Đang tải... (xem toàn văn)
.. .DESIGN AND PREPARATION OF NANOSTRUCTURES FOR APPLICATIONS NEO MIN SHERN (B.Sc.(Hons.), NUS) A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF CHEMISTRY NATIONAL... based on the bandgap size and relative position of the valence band and conduction band energy levels of the core and shell components Figure 1.5 Schematic diagram of the different type of core-shell... PbS and nm PbS/CdS, (c-d) 6nm PbS and nm PbS/CdS, and (e-f) nm PbS and nm PbS/CdS 111 4.2 XRD patterns of (a) nm PbS and nm PbS/CdS; (b) nm PbS and nm PbS/CdS; (c) nm PbS and nm PbS/CdS Standard
DESIGN AND PREPARATION OF NANOSTRUCTURES FOR APPLICATIONS NEO MIN SHERN NATIONAL UNIVERSITY OF SINGAPORE 2009 DESIGN AND PREPARATION OF NANOSTRUCTURES FOR APPLICATIONS NEO MIN SHERN (B.Sc.(Hons.), NUS) A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF CHEMISTRY NATIONAL UNIVERSITY OF SINGAPORE 2009 Acknowledgement I would like to express my gratitude to the many people who have made this thesis possible. Firstly, I want to thank my PhD supervisor, Associate Professor Chin Wee Shong for her advice and understanding throughout my PhD study. I also wish to thank my co-supervisor Associate Professor Sow Chorng Haur for his help in the physical aspects of the nanomaterials and for constant advice. I also want to specially thank Madam Liang Eping and Dr Xu Hairuo for their monumental help in all aspects of my PhD studies from experimental techniques to emotional relief. I also wish to thank Dr Zhang Zhihua and Dr Lim Wen Pei and Dr Liu Chenmin for being my important mentors. Special thanks to our collaborators, Dr Venkatram Nalla and his supervisor Professor JI Wei from the Department of Physics for their help in the nonlinear optics measurements and publication. Special thanks to our collaborators, Dr Yan Yongli and his supervisor Assistant Professor Xu Qing-Hua for their help in the transient absorption studies. Special thanks to our collaborators, Dr Han Yi-Fan, Dr Zhong Ziyi at Applied catalysis group, Institute of Chemical and Engineering Sciences for their help in the catalysis measurements. I want to sincerely thank Mr Li Guangshuo for the polymer film synthesis and our constant discussions. I also wish to thank, Ms Lee Ya Ling, Ms Tan Zhi Yi for i their tremendous help in the MnOx and PbS/Polystyrene synthesis. Also thanks to Mr Rajiv Ramanujam Prabhakar for his help in conductive-AFM measurement. Special mentions are also given to Dr Liu Binghai, Ms Tang Chui Ngoh and Dr Tian Lu for their assistance in the TEM and SEM measurements. I also appreciate the help from all the other related staffs in the Department of Chemistry, Department of Physics and Department of Biological Sciences in the characterizations of my samples. Furthermore, I would like to thank my seniors, Dr Ang Thiam Peng, Dr Kerk Wai Tat and Dr Yin Fenfang for their guidance throughout this project. Thank all my group members, Ms Shawn Low, Ms Teo Tingting Sharon, Ms Loh Pui Yee, Ms Tan Zhi Yi, Mr Huang Baoshi Barry, Mr Khoh Rong Lun, Mr Low Jia En, Mr Lee Kian Keat, Dr Binni Varghese and Mr Lim Zhi han for their support and for making my days in the lab always enjoyable. The National University of Singapore (NUS) is gratefully acknowledged for supporting this project and my Graduate Research Scholarship. Finally, I would like to express my heartfelt gratitude to my parents and my brother for their understanding. ii ii Table of Contents Summary.............................................................................................................viii List of Publications ..............................................................................................xi List of Tables ...................................................................................................... xii List of Figures ................................................................................................... xiv