each sample at a probe delay time of 400 fs. This data

indicate the occurrence of a transformation of the band gap, band structure and surface-trap

states as function of alloying.

At early time delays (0.400 ps Plot 1), the 1S band corresponding to the 2 hr alloy (green

line) is much more intense than all other samples and is significantly blue shifted compared the

core/shell nanorods (black line). After 2.47 ps (Plot 2) the 1S band in all samples is maximized

and the wavelength shift mentioned previously is much more evident. The 1 hr alloy remains

very broad even after 575 ps (Plot 4). The overall signal obtained from the alloys is not as

intense as the core/shell and the 1 and 2 hr alloys are not as intense as the 3 hr alloy.

Photoinduced absorption present in the core/shell materials does not appear in the alloys. This

will be discussed in further detail later. At 150 ps (Plot 3), the alloy bleach has noticeably

decayed; meanwhile the core/shell has not decayed significantly. Even after 575 ps, the









of the energy states of the materials.(125) Due to the quantization of nanocrystals, the spacing

between the energy levels for the electrons is quite large, reaching values much greater than

longitudinal optical (LO) phonons found in bulk semiconductors (~ 25 meV). Cooling via

phonons is possible; however, it requires a simultaneous emission of a substantial amount of

phonons, which has low probability. Therefore, it was assumed for many years that the

relaxation of these excitons should be inhibited, due to this "phonon bottleneck", resulting in

nanosecond cooling times.(122, 126) The electron relaxation from the 1P-to-1S in CdSe

nanocrystals occurs in the subpicosecond timescale (faster than even in bulk) thus bypassing this

bottleneck.(10, 17, 122) Klimov was able to extract population dynamics of the 1S and 1P states

determining that a 1P to 1S relaxation rate of ~ 300 fs and a 1S buildup time depends on the

confinement enhancement and decreases as the nanocrystal radius decreases.(18) This fast

relaxation process is Auger in nature in that the Coulomb interaction between the electron and

hole, which is increased due to quantum confinement, allows the electron to relax but transfer

excess energy to the hole, scattering it deeper into the valence band.(10, 31, 122) This strong

coupling between the electron and hole in quantum systems has allowed for predictions of

efficient electron-hole energy transfer to occur within 500 fs (12 7) to 2 ps (128) though it is

difficult to measure these values directly. Phonon assisted relaxation of the hole is more probable

due to smaller energy spacing within the valence band.(122)

Using infrared transient absorption(129) and terahertz spectroscopy(122) several groups

have attempted to correctly identify and confirm the Auger relaxation mechanism where the

electron transfers excess energy to the hole. The distinct photoabsorption features present in

transient spectra is very useful in identifying different photo-excited species. However, broad

band spectra are sometimes difficult to interpret and assignment of various species becomes









Detecting at 550 nm, there is no contribution from the fast component to the 35 PRU signal, but

it is still present in the 108 PRU signal.

The photoluminescence rise follows the response function of the experimental setup. At

very long wavelengths one would expect to see a build up of population (rise time) due to energy

transfer from the isolated states however, the data collected do not show this slow rise time.(156,

196) This indicates that the singlet exciton is transferred from shorter (high energy) chains to

traps within the chains, not to aggregates. The right column in Figure 4-17 shows the

intermediate and long decays, observed at different detection wavelengths for the 35 and 108

PRU samples. Detection at 430 nm (top graph) presents an intermediate decay constant of 11+1

ps and a long decay of 178 & 6 ps. For longer detection wavelengths, the intermediate time

constant is ~ 30 to 40 ps while the long time decay lies between 350 and 450 ps. Results are

summarized in Table 4-2. The qualitative trend observed for the intermediate component is the

same for the 35 and 108 PRU but not 8 PRU. In the 450 to 590 nm detection region, as the

wavelength increases, the amplitude of the intermediate decay increases but the lifetimes do not

change significantly. These time-resolved emission signals result from an ensemble of isolated

CPE chains with different conjugation lengths. Meanwhile, the slow decay time increases as the

wavelength increases. A 350 to 450 ps time constant corresponds to the isolated chain natural

fluorescence lifetime (extracted from the 8 PRU detected at 450 nm); therefore, this component

is assigned to isolated chains not participating in the energy transfer process. Energy transfer is

not only dependent on spectral overlap of the donor fluorescence (isolated chain) and acceptor

absorption (aggregate chain), but also on the molecular distance between the two species.(2)

After initial energy transfer (<1.5 ps) from short conjugated segments to traps, the energy is then

transferred to the aggregate species (intermediate decay time). The change in amplitudes of the









10,000. An additional voltage forces the electron cloud to hit the phosphor screen located at the

front of the fiber optic exit window. The voltages between the photocathode and MCP can be

controlled in a manner so that the image intensifier can be quickly turned off and on, effectively

creating an electronic gate.(89, 90)

As photons hit the surface of the CCD sensor, electrons are generated which are stored in

individual pixels. The maximum number of electrons that one pixel can accumulate during

integration is considered to be the full well capacity. A 16 bit analog to digital output converter

which is capable of digitizing 65,536 levels (216) Of light is used to read out the pixels.(89, 91-

93) The dynamic range is the number of steps or levels of light intensity that can be represented

per bit.(93) Without using an image intensifier (gain), the CCD dynamic range (maximum and

minimum signal intensities that can be measured simultaneously) is defined as the full well

capacity per pixel divided by the read noise.(89, 90, 92, 94) For this system, the full well

capacity per pixel is 300,000 electrons and the read noise is 4 electrons resulting in a potential

dynamic range of 75,000 to 1. This value exceeds the upper constraint of the digitizer so the

dynamic range is instead limited to 65,000 to 1. Dynamic ranges can vary since the read noise

depends on the read out rate. If the read out is fast, the read noise can be high and the dynamic

range can be low or vice versa. A slower read out rate will reduce the read noise (high read noise

will affect the quality of the image). Depending on the application, cameras that have a high well

capacity and low read noise (high dynamic range) in addition to a large analog to digital

conversion capability are optimal.(90, 92) The response of this camera is considered to be linear

(1) within its full dynamic range.(89)

The dynamic range of the CCD serves as the base dynamic range of the ICCD camera

system. As gain is added, the dynamic range is reduced.(90) For example, if the gain is set to 50











-1 12 345


I n l l e I n
0 100 200 300 400 500
time (ps)


-~------


r


CdSe Rods


B)


-1 0 2 a


CdSe/ZnSe


I I


I )


It~


200


300


Figure 3-4. Kinetic traces corresponding to the 1S (
rods and B) CdSe/ZnSe rods.


-), 1P ( ) and 2S (-) bands for A) CdSe


time (ps)










(80). W. E. Garner, Chentistry of the solid state (Butterworths Scientific Publications,
London, 1955).

(81). W. E. Martin, Photoluminescence Determinations of Cd Diffusion in ZnSe, Journal
ofApplied Physics (1973) 44, 5639.

(82). P. J. Parbrook, B. Henderson, K. P. Odonnell, P. J. Wright, B. Cockayne,
Interdiffusion in wide-bandgap Zn(Cd)S(Se) strained layer superlattices,
Semiconductor Science and Technology (1991) 6, 818.

(83). A. Rosenauer, T. Reisinger, E. Steinkirchner, J. Zweck, W. Gebhardt, High-resolution
transmission electron-microscopy determination of Cd diffusion in CdSe/ZnSe single-
quantum-well structures, Journal ofCrystal Gi 1,n thr (1995) 152, 42.

(84). M. Strassburg, M. Kuttler, U. W. Pohl, D. Bimberg, Diffusion of Cd, Mg and S in
ZnSe-based quantum well structures, Thin Solid Films (1998) 336, 208.

(85). W. Chen, J. O. Malm, V. Zwiller, R. Wallenberg, J. O. Bovin, Size dependence of
Eu2+ fluorescence in ZnS : Eu2+ nanoparticles, Journal ofAppliedPhysics (2001) 89,
2671.

(86). W. Chen, R. Sammynaiken, Y.N. Huang, Crystal field, phonon coupling and
emission shift of Mn2+ in ZnS : Mn nanoparticles, Journal of Applied Physics (2001)
89, 1120.

(87). C. X. Shan, X.W. Fan, J.Y. Zhang, Z.Z. Zhang, B.S. Li, Y.M Lu, Y.C. Liu, D.Z.
Shen, X.G. Kong, X.H. Wang, Growth and evolution of ZnCdSe quantum dots,
Journal of Vacuum Science & Technology B (2002) 20, 1 102.

(88). X. Y. Wang, J. Y. Zhang, A. Nazzal, M. Darragh, M. Xiao, Electronic structure
transformation from a quantum-dot to a quantum-wire system: Photoluminescence
decay and polarization of colloidal CdSe quantum rods, Applied Physics Letters
(2002) 81, 4829.

(89). U. Manual, A User's Guide to the Andor iStar (Andor Technology Limited, 2001).

(90). Andor, Andor Knowledge Lib rary, http://www.andor. com/l ib rary/di gital cam era s/

(91). J. Alford, Personal communication concerning the iStar CCD system,(2007)

(92). S. Cannistra, How to choose a CCD camera,
http://www. starrywonders. com/ccdcameraconsi derati ons. html

(93). M. C. Gino, Noise, Noise, Noise, http://www.astrophys-
as sist. com/educate/noi se/noi se.htm

(94). Apogee, CCD University, http ://www.ccd.com/ccdu.html












TABLE OF CONTENTS


page

ACKNOWLEDGMENTS .............. ...............4.....


LIST OF TABLES ............ ....._._. ...............9.....


LIST OF FIGURES .............. ...............10....


AB S TRAC T ............._. .......... ..............._ 13...


CHAPTER


1 INTRODUCTION .............. ...............15....


Study Overview ......_._ ................ ...............15.......
Photophysics Concepts ................. ...............16.......... .....
Energy Transfer ................. ...............18.......... ......
Radiative Energy Transfer............... ...............19
Non-radiative Energy Transfer............... .... ................1
Random Walk Migration (Intrachain Energy Transfer) ................. ........................20
Emission Measurements ................. ...............22.................
Transient Absorption .............. ...............24....


2 QUANTUM NANOPARTICLES .............. ...............27....

Overview. ............ _.... .....__ _.... .............2
Bulk vs. Quantum Semiconductors .............. ...............27....
Size and Shape Dependence .............. ...............33....
Passivation ........._._... .... ...._.__......_. .............3

Composition Changes: Interdiffusion................ ... ..... ... ..... ........3
Experimental Methods: Nanorod Synthesis and Composition Characterization ...................41
Preparation of ZnCdSe Nanorods. ............ .....__ .......__ ........ 4
Steady State Instrumentation ................. ......__ ..... ............4
Time-Resolved Photoluminescence Instrumentation ......___ .......___ ................43
Results and Discussion .............. .. ...............46...

Synthesi s of ZnCd Se Nanorods ............._. ...._... ...............46...
Structure of ZnCd Se Nanorods ............_. ...._... ...............47...
Effect of Alloying on the Phonon Spectra ....._.._._ ........___ .....__ ..........4
Photoluminescence and Absorption Properties ................. ...............................52
Time-Resolved Photoluminescence (TRPL) ................. ........__ ......... 55.... ....
Sum m ary ................. ...............62........ ......


3 QUANTUM PARTICLE ELECTRONIC STRUCTURE ................. ......__ ..........._._.63

Introducti on .........._.... .. .. .. _. ..... ... ...............63

Experimental Methods: Transient Absorption............... ...............6









0.1 (108 PRU). Zhao et al. suggest that conformational, vibrational and rotational degrees of

freedom creating non-radiative decay channels lead to decrease fluorescence.(181) On the

contrary, it is clear that the large absorption red shifts due primarily to the rigidity of the

polymers lead to large emission and absorption spectral overlap. Moreover, the conformational

restrictions induced by this rigidity give rise to conjugation lengths longer than expected

ultimately reducing the degrees of freedom within the polymer and thus potentially makes the

stated reason an invalid argument to explain small quantum yields for long PPE-CO2- chains.

Figure 4-11 shows the emission spectrum of the PPE-CO2- (3 5 PRU). The black line

corresponds to the polymer dissolved in methanol and it shows the sharp bands characteristic of

isolated chain emission. The red line corresponds to a methanol solution in which 6 pLM of Ca2+

has been added. Ca2+ is an effective cross linker with the 2 carboxyl groups inducing aggregation

of the PPE-CO2-.(155) Emission from the aggregate can be clearly observed on the shoulder at

520 nm, as it grows relative to the unaggregated emission at 436 nm. Finally, when the CPE is

dissolved in water, the aggregated emission is mostly observed (green line). Overall, the red

shift, quenching and band broadening are due to aggregate formation of the polymer chains since

Ca2+ is a closed-shell ion and does not act as an electron or energy acceptor. (191-194) The small

Stoke' s shift previously mentioned results in excellent overlap of the isolated chain emission

with the aggregate absorption enhancing energy transfer from the higher energy isolated species

to the lower energy aggregates.

The excitation spectrum of a given chromophore is determined by monitoring the

fluorophore emission as it is excited at different wavelengths. Abramowitz et al. from the

Olympus" Microscopy Resource provides an excellent description of the excitation spectra

collection process.









based on the number of exponentials and an extensive number of parameters required to fit the

decay curve resulting in exact decay lifetimes. This can become very complicated and can lead to

incorrect assignments of the photophysical processes occurring within the material. Jones et al.

found that they were able to fit their photoluminescence data collected from CdSe/ZnS core shell

quantum dot' s decays with a multi-exponential function, implying that there is an existence of

several discrete relaxation pathways, with individual lifetimes. However, they were not able to

claim the exact number and identity of such pathways.(117) For simultaneous measurements of a

large ensemble of relaxation times it is more advantageous to use a stretched exponential (non-

exponential) function to evaluate the distribution of relaxation times in such dispersive

systems.(43, 116) This type of equation encompasses both independent, single step processes in

addition to sequential multi-step processes.(116)


I(t) = Io exp -~l (2-6)


Where z is the characteristic lifetime and (0 < P < 1) is the dispersion exponent. Despite only

extracting an average lifetime from a non-exponential; the function provides a phenomenological

description that is considered purely empirical, fitting data with a minimum number of

parameters. These parameters can vary depending on the phenomenon of interest and external

variables such as temperature.(43) For the limiting case of P- 1, we get the single exponential

decay with the characteristic lifetime, z. For ideal, single quantum dots, we can expect P=1. It

should be mentioned that P<1 results from superposition of many exponential decays and as P

approaches zero, the distribution of decay times increases. This decay law can then be used to

compare different samples qualitatively in terms of non-uniformity or topological disorder.(42)










(counts/electron) a multiplication factor is employed which reduces the dynamic range from

65,000 counts to 13 10 counts (65,536 divided by 50). Gain can be advantageous if used in

appropriate amounts. For instance, the read noise produced by the CCD section of the camera is

no longer an issue. However, the dark current that is created thermally by the photocathode prior

to the amplification stage is still present and can also be amplified when gain is applied. Thus,

even though high gains will lead to enhanced signals, the noise is also increased. It is also

possible to have too little gain, which can sacrifice the well depth with no significant signal

amplification. A balance between the system gain and dynamic range is necessary to achieve the

best signal to noise ratio. Once gain is applied, the response of the CCD remains linear within its

dynamic range; however, the signal to noise equation is changed by the gain noise factor of

1.4.(89-92, 94)

Results and Discussion

Synthesis of ZnCdSe Nanorods

Combinations of surfactants such as TDPA and TOPO are generally used to prepare

nanocrystals since they have strong binding energies that ultimately raise the surface energies of

a crystal face compared to another. (42, 95, 96) Previously, Zn-TDPA and Cd-TDPA in a TOPO

solution were utilized in order to synthesize ZnCdSe nanorods. However, this method was not

successful, which most likely resulted due to the different reactivity with Se-TOP leading to a

lack of crystallite shape control.(42, 73, 97) It is also suggested that the temperature be higher

and reaction time be longer in order to promote a more thorough complexation of ZnO with

TDPA. When synthesizing the ZnCdSe ternary heterostructures, it is important to first prepare

CdSe nanorods. Once the CdSe rod has been grown, a Zn-Oleate and Se-TOP mixture was used

to grow the shell overtop the core. Zn-oleate and Se-TOP mixture was slowly added to prevent

homogeneous nucleation. It is extremely important that the temperature be controlled properly.















A)











150 200 250
Raman shift(cm l)






B) mOCdSe


CO C

C U. O ZnSe
i ZnCdSe
E *






150 175 200 225 250 2.'5
Raman shift(cm-1)

Figure 2-9. Raman spectra of LO phonon mode of A) CdSe nanorods and B) CdSe/ZnSe
core/shell nanorods. Adapted from H.Lee.(66)























h(n m)











0 20 40 60 80 100 120 140


1.0-

A) 0 8
0.6-

2 0.4-

0.2-

0.0-
600


620 640 660 680


Time (ns)

Figure 2-15. CdSe/ZnSe Core/Shell Photoluminesence: A) Broad band spectra at 8.8 ns (
25 ns (-) and B) Kinetic traces for 645 (-), 670 (-) and 630 ( ) nm.


-) and


0.01


0 20 40 60 80 100 120 140 160
Time (ns)


Figure 2-16. TRPL decay curve of CdSe/ZnSe nanorod (-),
3hr alloy (-)


1 hr alloy (-


-), 2 hr alloy (


), and









addition, the effects alloying has on the excited state of the ZnCdSe nanorods are discussed. Two

models, one for the core/shell and one for the alloy, are proposed to describe the relaxation

processes observed in each of the experiments.

The dynamics of the energy transfer from isolated to aggregated species in a series of

different polymer repeat unit sizes of a conjugated polyelectrolyte, PPE-CO2 Synthesized by

Xiaoyoung Zhao in Dr. Schanze's lab are discussed (Chapter 4). Anisotropy measurements

confirm that this polymer is very rigid and the conjugation length is longer than expected. The

data is analyzed to extract the influence aggregation has on the isolated chain emission.

Finally, Chapter 5 summarizes each project and states general conclusions drawn from the

results collected for this dissertation. Suggestions are made for potential applications for which

semiconductor nanoparticles presented in this dissertation may be useful. Also, an additional

molecule, similar to the PPE-CO2~, iS presented as the next step in a series of polymers to

investigate for chemo-or bio sensors.

Photophysics Concepts

Spectroscopy methods, whether they be time-resolved or steady state, provide numerous

ways to measure emission of materials that are intended to be used in a wide variety of

applications including opto-and electronic, biomedical, and chemical research. This dissertation

focuses on the photophysical properties of nanocrystals and energy transfer mechanisms that

induce the amplified quenching capabilities of conjugated polyelectrolytes.

It is important to know the multiple photophysical processes that an excited chromophore

can undergo between the absorption and emission of light. These processes are dictated by the

probability that a transition from an initial state to a Einal state can occur. By using time-

dependent perturbation theory, Fermi's Golden Rule for transitions between two states

corresponds to a transition rate equal to: (1, 2)









El-Sayed et al. synthesized and compared CdSe rods with aspect ratios ~ 3 (length/width)

and dots of 4.2 nm diameter.(53) TEM images show that the particles are different although the

steady state absorption spectra do not indicate significant differences. Electron-hole dynamics

measured by femtosecond pump-probe, although still not completely understood, show quite

different behavior for rods versus dots. This is confirmed by the increase in the number of bands

in the deconvoluted absorption spectra of the quantum rods.(1 7, 18, 53) Moreover, they observed









Traps Front
Zoomed
View

Figure 2-4. Drawing of a nanorod with each of the axis labeled. The front zoomed view shows
that the surface curvature is not smooth, leading to surface traps.

a significant increase in the carrier relaxation time in the quantum rods compared to the quantum

dots.(53) Nanodots have a higher order of symmetry, which is lost in the rods. Extension of the

c-axis results in a splitting of the degenerate level in the symmetric quantum dot (30, 53-60) and

that energy level splitting could be one reason for El-Sayed' s results. Due to the large surface-to-

volume ratio at the surface in nanorods, electron and holes have a high probability of being

trapped by surface impurities. However, the quantum dot curvature can create a larger number of

localized surface trap states than the elongated nanorods,(53) which allow for the carriers to have

"free" motion in the c-direction, reducing the probability of the carrier to be trapped as quickly.

In some cases, the impurities present enable the materials to be used in oxidation-reduction

chemistry, more specifically photocatalysis, (15) photodegradation and detoxification of










I, + 2 I
I
magic angle


(4-14)


Tilted


8,R





LIGHT


LIGHT


Figure 4-7. Berek polarization compensator. Tilting the crystal causes retardance and
birefringence. Adapted from New Focus.(185)


Input: Wave Plate
Linearly Setting: 3/2
Polarized


Output:
Linearly
Polarized -
900 rotated


Figure 4-8. Berek compensator used as a half-wave plate. I
orientation. Adapted from New Focus.(184)


retardance indicator, O


No Tilt


ne)


no



o e










(108). C. Ramkumar, K. P. Jain, S. C. Abbi, Resonant Raman scattering probe of alloying
effect in GaAsl-xPx ternary alloy semiconductors, Physical Review B (1996) 54, 7921.

(109). C. Ramkumar, K. P. Jain, S. C. Abbi, Raman-scattering probe of anharmonic effects
due to temperature and compositional disorder in III-V binary and ternary alloy
semiconductors, Physical Review B (1996) 53, 13672.

(110). T. Kummell et al., Size dependence of strain relaxation and lateral quantization in
deep etched CdxZn1-xSe/ZnSe quantum wires, Physical Review B (1998) 57, 15439.

(111). P. R. Yu, J. M. Nedeljkovic, P. A. Ahrenkiel, R. J. Ellingson, A. J. Nozik, Size
dependent femtosecond electron cooling dynamics in CdSe quantum rods, Nano
Letters (2004) 4, 1089.


(112). M. Achermann, J. A. Hollingsworth, V. I. Klimov, Multiexcitons confined within a
subexcitonic volume: Spectroscopic and dynamical signatures of neutral and charged
biexcitons in ultrasmall semiconductor nanocrystals, Physical Review B (2003) 68,
245302.

(113). Y. Kawakami, K. Omae, A. Kaneta, K. Okamoto, Y. Narukawa, T. Mukai, S. Fujita,
In inhomogeneity and emission characteristics oflInGaN, Journal ofPhysics-
Condensed Matter (2001) 13, 6993.

(114). H. S. Kim, R. A. Mair, J. Li, J. Y. Lin, H. X. Jiang, Time-resolved
photoluminescence studies of AlxGal-xN alloys, Applied Physics Letters (2000) 76,
1252.

(115). X. H. Zhong, Y. Y. Feng, W. Knoll, M. Y. Han, Alloyed ZnxCdl-xS nanocrystals with
highly narrow luminescence spectral width, Journal of the American Chemical
Society (2003) 125, 13559.

(116). R. F. Mahrt, T. Pauck, U. Lemmer, U. Siegner, M. Hopmeier, R. Hennig, H. Bassler,
E.O. Gobel, P.H. Bolivar, G. Wegmann, H. Kuz, U. Scherf, K.Mullen, Dynamics of
optical excitations in a ladder-type pi-conjugated polymer containing aggregate states,
Physical Review B (1996) 54, 1759.

(117). M. Jones, J. Nedeljkovic, R. J. Ellingson, A. J. Nozik, G. Rumbles,
Photoenhancement of luminescence in colloidal CdSe quantum dot solutions, Journal
of Physical Chemistry B (2003) 107, 1 1346.

(118). X. Chen, B. Henderson, K. P. Odonnell, Luminescence decay in disordered low-
dimensional semiconductors, Applied Physics Letters (1992) 60, 2672.

(119). R. Cingolani et al., Exciton spectroscopy in Znl-xCdxSe/ZnSe quantum-wells,
Physical Review B (1995) 51, 5176.












7x105


6x105


5x105


4x105


3x105


2x10 -


1x10 -




400 450 500


550 600


650 700


h (nm)


Figure 5-2. Photoluminescence spectra of hyperbranched PPE-CO2- in MeOH (-) and water









have not been able to exhibit as efficient CM efficiencies in the visible region.(199)

Investigations into the effect that shape has on this phenomenon is recommended.

PPE-CO2 COnclusions and Future Work

Conclusions

The synthesis, characterization and time-resolved measurements conducted on this series

of PPE-CO2~ pOlymers with different chain lengths and extent of aggregation have led to several

interesting observations and conclusions. A complete steady state and time-resolved study of

length, solvent and aggregated inducer has been conducted in our labs. In particular, isotropic

and anisotropic fluorescence up-conversion was utilized to understand the quenching mechanism

within PPE-CO2-. We conducted a wavelength detection study based on exciting on the blue side

(isolated chain) of the absorption. Upon excitation of the aggregates from energetically higher

lying isolated chains, the fluorescence lifetimes result in multi-exponential behavior due to the

competition between multiple decay pathways. (116, 153, 156, 165, 215)

In all samples, the emission is inhomogeneously broadened. Detection of time-resolved

signals at all wavelengths is associated with the isolated chains emission, despite the

superposition of both species present within the samples. In addition, unlike PPE-SO3-, PPE-

CO2~ is Very rigid resulting in longer conjugation lengths than 4.5 PRU. This was determined not

by evaluating the behavior of the polymer chains as a function of chain length using steady-state

absorption but by evaluating the extremely slow time-resolved anisotropy decay. From the

excitation spectra collected for each polymer chain, it is seen that even a dilute PPE-CO2- with

only 8 polymer repeat units exhibits a slight amount of aggregation when detected at very red

wavelengths. Moreover, we conclude that the energy transfer from isolated to aggregated chains

is extremely fast, occurring in 30 to 40 ps for all samples. This component was significantly

enhanced when a poor solvent (water) or an aggregate inducer was used. An even faster










Summary

Transient absorption spectroscopy was utilized to extract the exciton dynamics within

binary CdSe/ZnSe core/shell and ternary ZnCdSe nanorods. A comparison between the exciton

behavior in unpassivated CdSe core and CdSe/ZnSe core/shell materials, an excitation

wavelength dependence for the core/shell nanorods and the influence alloying has on the exciton

behavior are all presented. For all samples, at high energy excitation, a 1P decay and subsequent

1S rise is observed corresponding to a 1P to 1S relaxation process. Also, the introduction of a

midgap state in the core/shell material leads to photoinduced absorption after the 1P bleach

recovers. Upon low energy excitation, this midgap state is directly populated. Surface trap states

reappear in the alloyed heterostructures (no passivation) leading to faster bleach recoveries then

the core/shell materials.









The Model 5540 Berek polarization compensator from New Focus" was used in these

experiments to convert and control the pump polarization. A compensator such as this utilizes

the principal that different wavelengths of light propagate at different speeds through a medium

and that this velocity depends on the index of refraction. This compensator can cause a %4-wave

or '/-wave retardance for wavelengths in the ultraviolet (200 nm) to the infrared (1600 nm). The

compensator has a 12 mm aperture and was directly mounted to a post. The Berek compensator

is made up of a single birefringent uniaxial plate with an adjustable tilt angle to impose velocity

changes on incident light resulting in retardation. The velocity changes are both tilt angle and

wavelength dependent. The extraordinary axis, ne, is oriented perpendicular to the plate while the

ordinary axis, no, is parallel (Figure 4-7). If no tilt is imposed, the incident light remains normal

to the plate. As the light propagates through the medium, its velocity remains unaffected by the

polarization and is only dependent on the ordinary index of refraction. If the plate is tilted to a

particular angle, 8R, the velocity of the propagating light is changed. The axis oriented in the

plane of incidence is no longer ordinary, instead it has an "extraordinary" component, ne',

causing retardation. Polarized light that is perpendicular to the plane of incidence has a velocity

unaffected by the tilt. As a result, there is a retardance that is created between the ordinary and

extraordinary waves propagating in the polarization planes. The main advantage of using a Berek

compensator for polarization measurements is that it allows for simple and independent

adjustments for not only retardation but also plane of incidence orientation adjustments (which

are both wavelength dependent) as one unit. The retardation knob is used to set the tilt angle

while the orientation knob acts as a wave plate.(184, 185)

Due to group velocity dispersion, the retardance of the electric field is wavelength

dependent. Therefore, one must set the correct position of the Berek compensator polarization










component (1.5 ps) was also observed and it was assigned to the transfer from shorter (high

energy) isolated chains to longer chains and traps within the isolated chain backbone.

Outlook/ Future Work (Hyperbranched PPE-CO2?

Within our collaboration with the Schanze and Reynolds groups, we have an opportunity

to investigate excitation and relaxation mechanisms for materials in solutions and in films using

both fluorescence up-conversion and broad band transient absorption techniques. Xiaoyong

Zhao, a student in the Schanze group, has synthesized a new hyperbranched PPE-CO2- in which

there are three carboxylate side chains attached on each side of the polymer backbone. Based on

excitation spectra data presented in Chapter 4, even dilute solutions of the linear 8 polymer

repeat unit displays some sort of aggregation within the sample. This new, hyperbranched

polymer is said to have no aggregation present even if dissolved in water. Figures 5-1 show the

absorption spectra of this new polymer when dissolved in different solvents and compared to the

linear 8 PRU (data collected by Xiaoyong Zhao). From this figure, it is necessary to look into the

excitation spectra of the hyperbranched polymer to compare to the dilute 8 PRU. The small

shoulder present in Figure 5-1 on the red side of the absorption is a small indication that there

may still be some aggregates present within this hyperbranched sample. However, this shoulder

could simply be an artifact of the size and repeat unit distributions present within polymeric

samples. The only way to be sure is to collect excitation spectra at various wavelengths and look

for red shifts that are characteristic of aggregate species. The polymer dissolved in water does

not alter the absorption spectra significantly. However, photoluminescence is quenched by nearly

half of its original intensity when dissolved in methanol (Figure 5-2). This is another indication

that the polymer could have some aggregate present despite the lack of structure in the

absorption spectrum.










MOs = Ic Electron
Intraband

~5~3 Transition

Cd2' Enegy 1 Interband

EnergyTransition

h'

Sel- Intra aend

Se2- Tasto
MOs = Ici


Figure 3-1. Electronic structure in semiconductor nanoparticles. Adapted from M. El-Sayed.(13)

After the initial excitation, the electron can only be further excited to higher states of the

cation MOs while the hole can only be further excited to other anion MOs (intraband transitions).

The recombination of the electron and hole from the conduction band to the valence band

involves an interband transition which can be directly detected in the visible spectrum. The

transient absorption signals detected in the visible region only reveal the behavior of this bound

exciton. Intraband relaxation of either electron or hole dynamics for strongly confined

nanocrystals are detected independently in different spectral regions. Since the energy spacing

between the levels within the cation MOs is much larger than the energy spacing between the

anion MOs the intraband excitation or relaxation of the hole intraband transfers is detected at

lower energies (infrared region) than the electron intraband transfers.(13)

The relaxation processes from higher to lower excited states within nanocrystals are

extremely intriguing and counterintuitive. Unlike in bulk materials where the cooling of the

photo-generated carriers is rapid and occurs via lattice phonons through its conduction band

continuum, carrier cooling in quantum particles must occur in discrete steps based on the nature









where z represents the time delay between the arrival of the gate pulse with respect to the sample

fluorescence. This optical gating technique is very advantageous because the time resolution is

dependent only on the width of the gate and pump pulses, not the detection system.(182) The

optical path length was 2 mm and the concentration of samples did not exceed 30 pLM yielding an

optical density ~ 0.45/mm. A circulating cell was used to ensure that a fresh volume of sample

was excited with every laser shot and a maximum of 100 nJ of energy per shot were used.

Anisotropy is the measurement of the extent of polarization that a material maintains after

being excited with polarized light. When the emission anisotropy is nonzero, the emission of the

material is polarized. The transition dipole moment (ya) of a molecule dictates which orientation

or direction molecules will absorb light. Light that is polarized consists of an electric field (E)

that oscillates in a particular direction. Excitation of a material with linearly polarized light

results in an excitation probability function that is proportional to the square of the scalar product

of the molecules dipole moment and the electric field vector (ya. E or cos2 8A) (Figure 4-6). The

phenomenon of polarized emission is dependent on the absorption and emission transition dipole

moments which can be oriented at different angles to one another. When the angle between the

two vectors is 900, the excitation probability is zero and maximized if they are parallel. After

creating an exciton in a high energy electronic state of an anisotropic material, it relaxes to the

first singlet state (Kasha' s Rule (2)) via internal conversion. Regardless of the orientation of the

transition moment of the high energy initial state, the emissive transition moment at the first

singlet state will remain the same (Figure 4-5). If the absorption and emission moments are

identical, the anisotropy will not be lost; however, if they differ, Figure 4-5, the anisotropy value

will change.(2)









Several studies have investigated the influence aggregates have on the kinetics within

water soluble conjugated polymer systems. (139, 152, 154, 1 73-1 79) For example, Fakis et al.

has shown that energy transfer from isolated poly(fluorenevinylene-co-phenylenevinylene

(PFV-co-PV) (156) to aggregated chains is very rapid and efficient. They determined the isolated

chain fluorescence, the aggregate emission and energy transfer contributions to the overall decay.

In addition, the correlation between the concentration and energy transfer efficiency was

thoroughly examined. A reduction in the concentration causes the energy transfer efficiency and

energy transfer rates to decrease linearly.(156) In this thesis, we investigate the influence chain

length, solvent and metal cations have on the ultrafast emission of a carboxylated

poly(phenylene) vinylene (PPE-CO2 ) Shown in Figure 4-2 to determine the excitation transport

processes. The energy transfer mechanism between isolated and aggregated chains within the

PPE-CO2~ pOlymer is of particular interest.





SPolymer Q~tCuencher


Figure 4-1. Intrachain energy transfer of excitation to quencher molecule along polymer
backbone.


CO2Na+

O






PPE-CO2

Figure 4-2. PPE-CO2~ pOlymer repeat unit. PPE-CO2~ in methanol (left) and water (right)









shell exhibits an 11% smaller lattice parameter causing the core to be compressed, whereas the

lattice parameter in the c-axis for the core/shell material was ~ 1% smaller compared to the

CdSe. After interdiffusion of Zn into the core, the lattice parameter and the lattice mismatch

strain are reduced due to a lattice contraction. Also, after addition of Zn into the core, the

diffraction peaks shifted to a larger 26 indicating a smaller interplanar spacing.(38, 42)

A HR-TEM image of the ZnCdSe nanorods is shown in Figure 2-8. The diameter and

lengths of the nanorods measured from such images has been included in the histogram.
















20 nm scale bar 5 nm scale bar

Figure 2-8. HR-TEM image and histogram of size distribution of ZnCdSe nanorods. Lattice
fringe from a nanorod is shown in the lower right corner. Adapted from H. Lee.(66)

From this graph, the average diameter is ~6 nm and the average length is ~13 nm resulting in an

aspect ratio equal to ~ 2. 1 nm for the alloys.(42, 66) When using XRD, the diffraction patterns

can exhibit broadening effects due to particle size. Using the Debye-Scherrer formula, the

average crystallite size in A can be determined:


D, (2 -5)
/7 cos B

Where k is a correction factor to account for particle shapes, and P is the observed width at half

the maximum peak intensity and 6 is the Bragg angle. It must be noted that the observed width









Figure 1-1, several vibronic transitions of one state or in this case, a donor molecule, can be

isoenergetic to the corresponding transitions of the acceptor (D*- D and A A*). In general,

the non-radiative transfer rate is given by Eqn 1-1, (Fermi's Gold Rule) where the density (p) is

not only related to the coupling of the initial and final states capable of a transition (determined

by Frank-Condon factors) but also by the non-inhomogeneously broadened spectral overlap, J, of

the donor emission, ID(v),and acceptor absorption, E,(v), determined using Eq 1-2.(1, 2)


J =il ID 4 (v)dv (1-2)


This integral assumes that the relaxation within the excited state vibrational manifold is faster

than the energy transfer process and that energy transfer abides by the Franck-Condon principle

(vertical transition). As the number of resonant transitions between the donor and acceptor

increases, the likelihood for a non-radiative energy transfer process to occur increases since these

transitions are proportional to the overlap integral (Figure 1-2).(1, 2)

Random Walk Migration (Intrachain Energy Transfer)

In some cases, a quenching of the fluorescence occurs but can not be explained by a

bimolecular energy transfer mechanism. Upon excitation of a molecule, an excited state electron

and ground state hole pair are created, termed "exciton". If the molecule consists of multiple

segments that are equivalent in energy or are a cascade of energies (like a polymer with repeating

chromophores), this exciton can diffuse from one segment to another while remaining bound.

The exciton undergoes a mechanism that involves a "hopping" from one segment to another

within the same polymer. The "random walk" or intrachain energy transfer implies that the

electron and hole move together, and will always be located within the same chromophore thus

charge separation does not occur. Also, during the course of the energy diffusion, the energy is









ACKNOWLEDGMENTS

As I reflect on the number of years that have led up to this moment of earning a Ph.D. in

Physical Chemistry, I realize there are numerous people to recognize and say thank you to

because without their support and encouragement I would not have made it to this point.First and

foremost, I thank the Lord because without His mercy nothing is possible. I thank my advisor,

Professor Valeria D. Kleiman for her guidance, patience and consistent motivation throughout

my journey. I appreciate the effort and time she has put into helping me pursue this degree

including letting me explore new possibilities and career building activities such as working for a

summer at Corning, Inc as an intern. Valeria has continually believed in me and my work even

when I did not and it is this type of support and enthusiasm that has enabled me to finish this

proj ect.

I thank my supervisory committee members Professors Philip Brucat and Nico Omenetto

for their guidance and thought provoking discussions. My gratitude goes to Dr. Kirk Schanze,

Dr. Hui Jiang and Xiaoyoung Zhao for their CPE collaboration. I appreciate them providing the

polymers and their willingness to help satisfy the needs of the proj ect. I am also grateful for Dr.

Paul Holloway and Dr. Hyeokjin Lee asking for our assistance in their nanorod proj ect; it has

been an experience I truly enjoyed.

I express gratitude to the members, past and present, of the Kleiman Group. Thanks goes

to Dr. Juirgen Muiller for giving me a fundamental understanding of the transient absorption. I

thank Dr. Evrim Atas for her ongoing friendship, Turkish cooking and being my upconversion

mentor. There are no words that describe how much I appreciate Daniel Kuroda. I thank him for

not only his ability to answer all of my questions but also for his constant support and

encouragement and of course, his BBQing skills. I also want to thank "Cochuk", Aysun Altan,









LIST OF REFERENCES


(1). N. J. Turro, M~odern M~olecular Photochemistry (Univ Science Books, New York,
1991).

(2). B. Valeur, M~olecular Fluorescence Principles and Applications (Wiley-VCH,
Weinheim, 2002).

(3). G. S. H. Singhal, Janos; Rabinowitch, Eugene., Excitation-energy migration between
chlorophyll and b-carotene, Journal of Chemical Physics (1968) 49, 5206.

(4). Govindj ee, Excitation Energy Transfer and Energy M~igration : Some Basics and
Background, http:.//www.1ife.uiuc. edu/govindj ee/biochem494/foerster.htm, Online
Class Notes

(5). E. G. Rabinowitch, Phounpubesisl~\i (John Wiley & Sons, Inc., New York, 1969).

(6). G. D. Joly, L. Geiger, S. E. Kooi, T. M. Swager, Highly effective water-soluble
fluorescence quenchers of conjugated polymer thin films in aqueous environments,
Macromolecules (2006) 39, 7175.

(7). J. H. Wosnick, C. M. Mello, T. M. Swager, Synthesis and application of
poly(phenylene ethynylene)s for bioconjugation: A conjugated polymer-based
fluorogenic probe for proteases, Journal of the American Chemical Society (2005)
127, 3400.

(8). T. Kippeny, L. A. Swafford, S. J. Rosenthal, Semiconductor nanocrystals: A powerful
visual aid for introducing the particle in a box, Journal of Chemical Education (2002)
79, 1094.

(9). A. Hagfeldt, M. Gratzel, Light-induced redox reactions in nanocrystalline systems,
Chemical Reviews (1995) 95, 49.

(10). V. I. Klimov, D. W. McBranch, C. A. Leatherdale, M. G. Bawendi, Electron and hole
relaxation pathways in semiconductor quantum dots, Physical Review B (1999) 60,
13740.

(11). M. C. Schlamp, X. G. Peng, A. P. Alivisatos, Improved efficiencies in light emitting
diodes made with CdSe(CdS) core/shell type nanocrystals and a semiconducting
polymer, Journal ofApplied Physicsl (997) 82, 583 7.

(12). Y. Wang, N. Herron, Nanometer-sized semiconductor clusters Materials synthesis,
quantum size effects, and photophysical properties, Journal of Physical Chemistry
(1991) 95, 525.









chemical and environmental pollutants.(61) For optical applications such as photovoltaics or

LEDs, it is important that these surface traps do not contribute to the exciton trapping within the

material. Several groups have worked on developing passivation techniques that will enable

enhancement of their photophysical characteristics without altering their confinement behavior.

Passivation

Modiaication of semiconductor nanocrystal surfaces plays an important role in their

electronic and optical properties and has been the subj ect of extensive investigations. (34, 40, 62-

65) The dangling bonds present on the surface of the nanocrystals negatively influence the

optical properties but passivation has been proven to improve various confinement properties

such as high quantum efficiency and luminescence stability. Due to a high surface-to-volume

ratio, even pristine, bare CdSe quantum dots tend to result in low luminescent yields (0.6%) and

poor stability.(66) The ratio leads to augmentation of the electron and surface state

wavefunction overlap which creates localized midgap surface state traps resulting in non-

radiative decay and decreasing the overall photoluminescence quantum yield (Figure 2-4).(67)

However, in certain applications (68), in which the charge carrier-interface interaction is crucial,

the high surface-to-volume ratio is beneficial. (14, 15, 69)

If the surface of colloidal nanoparticles is coated with an appropriate passivating agent,

e.g., organic molecules, this competition may be sufficiently reduced to dramatically extend the

band-edge lifetime and enhance the luminescence efficiency. (37, 53, 64, 70) However, due to

several drawbacks including imperfect surface passivation and exchange reactions causing

photodegradation, organic coating is not sufficient for improving quantum yields.(16, 71)

Using the diffusion-controlled colloidal growth method developed by Bawendi and co-

workers, (37) CdSe quantum dots have been passivated with various shells, among these is

ZnS,(39) which narrows the fluorescence emission and improves their efficiency. Epitaxial









the diffusion of energy in molecules. In this dissertation only an overview of the radiative and

non-radiative (interchain) energy transfer and random walk diffusion (intrachain) are discussed.

Radiative Energy Transfer

A two step process that involves emission of a photon from the excited state of the donor

molecule followed by the same photon being absorbed by the ground state of the acceptor is

called radiative energy transfer.

Step 1: D* D + hu

Step 2: hu + A -A*

This type of energy transfer is the least complicated since it does not involve the interaction of

the donor and acceptor molecules. For this mechanism to be effective, the quantum yield of the

donor must be high in the spectral region of the absorption of the acceptor. To further enhance

radiative energy transfer, it is beneficial to have a high concentration and extinction coefficient

of the acceptor in addition to a large spectral overlap between the emission of the excited donor

and ground state absorption of the acceptor. The emission spectra of a donor molecule that

undergoes radiative transfer will experience a decrease in its fluorescence intensity in the

spectrally overlapped region and can lead to repeated absorption and emission if the donor and

acceptor molecules are identical (self-absorption/reabsorption). If there is adequate absorption

and emission overlap, the fluorescence lifetimes can increase.(1, 2) An example of this process is

shown in Figure 4-8, where we observe self-absorption in a conjugated polyelectrolyte solution

that is highly concentrated.

Non-radiative Energy Transfer

Non-radiative energy transfer occurs in a single step and just as radiative energy transfer,

depends on the spectral overlap between the donor' s emission and acceptor' s absorption spectra

but relies more on their coupled resonances (Figure 1-2). As seen in the Jablonski diagram in









wavelength increases to 450 nm (red line). The isolated chain time constant of 531 ps (Table 4-2)

is extracted from the mono-exponential decay of the 8 PRU at intermediate wavelengths 450 (red

line). As the detection wavelength is increased from 450 to 550 nm (not shown) the behavior of

the exponential decay does not change. At all detection wavelengths the rise times are

comparable to our instrument response function. The first panel of Figure 4-16 shows the same

detection wavelengths on a shorter time scale. An extremely fast decay (< 1.5 ps) is observed

Table 4-2. Detection dependence decay times

PRU Det h (nm) zl1 (o ps) Amplitude zz2AG tpS) Amplitude
8 PRU 430 2713 36% 490+10 64%
436 1415 29 624119 71
450 531+7 100

35 PRU 430 1111 65% 17816 35%
450 3713 39 363110 59
500 3315 39 402117 61
550 3512 53 454116 47

108 PRU
Det h (nm)
430 1412 62% 20119 38%
450 4314 39 395111 63
500 42 (fixed) 39 333125 62
550 3916 55 551149 45

35 w 50% Ca
Det h (nm)
430 1812 61% 450117 40%
450 4315 29 468111 71
550 2614 53 611145 47

when detected at 430 nm (blue line) and its contribution to the overall signal diminishes as the

wavelength increases from 430 to 436 (magenta line) to 450 nm (red line). Emission at

wavelengths below 450 nm spectrally overlaps with the aggregate species absorption resulting in










"An emission wavelength is chosen and only emission light at that wavelength is
allowed to reach the detector. Excitation is induced at various excitation
wavelengths and the intensity of the emitted fluorescence is measured as a function
of wavelength. The result is a graph or curve which depicts the relative fluorescence
intensity produced by excitation over the spectrum of excitation wavelengths."
(195)

Excitation experiments at different detection wavelengths can be employed to identify the

species contributing to the emission which can be hidden due to inhomogeneous broadening

caused by the polydispersity present within polymeric samples. The absorption spectra lead one
1.2 .~~~.


1.0


0.8-


S0.6





0.0 I I III



400 450 500 550 600 650 700
h(nm)
Figure 4-11. Emission of 10 CLM 35 PRU PPE-CO2- in methanol (-), methanol with ~ 6 CLM
Ca2+ (-) and in water ()

to believe that no aggregates are present in the shorter polymer samples since neither a shoulder

or broadening are observed; however, excitation spectra indicate that this is not the case. Figure

4-12 presents the excitation spectra of the 8 PRU CPE in methanol detected a four different

wavelengths (430, 475, 510 and 590 nm). Detection at 430 (blue line) and 475 nm (green line)

show broad, featureless excitation spectra peaked at 396 nm. Upon shifting the detection

wavelength to 510 nm the excitation spectra becomes even broader and a small red-edge shift










For longer PPE-CO2- PRU chains dissolved in methanol, the distinction between isolated

and aggregate emission bands based on the excitation spectra becomes clearer. For example,

Figure 4-13 shows the excitation spectra of 35 PRU PPE-CO2- in methanol. Detection at 430 nm

(blue line) results in a distinct peak at 390 nm. As the detection wavelength increases, the peak

broadens and shifts to the red. Detection at 590 nm (red line) clearly shows a new peak at 430

nm and this peak is attributed to the direct excitation of aggregated species. Figure 4-14 presents

the excitation spectra of the 3 5 PRU CPE dissolved in water detected a four different wave-

lengths (430, 475, 510 and 590 nm). Detection at 430 nm (blue line) show broad, featureless

excitation spectra peaked at 390 nm due to isolated chains. As the detection wavelength

increases, the peak broadens and shifts to the red. Upon shifting the detection wavelength to 510

nm (black line) the excitation spectra becomes even broader and a new peak appears (436 nm).

Detection at 510 (black line) and 590 nm (red line) exhibit more well-defined peaks at 436 nm,

comparable to the 185 PRU absorption spectrum in neat methanol.

1.0




0.8-





0.4-


0.2-


0.0 llIII
325 350 375 400 425 450 475 500
h (n m)

Figure 4-13. Excitation spectra 10 pLM 35 PRU PPE-CO2~ in methanol. Detection at 430 (-),
475 ( ), 510 (-), and 590 (-) nm









times per scan. When twenty scans were completed, the total number of laser shots per point was

equal to 5000. Kinetic traces at particular wavelengths can be extracted from the full spectrum

collected using the CCD camera.

from OPA4, delay stage




Pumpr Mnochromator

Chopper ,
From Reference
I ""

White Probe
Lig ht
Generation P,
SREF


Figure 3-2. Transient absorption schematic

Results

Insight can be gained by investigating the spectral evolution caused by population density

changes in different energy levels using broad band femtosecond time-resolved absorption.(17,

53) A continuum probe results in a collection of the entire transient spectra at each delay time in

a single experiment. Therefore, photo-excited species can be detected and in principle, identified

based on their characteristic transient absorption features. In even a simple system, assignment

and interpretation of such photo-excited species can be difficult due to convoluted absorption

features within the transient spectrum.

CdSe versus CdSe/ZnSe Core/Shell

CdSe nanorods were synthesized using the method described by Peng.(46) Raman

spectroscopy is a useful tool for evaluating the structure and compositional homogeneity of









includes Sophie, Merve, Roxy, Richard, Rob, Neil, Eric, Megan, Meg etc... Their laughter and

craziness will be greatly missed, especially during the fall at tailgating. Thanks to M.I.A and

Whoever Shows Up, the two best intramural softball teams in University of Florida history for

making me feel a little bit younger. I have enjoyed playing for five years and will miss the

teammates that have helped form our dynasty.

Finally, I thank my parents Craig Hardison and Susan Keller for allowing me to make my

own decisions so that I could become the independent woman I am today. They have always

believed that I could do anything I put my mind to. I thank my sister Brynn, for terrorizing me as

a child but growing up to become a wonderful young woman that I can call my friend.










1.0-

0.8-




0.4-

0.2-

0.0-


325 350 375 400 425 450 475 500

h (nm)

Figure 5-1. Absorption spectra of the hyperbranched PPE-CO2- in MeOH (-) and water (-)
and linear PPE-CO2- 8 PRU in MeOH ( )

If a complete understanding is necessary for these linear and hyperbranched polymers, it is

necessary to carry out similar methods of experimentation conducted in this dissertation. First, it

is imperative to make sure that the excitation spectra either does or does not indicate the presence

of aggregates. Second, a time-resolved detection wavelength dependence would be useful to

elucidate the dynamics of the system under various conditions. This data can then be compared

to the linear polymers. Once a comparison is made, further experiments can be designed to see

which polymer would be better for quenching and how calcium will affect the quenching

efficiency. Understanding these conjugated polyelectrolytes will help with designing new

materials for multiple applications including solar cells, LEDs and even chemo- and biosensors.

The hyperbranched CPE avenue is a new and exciting field and I recommend more time-resolved

experimentation be accomplished in this area.










(159). B. J. Schwartz, Conjugated polymers as molecular materials : How chain
conformation and film morphology influence energy transfer and interchain
interactions, Annual Review of Physical Chentistry (2003) 54, 141.

(160). R. Jakubiak, C. J. Collison, W. C. Wan, L. J. Rothberg, B. R. Hsieh, Aggregation
quenching of luminescence in electroluminescent conjugated polymers, Journal of
Physical Chentistry A (1999) 103, 2394.

(161). S. A. Jenekhe, J. A. Osaheni, Excimers and exciplexes of conjugated polymers,
Science (1994) 265, 765.

(162). I. D. W. Samuel, G. Rumbles, C. J. Collison, Efficient interchain photoluminescence
in a high-electron-affinity conjugated polymer, Physical Review B (1995) 52, 11573.

(163). C. J. Collison, L. J. Rothberg, V. Treemaneekarn, Y. Li, Conformational effects on
the photophysics of conjugated polymers: A two species model for MEH-PPV
spectroscopy and dynamics, Macrontolecules (2001) 34, 2346.

(164). G. H. Gelinck, J. M. Warman, E. G. J. Staring, Polaron pair formation, migration, and
decay on photoexcited poly(phenylenevinylene) chains, Journal of Physical
Chentistry (1996) 100, 5485.

(165). J. W. Blatchford, S.W. Jessen, L.B. Lin, T.L, Gustafson, D.K. Fu, H.L. Wang, T.M.
Swager, A.G. MacDiarmid, A.J. Epstein, Photoluminescence in pyridine-based
polymers: Role of aggregates, Physical Review B (1996) 54, 9180.

(166). U. Lemmer, S. Heun, R.F. Mahrt, U. Scherf M. Hopmeier, U. Siegner, E.O. Gobel, K.
Mullen, H. Bassler, Aggregate Fluorescence in Conjugated Polymers, Chemical
Physics Letters (1995) 240, 373.

(167). X. Y. Zhao, M.R. Pinto, L.M. Hardison, J. Mwuara, J. Muller, H. Jiang, D.Witker,
V.D. Kleiman, J.R. Reynolds, K.S. Schanze,Variable band gap poly(arylene
ethynylene) conjugated polyelectrolytes, Macronsolecules (2006) 39, 6355.

(168). C. Y. Tan, E. Atas, J.G. Muller, M.R. Pinto, V.D. Kleiman, K.S. Schanze Amplified
quenching of a conjugated polyelectrolyte by cyanine dyes, Journal of the American
Chemical Society (2004) 126, 13685.

(169). B. S. Harrison, M. B. Ramey, J. R. Reynolds, K. S. Schanze, Amplified fluorescence
quenching in a poly(p-phenylene)-based cationic polyelectrolyte, Journal of the
American Chemical Society (2000) 122, 8561.

(170). G. H. Gelinck, E.G.J. Staring, D.H. Hwang, G.C.W. Spencer, A.B. Holmes, J.M.
Warman, The effect of broken conjugation and aggregation on photo-induced charge
separation on polyphenylenevinylene chains, Synthetic Metals (1997) 84, 595.

(171). C. H. Fan, S.Wang, J.W. Hong, G.C. Bazan, K.W. Plaxco, A.J. Heeger, Beyond
superquenching: Hyper-efficient energy transfer from conjugated polymers to gold









of interchain species.(159) In interchain interactions, xn-electron density is delocalized among

numerous conjugated segments in different polymer chains. Depending on the physical

conformation of the chains, it is possible that the two interacting species be located on the same

chain. For example, if the polymer chains are extremely long, the conjugation segments from the

same chain can interact spatially as a result of xn-x stacking due to backfolding. Shared xn-

electrons between two polymer chromophores in their excited state that are next to each other

create a species named termed "excimer".(159-162) When neutral excitons are shared by two or

more adjacent chromophores in the ground and excited state, the intrachain species that is

formed is known as an "aggregate". The aggregate formations that interact electronically will

cause a significant change in the absorption spectra corresponding to an elongation of the xn-

electron delocalization resulting in lower energy peaks compared to isolated chains.(159) In

addition to aggregates, a "polaron pair" can be created after excitation resulting in an radical

cation (hole polaron) in one chromophore and a radical anion (electron polaron) on another.

(159, 163, 164) A significant redshift in the emission spectra is a photophysical indicator that an

excited interchain species is present within the conjugated sample due to delocalization of xn-

electrons creating a lower electronic state compared to the isolated chains. Since this

phenomenon occurs for each of the interchain interactions, it is hard to distinguish between the

various types. Detection and identification is further complicated for room temperature

fluorescence measurements due to the large numbers of non-radiative "trap" sites in conjugated

polymers resulting in very low emission quantum yields. (159, 160, 163) Aggregates can be

differentiated from the other species because they are the only ones that show a weak redshift in

the ground state detectable in the absorption spectrum. (159, 165, 166) This shift can be subtle,

especially if the aggregate absorption is symmetry of the transition is forbidden; therefore, the









subsequent Ti:Sa amplifier (Spectra-Physics, Spitfire) with a repetition rate of 1 k
produce excitation pulses. More specifically, the output of the Ti:Sa amplifier feeds an OPA, and

the fourth harmonic of the signal is tuned to 375 nm. The excitation beams is fed through a prism

compressor, yielding an instrument response function of 225 fs. The instrument response

function (IRF) is determined by the cross-correlation of the excitation and gate pulses.

The upconversion setup used for these experiments is described in detail elsewhere.(182,

183) Briefly, a fraction of the 800 nm Ti:Sa amplifier that is leftover from the OPA is used as a

time delayed gate pulse (30 CLJ/pulse). After excitation, the sample fluorescence is collected using

a pair of off-axis parabolic mirrors and focused and spatially overlapped with the gate pulse in a

nonlinear crystal (0.5 mm BBO), resulting in the sum frequency of the two electromagnetic

fields (Figure 4-4 A).

The up-conversion signal has a photon frequency given by:


Usuns =gate + luo .U + (4-3)

This is also written as:

111

sunt gate fluo(4)

Detection wavelength is chosen by tuning the nonlinear crystal to a particular angle. Table

4-1 includes a list of each detection wavelength and crystal angles used for the experiments

discussed in this chapter. The resultant signal is then focused into a monochromator, detected

with a photomultiplier and the signal is gated with an integrating boxcar. When the gate pulse is

temporally and spatially overlapped with the fluorescence signal, the nonlinear crystal behaves

as an optical gate. Therefore, scanning the gate pulse with respect to the excitation pulses enables










(27). N. Chestnoy, R. Hull, L. E. Brus, Higher excited electronic states in clusters of ZnSe,
CdSe, and ZnS spin-orbit, vibronic, and relaxation phenomena Journal of Chemical
Physics (1986) 85, 2237.

(28). Y. Kayanuma, Quantum-size effects of interacting electrons and holes in
semiconductor microcrystals with spherical shape, Physical Review B (1988) 38,
9797.

(29). M. Nirmal, L. Brus, Luminescence photophysics in semiconductor nanocrystals,
Accounts of Chemical Research ( 1999) 32, 407.

(30). A. L. Efros, F. G. Pikus, V. G. Burnett, Density of states of a 2-dimensional electron-
gas in a long-range random potential, Physical Review B (1993) 47, 2233.

(31). A. J. Nozik, Spectroscopy and hot electron relaxation dynamics in semiconductor
quantum wells and quantum dots, Annual Review of Physical Chentistry (2001) 52,
193.

(32). X. G. Peng, L. Manna, W.D. Yang, J. Wickam, A.Kadavanich, A.P. Alivisatos, Shape
control of CdSe nanocrystals, Nature (2000) 404, 59.

(33). L. Manna, E. C. Scher, A. P. Alivisatos, Synthesis of soluble and processable rod-,
arrow-, teardrop-, and tetrapod-shaped CdSe nanocrystals, Journal of the American
Chemical Society (2000) 122, 12700.

(34). L. Manna, E. C. Scher, L. S. Li, A. P. Alivisatos, Epitaxial growth and photochemical
annealing of graded CdS/ZnS shells on colloidal CdSe nanorods, Journal of the
American Chemical Society (2002) 124, 7136.

(35). C. D. Donega, S. G. Hickey, S. F. Wuister, D. Vanmaekelbergh, A. Meijerink,
Single-step synthesis to control the photoluminescence quantum yield and size
dispersion of CdSe nanocrystals, Journal of Physical Chentistry B (2003) 107, 489.

(36). L. A. Swafford, Homogeneously Alloyed CdSxSel-x Nanocrystals: Synthesis,
Characterization, and Composition/Size-Dependent Band Gap, Journal of the
American Chemical Society (2006).

(37). C. B. Murray, D. J. Norris, M. G. Bawendi, Synthesis and Characterization of Nearly
Monodisperse Cde (E = S, Se, Te) Semiconductor Nanocrystallites, Journal of the
American Chemical Society (1993) 115, 8706.

(38). T. Mokari, U. Banin, Synthesis and properties of CdSe/ZnS core/shell nanorods,
Chentistry of2aterials (2003) 15, 3955.

(39). B. O. Dabbousi, J. RodriguezViejo, F.V. Mikulec, J.R. Heine, H. Mattoussi, R. Ober,
K.F. Jensen, M.G. Bawendi, (CdSe)ZnS core-shell quantum dots: Synthesis and
characterization of a size series of highly luminescent nanocrystallites, Journal of
Physical Chentistry B (1997) 101, 9463.









CHAPTER 4
CONJUGATED POLYELECTROLYTES (CPES)

Introduction

xn-Conjugated polymers are an interesting class of materials with unique physical

characteristics that make them excellent candidates for various purposes including lasers (146),

LEDs (147), photovolatics (148), and transistors (149). To be useful for any application, a

fundamental understanding of their photophysical properties is necessary in order to continue to

improve their efficiency and efficacy. In recent years, xn conjugated polyelectrolytes (CPEs)

have been synthesized incorporating ionic solubilizing side groups enabling the polymer to be

dissolved in water and other polar solvents while preserving the photophysical properties

associated with the polymer backbone.(6, 7) In an effort to reduce exposure of the non-ionic

components to the environment, when CPEs are dissolved in a polar solvent such as water they

self-assemble into aggregates due to the interaction between the charged functional groups and

hydrophobic backbone. (150-154) This intra- or intermolecular xn-x stacking of the polymer chain

creates new, red shifted absorption and emission peaks, (155) decreases the overall fluorescence

quantum yield and competes with radiative emission processes from the isolated chains.(156)

Aggregates can also form in concentrated polymer solutions with nonpolar solvents.(14, 20, 26-

28) Addition of a metal cation such as Ca2+ acts as a cross-linking agent and it has been shown to

induce aggregation in methanol, improving the amplified quenching properties of CPEs.(155,

157-159)

Aggregation is easily confused with other types of interchain interactions of xn-electrons in

spatially close chromophores. It is extremely important to correctly define, understand and

identify the different types of interchain species that can be present within the conjugated

polymer solution or film. Within the literature, there is a discussion on the proper identification









The electronic and optical properties of a material are due to electron motion within

molecular orbitals. The energy absorbed by an unbound electron (not confined) within the

density of states in a bulk material is not quantized. Therefore, the energy released by this

electron is converted into kinetic energy. A semiconductor can be photoexcited with a photon,

exciting an electron from the valence band into the conduction band of the material, leaving a

hole of opposite electric charge behind, separated by distance consisting of several atoms within

the material. These distances are within the nanometer scale and are called the Bohr exciton

radii. This radius, combined with a high dielectric constant results in a small binding energy. The

electron can be bound to the hole due to Coulombic forces and if these interactions are strong

enough, a Coulomb correlated, bound 'quasi-particle' called an exciton (electron-hole pair) is

formed. If the size of the electron-hole pair is approximately the same as the Bohr radius, and it

is larger than the lattice spacing within the crystal, a Wannier-type exciton is formed. This

exciton can diffuse through the material until it is trapped, annihilated (under multi-excitation

conditions) or recombined. If the wavefunctions of the electron and hole extend over a large

number of atoms, the Coulombic attraction becomes negligible resulting in unbound charge

carriers which have slightly higher energies than the bound electron-hole pair. (13, 21, 22)

If the size of bulk semiconductor is significantly decreased to the point where is it similar

to the size of the Bohr exciton radius, then the motion of the exciton will become confined in

multiple dimensions (quantum confinement) since it will have less "room" to move. The energy

spacing between the various confined (bound) electron and hole states within the corresponding

bands becomes quantized and the separation between these energy states will increase as the size

of the particle (space) decreases due to stronger confinement.(13, 17, 21, 22) In addition, the

energy separation of the electron states is larger than the separation of the hole states since the









formation of ZnCdSe alloys is not perfectly uniform and contributes to broad, inhomogeneous

photoluminescence spectra and a distribution of decay rates.

Figure 2-14 shows the broad band luminescence spectra of the core/shell and alloyed

nanorod samples as a function of time. This data was collected on a nanosecond time scale using

a 4 nanosecond instrument response and 0.4 ns time step. The fast decays (< 4 ns) are not

detectable due to this limitation. From this data we were able to extract time traces at the

maximum wavelength to determine the corresponding decay rates of each sample. Figure 2-15 A

shows the broad band luminescence of the core/shell sample at two different time steps, 4.4 ns

(black line) and 20 ns (red line). As the signal decays at the maximum wavelengths, the

photoluminescence values do not shift significantly but the broadening is slightly reduced.

Broadening also occurs since measurements of the nanorods samples were carried out at room

temperature. The time traces corresponding to three different wavelengths, 645, 670 and 630 nm

(black, red and green lines, respectively) for the core/shell sample are shown Figure 2-15 B. At

early decay times, the emission at different wavelengths is not the same but ends up being

identical after ~ 100 ns. Moving from the core/shell (A) to the alloy 3 hr (D) the band gap shifts

to higher energies and the broad band signal narrows.

Figure 2-16 shows the log plot of the time-resolved photoluminescence decay curves for

the nanorods samples. These decays are the normalized kinetic traces at the wavelengths

corresponding to the maximums of the photoluminescence. Several differences can be

highlighted. The signal at decay times less than four nanoseconds have to be deconvoluted with

the instrument response function. Since we are interested in extracting a characteristic lifetime

we do not consider this for analysis. It is seen from this plot that the core/shell decay (black line)

is much faster than the alloys. Also, the core/shell decay curve is almost a straight line,











efficient energy transfer from isolated chains to aggregates. Interestingly, detection at 590 nm


does not yield a significant change in the rise time (not shown) or decay compared to 450 nm


which had been expected if detecting emission from the aggregates. Dynamics observed at


wavelengths between 430 and 450 nm exhibit an additional intermediate decay time of 30 to 40


ps (middle panel). The amplitude of these time constants decreases as the wavelength increases.

From the excitation spectra it is clear that there is a small, yet significant amount of aggregate


within the sample below 450 nm facilitating energy transfer; hence, the intermediate time


constant is assigned to the energy transfer from the ensemble isolated to the aggregated chains.

1.6 a ,, ~ ~ ..,

1.4 -- -- -

1.2 -- -- -

1.0 -- -

E 0.8-- -

0.6 -- -

0.4 -- -

0.2 -- -- -

0.0 a - -
0 2 4 0 50 100 150 20t) 100 200 300 400 500 600

time (ps)


Figure 4-16. Time-resolved fluorescence decay of 8 PRU PPE-CO2- in methanol at three
different detection wavelengths. 430 (-), 436 (-), and 450 (-) nm, Fits (-)

The results for solutions of 35 and 108 PRU PPE-CO2- (30 CLM) are shown in Figure 4-17


for three distinct wavelengths (430, 450 and 550 nm) upon excitation at 375 nm. Detecting at


430 nm, the contribution from the very fast component is larger for the 35 PRU sample (black


line) compared to the 108 PRU sample (red line). As the detection wavelength is shifted to the


lower energies, contribution from this fast component to the overall signal is reduced.


Interestingly, the changes are more pronounced on the 35 PRU than on the 108 PRU sample.

















































II I I I I


I I


indicating a higher degree of homogeneity and a smaller distribution of decay rates compared to

the alloys. The alloys (green, red, and blue lines) do not show the same behavior as the

core/shell; instead, they exhibit two or more decay components (similar dynamics prior to 50 ns

but deviate after that). The data collection for alloy 3 hr (blue line) is shorter due to the lack of

detectable signal after 60 ns.


640

620

E 600

580


15 20


A)


,' 640


620


0 5 10 15 20 25 0 5 10
time (ns) time (ns)


C)


D)


0 5 10 15 20 25 0 5 10 15
time (ns) time (ns)

Figure 2-14. Broad band Photoluminesce of A): CdSe/ZnSe core/shell nanorods B) 1 hr ZnCdSe
C) 2hr ZnCdSe D) 3hr ZnCdSe

In dispersed systems such as polymers (116) or colloidal nanoparticles, it is easy to

believe that the relaxation behaves non-exponentially and that the large distributions of local

environments lead to variations of relaxation times. Multi-exponential functions are useful and

are more commonly utilized for fitting decay curves; however, a model is generally proposed


E










spectrum to lower energies.(72, 78) This is an important factor in the analysis of our CdSe/ZnSe

core/shell nanorods which will be discussed in Chapter 3.

organic
molecule









offset *


band ~~t~~as~ ~ hN
offset**



A) B)

Figure 2-5. 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. Adapted from H. Lee.(66)

Atom dislocations induced by interfacial strain as a result of the lattice mismatch between

the core and shell can also have a negative effect on the luminescence quantum yield because

they can behave as sites that cause non-radiative recombination. The defects that arise from the

core/shell interface can be reduced resulting in higher quantum yields by either growing a

nanocrystal that is comprised of one core and two shells, such as a CdSe core passivated with a

CdS/ZnS shell/shell structure (34, 66, 79) or by "photoannealing"(34, 66) which reduce strain or

diffuse defects to the surface, respectively. For example, irradiating CdSe/ZnS core/shell

nanocrystals with UV light caused an increase in the photoluminescence quantum yield by

reducing the number of vacancies present at the interface.(34, 66)











are IRF-limited at all wavelengths. The augmented amount of aggregation increases the


amplitude of the energy transfer components. Jiang determined that Ca2+ induces the formation


of "loose" aggregates in methanol and more traps along the polymer backbone.(155) Our results


show that the addition of Ca2+ (red line) amplifies the energy transfer from the short isolated


chains to traps. Detecting at 430 nm, the long component increases from ~180 to ~ 450 ps while


the intermediate decays times remain relatively the same (Table 4-2).

1 : : 1.50 1.50 *
4.5-

4.0 -1 1.25 -1 1.25-

3.5 --
1.001 -1 1.00-
3.0-

E 2.5 --0.75 -U -1 0.75

Z2.0
0.50C C, -1 0.50-
1.5-

10-0.25 Vs.i -1 0.25-

0.5-
0.00 -1 0.00-
0.0 ~ ~ ~ 1 1 1 1 1
O 1 2 3 4 5 0 50 100 150 200 0 100 200 300 400 500 600

time (ps)

Figure 4-18. 30 ILM PPE-CO2- 35 PRU without (-) and with (-) Ca2+ at 430 nm.

Figure 4-19 presents the effect of the addition of Ca2+ to the dilute 8 PRU sample detected

at 450 nm upon excitation at 375 nm. The black line corresponds to the isolated chain emission


and the red line corresponds to the 8 PRU with the addition of 15 pM Ca2+. At 450 nm, the


addition of Ca2+ introduces a 1.5 ps decay time and an intermediate component (~ 30 to 50 ps)


not seen in the neat 8 PRU sample. The radiative decay rates measured with and without calcium


differ due to the presence of aggregate structures in conjugated polymers. As stated previously,


the addition of the dication calcium to the PPE-CO2-/methanol mixture causes aggregation,










by 80% (I/Io). The ultrafast decay that occurs in less than 1.5 ps is present in all samples

regardless of the presence of calcium (not shown). No significant changes in the excited state

lifetime at long timescales were detected despite a reduction in the photo-luminescence quantum

yield implying that the most important step in the decay mechanism facilitating amplified

quenching occurs in the first 2 ps in PPE-CO2~ pOlymers.


1.00-



0.75--



o 0.50--



0.25-



0.00--
) 50 100 150 200 250 300 350 400 450
time (ps)
Figure 4-20. 30 CLM PPE-CO2- 35 PRU with (-) Ca2+ at 450 nm 30 CLM PPE-CO2- 35 PRU with
MV2+ (80% quenched) at 450 nm and (-) 30 CLM PPE-CO2- 35 PRU with 15 CLM
Ca2+ and MV2+ (80% quenched)

Time-Resolved Anisotropy

To investigate the fluorescence depolarization, random walk migration, or intermolecular

energy transfer we measured the anisotropy dynamics of the 8 PRU polymers in methanol.

Figure 4-21 depicts the time-resolved fluorescence anisotropy of the 8 PRU detected at 430

(black line) and 450 (red line) nm upon excitation at 375 nm. An ultrafast loss of anisotropy (not

shown) followed by a constant anisotropy during the lifetime emission of the polymer is










particular emission wavelengths, the Auger decay can be stifled in rod-shaped nanocrystals due

to not only the dependence that the confinement potential has on the length and size but also the

linear scaling of the decay time with rod volume.(52, 199, 200) For higher energy (shorter

wavelengths) quantum dots, the increased surface-to-volume ratio inhibits the Auger decay

suppression. Elongation of the nanocrystal in the c-direction has successfully increased the

optical gain lifetime since the effect that Auger has on the recombination behavior in rods is

decreased.(199, 201) CdSe/ZnSe core/shell materials have been thought to be used for this

application but it is more suitable to use inverted ZnSe/CdSe heterostructures in order to control

the electron-hole wavefunction overlap to increase the confinement energies and reduce Auger

recombination.(199, 202, 203)

Formulation of new alloy nanorods has opened the door for new investigations for their

potential applications. Extension of the c-axis in these ternary materials enables for higher

confinement potential in the blue-green region. It may be worth trying to determine how to alloy

the materials and then coat them to increase their photoluminescence quantum yield to make

them comparable to the core/shell or shell/core materials. These new alloy/shell materials can be

tunable based on the diffusion and have extended lifetimes necessary for charge-separated or

optical gain applications.

Light emitting diodes (LEDs) (11, 204-206). Although several improvements using

organic molecules for organic light emitting diodes (OLEDs) make them comparable to current

technologies, there are ongoing drawbacks and problems that must be overcome before these

devices can be commercially utilized. These include: a) difficulties tuning the colors since the

fluorescence is broad and b) synthesis of multiple molecules is required to obtain a broad range

of colors. Nanocrystals are being considered as attractive candidates to be used for LEDs since









multiple processes detected simultaneously. This effect is not observed in the core sample. For

the core/shell sample, it is proposed that the hole migrates to the valence band of the ZnSe shell

(valence band offset = 0.07 eV) due to the energy transfer during Auger relaxation of the

electron resulting in longer bleach decaying times. The insets of Figure 3-4 A and B compare the

higher energy, 1P state negative absorption decay and photoinduced absorption to the 1S and 2S

states bleach decays. It is interesting to note that the rise of the negative signal from the 2S and

1P states (black and green respectively) are identical but the 1P bleach decays rapidly (~ 1 ps)

into a positive signal, matching the rise of the bleach of the lower 1S energy state (red line). This

confirms a 1P to 1S relaxation process in both CdSe and CdSe/ZnSe.

Core/Shell Excitation Dependence

An excitation wavelength dependence study (450, 575, 610, 630, 650 nm) was conducted

to elucidate the influence on the kinetic processes of the ZnCdSe interfacial state previously

detected by Raman spectroscopy. Figure 3-5 shows the steady state absorption spectrum, with

the horizontal arrows signaling the excitation wavelength for each row of time-resolved spectra

presented on the right two columns. The transient spectrum showed on top indicates the

detection wavelength for each column. For example, the kinetic data on the top left plot presents

the transient signal in the area of the 1P state after excitation at 650 nm whereas the middle right

plot presents the transient signal of the 1S state after excitation at 575 nm. After excitation at the

same energies, the 1S band bleach rises with a time constant corresponding to the decay of the

1P bleach. In addition, the 1P detection shows photoinduced absorption present at a time delay

greater than 1 ps due to the ZnCdSe interfacial state. For excitations greater than 600 nm (lower

energies), the P band does not contribute to the dynamics; instead, the excitation simultaneously

populates both the S band and trap states.










quantum yield decreasing after 1 and further after 2 hours 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 hours increased the

quantum yield over that from samples annealed for 1 or 2 hours. This increased quantum yield

and full-width-half maximum (FWHM) reduction can be attributed to annealing of crystalline

defects and reduction of stress, consistent with the Raman data. Defects found in the crystal are

known to act as "traps", reducing emission efficiency(114, 115). (42)
6x1 06


5x106A)1
4x1 06


3x1 06 B) \


2 x1061



1x106



400 450 500 550 600 650 700 750 800
Wavelength(nm)
Figure 2-13. Photoluminescence spectra from A) CdSe/ZnSe core/shell nanorods and ZnCdSe
nanorods alloyed at 270oC for B) 1, C) 2, and D) 3 hrs. Adapted from H. Lee.(66)

Time-Resolved Photoluminescence (TRPL)

The linewidths of optical transitions can be inhomogeneously broadened due to effects that

act differently on different radiating or absorbing particles.(2) Emission from an inhomogeneous

population (different sizes, shapes or composition) leads to the simultaneous probing of particles

with different decaying rates. Monitoring time-resolved photoluminescence at different

wavelengths not only identifies the states emitting but also extracts their decay rates. The










Photoluminescence and Absorption Properties

Significant differences can be seen in the absorption spectra for CdSe core, CdSe/ZnSe

core/shell and ZnCdSe alloyed nanorods. The absorption spectra of nanorods are shown in

Figure 2-1 1. It presents two absorption peaks on top of a broad absorption. These absorption

peaks correspond to confined states although they are not as sharp or as well resolved as peaks

reported for CdSe quantum dots.(17) Due to the loss of symmetry in nanorods, confinement

along the c-axis is not as strong as it can be in dots which results in a large distribution of energy

levels in the conduction and valence bands.(53, 111) In addition, the compositional disorder

indicated from the Raman data will lead to broader features in the absorption and emission

spectra. Therefore, it is difficult to determine the exact energy spacing between the first and

second absorption peaks. (42)
0.25


0.20 :


0.15-


S0.10-


0.05-
A)
0.oo C B

300 400 500 600 700
Wavelength
Figure 2-1 1. UV-Vis absorption spectra of A) CdSe nanorods, B) CdSe/ZnSe core-shell
nanorods, and C) ZnCdSe nanorods alloyed at 2700C for 3hrs. Adapted from H.
Lee.(66)

For CdSe core and CdSe/ZnSe core/shell nanorods, the absorption edge is at ~650 nm and

~645 nm, respectively. A second peak is observed at ~520 nm. These features correlate to optical












Re sults............... .... ........ ........ .. ............. .............6
CdSe versus CdSe/ZnSe Core/Shell .....__.....___ ..........__ ............6

Core/Shell Excitation Dependence ............_....._ ....._ ............7
Core/Shell versus Alloys .............. ...............75....
Discussion ............. ...... ._ ...............78...
Summary ............. ...... ._ ...............85...


4 CONJUGATED POLYELECTROLYTES (CPES) ................ ...............86................


Introducti on ................. ...............86.................

Quenching PPE-CO2- ............. ...... ...............93...
Experimental M ethods............... ... .. .......... ...............9
Synthesis of Variable Chain Lengths of PPE-CO2- ......____ ...... .___ ................95
Photophysical M ethods ................................................9
Photophysics of Variable Chain Length PPE-CO2- Polymers ................ ............ .........105
Steady State Characterization............... ...........10
Time-Resolved Fluorescence ............_ ..... ..__ ...............112...
Isotropic Up conversion ................. ...............112.....__ ......
Time-Resolved Anisotropy ..........._..._ ...............119.....__ ......
Potential Kinetic M odel ..........._...__........ ...............121....

Summary ..........._...__........ ...............122.....


5 CONCLUSIONS AND FUTURE WORK ..........._.._........_........__ ...........12


Nanoparticle Conclusions and Future Work............... ...............124.
Conclusions .............. ...............124....
Outlook/Future Work ...................... ...............125
PPE-CO2 COnclusions and Future Work ........._..__......_ .. ...............129.
Conclusions ........... .... .. .................... ...............12
Outlook/ Future Work (Hyperbranched PPE-CO2-) ................ .......... ...............130


LIST OF REFERENCES ................. ...............133................


BIOGRAPHICAL SKETCH ................. ...............150......... ......











me 0.2 0.'13 0.16 0.28


'1.0


o.o


CBO (eV)
VBO (eV)
o~o dS OdSe ZnSe ZnS;


-1.0 -0.7 0.5-1 0.6-1.1' 0.5 my


Figure 3-9. Valence and conduction band offsets for various materials. (75)

From the observations above we propose two models associated with the exciton dynamics

within the CdSe/ZnSe core/shell and ZnCdSe alloy nanorods. In Figure 3-10, high energy

excitation results in very fast 1P to 1S relaxation times (~ 1ps). State filling within the

conduction band of the CdSe occurs.


CdSe


1P,


1S,


hy= 2.75 eV




0 .4 V I S1/1S 3 2 I
0.4~ e1P3 2


Figure 3-10. CdSe/ZnSe core/shell potential kinetic model.


ZnSe
PIA









md
ition


n n7 eV' VB Offset









quantum dots have been the focus of several photophysical studies (18-20) stimulating interest in

other types of quantum particles. Fabrication of nearly spherical particles with various

compositions in addition to synthesis of rod shaped (32) and even multifaceted tetrapods (33) has

been achieved. Manufacturing such materials can be achieved by two different methods: 1.

bottom-up and 2. top-down. The first method utilizes synthetic routes that adjust ratios of the

chemicals needed to make the nanoparticles with passivation or capping materials.(33-42) In the

latter, the bulk semiconductor is "cut down" to scale using laser ablation-condensation or

lithographic techniques although these methods are extremely expensive.(43) The materials

presented in this thesis have been prepared by a "bottom-up" approach in an attempt to

synthesize better materials while enhancing their process-ability.

Dependence on the sensitivity to size and shape is important when considering the tunable

optical properties of quantum nanoparticles. In particular, the size dependence and

photoluminescence tunability in the visible region of CdSe quantum dots has been studied

extensively.(37, 44-48) New methods for synthesis of rod-shaped CdSe nanoparticles have

opened the door for shape-dependent applications such as polarized LEDs (49, 50). In particular,

Alivisatos' group synthesized quantum confined colloidal nanoparticles with rod-like

architectures by using various surfactants that bind to different faces of the crystal.(51) For

example, colloidal CdSe rod lengths can be varied from 5 nm up to 100 nm while maintaining a

2 to 10 nm diameter, which preserves lateral confinement of carriers in the nanocrystal.

Alivisatos determined that the band gap depends mainly on the width (a or b axis) and slightly on

the length (c-axis) (Figure 2-4). (52, 53) However, a comparison of the dynamics within CdSe

dots and nanorods has proven to be useful in understanding the electronic structure differences

that occur when the c-axis is elongated.









controversy between discriminating between aggregates and excited state interchain species is

still ongoing.(159) Based on the spectral signatures present in our photophysical character-

ization, the species present in PPE-CO2~ are COnsidered to be aggregates.

Correct identification of the types of interchains species is extremely important when

considering charge transport and light emission applications of conjugated polymers. In order to

fully understand a system and be able to make synthetic improvements it is necessary to

characterize each of the species accordingly. Also, CPEs are opening the door to various

biological applications but aggregation must be considered because it is an extremely important

factor that influences the polymer quenching capabilities and ultimately their performance as

chemo- or biosensors.(130, 155, 157, 160, 161)

Zhao et al. has reported the synthesis and characterization of a series of variable band gap

poly(arylene ethynylene) (PAE) water soluble conjugated polyelectrolytes dissolved in methanol,

water and methanol/water mixtures.(167) By only varying the anionic side group, they achieved

band gap tunability within the visible region. Photophysical data collected in their study correlate

the CPE side chain structure to the extent of polymer aggregation when dissolved in each

solvent. The work done for this thesis focuses on the role aggregation plays in the intra- and

intermolecular energy transport within varying polymer repeat units (PRU) of PPE-CO2-. From

previous research in this area it has been determined that the quenching efficiency increases as

the amount of controlled aggregation increases.(130, 150, 155)

Several investigations, (158-160) including those done by Chen et al. (157) have alluded to

the idea that quenching of a conjugated polymer emission is the fundamental property necessary

to understand and characterize these materials to be useful for chemo- and bio sensors. More

specifically, anionic polymeric electrolytes can be efficiently quenched by cationic systems in









decrease the overall intensity (not as low as bare CdSe rods) and bleach lifetimes (recovery

occurs faster than in the CdSe bare rods). In addition to surface state traps, crystal defects are

known to act as non-radiative recombination centers, reducing the emission efficiency and

enhancing the bleach recovery.(66, 114, 115) The band gap shifts, band narrowing and increase

in the overall bleach amplitude can be attributed to stress relaxation by thermal annealing.(66,

110) This is consistent with the weaker, broadened 1S band bleach signal observed after 1 hr and

further after 2 hrs of annealing, since diffusion alters the distribution of Zn throughout the

nanorod, i.e. decreasing the amount of Zn present at the surface and increasing the amount to-

wards the middle. However, annealing for 3 hrs enhanced and narrowed the 1S band bleach

compared to the samples annealed for 1 or 2 hrs. This increased change is attributed to annealing

of crystalline defects and reduction of stress, consistent with the Raman data reported.(42)

Tunneling of the electron wavefunction into the ZnS shell has been reported in CdSe/ZnS

core/shell structures by Mokari and Banin(38) resulting in a ~10 nm red shift. This tunneling led

to a delocalization of the electron, lowering its confinement energy and consequently the energy

of the exciton levels.(39, 42) Raman data presented previously indicate 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 (42, 135) resulting in increased localization and a blue shift in

emission. In our experiments, addition of the ZnSe shell does not alter the band gap significantly

(only a 4 nm blue shift observed in photoluminescence). We have successfully engineered the

material to create electron and hole wave functions that experience a confinement potential that

localizes (Type-I localization) the electron wave function within the CdSe core despite addition

of a shell.(136) It has been shown that the rise of the 2S and 1P bands are identical indicating

that the hole is delocalized within the density of states located in the valence band of the CdSe









principle, the interaction between atoms "far" from one another in nanocrystals is weak which

will reduce the activation energy ultimately enhancing the diffusion in these nanomaterials.(66,

85, 86)

In our samples, we were able to achieve interdiffusion of Zn from a ZnSe shell into a CdSe

core by alloying at a temperature (270~290oC) determined to be effective by Zhong et al.(48)

Despite the fact that ZnCdSe quantum dots have been grown on ZnSe (87) and GaAs (88)

substrates to determine how radius affects the fluorescence lifetime, little has been published on

the compositional affects on the dynamics of colloidal ZnCdSe nanoparticles.

From TEM images we observe that diffusion of Zn into the CdSe core does not change the

shape of the rods significantly and the single phonon mode observed by Raman backscattering

indicate that the ZnCdSe materials are complete quantum rod alloys, not composites.

The differences that arise in both steady state absorption and photoluminescence in

addition to time-resolved photoluminescence measurements make investigating these ternary

systems using ultrafast techniques extremely appealing (Chapter 3). In the current chapter, the

synthesis, characterization, steady state photophysics and time-resolved photoluminescence

measurements are described to begin to explain the carrier relaxation within CdSe, CdSe/ZnSe

core/shell and ZnCdSe alloy quantum rods.

Experimental Methods: Nanorod Synthesis and Composition Characterization

The synthesis, XRD, TEM, Raman, and some optical characterization of each of the

materials studied in this dissertation were carried out by Dr. Hyeokjin Lee in the Department of

Materials Science at the University of Florida.

Preparation of ZnCdSe Nanorods

CdSe nanorods were synthesized using the method described by Peng. (46) In this method,

CdO, trioctylphosphine oxide (TOPO) and tetradecylphosphonic acid (TDPA) were heated in a









axis. First, the tilt angle, OR, iS calculated then used to calculate the Retardation Indicator position

(I). From the Berek compensator manual we can derive the relationship between R and the tilt

angle. A summary is provided in this dissertation.

Consider a uniaxial crystal with an optical axis parallel to the plate surface. A normal

incident beam experiences a retardation (R) that is dependent on the path length (d), wavelength

(h) and the ordinary and extraordinary indices of refraction:(186)


R = (n n,) (4-8)


However, if the plate is tilted, the retardation equation becomes:(186)


R = (ne cos8, no cos 0) (4-9)


The tilt-induced extraordinary index of refraction from Figure 4-7 is determined by:(184, 185)

1 cosZ B sin" OR
-+ (4-10)
ne n; n;

The relationship between the optical axis of the medium, tilt angle, angle of incidence and

indices of refraction are used to derive the following equation for the retardance:(184)


2000 1 1- n sin O
R g s O 1 (4-11)


To use the Berek compensator as a half wave plate (1/2), R is fixed to 0.5. The tilt angle and

Retardation Indicator equations are purely empirical and are based on the crystal dimensions

only known by New Focus@ (187) and the dispersion relations for the indices of refraction

determined by Dodge (188) included in the Berek compensator manual. The indicator versus

wavelength graph corresponding to quarter and half wave retardance is included in the manual.

(184) The tilt angle is estimated using the following empirical equation:(184)









Photophysics of Variable Chain Length PPE-CO2~ POlymers

Steady State Characterization

The steady state photophysics pertaining to variable chain lengths of water soluble PPE-

CO2~ pOlyelectrolytes has been previously reported.(181) Inhomogeneous broadening (Figure 4-

9) is exhibited in the absorption spectra due to a distribution of excitation energies resulting from

slight variations and superpositions of absorptions of various segments with different

conjugation lengths. In addition, as the length of the polymer chain increases, the isolated xn -

x*" peak shifts towards the red possibly due to an extension of the conjugation length.(190) The

absorption maximums for the 8 PRU and 185 PRU are 404 and 432 nm respectively. Moreover,

the shoulder at 432 nm, which is assigned to aggregated species, becomes the absorption

maximum for the 108 and 185 PRU polymers. Structured vibronic features in the emission

spectrum are shown in Figure 4-10. The high energy emission corresponds to isolated chain

emission while the broader, low energy emission arises from the aggregate states (appearing as a

shoulder). The emission does not display the similar red shift seen in the absorption, instead the

fluorescence peak shifts are extremely small and decrease as the chain length is extended. The

SO S1 (0-0) transition corresponding to the 35 PRU is Stoke's shifted with respect to the 404

nm absorption by 20 nm. Meanwhile, the 185 PRU displays a Stoke's shift of only 4 nm between

the blue end of the emission (436 nm) and the red edge of the absorption (432 nm). The 185 PRU

sample undergoes self-absorption at 436 nm. A similar behavior has been observed in poly-

(para)-phenylene-ladder-type (LPPP) (116) in which the bridging present within the polymer

prevents the phenyl rings to twist, maintaining conjugation. The authors claim that the small

Stoke' s shift reflects the rigid geometry of the conjugated main chain resulting in reabsorption of

the So S1 (0-0) transition.(181) PPE-CO2~ pOlymers are geometrically rigid resulting in

comparable Stoke's shifts to the LPPP polymer. As the chain length increases from 35 to 185









2.47 ps (blue line), carriers quickly relax from the higher energy states to the band gap state

resulting in a corresponding 1P to 1S relaxation.


Al






-0.01-
A) CdSe Rods

0.00- All





-0.02-

B) CdSa/ZnSe CS Rods

500 550 600 650 700



Figure 3-3. Broad band transient absorption spectra for A) CdSe Rods and B) CdSe/ZnSe
Core/Shell Rods at various pump delay times: 0 (-), .400 (-), .800 ( ), 2.47
(-), 200 ( ), 575 (-) ps.

Extracting the kinetic information from the broad band spectrum enables the comparison of

carrier relaxation trends at each optical transition (Figure 3-4). In both systems (CdSe and

CdSe/ZnSe), the decay lifetimes corresponding to the 1S and 2S bands are identical (>200 ps)

(black and red lines). The 1P decays rapidly and results in photoinduced absorption at longer

delay times, the origins of which will be explained in further detail in the discussion. The rise of

the PIA signal in bare CdSe rods is slower (>10 ps) than the passivated rods (< 2 ps). A "dip" is

observed in the passivated sample for both the 1S and 2S bands, which results from overlap of









CHAPTER 3
QUANTUM PARTICLE ELECTRONIC STRUCTURE

Introduction

Investigations into the behaviors of the electrons and holes in quantum nanoparticles have

been of great interest for several years.(10, 18, 31, 120-123) As seen in the previous chapter,

steady state spectroscopy does not give a clear understanding of the exciton behavior when

comparing rods to dots. On the other hand, there are methods that can measure the temporal

dynamics and the kinetics of photophysical processes. These methods are called time-resolved

spectroscopy techniques,(124) and they are a powerful tool that can bridge fundamental

parameters such as size, shape, composition and passivation to increasing quantum yields and

stability, reducing photodegradation, lowering the cost of fabrication (deposition and lithography

methods are expensive), making the synthesis safer, and improving their process-ability.

Whether in a conjugated molecule or semiconductor nanoparticle, upon excitation, an

electron and hole are created. Figure 3-1 shows a cartoon of molecular orbitals (MO) as linear

combinations of atomic orbitals. In conjugated molecules, the MOs near the "band gap" are

linear combinations of the same type of atomic orbitals whereas in nanocrystals the MOs can be

linear combinations of atomic orbitals from different atoms. When an electron is excited from

the HOMO (valence band) to the LUMO (conduction band), a hole is left behind in the HOMO.

In a CdSe nanocrystal, the HOMO has contributions from atomic orbitals from the Se2- whereas

the LUMO is a linear combination of atomic orbitals from Cd2+. Therefore, the created hole will

be located within the anion MOs while the electron will occupy the cation MOs. This in addition

to a high dielectric constant present in semiconductor nanocrystals means that the electron and

hole are correlated although at the same time the individual carriers can behave, i.e. be excited,

trapped or relax nonradiatively, to some extent, independently.(13)









their emission is not only tunable but considerably narrower than that from organic materials.

Moreover, nanocrystals have a higher probability of resisting photodegradation.(66)

Hybrid OLEDs have been developed in the past fifteen years, incorporating a polymer such

as PEDOT (207) or PPV (204) to transport charge to various nanocrystals (CdSe (204),

CdSe/CdS (11), CdSe/ZnS (208)) that act as the emission layer resulting in more stable and

efficient devices. Development of better materials and manipulating their interactions are the

main goals when working towards designing products that result in high electroluminescence

efficiencies. To achieve commercial quality devices the functionality must be improved by

enhancing the charge transfer between the polymers to the nanocrystal emission layer and

increasing the surface quality so the "traps" which cause non-radiative recombination are

reduced.(66) Experimentation with different combinations of polymer/nanoparticle blends is a

standard methodology to find devices that get rid of such adverse consequences. Within the

literature, most nanoparticles are spheres and their band gaps are tuned by only changing their

diameter.(20, 39, 48, 75, 78, 136, 138, 209-212) As the diameters are decreased, the band gap

energy does increase and emission in the blue-green region is achieved, however, the surface-to-

volume ratio is significantly increased which can lead to more surface "traps". Therefore, some

investigations into hybrid LEDs should incorporate not only size distributions to obtain tunability

but to investigate how the shape, passivation thickness and composition will affect the overall

efficiencies of the devices. A wide range of colors in the blue-green region can be achieved

simply by altering the alloying times in ZnCdSe nanorods. If a simple technique was developed

to passivate these alloy rods to reduce surface traps, they would allow for blue-green emission

wavelengths via a simple synthetic route (one batch).









solution. The quenching efficiency is described by the conventional "Stern-Volmer"

relationship:(2)

rI
o __-1+ k ro[Q]= 1+ Ks [Q] (4-1)
OI

where 40 (lo) and O(1 are the steady state fluorescence quantum yields (fluorescence

intensities) in the absence and presence of the quencher molecule respectively, Ksy is the Stern-

Volmer constant, and [Q] is the quencher concentration. A "Stern-Volmer" plot is the

fluorescence intensity ratio (l0/1) versus Q. This plot is expected to be linear with the slope equal

to the Stern-Volmer constant. From this information, the quenching rate constant, k,, can be

calculated if the excited state lifetime, to of the neat sample is known.