List of Symbols................................................................................................. xxiv List of Abbreviation......................................................................................... xxvi Chapter 1 Introduction 1 1.1 Size-controlled preparation of nanomaterials 1 1.2 Shape-controlled synthesis of nanomaterials 5 1.2.1 Chemical synthesis of nanoparticles in solution 5 1.2.2 Template synthesis of nanowires 9 1.3 Core-shell quantum dots 14 1.4 Nonlinear optical limiting properties 18 1.4.1 Nonlinear scattering (NLS) 19 1.4.2 Free carrier absorption (FCA) 22 1.4.3 Multi-photon absorption (MPA) 23 1.4.4 Nonlinear Refraction (NLR) 25 1.5 Ultrafast electron and lattice dynamics 28 1.5.1 Carrier Excitation 29 1.5.2 Thermalization 31 iii 1.5.3 Carrier removal 1.5.4 Transient absorption (TA) 33 34 1.6 Nanoparticles in catalysis 34 1.7 Objective and scope of thesis 36 1.8 References 40 Chapter 2 Experimental 49 2.1 Chemical reagents used 49 2.2 Synthesis procedures 51 2.2.1 Preparation of nanosized PbS 51 2.2.2 Preparation of styrene pre-polymer 52 2.2.3 Preparation of PbS/PS nanocomposites 52 2.2.4 Preparation of CdS/PMMA nanocomposites 52 2.2.5 Preparation of heterogenous layered nanocomposites 54 2.2.6 56 Imprinting-thermal cross-linking QD-polymer film 2.2.7 Preparation of PbS/CdS core-shell nanoparticles 57 2.2.8 Electrochemical preparation of PbS and composite nanowires 59 2.2.9 Preparation of MnO Nanocrystals 60 2.2.10 Preparation of Mn3O4 nanorods 61 2.2.11 Preparation of MnO2 Nanocrystals 62 2.2.12 Preparation of nanosized Mn2O3 63 2.3 Characterization techniques 64 iv 2.3.1 Transmission Electron Microscopy (TEM) and High Resolution TEM (HRTEM) 64 2.3.2 Scanning Electron Microscopy (SEM) 64 2.3.3 Powder X-ray Diffraction (XRD) 65 2.3.4 Elemental Analysis (EA) 65 2.3.5 Ultraviolet-visible (UV-VIS-NIR) Absorption Spectroscopy 65 2.3.6 Steady-state Photoluminescence Spectroscopy (PL) 66 2.3.7 Profile meter 67 2.38 Thermal Gravimetric Analysis (TGA) 68 2.3.9 Fourier Transform Infrared (FT-IR) 68 2.3.10 Atomic force Microscopy 68 2.4 Nonlinear and transient optical studies 69 2.4.1 Z-scan technique 69 2.4.2 Transient absorption studies 71 2.5 CO oxidation catalytic studies 73 2.6 References 75 Chapter 3 Preparation of CdS and PbS polymer composite films and the study of their luminescence and nonlinear optical properties 3.1 The PbS/PS and CdS/PMMA composite films 76 78 3.1.1 PbS/PS composite films 78 3.1.2 CdS/PMMA composite films 85 3.2 Layered Nanocomposite films 87 v 3.3 Nonlinear optical properties of PbS in hexane and in polymer films 93 3.4 Summary 105 3.5 References 106 Chapter 4 Synthesis of PbS/CdS core-shell QDs and their nonlinear optical properties 108 4.1 Synthesis and characterization of core-shell PbS/CdS 110 4.2 Z-scan study for PbS/CdS QDs in hexane and PS polymer films 122 4.2.1 Effect of surface treatment on nonlinear scattering 126 4.22 Influence of film thickness on optical limiting 129 4.3 Femntosecond pump-probe transient absorption study of PbS and PbS/CdS QDs 132 4.3.1 Transient Absorption of PbS QDs 132 4.3.2 Transient Absorption of core-shell PbS/CdS QDs 135 4.4 Summary 145 4.5 References 146 Chapter 5 Fabrication of PbS and Metal/PbS Core-shell Nanowires via Electrochemical Methods 150 5.1 Deposition of pure PbS NWs 152 5.2 Deposition of metal (Au, Cu)/PbS core-shell NWs 155 5.3 Conductive AFM measurements 165 5.4 Summary 166 5.5 References 166 vi Chapter 6 Synthesis of MnO, Mn3O4, Mn2O3 and MnO2 nanocrystals and their catalytic studies 6.1 Synthesis of MnO nanocrystals and their catalytic studies 6.1.1 Synthesis of MnO nanoparticles 169 170 172 6.1.2 Size control: Oleic acid concentration 177 6.1.3 Size control: Temperature 180 6.14 Shape control 182 6.1.5 Catalytic activity of MnO 6.2 Synthesis of Mn3O4 nanocrystals and their