From a time-resolved measurements point of view, the quenching mechanism is "static" if

there is no change in z when a quencher is added to the solution. The following relationship is

used if the lifetime does change:(2)


a= 1+k ro[Q]. (4-2)


This relationship enables one to determine if the fluorescence decay under dynamic quenching

conditions in the presence of the quencher molecule and provides the value of k,. In CPEs, the

quenching mechanism is both static and dynamic. In fact, the dynamic component does not

necessarily arise from the diffusion of the quencher in the solution but from the diffusion of the

excitation within the polymer chain. (130, 168)

It is well known that fluorescence within the visible spectrum from low concentrations of

CPEs can be superlinearly quenched when placed in the presence of an oppositely charged

electron- or energy quencher molecule (superquenching (146); amplified quenching (167,

169)).(151, 152, 154, 157, 158, 162, 163, 165, 166, 170) Amplified quenching may lead to the










2 3cos2a_
ro (4-6)
5 2



O No Absorption




Maximum Absorption






OA Absorption COS28A



Figure 4-6. Photoselection. Adapted from (2)

For a spherical obj ect, if the absorption and emission transition dipole moments are

parallel (a = 00), ro should equal to 2/5 (0.4); however, if they are perpendicular (a = 900) the

lower limit is -1/5 (-0.2). These values correspond to the limiting values. If all emission

polarization is lost (due to any of the processes listed above) a value of zero anisotropy is

expected. The temporal behavior of the anisotropy can provide useful information regarding the

polarization loss mechanism.(2)

Fluorescence anisotropy decay measurements were conducted by rotating excitation

pulses with respect to a fixed polarization detection scheme. A Berek compensator is used to

excite the molecule with a beam polarized parallel and perpendicular with respect to the detected

fluorescence intensities. The anisotropy value (r) was then calculated using:

I, -I
r = "(4-7)
I, + 27,










(120). S. A. Empedocles, M. G. Bawendi, Quantum-confined stark effect in single CdSe
nanocrystallite quantum dots, Science (1997) 278, 2114.

(121). V. Esch, B. Fluegel, G. Khitrova, H.M. Gibbs, J.J. Xu, K.Kang, S.W. Koch, L.C. Liu,
S.H. Risbud, N. Peyghambarian, State filling, coulomb, and trapping effects in the
optical nonlinearity of CdTe quantum dots in glass, Physical Review B (1990) 42,
7450.

(122). E. Hendry, M. Koeberg, F. Wang, H. Zhang, C.D. Donega, D. Vanmaekelbergh, M.
Bonn, Direct observation of electron-to-hole energy transfer in CdSe quantum dots,
Physical Review Letters (2006) 96, 125201.

(123). A. Shabaev, A. L. Efros, 1D exciton spectroscopy of semiconductor nanorods, Nano
Letters (2004) 4, 1821.

(124). C. Rulliere, Femtosecond Laser Pulses: Principles and Experiments, 2nd Edition
(2004).

(125). C. Burda, X. B. Chen, R. Narayanan, M. A. El-Sayed, Chemistry and properties of
nanocrystals of different shapes, Chemical Reviews (2005) 105, 1025.

(126). K. Brunner, U. Bockelmann, G. Abstreiter, M. Walther, G. Bohm, G. Trankle, G.
Weimann, Photoluminescence from a single GaAs/AlGaAs quantum dot Physical
Review Letters (1992) 69, 3216.

(127). L. W. Wang, M. Califano, A. Zunger, A. Franceschetti, Pseudopotential theory of
Auger processes in CdSe quantum dots, Physical Review Letters (2003) 91, 56404.

(128). A. L. Efros, V. A. Kharchenko, M. Rosen, Breaking the phonon bottleneck in
nanometer quantum dots Role of Auger-like processes, Solid State Communications
(1995) 93, 281.

(129). P. Guyot-Sionnest, M. Shim, C. Matranga, M. Hines, Intraband relaxation in CdSe
quantum dots, Physical Review B (1999) 60, R2181.

(130). J. G. Muller, E. Atas, C. Tan, K. S. Schanze, V. D. Kleiman, The role of exciton
hopping and direct energy transfer in the efficient quenching of conjugated
polyelectrolytes, Journal of the American Chemical Society (2006) 128, 4007.

(131). V. I. Klimov, A. A. Mikhailovsky, D. W. McBranch, C. A. Leatherdale, M. G.
Bawendi, Quantization of multiparticle Auger rates in semiconductor quantum dots,
Science (2000) 287, 1011.

(132). J. L. Blackburn, D. C. Selmarten, A. J. Nozik, Electron transfer dynamics in quantum
dot/titanium dioxide composites formed by in situ chemical bath deposition, Journal
of Physical Chemistry B (2003) 107, 14154.









the electron, lowering the confinement energy and ultimately decreasing the energy of the

exciton levels.(112) Based on the Raman data presented, the formation of interfacial ZnCdSe

results in a decrease in size of the CdSe core (56) resulting in increased localization and a blue

shifted emission. This is further supported from alloy formation resulting in a blue shift of the

photoluminescence peak.(42)

The energies of the corresponding absorption features from alloyed ZnCdSe (3hrs at

2700C) is considerably blue shifted to ~555 nm and ~465nm (Figure 2-11 C). These features

originate from the states similar to those in the core and core/shell nanorods but with a larger

band gap due to the formation of ZnCdSe. Figure 2-13 presents the photoluminescence spectra of

the alloyed ZnCdSe samples. Upon annealing at 2700C the photoluminescence spectra shift to

higher energies. After one hour of annealing, the peak appears at 610 nm. Further annealing (2

and 3 hours) produces a much larger blue shift, 510 and 565 nm, respectively. This behavior is

consistent with the variation in composition indicated by the broad Raman peak (Figure 2-10). In

addition to the energy shift, the alloys present changes in bandwidth and intensity. As alloying

time increases, the width of the photoluminescence band is reduced. The change in intensity does

not follow a trend, with the 3 hour alloyed sample presenting a sharp increase in photo-

luminescence intensity. Quantum yield measurements are ~8, 5 and 10% for 1, 2 and 3 hrs,

respectively. These values are higher than CdSe rods (0.6%) but lower than the core/shell sample

(15%). Composition disorder in ternary alloy nanorods will lead to localization of excitons

compared to binary samples.(42, 113) Such localization effects are known to improve the

photoluminescence efficiency by increasing the overlap integral of the electron and hole

wavefunctions. On the contrary, the quantum yield values are lower compared to the core/shell

due to the lack of surface passivation on the ZnCdSe nanorods. This is consistent with the









Previously, a series of steady state, time-resolved, anisotropy measurements and numerical

models were conducted with a similar CPE, PPE-SO3-, to determine the rate and efficiency

exciton migration has on fluorescence quenching.(130) Using PPE-SO3~ aS a model polymer

which exhibited both long range and random walk kinetics, we have designed experiments in

which the results should indicate the type of energy transfer present within PPE-CO2-. Time-

resolved photoluminescence and time-resolved anisotropy measurements were employed to

monitor the potential exciton hopping that was previously observed in PPE-SO3-, determine the

rise time of the aggregate state emission and characterize the overall polymer decay.

We studied very short and long polymer repeat unit (PRU) PPE-CO2- chains with the

expectation that short chains would be less likely to aggregate. We find that even short chains (8

PRU) form solutions with both isolated and aggregated chains. Even though the only difference

between PPE-SO3- and PPE-CO2~ is their ionic group, their photophysics are quite different.

PPE-CO2~ Steady state photophysics correlate more with ladder-type (poly-paraphenylene)

(LPPP) polymers. (116, 166, 174-177)

In most cases, conjugated polymer chains are not "frozen" in one conformation, instead

they have a proclivity to twist and coil. A series of chromophores can be linked resulting in

different degrees of xn-electron delocalization depending on the planarity of the conjugated

segments. Even if there are slight twists or bends along the polymer backbone, it is possible for

the conjugation to not completely break, resulting in larger delocalization lengths.(130, 159) Just

as in semiconductor nanoparticles, a particle-in-a-box model (1-D for polymers) is used to

explain the delocalization of excitons along the polymer backbone. Conjugation lengths that are

long tend to have lower xn x~ transition energies and vice versa.(159) Longer conjugation

lengths can be due to polymer rigidity which can create small shifts between the absorption and

































To my family









When the temperature was > 210oC, the emission blue shifted because the shell began alloying

causing a blue shift in the emission. If the temperature was kept too low, for example <170oC,

only a small shell grew because the Zn-oleate complex reacted too slowly with TOP-Se to grow

a shell, resulting in very weak emission. Finally, diffusion of Zn from the shell to the core is

instigated by raising the temperature slowly to 270oC to form ZnCdSe alloys.(42)

Structure of ZnCdSe Nanorods

X-Ray diffraction patterns for hexagonal CdSe, CdSe/ZnSe, and ZnCdSe nanorods are

shown in Figure 2-7.

(0 02) (1 10) (1 03) (1 12)
(1 00)
(101)





B)I




A)

20 30 40 50 60
2Theta

Figure 2-7. Powder X-ray diffraction patterns of A) CdSe nanorods, B) CdSe/ZnSe core/shell
nanorods, and C) ZnCdSe alloyed nanorods. Adapted from H. Lee.(66).

The crystal structure for this series of rods can be extracted from this experiment. In each of the

materials measured, the (002) diffraction peak is not as broad as the (001) diffraction peak. The

(002) peak is assigned to the plane that is perpendicular to the extended c-axis in rod-shaped

materials. The lattice spacings for CdSe, CdSe/ZnSe, and ZnCdSe were 7.01 A+, 6.94 A+, and 6.77

A+, respectively. These values are extremely interesting since in CdSe/ZnS nanoparticles the ZnS










is generally slow (>10 ps) and is related to the rate at which the carrier is trapped either at the

surface or other defect sites. The trapped electron can then relax to the ground state via an

alternative radiative decay pathway or decay non-radiatively.(67, 72) The actual rates and overall

dynamics are indistinguishable because of the inhomogeneity and ensemble averages of the

samples; therefore, direct assignment of rates and pathways is extremely difficult. On the other

hand, photoinduced absorption within the core/shell rods arises from interface defects resulting

in states that can act as traps and/or non-radiative recombination sites caused by lattice strain

relaxation introduced between the core and shell.(38, 65, 72) It appears that after the initial

Auger cooling from the 1P to 1S state, some carrier populations sample the ZnCdSe interfacial

layer resulting in photoinduced absorption. The signal is strong and lasts more than 500 ps

(Figure 3-4).

Compositional disorder (42, 108, 109) in ternary alloy nanorod structures leads to

localization of excitons, (113) hence an increase in quantum yield for ZnCdSe versus CdSe

nanoparticles. Localization effects increase the overlap integral of the electron and hole

wavefunctions improving the luminescence efficiency of the material and decreasing the bleach

lifetime for the ternary alloy samples compared to the binary CdSe materials. The lower ZnCdSe

quantum yield versus the CdSe/ZnSe core/shell nanorods results from the lack of surface

passivation and crystal defects within the ZnCdSe nanorods due to Zn diffusion into the CdSe

core. Increasing the Zn character in the core causes a blue shift in the spectrum as alloying time

increases. Raman data presented in Figure 2-11 show that CdSe and ZnSe phonon modes are no

longer present; the only mode observed is the ZnCdSe state. The lack of or small amount of

photoinduced absorption confirms that the alloy nanorod composition is uniform and few

interfacial traps are present. Since the inorganic shell disappears, surface traps reoccur and









Photovoltaics (213, 214) Although the cost of making quantum dot based photovolatics is

small, the efficiencies, due to recombination loses, are still too low for them to be used on a large

scale. Hybrid photovoltaic devices are integrated within the polymers to transport charge for

such applications as solar cells. Achieving charge separation and positive transport of the hole

and electron to the indium tin oxide (ITO) and aluminum electrode, respectively is the main goal

in photovoltaics.(66, 214) Instead of focusing on a binary system, a device utilizing ternary

compositions with varying degrees of Zn diffused into the core, may serve as more suitable

materials for photovoltaics. The variable Zn diffusion will create a gradient from the core to the

surface enabling the exciton to hop from one rod to another ultimately reaching the aluminum

electrode. Also, work might be directed to achieve charge separation within the nanocrystals.

The fact that the hole created after excitation within our CdSe/ZnSe quantum rods potentially

tunnels into the shell could help sustain charge separation and inhibit premature charge carrier

recombination. Investigations into the kinetics and mechanisms for creating and maintaining

charge separation in these materials are recommended since these processes are not completely

understood; however, it has been shown that the interface between nanocrystals and porous TiO2

supports highly efficient charge separation.(132, 199, 202)

Carrier multiplication The carrier-carrier interactions in nanoparticles lead to improved

exciton (carrier) multiplication (CM) which results from direct photogeneration of multiple

electron-hole pairs by single photons. This process is relatively new and the actual mechanism

behind carrier multiplication (if it truly exists) from a single excitation are currently being

debated. Klimov showed that seven excitons are produced in PbSe nanocrystals (QE = 700%

where 100% means 1 photon creates 1 e-h pair). This unique finding will be good for

photovoltaic cells and improve solar fuel technologies in the IR region. However, CdSe dots










(184). New Focus, M~odel 5540 User's Manual: Thze Berek Polarization Conspensator.

(185). New Focus, Polarization and Polarization Control.

(186). A. Yariv, and Yeh, P., Optical Waves in Crystals: Propagation and Control ofLaser
Radiation (Wiley-Interscience 2002).

(187). R. Raj asekaran, Personal conanunication for Berek Conspensator Crystal
Specifications, (2007)

(188). M. Dodge, Refractive properties of magnesium fluoride, Applied Optics (1984) 23,
1980.

(189). J. R. Lakowicz, Principles ofFluorescence Spectroscopy (Kluwer Academic/Plenum
Publishers, New York, ed. 2, 1999).

(190). R. E. Martin, F. Diederich, Linear monodisperse pi-conjugated oligomers: Model
compounds for polymers and more, Angewandte Chenzie-hiternational Edition (1999)
38, 1350.

(191). U. H. F. Bunz, J.M. Imhof, R.K. Bly, C.G. Bangcuyo, L. Rozanski, D.A.V. Bout,
Photophysics of poly [p-(2,5-didodecylphenylene)ethynylene] in thin films,
Macrontolecules (2005) 38, 5892.

(192). C. E. Halkyard, M. E. Rampey, L. Kloppenburg, S. L. Studer-Martinez, U. H. F.
Bunz, Evidence of aggregate formation for 2,5-dialkylpoly(p-phenyleneethynylenes)
in solution and thin films, Macrontolecules (1998) 31, 8655.

(193). J. Kim, D. T. McQuade, S. K. McHugh, T. M. Swager, lon-specific aggregation in
conjugated polymers: Highly sensitive and selective fluorescent ion chemosensors,
Angewandte Chenzie-hiternational Edition (2000) 39, 3868.

(194). J. Kim, T. M. Swager, Control of conformational and interpolymer effects in
conjugated polymers, Nature (2001) 411, 1030.

(195). M. Abramowitz, Johnson I. D., and Davidson, M. W., Fluorescence filter spectral
transmission~rtrtrtrt~t~t~ profiles,
http://www. olympusmi cro. com/primer/j ava/fluorescence/fluorocub es/index.html,

(196). V. F. Kamalov, I. A. Struganova, K. Yoshihara, Temperature dependent radiative
lifetime of J-aggregates, Journal of Physical Chentistry (1996) 100, 8640.

(197). D. Anestopoulos, Fakis, M etal, Excitation energy transfer in a cationic water-soluble
conjugated co-polymer studied by time resolved anisotropy and fluorescence
dynamics, Chemical Physics Letters (2006) 421, 205.














-0.25



E -o.so



-0.75



-1.00

0 100 200 300 400 500

time (ps)
Figure 3-8. Comparison of the 1S band for the CdSe/ZnSe core/shell (-) and 3 hr ZnCdSe alloy


Discussion

In nanocrystals, optical transitions resulting in ground state absorption changes are due to

state-filling effects while extremely fast transitions (<1 ps) result from Coulomb interactions, i.e.,

Stark Effect.(1 7 72, 121) Red shifts are observed in CdSe quantum rods at longer delay times;

they are identified as a convolution of the S-type states near the band gap stemming from the

inherent size distribution present in colloidal nanoparticle samples.(72) The relaxation dynamics

within these systems are strongly influenced by ensemble dynamics collectively creating

inhomogeneities and also multiple photoinduced processes leading to multi-exponential or non-

exponential behaviors.(10)

In analogy to bulk materials, cooling of hot electrons could occur via emitting LO

phonons, and this mechanism would result in slow decay rates since the spacing of the intraband

states is large in quantum systems resulting in a phonon bottleneck.(31, 111, 122) However,

several studies (17, 111, 132) including our results, demonstrate a different behavior. High









includes additional sources of broadening, arising from the experimental setup and

instrumentation.(98, 99) Using Eq. 2-5, the particle sizes of ZnCdSe nanorods calculated were a

diameter of 5.5 nm and a length of 1 1.8 nm resulting in an aspect ratio of ~2. 1 nm, which agree

well with HR-TEM data shown in Figure 2-8.(42, 66)

Effect of Alloying on the Phonon Spectra

The compositional changes to the structure of a material that arise when adding a shell or

alloying by diffusion through the dependence of the phonon frequencies have been studied using

Raman spectroscopy.(42, 100, 101) The Raman peaks detected from CdSe nanorods are shown

in Figure 2-9 A. The peak at ~206 cml is from the CdSe LO phonon (42, 101, 102) which is 4

cm-l shifted compared to the bulk CdSe (210 cm- ) which is due to the quantum confinement of

the optical phonons in the nanorods.(42, 100-102) A broad "shoulder" (~180cm- ) appears to the

left of the main mode which arises from the non-spherical geometry of the CdSe nanorods. (42,

103, 104)

The Raman peak for CdSe/ZnSe core/shell nanorods is shown in Figure 2-9 B. The

original CdSe LO phonon mode is still detected with the addition of the ZnSe shell mode at ~247

cm- ). A new interfaciall ZnCdSe" is also detected and corresponds to a frequency ~23 5 cml

The small, unresolved Raman peaks on either side of the CdSe phonon mode can be assigned to

isolated atom-impurity modes when Zn and Cd atoms interchange with one another (Zn in CdSe

~190 cm-l and Cd in ZnSe ~218 cm )~. (42, 105)

The effects of alloying time (1, 2 or 3 hrs at 270oC) on the Raman spectra are shown in

Figure 2-10. After alloying, one mode is present at ~223 cm l, ~228 cm-l and ~226 cml (1, 2 and

3 hrs, respectively), which is similar to the interfacial layer observed in the core/shell material

and the one phonon-mode behavior for bulk ZnCdSe. (42, 106, 107)









rods and CdSe/ZnSe core/shell rods. In this work, we explore how addition of shell with only

one inorganic material with a small valence band offset affects the photophysical properties and

then compare this data to the exciton behavior in ternary alloy heterostructures.

Experimental Methods: Transient Absorption

Relaxation processes of colloidal nanocrystals were explored using femtosecond transient

absorption (TA). 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 k
excitation pulses. A portion of the amplifier output is split off to pump a 1 mm rotating CaF2

window to generate white light continuum probe with an effective bandwidth ranging from 310

to 750 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. A detailed description is available

elsewhere. (130) A general schematic is provided in Figure 3-2. The OPA idler/signal output is

used to produce excitation pulses (pump) through harmonic generation (450, 575, 610, 630 and

650 nm). This beam is then fed through a prism compressor, resulting in pulse lengths less than

100 fs (FWHM). The excitation beam is focused to a diameter of ~150 Clm at the sample position

and its energy was set to ~ 39 to 45 nJ yielding a fluence of 221 to 255 CIJ/cm2. Low fluences are

necessary to avoid multiple excitations (biexcitons). From previous works, it is known that a

signature of multiparticle interactions are decay rates that occur faster than 50 ps. Klimov also

observed that the decay rates increased as the number of excitons per nanoparticle

increased.(131) Experimentally, I verified the fact that multiple excitations per nanocrystal were

not initially created via a power dependence study. Our data show (Figure 3-4: no fast decay











4-10. The emission spectra 10CLM PPE-CO2- in methanol ................. ........__ ........._.._. 106

4-11. The emission of 10 CLM 35 PRU PPE-CO2- ............. .....................108

4-12. The excitation spectra 10 CLM 8 PRU PPE-CO2-. ............ ...............109.....

4-13. The excitation spectra 10 CLM 35 PRU PPE-CO2- in methanol .........._.._. ......_.._.......110

4-14. The excitation spectra 10 CLM 35 PRU PPE-CO2- in water ................. ........___.........111

4-15. The excitation spectra 10 CLM 35 PRU PPE-CO2- in methanol with Ca2+. ................... .....112

4-16. The time-resolved fluorescence decay of 8 PRU PPE-CO2- in methanol .........._............114

4-17. The time-resolved fluorescence decay of 30 CLM PPE-CO2- with different polymer
repeat units in methanol. ........._.. ........... ...............116..

4-18. The time-resolved fluorescence decay of 30 CLM PPE-CO2- (35 PRU) with and without
Ca2+ .......... ...............117......

4-19. The time-resolved fluorescence decay of 10 CLM PPE-CO2- (8 PRU) with and without
Ca2+ ............ ... ...............118..............

4-20. The time-resolved fluorescence decay of 30 CLM PPE-CO2- 35 PRU with different
quenchers ................. ...............119......... ......

4-21. The anisotropy of 8 PRU PPE-CO2-.........._._. ......... ...............12

4-22. Possible kinetic model for all PPE-CO2- PRU chains .............. ...............123....

5-1. The absorption spectra of the hyperbranched PPE-CO2- and the linear PPE-CO2-.............13 1

5-2. The photoluminescence spectra of hyperbranched PPE-CO2- ................ ......._... .......132










(13). M. A. El-Sayed, Small is different: Shape-, size-, and composition-dependent
properties of some colloidal semiconductor nanocrystals, Accounts of Chemical
Research (2004) 37, 326.

(14). M. A. Fox, M. T. Dulay, Heterogeneous photocatalysis, Chemical Reviews (1993) 93,
341.

(15). L. N. Lewis, Chemical catalysis by colloids and clusters, Chemical Reviews (1993)
93, 2693.

(16). J. Aldana, Y.A. Wang, and X.G. Peng, Photochemical instability of CdSe
nanocrystals coated by hydrophilic thiols, Journal of the American Chemical Society
(2001) 123, 8844.

(17). V. I. Klimov, Optical nonlinearities and ultrafast carrier dynamics in semiconductor
nanocrystals, Journal of Physical Chemistry B (2000) 104, 61 12.

(18). V. I. Klimov, D. W. McBranch, Femtosecond 1P-to-1S electron relaxation in strongly
confined semiconductor nanocrystals, Physical Review Letters (1998) 80, 4028.

(19). N. Le Thomas, E. Herz, O. Schops, U. Woggon, M. V. Artemyev, Exciton fine
structure in single CdSe nanorods, Physical Review Letters (2005) 94, 016803.

(20). C. A. Leatherdale, W. K. Woo, F. V. Mikulec, M. G. Bawendi, On the absorption
cross section of CdSe nanocrystal quantum dots, Journal of Physical Chemistry B
(2002) 106, 7619.

(21). M. L. Steigerwald, L. E. Brus, Semiconductor crystallites A class of large
molecules, Accounts of Chemical Research (1990) 23, 183.

(22). A. P. Alivisatos, Perspectives on the physical chemistry of semiconductor
nanocrystals, Journal of Physical Chemistry (1996) 100, 13226.

(23). A. L. Efros, A. L. Efros, Interband absorption of light in a semiconductor sphere,
Soviet Physics Semiconductors-Ussr (1982) 16, 772.

(24). A. L. Efros, M. Rosen, The electronic structure of semiconductor nanocrystals,
Annual Review of2aterials Science (2000) 30, 475.

(25). A. I. Ekimov, F. Hache, M. C. Schanne-Klein, D. Ricard, C. Flytzanis, I. A.
Kudryaytsev, T. V. Yazeva, A. V. Rodina, and Al. L. Efros, Absorption and intensity-
dependent photoluminescence measurements on CdSe quantum dots- Assingment of
the 1 st electronic-transitions, Journal of the Optical Society of America B-Optical
Physics (1993) 10, 100.

(26). L. E. Brus, Electron electron and electron-hole interactions in small semiconductor
crystallites- The size dependence of the lowest excited electronic state Journal of
Chemical Physics (1984) 80, 4403.









the 532 nm line from a Verdi 8 doubled Nd-YAG solid state laser in a Ramanor U-1000 Jobin-

Yvon Raman spectrometer.(42)

Absorption spectra were collected with a Shimadzu UV-2401PC spectrophotometer.

Photoluminescence was measured at room temperature using nanorods suspended in toluene

using a Fluorolog Tau 3 spectrofluorometer (Jobin Yvon Spex instruments, S.A. Inc.). The

photoluminescence quantum yield was determined using Rhodamine 6G organic dye standard.

(42)

Time-Resolved Photoluminescence Instrumentation

Relaxation processes of colloidal nanocrystals were explored using time-resolved

photoluminescence. 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 k
parametric amplifier (OPA) to generate excitation pulses. For this experiment, since 400 nm is at

the limit of both the signal and idler, we must use the second harmonic of the amplifier (800 nm)

to achieve stable and high energy pulses. The residual 800 nm is directed to a horizontal BBO

crystal (output of the Spitfire is polarized in the horizontal direction). A general schematic is

provided in Figure 2-6. The second harmonic (400 nm) is then fed through a prism compressor,

resulting in pulse lengths less than 100 fs (FWHM). The excitation beam is focused to a diameter

of ~150 Clm at the sample position and its energy was set to~- 56 nJ yielding a fluence of 3 17

CLJ/cm2. The optical density of each solution was 0.075/mm at 400 nm. Sample solutions of

colloidal nanorods dissolved in toluene 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.

Broad band luminescence (grating range: 438 to 718 nm) from the sample was collected using a










semiconductor quantum dots, ultimately changing the behavior of the excitons.(1 7) The

absorption spectra of five colloidal CdSe nanocrystals with different radii (1.2, 1.7, 2.3, 2.8, and

4.1 nm) is shown in Figure 2-3 (17) and illustrates not only the quantum dot band gap

dependence but the features corresponding to the optical transitions that arise from the coupled

electron and hole electronic states previously discussed.(16-19)



R = I.2 nm (6 = 4 4%)


1.7 nm lr


10 2.3 nm (6%)












4 ~e~ -7I:h



O f ISe)- 2S.~Ilt IP(e) IP3,2h)
0 ~ e- : ISlc)25.,

2.0 2.2 2.4 2.6 2.8 3.0 3.2
Photon energy leV)

Figure 2-3. Absorption spectra of TOPO/TOP passivated CdSe nanocrystals with radii from 1.2
to 4. 1 nm.(1 7)

Size and Shape Dependence

Quantum confinement or the "quantum size effect" is a property that in recent years has

revolutionized the semiconductor industry. This effect leads to unique electronic and optical

properties making quantum dots differ from their bulk counterparts. II-VI semiconductor










(198). A. A. Mikhailovsky, A. V. Malko, J. A. Hollingsworth, M. G. Bawendi, V. I. Klimov,
Multiparticle interactions and stimulated emission in chemically synthesized quantum
dots, Applied Physics Letters (2002) 80, 2380.

(199). V. I. Klimov, Mechanisms for photogeneration and recombination of multiexcitons in
semiconductor nanocrystals: Implications for lasing and solar energy conversion,
Journal of Physical Chentistry B (2006) 110, 16827.

(200). D. Katz, T. Wizansky, O. Millo, E. Rothenber, T. Mokari, U. Banin, Size-dependent
tunneling and optical spectroscopy of CdSe quantum rods, Physical Review Letters
(2002) 89, 086801.

(201). H. Htoon, J. A. Hollingworth, A. V. Malko, R. Dickerson, V. I. Klimov, Light
amplification in semiconductor nanocrystals: Quantum rods versus quantum dots,
Applied Physics Letters (2003) 82, 4776.

(202). S. A. Ivanov, J. Nanda, A. Piryatinski, M. Achermann, L.P. Balet, I.V. Bezel, P.O.
Anikeeva, S. Tretiak, V.I. Klimov, Light amplification using inverted core/shell
nanocrystals: Towards lasing in the single-exciton regime, Journal of Physical
Chentistry B (2004) 108, 10625.

(203). J. Nanda, S.A. Ivanov, H. Htoon, I. Bezel, A Piratinski, S. Tretiak, V.I. Klimov,
Absorption cross sections and Auger recombination lifetimes in inverted core-shell
nanocrystals: Implications for lasing performance, Journal ofAppliedPhysics (2006)
99, 013707.

(204). V. L. Colvin, M. C. Schlamp, A. P. Alivisatos, Light-emitting-diodes made from
cadmium selenide nanocrystals and a semiconducting polymer Nature (1994) 370,
354.

(205). B. O. Dabbousi, M. G. Bawendi, O. Onitsuka, M. F. Rubner, Electroluminescence
from CdSe quantum-dot polymer composites, ApplieadPhysics Letters (1995) 66,
1316.

(206). H. Mattoussi, L.H. Radzilowski, B.O. Dabbousi, E.L. Thomas, M.G. Bawendi, M.F.
Rubner, Electroluminescence from heterostructures of poly(phenylene vinylene) and
inorganic CdSe nanocrystals, Journal ofAppliedPhysics (1998) 83, 7965.

(207). H. S. Yang, P. H. Holloway, Electroluminescence from hybrid conjugated polymer -
CdS : Mn/ZnS core/shell nanocrystals devices, Journal of Physical Chentistry B
(2003) 107, 9705.

(208). S. Coe, W. K. Woo, M. Bawendi, V. Bulovic, Electroluminescence from single
monolayers of nanocrystals in molecular organic devices, Nature (2002) 420, 800.

(209). C. Bonati, M.B. Mohamed, D. Tonti, G. Zgreblic, S. Haacke, F. van Mourik, M.
Chergui, Spectral and dynamical characterization of multiexcitons in colloidal CdSe
semiconductor quantum dots, Physical Review B (2005) 71, 205317.










Thus, the greater the non-radiative decay rate constant, the smaller the quantum yield, and vice

versa. The time that it takes for the excited state of a molecule to decay to 1/e of its initial value

is the lifetime of the excited state, which is given by:


S=~ (1-5)
kvad + k,,,

The fluorescence emission intensity, quantum yield, and lifetime can be negatively

affected by numerous quenching processes including collisions with heavy atoms, electron

transfer, energy transfer, excimer formations, aggregate formation, and dynamic collisions.(2)

The quenching processes discussed above can be measured but the results of these experiments

can be hard to interpret. Instead, fluorescence anisotropy is employed to understand amplified

quenching observed in conjugated polyelectrolytes.

Direct measurement of the random walk "hopping" of excitons is a difficult task since this

process can compete with other energy transfer processes. Time-resolved anisotropy is type of

measurement that measures the decay of polarized fluorescence, which gives a better

understanding of the random intrachain energy migration in a material. The sensitivity to

depolarizing the transition dipole moments between an absorbing and emitting molecule is

directly related to the loss of anisotropy. Excitation with light polarized in a particular direction

will only excite molecules with the same orientation. For example, a vertical excitation will

preferentially excite molecules with vertical transition dipole moments. Anisotropy values will

not change if as the exciton migrates there is no change in the dipole moments between the

chains that absorb and emit. However, if these dipoles do change, as the exciton "hops", it loses

its original orientation and the fluorescence signal depolarizes.(4, 5)









development of more sensitive sensors but a complete explanation responsible for such high

quenching efficiencies within CPEs has yet to be determined. This intriguing effect is not only

due to ion-pairing between the polymer and quenchers (157, 163, 164, 1 70) as in typical Stern-

Volmer kinetics, but also inter- and intrachain energy transport mechanisms. More specifically,

the random walk diffusion of the excitation energy along the polymer backbone, (154, 157, 1 71,

1 72) energy transfer between the polymer and quencher (1 71) and energy transfer between the

isolated polymer species and aggregated chains all contribute to such unique behavior.(143, 146,

162, 165-167)

Energy transfer is strongly dependent on the spectral overlap between the donor emission

and acceptor absorption. The energy transfer is also very rapid. (168-1 71) If an aggregate inducer

or quencher is added to the polymer solution, the conformation can change resulting in a spatial

redistribution of several chromophores. This enables the excitation located on the polymer

backbone to easily migrate to the quencher located at a particular site lower in energy. Therefore,

one quencher molecule can have the ability to reduce the emission from a large number of

chromophores. (157, 1 73) The intrachain random walk model, which leads to excitation

migration towards the quencher molecule, is strongly dependent on the conjugation, polymer

chain lengths and transition dipole orientations (Figure 4-1). Using time-resolved anisotropy

fluorescence measurements, it is seen that after excitation, the energy or exciton "hops"/migrates

from shorter (high energy segments) to longer (low energy ones) and depolarizes along the way,

reducing the anisotropy value. The exciton will continue to funnel through the cascade of

chromophores until it is either "trapped" or it reaches the lowest energy level where it can

fluoresce or non-radiatively decay.










overgrowth of a higher band gap inorganic shell (Figure 2-5) creates a "step" to confine the

exciton to the core.(39-41, 65) This increased confinement has been employed to enhance the

quantum yield of the dots to over one order of magnitude and to increase its stability against

surface oxidation. (34, 38, 72) Various examples using ZnS, CdS and ZnSe as shell layers

include CdSe/ZnS (39, 40), CdSe/CdS (11), CdSe/ZnSe (73), CdS/ZnS (34) and InAs/ZnSe (74).

Passivation using multiple shells (34, 70, 75) or "onion like" structures (76) has also been

achieved.

It has been observed that the absorption and emission of a nanocrystal that is passivated

with a ZnS shell exhibits a shift to longer wavelengths by approximately 10 to 20 nm as

compared to the unpassivated core.(39, 40, 66, 77) Dabbousi et al. explained this phenomenon

by considering charge carriers in a spherical box. The observed shifts to lower energies result

from the tunneling of the lighter electron wave function into the shell while the hole remains in

the core. If this happens, the exciton is "delocalized" in the particle resulting in decreased

confinement and excited state energy. For the electron to be able to penetrate into the shell, it

must be able to overcome the valence band offset (barrier height), the energy difference between

the valance band of the core and the valence band of the shell, that is present between the core

and shell. If this offset is small, the shifts towards lower energies can be large.(39, 66) For

nanoparticles passivated with an inorganic material, a critical thickness is present that is

dependent on the size of the core and lattice mismatch between the core and shell. (34, 38, 72)

This thickness influences the ability of the electron to tunnel to the surface potentially resulting

in little to no shifts in band gaps or confinement potentials compared to unpassivated particles.

However, it is possible to surpass this thickness allowing the electron to tunnel into the shell

layer which can decrease the overall quantum yield and shift the absorption and emission









for helping me out in so many ways, especially in the last few weeks, it has been greatly

appreciated.

I have dedicated this dissertation to my family. To me, family means more than just blood

relatives, it is who you believe supports you and will be there for you always. To start, I want to

thank Coach Dr. Nancy Bottge. She was such an influential person in my life and taught me that

I must "stick to the fight when hardest hit". Her "don't quit" attitude is one of the reasons for this

success. She had not only been a mentor and a coach but a friend that I could turn to and who

taught me so many invaluable lessons. I extend extreme gratitude to Chad Mair who has been

beside me each step of my life in the past five years. I have been blessed with having such a

good friend that I love so much that I consider him to be my brother. I can count on him for

anything and know that without him, this achievement in my life may not have been possible. He

has been the best work out partner, best friend and best colleague a girl could ask for. Jana

Vanderloop, my best friend of ten, going on forever, years has also been a rock for me to lean on.

I will always be sure to appreciate our "inner randomness" because without it, life is too serious.

I enj oy the fun we have on a daily basis, it keeps me sane. From day one, Richard Farley and I

have battled our way through the trials and tribulations of grad school. I thank him for his

companionship, sense of humor and open mindedness. I thank Roxy "Rory" Lowry and Todd

Prox for their friendship, laughs and their ability to give me different perspectives on all

situations I run into in life. I also appreciate the encouragement and advice that I received from

my friend Jim Reynolds. I thank Megan Meyer for her intense sarcasm because no matter what

mood I am in, it always puts a huge smile on my face.

My time here would not have been the same without the social activities provided by all

my friends in Gainesville. I am extremely grateful that I was a part of such a fun group that









CHAPTER 5
CONCLUSIONS AND FUTURE WORK

Nanoparticle Conclusions and Future Work

Conclusions

A complete steady state and time-resolved study of size, shape passivation and

composition dependence on colloidal semiconductor nanoparticles has been conducted in our

labs. Using pump-probe spectroscopy we were able to detect traps and interfacial states.

Confirmation of Auger-like cooling resulting in 1P to 1S relaxation (~ 1 ps) has been shown in

addition to interband relaxation (> 200 ps) have been measured for each nanoparticle system.

Comparisons between materials with different compositions were made finding higher

confinement potential in ZnCdSe alloys than in CdSe resulting in a lower probability of the

exciton to sample the surface and be trapped. Finally, utilizing each of the spectroscopic tools

available we were able to combine steady state and time-resolved data to construct qualitative

models describing the nanorod systems.

From this work we can conclude that passivation and alloying result in quantum yields

higher than for bare CdSe. The excitons are more confined in the alloy particles than in CdSe

rods. It is well known that the band gap is size (17) and shape dependent.(53) Despite breaking

the symmetry within the nanoparticle, confinement properties were maintained in our samples.

Passivation with an inorganic shell results in increased quantum yields and bleach signals

because the surface traps are eliminated. Finally, modifying the composition using one additional

synthetic alloying step has greatly improved the process-ability without drastically sacrificing

confinement characteristics.









UJLTRAFAST SPECTROSCOPY OF NOVEL MATERIALS


By

LINDSAY M. HARDISON


















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

2007









three-neck flask on a Schlenk line under a N2 atmosphere to 350oC 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 280oC while stirring, and selenium

dissolved in trioctylphosphine (TOP) was inj ected quickly. After inj section, the temperature of the

mixture was kept at 250oC for the 30 min growth of CdSe nanorods, and then cooled to 180oC.

(42)

For shell growth, ZnO was dissolved in oleic acid (Zn-oleate) at 350oC and cooled to room

temperature, and then TOP was added to prevent solidification. In addition, Se was dissolved in

TOP (Se-TOP). The Zn-oleate and Se-TOP solutions were mixed by stirring for ten minutes at

room temperature, and this mixture was loaded into a syringe and inj ected drop-by-drop into the

reaction flask over 1.5 hr. After inj section was complete, the solution was stirred at room

temperature for another ten minutes. For alloying, the reaction vessel was heated with stirring to

270oC 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. (42)

Steady State Instrumentation

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









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

UJLTRAFAST SPECTROSCOPY OF NOVEL MATERIALS

By

Lindsay M. Hardison

December 2007

Chair: Valeria D. Kleiman
Major: Chemistry

My research focused on steady state and time-resolved photophysical characterization of a

series of semiconductor nanoparticles and water soluble conjugated polyelectrolytes. Several

studies have shown that the electronic structure and relaxation dynamics in CdSe nanocrystals

are not only size but are also shape and passivation dependent; however, there is no detailed

comparison of the photophysical properties of ZnCdSe particles with different relative amounts

of Zn. This dissertation presents data collected for colloidal CdSe, CdSe/ZnSe and ZnCdSe

nanoparticles with rod-like architectures synthesized and investigated in our labs to determine

how size, shape, passivation and composition affect the quantum confinement and dynamics. In

addition, a series of different polymer repeat unit lengths of a linear conjugated polyelectrolyte

(CPE) with a carboxylate ionic side chain have been synthesized and their photophysical

properties have been explored.

Spectral shifts and line broadening exhibited within the Raman spectroscopy, UV-Vis

spectroscopy and photoluminescence aided in determining the extent of alloying and

compositional disorder created during the alloying process. The photoluminescence quantum

yield of ZnCdSe nanorods is higher than that from pristine CdSe nanorods indicating a higher

binding energy of the exciton. This effect is speculated to be due to increased localization of the









radiative energy transfer process in which the photon emitted by the donor is then absorbed by

the acceptor. Therefore, it does not compete with other decay mechanisms and the fluorescence

decay time of the donor remains unchanged (refer to Chapter 1). Following the energy transfer,

the detected emission is dominated by the ensemble of decays from isolated chains and traps

located along the polymer backbone in addition to competing with non-radiative decay channels.

In summary, upon excitation of the aggregates from energetically higher lying isolated chains,

the fluorescence lifetimes result in multi-exponential behavior due to the competition between

the radiative and non-radiative decay. The integrated fluorescence collected for these

experiments does show emission from aggregates but it cannot be observed in the ultrafast time-

resolved experiments because of very long decay time constants and small contribution to the

overall signal.

Summary

The ultrafast time-resolved fluorescence of a series of PPE-CO2~ pOlymer repeat units is

presented. Using steady state UV-Vis, photoluminescence and excitation resources we

distinguished the species present in each solution. It was shown that even dilute, short PRU

chains do exhibit a small amount of aggregation. The addition of calcium or using water as the

solvent induces aggregation resulting in broad absorption/excitation spectra and the growth of a

red shoulder in the emission. To investigate the influence aggregation has on the fluorescence of

the polymers, we conducted a detection wavelength study using fluorescence upconversion. The

isolated chain emission was extracted from the 8 PRU at 450 nm (isolated chain emission and

aggregate absorption is minimal). In the presence of aggregates, an intermediate time constant on

the order of 30 to 40 ps is observed and is assigned to the energy transfer from the isolated to

aggregate species. At bluer wavelengths, a fast decay (< 1.5 ps) is observed and is attributed to

the transfer of excitation from shorter, high energy chains to longer, low energy chains and traps.









From the evidence presented in this thesis, it is conceivable that during the Auger cooling of the

hot electron from the 1P to 1S state, excess energy could cause the hole to become delocalized

within the ZnCdSe interface or even in the ZnSe shell. The bleach recovery from the conduction

band to ground state valence band is the longest in the core/shell materials due to passivation of

surface traps and potentially due to the position of the hole wavefunction.

Once the materials are alloyed, the ZnCdSe in Figure 3-11 becomes the only inorganic

material present. Again, 1P to 1S relaxation is observed with high energy excitation followed by

subsequent interband relaxation from the conduction to valence band. However, based on the

work completed by Rosenthal et al.,(67) midgap surface states involving selenium dangling

bonds are present due to the lack of inorganic passivation. An electron relaxing from the surface

Se atoms to the valence band can immediately fill the vacancy left by the photogenerated

electron contributing to the deep trap emission observed at 700 nm.

1Pe

<1ps

1Se V \ Trap State





Interband
hy= 2.75 el Relaxation /Surface Trap





0.4 eV 1s/
1P3/2
S2S1/2

Figure 3-11. ZnCdSe alloy potential kinetic model.










Experimental Methods

Synthesis of Variable Chain Lengths of PPE-CO2

Xiaoyong Zhao, a member of the Schanze group, is responsible for the synthesis and some

of the steady state characterization of the various chain lengths of PPE-CO2~ inVCStigated within

this dissertation.

To polymerize a stoi chi ometri c mixture of 2, 5 -bi s-(dodecyl oxy-carb onylmethoxy)- 1,4-

diiodobenzene and 1 ,4-di-ethynylbenzene a "precursor route" in which a Sonagashira coupling

reaction is used to produce a poly(phenylene ethynylene) with a dodecyl ester protecting the

carboxyl group. Gel permeation chromatography of the ester precursor polymers showed that

the molecular weight (Mn) for the four polymer chain lengths investigated in this dissertation are

S5000, 24000, 74000 and 127000 g-moll corresponding to average degrees of polymerizations

(Xn) of 8, 35, 108, 185, respectively. The protected ester polymer precursor was then hydrolyzed

with (n-Bu)4NOH to provide for the water-soluble conjugated polyelectrolyte PPE-CO2-. The

final polymer product was purified using dialysis against DI water for 4 days. All of the

polymers have polydispersity indices of ~ 2.(181)

Photophysical Methods

UV-Visible absorption spectra were recorded using a Lambda 25 spectrophotometer form

Perkin Elmer. Steady-state excitation and emission spectra were obtained with a Fluorolog-3

spectrofluorometer from Jobin Yvon. A 1-cm square quartz cuvette was used for all spectral

measurements. Concentrations varied from 10 to 30 CLM and were dissolved in spectroscopic

grade methanol.

Time-resolved anisotropy and fluorescence dynamics measurements were performed using

a femtosecond upconversion apparatus. An optical parametric amplifier (OPA) pumped by a

commercial Ti:Sa laser system consisting of a Ti:Sa oscillator (Spectra-Physics, Tsunami) and a
















Atomic
Diatomic
Hybrid Orbitals





s Antibondin




Bonding


the created exciton and electronic properties is of significant interest for the design and

engineering of useful bulk and nanoscale semiconductor materials. Consider a summary of the

band theory of solids presented in Figure 2-1.


Mol. Bulk Quantum
)rbitalsl Solid Solids

,UMO CB" IC



iOMO VB g


Discrete States Density of States Discrete States

Figure 2-1. Band theory of solids

Silicon has four sp3 hybridized atomic orbitals. Neighboring atoms contribute orbitals

which combine to form highest occupied molecular orbitals (bonding orbitals, o) and lowest

unoccupied molecular orbitals (antibonding orbitals, o*). The total number of occupied and

unoccupied orbitals is equal to the number of atomic orbitals present within the crystal. As more

atoms are added, a density of orbital energies develops reducing the spacing between the states in

each band. This increase in density results in a continuum of energies separated by a gap. In a

bulk solid, the highest occupied orbitals form the valence band and the lowest unoccupied

orbitals form the conduction band. The minimum energy required to excite an electron from the

top of the valence band to bottom of the conduction band is the band gap energy of the

semiconductor (E,).(21)











observed. After the initial change in the first 5 ps from r ~ 0.4 to 0.2, both curves then remain

parallel to one another.

0.1 4gggg


0.3-


0.2--


0.1-


0.0


-0.1 --


-0.2 gggg.
0 100 200 300 400 500 600

time (ps)

Figure 4-21. Anisotropy of 8 PRU PPE-CO2-. Detection wavelengths 430 (-) and 450 (-) nm

Detection at 430 nm occurs in a region in which shorter conjugation lengths are present

allowing for a higher number of hops before finding traps along the polymer backbone resulting

in smaller anisotropy value (r~ 0.10) compared to the 450 nm (r~ 0.18). The 450 nm scan

corresponds to slightly longer conjugation and fewer hops. As shown in other works (130, 197) a

long polarization decay corresponds to aggregates emitting almost randomly polarized light

reducing the total polarization. However, in this polymer there is little to no depolarization due to

reorientation at 450 nm. Random walk migration is considered to occur at intermediate decay

times in PPE-SO3-, (130) but is not observed in PPE-CO2-. If random walk of excitations along

the polymer backbone were to occur, as the wavelength increased the hopping rate would

decrease due to a lack of lower lying states that are available for the excitation to jump. The

following factors clearly eliminate the possibility of random exciton migration in this PPE-CO2:










Composition Changes: Interdiffusion

It has been an on-going goal to develop synthetic methods to produce highly luminescent

quantum confined materials with increased stability in the blue-green spectral region. Therefore,

some focus on synthesizing binary or core/shell materials has shifted to mixed ternary

heterostructures. This would allow for an extra degree of freedom (size and composition) to

achieve particular confinement characteristics, such as photoluminescence tunability in fewer

synthetic steps. If the cations in the shell were to exchange with the cations in the core

(interdiffusion), the optical properties can be changed significantly. More specifically, changes to

the energies of the valence and conduction bands, band gap and confinement potentials are to be

expected.

Temperature can influence the rates of chemical reactions and can be described using an

Arrhenius equation. In solid state, diffusion is thermally activated thus the diffusion coefficient,

D, can be determined using Eq. 2-4.(80)


D = D,e (2-4)

E4 represents the activation energy and Do is the diffusion coefficient when the temperature is

considered to be infinite. (80) Since temperature influences diffusion in solid state materials it is

important to determine the optimal conditions that will produce the desired ternary

heterostructure when alloying a core/shell material. This poses a problem for II-VI band gap

materials since the experimental values for D, and E4 for interdiffusion are limited.

Several groups have determined experimental values for diffusion lengths in bulk and

quantum well structures. For example, Martin (81) investigated the diffusion lengths for Cd

diffusion into ZnSe that were annealed for 1 hour at temperatures in the range of 300 to 550oC

and concluded that the optical properties, such as photoluminescence, should exhibit changes at











LIST OF FIGURES


Figure page

1-1. Jablonski diagram. .............. ...............17....

1-2. Generalized diagram for spectral overlap of donor emission and acceptor absorption
and the energy transfer between resonant transitions of donor and acceptor. Adapted
from B. Valeur.(2)............... ...............2

1-3. Signals in transient absorption measurements ................. ...............26..............

2-1. The band theory of solids .............. ...............28....

2-2. The nanocrystal band gap size dependence ................. ...............30..............

2-3. The absorption spectra of TOPO/TOP passivated CdSe nanocrystals with radii from 1.2
to 4. 1 nm .............. ...............33....

2-4. Nanorod with each axi s labeled ................. ...............35..............

2-5. Electronic potential step of valence and conduction bands ................. ................. ...._3 8

2-6. Time-resolved photoluminescence ................. ...............44........... ....

2-7. Powder X-ray diffraction patterns of CdSe nanorods, CdSe/ZnSe core/shell nanorods,
and ZnCdSe alloyed nanorods .............. ...............47....

2-8. High resolution-transmission electron microscopy image and histogram of size
distribution of ZnCdSe nanorods. ..........._......_ .....__ ...........4

2-9. Raman spectra of LO phonon mode of CdSe nanorods and CdSe/ZnSe core/shell
nanorods ........... __..... ._ ...............50....

2-10. Raman LO phonon spectra of ZnCdSe nanorods after annealing at 2700C.............._._......5 1

2-11. The UV-Vis absorption spectra CdSe, CdSe/ZnSe core-shell and ZnCdSe nanorods......52

2-12. The photoluminescence spectra of CdSe/ZnSe core/shell and CdSe nanorods...................53

2-13. The photoluminescence spectra of CdSe/ZnSe core/shell and ZnCdSe nanorods .............55

2-14. The broad band photoluminesce ................. ...............57........... ..

2-15. The CdSe/ZnSe core/shell photoluminesence .............. ...............59....

2-16. The time-resolved photoluminescence decay curves .............. ...............59....









nanocrystals.(16, 1 7) In the previous chapter, the effect of adding an inorganic shell and

subsequent alloying have on phonon spectra was presented. It is clear that new modes appear

after the addition of the ZnSe shell. Despite no significant changes between the bare and

passivated samples in the steady state absorption, the Raman data shows that the structure of the

system has changed; therefore it is necessary to investigate the system using time-resolved

methods. More specifically, the linear absorption spectra indicate that confinement is maintained

after addition of the ZnSe shell. Meanwhile, the quantum yield is increased from 0.6% to 15% in

bare and core/shell materials respectively.(42) Clearly, surface traps are reduced, prompting a

time-resolved investigation into the changes in the carrier relaxation due to the change in

electronic structure.

Figure 3-3 depicts the transient signal collected at 0 fs, 400 fs, 800 fs, 2.47 ps, 200 ps, and

575 ps for bare CdSe and passivated CdSe/ZnSe core/shell samples. Both samples are excited at

450 nm well above the band gap and the 1S and 1P absorption bands seen in Figure 2-12.

Multiple transitions dominated by state-filling are observed, leading to transient bands at the

energies of the allowed optical transitions. Exact determination of the electron-hole transitions

which give rise to different resonances need to be determined by comparison with the states

theoretically calculated by an effective mass theory. (72) In this work, we assign the transitions

based on works done by Klimov,(1 7) Efros,(25) and Guyot-Sionnest(129). Using their notation,

B1 and B2 are assigned to the photobleach of the 1S [1 S(e)-1S3/2(h)] and 2S [1S(e)- 2S1/2(h)]

states respectively while B3 corresponds to the bleach absorption of the 1P [1P(e)-1P3/2(h)] state.

Meanwhile, the Al band is assigned to the photoinduced absorption that grows in after high

energy excitations cool from the 1P to the l S. Within 400 fs (red line), the carriers are

distributed throughout the cascade of energy states. As the delay time increases from 400 fs to









Time-resolved anisotropy confirmed that this polymer, no matter the PRU size, is extremely

rigid and has long conjugation lengths.

t,2~ 30 to 40 ps

,z< 1.5 ps






375 nmn z \,350 to 450









Figure 4-22. Possible kinetic model for all PPE-CO2~ PRU chains




Full Text

PAGE 1

1 ULTRAFAST SPECTROSCOPY OF NOVEL MATERIALS By LINDSAY M. HARDISON 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 2007

PAGE 2

2 2007 Lindsay M. Hardison

PAGE 3

3 To my family

PAGE 4

4 ACKNOWLEDGMENTS As I reflect on the number of years that have led up to this moment of earning a Ph.D. in Physical Chemistry, I realize there are numer ous people to recognize and say thank you to because without their support and encouragement I w ould not have made it to this point.First and foremost, I thank the Lord because without His me rcy nothing is possible. I thank my advisor, Professor Valeria D. Kleiman for her guidance, patience and consiste nt motivation throughout my journey. I appreciate the effort and time sh e has put into helping me pursue this degree including letting me explore new possibilities and career building act ivities such as working for a summer at Corning, Inc as an intern. Valeria has continually believed in me and my work even when I did not and it is this type of support and enthusiasm that has enabled me to finish this project. I thank my supervisory committee members Prof essors Philip Brucat and Nico Omenetto for their guidance and thought provoking discussions. My gratitude goes to Dr. Kirk Schanze, Dr. Hui Jiang and Xiaoyoung Zhao for their CPE co llaboration. I apprecia te them providing the polymers and their willingness to help satisfy the need s of the project. I am also grateful for Dr. Paul Holloway and Dr. Hyeokjin Lee asking for our assistance in thei r nanorod project; it has been an experience I truly enjoyed. I express gratitude to the members, past a nd present, of the Kleiman Group. Thanks goes to Dr. Jrgen Mller for giving me a fundament al understanding of the transient absorption. I thank Dr. Evrim Atas for her ongoing friendshi p, Turkish cooking and being my upconversion mentor. There are no words that describe how mu ch I appreciate Daniel Kuroda. I thank him for not only his ability to answer all of my questions but also for his constant support and encouragement and of course, his BBQing skills. I also want to thank Cochuk, Aysun Altan,

PAGE 5

5 for helping me out in so many ways, especially in the last few week s, it has been greatly appreciated. I have dedicated this disserta tion to my family. To me, fami ly means more than just blood relatives, it is who you believe supports you and will be there for you always. To start, I want to thank Coach Dr. Nancy Bottge. She was such an in fluential person in my life and taught me that I must stick to the fight when hardest hit. Her dont quit attitude is one of the reasons for this success. She had not only been a mentor and a co ach but a friend that I could turn to and who taught me so many invaluable lessons. I extend ex treme gratitude to Chad Mair who has been beside me each step of my life in the past five years. I have been blessed with having such a good friend that I love so much that I consider him to be my brother. I can count on him for anything and know that without him, this achievement in my life may not have been possible. He has been the best work out partner, best friend and best colleague a girl could ask for. Jana Vanderloop, my best friend of ten, going on forever, years has also been a rock for me to lean on. I will always be sure to appreciate our inner ra ndomness because without it, life is too serious. I enjoy the fun we have on a daily basis, it k eeps me sane. From day one, Richard Farley and I have battled our way through the trials and tribulations of gr ad school. I thank him for his companionship, sense of humor and open mindedness. I thank Roxy Rory Lowry and Todd Prox for their friendship, laughs and their ability to give me different perspectives on all situations I run into in life. I also appreciate the encouragement and advice that I received from my friend Jim Reynolds. I thank Megan Meyer fo r her intense sarcasm because no matter what mood I am in, it always puts a huge smile on my face. My time here would not have been the same without the social ac tivities provided by all my friends in Gainesville. I am extremely gratef ul that I was a part of such a fun group that

PAGE 6

6 includes Sophie, Merve, Roxy, Richard, Rob, Neil, Eric, Megan, Meg etcTheir laughter and craziness will be greatly missed, especially during the fall at tailgating. Thanks to M.I.A and Whoever Shows Up, the two best in tramural softball teams in Univ ersity of Florida history for making me feel a little bit younge r. I have enjoyed playing fo r five years and will miss the teammates that have helped form our dynasty. Finally, I thank my parents Craig Hardison and Susan Keller for allowi ng me to make my own decisions so that I could become the inde pendent woman I am today. They have always believed that I could do anything I put my mind to. I thank my sist er Brynn, for terrorizing me as a child but growing up to become a wonderf ul young woman that I can call my friend.

PAGE 7

7 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES................................................................................................................ .......10 ABSTRACT....................................................................................................................... ............13 CHAPTER 1 INTRODUCTION................................................................................................................... ...15 Study Overview................................................................................................................. .....15 Photophysics Concepts.......................................................................................................... .16 Energy Transfer...............................................................................................................18 Radiative Energy Transfer...............................................................................................19 Non-radiative Energy Transfer........................................................................................19 Random Walk Migration (Intr achain Energy Transfer)..................................................20 Emission Measurements..................................................................................................22 Transient Absorption........................................................................................................... ...24 2 QUANTUM NANOPARTICLES..............................................................................................27 Overview....................................................................................................................... ..........27 Bulk vs. Quantum Semiconductors........................................................................................27 Size and Shape Dependence...................................................................................................33 Passivation.................................................................................................................... ..........36 Composition Changes: Interdiffusion.....................................................................................39 Experimental Methods: Nanorod Synthesi s and Composition Characterization...................41 Preparation of ZnCdSe Nanorods....................................................................................41 Steady State Instrumentation...........................................................................................42 Time-Resolved Photoluminescence Instrumentation......................................................43 Results and Discussion......................................................................................................... ..46 Synthesis of ZnCdSe Nanorods.......................................................................................46 Structure of ZnCdSe Nanorods.......................................................................................47 Effect of Alloying on the Phonon Spectra.......................................................................49 Photoluminescence and Absorption Properties...............................................................52 Time-Resolved Photoluminescence (TRPL)...................................................................55 Summary........................................................................................................................ .........62 3 QUANTUM PARTICLE ELECTRONIC STRUCTURE..........................................................63 Introduction................................................................................................................... ..........63 Experimental Methods: Transient Absorption........................................................................67

PAGE 8

8 Results........................................................................................................................ .............69 CdSe versus CdSe/ZnSe Core/Shell................................................................................69 Core/Shell Excitation Dependence..................................................................................72 Core/Shell versus Alloys.................................................................................................75 Discussion..................................................................................................................... ..........78 Summary........................................................................................................................ .........85 4 CONJUGATED POLYELECTROLYTES (CPES)...................................................................86 Introduction................................................................................................................... ..........86 Quenching PPE-CO2 -..............................................................................................................93 Experimental Methods........................................................................................................... .95 Synthesis of Variable Chain Lengths of PPE-CO2 -.........................................................95 Photophysical Methods...................................................................................................95 Photophysics of Variable Chain Length PPE-CO2 Polymers..............................................105 Steady State Characterization........................................................................................105 Time-Resolved Fluorescence................................................................................................112 Isotropic Upconversion.................................................................................................112 Time-Resolved Anisotropy...........................................................................................119 Potential Kinetic Model.................................................................................................121 Summary........................................................................................................................ .......122 5 CONCLUSIONS AND FUTURE WORK...............................................................................124 Nanoparticle Conclusions and Future Work.........................................................................124 Conclusions...................................................................................................................124 Outlook/Future Work....................................................................................................125 PPE-CO2 Conclusions and Future Work..............................................................................129 Conclusions...................................................................................................................129 Outlook/ Future Work (Hyperbranched PPE-CO2 -)......................................................130 LIST OF REFERENCES.............................................................................................................133 BIOGRAPHICAL SKETCH.......................................................................................................150

PAGE 9

9 LIST OF TABLES Table page 2-1 Comparison of and value of CdSe/ZnSe and ZnCdSe nanorods......................................61 4-1 Experimental conditions for wavelength dependence study..................................................97 4-2 Detection dependence decay times.......................................................................................113

PAGE 10

10 LIST OF FIGURES Figure page 1-1. Jablonski diagram........................................................................................................ ..........17 1-2. Generalized diagram for spectral overl ap of donor emission and acceptor absorption and the energy transfer between resonant transitions of donor and acceptor. Adapted from B. Valeur.( 2 )..............................................................................................................21 1-3. Signals in transien t absorption m easurements.......................................................................26 2-1. The band theory of solids................................................................................................ ......28 2-2. The nanocrystal band gap size dependence...........................................................................30 2-3. The absorption spectra of TOPO/TOP pa ssivated CdSe nanocrystals with radii from 1.2 to 4.1 nm...................................................................................................................... ......33 2-4. Nanorod with each axis labeled........................................................................................... ..35 2-5. Electronic potential step of valence and conduction bands...................................................38 2-6. Time-resolved photoluminescence........................................................................................44 2-7. Powder X-ray diffraction patterns of CdSe nanorods, CdSe/ZnSe core/shell nanorods, and ZnCdSe alloyed nanorods...........................................................................................47 2-8. High resolution-transmission electron microscopy image and histogram of size distribution of ZnCdSe nanorods.......................................................................................48 2-9. Raman spectra of LO phonon mode of CdSe nanorods and CdSe/ZnSe core/shell nanorods....................................................................................................................... ......50 2-10. Raman LO phonon spectra of ZnCdSe nanorods after annealing at 270C........................51 2-11. The UV-Vis absorption spectra CdSe, CdSe/ZnSe core-shell and ZnCdSe nanorods........52 2-12. The photoluminescence spectra of CdSe/ZnSe core/shell and CdSe nanorods...................53 2-13. The photoluminescence spectra of CdSe/ZnSe core/shell and ZnCdSe nanorods..............55 2-14. The broad band photoluminesce..........................................................................................57 2-15. The CdSe/ZnSe core/shell photoluminesence.....................................................................59 2-16. The time-resolved photoluminescence decay curves..........................................................59

PAGE 11

11 2-17. The ln[ln(Io/It)] versus ln(time) of Cd Se/ZnSe coreshell nanorods and ZnCdSe alloy nanorods....................................................................................................................... ......60 2-18. High resolution-transmission electron mi croscopy images of CdSe/ZnSe Core/Shell and ZnCdSe Nanorods.......................................................................................................61 3-1. Electronic structure in semiconductor nanoparticles.............................................................64 3-2. Transient absorption..................................................................................................... .........69 3-3. The broad band transient absorption spectra for CdSe and CdSe/ZnSe core/shell rods at various pump delay times..................................................................................................71 3-4. The kinetic traces correspond ing to the 1S, 1P and 2S bands...............................................73 3-5. The time-resolved excitation dependence collected for the core/shell sample.....................74 3-6. The broad band transient absorption spec tra for CdSe/ZnSe and ZnCdSe nanorods at various time delays............................................................................................................76 3-7. The 1S and 1P composition dependence...............................................................................77 3-8. The comparison of the 1S band for the Cd Se/ZnSe core/shell and 3 hr ZnCdSe alloy ........78 3-9. Valence and conduction band offsets for various materials. ( 75 ).........................................83 3-10. CdSe/ZnSe core/shell potential kinetic model.....................................................................83 3-11. ZnCdSe alloy potential kinetic model.................................................................................84 4-1. The intrachain energy transfer of ex citation to quencher molecule along polymer backbone....................................................................................................................... .....91 4-2. The PPE-CO2 polymer repeat unit........................................................................................91 4-3. The Stern-Volmer plot of 10 M 185 PRU PPE-CO2 -..........................................................94 4-4. Fluorescence up-conversion............................................................................................... ...97 4-5. Transition moments....................................................................................................... ........99 4-6. Photoselection........................................................................................................... ...........100 4-7. Berek polarization compensator..........................................................................................104 4-8. Berek compensator used as a half-wave plate.....................................................................104 4-9. The chain length absorption shift for PPE-CO2 in methanol..............................................106

PAGE 12

12 4-10. The emission spectra 10 M PPE-CO2 in methanol.........................................................106 4-11. The emission of 10 M 35 PRU PPE-CO2 ......................................................................108 4-12. The excitation spectra 10 M 8 PRU PPE-CO2 -...............................................................109 4-13. The excitation spectra 10 M 35 PRU PPE-CO2 in methanol.........................................110 4-14. The excitation spectra 10 M 35 PRU PPE-CO2 in water...............................................111 4-15. The excitation spectra 10 M 35 PRU PPE-CO2 in methanol with Ca2+.........................112 4-16. The time-resolved fluorescence decay of 8 PRU PPE-CO2 in methanol.........................114 4-17. The time-resolved fluorescence decay of 30 M PPE-CO2 with different polymer repeat units in methanol...................................................................................................116 4-18. The time-resolved fluorescence decay of 30 M PPE-CO2 (35 PRU) with and without Ca2+ .............................................................................................................................. ...117 4-19. The time-resolved fluorescence decay of 10 M PPE-CO2 (8 PRU) with and without Ca2+............................................................................................................................... ...118 4-20. The time-resolved fluorescence decay of 30 M PPE-CO2 35 PRU with different quenchers...................................................................................................................... ...119 4-21. The anisotropy of 8 PRU PPE-CO2 -..................................................................................120 4-22. Possible kinetic model for all PPE-CO2 PRU chains.......................................................123 5-1. The absorption spectra of the hyperbranched PPE-CO2 and the linear PPE-CO2 -.............131 5-2. The photoluminescence spectra of hyperbranched PPE-CO2 ............................................132

PAGE 13

13 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ULTRAFAST SPECTROSCOPY OF NOVEL MATERIALS By Lindsay M. Hardison December 2007 Chair: Valeria D. Kleiman Major: Chemistry My research focused on steady state and time -resolved photophysical characterization of a series of semiconductor nanoparticles and water soluble conjugated polyelectrolytes. Several studies have shown that the electronic structur e and relaxation dynamics in CdSe nanocrystals are not only size but are also shape and passi vation dependent; however, there is no detailed comparison of the photophysical properties of Zn CdSe particles with different relative amounts of Zn. This dissertation presents data collected for colloidal CdSe, CdSe/ZnSe and ZnCdSe nanoparticles with rod-like architectures synthesized and inves tigated in our labs to determine how size, shape, passivation and composition aff ect the quantum confinement and dynamics. In addition, a series of different polymer repeat uni t lengths of a linear conj ugated polyelectrolyte (CPE) with a carboxylate ionic side chain ha ve been synthesized and their photophysical properties have been explored. Spectral shifts and line broadening exhi bited within the Raman spectroscopy, UV-Vis spectroscopy and photoluminescen ce aided in determining th e extent of alloying and compositional disorder created during the alloying process. The photoluminescence quantum yield of ZnCdSe nanorods is higher than that from pristine CdSe nanor ods indicating a higher binding energy of the exciton. This effect is specul ated to be due to increased localization of the

PAGE 14

14 exciton as a result of fluctuat ions in the composition, ultimate ly resulting in increases in luminescence efficiencies. Moreover, time-resolved photoluminescence char acterized lifetimes of nanoparticles with similar shape but different composition. Emissi on of an inhomogeneous population distribution (different sizes, shapes or composition) leads to the simultaneous probing of particles with different decaying rates. A stretched exponential function, I(t)= A*exp[-(t/ )], can be used to describe these systems, where <1 corresponds to disper se populations. In the experiments presented here, the photolumi nescence data yields small values, independent of the emitted photon energy. Photoluminescence decay lifetime, of the samples increased with alloying time due to compositional disorder l eading to exciton localization. The dynamics of each nanorod was studied by absorption changes using ultrafast pumpprobe spectroscopy. An excitation wavelength dependence study has been conducted to gain insight into the intraband/interband relaxation in core/shell nanorods with small valence band offsets. Determination of the dynamics and mechan isms of these systems will be useful for the study of fundamental physics and light emitting applications such as LEDs, photovoltaic devices, lasing and fluorescence tagging. CPEs are soluble in polar solvents and thei r conformational properties can be tuned to enhance their emissive behavior for sensing a nd device applications. It was found that polymer concentration, solvent, aggregati on inducer and chain length, all affect the quenching efficiency; therefore, this dissertation examines energy tr ansfer mechanism responsible for this behavior using ultrafast upconversion. Upon excitation of the aggregates from energetically higher lying isolated chains, the fluorescence lifetimes resu lt in multi-exponential behavior due to the competition between the radiativ e and non-radiatve decay.

PAGE 15

15 CHAPTER 1 INTRODUCTION Study Overview The overall goal of my study was to investigat e light-matter interactions in a series of novel materials. The projects presented are an in teresting opportunity to explore the exciton dynamics and energy transfer processes in inor ganic semiconductor nanoparticles and organic conjugated polymers. The first ch apter will briefly discuss fluores cence principles and excitation energy transfer (interchain) and random walk (i ntrachain) energy transfer. This discussion is helpful to understand quenching of conjugated po lyelectrolytes. Also, th e signals that can be detected using femtosecond transient absorpti on (pump-probe) spectros copy are described in detail. The information provided is advantageous for the reader to understand the data presented concerning the excited state of the nanorods. Chapter 2 describes the motivation for the sy nthesis and characterizat ion of an array of CdSe/ZnSe core/shell and ZnCdSe alloyed nanorods. Background into the size, shape, passivation and composition dependence is pres ented. Moreover, synthetic steps, x-ray diffraction, high-resolution transmission electr on microscopy, Raman spectroscopy, steady state absorption, photoluminescence and time-resolved photoluminescence data are included. The ternary alloy composition is confirmed and th e qualitative trends based on compositional disorder are discussed. Femtosecond transient absorption (Chapter 3) was used to explore the excited state behavior of the same series of nanorods. A co mparison between CdSe and CdSe/ZnSe core/shell nanorods is made to show how passivation alte rs the exciton behavior. Also, excitation wavelength dependence is presented for the CdSe /ZnSe shell to elucidate the influence an interfacial state (determined in Chapter 2) has on the dynamics of the photo-generated exciton. In

PAGE 16

16 addition, the effects alloying has on the excited st ate of the ZnCdSe nanorods are discussed. Two models, one for the core/shell a nd one for the alloy, are propose d to describe the relaxation processes observed in ea ch of the experiments. The dynamics of the energy transfer from isolat ed to aggregated species in a series of different polymer repeat unit sizes of a conjugated polyelectrolyte, PPE-CO2 -, synthesized by Xiaoyoung Zhao in Dr. Schanzes lab are discus sed (Chapter 4). Anis otropy measurements confirm that this polymer is very rigid and th e conjugation length is long er than expected. The data is analyzed to extract the influence a ggregation has on the isol ated chain emission. Finally, Chapter 5 summarizes each project and states general conclusions drawn from the results collected for this dissert ation. Suggestions are made for pot ential applications for which semiconductor nanoparticles presented in this dissertation may be useful. Also, an additional molecule, similar to the PPE-CO2 -, is presented as the next step in a series of polymers to investigate for chemo-or bio sensors. Photophysics Concepts Spectroscopy methods, whether they be time-re solved or steady state, provide numerous ways to measure emission of materials that ar e intended to be used in a wide variety of applications including opto-and el ectronic, biomedical, and chemi cal research. This dissertation focuses on the photophysical propert ies of nanocrystals and energy transfer mechanisms that induce the amplified quenching capabilitie s of conjugated polyelectrolytes. It is important to know the multiple photophysic al processes that an excited chromophore can undergo between the absorption and emission of light. These processes are dictated by the probability that a transition from an initial stat e to a final state can occur. By using timedependent perturbation theory, Fermis Golden Rule for transitions between two states corresponds to a transi tion rate equal to: ( 1, 2 )

PAGE 17

17 S0 S1S2Sn A F IC T1T2 ISC IC PElectronic ground state IVR S0 S1S2Sn A F IC T1T2 ISC IC PElectronic ground state IVR 2 2'22TifkVH (1-1) where corresponds to the density of final states that are available to interact with the initial states via the perturbation, H This perturbation can alter the positions or motions of particles and restructure the initial state so that is looks li ke the final state. Thus, Fermis Golden Rule is simply a transition rate probability between an initial and final state which depends on the magnitude of a perturbation.( 1 ) The electronic transitions can be visualized along with the processes that can occur between these states in a general Jabl onski diagram seen in Figure 1-1. Figure 1-1. Jablonski diagra m. A = photon absorption; F = fluorescence (emission); P = phosphorescence; S = singlet stat e; T = triplet state; IC = internal conversion; ISC = intersystem crossing, IVR = in ternal vibrational relaxation. Adapted from B. Valeur. ( 2 ) Absorption of a photon by the ground state, S0, promotes an electr on to the vibrational levels of an upper singlet excited state, S1, S2, or higher, via a spin-cons erved, allowed transition. Subsequently, the excitation can be transferred to an isoenergetic vibrational manifold of a lower singlet excited state with the same spin multiplicity, for example S2 to S1. This process is aided

PAGE 18

18 by the overlap between the wavefunctions of the vi brational levels particip ating in the process. This radiationless passage is termed internal c onversion and occurs quite rapidly, usually within a few picoseconds or less after light absorption, which is significantl y faster than typical fluorescence lifetimes. Based on Kashas rule,( 2 ) the excitation will rapidly relax to the lowest vibrational level of the firs t singlet excited state, S1, via internal vibrational relaxation (IVR); therefore, the fluorescence emission, in most orga nic molecules, comes from the lowest excited vibrational level. Radiationless decay (which releases heat) and intersystem crossing (ISC) to a triplet excited st ate (resulting in phosphor escence) can also occur. ISC is a non-radiative transition that involves two electronic states that are equally energe tic but have different multiplicities. However, the magnitude of coupl ing between the orbital magnetic moment and spin magnetic moment (spin-orbit coupling) can be larg e enough so that this normally forbidden transition may occur.( 2 ) Energy Transfer One of the motivations behind the work presen ted in this dissertation is to identify the mechanisms responsible for the amplified quenc hing observed in conjugated polyelectrolytes. Aside from relaxation, the excited state of a ch romophore, D*, can relax to its ground state after transferring the photoexcited energy to an acceptor molecule, A, via a bimolecular process: D* + A D +A* This process strongly depends on two conditions: (1) the emission of the excited donor should overlap with the absorption of th e acceptor and (2) the natural lifetime of the excited donor must be slower than the energy transfer process. Once energy transfer has occurred, the photoexcited chromophore, A*, has the ability to play a part in photochemical reactions or display sensitized emission.( 1, 2 ) There are several different types of ener gy transfer mechanisms used to describe

PAGE 19

19 the diffusion of energy in molecules. In this dissertation only an overvie w of the radiative and non-radiative (interchain) energy transfer and rand om walk diffusion (intrachain) are discussed. Radiative Energy Transfer A two step process that involves emission of a photon from the excited state of the donor molecule followed by the same photon being ab sorbed by the ground state of the acceptor is called radiative energy transfer. Step 1: D* D + h Step 2: h + A A* This type of energy transfer is the least compli cated since it does not i nvolve the interaction of the donor and acceptor molecules. For this mechanis m to be effective, the quantum yield of the donor must be high in the spectra l region of the absorption of th e acceptor. To further enhance radiative energy transfer, it is beneficial to have a high conc entration and extinction coefficient of the acceptor in addition to a large spectral ove rlap between the emission of the excited donor and ground state absorption of th e acceptor. The emission spectra of a donor molecule that undergoes radiative transfer will experience a decrease in its fluorescence intensity in the spectrally overlapped region and can lead to re peated absorption and emission if the donor and acceptor molecules are identical (self-absorption/r eabsorption). If there is adequate absorption and emission overlap, the fluorescence lifetimes can increase.( 1, 2 ) An example of this process is shown in Figure 4-8, where we observe self-abs orption in a conjugated polyelectrolyte solution that is highly concentrated. Non-radiative Energy Transfer Non-radiative energy transfer occu rs in a single step and just as radiative energy transfer, depends on the spectral overlap between the donors emission and acceptors absorption spectra but relies more on their coupled re sonances (Figure 1-2). As seen in the Jablonski diagram in

PAGE 20

20 Figure 1-1, several vibronic transitions of one state or in this case, a donor molecule, can be isoenergetic to the corresponding transitions of the acceptor (D* D and A A*). In general, the non-radiative transfer rate is given by Eqn 1-1, (Fermis Gold Rule) where the density ( ) is not only related to the coupling of the initial and final states capable of a transition (determined by Frank-Condon factors) but also by the non-in homogeneously broadened spectral overlap, J of the donor emission,()DI ,and acceptor absorption,()A determined using Eq 1-2.( 1, 2 ) 0()()DAJId (1-2) This integral assumes that the relaxation within the excited state vibrational manifold is faster than the energy transfer proce ss and that energy transfer abid es by the Franck-Condon principle (vertical transition). As the number of resonant transitions between the donor and acceptor increases, the likelihood for a non-ra diative energy transfer process to occur increases since these transitions are proportional to the overlap integral (Figure 1-2).( 1, 2 ) Random Walk Migration (Intrachain Energy Transfer) In some cases, a quenching of the fluorescen ce occurs but can not be explained by a bimolecular energy transfer mechanism. Upon exc itation of a molecule, an excited state electron and ground state hole pair are creat ed, termed exciton. If the molecule consists of multiple segments that are equivalent in energy or are a cascade of energies (like a polymer with repeating chromophores), this exciton can diffuse from one segment to another while remaining bound. The exciton undergoes a mechanism that involv es a hopping from one segment to another within the same polymer. The random walk or intrachain energy transfer implies that the electron and hole move together, and will always be located within the same chromophore thus charge separation does not occur. Also, during th e course of the energy diffusion, the energy is

PAGE 21

21 not dissipated. This type of energy transfer between chromophores is difficult to measure primarily because directly collecting the fluor escence of an acceptor after exciting a donor chromophore of the same species with similar energies can be complicated. Figure 1-2. Generalized diagram for spectral overlap of donor emission and acceptor absorption and the energy transfer between resonant transitions of donor and acceptor. Adapted from B. Valeur.( 2 ) Eugene Rabinowitch, a biophysicist, likened the random walk to a steel ball being shot into a pinball machine where the ball bounces around with in the machine but eventu ally either falls to the bottom (fluorescence) or falls into a play hole, (trap).( 3 ) Scientifically speaking, the exciton can either fluoresce by recombining the el ectron and hole, can be dissipated by internal

PAGE 22

22 conversion in one of the chromophores or it can reach a trap. A trap (or "energy sink") is considered a lower energy state that is too deep for the exciton to overcome so from there it will recombine radiatively or non-radiatively but it will not undergo any more diffusion.( 4, 5 ) In the case of conjugated polymers, this trap could be due to a kink in the polymer chain, a defect on a chromophore, or a particularly low energy chro mophore. If trapping happens the exciton cannot reach its desired destination, an external quenche r, an analyte or a material that can separate charges, which would be detrimental for bioand chemo sensors and photovoltaics. Emission Measurements Conjugated polyelectrolytes are intended for use as fluorescence based-sensors and rely on changes of emission intensity and/or lifetime. Se nsors of this nature are some of the most common due to the ease of measur ement and low detection limits.( 6, 7 ) The fluorescence quantum yield ( F) is an important parameter that is de fined as the ratio of the numbers of emitted photons, Nemit, to the number of absorbed photons, Nabs. If all possible pathways are considered, the quantum yield is calculated as follows:( 2 ) 1 1emitradradrad F absallradICISCET f lNkk Nkkkkk (1-3) where krad is the rate constant of the fluorescence emission, kIC, kISC, and kET are the internal conversion, intersystem crossing and energy transfer rate consta nts, respectively. If the nonradiative decay ( knr) is the only competing process with the fluorescence emission, the quantum yield is given by: ( 2 ) rad F radnrk kk (1-4)

PAGE 23

23 Thus, the greater the non-radiative decay rate co nstant, the smaller the quantum yield, and vice versa. The time that it takes for th e excited state of a molecule to decay to 1/e of its initial value is the lifetime of the excite d state, which is given by: 1radnrkk (1-5) The fluorescence emission intensity, quantum yield, and lifetime can be negatively affected by numerous quenching processes incl uding collisions with heavy atoms, electron transfer, energy transfer, excimer formations, aggregate formation, and dynamic collisions.( 2 ) The quenching processes discussed above can be m easured but the results of these experiments can be hard to interpret. Instead, fluorescen ce anisotropy is employed to understand amplified quenching observed in conj ugated polyelectrolytes. Direct measurement of the random walk hopping of excitons is a difficult task since this process can compete with other energy transfer pr ocesses. Time-resolved anisotropy is type of measurement that measures the decay of polarized fluorescence, which gives a better understanding of the random in trachain energy migration in a material. The sensitivity to depolarizing the transition dipole moments between an absorbing and emitting molecule is directly related to the loss of anisotropy. Excita tion with light polarized in a particular direction will only excite molecules with the same orient ation. For example, a vertical excitation will preferentially excite molecules with vertical transition dipole moments. Anisotropy values will not change if as the exciton migrates there is no change in the dipole moments between the chains that absorb and emit. However, if these dipoles do change, as the exciton hops, it loses its original orientation and the fluorescence signal depolarizes.( 4, 5 )

PAGE 24

24 Transient Absorption Changes in population of differe nt energy states can be examined using femtosecond timeresolved pump-probe spectroscopy. A pump pulse excites the sample, cau sing a depopulation of the ground state. A broad-band probe is transmitted through the excited volume of the sample, monitoring the populations of vari ous excited states. The time-dep endency of the technique is introduced by varying the dela y time between the probe pulse and the excitation pulse. Simple absorption measurements can be accomplished by measuring the log of the ratio of the intensity of an incident beam that enters the solution, I0, and intensity of the beam exiting the solution, I 0loglog I AT I (1-6) where A is the absorbance and T is the transmittance. This ratio leads to the number of photons absorbed which is based on the sum of the absorption cross section, (cm2), of all the molecules in the path of the incident beam.( 2 ) 00 3 31 ln or log 10002303 (molecules/mol) (mol/L) since N(molecules/cm) 1000 (cm/L)a a aINclI Ncl II Nc (1-7) where, Na is Avogadros number, c is the sample concentration (M) and l is the optical path length (cm) within the sample. The molar absorption coefficient (extinction coefficient) is defined as:( 2 ) 2303aN Units = M-1 cm-1. (1-8) So, for a dilute solution, the absorption of light can be described using Beers law: or 2303aN A clAcl (1-9)

PAGE 25

25 In ultrafast experiments, shot-t o-shot laser fluctuations can hinder the detection of small transient signals. To overcome this limitation, a shot-to-shot normalization is utilized by having a second beam (reference) transmitted through the sa mple without overlapping with the pump. For transient absorption, the change in transmission ( T ) is defined as the transmission of the probe in the presence of the pump ( It, pump) minus the transmission of the probe in the absence of the pump ( It, no pump). The normalized change in transmission is given by: ,, ,,tpumptnopump referencereference pumpnopump tnopump nopump referenceII II TT T I TT I (1-10) this can be converted to cha nges in absorption using Eq 1-11. log1 T A T (1-11) There are multiple types of signals that can be observed in data collected using this technique. In Figure 1-3 the po ssible transitions associated wi th particular changes in the absorption spectra are shown. Bleach (1) The ground state of the sample will absorb photons creating a bleach of the ground state. When the pump is on, the deplet ed ground state will absorb less photons, and the measured A will be negative. If measuring transmissi on, the probe will transmit more, creating a positive change in T/T. This bleach signal will appear at transitions observed in the steady state absorption spectrum. A bleach signal appe ars instantaneously after the excitation. Photoinduced absorption (2) If the photoexcited state absorbs a photon, the beam probing that state will be attenuated, creating a positive change in absorption and a negative change in transmission.

PAGE 26

26 S01 t<0 t=0 A<0 Excitation t = 0 t > 0 A 2 probeS0 A > 0 probe A3 1 2 probe A<0 probe ASteady State Time ResolvedAbs Fluo S01 t<0 t=0 A<0 Excitation t = 0 t > 0 A 2 probeS0 A > 0 probe A3 1 2 probe A<0 probe ASteady State Time ResolvedAbs FluoStimulated emission (3) The pump will excite a population to some higher lying states and the probe will stimulate them to emit. If measuring the change in transmission, this signal will be positive since more light is apparently being transmitted through the sample. If measuring A, since the probe is stimula ting emission it corresponds to a ne gative absorption. A stimulated emission signal might not appear instantaneously since the relaxation of the initially excited state will occur before stimulated emission. Although stimulated emission can appear at wavelengths other than the steady state absorptio n spectrum, it can be difficult at times to distinguish between the bleach and stimulated em ission signals due to overlap of the absorption and emission positions. Figure 1-3. Signals in transient absorption measurements

PAGE 27

27 CHAPTER 2 QUANTUM NANOPARTICLES Overview Continuous advancements in the synthetic methods for the production of colloidal semiconductor nanoparticles of different size, shape, and composition have greatly improved their process-ability and functionality. The effect s that the physical features of these materials have on their corresponding photophysics have been the focus of nu merous scientific investigations in hopes that the inherent character istics of the new systems will make them useful for applications extending from basic fundamental physics studies ( 8 ), photovoltaics ( 9 ), optoelectronics ( 10-12 ) to photocatalysis.( 13-15 ) Many researchers have employed ultrafast time-resolved techniques to elucidate the dyna mics within various colloidal semiconductor quantum confined materials.( 16, 17 ) These fundamental investig ations are important for understanding and influencing the di rection and development of the area of nanoparticle science. New and precise synthetic methods have provided the ability to control the size, shape, and composition in order to manipulate the electronic or optical propert ies of nanoparticle materials. In 2004, we started collaborating with Profe ssor Paul Holloways research group to focus on the study of exciton dynamics in various semiconductor nanomaterial s. A member of Dr. Holloways group, Dr. Hyeokjin Lee, had synthesi zed CdSe, CdSe/ZnSe core/shell, ZnCdSe rods and ZnCdSe dots. Aside from CdSe, ( 13, 16-20 ) the literature lacked any information concerning the dynamics within these materials; it was an excellent opportunity fo r our lab to conduct cutting edge research in this area of materials science. Bulk vs. Quantum Semiconductors The primary bulk semiconductors used in solid -state electronic appl ications include materials such as silicon, gallium arsenide and cadmium selenide.( 21 ) The relationship between

PAGE 28

28 sp3+* Atomic Hybrid Orbitals Diatomic Orbitals Bonding Antibonding Mol. Orbitals HOMO LUMO Bulk Solid CB VB EgDensity of States Discrete States Discrete States Quantum Solids EgCB VB sp3+* Atomic Hybrid Orbitals Diatomic Orbitals Bonding Antibonding Mol. Orbitals HOMO LUMO Bulk Solid CB VB EgDensity of States Discrete States Discrete States Quantum Solids EgCB VB the created exciton and electroni c properties is of significant interest for the design and engineering of useful bulk and nanoscale semiconductor materials. Consider a summary of the band theory of solids presented in Figure 2-1. Figure 2-1. Band theory of solids Silicon has four sp3 hybridized atomic orbitals. Nei ghboring atoms contribute orbitals which combine to form highest occupied molecular orbitals (bonding orbitals, ) and lowest unoccupied molecular orbita ls (antibonding orbitals, *). The total number of occupied and unoccupied orbitals is equal to th e number of atomic orbitals present within the crystal. As more atoms are added, a density of orbital energies de velops reducing the spacing between the states in each band. This increase in density results in a continuum of ener gies separated by a gap. In a bulk solid, the highest occupied orbitals form the valence band and the lowest unoccupied orbitals form the conduction band. The minimum en ergy required to excite an electron from the top of the valence band to bottom of the conduction band is the band gap energy of the semiconductor (Eg).( 21 )

PAGE 29

29 The electronic and optical properties of a material are due to electron motion within molecular orbitals. The energy absorbed by an unbound electron (not co nfined) within the density of states in a bulk material is not quant ized. Therefore, the energy released by this electron is converted into kinetic energy. A semiconducto r can be photoexcited with a photon, exciting an electron from the valence band into the conduction band of the material, leaving a hole of opposite electric charge be hind, separated by distance cons isting of severa l atoms within the material. These distances are within the na nometer scale and are called the Bohr exciton radii. This radius, combined with a high dielectri c constant results in a small binding energy. The electron can be bound to the hole du e to Coulombic forces and if these interactions are strong enough, a Coulomb correlated, bound quasi-particl e called an exciton (electron-hole pair) is formed. If the size of the electronhole pair is approximately the same as the Bohr radius, and it is larger than the lattice spacing within the cr ystal, a Wannier-type ex citon is formed. This exciton can diffuse through the material until it is trapped, annihilated (under multi-excitation conditions) or recombined. If the wavefunctions of the electron and hole extend over a large number of atoms, the Coulombic attraction becomes negligible re sulting in unbound charge carriers which have slightly higher en ergies than the bound electron-hole pair. ( 13, 21, 22 ) If the size of bulk semiconductor is significan tly decreased to the poi nt where is it similar to the size of the Bohr exciton radius, then th e motion of the exciton wi ll become confined in multiple dimensions (quantum confinement) since it will have less room to move. The energy spacing between the various confined (bound) elect ron and hole states within the corresponding bands becomes quantized and the separation between these energy states will increase as the size of the particle (space) decreases due to stronger confinement.( 13, 17, 21, 22 ) In addition, the energy separation of the electron states is larger than the separation of the hole states since the

PAGE 30

30 EgEgEg VB VB VB CB CB CB EgEgEg VB VB VB CB CB CBhole has a larger reduced mass compared to the el ectron and the density of states in the valence band is larger than in the conduction band. Overa ll, due to quantum confinement, as the size of the semiconductor nanoparticle d ecreases, the band gap increases resulting in changes in the absorption and emission colors.( 13, 22, 23 ) (Figure 2-2) Figure 2-2. Nanocrystal band gap size dependence Just as in most quantum systems, there are multiple attempts to describe the electronic states mathematically. One particular way, us ed in quantum confinement of quantum dots, assumes that the quantum dots them selves are larger than the lattice constants of the crystal structure, which implies that the effective mass of the charge carriers remains unchanged despite the difference in size of the quantum dot compared to bulk. This is known as the effective mass approximation (EMA) and is utilized by most researchers in this area.( 22-25 ) Since, the effective masses of the carriers are considered to be cons tant; any modifications to the optical properties of the quantum dots observed will be due to quantum confinement. II-VI and III-V semiconductor quantum dots are considered to be in the strong confinement regime because their dimensions are generally larger than the lattice c onstant but less than or e qual to the Bohr radius

PAGE 31

31 size (11.2 nm for CdSe). Strong confinement of the electronic wavefunc tions results in an increase in the Coulomb interactions between th e electron and hole. If the materials are larger than this radius but smaller than their bulk count erparts, they are consid ered to be within the weak confinement regime.( 24 ) Nanowires, are a perfect example of materials in this regime due to the extension of the c-axis and smaller confinement potential in that direction. Considering EMA and the augmentation of C oulomb interactions between electrons and holes, excitons within nanometer sized semiconductors can be compared to motion of a particle in a 3-D box; as the size of th e box decreases, the kinetic energy and excitation energy increases. Figure 2-2 depicts how the quantum dot band gap varies as the size of the dot (box) changes. The enhanced exciton confinement w ithin smaller dots increases the amount of energy necessary to promote an electron to the conduction ba nd, overcoming the band gap barrier. The Schrdinger equation is used to consider the energy of the electronic states in quantum dots: ) ( ) ( r E r H (2-1) where H is the hydrogenic Hamiltonian for a Wannier-type exciton:( 21, 26 ) 222 2 **22eh ehehe H mmrr (2-2) where me and mh represent the electron and hole mass resp ectively, the distances between the electron and hold from the center of the quantum dot are er and hr and is the dielectric constant of the semiconductor. Since the center of mass and reduced mass motions cannot be separated into independent coordinates, analyti cal solutions for Eq. 2-1 and 2-2 are impossible. Therefore, different methods su ch as perturbation theory ( 26, 27 ) or a variational calculation ( 28 ) are utilized to describe energy in quantum systems resulting in the following equation:( 28 )

PAGE 32

32 222 min 2111.8 0.25 2Ryd ehe EE RmmR (2-3) Due to the quantum dot size dependence resulting in blue shifts as the dot size decreases, this equation evaluates the adju sted quantum dot band gap,minE, with respect to the quantum dot radius, R and the bulk exciton binding energy, *RydE.(26-28) As a result of quantum confinement propert ies exhibited by semiconductor nanoparticles, their electronic states are discrete and well-defi ned; therefore, their electronic states can be described in an atomic-like fashion. Three quant um numbers including spin are derived from the Schrdinger equation (Eqn 2-2) to evaluate the electronic states of the quantum dots. Using the effective mass model,(25) the electron and hole state notation is nLe and nLF, respectively where n is the principal quantum number (1, 2, 3, et c), and L is the envelope wave function angular momentum (S, P, D, etc) used to distingui sh energy states of the electrons and holes.(23) The hole total angular momentum, F [for a value of F, the state is (2 F+1)-fold degenerate], where F =2L+S and S represents spin.(17) In CdSe, the valence band is sixfold degenerate if the spin is considered since this band originates from p-atomic orbitals from within the selenium atoms.(25) The fine structure of the lowest exciton stat e within the valance band can be revealed (29) if the nanocrystal is non-spherical(30) or the crystal field effects(23) and exchange interactions(29) are considered.(17) It is well known that within CdSe quant um dots, the three lowest electron energy states are 1Se, 1Pe, and 1De and the first three hole states are 1S3/2, 1P3/2, 2S3/2.(25) Thus, the three lowest energy bands in ideal CdSe quantum dots are labeled as 1S [1S(e)-1S3/2(h)], 2S [1S(e)-2S3/2(h)] and 1P [1P(e)-1P3/2(h)]. It is possible to disrupt the ideal selection rules for spherical quantum dots, n=0, L=0, 2 and F=0, (31) via strong hole-state mixing (31) or by breaking the symmetry, which alters the degener acy and splitting of the excited states of the

PAGE 33

33 semiconductor quantum dots, ultimately changing the behavior of the excitons.(17) The absorption spectra of five colloidal CdSe nanocrystal s with different radii (1.2, 1.7, 2.3, 2.8, and 4.1 nm) is shown in Figure 2-3 (17) and illustrates not only the quantum dot band gap dependence but the features corres ponding to the optical transitions that arise from the coupled electron and hole electronic states previously discussed.(16-19) Figure 2-3. Absorption spectra of TOPO/TOP passi vated CdSe nanocrystals with radii from 1.2 to 4.1 nm.(17) Size and Shape Dependence Quantum confinement or the quantum size eff ect is a property that in recent years has revolutionized the semiconductor industry. This e ffect leads to unique electronic and optical properties making quantum dots differ from their bulk counterparts. II-VI semiconductor

PAGE 34

34 quantum dots have been the focus of several photophysical studies (18-20) stimulating interest in other types of quantum particle s. Fabrication of nearly spherical particles with various compositions in addition to synthesis of rod shaped (32) and even multifaceted tetrapods (33) has been achieved. Manufacturing such materials can be achieved by two different methods: 1. bottom-up and 2. top-down. The first method utilizes synthetic routes that adjust ratios of the chemicals needed to make the nanoparticl es with passivation or capping materials.(33-42) In the latter, the bulk semiconductor is cut down to scale using laser ablation-condensation or lithographic techniques a lthough these methods are extremely expensive.(43) The materials presented in this thesis have been prepared by a bottom-up approach in an attempt to synthesize better materials while en hancing their process-ability. Dependence on the sensitivity to size and shap e is important when considering the tunable optical properties of quant um nanoparticles. In partic ular, the size dependence and photoluminescence tunability in the visible regi on of CdSe quantum dots has been studied extensively.(37, 44-48) New methods for synthesis of r od-shaped CdSe nanoparticles have opened the door for shape-dependent a pplications such as polarized LEDs (49, 50). In particular, Alivisatos group synthesized quantum confin ed colloidal nanopart icles with rod-like architectures by using various surfactants that bind to different faces of the crystal.(51) For example, colloidal CdSe rod lengths can be vari ed from 5 nm up to 100 nm while maintaining a 2 to 10 nm diameter, which preserves lateral confinement of carriers in the nanocrystal. Alivisatos determined that the band gap depends ma inly on the width (a or b axis) and slightly on the length (c-axis) (Figure 2-4). (52, 53) However, a comparison of the dynamics within CdSe dots and nanorods has proven to be useful in und erstanding the electronic structure differences that occur when the c-axis is elongated.