catalytic studies 189 197 6.2.1 Synthesis of Mn3O4 nanocrystals 198 6.2.2 Catalytic activity of nanocrystalline Mn3O4 209 6.3 Synthesis, characterization and catalytic activity of nanocrystalline MnO2 213 6.4 Catalytic studies of nanocrystalline Mn2O3 226 6.5 Comparison of the catalytic activities of different manganese oxides 232 6.6 References 237 Chapter 7 Conclusions and Outlook 7.1 References Appendices 243 248 249 A Integrated PL intensity versus absorbance for (a) 5.0nm PbS and (b) IR125 dye. 249 B Integrated PL intensity versus absorbance for (a) 5.0nm PbS/CdS and (b) 6.0nm PbS/CdS and (c) IR125 dye. 249 vii Summary Summary This thesis reports the synthesis and investigation on several nanomaterials with different functional properties that can be tuned by changing the particle size, shape or composition. In Chapter 3, we developed two methods (drop casted and imprinting-thermal cross-linking) for the formation of single layer and multilayer polystyrene and poly(methylmethacrylate) thin films uniformly embedded with CdS or PbS QDs of varying sizes. Using PbS QDs of different sizes, the drop casted layered nanocomposites revealed interesting PL properties dependent on the orientation of excitation. Multilayer CdS/PMMA-PbS/PS prepared by drop casted method revealed PL that was dependent on the thickness of the PbS layer and the orientation of the excitation source. Micrometer thickness PbS/PS thin films were also successfully prepared by a novel imprinting-thermal cross-linking method. The PbS QDs sizes and NIR luminescence properties were preserved in these films. The NLO responses of these films were studied with femtosecond Zscan technique and compared to the solution phase properties. The free-carrier absorption cross-section, free-carrier refraction cross-section, and optical Kerr nonlinearity determined at laser excitation wavelength of 780 nm were found to be inversely related to the particle size. Nonlinear scattering was found to play an important role especially for larger particles in the solution study at high excitations. viii Summary In Chapter 4, we synthesized different sizes of core-shell PbS/CdS QDs using a cationic exchange method and studied the changes in their absorption and luminescence. The nonlinear properties of core-shell QDs in hexane and polymer films were studied by either transient absorption and/or Z-scan technique and compared to the results for pure PbS obtained in Chapter 3. We also studied the effect of free surface ligands and thickness of the polymer film on their optical limiting properties. The influence of the excitation intensity, pump wavelength and probe delay time on the transient differential transmittance spectra and relaxation kinetics of the PbS and core-shell PbS/CdS QDs were studied in details. Chapter 5 presents the formation of PbS nanowires using potentiostatic or cyclic electrochemical deposition into the channels of anodized alumina template. Different segmented and core-shell metal/PbS nanowires were fabricated using a step-wise pore widening deposition method and a study of the mechanism for their formation was carried out. Deposition of copper as core wires led to the formation of composite nanowires containing an intermediate layer of copper sulfide. CV deposition gave rise to a Cu core wire that was covered with a layer of copper sulfide along its entire length and with PbS deposited along the upper 34 μm length. The constant potential method produced a more regular Cu/PbS core-shell structure. The conductivity of the PbS nanowires were studied using conductive AFM analysis. ix Summary Chapter 6 reports our synthesis of a series of different phases of nanocrystalline MnOx to study their applications as catalysts for the oxidation of CO in comparison with their bulk counterparts. We explored in details the crystal growth of MnO and Mn3O4 nanocrystals. By varying the