PAGE 35

35 a b c Front Zoomed View Traps + a b c a b c Front Zoomed View Traps + Front Zoomed View Traps +El-Sayed et al. synthesized and compared CdSe rods with aspect ratios ~ 3 (length/width) and dots of 4.2 nm diameter.(53) TEM images show that the pa rticles are different although the steady state absorption spectra do not indicate sign ificant differences. Electron-hole dynamics measured by femtosecond pump-probe, although s till not completely un derstood, show quite different behavior for rods versus dots. This is confirmed by the increase in the number of bands in the deconvoluted absorption spectra of the quantum rods.(17, 18, 53) Moreover, they observed Figure 2-4. Drawing of a nanorod with each of th e axis labeled. The front zoomed view shows that the surface curvature is not smooth, leading to surface traps. a significant increase in the car rier relaxation time in the quantum rods compared to the quantum dots.(53) Nanodots have a higher order of symmetry, whic h is lost in the r ods. Extension of the c-axis results in a splitting of the degene rate level in the symmetric quantum dot (30, 53-60) and that energy level splitting could be one reason for El-Sayeds results. Due to the large surface-tovolume ratio at the surface in nanorods, electr on and holes have a high probability of being trapped by surface impurities. However, the quantu m dot curvature can create a larger number of localized surface trap states than the elongated nanorods,(53) which allow for the carriers to have free motion in the c-direction, reducing the probab ility of the carrier to be trapped as quickly. In some cases, the impurities present enable th e materials to be used in oxidation-reduction chemistry, more specifically photocatalysis, (15) photodegradation and detoxification of

PAGE 36

36 chemical and environmental pollutants.(61) For optical applications such as photovoltaics or LEDs, it is important that these surface traps do not contribute to the exci ton trapping within the material. Several groups have worked on develo ping passivation techniques that will enable enhancement of their photophysical characteristics without alteri ng their confinement behavior. Passivation Modification of semiconductor nanocrystal su rfaces plays an important role in their electronic and optical propertie s and has been the subject of extensive investigations. (34, 40, 6265) The dangling bonds present on the surface of the nanocrystals negatively influence the optical properties but passivation has been pr oven to improve various confinement properties such as high quantum efficiency and lumines cence stability. Due to a high surface-to-volume ratio, even pristine, bare CdSe quantum dots tend to result in low luminescent yields (0.6%) and poor stability.(66) The ratio leads to augmentati on of the electron and surface state wavefunction overlap which creates localized midgap surface state tr aps resulting in nonradiative decay and decreasing the overall phot oluminescence quantum yield (Figure 2-4).(67) However, in certain applications (68), in which the charge carrier-interface interaction is crucial, the high surface-to-volume ratio is beneficial. (14, 15, 69) If the surface of colloidal na noparticles is coated with an appropriate passivating agent, e.g., organic molecules, this competition may be sufficiently reduced to dramatically extend the band-edge lifetime and enhance th e luminescence efficiency. (37, 53, 64, 70) However, due to several drawbacks including imperfect surface passivation and exchange reactions causing photodegradation, organic coating is not sufficient for improving quantum yields.(16, 71) Using the diffusion-controlled colloidal gr owth method developed by Bawendi and coworkers, (37) CdSe quantum dots have been passivat ed with various shells, among these is ZnS,(39) which narrows the fluorescence emission and improves their efficiency. Epitaxial

PAGE 37

37 overgrowth of a higher band gap inorganic shell (F igure 2-5) creates a step to confine the exciton to the core.(39-41, 65) This increased confinement has been employed to enhance the quantum yield of the dots to over one order of magnitude and to increase its stability against surface oxidation. (34, 38, 72) Various examples using ZnS, CdS and ZnSe as shell layers include CdSe/ZnS (39, 40), CdSe/CdS (11), CdSe/ZnSe (73), CdS/ZnS (34) and InAs/ZnSe (74). Passivation using multiple shells (34, 70, 75) or onion like structures (76) has also been achieved. It has been observed that the absorption and emission of a nanocrystal that is passivated with a ZnS shell exhibits a shift to longer wavelengths by approximately 10 to 20 nm as compared to the unpassivated core.(39, 40, 66, 77) Dabbousi et al. explained this phenomenon by considering charge carriers in a spherical box. The observed shifts to lower energies result from the tunneling of the lighter electron wave f unction into the shell wh ile the hole remains in the core. If this happens, the exciton is delo calized in the particle resulting in decreased confinement and excited state ener gy. For the electron to be able to penetrate into the shell, it must be able to overcome the valence band offset (barrier height), the energy difference between the valance band of the core and the valence band of the shell, that is present between the core and shell. If this offset is small, the shifts towards lower energies can be large.(39, 66) For nanoparticles passivated with an inorganic material, a critical thickness is present that is dependent on the size of the core and latti ce mismatch between the core and shell. (34, 38, 72) This thickness influences the abili ty of the electron to tunnel to the su rface potentially resulting in little to no shifts in band gaps or confin ement potentials compared to unpassivated particles. However, it is possible to surpass this thickne ss allowing the electron to tunnel into the shell layer which can decrease the overall quantum yi eld and shift the absorption and emission

PAGE 38

38 spectrum to lower energies.(72, 78) This is an important factor in the analysis of our CdSe/ZnSe core/shell nanorods which will be discussed in Chapter 3. A) B) Figure 2-5. Electronic potentia l step of valence and conducti on bands, HOMO and LUMO levels of A) inorganic core and B) inorganic core/shell nanocrystals, both with surface attachment of organic molecules. Adapted from H. Lee.(66) Atom dislocations induced by inte rfacial strain as a result of the lattice mismatch between the core and shell can also have a negative e ffect on the luminescence quantum yield because they can behave as sites that cause non-radiativ e recombination. The defects that arise from the core/shell interface can be reduced resulting in higher quantum yields by either growing a nanocrystal that is comprised of one core and two shells, such as a CdSe core passivated with a CdS/ZnS shell/shell structure (34, 66, 79) or by photoannealing(34, 66) which reduce strain or diffuse defects to the surface, respectively. For example, irradiating CdSe/ZnS core/shell nanocrystals with UV light caused an increa se in the photoluminescence quantum yield by reducing the number of vacanci es present at the interface.(34, 66) organic molecule Eg(core) Eg(shell) band offset band offset Organic molecule organic molecule Eg(core) Eg(shell) band offset band offset organic molecule Eg(core) Eg(shell) band offset band offset Organic molecule

PAGE 39

39 Composition Changes: Interdiffusion It has been an on-going goal to develop synt hetic methods to produce highly luminescent quantum confined materials with increased stability in the blue-green spec tral region. Therefore, some focus on synthesizing binary or core/s hell materials has shifted to mixed ternary heterostructures. This would allow for an extra degree of freedom (size and composition) to achieve particular confinement characteristics, such as photoluminescence tunability in fewer synthetic steps. If the cations in the shell we re to exchange with the cations in the core (interdiffusion), the optical properties can be ch anged significantly. More specifically, changes to the energies of the valence and conduction bands, band gap and c onfinement potentials are to be expected. Temperature can influence the rates of chemi cal reactions and can be described using an Arrhenius equation. In solid state, diffusion is th ermally activated thus the diffusion coefficient, D, can be determined using Eq. 2-4.(80) 0AE RTDDe (2-4) EA represents the activation energy and D0 is the diffusion coefficient when the temperature is considered to be infinite. (80) Since temperature influences diffusi on in solid state materials it is important to determine the optimal conditi ons that will produce the desired ternary heterostructure when alloying a core/shell material. This poses a problem for II-VI band gap materials since the experimental values for 0D and EA for interdiffusion are limited. Several groups have determined experiment al values for diffusion lengths in bulk and quantum well structures. For example, Martin (81) investigated the diffusion lengths for Cd diffusion into ZnSe that were annealed for 1 hour at temperatures in the range of 300 to 550oC and concluded that the optical properties, such as photoluminescence, shou ld exhibit changes at

PAGE 40

40 temperatures as low as 350oC for a diffusion distance required for one monolayer of CdSe (~0.3 nm).(66) ZnSe/CdSe and ZnS/CdS superlattice struct ures investigated by Parbrook et al. exhibited diffusion at annealing te mperatures greater than 400 and 450oC rather than below 400oC.(66, 82) The negligible changes obs erved in the range of 340oC~400oC are further confirmed by investigations to determine diffus ion lengths for Cd in ZnSe/CdSe quantum wells by Rosenauer et al.(66, 83) and Stra burg et al.(66, 84) If the same experiments are conducted in nanocrystals the diffusion behavior is quite different compared to quantum wells or bulk materials; the alloying point (t he temperature at which the co re/shell nanocrystals begin to alloy (48)) in nanocrystals has been shown to occur at lower temperatures. For example, Zhong et al.(48, 66) observed alterations to the band gap and blue shifts in the photoluminescence in the range of 270~290oC in colloidal CdSe/ZnSe core/shell nanoparticles.(66) Additionally, at temperatures greater than 290oC, the alloying process can take as little as five minutes to complete.(48) There are several reasons that can account for the smaller alloying point temperature. Recall that as the size of the na noparticle decreases, the surface-to-volume ratio increases resulting in a large number of atoms be ing exposed to the surface or located in the interfacial region betwee n the core and shell; therefore, diffusion at lower temperatures compared to bulk materials can be expected. Diffu sion in colloidal nanopart icles can be also be aided by the desire for the surface atoms to mini mize their energy by reorganizing to reduce their surface area.(22, 66) The interface between the core and sh ell can have imperfections or defects that will also increase the diffusion rates. However, these imperfections can lead to a decrease in the photoluminescence quantum yields Finally, diffusion can be in fluenced by the crystal field strength of the nanoparticle. The crystal field stre ngth will scale with the size of the nanoparticle; therefore, bulk crystals have larger crystal fiel d strengths than quantum particles. Based on this

PAGE 41

41 principle, the interaction between atoms far fr om one another in nanocrystals is weak which will reduce the activation energy ultimately e nhancing the diffusion in these nanomaterials.(66, 85, 86) In our samples, we were able to achieve interd iffusion of Zn from a ZnSe shell into a CdSe core by alloying at a temperature (270~290oC) determined to be effective by Zhong et al.(48) Despite the fact that ZnCdSe quant um dots have been grown on ZnSe (87) and GaAs (88) substrates to determine how radius affects the fluorescence lifetime, little has been published on the compositional affects on the dynamics of colloidal ZnCdSe nanoparticles. From TEM images we observe that diffusion of Zn into the CdSe core does not change the shape of the rods si gnificantly and th e single phonon mode observed by Raman backscattering indicate that the ZnCdSe materials are comp lete quantum rod alloys, not composites. The differences that arise in both stea dy state absorption and photoluminescence in addition to time-resolved photoluminescence meas urements make investigating these ternary systems using ultrafast techniques extremely appea ling (Chapter 3). In th e current chapter, the synthesis, characteri zation, steady state photophysics a nd time-resolved photoluminescence measurements are described to begin to explai n the carrier relaxation within CdSe, CdSe/ZnSe core/shell and ZnCdSe alloy quantum rods. Experimental Methods: Nanorod Synthesi s and Composition Characterization The synthesis, XRD, TEM, Raman, and some optical characterization of each of the materials studied in this dissertation were carried out by Dr. Hyeokjin Lee in the Department of Materials Science at the University of Florida. Preparation of ZnCdSe Nanorods CdSe nanorods were synthesized us ing the method described by Peng. (46) In this method, CdO, trioctylphosphine oxide (TOP O) and tetradecylphosphonic acid (TDPA) were heated in a

PAGE 42

42 three-neck flask on a Schlenk line under a N2 atmosphere to 350oC 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 280oC while stirring, and selenium dissolved in trioctylphosphine (T OP) was injected quickly. After injection, the temperature of the mixture was kept at 250oC for the 30 min growth of CdSe nanorods, and then cooled to 180oC. (42) For shell growth, ZnO was dissolved in oleic acid (Zn-oleate) at 350oC and cooled to room temperature, and then TOP was added to prevent so lidification. In addition, Se was dissolved in TOP (Se-TOP). The Zn-oleate and Se-TOP soluti ons were mixed by stirring for ten minutes at room temperature, 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 room temperature for another ten minutes For alloying, the reaction vessel was heated with stirring to 270oC 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 cosolvents.(42) Steady State Instrumentation High-resolution transmission el ectron microscope (HR-TEM) images were collected using a JEOL 2010F microscope for imaging and direct determination of the av erage 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 di ffractometer and used for determination of both the crystal structure and size. Raman spectra were measured at 300K in the backsc attering geometry, using

PAGE 43

43 the 532 nm line from a Verdi 8 doubled Nd-YAG solid state laser in a Ramanor U-1000 JobinYvon Raman spectrometer.(42) Absorption spectra were collected with a Shimadzu UV-2401PC spectrophotometer. Photoluminescence was measured at room temp erature using nanorods suspended in toluene using a Fluorolog Tau 3 spectrofluorometer (Job in Yvon Spex instruments, S.A. Inc.). The photoluminescence quantum yield was determined using Rhodamine 6G organic dye standard. (42) Time-Resolved Photolumin escence Instrumentation Relaxation processes of colloidal nanocrys tals were explored using time-resolved photoluminescence. A commercial Ti-Sapphire (T i-Sa) laser system consisting of a Ti-Sa oscillator (Tsunami, Spectra-Physics) and subseque nt amplifier (Spitfire, Spectra-Physics) with a repetition rate of 1 kHz. The setup in our lab di rects the output of the amplifier into an optical parametric amplifier (OPA) to generate excitation pulses. For this experiment, since 400 nm is at the limit of both the signal and idler, we must use the second harmonic of the amplifier (800 nm) to achieve stable and high energy pulses. The residual 800 nm is directed to a horizontal BBO crystal (output of the Spitfire is polarized in the horizontal direction). A general schematic is provided in Figure 2-6. The sec ond harmonic (400 nm) is then fed through a prism compressor, resulting in pulse lengths less th an 100 fs (FWHM). The excitation beam is focused to a diameter of ~150 m at the sample position an d its energy was set to ~ 56 nJ yielding a fluence of 317 J/cm2. The optical density of each solution wa s 0.075/mm at 400 nm. Sample solutions of colloidal nanorods dissolved in toluene were place d 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. Broad band luminescence (grating range: 438 to 718 nm) from the sample was collected using a

PAGE 44

44 Pump from OPA, delay stage Lens Pump C C D from OPA, delay stage Lens Monochromator Pump from OPA, delay stage Lens Pump C C D from OPA, delay stage Lens Monochromator 2 in. lens and then focusing this into the en trance slit of a monochromator. Time-resolved photoluminescence spectra were r ecorded with an in tensified charge-c oupled device (ICCD) (Andor iStar coupled to a Shamrock 303i spectrog raph) with a 4 ns gate. The 4 ns collection window electronically scans to map the tempor al evolution of the photoluminescence. The exposure time at each time step is 0.5 seconds. Figure 2-6. Time-resolved photoluminescence The standard mechanical shutters commercially available are unable to gate at ultrafast speeds. Instead, the image intensifier found in th e ICCD acts not only as an amplifier but as an electronic gate, opening and closing on a nanosecond timescal e. There are three major components of an image intens ifier: photocathode, microcha nnel plate (MCP) and phosphor screen. The limitations of each of these determ ine how well the intensifier can perform. The incident image is first captured by the photocat hode which subsequently emits a photoelectron and is then pulled to the MCP by an electric fi eld. A high voltage is applied to the MCP causing the photoelectron to ri cochet along the channel walls creat ing an avalanche of secondary electrons which exit the MCP as an electron clou d. Typical intensifications can be as high as

PAGE 45

45 10,000. An additional voltage forces the electron cloud to hit the phosphor screen located at the front of the fiber optic exit window. The volta ges between the photocathode and MCP can be controlled in a manner so that the image intensifie r can be quickly turned off and on, effectively creating an electronic gate.(89, 90) As photons hit the surface of the CCD sensor, electrons are generated which are stored in individual pixels. The maximum number of el ectrons that one pixel can accumulate during integration is considered to be the full well ca pacity. A 16 bit analog to digital output converter which is capable of digitizing 65,536 levels (216) of light is used to read out the pixels.(89, 9193) The dynamic range is the number of steps or leve ls of light intensity that can be represented per bit.(93) Without using an image intensifier (g ain), the CCD dynamic range (maximum and minimum signal intensities that can be measured simultaneously) is defined as the full well capacity per pixel divided by the read noise.(89, 90, 92, 94) For this system, the full well capacity per pixel is 300,000 electrons and the read noise is 4 electrons resulting in a potential dynamic range of 75,000 to 1. This value exceeds the upper constraint of the digitizer so the dynamic range is instead limited to 65,000 to 1. Dynamic ranges can vary since the read noise depends on the read out rate. If the read out is fast, the read noise can be high and the dynamic range can be low or vice versa. A slower read ou t rate will reduce the read noise (high read noise will affect the quality of the image). Depending on the application, cameras that have a high well capacity and low read noise (high dynamic range) in addition to a large analog to digital conversion capability are optimal.(90, 92) The response of this camera is considered to be linear (1) within its full dynamic range.(89) The dynamic range of the CCD serves as th e base dynamic range of the ICCD camera system. As gain is added, the dynamic range is reduced.(90) For example, if the gain is set to 50

PAGE 46

46 (counts/electron) a multiplication factor is em ployed which reduces the dynamic range from 65,000 counts to 1310 counts (65,536 divided by 50) Gain can be advantageous if used in appropriate amounts. For instance, the read noise produced by the CCD section of the camera is no longer an issue. However, the dark current th at is created thermally by the photocathode prior to the amplification stage is still present and can also be amplified when gain is applied. Thus, even though high gains will lead to enhanced signa ls, the noise is also increased. It is also possible to have too little gain, which can sa crifice the well depth with no significant signal amplification. A balance between the system gain and dynamic range is ne cessary to achieve the best signal to noise ratio. Once ga in is applied, the response of the CCD remains linear within its dynamic range; however, the signal to noise equa tion is changed by the gain noise factor of 1.4.(89-92, 94) Results and Discussion Synthesis of ZnCdSe Nanorods Combinations of surfactants such as TDPA and TOPO are generally used to prepare nanocrystals since they have strong binding energies that ultimately raise the surface energies of a crystal face compared to another. (42, 95, 96) Previously, Zn-TDPA and Cd-TDPA in a TOPO solution were utilized in orde r to synthesize ZnCdSe nanorods. However, this method was not successful, which most likely resulted due to the different reactivity with Se-TOP leading to a lack of crystallite shape control.(42, 73, 97) It is also suggested that the temperature be higher and reaction time be longer in order to promot e a more thorough complexation of ZnO with TDPA. When synthesizing the ZnCdSe ternary hete rostructures, it is important to first prepare CdSe nanorods. Once the CdSe rod has been grow n, a Zn-Oleate and Se-TOP mixture was used to grow the shell overtop the core. Zn-oleate an d Se-TOP mixture was slowly added to prevent homogeneous nucleation. It is ex tremely important that the temp erature be controlled properly.

PAGE 47

47 2030405060 2030405060 2030405060 Counts (112) (110) (101) (102) (103) (002) (100) (c) (b) (a) 2Theta C) B) A)2030405060 2030405060 2030405060 Counts (112) (110) (101) (102) (103) (002) (100) (c) (b) (a) 2Theta C) B) A)When the temperature was 210oC, the emission blue shifted because the shell began alloying causing a blue shift in the emission. If the te mperature was kept too low, for example <170oC, only a small shell grew because the Zn-oleate comp lex reacted too slowly with TOP-Se to grow a shell, resulting in very weak emission. Finally diffusion of Zn from the shell to the core is instigated by raising the temperature slowly to 270oC to form ZnCdSe alloys.(42) Structure of ZnCdSe Nanorods X-Ray diffraction patterns for hexagonal Cd Se, CdSe/ZnSe, and ZnCdSe nanorods are shown in Figure 2-7. Figure 2-7. Powder X-ray diffraction patterns of A) CdSe nanorods, B) CdSe/ZnSe core/shell nanorods, and C) ZnCdSe alloyed nanorods. Adapted from H. Lee.(66). The crystal structure for this series of rods can be extracted from this experiment. In each of the materials measured, the (002) diffr action peak is not as broad as the (001) diffraction peak. The (002) peak is assigned to the pl ane that is perpendicular to th e extended c-axis in rod-shaped materials. The lattice spacings for CdSe, CdSe/ZnSe, and ZnCdSe were 7.01 6.94 and 6.77 respectively. These values are extremely inte resting since in CdSe/ZnS nanoparticles the ZnS

PAGE 48

48 01234567891011121314151617181920 0 10 20 30 40 50 60 70 80 90 100 Length Diameter Count01234567891011121314151617181920 0 10 20 30 40 50 60 70 80 90 100 nm 01234567891011121314151617181920 0 10 20 30 40 50 60 70 80 90 100 Length Diameter Count01234567891011121314151617181920 0 10 20 30 40 50 60 70 80 90 100 nmshell exhibits an 11% smaller lattice parameter causing the core to be compressed, whereas the lattice parameter in the c-axis for the core/she ll material was ~ 1% smaller compared to the CdSe. After interdiffusion of Zn into the core, the lattice parameter and the lattice mismatch strain are reduced due to a lat tice contraction. Also, after addition of Zn into the core, the diffraction peaks shifted to a larger 2 indicating a smaller interplanar spacing.(38, 42) A HR-TEM image of the ZnCdSe nanorods is shown in Figure 2-8. The diameter and lengths of the nanorods measured from such images has been included in the histogram. 20 nm scale bar 5 nm scale bar Figure 2-8. HR-TEM image and histogram of si ze distribution of ZnCdSe nanorods. Lattice fringe from a nanorod is shown in the lo wer right corner. Adapted from H. Lee.(66) From this graph, the average diameter is ~6 nm and the average length is ~13 nm resulting in an aspect ratio equal to ~ 2.1 nm for the alloys.(42, 66) When using XRD, the diffraction patterns can exhibit broadening effects due to particle size. Using the Debye-Scherrer formula, the average crystallite size in can be determined: coshklk D (2-5) Where k is a correction factor to account for particle shapes, and is the observed width at half the maximum peak intensity and is the Bragg angle. It must be noted that the observed width

PAGE 49

49 includes additional sources of broadening, arising from the experimental setup and instrumentation.(98, 99) Using Eq. 2-5, the particle sizes of ZnCdSe nanorods calculated were a diameter of 5.5 nm and a length of 11.8 nm resul ting in an aspect ratio of ~2.1 nm, which agree well with HR-TEM data shown in Figure 2-8.(42, 66) Effect of Alloying on the Phonon Spectra The compositional changes to the structure of a material that arise when adding a shell or alloying by diffusion through the dependence of th e phonon frequencies have been studied using Raman spectroscopy.(42, 100, 101) The Raman peaks detected from CdSe nanorods are shown in Figure 2-9 A. The peak at ~206 cm-1 is from the CdSe LO phonon (42, 101, 102) which is 4 cm-1 shifted compared to the bulk CdSe (210 cm-1) which is due to the quantum confinement of the optical phonons in the nanorods.(42, 100-102) A broad shoulder (~180cm-1) appears to the left of the main mode which arises from the non-spherical geometry of the CdSe nanorods. (42, 103, 104) The Raman peak for CdSe/ZnSe core/shell nanorods is shown in Figure 2-9 B. The original CdSe LO phonon mode is st ill detected with th e addition of the ZnSe shell mode at ~247 cm-1). A new interfacial ZnCdSe is also de tected and corresponds to a frequency ~235 cm-1. The small, unresolved Raman peaks on either side of the CdSe phonon mode can be assigned to isolated atom-impurity modes when Zn and Cd at oms interchange with one another (Zn in CdSe ~190 cm-1 and Cd in ZnSe ~218 cm-1). (42, 105) The effects of alloying time (1, 2 or 3 hrs at 270oC) on the Raman spectra are shown in Figure 2-10. After alloying, one mode is present at 223 cm-1, 228 cm-1 and 226 cm-1 (1, 2 and 3 hrs, respectively), which is similar to the interf acial layer observed in the core/shell material and the one phonon-mode behavior for bulk ZnCdSe. (42, 106, 107)

PAGE 50

50 A) B) Figure 2-9. Raman spectra of LO phonon mode of A) CdSe nanorods and B) CdSe/ZnSe core/shell nanorods. Adapted from H.Lee.(66) 150175200225250275 iZnCdSeZnSeCdSei(b) Raman IntensityRaman shift(cm-1)150200250 (a) IntensityRaman shift(cm-1)150175200225250275 iZnCdSeZnSeCdSei(b) Raman IntensityRaman shift(cm-1)150200250 (a) IntensityRaman shift(cm-1) A) B)150175200225250275 iZnCdSeZnSeCdSei(b) Raman IntensityRaman shift(cm-1)150200250 (a) IntensityRaman shift(cm-1)150175200225250275 iZnCdSeZnSeCdSei(b) Raman IntensityRaman shift(cm-1)150200250 (a) IntensityRaman shift(cm-1) 150175200225250275 iZnCdSeZnSeCdSei(b) Raman IntensityRaman shift(cm-1)150200250 (a) IntensityRaman shift(cm-1)150175200225250275 iZnCdSeZnSeCdSei(b) Raman IntensityRaman shift(cm-1)150200250 (a) IntensityRaman shift(cm-1) 150175200225250275 iZnCdSeZnSeCdSei(b) Raman IntensityRaman shift(cm-1)150200250 (a) IntensityRaman shift(cm-1)150175200225250275 iZnCdSeZnSeCdSei(b) Raman IntensityRaman shift(cm-1)150200250 (a) IntensityRaman shift(cm-1) A) B)

PAGE 51

51 A) B) C) A) A) A) B) C) Figure 2-10. Raman LO phonon spectra of ZnCdSe nanorods after annealing at 270C for A) 1, B) 2, or C) 3 hrs. Adapted from H. Lee.(66) Detection of only one mode that is only slig htly shifted compared to the bulk in these materials indicates that the inte rface between the CdSe core and Zn Se shell is no longer present, implying that the material is in fact an a lloy and not a composite. The narrow particle distribution and uniform compos ition observed in the XRD is confirmed from the relatively sharp single-mode peak. The broader peak observe d after only one hour of alloying compared to the 2 and 3 hour is primarily due to compositiona l disorder. As the alloying continues from one to two hours, the Zn continues to diffuse resulting in a 5 cm-1 shift of the Raman peak. After 3 hours of alloying this peak shifts back 2 cm-1 due to compositional disorder (42, 108, 109) and stress relaxation by thermal annealing(110).(42)

PAGE 52

52 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 Wavelength(c) (b) (a) Absorbance A) B) C) 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 Wavelength(c) (b) (a) Absorbance A) B) C)300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 Wavelength(c) (b) (a) Absorbance 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 Wavelength(c) (b) (a) Absorbance 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 Wavelength(c) (b) (a) Absorbance 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 300400500600700 0.00 0.05 0.10 0.15 0.20 0.25 Wavelength(c) (b) (a) Absorbance A) B) C) Photoluminescence and Absorption Properties Significant differences can be seen in the absorption spectra for CdSe core, CdSe/ZnSe core/shell and ZnCdSe alloyed nanorods. The absorption spectra of nanorods are shown in Figure 2-11. It presents two absorption peaks on top of a broad absorp tion. These absorption peaks correspond to confined states although they are not as sharp or as well resolved as peaks reported for CdSe quantum dots.(17) Due to the loss of symmetry in nanorods, confinement along the c-axis is not as strong as it can be in dots which results in a la rge distributio n of energy levels in the conduction and valence bands.(53, 111) In addition, the compositional disorder indicated from the Raman data will lead to broader features in the absorption and emission spectra. Therefore, it is difficult to determin e the exact energy spacing between the first and second absorption peaks. (42) Figure 2-11. UV-Vis absorption spectra of A) CdSe nanorods, B) CdSe/ZnSe core-shell nanorods, and C) ZnCdSe nanorods alloyed at 270C for 3hrs. Adapted from H. Lee.(66) For CdSe core and CdSe/ZnSe core/shell nanorods, the absorp tion edge is at 650 nm and 645 nm, respectively. A second peak is observed at ~520 nm. These features correlate to optical

PAGE 53

53 450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 IntensityWavelength(nm)(b) (a) A) B)450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 IntensityWavelength(nm)(b) (a) 450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 IntensityWavelength(nm)(b) (a) 450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 IntensityWavelength(nm)(b) (a) 450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 IntensityWavelength(nm)(b) (a) A) B)transitions involving the electr on and hole quantized states.(22) Direct assignment of these peaks is difficult and needs to be evaluate d using the effective mass approximation. The photoluminescence spectra from CdSe core and Cd Se/ZnSe core/shell nanorods, with peaks at 642 nm and 638 nm respectively, are shown in Figure 2-12. With the addition of the ZnSe shell, CdSe photoluminescence quantum yields increased from 0.6% to 15% due to passivation of nonradiative surface states.(42) Increasing the shell thickness up to a critical thickness of an inorganic shell with a higher bandgap has b een shown to increase the photoluminescence quantum yield in rods.(34, 38, 72) Defects at the core/shell in terface due to lattice strain relaxation from shells thicker than the critical value will actually decrease the quantum yield.(39, 42, 65, 72) Figure 2-12. Photoluminescence spectra of A) CdSe/ZnSe core/shell nanorods and B) CdSe nanorods. Adapted from H. Lee.(66) It is intriguing that a 4 nm blue shift from core to core-shell emission occurs because Mokari and Banin (78) have reported a ~10 nm red shift for CdSe/ZnS core/shell quantum rods. They attribute this shift to t unneling of the electron wave functi on into the ZnS shell delocalizing

PAGE 54

54 the electron, lowering the confinement energy and ultimately decreasing the energy of the exciton levels.(112) Based on the Raman data presented, the formation of interfacial ZnCdSe results in a decrease in size of the CdSe core (56) resulting in increased localization and a blue shifted emission. This is further supported from alloy formation re sulting in a blue shift of the photoluminescence peak.(42) The energies of the corresponding absorption features from alloyed ZnCdSe (3hrs at 270C) is considerably blue sh ifted to ~555 nm and ~465nm (Fi gure 2-11 C). These features originate from the states similar to those in th e core and core/shell nano rods but with a larger band gap due to the formation of ZnCdSe. Figur e 2-13 presents the photol uminescence spectra of the alloyed ZnCdSe samples. Upon annealing at 270C the photoluminescence spectra shift to higher energies. After one hour of annealing, the peak appears at 610 nm. Further annealing (2 and 3 hours) produces a much larger blue shift, 510 and 565 nm, respectively. This behavior is consistent with the variation in composition indi cated by the broad Raman peak (Figure 2-10). In addition to the energy shift, the alloys present changes in bandwidth and intensity. As alloying time increases, the width of the photoluminescence band is reduced. The change in intensity does not follow a trend, with the 3 hour alloyed sample presenting a sharp increase in photoluminescence intensity. Quantum yield measurements are ~8, 5 and 10% for 1, 2 and 3 hrs, respectively. These values are higher than CdSe rods (0.6%) but lower than the core/shell sample (15%). Composition disorder in ternary alloy na norods will lead to localization of excitons compared to binary samples.(42, 113) Such localization effect s are known to improve the photoluminescence efficiency by increasing the ov erlap integral of the electron and hole wavefunctions. On the contrary, the quantum yield values are lower compared to the core/shell due to the lack of surface passi vation on the ZnCdSe nanorods. This is consistent with the

PAGE 55

55 400450500550600650700750800 400450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 400450500550600650700750800 400450500550600650700750800 400450500550600650700750800 Wavelength(nm) Intensity(d) (c) (b) (a) A) B) C) D) 400450500550600650700750800 400450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 400450500550600650700750800 400450500550600650700750800 400450500550600650700750800 Wavelength(nm) Intensity(d) (c) (b) (a) A) B) C) D)400450500550600650700750800 400450500550600650700750800 0 1x1062x1063x1064x1065x1066x106 400450500550600650700750800 400450500550600650700750800 400450500550600650700750800 Wavelength(nm) Intensity(d) (c) (b) (a) A) B) C) D) quantum yield decreasing after 1 and further afte r 2 hours of annealing, since diffusion will be reducing the gradient in Zn (i.e. reducing the hi gh concentration at the surface and increasing the low concentration in the middle of the nanorods). However, annealing for 3 hours increased the quantum yield over that from samples annealed fo r 1 or 2 hours. This increased quantum yield and full-width-half maxi mum (FWHM) reduction can be attribut ed to annealing of crystalline defects and reduction of stress, c onsistent with the Raman data. Defects found in the crystal are known to act as traps, re ducing emission efficiency(114, 115). (42) Figure 2-13. Photoluminescence spectra from A) CdSe/ZnSe core/shell nanorods and ZnCdSe nanorods alloyed at 270oC for B) 1, C) 2, and D) 3 hrs. Adapted from H. Lee.(66) Time-Resolved Photoluminescence (TRPL) The linewidths of optical transitions can be inhomogeneously broadened due to effects that act differently on different radi ating or absorbing particles.(2) Emission from an inhomogeneous population (different sizes, shapes or composition) leads to the simultane ous probing of particles with different decaying rates. Monitoring ti me-resolved photoluminescence at different wavelengths not only identifies the states emittin g but also extracts their decay rates. The

PAGE 56

56 formation of ZnCdSe alloys is not perfectly un iform and contributes to broad, inhomogeneous photoluminescence spectra and a distribution of decay rates. Figure 2-14 shows the broad band luminescence spectra of the core/shell and alloyed nanorod samples as a function of time. This data was collected on a nanosecond time scale using a 4 nanosecond instrument response and 0.4 ns time step. The fast decays (< 4 ns) are not detectable due to this limitation. From this data we were able to extract time traces at the maximum wavelength to determine the corresponding decay rates of each sample. Figure 2-15 A shows the broad band luminescence of the core/shell sample at two different time steps, 4.4 ns (black line) and 20 ns (red line). As the si gnal decays at the maximum wavelengths, the photoluminescence values do not shift significantly but the broadening is slightly reduced. Broadening also occurs since measurements of the nanorods samples were carried out at room temperature. The time traces corresponding to three different wavelengths, 645, 670 and 630 nm (black, red and green lines, respectively) for th e core/shell sample are shown Figure 2-15 B. At early decay times, the emission at different wa velengths is not the same but ends up being identical after ~ 100 ns. Moving from the core/shell (A) to the alloy 3 hr (D) the band gap shifts to higher energies and the br oad band signal narrows. Figure 2-16 shows the log plot of the time-resolved photoluminescence decay curves for the nanorods samples. These decays are the nor malized kinetic traces at the wavelengths corresponding to the maximums of the photol uminescence. Several differences can be highlighted. The signal at decay times less than four nanoseconds have to be deconvoluted with the instrument response function. Since we are in terested in extracting a characteristic lifetime we do not consider this for analysis It is seen from this plot that the core/shell decay (black line) is much faster than the alloys. Also, the core /shell decay curve is almost a straight line,

PAGE 57

57 0510152025 620 640 660 680 D) C) B)nmnmtime (ns) time (ns)nm nmtime (ns)A)05101520 560 580 600 620 640 time (ns)0510152025 540 560 580 600 051015 540 560 580 600 A) C) D) B) 0510152025 620 640 660 680 D) C) B)nmnmtime (ns) time (ns)nm nmtime (ns)A)05101520 560 580 600 620 640 time (ns)0510152025 540 560 580 600 051015 540 560 580 600 0510152025 620 640 660 680 D) C) B)nmnmtime (ns) time (ns)nm nmtime (ns)A)05101520 560 580 600 620 640 time (ns)0510152025 540 560 580 600 051015 540 560 580 600 A) C) D) B)indicating a higher degree of hom ogeneity and a smaller distributi on of decay rates compared to the alloys. The alloys (green, red, and blue lines) do not show the same behavior as the core/shell; instead, they exhibi t two or more decay components (similar dynamics prior to 50 ns but deviate after that). The data collection for alloy 3 hr (blue line) is shorter due to the lack of detectable signal after 60 ns. Figure 2-14. Broad band Photoluminesce of A): Cd Se/ZnSe core/shell nanorods B) 1 hr ZnCdSe C) 2hr ZnCdSe D) 3hr ZnCdSe In dispersed systems such as polymers (116) or colloidal nanoparticles, it is easy to believe that the relaxation beha ves non-exponentially and that th e large distributions of local environments lead to variations of relaxation times. Multi-expone ntial functions are useful and are more commonly utilized for fitting decay curves; however, a model is generally proposed

PAGE 58

58 based on the number of exponentials and an extens ive number of parameters required to fit the decay curve resulting in exact de cay lifetimes. This can become ve ry complicated and can lead to incorrect assignments of the photophysical proce sses occurring within the material. Jones et al. found that they were able to f it their photoluminescence data colle cted from CdSe/ZnS core shell quantum dots decays with a multi-exponential functi on, implying that there is an existence of several discrete relaxation pathways, with individual lifetimes. However, they were not able to claim the exact number and identity of such pathways.(117) For simultaneous measurements of a large ensemble of relaxation times it is more advantageous to use a stretched exponential (nonexponential) function to evaluate the distribution of relaxation times in such dispersive systems.(43, 116) This type of equation encompasses bot h independent, single step processes in addition to sequential multi-step processes.(116) t I t Ioexp ) ( (2-6) Where is the characteristic lifetime and (0 1) is the dispersion exponent. Despite only extracting an average lifetime from a non-expone ntial; the function provides a phenomenological description that is consider ed purely empirical, fitting data with a minimum number of parameters. These parameters can vary dependi ng on the phenomenon of interest and external variables such as temperature.(43) For the limiting case of 1, we get the single exponential decay with the characteristic lifetime, For ideal, single quantum dots, we can expect =1. It should be mentioned that <1 results from superposition of many exponential decays and as approaches zero, the distribution of decay times in creases. This decay law can then be used to compare different samples qual itatively in terms of non-uniform ity or topological disorder.(42)

PAGE 59

59 Figure 2-15. CdSe/ZnSe Core/Shell Photoluminesen ce: A) Broad band spectra at 8.8 ns () and 25 ns ( ) and B) Kinetic tra ces for 645 (), 670 ( ) and 630 ( ) nm. Figure 2-16. TRPL decay curve of CdSe/ZnSe nanorod ( ), 1 hr alloy ( ), 2 hr alloy ( ), and 3hr alloy ( ) 0204060801001201400 1 2 3 4 Log PLTime (ns)600620640660680700 0.0 0.2 0.4 0.6 0.8 1.0 A) Norm PL(nm)B) 020406080100120140160 0.01 0.1 1 Norm Log PLTime (ns)

PAGE 60

60 012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d) A) B) C)D)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d) 012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d) A) B) C)D) A) C) D) B)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d) A) B) C)D)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d) 012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d) A) B) C)D) 012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d) A) B) C)D)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d) 012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d)012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 2hr0.00.51.01.52.02.53.03.54.04.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 3hr012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) CdSe/ZnSe Coreshell012345 -4 -3 -2 -1 0 Ln(LnIo/It)Ln(Time(ns)) ZnCdSe alloy 1hrT=300K = 0.75 = 173ns T=300K = 0.58 = 277ns T=300K = 0.48 = 501ns T=300K = 0.58 = 276ns(a) (b) (c) (d) A) B) C)D) A) C) D) B)Figure 2-17 shows a plot of the data for the nano rod samples in the form of an ln[ln(Io/It)] versus ln(time) plot and fitting using a linear function. This type of plot is useful to determine the extent of exponential behavior present in the material.(116) Starting with Eq. 2-6: () lnoItt I (2-7) 0lnlnln ()I t It (2-8) Figure 2-17. Equation ln[ln(Io/It) ] versus ln(time) of A) CdSe /ZnSe coreshell nanorods, B) ZnCdSe alloy nanorods 1hr, C) ZnCdSe a lloy nanorods 2hr, and D) ZnCdSe alloy nanorods 3hr. Adapted from H. Lee.(66) When the slope, equals 1, the line should be stra ight indicating an expo nential decay with a well-defined rate.(116) Obtained values are summarized in Table 2-1. The fitted of CdSe/ZnSe coreshell nanorod is ~0.75, which reflects high er degree of ordered crystals. Difference of

PAGE 61

61 CdSe/ZnSeCore/Shell Nanorods (Aspect Ratio ~ 2.5) Alloying ~280C 3 Hours ZnCdSeNanorods (Aspect Ratio ~ 2.1) 20nm CdSe/ZnSeCore/Shell Nanorods (Aspect Ratio ~ 2.5) Alloying ~280C 3 Hours ZnCdSeNanorods (Aspect Ratio ~ 2.1) 20nmvalues between 1 and ~0.75 might be mainly due to size distribution. CdSe/ZnSe quantum wells have demonstrated similar values.(42, 66, 118) Table 2-1 Comparison of and value of CdSe/ZnSe and ZnCdSe nanorods. (42) A comparison of the TEM images of the core/s hell and the 3 hr all oy are shown in Figure 2-18. The aspect ratio of the core/shell nanorods is 2.5 while for the alloy it is 2.1, indicating that the size distribution is not significantly change d during alloying process (Zn diffusion does not change the shape or size signifi cantly). Thus, the photophys ical changes observed are due to the annealing process altering the composition. Comp aring CdSe/ZnSe coreshell nanorods to alloy ZnCdSe nanorods, the values are significantly decrease d from 0.75 to 0.48~0.58 due to increased disorder as a result of spatial fluctuations of Zn concentrations after the annealing process. After three hours of annealing, the value did increase, which is consistent with the increased compositional homogeneity observed in the photoluminescence and Raman data.(42) Figure 2-18. TEM images of CdSe/ZnSe Core/Shell and ZnCdSe Nanorods Nanorods e m (nm) (ns) Core/Shell ZnCdSe Alloy 1hr ZnCdSe Alloy 2hr ZnCdSe Alloy 3hr 645 625 570 566 0.75 0.58 0.48 0.58 173 277 501 276

PAGE 62

62 Theoretical arguments predict that the radia tive lifetime of bound excitons increases with binding energy (114) and this is observed in our samples where the s increase with alloying time from 173 ns to 276 to ~501ns (Table 2-1). Th ese time constants imply that in order to fully characterize the long decay presen t in these samples, experiments on a much longer time scale should be conducted. Exciton binding energy is increased by exciton confinement which is obtained in this experiment by localization of the carrier wave function by changing the composition of the quantum rods.(119) Therefore, it is expected that the luminescence decay time ( ) in these alloy nanorods will increase due to increased localization of excitons, i.e. binding energies induced by co mpositional fluctuations.(42) Summary Green-yellow emitting ZnCdSe ternary alloy na norods with relatively high quantum yield are presented. The nanorods size and shape were characterized by XRD, TEM while the limited alloying in core/shell nanorods and composition disorder was detected by Raman spectroscopy. It has been shown that the quant um yield of ZnCdSe nanorods is a function of alloying time and is significantly higher compared to CdSe nanorods, but is still lower than the core/shell nanorods. The luminescent efficiency of these materials wa s 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 su rface. Time resolved emission provided information regarding the role of di ffused Zn. A stretched exponential function was used to describe these systems, where <1 corresponds to a distribution of decay rates. Comparing CdSe/ZnSe core/shell nanorods to ZnCdSe alloy nanorods we found a significant decrease in the value. Photolumines cence decay lifetime, of the samples increased with alloying time due to compositional disord er leading to exciton localization.(42)

PAGE 63

63 CHAPTER 3 QUANTUM PARTICLE ELECTRONIC STRUCTURE Introduction Investigations into the behavi ors of the electrons and holes in quantum nanoparticles have been of great interest for several years.(10, 18, 31, 120-123) As seen in the previous chapter, steady state spectroscopy does not give a clear understanding of the exciton behavior when comparing rods to dots. On the other hand, th ere are methods that can measure the temporal dynamics and the kinetics of photophysical proce sses. These methods are called time-resolved spectroscopy techniques,(124) and they are a powerful tool that can bridge fundamental parameters such as size, shape, composition a nd passivation to increasi ng quantum yields and stability, reducing photodegradation, lowering the cost of fabrication (d eposition and lithography methods are expensive), making the synthesis safer, and improving their process-ability. Whether in a conjugated molecule or se miconductor nanoparticle upon excitation, an electron and hole are created. Fi gure 3-1 shows a cartoon of mol ecular orbitals (MO) as linear combinations of atomic orbitals. In conjugate d molecules, the MOs near the band gap are linear combinations of the same type of atomic orbitals whereas in nanocrystals the MOs can be linear combinations of atomic orbitals from diffe rent atoms. When an electron is excited from the HOMO (valence band) to the LUMO (conduction band), a hole is left behind in the HOMO. In a CdSe nanocrystal, the HOMO has contribu tions from atomic orbitals from the Se2whereas the LUMO is a linear combination of atomic orbitals from Cd2+. Therefore, the created hole will be located within the anion MO s while the electron will occupy the cation MOs. This in addition to a high dielectric constant pr esent in semiconductor nanocrysta ls means that the electron and hole are correlated although at the same time the individual carrier s can behave, i.e. be excited, trapped or relax nonradiatively, to some extent, independently.(13)

PAGE 64

64 Cd1+ Se1-h+ eInterband Transition Electron Intraband Transition Hole Intraband Transition Cd2+ Se2MOs = ci MOs = ci Energy Cd1+ Se1-h+ eInterband Transition Electron Intraband Transition Hole Intraband Transition Cd2+ Se2MOs = ci MOs = ci Energy Figure 3-1. Electronic structur e in semiconductor nanoparticles. Adapted from M. El-Sayed.(13) After the initial excitation, the electron can only be further excited to higher states of the cation MOs while the hole can only be further ex cited to other anion MOs (intraband transitions). The recombination of the electron and hole from the conduction band to the valence band involves an interband transition which can be di rectly detected in the visible spectrum. The transient absorption signals detect ed in the visible region only re veal the behavior of this bound exciton. Intraband relaxation of either electron or hole dyna mics for strongly confined nanocrystals are detected independently in diffe rent spectral regions. Since the energy spacing between the levels within the cation MOs is mu ch larger than the energy spacing between the anion MOs the intraband excitation or relaxation of the hole intrab and transfers is detected at lower energies (infrared region) than the electron intraband transfers.(13) The relaxation processes from higher to lowe r excited states within nanocrystals are extremely intriguing and counterin tuitive. Unlike in bulk materi als where the cooling of the photo-generated carriers is rapid and occurs via lattice phonons through its conduction band continuum, carrier cooling in quant um particles must occur in disc rete steps based on the nature