temperature, monomer concentration and reaction time, the resultant nanoparticles could have shapes ranging from cubic, spherical nanocrystals, long nanorods and rice shaped nanocrystals. The nanocrystalline MnO, α-MnO2, γ-MnO2, α-Mn2O3 and Mn3O4 catalysts were characterized using XRD, TEM, TGA, FT-IR and BET. This allowed the study of the influence of phase, structure and surface area on their catalytic properties. The intrinsic reaction rates and apparent activation energies were compared to the bulk MnOx. The phase transformations of the MnOx nanocrystals during storage and during the CO oxidation process were determined to be important factors in influencing its catalytic activity. Surface area was found to be less important in influencing the overall catalytic activity in comparison to the phase, structure and stability of the MnOx. x List of Publications Size-Dependent Optical Nonlinearities and Scattering Properties of PbS Nanoparticles , M.S. Neo, N. Venkatram, G.S. Li, W.S. Chin, and Ji Wei , J. Phys. Chem. C 2009, 113, 19055. Fabrication of PbS and metal/PbS core/shell and composite nanowires, M.S. Neo, R.P. Rajiv, C.H. Sow, and W.S. Chin , Submitted for publication. Synthesis of PbS/CdS core-shell QDs and their nonlinear optical properties, M.S. Neo, N. Venkatram, G.S. Li, W.S. Chin, and Ji Wei, manuscript in preparation. xi List of Tables 2.1 Chemical and solvents used in the work described in this thesis; their purity and sources. 49 2.2 Sizes of core-shell QDs and the corresponding reaction time 58 2.3 Pre-treatment temperatures of nanocatalysts 74 3.1 Absorption and emission wavelength maxima of PbS nanoparticles in 84 solvent and in PbS/PS composite films. 3.2 CdS/PMMA bandgap emission, PL shift and the luminescent enhancement with excitation at 375 nm. 87 3.3 A summary of the average sizes of PbS QDs estimated from TEM, the respective first excitonic position and bandgap luminescence measured in hexane. 96 3.4 Fitted FCA cross-section (σc), nonlinear refractive index (n2), FCR 101 cross-section (σr), and scattering coefficient (αS) of the PbS QDs in solution. (All values listed are within 10% error). 3.5 Fitted FCA cross-section (σc), nonlinear refractive index (n2), FCR cross-section (σr) of the PbS QDs in PS films (All values listed are within 10% error). 101 4.1 Average sizes of PbS QDs and PbS/CdS core-shell QDs obtained at the optimum reaction time. 112 4.2 Elemental ratios of the different elements from ICP-OES, estimated sizes of PbS core and CdS shell from Eqn 4.1, and the first excitonic peak of the different PbS/CdS core-shell QDs. 115 4.3 A comparison of the NIR absorption PL peak positions, Stoke shifts and QY for the various PbS QDs and PbS/CdS core-shell QDs prepared 116 4.4 Fitted FCA cross-section (σc), nonlinear refractive index (n2), FCR cross-section (σr), and scattering coefficient (αS) of the PbS/CdS core-shell QDs in solution compared to PbS QDs studied in Chapter 3. 126 4.5 Sizes of QDs and scattering coefficient (αS) of the PbS/CdS QDs in 128 xii Sample I and II. 4.6 Size, thickness of films, fitted FCA cross-section (σc), nonlinear refractive index (n2), and FCR cross-section (σr) of the prepared PbS/CdS QDs in PS films 131 4.7 Intensity of the pump; fast and slow lifetime relaxation dynamics of different core-shell PbS/CdS QDs and referenced PbS QDs. 139 5.1 Average length of PbS NWs obtained using different total CV 154 deposition time. 5.2 Relative elemental ratio of Pb, Cu and S detected from EDX analysis 158 on areas I and II in Figure 5.6(a) and areas A, B and C in Figure 5.7(a). 6.1 The average size and shape of MnO produced using different reaction 176 conditions. 6.2 Main peaks in the IR spectrum of MnO (a) as prepared; (b) heated in 194 air; (c) after pretreatment at 200oC in N2 and (d) after pretreatment at 300oC in N2. 6.3 Major peaks observed in the IR spectrum of Mn3O4 prepared with 200 different methods. 