PAGE 65

65 of the energy states of the materials.(125) Due to the quantization of nanocrystals, the spacing between the energy levels for the electrons is qui te large, reaching values much greater than longitudinal optical (LO) phonons found in bul k semiconductors (~ 25 meV). Cooling via phonons is possible; however, it requires a simu ltaneous emission of a substantial amount of phonons, which has low probability. Therefore, it was assumed for many years that the relaxation of these excitons should be inhibite d, due to this phonon bo ttleneck, resulting in nanosecond cooling times.(122, 126) The electron relaxation from the 1P-to-1S in CdSe nanocrystals occurs in the subpico second timescale (faster than even in bulk) thus bypassing this bottleneck.(10, 17, 122) Klimov was able to extr act population dynamics of the 1S and 1P states determining that a 1P to 1S relaxation rate of ~ 300 fs a nd a 1S buildup time depends on the confinement enhancement and decreases as the nanocrystal radius decreases.(18) This fast relaxation process is Auger in nature in that the Coulomb interaction between the electron and hole, which is increased due to quantum confinement, allows the electron to relax but transfer excess energy to the hole, scattering it deeper into the valence band.(10, 31, 122) This strong coupling between the electron and hole in quant um systems has allowed for predictions of efficient electron-hole energy tran sfer to occur within 500 fs (127) to 2 ps (128) though it is difficult to measure these values directly. Phonon assisted relaxation of th e hole is more probable due to smaller energy spacing within the valence band.(122) Using infrared transient absorption(129) and terahertz spectroscopy(122) several groups have attempted to correctly identify and conf irm the Auger relaxation mechanism where the electron transfers excess energy to the hole. The distinct photoabsorption features present in transient spectra is very useful in identifying different photo-excited species. However, broad band spectra are sometimes difficult to interp ret and assignment of various species becomes

PAGE 66

66 tedious and complicated. Moreover, luminescen ce up-conversion does provide for better time resolution then other time-resolv ed techniques which allow for de tection of ultrafast dynamic processes. Time-resolved luminescence is not us eful when trying to extract information about charge carrier mobility and density or even ca rrier cooling since the dynamics of each carrier must be measured separately. Although pump-pr obe in the infrared region has shown evidence for the Auger cooling mechanism,(18) Terahertz spectroscopy is the first method that allowed for direct measurement and quan tification of hot cooling.(122) Hendry et al. determined that the hole relaxation rate strongly depends on the amou nt of excess energy the electron provides. They were able to confirm the Auger cooling mechanism and claim that this decay occurs on a 1 ps timescale. (122) This dissertation utilizes the information ga thered by Hendry concerning the Auger cooling mechanism, focusing on the exciton dynamics, i.e. interband relaxation fo r bare, core/shell and ternary nanorods. From the experiments complete d by Hendry, and our data, it is concluded that the excess energy that the electr on transfers to the hole, in addition to the small valence band offset between the core and shell (0.07 eV), coul d be sufficient to cause the hole to tunnel into the ZnSe shell and extend the elect ron and hole recombination times. The ultrafast carrier dynamics in bare CdSe and core/shell CdSe/CdS/ZnS quantum rods using femtosecond pump-probe spectrosc opy has been conducted previously (72) to see how the interface between the core and shell affects the competition between photoinduced absorption and stimulated emission for lasing applications Faster relaxation in the core samples was observed due to surface traps. U pon passivation, the number of surf ace trap states decreases, at the same time the shell introduces interfacial stat es due to the lattice strain mismatch between core and shell.(72) To date, no studies have compared the ultrafast carrier dynamics in bare CdSe

PAGE 67

67 rods and CdSe/ZnSe core/shell rods. In this wo rk, we explore how addition of shell with only one inorganic material with a small valence band offset affects the photophysical properties and then compare this data to the exciton beha vior in ternary all oy heterostructures. Experimental Methods: Transient Absorption Relaxation processes of colloidal nanocrystals were explored using femtosecond transient absorption (TA). A commercial Ti -Sapphire (Ti-Sa) laser system consisting of a Ti-Sa oscillator (Tsunami, Spectra-Physics) and subsequent amplifie r (Spitfire, Spectra-Physics) with a repetition rate of 1 kHz was used to pump an optical pa rametric amplifier (OPA) to generate tunable excitation pulses. A portion of the amplifier output is split off to pump a 1 mm rotating CaF2 window to generate white light continuum probe with an effective bandwidth ranging from 310 to 750 nm. Prior to white-light ge neration, the probe polarization is tilted by 45 degrees with respect to the pump pulse using a thin-film polarizer. A detailed description is available elsewhere. (130) A general schematic is provided in Fi gure 3-2. The OPA idler/signal output is used to produce excitation pulses (pump) th rough harmonic generation (450, 575, 610, 630 and 650 nm). This beam is then fed through a prism compressor, resulting in pulse lengths less than 100 fs (FWHM). The excitation beam is focused to a diameter of ~150 m at the sample position and its energy was set to ~ 39 to 45 nJ yielding a fluence of 221 to 255 J/cm2. Low fluences are necessary to avoid multiple exc itations (biexcitons). From previous works, it is known that a signature of multiparticle interactions are decay rates that occur faster than 50 ps. Klimov also observed that the decay rates increased as the number of excitons per nanoparticle increased.(131) Experimentally, I verified the fact that multiple excitations per nanocrystal were not initially created vi a a power dependence study. Our data show (Figure 3-4: no fast decay

PAGE 68

68 observed in the kinetic traces prio r to 100 ps) that we were able to assume that for the power range utilized, multi-excitons are not initially generated. Prior to interaction at the sample, a fraction of the probe pulse is split off. This reference beam follows the same optical path as the probe but it probes only a sample volume that is not excited by the pump pulse. The pump pulses were modulated by an optical chopper at a frequency of 500 Hz and passed through a comput er-controlled optical de lay line to delay the probe arrival time relative to the excitation. The pump and probe beams were spatially overlapped at the sample. Probe and reference signals are collected in the presence and absence of the excitation pulse and the ratio: probeprobe referencereference pumponpumpoffII II (3-1) is recorded for each wavelength at every time step. A Glan-Thompson polarizer splits the transmitted signal, into its polar ization components, parallel (A| |) and perpendicular (A) with respect to the pump. The intensity at magic angle is calculated from the parallel and perpendicular components measured simultaneously: magic angle2 3 AA A (3-2) eliminating the influence of rotational times on the signal. Parallel and perpendicular transmitted probe and reference signals were focused into a spectrograph attached to a charge-coupled device (CCD) (Andor iStar coupled to Shamrock 303i spectrograph) for detection. Sample solutions for TA measurements were placed in a quartz cuvette with a 2 mm path length and continuously stirred to guarant ee excitation of a new sample vo lume with every laser shot. Changes in optical density (OD) were in the range of 1 to 50 mOD, and scans were repeated multiple times to achieve acceptable signal-to-n oise ratio. Each time step was averaged 250

PAGE 69

69 times per scan. When twenty scans were completed, the total number of laser shots per point was equal to 5000. Kinetic traces at particular wa velengths can be extracted from the full spectrum collected using the CCD camera. CCD ARRAY White Light Generation PS PP REFChopper Pump Reference Probe from OPA, delay stage Chopper Pump Reference Probe from OPA, delay stage C C D From Amplifier Monochromator CCD ARRAY White Light Generation PS PP REFChopper Pump Reference Probe from OPA, delay stage Chopper Pump Reference Probe from OPA, delay stage C C D C C D From Amplifier Monochromator Figure 3-2. Transient absorption schematic Results Insight can be gained by inve stigating the spectral evoluti on caused by population density changes in different energy levels using br oad band femtosecond time-resolved absorption.(17, 53) A continuum probe results in a collection of the entire transient spectra at each delay time in a single experiment. Therefore, photo-excited species can be detected and in principle, identified based on their characteristic transient absorption features. In even a simple system, assignment and interpretation of such photo-excited species can be difficult due to convoluted absorption features within the transient spectrum. CdSe versus CdSe/ZnSe Core/Shell CdSe nanorods were synthesized us ing the method described by Peng.(46) Raman spectroscopy is a useful tool for evaluating the structure and compositional homogeneity of

PAGE 70

70 nanocrystals.(16, 17) In the previous chapter, the eff ect of adding an inorganic shell and subsequent alloying have on phonon sp ectra was presented. It is clear that new modes appear after the addition of the ZnSe shell. Despite no significant changes between the bare and passivated samples in the steady state absorption, the Raman data shows that the structure of the system has changed; therefore it is necessary to investigate the system using time-resolved methods. More specifically, the linear absorption spectra indicate that confinement is maintained after addition of the ZnSe shell. Meanwhile, the qua ntum yield is increased from 0.6% to 15% in bare and core/shell materials respectively.(42) Clearly, surface traps are reduced, prompting a time-resolved investigation into the changes in the carrier relaxation due to the change in electronic structure. Figure 3-3 depicts the transient signal collect ed at 0 fs, 400 fs, 800 fs, 2.47 ps, 200 ps, and 575 ps for bare CdSe and passivated CdSe/ZnSe core /shell samples. Both samples are excited at 450 nm well above the band gap and the 1S and 1P absorption bands seen in Figure 2-12. Multiple transitions dominated by state-filling ar e observed, leading to transient bands at the energies of the allowed optical transitions. Exact determination of the electron-hole transitions which give rise to different resonances need to be determined by comparison with the states theoretically calculated by an effective mass theory. (72) In this work, we assign the transitions based on works done by Klimov,(17) Efros,(25) and Guyot-Sionnest(129). Using their notation, B1 and B2 are assigned to the photobleach of the 1S [1S(e)-1S3/2(h)] and 2S [1S(e)2S1/2(h)] states respectively while B3 corresponds to the bleach absorption of the 1P [1P(e)-1P3/2(h)] state. Meanwhile, the A1 band is assigned to the photo induced absorption that grows in after high energy excitations cool from the 1P to the 1S Within 400 fs (red li ne), the carriers are distributed throughout the cascade of energy states. As the delay time increases from 400 fs to

PAGE 71

71 500550600650700 -0.03 -0.02 -0.01 0.00 B3 B2 B) CdSe/ZnSe CS Rods Anm)B3-0.01 0.00 A1 B1 B2 B1 A) CdSe Rods A12.47 ps (blue line), carriers quickly relax from the higher energy states to the band gap state resulting in a corresponding 1P to 1S relaxation. Figure 3-3. Broad band transient absorp tion spectra for A) CdSe Rods and B) CdSe/ZnSe Core/Shell Rods at various pump delay times: 0 (), .400 ( ), .800 ( ), 2.47 ( ), 200 ( ), 575 ( ) ps. Extracting the kinetic information from the broad band spectrum enables the comparison of carrier relaxation trends at each optical tran sition (Figure 3-4). In both systems (CdSe and CdSe/ZnSe), the decay lifetimes corresponding to the 1S and 2S bands are identical (>200 ps) (black and red lines). The 1P decays rapidly a nd results in photoinduced absorption at longer delay times, the origins of which will be explained in further detail in the discussion. The rise of the PIA signal in bare CdSe rods is slower (>10 ps) than the passivated rods (< 2 ps). A dip is observed in the passivated sample for both the 1S and 2S bands, which results from overlap of

PAGE 72

72 multiple processes detected simultaneously. This effect is not observed in the core sample. For the core/shell sample, it is proposed that the hole migrates to the valence band of the ZnSe shell (valence band offset = 0.07 eV) due to the en ergy transfer during A uger relaxation of the electron resulting in longer ble ach decaying times. The insets of Figure 3-4 A and B compare the higher energy, 1P state negative absorption decay and photoinduced absorption to the 1S and 2S states bleach decays. It is interesting to note that the rise of the negative signal from the 2S and 1P states (black and green resp ectively) are identical but the 1P bleach decays rapidly (~ 1 ps) into a positive signal, matching the rise of the bl each of the lower 1S energy state (red line). This confirms a 1P to 1S relaxation process in both CdSe and CdSe/ZnSe. Core/Shell Excitation Dependence An excitation wavelength dependence study (450, 575, 610, 630, 650 nm) was conducted to elucidate the influence on the kinetic processe s of the ZnCdSe interfacial state previously detected by Raman spectroscopy. Figure 3-5 show s the steady state absorption spectrum, with the horizontal arrows signaling the excitation wa velength for each row of time-resolved spectra presented on the right two columns. The tran sient spectrum showed on top indicates the detection wavelength for each column. For example, the kinetic data on the top left plot presents the transient signal in the area of the 1P state after excitation at 650 nm whereas the middle right plot presents the transien t signal of the 1S state after excitation at 575 nm. After excitation at the same energies, the 1S band bleach rises with a ti me constant corresponding to the decay of the 1P bleach. In addition, the 1P detection shows photoinduced absorption present at a time delay greater than 1 ps due to the ZnCdSe interfacial state. For excitations greater than 600 nm (lower energies), the P band does not contribute to the dynamics; instead, the excitation simultaneously populates both the S band and trap states.

PAGE 73

73 Figure 3-4. Kinetic traces corresponding to the 1S ( ), 1P ( ) and 2S () bands for A) CdSe rods and B) CdSe/ZnSe rods. 0100200300400500 -1 0 1 -1012345-1 0 1 norm Atime (ps)CdSe/ZnSe Dip time (ps) B)0100200300400500 -1 0 1 -1012345-1 0 norm Atime (ps)CdSe RodsA) time (ps)

PAGE 74

74 0 0 2 0 0 0123 0 0123 0 A400 450 500 550 600 650 700 01 Excitation Wavelength 500550600650700 1A Detection Wavelength APS time (ps)Considering the kinetic traces in the first column (P detection) as the excitation energy is decreased the bleach decay faster and the onset of the photoinduced absorption occurs at earlier delay time. Eventually, (excitation at 650 nm) the bleach is no longer presen t, indicating that the P transition is not being accessed. As seen in the kinetic traces in the second column (S detection), as excitation energy is increased, the rise time of the negative absorptions are slower. When the sample is excited at 650 nm, the rise is instantaneous, while at 450 it is greater than 1 ps. The energy level associated with the ZnCdSe interfacial state is then considered to expand through the density of st ates between the CdSe and ZnSe band gaps. Figure 3-5. Time-resolved excitation dependence coll ected for the core/shell sample. Left graph: linear absorption spectra and indicates th e excitation wavelengths (450, 575 and 650 nm). Top graph: transient spectra at 2.47 ps. Kine tic traces in column 1 correspond to the 1P band while column 2 corre late with the 1S band behavior

PAGE 75

75 Core/Shell versus Alloys Prior to our work, limited information appeared in the literature involving the synthesis of colloidal ZnCdSe nanorods for use in optoelectronic devices. Green-yellow emitting ZnCdSe nanorods were prepared by diffusion of Zn into the CdSe core. For alloying, the reaction vessel containing CdSe/ZnSe nanorods was heated and stirred for up to 3 hrs. An aliquot was removed after heating for 1, 2 and 3 hrs, immediately c ooled and diluted with toluene to terminate the alloying process and then precipitated with me thanol/toluene co-solvents. The Raman data presented in Chapter 2 indicat e the enhancement of the ZnCdSe phonon mode but disappearance of the CdSe and ZnSe modes, confirming the alloyed nanorod composition.(42) Figure 3-6 compares the bleach spectra of the core/shell and ZnCdSe alloyed samples using the delay times of 0.400, 2.47, 150 and 575 ps. Note that this data is presented differently than in Figure 3-3. Each plot includes the four samples at one pa rticular delay time. For example, Plot 1 corresponds to the transients for each samp le at a probe delay time of 400 fs. This data indicate the occurrence of a transformation of the band gap, band structure and surface-trap states as function of alloying. At early time delays (0.400 ps Plot 1), the 1S band correspo nding to the 2 hr alloy (green line) is much more intense than all other samples and is significantly blue shifted compared the core/shell nanorods (black line). After 2.47 ps (P lot 2) the 1S band in all samples is maximized and the wavelength shift mentione d previously is much more evident. The 1 hr alloy remains very broad even after 575 ps (Plot 4). The overa ll signal obtained from the alloys is not as intense as the core/shell and th e 1 and 2 hr alloys are not as intense as the 3 hr alloy. Photoinduced absorption present in the core/shell materials does not appear in the alloys. This will be discussed in further detail later. At 150 ps (Plot 3), the alloy bleach has noticeably decayed; meanwhile the core/shell has not d ecayed significantly. Even after 575 ps, the

PAGE 76

76 core/shell decay is very small (compare bl ack line in Plots 2, 3 and 4) and photoinduced absorption is still present. For the ZnCdSe sample s there is no distinct band gap associated with CdSe or ZnSe. The only band present corresponds to the ternary composition; thus, it does not exhibit excited state absorption. As alloying time increases, the amount of Zn diffusing into the CdSe increases resulting in the band gap shifts observed in both steady state and transient measurements. Figure 3-6. Broad band transi ent absorption spectra for co re/shell(), Alloy 1 hr ( ), 2 hr ( ), 3 hr ( ) at various time delays. A comparison between the 1S and 1P bands of each of the samples is shown in Figure 3.7. All samples were excited at 450 nm but the dete ction wavelengths are not the same due to the band gap shifts induced by Zn diffusion. The tran sient spectrum at the top is shown as a guide indicating the detection wavelengt h for the 3 hr alloy. The firs t column of kinetic traces corresponds to the detection of the P band while the second column depicts the dynamics of the 500550600650700 -0.04 -0.03 -0.02 -0.01 0.00 PLOT 4 PLOT 3 PLOT 2 400 fs PLOT 1 2hr 1S C/S 1S 500550600650700 -0.04 -0.03 -0.02 -0.01 0.00 1S 1S 1S 2.47 ps 1S500550600650700 -0.04 -0.03 -0.02 -0.01 0.00 A (nm)150 ps500550600650700 -0.04 -0.03 -0.02 -0.01 0.00 (nm)575 ps

PAGE 77

77 475500525550575600 P P S A S Detection Wavelength Time (ps)0 0 0 0 0 0 01230 01230 3 HR 2 HR Alloy 1 HR Alloy Core/ShellS band. The amount of Zn diffusion into the CdSe increases from top to bottom as alloying time increases. When Zn is mixed in with the core, the material has more ZnSe character. Looking at the P band, when Zn diffusion occurs, the photoi nduced absorption is eliminated. Also, the rise of the P band bleach gets faster as more Zn is diffused into the CdSe core. The lack of photoinduced absorption in each of the alloys also confirms that the interf acial state present in the core/shell material is no longe r present. This is in agreement with the Raman data in Chapter 2. As seen in column 2, the rise of the bleach also gets faster as alloying time is increased. The S bleach decay of the alloys (Figure 3-8 red line) is significantly faster than the core/shell material (black line) due to surface trap s, since the inorganic passivati on layer is no longer present. Figure 3-7. The 1S and 1P composition dependence.

PAGE 78

78 Figure 3-8. Comparison of the 1S band for the Cd Se/ZnSe core/shell () a nd 3 hr ZnCdSe alloy ( ) Discussion In nanocrystals, optical transitions resulting in ground state absorpti on changes are due to state-filling effects whil e extremely fast transitions (<1 ps) re sult from Coulomb interactions, i.e., Stark Effect.(17, 72, 121) Red shifts are observed in CdSe quantum rods at longer delay times; they are identified as a convolution of the S-t ype states near the band gap stemming from the inherent size distribution present in colloidal nanoparticle samples.(72) The relaxation dynamics within these systems are str ongly influenced by ensemble dynamics collectively creating inhomogeneities and also multiple photoinduced processes leading to multi-exponential or nonexponential behaviors.(10) In analogy to bulk materials, cooling of hot electrons could occur via emitting LO phonons, and this mechanism would result in slow decay rates since the spacing of the intraband states is large in quantum syst ems resulting in a phonon bottleneck.(31, 111, 122) However, several studies (17, 111, 132) including our results, demonstrate a different behavior. High 0100200300400500-1.00 -0.75 -0.50 -0.25 0.00 Norm Atime (ps)

PAGE 79

79 energy relaxation in quantum systems occurs via Auger cooling in 2 ps (1P to 1S relaxation).(18, 53, 122, 128, 129) In each of the samples measured, relaxation was successfully observed as a corresponding decay of the 1P band and subsequent rise of the 1S band. Other possible cooling mechanisms include multi-phonon re laxation and polaron effects.(128, 133, 134) Under our experimental conditions, we create less than one electron-hole pair per nanorod, and thus those mechanisms are not plausible. In quantum dots, strong electron-hole Coul omb attractions favor energy transfer from the electron to the hole. (111, 122) Due to increased size dependence on carrier cooling, hot electrons tr ansfer kinetic energy to holes wh ich can quickly and efficiently undergo intraband cooling due to their relatively larger effective mass and smaller electronic energy level spacing.(18, 122) Enhancement of the Auger electr on-hole interaction would then bypass any phonon bottleneck.(122, 127-129, 131) This can not be conf irmed without directly measuring the electron-to-hole energy transfer which has been successfully observed by Hendry et al. using THz spectroscopy.(18, 122) Defect states can be resonant with high en ergy levels as observed by Rosenthal et al.(67, 122) The presence of these states results in an efficient trappi ng mechanism and faster bleach recoveries in core samples. These trap stat es are reduced by passivation with an inorganic material as in core/shell rods.(72) The shell reduces the number of surface states present in the bare rods preventing the exciton to sample th ese higher excited states resulting in higher photoluminescence quantum yields.(78) The core and core/shell systems both exhibit photoinduced absorption indicati ng the presence of an alternat e state contri buting to the excited state signal. The high surfacetovolume ratio within the bare rods creates additional electronic trap states leading to small photoinduced absorption f eatures at energies higher than the band gap but similar to the high energy 1P st ate. The onset of this photoinduced absorption

PAGE 80

80 is generally slow (>10 ps) and is related to the rate at which the carrier is trapped either at the surface or other defect sites. The trapped electr on can then relax to the ground state via an alternative radiative decay path way or decay non-radiatively.(67, 72) The actual rate s and overall dynamics are indistinguishable be cause of the inhomogeneity a nd ensemble averages of the samples; therefore, direct assignment of rates an d pathways is extremely difficult. On the other hand, photoinduced absorption within the core/shell rods arises from interface defects resulting in states that can act as traps and/or non-radi ative recombination sites caused by lattice strain relaxation introduced between the core and shell.(38, 65, 72) It appears that after the initial Auger cooling from the 1P to 1S state, some carrier populations sample the ZnCdSe interfacial layer resulting in photoinduced ab sorption. The signal is strong and lasts more than 500 ps (Figure 3-4). Compositional disorder (42, 108, 109) in ternary alloy nanor od structures leads to localization of excitons, (113) hence an increase in quantum yield for ZnCdSe versus CdSe nanoparticles. Localization eff ects increase the overlap integr al of the electron and hole wavefunctions improving the luminescence efficien cy of the material and decreasing the bleach lifetime for the ternary alloy samples compared to the binary CdSe materials. The lower ZnCdSe quantum yield versus the CdSe/ZnSe core/shell nanorods results from the lack of surface passivation and crystal defects within the ZnCdSe nanorods due to Zn diffusion into the CdSe core. Increasing the Zn character in the core caus es a blue shift in the spectrum as alloying time increases. Raman data presented in Figure 211 show that CdSe and ZnSe phonon modes are no longer present; the only mode obser ved is the ZnCdSe state. The lack of or small amount of photoinduced absorption confirms that the alloy nanorod composition is uniform and few interfacial traps are present. Since the inorganic shell disappears, surface traps reoccur and

PAGE 81

81 decrease the overall intensity (not as low as bare CdSe rods) and bleach lifetimes (recovery occurs faster than in the CdSe bare rods). In addition to surface state traps, crystal defects are known to act as non-radiative recombination cen ters, reducing the emission efficiency and enhancing the bleach recovery.(66, 114, 115) The band gap shifts, band narrowing and increase in the overall bleach amplitude can be attributed to stress relaxation by thermal annealing.(66, 110) This is consistent with the weaker, broadene d 1S band bleach signal observed after 1 hr and further after 2 hrs of anneali ng, since diffusion alters the di stribution of Zn throughout the nanorod, i.e. decreasing the amount of Zn presen t at the surface and in creasing the amount towards the middle. However, annealing for 3 hrs enhanced and narrowed the 1S band bleach compared to the samples annealed for 1 or 2 hrs. This increased change is attributed to annealing of crystalline defects and reduc tion of stress, consistent w ith the Raman data reported.(42) Tunneling of the electron wavefunction into th e ZnS shell has been reported in CdSe/ZnS core/shell structures by Mokari and Banin(38) resulting in a ~10 nm red shift. This tunneling led to a delocalization of the electron, lowering it s confinement energy and consequently the energy of the exciton levels.(39, 42) Raman data presented previously indicate formation of interfacial ZnCdSe in as-grown CdSe/ZnSe core/shell nano rods. This reaction would be expected to decrease the size of the CdSe core (42, 135) resulting in increased localization and a blue shift in emission. In our experiments, addition of the Zn Se shell does not alter the band gap significantly (only a 4 nm blue shift observed in photolum inescence). We have successfully engineered the material to create electron and hole wave functions that experien ce a confinement potential that localizes (Type-I localization) th e electron wave function within the CdSe core despite addition of a shell.(136) It has been shown that the rise of th e 2S and 1P bands ar e identical indicating that the hole is delocalized within the density of states located in the valence band of the CdSe

PAGE 82

82 core due to the small intraband spacing. Additiona lly, the 2S and 1S decays are identical. In fact, the dip observed for each of these bands in the core/shell sample after several ps is likely due to the hole tunneling into the interfacial ZnCdSe and/or ZnSe valence band due to the small valence band offset (the energy difference between the valence ba nd of the core and the valence band of the shell). The small valence band offs et between CdSe and ZnSe does not guarantee that the electron hole energy transfer does not cause the hole to transfer to the shell. If this were the case, it could account for the extended recovery rate observed in the core/shell versus core sample. However, we are unable to measure the m obility of the hole directly, so this is only a prediction. Figure 3-9 shows the valence and conduction band offsets of various bu lk materials. Most of the work presented in the literatu re deals with nanoparticles with ZnS (39, 137-140) or CdS (65) as the inorganic shell. CdS ha s been used to passivate CdSe(65) since it has a small lattice mismatch value (0.04 eV) compared to ZnSe (0. 07 eV) or ZnS (0.11 eV). In order to prevent penetration of the carriers into the shell or potential well barr ier, the photo-generated exciton created in the CdSe core should be confined by a material with valence and conduction potential wells that are comparable. Several studies have attempted to calculate to the first order approximation, a penetration length (L) that is proportional to:(75) 1/2()OLmV (3-5) where m represents the effective mass of the charge carrier (electron or hole) and VO is the band offset. These values can be seen in Figure 3-9. ( 141-145 ) Light electrons are more easily confined in heterostructures that have the same anion; therefore, the CdSe/ZnSe heterostructure is better balanced than the CdSe/CdS resulti ng in larger oscillator st rengths from enhanced electron-hole wavefunction overlap.( 75 )

PAGE 83

83 ZnSe CdSe 1Se 1Pe 1S3/2 1P3/2 2S1/2 0.07 eV V.B. Offset ~ 1ps Core/shell interface PIA Interband Relaxation h = 2.75 eV 0.4 eV Figure 3-9. Valence and conduction band offsets for various materials. ( 75 ) From the observations above we propose two mo dels associated with the exciton dynamics within the CdSe/ZnSe core/shell and ZnCdSe alloy nanorods. In Figure 3-10, high energy excitation results in very fast 1P to 1S rela xation times (~ 1ps). State filling within the conduction band of the CdSe occurs. Figure 3-10. CdSe/ZnSe core/shell potential kinetic model.

PAGE 84

84 0.4 eV 1Se 1Pe 1S3/2 1P3/2 2S1/2 < 1ps Interband Relaxation h = 2.75 eV Trap State Surface Trap From the evidence presented in this thesis, it is conceivable that during the Auger cooling of the hot electron from the 1P to 1S state, excess en ergy could cause the hole to become delocalized within the ZnCdSe interface or even in the ZnSe shell. The bleach recovery from the conduction band to ground state valence band is the longest in the core/shell materials due to passivation of surface traps and potentially due to the position of the hole wavefunction. Once the materials are alloyed, the ZnCdSe in Figure 3-11 becomes the only inorganic material present. Again, 1P to 1S relaxation is observed with high en ergy excitation followed by subsequent interband relaxati on from the conduction to valence band. However, based on the work completed by Rosenthal et al.,( 67 ) midgap surface states involving selenium dangling bonds are present due to the lack of inorganic passivation. An electron relaxing from the surface Se atoms to the valence band can immediatel y fill the vacancy left by the photogenerated electron contributing to the deep tr ap emission observed at 700 nm. Figure 3-11. ZnCdSe alloy pot ential kinetic model.

PAGE 85

85 Summary Transient absorption spectrosc opy was utilized to extract the exciton dynamics within binary CdSe/ZnSe core/shell a nd ternary ZnCdSe nanorods. A comparison between the exciton behavior in unpassivated CdSe core and CdSe/ZnSe core/shell materials, an excitation wavelength dependence for the core/shell nanorods and the influence allo ying has on the exciton behavior are all presented. For all samples, at high energy excitation, a 1P decay and subsequent 1S rise is observed corresponding to a 1P to 1S relaxation pro cess. Also, the introduction of a midgap state in the core/shell material leads to photoinduced absorption after the 1P bleach recovers. Upon low energy excitation, this midgap st ate is directly populated. Surface trap states reappear in the alloyed heterostructures (no passiv ation) leading to faster bleach recoveries then the core/shell materials.

PAGE 86

86 CHAPTER 4 CONJUGATED POLYELECTR OLYTES (CPES) Introduction -Conjugated polymers are an interesting cl ass of materials with unique physical characteristics that make them excellent can didates for various purpos es including lasers ( 146 ), LEDs ( 147 ), photovolatics ( 148 ), and transistors ( 149 ). To be useful for any application, a fundamental understanding of their photophysical prope rties is necessary in order to continue to improve their efficiency and efficacy. In recent years, conjugated polyel ectrolytes (CPEs) have been synthesized incorporating ionic sol ubilizing side groups enabling the polymer to be dissolved in water and other polar solvents while preserving the photophysical properties associated with the polymer backbone.( 6, 7 ) In an effort to reduce exposure of the non-ionic components to the environment, when CPEs are di ssolved in a polar solven t such as water they self-assemble into aggregates due to the inte raction between the charged functional groups and hydrophobic backbone. ( 150-154 ) This intraor intermolecular stacking of the polymer chain creates new, red shifted abso rption and emission peaks, ( 155 ) decreases the overall fluorescence quantum yield and competes with radiative em ission processes from the isolated chains.( 156 ) Aggregates can also form in concentrated polymer solutions w ith nonpolar solvents. (14, 20, 2628) Addition of a metal cation such as Ca2+ acts as a cross-linking agent and it has been shown to induce aggregation in methanol, improving th e amplified quenching properties of CPEs.( 155, 157-159 ) Aggregation is easily confused with othe r types of interchain interactions of -electrons in spatially close chromophores. It is extremely important to correctly define, understand and identify the different types of interchain spec ies that can be present within the conjugated polymer solution or film. Within the literature, there is a discussion on the proper identification

PAGE 87

87 of interchain species.( 159 ) In interchain interactions, -electron density is delocalized among numerous conjugated segments in different polymer chains. Depending on the physical conformation of the chains, it is possible that the two interacting species be located on the same chain. For example, if the polymer chains are ex tremely long, the conjugation segments from the same chain can interact spatially as a result of stacking due to backfolding. Shared electrons between two polymer chro mophores in their exci ted state that are next to each other create a species named termed excimer.( 159-162 ) When neutral excitons are shared by two or more adjacent chromophores in the ground and ex cited state, the intrach ain species that is formed is known as an aggregate. The aggregate formations that interact electronically will cause a significant change in the absorption spectra corresponding to an elongation of the electron delocalization resulti ng in lower energy peaks compared to isolated chains.( 159 ) In addition to aggregates, a polaron pair can be created after excitation resulting in an radical cation (hole polaron) in one ch romophore and a radical anion (e lectron polaron) on another. ( 159, 163, 164 ) A significant redshift in the emission sp ectra is a photophysical indicator that an excited interchain species is present within the conjugated sample due to delocalization of electrons creating a lower electr onic state compared to the isolated chains. Since this phenomenon occurs for each of the interchain inte ractions, it is hard to distinguish between the various types. Detection and identification is further complicated for room temperature fluorescence measurements due to the large numbers of non-radiative trap sites in conjugated polymers resulting in very lo w emission quantum yields. ( 159, 160, 163 ) Aggregates can be differentiated from the other spec ies because they are the only ones that show a weak redshift in the ground state detectable in the absorption spectrum. ( 159, 165, 166 ) This shift can be subtle, especially if the aggregat e absorption is symmetry of the tran sition is forbidden; therefore, the

PAGE 88

88 controversy between discriminating between aggreg ates and excited state interchain species is still ongoing.( 159 ) Based on the spectral signatures pr esent in our photophysical characterization, the species present in PPE-CO2 are considered to be aggregates. Correct identification of the types of interc hains species is extremely important when considering charge transport and light emission applications of c onjugated polymers. In order to fully understand a system and be able to make synthetic improvements it is necessary to characterize each of the species accordingly. Also, CPEs are opening the door to various biological applications but aggreg ation must be considered because it is an extremely important factor that influences the polymer quenching capabilities and ultimately their performance as chemoor biosensors.( 130, 155, 157, 160, 161 ) Zhao et al. has reported the synthesis and char acterization of a series of variable band gap poly(arylene ethynylene) (PAE) water soluble conjuga ted polyelectrolytes di ssolved in methanol, water and methanol/water mixtures.( 167 ) By only varying the anionic side group, they achieved band gap tunability within the vi sible region. Photophysical data co llected in their study correlate the CPE side chain structure to the extent of polymer aggregation when dissolved in each solvent. The work done for this thesis focuses on the role aggregation plays in the intraand intermolecular energy transport within vary ing polymer repeat units (PRU) of PPE-CO2 -. From previous research in this area it has been determ ined that the quenching efficiency increases as the amount of controlled aggregation increases.( 130, 150, 155 ) Several investigations, ( 158-160 ) including those done by Chen et al. ( 157 ) have alluded to the idea that quenching of a conjugated polymer emission is the fundamental property necessary to understand and characterize these materials to be useful for chemoand bio sensors. More specifically, anionic polymeric electrolytes can be efficiently quenched by cationic systems in

PAGE 89

89 solution. The quenching efficiency is desc ribed by the conventi onal Stern-Volmer relationship:( 2 ) 00 01[]1[]qSVI kQKQ I (4-1) where 0 ( I0) and ( I ) are the steady state fluorescence quantum yields (fluorescence intensities) in the absence and presence of the quencher molecule respectively, KSV is the SternVolmer constant, and [Q] is the quencher conc entration. A Stern-Volmer plot is the fluorescence intensity ratio ( I0/ I ) versus Q. This plot is expected to be linear with the slope equal to the Stern-Volmer constant. From this information, the quenching rate constant, kq, can be calculated if the exc ited state lifetime, 0 of the neat sample is known. From a time-resolved measurements point of view, the quenching mechanism is static if there is no change in when a quencher is added to the so lution. The following relationship is used if the lifetime does change:( 2 ) 0 01[]qkQ (4-2) This relationship enables one to determine if the fluorescence decay under dynamic quenching conditions in the presence of the quencher molecule and provides the value of kq. In CPEs, the quenching mechanism is both static and dynami c. In fact, the dynamic component does not necessarily arise from the diffusion of the quenche r in the solution but from the diffusion of the excitation within the polymer chain. ( 130, 168 ) It is well known that fluorescence within the visible spectrum from low concentrations of CPEs can be superlinearly quenched when pl aced in the presence of an oppositely charged electronor energy quencher molecule (superquenching ( 146 ); amplified quenching ( 167, 169 )).( 151, 152, 154, 157, 158, 162, 163, 165, 166, 170 ) Amplified quenching may lead to the

PAGE 90

90 development of more sensitive sensors but a complete explanation responsible for such high quenching efficiencies within CPEs has yet to be determined. This intriguing effect is not only due to ion-pairing between the polymer and quenchers ( 157, 163, 164, 170 ) as in typical SternVolmer kinetics, but also interand intrachain energy transport mechanisms. More specifically, the random walk diffusion of the excitation energy along the polymer backbone, ( 154, 157, 171, 172 ) energy transfer between the polymer and quencher ( 171 ) and energy transfer between the isolated polymer species and aggregated chai ns all contribute to such unique behavior.( 143, 146, 162, 165-167 ) Energy transfer is strongly dependent on the spectral overlap between the donor emission and acceptor absorption. The energy transfer is also very rapid. ( 168-171 ) If an aggregate inducer or quencher is added to the polymer solution, th e conformation can change resulting in a spatial redistribution of several chromophores. This enables the excitation located on the polymer backbone to easily migrate to the quencher located at a particular site lowe r in energy. Therefore, one quencher molecule can have the ability to reduce the emission from a large number of chromophores. ( 157, 173 ) The intrachain random walk model, which leads to excitation migration towards the quencher molecule, is strongly dependent on the conjugation, polymer chain lengths and transition dipol e orientations (Figure 4-1). Using time-resolved anisotropy fluorescence measurements, it is seen that after ex citation, the energy or exciton hops/migrates from shorter (high energy segments) to longer (l ow energy ones) and depolarizes along the way, reducing the anisotropy value. The exciton wi ll continue to funnel through the cascade of chromophores until it is either trapped or it reaches the lowest energy level where it can fluoresce or non-radiatively decay.

PAGE 91

91 O O CO2 -Na+ CO2 -Na+ nPPE-CO2 Several studies have investigated the influe nce aggregates have on the kinetics within water soluble conjugated polymer systems. ( 139, 152, 154, 173-179 ) For example, Fakis et al. has shown that energy transfer fr om isolated poly(fluorenevinyleneco -phenylenevinylene) (PFV-co-PV) ( 156 ) to aggregated chains is very rapid and efficient. They determined the isolated chain fluorescence, the aggregate emission and ener gy transfer contributions to the overall decay. In addition, the correlation between the con centration and energy transfer efficiency was thoroughly examined. A reduction in the concentrati on causes the energy transfer efficiency and energy transfer rates to decrease linearly.( 156 ) In this thesis, we investigate the influence chain length, solvent and metal cations have on the ultrafast emissi on of a carboxylated poly(phenylene) vinylene (PPE-CO2 -) shown in Figure 4-2 to dete rmine the excitation transport processes. The energy transfer mechanism between isolated and aggregated chains within the PPE-CO2 polymer is of particular interest. Figure 4-1. Intrachain energy tr ansfer of excitation to quencher molecule along polymer backbone. Figure 4-2. PPE-CO2 polymer repeat unit. PPE-CO2 in methanol (left) and water (right) Polymer Quencher

PAGE 92

92 Previously, a series of steady state, time-re solved, anisotropy measurements and numerical models were conducted with a similar CPE, PPE-SO3 -, to determine the rate and efficiency exciton migration has on fluorescence quenching.( 130 ) Using PPE-SO3 as a model polymer which exhibited both long range and random walk kinetics, we have designed experiments in which the results should indicate the type of energy transfer present within PPE-CO2 -. Timeresolved photoluminescence and time-resolved anisotropy measurements were employed to monitor the potential exc iton hopping that was previ ously observed in PPE-SO3 -, determine the rise time of the aggregate state emission a nd characterize the overall polymer decay. We studied very short and long polymer repeat unit (PRU) PPE-CO2 chains with the expectation that short chains woul d be less likely to aggregate. We find that even short chains (8 PRU) form solutions with both isolated and a ggregated chains. Even though the only difference between PPE-SO3 and PPE-CO2 is their ionic group, their p hotophysics are quite different. PPE-CO2 steady state photophysics correlate more with ladder-type (p oly-paraphenylene) (LPPP) polymers. ( 116, 166, 174-177 ) In most cases, conjugated polymer chains ar e not frozen in one conformation, instead they have a proclivity to twist and coil. A seri es of chromophores can be linked resulting in different degrees of -electron delocalization depending on the planarity of the conjugated segments. Even if there are slight twists or bends along the polymer backbone, it is possible for the conjugation to not completely break, resulting in larger delocalization lengths.( 130, 159 ) Just as in semiconductor nanoparticle s, a particle-in-a-box model (1-D for polymers) is used to explain the delocalizatio n of excitons along the polymer b ackbone. Conjugation lengths that are long tend to have lower transition energies and vice versa.( 159 ) Longer conjugation lengths can be due to polymer rigidity which can create small shifts between the absorption and

PAGE 93

93 emission maximums (Stokes shift). Both LPPP ( 116, 166, 174-177 ) and PPE-CO2 exhibit very small Stokes shifts alluding to their rigidity result ing in long conjugation lengths and therefore altering the dynamics and excit on hopping compared to PPE-SO3 -. Quenching PPE-CO2 As mentioned previously, addi tion of a cation such as calci um has been shown to induce various amounts of aggregation in CPE solutions. ( 178, 179 ) A perfect example of this effect is observed when poly (2-methoxy-5-propyloxy sulfonate phenylene vinylene) (MPS-PPV) fluorescence is quenched due to induced aggr egation created by the divalent cation, Ca2+.( 157 ) Dr. Hui Jiang, as a part of Dr. Kirk Schanze s lab in the Department of Chemistry at the University of Florida, has investigated the eff ects that additions of a divalent cation and an electron acceptor quencher, methyl violagen (MV2+) have on the quenching of PPE-CO2 -.( 155 ) Preliminary steady state absorption a nd quenching experiments of PPE-CO2 with MV2+ were conducted by Dr. Jiang and presented here to provide a better unders tanding of the CPE presented in this thesis. Figure 4-3 depicts the Stern-Volmer plots for PPE-CO2 fluorescence quenched by MV2+ in various solutions with increasing Ca2+ concentrations. As seen in previous works, ( 168, 180 ) the quenching efficiency varies depend ing whether the polymer is dissolv ed in methanol or water. In a water solution (closed squares) the quenching is extremely effi cient, requiring less than 1 M MV2+ to quench the fluorescence by 90%. More importa ntly, the superlineararity begins at very low quencher concentrations. On the other hand, if the polymer is dissolved in methanol (open squares), a well-define d induction region in which the slope ( KSV) is almost linear is observed, and the quenching efficiency does not become superl inear until much higher concentrations (~ > 3 M).( 155 )

PAGE 94

94 [MV2+]/ M Increasing [Ca2+]90% Quenching 40 30 20 10I0/I 0 123 45 [MV2+]/ M Increasing [Ca2+]90% Quenching 40 30 20 10I0/I [MV2+]/ M Increasing [Ca2+]90% Quenching 40 30 20 10I0/I [MV2+]/ M Increasing [Ca2+]90% Quenching [MV2+]/ M Increasing [Ca2+]90% Quenching [MV2+]/ M Increasing [Ca2+]90% Quenching [MV2+]/ M Increasing [Ca2+] [MV2+]/ M [MV2+]/ M Increasing [Ca2+]90% Quenching 40 30 20 10I0/I 0 123 45 0 123 45 Figure 4-3. Stern-Vo lmer plot of 10 M 185 PRU PPE-CO2 -. Quenching due to MV2+ in water ( ) and in methanol with 0 M ( ), 2.5 M ( ), 5.0 M ( ), 7.5 M ( ), or 10.0 M () CaCl2. ( 155 ) Addition of various amounts of calcium to meth anol solutions is also shown in Figure 43. As the concentration of calcium is increa sed, the induction regi on is reduced and the superlinear form occurs at smaller quencher amounts.( 155 ) A 1:1 stoichiometr ic ratio of 185PRU PPE-CO2 to calcium results in identical quen ching behavior as if the polymer was dissolved in water. In addition, the absorption was collected for the PPE-CO2 dissolved in methanol in the presence of various MV2+ concentrations. As the amount of quencher increases, the amount of aggregation increases as indicat ed by a redshift in the absorption spectra.( 155 ) This behavior follows suit with other CPE-quencher systems. ( 154, 157, 180 ) The mechanism for the observed amplified quenching can not be determ ined from steady state alone, hence the need for time-resolved measurements.