6.4 The particle sizes and morphologies of Mn3O4 obtained at different 202 reaction conditions. 6.5 FT-IR results of nanocrystalline Mn2O3 and reference Mn2O3. 227 6.6 Light-off temperatures of the nanocrystalline and bulk MnOx. 233 6.7 Apparent activation energies of nanocrystalline and bulk MnOx 235 Apparent activation energies of nanocrystalline and bulk MnOx. xiii List of Figures 1.1 Sketch of the solubility product [Cd][Se] as a function of temperature. 2 Solid line: thermodynamic curve for the equilibrium between the monomers Cd-TOPO and Se-TOP and a macroscopic CdSe crystal. Dashed line: the solubility product for the equilibrium between the monomers and the critical nuclei (CdSe)c indicative of supersaturation. The points indicate: nucleation (1), cooling (1–2), and growth of the nuclei at two different temperatures (3 and 3’). 1.2 The LaMer model for monodispersed particle formation (Cs:solubility; 3 C*min: minimum concentration for nucleation; C*max: maximum concentration for nucleation; I: prenucleation period; II: nucleation period; III: growth period). 1.3 (a) Kinetic shape control at high growth rate. The high-energy surface 7 grows more quickly than low energy surface in a kinetic regime. (b) Kinetic shape control through selective adhesion. (c) Complex shaped NCs of CdSe and TiO2 can be formed by sequential elimination of a high-energy facet. (d) High monomer concentration coupled with presence of two different crystal structure allowed the growth of branched nanostructures like tetrapods. 1.4 Fabrication process of thin film of p-CdTe on highly ordered, single- 10 crystalline nanopillars of n-CdS. 1.5 Schematic diagram of the different type of core-shell QDs. The upper 14 and lower edges of the rectangles correspond to the positions of the conduction- and valence-band edge of the core (center) and shell materials, respectively. 1.6 (a) Scattering intensity versus angle from a Rayleigh scatterer; (b) 20 scattering intensity versus angle from a Mie type scatterer. 1.7 Schematic representation of z-scan results for a negative refractive 27 nonlinearity (dashed curve) and a positive refractive nonlinearity (dotted curve). Both curves have been corrected for absorption. The solid curve shows the result of removing the aperture from the measurement apparatus and collecting all the transmitted light, thus isolating the nonlinear absorption. xiv 1.8 (a) Single photon and multiphoton excitation, (b) free carrier absorption 30 (Intraband absorption); (c) multiexciton generation (Impact ionization); (d) carrier-carrier scattering (e) carrier–phonon intravalley scattering; (f) carrier–phonon intervalley scattering, (g) Auger-like electron-hole cooling; (h) surface trapped electron cooling via intermediate trap states; (i) electron cooling via resonant high-frequency vibration coupling; (j) radiative recombination; (k) Auger recombination; (l) trapping to surface/defect states, and (m) diffusion of excited carriers. 1.9 Typical photoinduced absorption changes in a NC sample which allow 35 the exciton population dynamics to be obtained. 2.1 Schematic diagram of layered PbS-PS nanocomposites studied in Chapter 3. 54 2.2 Schematic diagram of layered PbS/PS-CdS/PMMA nanocomposites studied in Chapter 3. 55 2.3 Z-Scan setup: L1, L2 - Lenses, BS - Beam Splitter, S - Sample, PD – Photo detector. 70 2.4 Nonlinear scattering setup: L - Lens, S - Sample, PD - Photodiode. θ - 70 angle at which scattered light was collected. 2.5 Pump-probe setup: L- Lens, S - Sample, PD - Photodiode. M - mirror, 71 BS - beam splitter, SHG - second harmonic generator. 2.6 Schematic diagram of the set up for catalytic activity measurement. 3.1 Typical XRD pattern of a sample of PbS QDs indexed to JCPDS 05- 79 0592. 3.2 TEM images of PbS nanoparticles obtained at synthesis time of (a) 1, (b) 80 30 and (c) 120 minutes with averaged sizes estimated at 5.2 ±0.6, 6.1± 0.5 and 6.4± 0.6 nm respectively. 