PAGE 95

95 Experimental Methods Synthesis of Variable Chain Lengths of PPE-CO2 Xiaoyong Zhao, a member of the Schanze group, is responsible for the synthesis and some of the steady state characterization of the various chain lengths of PPE-CO2 investigated within this dissertation. To polymerize a stoichiometric mixture of 2, 5-bis-(dodecyloxy-carbonylmethoxy)-1,4diiodobenzene and 1,4-di-ethynylbenzene a precu rsor route in which a Sonagashira coupling reaction is used to produce a poly(phenylene et hynylene) with a dodecyl ester protecting the carboxyl group. Gel permeation chromatography of the ester precursor polymers showed that the molecular weight (Mn) for the four polymer chain lengths i nvestigated in this dissertation are 5000, 24000, 74000 and 127000 gmol-1 corresponding to average degrees of polymerizations ( Xn) of 8, 35, 108, 185, respectively. The protecte d ester polymer precursor was then hydrolyzed with ( n -Bu)4NOH to provide for the water-soluble conjugated polyelectrolyte PPE-CO2 -. The final polymer product was purified using dialysis against DI water for 4 days. All of the polymers have polydispers ity indices of ~ 2.( 181 ) Photophysical Methods UV-Visible absorption spectra were recorded using a Lamb da 25 spectrophotometer form Perkin Elmer. Steady-state excitation and emi ssion spectra were obtai ned with a Fluorolog-3 spectrofluorometer from Jobin Yvon. A 1-cm sq uare quartz cuvette was used for all spectral measurements. Concentrations varied from 10 to 30 M and were dissolved in spectroscopic grade methanol. Time-resolved anisotropy and fluorescence dyna mics measurements were performed using a femtosecond upconversion apparatus. An optical parametric amplifier (OPA) pumped by a commercial Ti:Sa laser system consisting of a Ti :Sa oscillator (Spectra-P hysics, Tsunami) and a

PAGE 96

96 subsequent Ti:Sa amplifier (Spectra-Physics, Spitfi re) with a repetition rate of 1 kHz is used to produce excitation pulses. More sp ecifically, the output of the Ti:S a amplifier feeds an OPA, and the fourth harmonic of the signal is tuned to 375 nm. The excitati on beams is fed through a prism compressor, yielding an instrument response function of 225 fs. The instrument response function (IRF) is determined by the cross-corre lation of the excitati on and gate pulses. The upconversion setup used for these experi ments is described in detail elsewhere.( 182, 183 ) Briefly, a fraction of the 800 nm Ti:Sa amplifie r that is leftover from the OPA is used as a time delayed gate pulse (30 J/pulse). After excitation, the sample fluorescence is collected using a pair of off-axis parabolic mirrors and focused a nd spatially overlapped with the gate pulse in a nonlinear crystal (0.5 mm BBO), resulting in th e sum frequency of the two electromagnetic fields (Figure 4-4 A). The up-conversion signal has a photon frequency given by: s umgatefluo (4-3) This is also written as: 111 s umgatefluo (4-4) Detection wavelength is chosen by tuning the non linear crystal to a part icular angle. Table 4-1 includes a list of each detection wavelength and crystal angles used for the experiments discussed in this chapte r. The resultant signal is then focu sed into a monochromator, detected with a photomultiplier and the signal is gated with an integrating boxcar. When the gate pulse is temporally and spatially overlapped with the fl uorescence signal, the nonlinear crystal behaves as an optical gate. Therefore, scanning the gate pulse with respect to the excitation pulses enables

PAGE 97

97 this optical gate to integrate different windows of time. The fl uorescence signal is temporally mapped at these varying time delays (Figure 4-4 B). ( 182 ) Figure 4-4. Fluorescence Up-Convers ion Technique A) Illustration of the upconversion principle B) Up-converted fluorescence signal generate d in a nonlinear crystal only while the delayed gate pulse is present. ( 182 ) Table 4-1.Experimental conditions for wavelength dependence study Micrometer PMT Emission (nm)Monochromator (nm) Micrometer Position (filters) Position w spacer 430 450 475 500 515 533 550 590 279.5 287.5 296.5 307.0 313.5 320.0 325.0 340.0 4.23 2.94 1.70 0.45 7.50 6.19 4.95 3.79 2.45 1.86 1.60 1.46 Blind Blind Blind Blind/Vis Vis(1 UG-11) Vis(1 UG-11) Vis(2 UG-11) Vis(2 UG-11) Micrometer PMT (nm)Monochromator(nm) Micrometer Position (filters) Position w spacer 430 450 475 500 515 533 550 590 279.5 287.5 296.5 307.0 313.5 320.0 325.0 340.0 4.23 2.94 1.70 0.45 7.50 6.19 4.95 3.79 2.45 1.86 1.60 1.46 Blind Blind Blind Blind/Vis Vis(1 UG-11) Vis(1 UG-11) Vis(2 UG-11) Vis(2 UG-11) Detected Micrometer PMT Emission (nm)Monochromator (nm) Micrometer Position (filters) Position w spacer 430 450 475 500 515 533 550 590 279.5 287.5 296.5 307.0 313.5 320.0 325.0 340.0 4.23 2.94 1.70 0.45 7.50 6.19 4.95 3.79 2.45 1.86 1.60 1.46 Blind Blind Blind Blind/Vis Vis(1 UG-11) Vis(1 UG-11) Vis(2 UG-11) Vis(2 UG-11) Micrometer PMT (nm)Monochromator(nm) Micrometer Position (filters) Position w spacer 430 450 475 500 515 533 550 590 279.5 287.5 296.5 307.0 313.5 320.0 325.0 340.0 4.23 2.94 1.70 0.45 7.50 6.19 4.95 3.79 2.45 1.86 1.60 1.46 Blind Blind Blind Blind/Vis Vis(1 UG-11) Vis(1 UG-11) Vis(2 UG-11) Vis(2 UG-11) Detected The upconverted fluorescence signal intensity is determined by the convolution of the fluorescence and gate pulse intensities: ()()()sumfluogate I ItItdt (4-5) gate flu sum Luminescence Gate Pulse Up-converted signal Non-linea r Crystal Excitation pulse Luminescence Gate pulse Up-converted signal A) B)

PAGE 98

98 where represents the time delay between the arrival of the gate pulse with respect to the sample fluorescence. This optical gating t echnique is very advantageous because the time resolution is dependent only on the width of the gate and pump pulses, not the detection system.( 182 ) The optical path length was 2 mm and the c oncentration of samples did not exceed 30 M yielding an optical density ~ 0.45/mm. A circul ating cell was used to ensure that a fresh volume of sample was excited with every laser shot and a maxi mum of 100 nJ of energy per shot were used. Anisotropy is the measurement of the extent of polarization that a ma terial maintains after being excited with polarized light. When the em ission anisotropy is nonzero, the emission of the material is polarized. The transition dipole moment (a) of a molecule dictates which orientation or direction molecules will absorb light. Light that is polarized consists of an electric field (E) that oscillates in a particular direction. Excita tion of a material with linearly polarized light results in an excitation probability function that is propor tional to the square of the scalar product of the molecules dipole moment and the electric field vector (a E or cos2 A) (Figure 4-6). The phenomenon of polarized emission is dependent on the absorption and emission transition dipole moments which can be oriented at different angles to one another. When the angle between the two vectors is 90, the excitation probability is zero and maximized if they are parallel. After creating an exciton in a high ener gy electronic state of an anisotr opic material, it relaxes to the first singlet state (Kashas Rule ( 2 )) via internal conversion. Regard less of the orientation of the transition moment of the high ener gy initial state, the emissive transition moment at the first singlet state will remain the same (Figure 45). If the absorption and emission moments are identical, the anisotropy will not be lost; however, if they differ, Figure 4-5, the anisotropy value will change.( 2 )

PAGE 99

99 S 2 S 1 S 0 Absorption Fluorescence S 0 S2S 0 S1 Transition moments S 2 S 1 S 0 Absorption Fluorescence S 2 S 1 S 0 S 2 S 1 S 0 Absorption Fluorescence S 0 S2S 0 S1 Molecules can be excited selectively simply by arranging the electric field vector of the incident light so that its orient ation is relatively similar to thei r dipole moments. This method is referred to as photoselection. For ex ample, if a laser pulse with a polarization set to vertical is used to pump a sample, only the molecules with vertical dipole moments will be excited. Multiple processes can cause depolarization within molecules.( 2 ) These include: Adapted from B.Valeur. (2) Absorption and emission transiti on moments are not parallel twisting vibrations Brownian motion Energy transfer to other molecules with di fferent transition mome nt orientations Molecular rotations Figure 4-5. Transition moments. Adapted from B. Valeur.( 2 ) Anisotropy measurements become very us eful tools for determining information concerning molecular size, shape an d flexibility in addition to the viscosity of the solvent. The fundamental anisotropy ( r0) is the theoretical anisotropy of a material that does not undergo any motion or loss of polarization.( 2 )

PAGE 100

100 No Absorption Maximum Absorption Absorption cos2A ANo Absorption Maximum Absorption Absorption cos2A A 2 023cos1 52 r (4-6) Figure 4-6. Photoselection. Adapted from ( 2 ) For a spherical object, if the absorption and emission transition dipole moments are parallel ( = 0), r0 should equal to 2/5 (0.4); howev er, if they are perpendicular ( = 90) the lower limit is -1/5 (-0.2). These values corre spond to the limiting valu es. If all emission polarization is lost (due to a ny of the processes listed above) a value of zero anisotropy is expected. The temporal behavior of the anisotro py can provide useful information regarding the polarization loss mechanism.( 2 ) Fluorescence anisotropy decay measurements were conducted by rotating excitation pulses with respect to a fixed polarization detection scheme. A Berek compensator is used to excite the molecule with a beam polarized paralle l and perpendicular with respect to the detected fluorescence intensities. The anisotropy value ( r ) was then calculated using: 2 I I r I I (4-7)

PAGE 101

101 The Model 5540 Berek polarization compensator from New Focus was used in these experiments to convert and control the pump polariz ation. A compensator su ch as this utilizes the principal that different wavelengths of light propagate at different speeds through a medium and that this velocity depends on the index of refraction. This compensator can cause a -wave or -wave retardance for wavelengths in the ultr aviolet (200 nm) to the in frared (1600 nm). The compensator has a 12 mm aperture and was direc tly mounted to a post. The Berek compensator is made up of a single birefringent uniaxial plate with an adjustab le tilt angle to impose velocity changes on incident light resulti ng in retardation. The velocity changes are both tilt angle and wavelength dependent. The extraordinary axis, ne, is oriented perpendicular to the plate while the ordinary axis, no, is parallel (Figure 4-7). If no tilt is imposed, the incident light remains normal to the plate. As the light pr opagates through the medium, its velocity remains unaffected by the polarization and is only de pendent on the ordinary index of refr action. If the plate is tilted to a particular angle, R, the velocity of the propagating light is changed. The axis oriented in the plane of incidence is no longer ordinary, in stead it has an extraordinary component, ne, causing retardation. Polarized light that is perpen dicular to the plane of incidence has a velocity unaffected by the tilt. As a result, there is a reta rdance that is created between the ordinary and extraordinary waves propagating in the polarizati on planes. The main advantage of using a Berek compensator for polarization measurements is that it allows for simple and independent adjustments for not only retardati on but also plane of incidence orientation adjustments (which are both wavelength dependent) as one unit. The retardation knob is used to set the tilt angle while the orientation knob acts as a wave plate.( 184, 185 ) Due to group velocity dispersion, the reta rdance of the electric field is wavelength dependent. Therefore, one must set the correct position of the Berek compensator polarization

PAGE 102

102 axis. First, the tilt angle,R is calculated then used to calcu late the Retardation Indicator position (I). From the Berek compensator manual we can derive the relationship between R and the tilt angle. A summary is provide d in this dissertation. Consider a uniaxial crystal with an optical axis parallel to the plate surface. A normal incident beam experiences a retardation (R ) that is dependent on the path length ( d ), wavelength ( ) and the ordinary and extraord inary indices of refraction:( 186 ) 0()ed R nn (4-8) However, if the plate is tilted, the retardation equation becomes:( 186 ) (coscos)eeood Rnn (4-9) The tilt-induced extraordinary index of refraction from Fi gure 4-7 is determined by:( 184, 185 ) 22 '22cossin 1RR eoennn (4-10) The relationship between the optical axis of the medium, tilt angle, angle of incidence and indices of refraction are used to derive the following equation for the retardance:( 184 ) 22 22 221sin 2000 sin1 1sineR oR oRn Rn n (4-11) To use the Berek compensator as a half wave plate ( /2), R is fixed to 0.5. The tilt angle and Retardation Indicator equations are purely empirical and are based on the crystal dimensions only known by New Focus ( 187 ) and the dispersion relations for the indices of refraction determined by Dodge ( 188 ) included in the Berek compensator manual. The indicator versus wavelength graph corresponding to quarter and half wave retardance is included in the manual. ( 184 ) The tilt angle is estimated usin g the following empirical equation:( 184 )

PAGE 103

103 1sin(0.284)R R (4-12) where is the wavelength in micrometers. OnceR is determined, the Retardation indicator setting on the compensator can be calculated from the following empirical relationship:( 184 ) 50.2271sin 4RI (4-13) After the retardation indicator is set, the Orientation (O) positi on must be rotated to the proper position (Figure 4-8). For excitation at 375 nm, th e tilt angle in radians was calculated to be 0.1238 which led to a Retardation Indicator setting (I) of 6.57. To rotate polarizations by 90, the retardance is turned to a -wave setting a nd the orientation positioned at 45 since the /2-wave plate causes rotation of the plane of polarization by twi ce the orientation angle.(184) A polarizer was used to verify the polarization of the beam exiting the compensator. The Berek compensator was set to magic angle conditions to measure isotropic fluorescence decay curves. Magic angle is a set condition that enables de tection of the total fluorescence intensity, not just emission proportional to I or I The emission monochromator depends on polarization; the obser ved signal is not proportional to the total intensity (which is equal to2 I I. In order to achieve the correct ratio, th e excitation is oriented 54.7 from the vertical since the cos2 (54.7) is 0.333 and sin2 (54.7) is 0.667 forming the correct sum for the total intensity. This is especially important for fluorescence decay measurements because the vertical and horizontal signals are usually very distinct due to molecular rotations, energy transfer or some other polarizat ion dependent process and if th eir intensities are not properly weighted then incorrect populati on decay times are recovered.(189) The intensity at magic angle is calculated as: (189)

PAGE 104

104 nonone ne none LIGHT LIGHTR= 0 RNo Tilt Tilted nonone ne none LIGHT LIGHTR= 0 RNo Tilt Tilted I O y xInput: Linearly Polarized Wave Plate Setting: /2 Output: Linearly Polarized 90rotated y x I O y xInput: Linearly Polarized Wave Plate Setting: /2 Output: Linearly Polarized 90rotated y x magic angle2 3 I I I (4-14) Figure 4-7. Berek polarization compensator. Tilting the crystal causes retardance and birefringence. Adapte d from New Focus.(185) Figure 4-8. Berek compensator used as a ha lf-wave plate. I = re tardance indicator, O = orientation. Adapte d from New Focus.(184)

PAGE 105

105 Photophysics of Variable Chain Length PPE-CO2 Polymers Steady State Characterization The steady state photophysics pertaining to va riable chain lengths of water soluble PPECO2 polyelectrolytes has been previously reported.(181) Inhomogeneous br oadening (Figure 49) is exhibited in the absorption spectra due to a distribution of excitation energies resulting from slight variations and superpos itions of absorptions of vari ous segments with different conjugation lengths. In addition, as the length of the polymer ch ain increases, the isolated peak shifts towards the red possibly due to an extension of the conjugation length.(190) The absorption maximums for the 8 PRU and 185 PR U are 404 and 432 nm respectively. Moreover, the shoulder at 432 nm, which is assigned to aggregated species, b ecomes the absorption maximum for the 108 and 185 PRU polymers. Stru ctured vibronic featur es in the emission spectrum are shown in Figure 4-10. The high ener gy emission corresponds to isolated chain emission while the broader, low en ergy emission arises from the a ggregate states (appearing as a shoulder). The emission does not disp lay the similar red sh ift seen in the absorption, instead the fluorescence peak shifts are extremely small and d ecrease as the chain length is extended. The S0 S1 (0-0) transition correspondin g to the 35 PRU is Stokes shifted with respect to the 404 nm absorption by 20 nm. Meanwhile, the 185 PRU disp lays a Stokes shift of only 4 nm between the blue end of the emission (436 nm) and the re d edge of the absorption (432 nm). The 185 PRU sample undergoes self-absorption at 436 nm. A similar behavior has been observed in poly(para)-phenyleneladder-type (LPPP) (116) in which the bridging present within the polymer prevents the phenyl rings to twist, maintaini ng conjugation. The authors claim that the small Stokes shift reflects the rigid ge ometry of the conjugated main chain resulting in reabsorption of the S0 S1 (0-0) transition.(181) PPE-CO2 polymers are geometrically rigid resulting in comparable Stokes shifts to the LPPP polymer As the chain length in creases from 35 to 185

PAGE 106

106 300325350375400425450475500 0.0 0.2 0.4 0.6 0.8 1.0 norm Abs (nm) PRU the weight of the aggregate emission at 520 nm increases due to an increase of the amount of aggregate present in solution. Figure 4-9. Chain length absorption shift for PPE-CO2 in methanol. Polymer repeat units 8 ( ), 35 ( ), 108 (), 185 ( ) Figure 4-10. Emission spectra 10 M PPE-CO2 (methanol) excited at 380 nm for 35 PRU ( ) and 185 PRU ( ). Despite the direct relationship between the increase in molar extinction values to the increase in repeat units, the qua ntum yield for fluorescence decreases from ~ 0.6 (8 PRU) to ~ 400450500550600650700 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Norm PL (nm)

PAGE 107

107 0.1 (108 PRU). Zhao et al. sugge st that conformational, vibra tional and rotational degrees of freedom creating non-radiative decay channe ls lead to decrease fluorescence.(181) On the contrary, it is clear that the large absorption red shifts due primarily to the rigidity of the polymers lead to large emission and absorption spectral overlap. Moreover, the conformational restrictions induced by this rigi dity give rise to conjugation lengths longer than expected ultimately reducing the degrees of freedom within the polymer and thus potentially makes the stated reason an invalid argument to expl ain small quantum yields for long PPE-CO2 chains. Figure 4-11 shows the emission spectrum of the PPE-CO2 (35 PRU). The black line corresponds to the polymer dissolved in methanol and it shows the sharp bands characteristic of isolated chain emission. The red line corre sponds to a methanol solution in which 6 M of Ca2+ has been added. Ca2+ is an effective cross linker with th e 2 carboxyl groups inducing aggregation of the PPE-CO2 -.(155) Emission from the aggregate can be clearly observed on the shoulder at 520 nm, as it grows relative to the unaggregated emission at 436 nm. Finally, when the CPE is dissolved in water, the aggregated emission is mostly observed (green line). Overall, the red shift, quenching and band broadening are due to a ggregate formation of th e polymer chains since Ca2+ is a closed-shell ion and does not act as an electron or energy acceptor. (191-194) The small Stokes shift previously mentione d results in excellent overlap of the isolated chain emission with the aggregate absorption enhancing energy tr ansfer from the higher energy isolated species to the lower energy aggregates. The excitation spectrum of a given chromophore is determined by monitoring the fluorophore emission as it is exci ted at different wavelengths. Abramowitz et al. from the Olympus Microscopy Resource provides an excellent description of the excitation spectra collection process.

PAGE 108

108 An emission wavelength is chosen and onl y emission light at that wavelength is allowed to reach the detector. Excita tion is induced at various excitation wavelengths and the intensity of the emitted fluorescence is measured as a function of wavelength. The result is a graph or curve which depicts the relative fluorescence intensity produced by excitation over the spectrum of excitation wavelengths. (195) Excitation experiments at differe nt detection wavelengths can be employed to identify the species contributing to the emission which can be hidden due to inhomogeneous broadening caused by the polydispersity presen t within polymeric samples. The absorption spectra lead one Figure 4-11. Emission of 10 M 35 PRU PPE-CO2 in methanol (), methanol with ~ 6 M Ca2+ ( ) and in water ( ) to believe that no aggregates ar e present in the shor ter polymer samples since neither a shoulder or broadening are observed; howev er, excitation spectra indicate th at this is not the case. Figure 4-12 presents the excitation spectra of the 8 PR U CPE in methanol det ected a four different wavelengths (430, 475, 510 and 590 nm). Detecti on at 430 (blue line) and 475 nm (green line) show broad, featureless exci tation spectra peaked at 396 nm Upon shifting the detection wavelength to 510 nm the excitati on spectra becomes even broade r and a small red-edge shift 400450500550600650700 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Norm PL (nm)

PAGE 109

109 325350375400425450475500 0.0 0.2 0.4 0.6 0.8 1.0 1.2 a.u.(nm)begins to appear (440 nm). By shifting detection to 590 nm (red line), this red edge shift appears more pronounced and it is due to the aggregate species. From this data, we conclude that even in dilute solutions of CPE with small PRU lengths, some aggregate is present and contributes to the overall energy transfer mechanism due to spectral overlap. A small amount of the short, ri gid chains are likely to stack on top of each other rather than cluster up creating this red edge shift. It is also suggested in Zhaos work that the reduction in quantum yield could be a resu lt of the presence of aggregates which compete with radiative decay channels.(181) Photophysical data presented he re suggest that the presence of aggregates is indeed a more suitable expl anation because the pres ence of isolated and aggregated chains is detected in small, dilute PRU samples in MeOH. Figure 4-12. Excitation spectra 10 M 8 PRU PPE-CO2 -. Detection at 430 ( ), 475 ( ), 510 (), and 590 ( ) nm

PAGE 110

110 325350375400425450475500 0.0 0.2 0.4 0.6 0.8 1.0 1.2 a.u. (nm)For longer PPE-CO2 PRU chains dissolved in methanol the distinction between isolated and aggregate emission bands based on the exc itation spectra becomes clearer. For example, Figure 4-13 shows the excitati on spectra of 35 PRU PPE-CO2 in methanol. Detection at 430 nm (blue line) results in a distinct peak at 390 nm. As the detection wavelength increases, the peak broadens and shifts to the red. Detection at 590 nm (red line) clearly shows a new peak at 430 nm and this peak is attributed to the direct excitation of aggregat ed species. Figure 4-14 presents the excitation spectra of the 35 PRU CPE dissolved in water detected a four different wavelengths (430, 475, 510 and 590 nm). Detection at 430 nm (blue line) show broad, featureless excitation spectra peaked at 390 nm due to is olated chains. As the detection wavelength increases, the peak broadens and shifts to the red. Upon shifting the det ection wavelength to 510 nm (black line) the excitation spectra becomes even broader a nd a new peak appears (436 nm). Detection at 510 (black line) a nd 590 nm (red line) exhibit more well-defined peaks at 436 nm, comparable to the 185 PRU absorption spectrum in neat methanol. Figure 4-13. Excitation spectra 10 M 35 PRU PPE-CO2 in methanol. De tection at 430 ( ), 475 ( ), 510 (), and 590 ( ) nm

PAGE 111

111 325350375400425450475500 0.0 0.2 0.4 0.6 0.8 1.0 1.2 a.u (nm)A similar trend is observed in Figure 4-15. Af ter addition of 60% calcium to the methanol solution, noticeable changes within the excitation spectra (at blue r wavelengths) indicate that the calcium induces more aggregation within the poly mer solution. Contribution from aggregates is clearly evident when det ecting at 475 (green line), 510 nm (b lack line), and 590 nm (red line), although the signal collected at 590 nm does not increase significantly compared to the neat methanol sample. When 35 PRU PPE-CO2 is dissolved in water, it is confirmed that CPEs do exist in water as aggregates, however; isolated chains are still present although their contribution to the overall steady state fluorescence is reduced due to fewer free chains in solution, reabsorption and energy transfer. The excitation sp ectra results presented here provide evidence to support our assumption that multiple specie s contribute to the emission and overall fluorescence decay in even dilute PPE-CO2 samples. Figure 4-14. Excitation spectra 10 M 35 PRU PPE-CO2 in water. Detection at 430 ( ), 475 ( ), 510 (), and 590 ( ) nm

PAGE 112

112 325350375400425450475500 0.0 0.2 0.4 0.6 0.8 1.0 1.2 a.u. (nm)Time-Resolved Fluorescence Isotropic Upconversion The fluorescence dynamics of PPE-CO2 with repeat unit lengths equal to 8, 35, 108 in MeOH were excited at 375 nm and detected at ma gic angle at several different wavelengths. Data was fit with a sum of exponent ials using the following equation: ()expi i it ItA (4-15) whereiA represents the weight of each rate constant and i is the associated time constant. Data from these fits are summarized in Table 4-2. Figure 4-15. Excitation spectra 10 M 35 PRU PPE-CO2 in methanol with ~ 6 M Ca2+. Detection at 430 ( ), 475 ( ), 510 (), and 590 ( ) nm Figure 4-16 shows the time-resolved fluorescence decay of 8 PRU PPE-CO2 in methanol at three different detection wa velengths (430, 436 and 450 nm) ex cited at 375 nm. Detection at 430 nm (blue line) shows a multi-exponential behavior which disappears as the detection

PAGE 113

113 wavelength increases to 450 nm (red line). The isolat ed chain time constant of 531 ps (Table 4-2) is extracted from the mono-expone ntial decay of the 8 PRU at intermediate wavelengths 450 (red line). As the detection wavelength is increased from 450 to 550 nm (not shown) the behavior of the exponential decay does not change. At a ll detection wavelengths the rise times are comparable to our instrument response function. The first panel of Fi gure 4-16 shows the same detection wavelengths on a shorter time scale. An extremely fast decay (< 1.5 ps) is observed Table 4-2. Detection dependence decay times PRU Det (nm) 1 (ps) Amplitude 2 (ps) Amplitude 8 PRU 430 27 36% 490 64% 436 14 29 624 71 450 531 100 35 PRU 430 11 65% 178 35% 450 37 39 363 59 500 33 39 402 61 550 35 53 454 47 108 PRU Det (nm) 430 14 62% 201 38% 450 43 39 395 63 500 42 (fixed) 39 333 62 550 39 55 551 45 35 w 50% Ca Det (nm) 430 18 61% 450 40% 450 43 29 468 71 550 26 53 611 47 when detected at 430 nm (blue line) and its cont ribution to the overall signal diminishes as the wavelength increases from 430 to 436 (magenta line) to 450 nm (red line). Emission at wavelengths below 450 nm spectrally overlaps with the aggregate species absorption resulting in

PAGE 114

114 024 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Norm PLtime (ps)050100150200 0100200300400500600 efficient energy transfer from is olated chains to aggregates. In terestingly, detection at 590 nm does not yield a significant change in the rise time (not shown) or decay compared to 450 nm which had been expected if detecting emissi on from the aggregates. Dynamics observed at wavelengths between 430 and 450 nm exhibit an a dditional intermediate decay time of 30 to 40 ps (middle panel). The amplitude of these time c onstants decreases as th e wavelength increases. From the excitation spectr a it is clear that there is a small, yet significant amount of aggregate within the sample below 450 nm facilitating ener gy transfer; hence, the intermediate time constant is assigned to the energy transfer from the ensemble isol ated to the aggregated chains. Figure 4-16. Time-resolved fluor escence decay of 8 PRU PPE-CO2 in methanol at three different detection wavelengths. 430 ( ), 436 ( ), and 450 ( ) nm, Fits () The results for solutions of 35 and 108 PRU PPE-CO2 (30 M) are shown in Figure 4-17 for three distinct wavelengths (430, 450 and 550 nm) upon excitation at 375 nm. Detecting at 430 nm, the contribution from the very fast compone nt is larger for the 35 PRU sample (black line) compared to the 108 PRU sample (red line). As the detection wavelength is shifted to the lower energies, contribution from this fast component to the overal l signal is reduced. Interestingly, the changes are more pronounced on the 35 PRU than on the 108 PRU sample.

PAGE 115

115 Detecting at 550 nm, there is no contribution from the fast component to the 35 PRU signal, but it is still present in the 108 PRU signal. The photoluminescence rise follows the response function of the expe rimental setup. At very long wavelengths one would expect to see a build up of population (rise time) due to energy transfer from the isolated states however, the da ta collected do not show this slow rise time.(156, 196) This indicates that the singlet exciton is tr ansferred from shorter (h igh energy) chains to traps within the chains, not to aggregates. The right column in Figure 4-17 shows the intermediate and long decays, observed at diffe rent detection wavele ngths for the 35 and 108 PRU samples. Detection at 430 nm (top graph) pres ents an intermediate decay constant of 11 ps and a long decay of 178 6 ps. For longer detection wavelengths, the intermediate time constant is ~ 30 to 40 ps while the long time decay lies between 350 and 450 ps. Results are summarized in Table 4-2. The qualitative trend ob served for the intermediate component is the same for the 35 and 108 PRU but not 8 PRU. In the 450 to 590 nm detection region, as the wavelength increases, the amplitude of the interm ediate decay increases but the lifetimes do not change significantly. These time-resolved emission signals result from an ensemble of isolated CPE chains with different conjugation lengths. Meanwhile, the slow decay time increases as the wavelength increases. A 350 to 450 ps time consta nt corresponds to the is olated chain natural fluorescence lifetime (extracted from the 8 PRU dete cted at 450 nm); therefore, this component is assigned to isolated chains not participating in the energy transf er process. Energy transfer is not only dependent on spectral overlap of the donor fluorescence (isolated chain) and acceptor absorption (aggregate chain), but also on the molecular distance betw een the two species.(2) After initial energy transfer (<1.5 ps) from short conjugated segments to traps, the energy is then transferred to the aggregate species (intermediat e decay time). The change in amplitudes of the

PAGE 116

116 intermediate decay time depends on the percentage of isolated chains transferring energy to the aggregate species. Figure 4-17. 30 M PPE-CO2 35 PRU () and 108 PRU ( ) with fits ( ) at various detection wavelengths A) 430 nm, B) 450 nm, C) 550 nm Figure 4-18 shows photoluminescence of a solution of 35 PRU PPE-CO2 in methanol detected at 430 nm and the influence of addition of Ca2+. The first panel shows the fast decays, the second panel shows the intermediate decays and the third panel s hows the long decays. Ca2+ greatly influences the dynamics, dominating the u ltrafast decay signal, altering the intermediate decay amplitude and increasing the long decay as the detection wavelength increases. Rise times -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 C) A) B)0.0 0.4 0.8 1.2 1.6 2.0 2.4 Norm PL0.0 0.4 0.8 1.2 1.6 2.0 Norm PL-0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0123450.0 0.4 0.8 1.2 Norm PLtime (s)01002003004005006000.0 0.2 0.4 0.6 0.8 1.0 time (ps)

PAGE 117

117 012345 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Norm PL050100150200 0.00 0.25 0.50 0.75 1.00 1.25 1.50 time (ps)0100200300400500600 0.00 0.25 0.50 0.75 1.00 1.25 1.50 are IRF-limited at all wavelengths. The augm ented amount of aggreg ation increases the amplitude of the energy transfer components. Jiang determined that Ca2+ induces the formation of loose aggregates in methanol a nd more traps along the polymer backbone.(155) Our results show that the addition of Ca2+ (red line) amplifies the energy tr ansfer from the short isolated chains to traps. Detecting at 430 nm, the long component increases from ~180 to ~ 450 ps while the intermediate decays times remain relatively the same (Table 4-2). Figure 4-18. 30 M PPE-CO2 35 PRU without () and with ( ) Ca2+ at 430 nm. Figure 4-19 presents the eff ect of the addition of Ca2+ to the dilute 8 PRU sample detected at 450 nm upon excitation at 375 nm. The black lin e corresponds to the isolated chain emission and the red line corresponds to the 8 PRU with the addition of 15 M Ca2+. At 450 nm, the addition of Ca2+ introduces a 1.5 ps decay time and an intermediate component (~ 30 to 50 ps) not seen in the neat 8 PRU sample. The radiativ e decay rates measured w ith and without calcium differ due to the presence of aggregate structures in conjugated polymers. As stated previously, the addition of the dication calcium to the PPE-CO2 -/methanol mixture causes aggregation,

PAGE 118

118 bringing the molecules within the Coulombic c oupling transfer radius facilitating extremely efficient energy transfer.(197) Cation-induced aggregation plays a critical role in amplified quenching (154, 155, 157, 168, 172, 180) due to ultrafast energy transfer. In addi tion, the Stern-Volmer behavior of CPEs in the presence of polyvalent quencher ions such as MV2+ and a cation in solution have proven to be superlinear. In fact, due to the loose type of aggregates formed when calcium is added to PPE-CO2solutions, small quenchers, like MV2+ have higher quenching e fficiencies than in other aggregate-inducing solutions, i.e., water. (155) Figure 4-19. 10 M PPE-CO2 8 PRU detected at 450 nm without () and with ( ) Ca2+ We examined the effects on the energy transfer after addition of a quencher molecule to a solution of PPE-CO2 with and without calcium. Figure 4-20 compares the fluorescence dynamics a solution of 30 M 35 PRU PPE-CO2 in methanol excited at 375 nm and detected at 450 nm when 15 M calcium, MV2+, and a mixture of the two are added. The concentration used for MV2+ samples was equivalent to the amount need ed to quench the steady state fluorescence 0100200300400500600 0.0 0.2 0.4 0.6 0.8 1.0 Norm PLtime (ps)

PAGE 119

119 050100150200250300350400450 0.00 0.25 0.50 0.75 1.00 Norm PLtime (ps) by 80% (I/I0). The ultrafast decay that occurs in less than 1.5 ps is present in all samples regardless of the presence of calcium (not show n). No significant changes in the excited state lifetime at long timescales were detected desp ite a reduction in the photo-luminescence quantum yield implying that the most important step in the decay mechanism facilitating amplified quenching occurs in the first 2 ps in PPE-CO2 polymers. Figure 4-20. 30 M PPE-CO2 35 PRU with () Ca2+ at 450 nm 30 M PPE-CO2 35 PRU with MV2+ (80% quenched) at 450 nm and ( ) 30 M PPE-CO2 35 PRU with 15 M Ca2+ and MV2+ (80% quenched) Time-Resolved Anisotropy To investigate the fluorescence depolarization, random walk migration, or intermolecular energy transfer we measured the anisotropy dynamics of the 8 PRU polymers in methanol. Figure 4-21 depicts the time-resolved fluorescence anisotropy of the 8 PRU detected at 430 (black line) and 450 (red line) nm upon excitation at 375 nm. An ultrafast loss of anisotropy (not shown) followed by a constant anisotropy duri ng the lifetime emission of the polymer is

PAGE 120

120 0100200300400500600 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 rtime (ps)observed. After the initial change in the first 5 ps from r ~ 0.4 to 0.2, both curves then remain parallel to one another. Figure 4-21. Anisotr opy of 8 PRU PPE-CO2 -. Detection wavelengths 430 () and 450 ( ) nm Detection at 430 nm occurs in a region in which shorter conjugati on lengths are present allowing for a higher number of hops before finding traps along the polymer backbone resulting in smaller anisotropy value (r~ 0.10) compar ed to the 450 nm (r~ 0.18). The 450 nm scan corresponds to slightly longer conjugation and fewer hops As shown in other works (130, 197) a long polarization decay corresponds to aggregat es emitting almost randomly polarized light reducing the total polarization. However, in this po lymer there is little to no depolarization due to reorientation at 450 nm. Random wa lk migration is considered to occur at intermediate decay times in PPE-SO3 -, (130) but is not observed in PPE-CO2 -. If random walk of excitations along the polymer backbone were to occur, as th e wavelength increased the hopping rate would decrease due to a lack of lower lying states th at are available for the excitation to jump. The following factors clearly eliminate the possibility of random exciton migr ation in this PPE-CO2 -:

PAGE 121

121 rigidity of the polymer, longer conjugation length, spectral over lap and proximity leading to energy transfer from isolated to aggregated ch ains, no detection wavelength dependence for the decays measured at various wavelengths and no depolarization at intermediate decay times even in the presence of aggregates. Potential Kinetic Model The scheme presented in Fi gure 4-22 depicts a proposed model concerning the dynamics mechanism in PPE-CO2 -. This figure presents the model for the different polymer repeat units collectively. The 8 PRU consists of mostly isolated chains. Afte r excitation at 375 nm, energy is transferred from the shorter conjugation lengths to traps located in the isol ated chains. Due to the spectral overlap between the isol ated chains emission and aggregate absorption, this energy transfer is detected primarily at wavelengths shorter than 450 nm. These detected emission signals are convoluted with decays from an ensemble of differen t conjugation lengths, kinks and other non-radiative recomb ination pathways within the isolat ed species. From the excitation spectra of the 8 PRU (Figure 4-12) we can see th at the red edge shift at 440 nm does not overlap significantly with wavelengths grea ter than 450 nm. Energy transfer from the isolated chains to the aggregates is not efficien t resulting in a mono-exponential fluorescence lifetime from an ensemble of isolated chains. Longer polymer chains (35, 108, 185) contain a larger mix of isolated and aggregated chains. After excitation at 375 nm, energy is tran sferred from the shorte r conjugation lengths to traps located in the isolated chains similar to th e 8 PRU. Subsequently, energy transfer from this ensemble to the aggregate species is observed wi thin a wide spectral range. The number of traps is increased upon addition of cal cium resulting in a sharp incr ease in the ultrafast time component but the energy transfer to the aggregate does not change significantly. Self-absorption is observed in the 108 and 185 polymer repeat unit samples (more aggregated); however, this is a

PAGE 122

122 radiative energy transfer process in which the photon emitted by the donor is then absorbed by the acceptor. Therefore, it does not compete w ith other decay mechanisms and the fluorescence decay time of the donor remains unchanged (refer to Chapter 1). Following the energy transfer, the detected emission is dominated by the ensemb le of decays from isolated chains and traps located along the polymer backbone in addition to competing with non-radiative decay channels. In summary, upon excitation of the aggregates fro m energetically higher lying isolated chains, the fluorescence lifetimes result in multi-exponen tial behavior due to the competition between the radiative and non-radiative decay. The in tegrated fluorescence collected for these experiments does show emission fr om aggregates but it cannot be observed in the ultrafast timeresolved experiments because of very long decay time constants and small contribution to the overall signal. Summary The ultrafast time-resolved fluor escence of a series of PPE-CO2 polymer repeat units is presented. Using steady stat e UV-Vis, photoluminescence a nd excitation resources we distinguished the species present in each soluti on. It was shown that even dilute, short PRU chains do exhibit a small amount of aggregation. The addition of calcium or using water as the solvent induces aggregation result ing in broad absorption/excitati on spectra and the growth of a red shoulder in the emission. To investigate th e influence aggregation has on the fluorescence of the polymers, we conducted a detection wavele ngth study using fluorescence upconversion. The isolated chain emission was extracted from th e 8 PRU at 450 nm (isolated chain emission and aggregate absorption is minimal). In the presence of aggregates, an intermediate time constant on the order of 30 to 40 ps is observed and is assigne d to the energy transfer from the isolated to aggregate species. At bluer wavelengths, a fast de cay (< 1.5 ps) is observed and is attributed to the transfer of excitation from shorter, high ener gy chains to longer, low energy chains and traps.

PAGE 123

123 Time-resolved anisotropy confirmed that this polymer, no matter the PRU size, is extremely rigid and has long c onjugation lengths. 375 nm Isolated Aggregate NR Isolated/traps1 < 1.5 ps 2 ~ 30 to 40 ps 3 ~ 350 to 450 ps 375 nm Isolated Aggregate NR Isolated/traps 375 nm Isolated Aggregate NR Isolated/traps1 < 1.5 ps 2 ~ 30 to 40 ps 3 ~ 350 to 450 ps Figure 4-22. Possible kinetic model for all PPE-CO2 PRU chains

PAGE 124

124 CHAPTER 5 CONCLUSIONS AND FUTURE WORK Nanoparticle Conclusions and Future Work Conclusions A complete steady state and time-resolv ed study of size, sh ape passivation and composition dependence on colloidal semiconducto r nanoparticles has been conducted in our labs. Using pump-probe spectroscopy we were ab le to detect traps and interfacial states. Confirmation of Auger-like cooling resulting in 1P to 1S relaxation (~ 1 ps) has been shown in addition to interband relaxation (> 200 ps) have been measured for each nanoparticle system. Comparisons between materials with differe nt compositions were made finding higher confinement potential in ZnCdSe alloys than in CdSe resulting in a lower probability of the exciton to sample the surface and be trapped. Finally, utilizing each of the spectroscopic tools available we were able to combine steady state and time-resolved data to construct qualitative models describing the nanorod systems. From this work we can conclude that passi vation and alloying resu lt in quantum yields higher than for bare CdSe. The excitons are more confined in the alloy particles than in CdSe rods. It is well known that the band gap is size (17) and shape dependent.(53) Despite breaking the symmetry within the nanopart icle, confinement properties were maintained in our samples. Passivation with an inorganic sh ell results in increased quant um yields and bleach signals because the surface traps are eliminated. Finally, modifying the composition using one additional synthetic alloying step has grea tly improved the process-ability without drastically sacrificing confinement characteristics.

PAGE 125

125 Outlook/Future Work Lasing/ Optical gain A population inversion in lasing me dia, in which the population of electrons in the excited state mu st be greater than the populati on left in the ground state, is necessary to achieve optical gain It is clear that the size of these nanocrystals leads to an enhancement of the carrier-carri er interactions leading to impressive optical properties in systems with single and multiexciton states. It has been shown that new energy relaxation pathways are created in nanocrystals compared to the bulk. For example, due to the high density of states present in bulk mate rials, there is a la ck of a phonon bottleneck which is also bypassed in nanocrystals because of ultraf ast non-radiative Auger recombin ation. Therefore, to attain inversion within nanocryst als, a simultaneous exc itation of the two electr ons in the ground state to an excited state must occur resulting in emi ssion of multi-excitons via an Auger process which dictates the decay of optical gain. (122, 198, 199) Recently, Klimov has investigated the mechan isms for photogeneration and recombination of multi-excitons in nanocrystals necessary fo r lasing and solar energy conversion. However, since this recombination occurs in less than 1 ps, the optical gain lifetimes are in the picosecond regime which is a drawback when designing materi als for lasing. To improve optical gain it is important to develop new materi als that inherently diminish this phenomenon. It has been suggested that increasing the nanoc rystal volume fracti on (packing density) in the optical gain medium or using quantum rods instead of spheres will help reduce the influence that Auger recombination has on these systems.(199) Klimov has been using quantum rods to try to achieve population i nversion for optical gain. In order to do this, the Auger effects mu st be significantly reduced. He showed that especially for CdSe based nanocry stal, the rods have slower A uger rates compared to quantum dots with the same volume that emit in the red and orange spectral region.(199) It is hoped at

PAGE 126

126 particular emission wavelengths, the Auger decay can be stifled in rodshaped nanocrystals due to not only the dependence that th e confinement potential has on th e length and size but also the linear scaling of the deca y time with rod volume.(52, 199, 200) For higher energy (shorter wavelengths) quantum dots, the increased surfac e-to-volume ratio inhibits the Auger decay suppression. Elongation of the na nocrystal in the c-direction has successfully increased the optical gain lifetime since the effect that Auge r has on the recombination behavior in rods is decreased.(199, 201) CdSe/ZnSe core/shell materials have been thought to be used for this application but it is more suitable to use inverted ZnSe/CdSe heterostructures in order to control the electron-hole wavefunction overlap to increas e the confinement energies and reduce Auger recombination.(199, 202, 203) Formulation of new alloy nanorods has opened the door for new investigations for their potential applications. Extensi on of the c-axis in these tern ary materials enables for higher confinement potential in the blue -green region. It may be worth trying to determine how to alloy the materials and then coat them to increas e their photoluminescence quantum yield to make them comparable to the core/shell or shell/core materials. These new alloy/shell materials can be tunable based on the diffusion and have extended lifetimes necessary for charge-separated or optical gain applications. Light emitting diodes (LEDs) (11, 204-206). Although several improvements using organic molecules for organic light emitting diodes (OLEDs) make them comparable to current technologies, there are ongoing drawbacks and pr oblems that must be overcome before these devices can be commercially utilized. These incl ude: a) difficulties tuning the colors since the fluorescence is broad and b) synthesis of multiple molecules is required to obtain a broad range of colors. Nanocrystals are bei ng considered as attractive candida tes to be used for LEDs since

PAGE 127

127 their emission is not only tunabl e but considerably narrower than that from organic materials. Moreover, nanocrystals have a higher proba bility of resisti ng photodegradation.(66) Hybrid OLEDs have been developed in the past fifteen years, incorporating a polymer such as PEDOT (207) or PPV (204) to transport charge to various nanocrystals (CdSe (204), CdSe/CdS (11), CdSe/ZnS (208)) that act as the emission layer resulting in more stable and efficient devices. Development of better materials and manipulating their interactions are the main goals when working towards designing produc ts that result in hi gh electroluminescence efficiencies. To achieve commer cial quality devices the functionality must be improved by enhancing the charge transfer between the pol ymers to the nanocrystal emission layer and increasing the surface quality so the traps which cause non-radiative recombination are reduced.(66) Experimentation with different combinati ons of polymer/nanoparticle blends is a standard methodology to find devices that get ri d of such adverse consequences. Within the literature, most nanoparticles are spheres and th eir band gaps are tuned by only changing their diameter.(20, 39, 48, 75, 78, 136, 138, 209-212) As the diameters are decreased, the band gap energy does increase and emission in the blue-gre en region is achieved, however, the surface-tovolume ratio is significantly increased which can lead to more surface traps. Therefore, some investigations into hybrid LEDs s hould incorporate not only size dist ributions to obtain tunability but to investigate how the shape, passivation thickness and composition will affect the overall efficiencies of the devices. A wide range of colors in the bl ue-green region can be achieved simply by altering the alloying times in ZnCdSe nanorods. If a simple technique was developed to passivate these alloy rods to reduce surface traps, they would allow for blue-green emission wavelengths via a simple s ynthetic route (one batch).

PAGE 128

128 Photovoltaics (213, 214) Although the cost of making quant um dot based photovolatics is small, the efficiencies, due to recombination loses, are still too low for them to be used on a large scale. Hybrid photovoltaic devices are integrated within the polymers to transport charge for such applications as so lar cells. Achieving charge separati on and positive transport of the hole and electron to the indium tin oxide (ITO) and aluminum electrode respectively is the main goal in photovoltaics.(66, 214) Instead of focusing on a binary system, a device utilizing ternary compositions with varying degrees of Zn diffus ed into the core, may serve as more suitable materials for photovoltaics. The variable Zn diffusi on will create a gradient from the core to the surface enabling the exciton to hop from one rod to another ultimately reaching the aluminum electrode. Also, work might be directed to achie ve charge separation w ithin the nanocrystals. The fact that the hole created after excitation within our CdSe /ZnSe quantum rods potentially tunnels into the shell could help sustain charge separation and i nhibit premature charge carrier recombination. Investigations into the kinetics and mechanis ms for creating and maintaining charge separation in these materials are recomm ended since these processes are not completely understood; however, it has been shown that the interface between nanocrystals and porous TiO2 supports highly efficien t charge separation.(132, 199, 202) Carrier multiplication The carrier-carrier interactions in nanoparticles lead to improved exciton (carrier) multiplication (CM) which resu lts from direct photogeneration of multiple electron-hole pairs by single photon s. This process is relatively new and the actual mechanism behind carrier multiplication (if it truly exists) from a single excitation are currently being debated. Klimov showed that seven excitons are produced in PbSe nanocrystals (QE = 700% where 100% means 1 photon creates 1 e-h pa ir). This unique finding will be good for photovoltaic cells and improve solar fuel technol ogies in the IR region. However, CdSe dots

PAGE 129

129 have not been able to exhibit as efficien t CM efficiencies in the visible region.(199) Investigations into the e ffect that shape has on this phenomenon is recommended. PPE-CO2 Conclusions and Future Work Conclusions The synthesis, characterization and time-reso lved measurements c onducted on this series of PPE-CO2 polymers with different chai n lengths and extent of aggr egation have led to several interesting observations and conclusions. A co mplete steady state and time-resolved study of length, solvent and aggregated inducer has been conducted in our labs. In particular, isotropic and anisotropic fluorescence upconversion was utilized to u nderstand the quenching mechanism within PPE-CO2 -. We conducted a wavelength detection st udy based on exciting on the blue side (isolated chain) of the absorpti on. Upon excitation of the aggreg ates from energetically higher lying isolated chains, the fluorescence lifetimes result in multi-exponential behavior due to the competition between multiple decay pathways. (116, 153, 156, 165, 215) In all samples, the emission is inhomogeneous ly broadened. Detection of time-resolved signals at all wavelengths is associated with the isolated chains emission, despite the superposition of both species present within the samples. In addition, unlike PPE-SO3 -, PPECO2 is very rigid resulting in l onger conjugation lengths than 4.5 PRU. This was determined not by evaluating the behavior of the polymer chains as a function of chain length using steady-state absorption but by evaluating the extremely slow time-resolved anisotropy decay. From the excitation spectra collected for each polymer chain, it is seen that even a dilute PPE-CO2 with only 8 polymer repeat units exhibits a slight amo unt of aggregation when detected at very red wavelengths. Moreover, we conclude that the energy transfer from isolated to aggregated chains is extremely fast, occurring in 30 to 40 ps fo r all samples. This component was significantly enhanced when a poor solvent (water) or an aggregate inducer was us ed. An even faster

PAGE 130

130 component (1.5 ps) was also observed and it was assigned to the transfer from shorter (high energy) isolated chains to l onger chains and traps within th e isolated chain backbone. Outlook/ Future Work (Hyperbranched PPE-CO2 -) Within our collaboration with the Schanze a nd Reynolds groups, we have an opportunity to investigate excitation and rela xation mechanisms for materials in solutions and in films using both fluorescence up-conversion and broad band transient absorption techniques. Xiaoyong Zhao, a student in the Scha nze group, has synthesized a new hyperbranched PPE-CO2 in which there are three carboxylate side chains attached on each side of the polymer backbone. Based on excitation spectra data presented in Chapter 4, even dilute solutions of the linear 8 polymer repeat unit displays some sort of aggregati on within the sample. This new, hyperbranched polymer is said to have no aggreg ation present even if dissolved in water. Figures 5-1 show the absorption spectra of this new polymer when dissol ved in different solvents and compared to the linear 8 PRU (data collected by Xiaoyong Zhao). From this figure, it is nece ssary to look into the excitation spectra of the hyperbra nched polymer to compare to the dilute 8 PRU. The small shoulder present in Figure 5-1 on the red side of the absorption is a small indication that there may still be some aggregates pres ent within this hyperbranched sample. However, this shoulder could simply be an artifact of the size and re peat unit distributions pr esent within polymeric samples. The only way to be sure is to collect excitation spectra at vari ous wavelengths and look for red shifts that are characteristic of aggreg ate species. The polymer dissolved in water does not alter the absorption spectra significantly. However, photoluminescence is quenched by nearly half of its original intensity when dissolved in methanol (Figure 5-2). Th is is another indication that the polymer could have some aggregate pr esent despite the lack of structure in the absorption spectrum.

PAGE 131

131 Figure 5-1. Absorption spectra of the hyperbranched PPE-CO2 in MeOH () and water ( ) and linear PPE-CO2 8 PRU in MeOH ( ). If a complete understanding is necessary for these linear and hyperbranched polymers, it is necessary to carry out si milar methods of experimentation conducte d in this dissertation. First, it is imperative to make sure that the excitation spec tra either does or does not indicate the presence of aggregates. Second, a time-resolved detec tion wavelength dependence would be useful to elucidate the dynamics of the system under variou s conditions. This data can then be compared to the linear polymers. Once a comparison is made further experiments can be designed to see which polymer would be better for quenching and how calcium will affect the quenching efficiency. Understanding these conjugated poly electrolytes will help with designing new materials for multiple applications including sola r cells, LEDs and even chemoand biosensors. The hyperbranched CPE avenue is a new and exc iting field and I recommend more time-resolved experimentation be accomplished in this area. 325350375400425450475500 0.0 0.2 0.4 0.6 0.8 1.0 Norm Abs (nm)

PAGE 132

132 Figure 5-2. Photoluminescence sp ectra of hyperbranched PPE-CO2 in MeOH () and water ( ) 400450500550600650700 0 1x1052x1053x1054x1055x1056x1057x105 PL(nm) Methanol Water

PAGE 133

133 LIST OF REFERENCES (1). N. J. Turro, Modern Molecular Photochemistry (Univ Science Books, New York, 1991). (2). B. Valeur, Molecular Fluorescence Principles and Applications (Wiley-VCH, Weinheim, 2002). (3). G. S. H. Singhal, Janos; Rabinowitch, Eugene., Excitation-energy migration between chlorophyll and b-carotene, Journal of Chemical Physics (1968) 49, 5206. (4). Govindjee, Excitation Energy Transfer and Ener gy Migration : Some Basics and Background, http://www.life.uiuc.edu/govindj ee/biochem494/foerster.htm Online Class Notes (5). E. G. Rabinowitch, Photosynthesis (John Wiley & Sons, Inc., New York, 1969). (6). G. D. Joly, L. Geiger, S. E. Kooi, T. M. Swager, Highly effective water-soluble fluorescence quenchers of conj ugated polymer thin films in aqueous environments, Macromolecules (2006) 39, 7175. (7). J. H. Wosnick, C. M. Mello, T. M. Swager, Synthesis and application of poly(phenylene ethynylene)s for bioconj ugation: A conjugated polymer-based fluorogenic probe for proteases, Journal of the American Chemical Society (2005) 127, 3400. (8). T. Kippeny, L. A. Swafford, S. J. Ro senthal, Semiconductor na nocrystals: A powerful visual aid for introducing the particle in a box, Journal of Chemical Education (2002) 79, 1094. (9). A. Hagfeldt, M. Gratze l, Light-induced redox reactions in nanocrystalline systems, Chemical Reviews (1995) 95, 49. (10). V. I. Klimov, D. W. McBranch, C. A. Leatherdale, M. G. Bawendi, Electron and hole relaxation pathways in semiconductor quantum dots, Physical Review B (1999) 60, 13740. (11). M. C. Schlamp, X. G. Peng, A. P. A livisatos, Improved efficiencies in light emitting diodes made with CdSe(CdS) core/shell type nanocrystals and a semiconducting polymer, Journal of Applied Physics (1997) 82, 5837. (12). Y. Wang, N. Herron, Nanometer-sized se miconductor clusters Materials synthesis, quantum size effects, an d photophysical properties, Journal of Physical Chemistry (1991) 95, 525.

PAGE 134

134 (13). M. A. El-Sayed, Small is different: Shape-, size-, and composition-dependent properties of some colloidal semiconductor nanocrystals, Accounts of Chemical Research (2004) 37, 326. (14). M. A. Fox, M. T. Dula y, Heterogeneous photocatalysis, Chemical Reviews (1993) 93, 341. (15). L. N. Lewis, Chemical catalysis by colloids and clusters, Chemical Reviews (1993) 93, 2693. (16). J. Aldana, Y.A. Wang, and X.G. Peng, Photochemical instability of CdSe nanocrystals coated by hydrophilic thiols, Journal of the American Chemical Society (2001) 123, 8844. (17). V. I. Klimov, Optical nonlinearities and ultrafast car rier dynamics in semiconductor nanocrystals, Journal of Physical Chemistry B (2000) 104, 6112. (18). V. I. Klimov, D. W. McBranch, Femtos econd 1P-to-1S electron re laxation in strongly confined semiconductor nanocrystals, Physical Review Letters (1998) 80, 4028. (19). N. Le Thomas, E. Herz, O. Schops, U. Woggon, M. V. Artemyev, Exciton fine structure in single CdSe nanorods, Physical Review Letters (2005) 94, 016803. (20). C. A. Leatherdale, W. K. Woo, F. V. Mikulec, M. G. Bawendi, On the absorption cross section of CdSe nanocrystal quantum dots, Journal of Physical Chemistry B (2002) 106, 7619. (21). M. L. Steigerwald, L. E. Brus, Se miconductor crystallites A class of large molecules, Accounts of Chemical Research (1990) 23, 183. (22). A. P. Alivisatos, Perspectives on the physical chemistry of semiconductor nanocrystals, Journal of Physical Chemistry (1996) 100, 13226. (23). A. L. Efros, A. L. Efros, Interband absorption of light in a semiconductor sphere, Soviet Physics Semiconductors-Ussr (1982) 16, 772. (24). A. L. Efros, M. Rosen, The electron ic structure of semiconductor nanocrystals, Annual Review of Materials Science (2000) 30, 475. (25). A. I. Ekimov, F. Hache, M. C. Scha nne-Klein, D. Ricard, C. Flytzanis, I. A. Kudryavtsev, T. V. Yazeva, A. V. Rodina, and Al. L. Efros, Absorption and intensitydependent photoluminescence measurements on CdSe quantum dotsAssingment of the 1st electronic-transitions, Journal of the Optical Society of America B-Optical Physics (1993) 10, 100. (26). L. E. Brus, Electron el ectron and electron-hole interactions in small semiconductor crystallitesThe size dependence of th e lowest excited electronic state Journal of Chemical Physics (1984) 80, 4403.

PAGE 135

135 (27). N. Chestnoy, R. Hull, L. E. Brus, Higher excited electronic states in clusters of ZnSe, CdSe, and ZnS spin-orbit, vibronic, and relaxation phenomena Journal of Chemical Physics (1986) 85, 2237. (28). Y. Kayanuma, Quantum-size effects of interacting elec trons and holes in semiconductor microcrystals with spherical shape, Physical Review B (1988) 38, 9797. (29). M. Nirmal, L. Brus, Luminescence photophysics in semiconductor nanocrystals, Accounts of Chemical Research (1999) 32, 407. (30). A. L. Efros, F. G. Pikus, V. G. Burnett, Density of states of a 2-dimensional electrongas in a long-range random potential, Physical Review B (1993) 47, 2233. (31). A. J. Nozik, Spectroscopy and hot electron relaxation dynamics in semiconductor quantum wells and quantum dots, Annual Review of Physical Chemistry (2001) 52, 193. (32). X. G. Peng, L. Manna, W. D. Yang, J. Wickam, A.Kadavanich, A.P. Alivisatos, Shape control of CdSe nanocrystals, Nature (2000) 404, 59. (33). L. Manna, E. C. Scher, A. P. Alivisat os, Synthesis of soluble and processable rod-, arrow-, teardrop-, and tetrapod-shaped CdSe nanocrystals, Journal of the American Chemical Society (2000) 122, 12700. (34). L. Manna, E. C. Scher, L. S. Li, A. P. Alivisatos, Epitaxial growth and photochemical annealing of graded CdS/ZnS sh ells on colloidal CdSe nanorods, Journal of the American Chemical Society (2002) 124, 7136. (35). C. D. Donega, S. G. Hickey, S. F. Wuister, D. Vanmaekelbergh, A. Meijerink, Single-step synthesis to control the photoluminescen ce quantum yield and size dispersion of CdSe nanocrystals, Journal of Physical Chemistry B (2003) 107, 489. (36). L. A. Swafford, Homogeneously Alloyed CdSxSe1-x Nanocrystals: Synthesis, Characterization, and Compos ition/Size-Dependent Band Gap, Journal of the American Chemical Society (2006). (37). C. B. Murray, D. J. Norris, M. G. Ba wendi, Synthesis and Char acterization of Nearly Monodisperse Cde (E = S, Se, Te) Semiconductor Nanocrystallites, Journal of the American Chemical Society (1993) 115, 8706. (38). T. Mokari, U. Banin, S ynthesis and properties of Cd Se/ZnS core/shell nanorods, Chemistry of Materials (2003) 15, 3955. (39). B. O. Dabbousi, J. Rodr iguezViejo, F.V. Mikulec, J.R. Heine, H. Mattoussi, R. Ober, K.F. Jensen, M.G. Bawendi, (CdSe)ZnS core-shell quantum dots: Synthesis and characterization of a size series of highly luminescent nanocrystallites, Journal of Physical Chemistry B (1997) 101, 9463.

PAGE 136

136 (40). M. A. Hines, P. Guyot-Sionnest, Sy nthesis and characte rization of strongly luminescing ZnS-Capped CdSe nanocrystals, Journal of Physical Chemistry (1996) 100, 468. (41). Y. W. Cao, U. Banin, Synthesis and characterizati on of InAs/InP and InAs/CdSe core/shell nanocrystals, Angewandte Chemie-International Edition (1999) 38, 3692. (42). H. Lee, L. M. Hardison, V.D. Kleima n, P.H. Holloway, H. Yang, Synthesis and characterization of colloidal tern ary ZnCdSe semiconductor nanorods, Journal of Chemical Physics (2006) 125, 029901. (43). M. Dovrat, Y. Goshen, J. Jedrzejewski I. Balberg, A. Sa'ar, Radiative versus nonradiative decay processes in silicon nanocrystals probed by time-resolved photoluminescence spectroscopy, Physical Review B (2004) 69 155311. (44). X. G. Peng, Mechanisms for the Shap e-Control and Shape-E volution of Colloidal Semiconductor Nanocrystals Chem. Euro J (2002) 8, 334. (45). Z. A. Peng, X. G. Peng, Formation of high-quality Cd Te, CdSe, and CdS nanocrystals using CdO as precursor, Journal of the American Chemical Society (2001) 123, 183. (46). Z. A. Peng, X. G. Peng, Nearly monodisperse and shape-controlled CdSe nanocrystals via alternative r outes: Nucleation and growth, Journal of the American Chemical Society (2002) 124, 3343. (47). L. H. Qu, Z. A. Peng, X. G. Peng, Alternative routes toward high quality CdSe nanocrystals, Nano Letters (2001) 1, 333. (48). X. H. Zhong, M. Y. Han, Z. L. Dong, T. J. White, W. Knoll, Composition-tunable ZnxCd1-xSe nanocrystals with high luminescence and stability, Journal of the American Chemical Society (2003) 125, 8589. (49). X. Chen, A. Nazzal, D. Goorsk ey, M. Xiao, Z.A. Peng, X.G. Peng., Polarization spectroscopy of single CdSe quantum rods, Physical Review B (2001) 64, 245304. (50). J. T. Hu, L.S. Li, W. D. Yang, L. Manna, L.W. Wang, A.P.Alivisatos, Linearly polarized emission from colloid al semiconductor quantum rods, Science (2001) 292, 2060. (51). L. Manna, E. C. Scher, A. P. Alivis atos, Shape control of colloidal semiconductor nanocrystals, Journal of Cluster Science (2002) 13, 521. (52). L. S. Li, J. T. Hu, W. D. Yang, A. P. Alivisatos, Band gap variation of sizeand shape-controlled colloidal CdSe quantum rods, Nano Letters (2001) 1, 349. (53). M. B. Mohamed, C. Burda, M. A. El -Sayed, Shape dependent ultrafast relaxation dynamics of CdSe nanocrystal s: Nanorods vs nanodots, Nano Letters (2001) 1, 589.

PAGE 137

137 (54). L. Brus, Electronic wa ve-functions in semiconductor clusters Experiment and theory, Journal of Physical Chemistry (1986) 90, 2555. (55). G. Cantele, D. Ninno, G. Iadonisi, Conf ined states in ellipsoidal quantum dots, Journal of Physics-Condensed Matter (2000) 12, 9019. (56). G. Cantele, D. Ninno, G. Iadonisi, Calcul ation of the infrared optical transitions in semiconductor ellipsoidal quantum dots, Nano Letters (2001) 1, 121. (57). Y. Kayanuma, Wannier excitons in low-dimensiona l microstructures Shape dependence of the quantum size effect, Physical Review B (1991) 44, 13085. (58). S. Legoff, B. Stebe, Influence of longit udinal and lateral confinements on excitons in cylindrical quantum dots of semiconductors, Physical Review B (1993) 47, 1383. (59). D. Ninno, G. Iadonisi, F. Buonocore, Carrier localization and photoluminescence in porous silicon, Solid State Communications (1999) 112, 521. (60). A. D. Yoffe, Low-dimensional syst ems Quantum-size effects and electronicproperties of semiconductor microcrystallite s (zero-dimensional systems) and some quasi-2-dimensional systems, Advances in Physics (1993) 42, 173. (61). C. E. Tyner, Application of solar th ermal technology to the destruction of hazardous wastes, Solar Energy Materials (1990) 21, 113. (62). K. Hashizume, M. Vacha, T. Tani, Prep aration and optical prope rties of capped-CdSe nanocrystals, Journal of Luminescence (2000) 87-9, 402. (63). L. Manna, L. W. Wang, R. Cingolani, A. P. Alivisatos, First-principles modeling of unpassivated and surfactant-passivated bulk f acets of wurtzite CdSe. A model system for studying the anisotropic grow th of CdSe nanocrystals, Journal of Physical Chemistry B (2005) 109, 6183. (64). D. J. Norris, A. Sacra, C. B. Murra y, M. G. Bawendi, Measurement of the sizedependent hole spectrum in CdSe quantum dots, Physical Review Letters (1994) 72, 2612. (65). X. G. Peng, M. C. Schlam p, A. V. Kadavanich, A. P. Alivisatos, Epitaxial growth of highly luminescent CdSe/CdS core/shell na nocrystals with photostability and electronic accessibility, Journal of the Americ an Chemical Society (1997) 119, 7019. (66). H. Lee, University of Florida (2005). (67). D. F. Underwood, T. Kippeny, S. J. Ro senthal, Ultrafast carri er dynamics in CdSe nanocrystals determined by femtosecond fluorescence upconversion spectroscopy, Journal of Physical Chemistry B (2001) 105, 436.

PAGE 138

138 (68). T. W. Roberti, N. J. Cherepy, J. Z. Zhang, Nature of the power-dependent ultrafast relaxation process of photoexcited charge carriers in II-VI semiconductor quantum dots: Effects of particle size, su rface, and electronic structure, Journal of Chemical Physics (1998) 108, 2143. (69). B. Oregan, M. Gratzel, A low-cost, high-efficiency sola r-cell based on dye-sensitized colloidal TiO2 films, Nature (1991) 353, 737. (70). D. V. Talapin, A. L. Rogach, A. Kornowski, M. Haase, H. Weller, Highly luminescent monodisperse CdSe and CdSe /ZnS nanocrystals synthesized in a hexadecylamine-trioctylphosphine oxide-trioctylphospine mixture, Nano Letters (2001) 1, 207. (71). J. E. B. Katari, V.L. Colvin, and A. P. Alivisatos, X-Ray P hotoelectron-Spectroscopy of CdSe nanocrystals with applications to studies of the nanocrystal surface, Physical Chemistry (1994) 98, 4109. (72). A. Creti, M. Anni, M.Z. Rossi, G. Lanzani, G. Leo, F. Della Sala, L. Manna, M. Lomascolo., Ultrafast carrier dynamics in core and core/shell CdSe quantum rods: Role of the surface and interface defects, Physical Review B (2005) 72, 125346. (73). P. Reiss, J. Bleuse, A. Pron, Highly luminescent CdSe /ZnSe core/shell nanocrystals of low size dispersion, Nano Letters (2002) 2, 781. (74). Y. W. Cao, U. Banin, Growth and properties of semic onductor core/shell nanocrystals with InAs cores, Journal of the American Chemical Society (2000) 122, 9692. (75). J. Bleuse, S. Carayon, P. Reiss, Optical properties of core/multishell CdSe/Zn(S,Se) nanocrystals, Physica E-Low-Dimensional Systems & Nanostructures (2004) 21, 331. (76). D. C. Pan, Q. Wang, J. B. Pang, S.C. Jiang, X.L. Ji, L.J. An, Semiconductor "nanoonions" with multifold alternating CdS/CdSe or CdSe/CdS structure, Chemistry of Materials (2006) 18, 4253. (77). J. J. Li, J.T. Hu, W.D. Yang, A.P. Alivisatos, Large-scale synthesis of nearly monodisperse CdSe/CdS core/shell nanoc rystals using air-stable reagents via successive ion layer adsorption and reaction, Journal of the American Chemical Society (2003) 125, 12567. (78). L. P. Balet, S. A. Ivanov, A. Piryatin ski, M. Achermann, V. I. Klimov, Inverted core/shell nanocrystals continuously tunabl e between type-I and type-II localization regimes, Nano Letters (2004) 4, 1485. (79). D. V. Talapin, A.L. Rogach, A. Kornowski, H. Weller, CdSe/CdS/ZnS and CdSe/ZnSe/ZnS core-shell-shell nanocrystals, Journal of Physical Chemistry B (2004) 108, 18826.

PAGE 139

139 (80). W. E. Garner, Chemistry of the solid state (Butterworths Scientific Publications, London, 1955). (81). W. E. Martin, Photoluminescence De terminations of Cd Diffusion in ZnSe, Journal of Applied Physics (1973) 44, 5639. (82). P. J. Parbrook, B. Henderson, K. P. Odonnell, P. J. Wright, B. Cockayne, Interdiffusion in wide-bandgap Zn(Cd) S(Se) strained layer superlattices, Semiconductor Science and Technology (1991) 6, 818. (83). A. Rosenauer, T. Reisinger, E. Steink irchner, J. Zweck, W. Gebhardt, High-resolution transmission electron-microscopy determina tion of Cd diffusion in CdSe/ZnSe singlequantum-well structures, Journal of Crystal Growth (1995) 152, 42. (84). M. Strassburg, M. Kuttler, U. W. Pohl D. Bimberg, Diffusi on of Cd, Mg and S in ZnSe-based quantum well structures, Thin Solid Films (1998) 336, 208. (85). W. Chen, J. O. Malm, V. Zwiller, R. Wallenberg, J. O. Bovin, Size dependence of Eu2+ fluorescence in ZnS : Eu2+ nanoparticles, Journal of Applied Physics (2001) 89, 2671. (86). W. Chen, R. Sammynaiken, Y.N. Hu ang, Crystal field, phonon coupling and emission shift of Mn2+ in ZnS : Mn nanoparticles, Journal of Applied Physics (2001) 89, 1120. (87). C. X. Shan, X.W. Fan, J.Y. Zhang, Z. Z. Zhang, B.S. Li, Y.M Lu, Y.C. Liu, D.Z. Shen, X.G. Kong, X.H. Wang, Growth a nd evolution of ZnCdSe quantum dots, Journal of Vacuum Science & Technology B (2002) 20, 1102. (88). X. Y. Wang, J. Y. Zhang, A. Nazzal M. Darragh, M. Xiao, Electronic structure transformation from a quantum-dot to a quantum-wire system: Photoluminescence decay and polarization of colloidal CdSe quantum rods, Applied Physics Letters (2002) 81, 4829. (89). U. Manual, A User's Guide to the Andor iStar (Andor Technology Limited, 2001). (90). Andor, Andor K nowledge Library, http://www.andor.com/library/digital_cameras/ (91). J. Alford, Personal communication concering the iStar CCD system,(2007) (92). S. Cannistra, How to choose a CCD camera, http://www.starrywonders.com/ccdcameraconsiderations.html (93). M. C. Gino, Noise, Noise, Noise, http://www.astrophysassist.com/educate/noise/noise.htm (94). Apogee, CCD University, http://www.ccd.com/ccdu.html

PAGE 140

140 (95). E. C. Scher, L. Manna, A. P. Aliv isatos, Shape control and applications of nanocrystals, Philosophical Transactions of the Royal Society of London Series aMathematical Physical and Engineering Sciences (2003) 361, 241. (96). A. Puzder, A.J. Williamson, N. Zaitseva, G. Galli, L. Manna and A.P. Alivisatos, The effect of organic ligand bi nding on the growth of CdSe nanoparticles probed by Ab initio calculations, Nano Letters (2004) 4, 2361. (97). M. A. Hines, P. Guyot-Sionnest, Br ight UV-blue luminescent colloidal ZnSe nanocrystals, Journal of Physical Chemistry B (1998) 102, 3655. (98). B. D. Cullity, and Stock, S. R., Elements of X-ray Diffraction (Prentice Hall, New York, ed. Third, 2001). (99). M. M. a. F. Rashad, O. A., Synthesi s and characterization of nano-sized nickel ferrites from fly ash for cat alytic oxidation of CO, Materials Chemistry and Physics (2005) 94, 365. (100). R. W. Meulenberg, T. Je nnings, G. F. Strouse, Compre ssive and tensile stress in colloidal CdSe semiconductor quantum dots, Physical Review B (2004) 70 235311. (101). A. V. Baranov, Y.P. Rakovi ch, J.F. Donegan, T.S. Perova R.A. Moore, D.V. Talapin, A.L. Rogach, Y. Masumoto, I. Nabiev, Effect of ZnS shell thickness on the phonon spectra in CdSe quantum dots, Physical Review B (2003) 68, 165306. (102). Y. N. Hwang, S. Shin, H.L. Park, SH Pa rk, U. Kim, H.S. Jeong, E. Shin, D. Kim, Effect of lattice contraction on the Raman shifts of CdSe quantum dots in glass matrices, Physical Review B (1996) 54, 15120. (103). F. Comas, C. Trallero-Giner, N. Studa rt, G. E. Marques, Interface optical phonons in spheroidal dots: Raman selection rules, Physical Review B (2002) 65, 073303. (104). C. Trallero-Giner, A. Debernardi, M. Cardona, E. Me nendez-Proupin, A. I. Ekimov, Optical vibrons in CdSe dots and disp ersion relation of the bulk material, Physical Review B (1998) 57, 4664. (105). R. G. Alonso et al., Raman-spectroscopy of 2 novel semiconductors and related superlattices Cubic Cd1-XMnxSe and Cd1-XZnxSe, Physical Review B (1989) 40, 3720. (106). V. V. Travnikov, V. K. Kaibyshev, Resonance ex citon-phonon spectra in open ZnCdSe/ZnSe nanowires: Raman scattering and hot luminescence, extended and localized exciton states, Physics of the Solid State (2003) 45, 1379. (107). R. Bhushan, V. Prasad, W. Meredith G. Horsburgh, G.D. Brownlie, K.A. Prior, B.C. Cavenett, W. Rothwell, A.J. Dann, Microp robe Raman study of the variation of LO phonon frequency with the Cd concen tration in the ternary compound Zn1-xCdxSe, Journal of Crystal Growth (1996) 159, 103.

PAGE 141

141 (108). C. Ramkumar, K. P. Jain, S. C. Abbi Resonant Raman scatte ring probe of alloying effect in GaAs1-xPx ternary alloy semiconductors, Physical Review B (1996) 54, 7921. (109). C. Ramkumar, K. P. Jain, S. C. Abbi Raman-scattering probe of anharmonic effects due to temperature and compositional disord er in III-V binary and ternary alloy semiconductors, Physical Review B (1996) 53, 13672. (110). T. Kummell et al., Size dependence of strain relaxa tion and lateral quantization in deep etched CdxZn1-xSe/ZnSe quantum wires, Physical Review B (1998) 57, 15439. (111). P. R. Yu, J. M. Nedeljkovic, P. A. Ah renkiel, R. J. Ellingson, A. J. Nozik, Size dependent femtosecond electron cooling dynamics in CdSe quantum rods, Nano Letters (2004) 4, 1089. (112). M. Achermann, J. A. Hollingsworth, V. I. Klimov, Multiexcitons confined within a subexcitonic volume: Spectro scopic and dynamical signature s of neutral and charged biexcitons in ultrasmall semiconductor nanocrystals, Physical Review B (2003) 68, 245302. (113). Y. Kawakami, K. Omae, A. Kaneta, K. Okamoto, Y. Narukawa, T. Mukai, S. Fujita, In inhomogeneity and emission characteristics of InGaN, Journal of PhysicsCondensed Matter (2001) 13, 6993. (114). H. S. Kim, R. A. Mair, J. Li, J. Y. Lin, H. X. Jiang, Time-resolved photoluminescence studies of AlxGa1-xN alloys, Applied Physics Letters (2000) 76, 1252. (115). X. H. Zhong, Y. Y. Feng, W. Knoll, M. Y. Han, Alloyed ZnxCd1-xS nanocrystals with highly narrow luminescen ce spectral width, Journal of the American Chemical Society (2003) 125, 13559. (116). R. F. Mahrt, T. Pauck, U. Lemmer, U. Siegner, M. Hopmeier, R. Hennig, H. Bassler, E.O. Gobel, P.H. Bolivar, G. Wegmann, H. Kuz, U. Scherf, K.Mullen, Dynamics of optical excitations in a ladder-type pi-conj ugated polymer containi ng aggregate states, Physical Review B (1996) 54, 1759. (117). M. Jones, J. Nedeljkovic, R. J. Ellingson, A. J. Nozik, G. Rumbles, Photoenhancement of luminescence in co lloidal CdSe quantum dot solutions, Journal of Physical Chemistry B (2003) 107, 11346. (118). X. Chen, B. Henderson, K. P. Odonnell, Luminescence decay in disordered lowdimensional semiconductors, Applied Physics Letters (1992) 60, 2672. (119). R. Cingolani et al., Exciton spectroscopy in Zn1-XCdxSe/ZnSe quantum-wells, Physical Review B (1995) 51, 5176.

PAGE 142

142 (120). S. A. Empedocles, M. G. Bawendi, Quan tum-confined stark effect in single CdSe nanocrystallite quantum dots, Science (1997) 278, 2114. (121). V. Esch, B. Fluegel, G. Khitrova, H.M. Gibbs, J.J. Xu, K.Kang, S.W. Koch, L.C. Liu, S.H. Risbud, N. Peyghambarian, State filli ng, coulomb, and trapping effects in the optical nonlinearity of CdTe quantum dots in glass, Physical Review B (1990) 42, 7450. (122). E. Hendry, M. Koeberg, F. Wang, H. Zhang, C.D. Donega, D. Vanmaekelbergh, M. Bonn, Direct observation of electron-to-hol e energy transfer in CdSe quantum dots, Physical Review Letters (2006) 96, 125201. (123). A. Shabaev, A. L. Efros, 1D exc iton spectroscopy of semiconductor nanorods, Nano Letters (2004) 4, 1821. (124). C. Rulliere, Femtosecond Laser Pulses: Principles and Experiments, 2nd Edition (2004). (125). C. Burda, X. B. Chen, R. Narayanan, M. A. El-Sayed, Chemistry and properties of nanocrystals of different shapes, Chemical Reviews (2005) 105, 1025. (126). K. Brunner, U. Bockelmann, G. Abstreit er, M. Walther, G. Bohm, G. Trankle, G. Weimann, Photoluminescence from a single GaAs/AlGaAs quantum dot Physical Review Letters (1992) 69, 3216. (127). L. W. Wang, M. Califano, A. Zunger, A. Franceschetti Pseudopotential theory of Auger processes in CdSe quantum dots, Physical Review Letters (2003) 91, 56404. (128). A. L. Efros, V. A. Kharchenko, M. Rosen, Breaking the phonon bottleneck in nanometer quantum dots Role of Auger-like processes, Solid State Communications (1995) 93, 281. (129). P. Guyot-Sionnest, M. Shim, C. Matranga M. Hines, Intraband relaxation in CdSe quantum dots, Physical Review B (1999) 60, R2181. (130). J. G. Muller, E. Atas, C. Tan, K. S. Schanze, V. D. Kleiman, The role of exciton hopping and direct energy tr ansfer in the efficien t quenching of conjugated polyelectrolytes, Journal of the American Chemical Society (2006) 128, 4007. (131). V. I. Klimov, A. A. Mikhailovsky, D. W. McBranch, C. A. Leatherdale, M. G. Bawendi, Quantization of multiparticle Auge r rates in semiconductor quantum dots, Science (2000) 287, 1011. (132). J. L. Blackburn, D. C. Selmarten, A. J. Nozik, Electron transfer dynamics in quantum dot/titanium dioxide composites formed by in situ chemical bath deposition, Journal of Physical Chemistry B (2003) 107, 14154.

PAGE 143

143 (133). S. Sauvage, P. Boucaud, Rpsm Lobo, F. Bras, G. Fishman, R. Prazeres, F. Glotin, J.M. Ortega, J.M Gerard, Long polaron lif etime in InAs/GaA s self-assembled quantum dots, Physical Review Letters (2002) 88, 177402. (134). P. C. Sercel, Multiphonon-assisted tu nneling through deep levels A rapid energyrelaxation mechanism in nonideal quantum-dot heterostructures, Physical Review B (1995) 51, 14532. (135). M. Ohishi, K. Tanaka, T. Fujimoto, M. Yoneta, H. Saito, Alloying of CdSe/ZnSe quantum dot grown by an altern ate molecular beam supply, Journal of Crystal Growth (2002) 237, 1320. (136). X. H. Zhong, R. G. Xie, Y. Zhang, T. Basche, W. Knoll, High-quality violetto redemitting ZnSe/CdSe core/shell nanocrystals, Chemistry of Materials (2005) 17, 4038. (137). B. P. Zhang, D. D. Manh, K. Waka tsuki, Y. Segawa, Nanostructures formed on CdSe/ZnSe surfaces, Journal of Crystal Growth (2001) 227, 645. (138). V. V. Nikesh, S. Mahamuni, Highl y photoluminescent ZnSe/ZnS quantum dots, Semiconductor Science and Technology (2001) 16, 687. (139). J. S. Steckel, J.P. Zimmer, S. Coe-Sulliv an, N.E. Stott, V. Bulovic, M.G. Bawendi, Blue luminescence from (CdS)Z nS core-shell nanocrystals, Angewandte ChemieInternational Edition (2004) 43, 2154. (140). Z. H. Yu, L. Guo, H. Du, T. Krauss, J. Silcox, Shell distribution on colloidal CdSe/ZnS quantum dots, Nano Letters (2005) 5, 565. (141). C. Guenaud, E. Deleporte, A. Filoramo P. Lelong, C. Delalande, C. Morhain, E. Tournie, J.P. Faurie, Study of the band a lignment in (Zn, Cd)Se/ZnSe quantum wells by means of photoluminescence excitation spectroscopy, Journal of Applied Physics (2000) 87, 1863. (142). V. I. Kozlovsky, Sadofyev, Yu G., Litvinov, V. G., Crystal and Solid State Physics, Physics of II-VI Compounds, Landolt-Borstein Series (Springer, Berlin, 1982). (143). V. I. Kozlovsky, Y. G. Sadofyev, V. G. Litvinov, Band alignment in ZnCdTe/ZnTe and ZnCdSe/ZnSe SQW structures grown on GaAs(100) by MBE, Nanotechnology (2000) 11, 241. (144). S. H. Wei, S. B. Zhang, A. Zunger, Firs t-principles calculation of band offsets, optical bowings, and defects in CdS, CdSe, CdTe, and their alloys, Journal of Applied Physics (2000) 87, 1304. (145). S. H. Wei, A. Zunger, Calculated natural band offsets of all II-VI and Ill-V semiconductors: Chemical trends and the role of cation d orbitals, Applied Physics Letters (1998) 72, 2011.

PAGE 144

144 (146). M. D. McGehee, A. J. Heeger, Semi conducting (conjugated) pol ymers as materials for solid-state lasers, Advanced Materials (2000) 12, 1655. (147). R. H. Friend, R. W. Gymer, A.B. Holmes, J.H. Bu rroughes, R.N. Marks, C. Taliani, D.D.C. Bradley, D.A. Dos Santos, J.L. Bredas, M. Logdlund, W.R. Salaneck, Electroluminescence in conjugated polymers, Nature (1999) 397, 121. (148). H. Spanggaard, F. C. Krebs, A brief history of the devel opment of organic and polymeric photovoltaics, Solar Energy Materials and Solar Cells (2004) 83, 125. (149). H. Sirringhaus, N. Tessler, R. H. Friend, Integrated optoelectronic devices based on conjugated polymers, Science (1998) 280, 1741. (150). M. Fakis, Anestopoulos, D ., Giannetas, V., et al, Influe nce of aggregates and solvent aromaticity on the emission of conjugated polymers, Journal of Physical Chemistry B (2006) 49, 24897. (151). J. H. Hsu, W.S. Fann, H.F. Meng, E.S. Chen, E.C. Chang, S.A. Chen, K.W. To, Decay dynamics of interchain excited states in lumine scent conjugated polymer CNPPV, Chemical Physics (2001) 269, 367. (152). F. J. Hua, E. Ruckenstein, Fluorescen ce study of aggregation in water of PEO-grafted polydiphenylamine, Langmuir (2004) 20, 3954. (153). L. O. Palsson et al., Photophysics of a fluorene co-polymer in solution and films, Chemical Physics (2002) 279, 229. (154). P. Wang, C. J. Collison, L. J. Rot hberg, Origins of aggregation quenching in luminescent phenylenevinylene polymers, Journal of Photochemistry and Photobiology a-Chemistry (2001) 144, 63. (155). H. Jiang, X. Y. Zhao, K. S. Sc hanze, Amplified fluorescence quenching of a conjugated polyelectroly te mediated by Ca2+, Langmuir (2006) 22, 5541. (156). M. Fakis, Anestopoulos D., Giannetas, V., et al, Femtosecond time resolved fluorescence dynamics of a cationic wate r soluble poly(fluorenevinylene-cophenylenevinylene), Journal of Physical Chemistry B (2006) 110, 12916. (157). L. H. Chen, D.W. McBranch, H.L. Wa ng, R. Helgeson, F. Wudl, D.G. Whitten, Highly sensitive biological and chemical sensors based on reversible fluorescence quenching in a conjugated polymer, Proceedings of the National Academy of Sciences of the United States of America (1999) 96, 12287. (158). N. GronbechJensen, R. J. Mashl, R. F. Bruinsma, W. M. Gelbar t, Counterion-induced attraction between rigi d polyelectrolytes, Physical Review Letters (1997) 78, 2477.

PAGE 145

145 (159). B. J. Schwartz, Conjugated polymer s as molecular materials : How chain conformation and film morphology influe nce energy transfer and interchain interactions, Annual Review of Physical Chemistry (2003) 54, 141. (160). R. Jakubiak, C. J. Collison, W. C. Wan, L. J. Rothberg, B. R. Hsieh, Aggregation quenching of luminescence in electro luminescent conjugated polymers, Journal of Physical Chemistry A (1999) 103, 2394. (161). S. A. Jenekhe, J. A. Osaheni, Excime rs and exciplexes of conjugated polymers, Science (1994) 265, 765. (162). I. D. W. Samuel, G. Ru mbles, C. J. Collison, Efficient interchain photoluminescence in a high-electron-affi nity conjugated polymer, Physical Review B (1995) 52, 11573. (163). C. J. Collison, L. J. Rothberg, V. Tr eemaneekarn, Y. Li, Conformational effects on the photophysics of conjugated polymers: A two species model for MEH-PPV spectroscopy and dynamics, Macromolecules (2001) 34, 2346. (164). G. H. Gelinck, J. M. Warman, E. G. J. St aring, Polaron pair form ation, migration, and decay on photoexcited poly(phenylenevinylene) chains, Journal of Physical Chemistry (1996) 100, 5485. (165). J. W. Blatchford, S.W. Jessen, L.B. Lin, T.L, Gust afson, D.K. Fu, H.L. Wang, T.M. Swager, A.G. MacDiarmid, A.J. Epstein, Photoluminescence in pyridine-based polymers: Role of aggregates, Physical Review B (1996) 54, 9180. (166). U. Lemmer, S. Heun, R.F. Mahrt, U. Scherf M. Hopmeier, U. Siegner, E.O. Gobel, K. Mullen, H. Bassler, Aggregate Fluo rescence in Conjugated Polymers, Chemical Physics Letters (1995) 240, 373. (167). X. Y. Zhao, M.R. Pinto, L.M. Hardison, J. Mwuara, J. Muller, H. Jiang, D.Witker, V.D. Kleiman, J.R. Reynolds, K.S. Sc hanze,Variable band gap poly(arylene ethynylene) conjugated polyelectrolytes, Macromolecules (2006) 39, 6355. (168). C. Y. Tan, E. Atas, J.G. Muller, M.R. Pinto, V.D. Kleiman, K.S. Schanze Amplified quenching of a conjugated polyelectrolyte by cyanine dyes, Journal of the American Chemical Society (2004) 126, 13685. (169). B. S. Harrison, M. B. Ramey, J. R. Reynolds, K. S. Schanze, Amplified fluorescence quenching in a poly(p-phenylene)-b ased cationic polyelectrolyte, Journal of the American Chemical Society (2000) 122, 8561. (170). G. H. Gelinck, E.G.J. Staring, D.H. Hwang, G.C.W. Spencer, A.B. Holmes, J.M. Warman, The effect of broken conjugation and aggregation on photo-induced charge separation on polyphenylenevinylene chains, Synthetic Metals (1997) 84, 595. (171). C. H. Fan, S.Wang, J.W. Hong, G. C. Bazan, K.W. Plaxco, A.J. Heeger, Beyond superquenching: Hyper-efficient energy tr ansfer from conjugated polymers to gold

PAGE 146

146 nanoparticles, Proceedings of the National Academy of Sciences of the United States of America (2003) 100, 6297. (172). B. S. Gaylord, A. J. Heeger, G. C. Bazan, DNA detection using water-soluble conjugated polymers and peptide nucleic acid probes, Proceedings of the National Academy of Sciences of th e United States of America (2002) 99, 10954. (173). M. Stork, B. S. Gaylord, A. J. Heeger, G. C. Bazan, Energy transfer in mixtures of water-soluble oligomers: Effect of charge, aggregation, and surfactant complexation, Advanced Materials (2002) 14, 361. (174). A. Haugeneder, U. Le mmer, U. Scherf, Exciton dissociation dynamics in a conjugated polymer contai ning aggregate states, Chemical Physics Letters (2002) 351, 354. (175). G. Petekidis, G. Fytas, U. Scherf, K. Mullen, G. Fleischer, Dynamics of poly(pphenylene) ladder polymers in solution, Journal of Polymer Science Part B-Polymer Physics (1999) 37, 2211. (176). J. G. Muller, U. Lemmer, G. Raschke, M. Anni, U. Scherf, J.M. Lupton, J. Feldman, Linewidth-limited energy transfer in single conjugated polymer molecules, Physical Review Letters (2003) 91, 267403. (177). B. Schweitzer, G. Wegmann, D. Hertel, R.F. Mahrt, H. Ba ssler, F. Uckert, U. Scherf, K. Mullen, Spontaneous and stimulated emission from a ladder-type conjugated polymer, Physical Review B (1999) 59, 4112. (178). H. P. Gregor, L. B. Luttinger, E. M. Loebl, Metal-polyelectrolyte complexes .4. complexes of polyacrylic acid with magnesi um, calcium, manganese, cobalt and zinc, Journal of Physical Chemistry (1955) 59, 990. (179). I. B. Kim, U. H. F. Bunz, Modulating the sensory response of a conjugated polymer by proteins: An agglutination assa y for mercury ions in water, Journal of the American Chemical Society (2006) 128, 2818. (180). C. Y. Tan, M. R. Pint o, K. S. Schanze, Photophysic s, aggregation and amplified quenching of a water-soluble pol y( phenylene ethynylene), Chemical Communications (2002), 5, 446. (181). X. Y. Zhao, Jiang, H., Schanze, K.S., Water-Soluble Poly(phenylene ethynylene)s of Variable Chain Length: Synthesis, Photophysics and Amplified Quenching, In press (2007). (182). E. Atas, University of Florida (2006). (183). E. Atas, Z. H. Peng, V. D. Kleiman, Energy transfer in unsymmetrical phenylene ethynylene dendrimers, Journal of Physical Chemistry B (2005) 109, 13553.

PAGE 147

147 (184). New Focus, Model 5540 User's Manual: The Berek Polarization Compensator. (185). New Focus, Polarization and Polarization Control. (186). A. Yariv, and Yeh, P., Optical Waves in Crystals: Propagation and Control of Laser Radiation (Wiley-Interscience 2002). (187). R. Rajasekaran, Personal communication for Berek Compensator Crystal Specifications, (2007) (188). M. Dodge, Refractive prope rties of magnesium fluoride, Applied Optics (1984) 23, 1980. (189). J. R. Lakowicz, Principles of Fluorescence Spectroscopy (Kluwer Academic/Plenum Publishers, New York, ed. 2, 1999). (190). R. E. Martin, F. Diederich, Linear monodisperse pi-conjugated oligomers: Model compounds for polymers and more, Angewandte Chemie-International Edition (1999) 38, 1350. (191). U. H. F. Bunz, J.M. Im hof, R.K. Bly, C.G. Bangcuyo, L. Rozanski, D.A.V. Bout, Photophysics of poly [p-(2,5-didodecylphe nylene)ethynylene] in thin films, Macromolecules (2005) 38, 5892. (192). C. E. Halkyard, M. E. Rampey, L. Kl oppenburg, S. L. Studer-Martinez, U. H. F. Bunz, Evidence of aggregate formation fo r 2,5-dialkylpoly(pphenyleneethynylenes) in solution and thin films, Macromolecules (1998) 31, 8655. (193). J. Kim, D. T. McQuade, S. K. McHugh, T. M. Swager Ion-specific aggregation in conjugated polymers: Highly se nsitive and selective fluorescent ion chemosensors, Angewandte Chemie-International Edition (2000) 39, 3868. (194). J. Kim, T. M. Swager, Control of conformational and interpolymer effects in conjugated polymers, Nature (2001) 411, 1030. (195). M. Abramowitz, Johnson I. D., and Davidson, M. W., Fluorescence filter spectral transmission profiles, http://www.olympusmicro.com/primer/java/fluorescence/fluorocubes/index.html, (196). V. F. Kamalov, I. A. Struganova, K. Yoshihara, Temperature dependent radiative lifetime of J-aggregates, Journal of Physical Chemistry (1996) 100, 8640. (197). D. Anestopoulos, Fakis, M etal, Excitation energy transfer in a cationic water-soluble conjugated co-polymer studied by time resolved anisot ropy and fluorescence dynamics, Chemical Physics Letters (2006) 421, 205.

PAGE 148

148 (198). A. A. Mikhailovsky, A. V. Malko, J. A. Hollingsworth, M. G. Bawendi, V. I. Klimov, Multiparticle interactions and stimulated emission in chemically synthesized quantum dots, Applied Physics Letters (2002) 80, 2380. (199). V. I. Klimov, Mechanisms for photogenera tion and recombination of multiexcitons in semiconductor nanocrystals: Implications for lasing and solar energy conversion, Journal of Physical Chemistry B (2006) 110, 16827. (200). D. Katz, T. Wizansky, O. Millo, E. Ro thenber, T. Mokari, U. Banin, Size-dependent tunneling and optical spectroscopy of CdSe quantum rods, Physical Review Letters (2002) 89, 086801. (201). H. Htoon, J. A. Hollingworth, A. V. Ma lko, R. Dickerson, V. I. Klimov, Light amplification in semiconductor nanocrysta ls: Quantum rods versus quantum dots, Applied Physics Letters (2003) 82, 4776. (202). S. A. Ivanov, J. Nanda, A. Piryatinski, M. Achermann, L.P. Balet, I.V. Bezel, P.O. Anikeeva, S. Tretiak, V.I. Klimov, Light amplification using inverted core/shell nanocrystals: Towards lasing in the single-exciton regime, Journal of Physical Chemistry B (2004) 108, 10625. (203). J. Nanda, S.A. Ivanov, H. Htoon, I. B ezel, A Piratinski, S. Tretiak, V.I. Klimov, Absorption cross sections and Auger recombination lifetimes in inverted core-shell nanocrystals: Implications for lasing performance, Journal of Applied Physics (2006) 99, 013707. (204). V. L. Colvin, M. C. Schlamp, A. P. Alivisatos, Light-emit ting-diodes made from cadmium selenide nanocrystals and a semiconducting polymer Nature (1994) 370, 354. (205). B. O. Dabbousi, M. G. Bawendi, O. Onitsuka, M. F. Rubner, Electroluminescence from CdSe quantum-dot polymer composites, Applied Physics Letters (1995) 66, 1316. (206). H. Mattoussi, L.H. Radz ilowski, B.O. Dabbousi, E.L. Thomas, M.G. Bawendi, M.F. Rubner, Electroluminescence from heterost ructures of poly(phenylene vinylene) and inorganic CdSe nanocrystals, Journal of Applied Physics (1998) 83, 7965. (207). H. S. Yang, P. H. Holloway, Electro luminescence from hybrid conjugated polymer CdS : Mn/ZnS core/shell nanocrystals devices, Journal of Physical Chemistry B (2003) 107, 9705. (208). S. Coe, W. K. Woo, M. Bawendi, V. Bulovic, El ectroluminescence from single monolayers of nanocrystals in molecular organic devices, Nature (2002) 420, 800. (209). C. Bonati, M.B. Mohamed, D. Tonti, G. Zgreblic, S. Haacke, F. van Mourik, M. Chergui, Spectral and dynamical characterization of multiexcitons in colloidal CdSe semiconductor quantum dots, Physical Review B (2005) 71, 205317.

PAGE 149

149 (210). C. Burda, M. A. El-Sayed, Highdensity femtosecond transient absorption spectroscopy of semiconductor nanoparticles. A tool to investigate surface quality, Pure and Applied Chemistry (2000) 72, 165. (211). C. Burda, S. Link, M. Mohamed, M. El-Sayed, The relaxation pathways of CdSe nanoparticles monitored with femtosecond timeresolution from the visible to the IR: Assignment of the transient f eatures by carrier quenching, Journal of Physical Chemistry B (2001) 105, 12286. (212). B. P. Zhang, Y.Q. Li, T. Yasuda, Y. Segawa, K.Edamatsu, T. Itoh, Time-resolved photoluminescence of ZnCdSe single quantum dots, Journal of Crystal Growth (2000) 214, 765. (213). N. C. Greenham, X. G. Peng, A. P. Alivisatos, Charge separation and transport in conjugated-polymer/semiconductor-nanocrystal composites studied by photoluminescence quenching and photoconductivity, Physical Review B (1996) 54, 17628. (214). N. C. Greenham, X. G. Peng, A. P. Alivisatos, Charge separation and transport in conjugated polymer cadmium selenide nanocrystal composites studied by photoluminescence quenching and photoconductivity, Synthetic Metals (1997) 84, 545. (215). L. M. Herz, C. Silva, R. T. Phillips, S. Setayesh, K. Mullen, Exciton migration to chain aggregates in conjuga ted polymers: influence of side-chain substitution, Chemical Physics Letters (2001) 347, 318.

PAGE 150

150 BIOGRAPHICAL SKETCH Lindsay Michelle Hardison was born on July 3, 1979, in Torrejon, Spain, where her father, Craig Hardison was stationed as a member of the United States Air Force. She lived there with her father and mother, Susan, until she was 16 m onths old. She then moved to Washington until she started elementary school. Lindsays father was relocated to Hampton, Virginia, where she continued her early academic studies. After attending Hampt on Christian High School for 3 years, Lindsay moved to Melbourne, Florida, in 1997 to begin her undergra duate studies at the Florida Institute of Technology. In 2001, she graduated magna cum laude with a Bachelor of Science degree in research chemistry. She then took a 1-year break from school and worked as an analytical chemist at Midwest Research Institute in Palm Bay. Lindsays drive to continue her education and desire to lear n brought her to the Un iversity of Florida in 2002. She began her doctoral work under the supervision of Professor Valeria D. Kleiman in the area of ultrafast laser spectroscopy of semiconductor na noparticles and conjugated polym ers. Her professional career as a Ph.D. will begin in Hillsboro, Oregon, as a t echnologies development engineer at Intel.