3.3 Absorption spectra of PbS nanoparticles withdrawn and isolated at 81 different reaction times from 1 min to 120 min & re-dispersed in hexane. (Also attached at the bottom is the spectrum of ethanol in hexane). 3.4 A schematic showing the steps involved for the preparation of 82 homogeneous PbS/PS nanocomposite. 3.5 SEM images of PbS/PS nanocomposite film prepared without adding 83 decanthiol compatibilizer at (a) 1600X and (b) 4300X magnifications. 73 xv 3.6 SEM images of PS/PbS nanocomposite film prepared with decanethiol at 83 (a) 65000X and (b) 8000X and (c) 850X magnifications. 3.7 Absorption (a) and PL (b) spectra of PbS/PS composite films containing 84 different sized nanoparticles. Excitation wavelength = 532 nm. 3.8 Typical (a) TEM image and (b) absorption spectrum of CdS 85 nanoparticles prepared at 120oC. 3.9 PL spectra of CdS/PMMA composite films before and after uv 86 irradiation with excitation at 375nm. (The two vertical lines are visual guides to show the blue shifting of the luminescence upon uv irradiation). 3.10 A schematic diagram showing the formation of layered PbS/PS 88 nanocomposite. The bottom layer comprises 6 nm PbS/PS, while the top layer comprising 5.2 nm PbS/PS. The centre PS was added as a barrier layer. 3.11 Cross-sectional side-view SEM images of the layered PbS/PS nanocomposite at (a) 50X and (b) 16000X magnifications. 88 3.12 Typical (a) absorption and (b) PL spectra of layered PbS/PS nanocomposite irradiated from top and bottom (refer to Fig. 3.10 for orientation) surface. Also included in (b) is the PL spectrum of a film comprising of mixed QDs of the two sizes. Excitation wavelength = 532 nm. 89 3.13 (a) Schematic diagram of layered PbS/PS-CdS/PMMA nanocomposites. 91 (b) NIR PL spectra of layered PbS/PS-CdS/PMMA nanocomposite excited at 532nm, (c-d) PL spectra of layered PbS/PS-CdS/PMMA nanocomposite excited at 375 nm before and after uv irradiation with PbS/PS of 10 μm and 150 μm thickness respectively. 3.14 PL of multilayer CdS/PS- PbS/PS nanocomposite films prepared by imprinting method with excitation at 375 nm from the (a) top surface and (b) bottom surface. 93 3.15 TEM micrographs of PbS QDs with average sizes of: (a) 4.6 ± 0.5 nm, 94 (b) 5.3 ± 0.5 nm, (c) 6.0 ± 0.6 nm, and (d) 11.0 ± 2.0 nm. Inserts give histograms of the respective size distributions. 3.16 Typical UV-VIS-NIR absorption spectra of the PbS QDs in solution. The 95 spectra are labeled with the corresponding sample sizes determined from Figure 3.15. xvi 3.17 Typical NIR luminescent spectra of the PbS QDs in (a) hexane, and (b) polystyrene composite film. Excitation wavelength = 532 nm. 96 3.18 Open-aperture Z-scans at varying laser intensities for different sized PbS QDs in (a) hexane and (b) PS film. Solid lines represent the theoretical fits. 98 3.19 Nonlinear scattering measurements of different sized QD collected at 99 10o. 3.20 Closed-aperture Z-scan curves of different sized PbS QDs in (a) hexane and (b) PS film with theoretical fits. 102 3.21 Size dependent FCA cross-section of PbS QDsin hexane and in PS film and its linear fits with 15% error bars. 104 4.1 TEM of PbS QDs and the corresponding core-shell QDs: (a-b) 5 nm PbS and 5 nm PbS/CdS, (c-d) 6nm PbS and 6 nm PbS/CdS, and (e-f) 7 nm PbS and 7 nm PbS/CdS. 111 4.2 XRD patterns of (a) 5 nm PbS and 5 nm PbS/CdS; (b) 6 nm PbS and 6 nm PbS/CdS; (c) 7 nm PbS and 7 nm PbS/CdS. Standard patterns of cubic PbS and CdS (JCPDS 5-0592 and JCPDS 800019) are marked for comparison. 113 4.3 (a) Absorption and (b) PL spectra of 5 nm PbS QDs (black line) and core-shell 5 nm PbS/CdS QDs with 1hrs (red line) or 20hrs of (blue line) of cationic exchange reaction. Excitation wavelength for PL = 532 nm. 117 4.4 (a) absorption and (b) PL spectra of (i) 6nm PbS QDs (dotted line) and 119 its core-shell PbS/CdS QDs (straight line), (ii) 7 nm PbS QDs (dotted line) and its core-shell PbS/CdS QDs (straight line). Excitation wavelength for PL = 532 nm. 4.5 HRTEM of a 6 nm PbS/CdS core-shell QDs showing crystalline PbS core with (200) planes and an unresolved shell layer. 4.6 (a) Absorption and (b) PL spectra of composite films of 5 nm PbS/CdS 121 core-shell QDs dispersed in polystyrene with 4.5 μm, 6 μm and 18.3 μm film thickness respectively. 4.7 (a) Absorption and (b) PL spectra of composite films of 6 nm PbS/CdS 122 core-shell QDs dispersed in polystyrene with 2.9 μm and 6.6 μm film thickness respectively. 120 xvii 4.8 Open aperture Z scan curves at different irradiance for different sizes of 123 PbS/CdS core-shell QDs in hexane: (a) 5 nm, (b) 6 nm, and (c) 7 nm. 4.9 Close aperture Z scan curves at different irradiance for different sizes of 124 PbS/CdS core-shell QDs in hexane: (a) 5 nm, (b) 6 nm and (c) 7 nm. 4.10 Open aperture Z scan curves at different irradiance for 5.0 nm PbS/CdS 128 QDs: Sample I (▪) and Sample II (◦)in solution at two different intensities. 4.11 (a) Open aperture and (b) close aperture Z scan curves at different 130 irradiance for 5 nm PbS/CdS QDs in PS polymer films of different thickness (i-iii): 4.5 μm, 6 μm and 18.3 μm. 4.12 (a) Open aperture and (b) close aperture Z scan curves at different 131 irradiance for 6 nm PbS/CdS QDs in PS polymer films of different thickness (i-ii): 2.9 μm and 6.6 μm. 4.13 Pump-probe signal of 4.6 nm PbS QDs excited at 780 nm with different 134 pump intensity and probed at 780 nm. 4.14 Pump-probe signal and life times of PbS QDs, pump at 780 nm probe at 135 780 nm. 4.15 Pump-probe signal of PbS/CdS core-shell QDs of different sizes: (a) 5 137 nm (b) 6 nm, excited at 780 nm with different pump intensity and probed at 780 nm. 4.16 Schematic diagram of pump generated absorption and subsequent 137 relaxation processes. After excitation by the pump, the exciton relaxes by (a) intraband relaxation, (b) excited state absorption, (c) biexciton formation, (d) trapping to surface/defect states, (e) radiative recombination and (f) Auger recombination (AR) 4.17 Relaxation dynamics of (a) 5 nm PbS; (b) 5 nm PbS/CdS and (c) 6 nm 141 PbS/CdS QDs in hexane at 600 nm probe wavelength at different pump intensity with 400 nm pump. 4.18 (a) Fast (τf) component and (b) slow (τf) component of relaxation 142 kinetics of 5 nm PbS, 5 nm PbS/CdS and 6 nm PbS/CdS QDs at 600 nm probe wavelength in hexane at different pump excitation intensity using 400 nm pump. 4.19 Transient differential transmittance spectra of (a) 5 nm PbS; (b) 5 nm 144 PbS/CdS and (c) 6 nm PbS/CdS measured at different delay times under 50 nJ/pulse excitation intensity with 400 nm pump. xviii 5.1 (a) Representative SEM image of the PbS NWs grown by CV deposition 153 at 0.1V/s for 500 scans; (b) XRD pattern indexed to JCPDS 05-0592; (ce) respectively the SAED, TEM and HRTEM images of the PbS NWs produced. 5.2 EDX spectrum revealing that the NWs are composed of Pb and S. Cu 153 peak present is due to the copper grid used for TEM analysis. 5.3 Typical CV scans obtained at different stages of the deposition of PbS 154 NWs at 0.1V/s between -0.4V and -0.95V. 5.4 Schematic showing the deposition of core-shell NWs: (a) Core metal 156 wire is deposited into AAO channel, (b) pore widening to create the annular gaps around the core wires, (c) PbS shell is then deposited onto the metal core wires. 5.5 (a) Representative SEM image of Au/PbS core-shell NWs prepared; (b) EDX collected near the top segment of the NWs; (c) XRD pattern fitted to the standard PbS (JCPDS 05-0592). 157 5.6 Representative SEM images of (a) Cu/PbS NWs prepared by CV deposition, (b) bundles of NWs detected breaking off from the base electrode. Regions marked in (a) indicate (I) the intact NWs arrays and (II) the remains of broken bundles, respectively. 158 5.7 (a) TEM image on a single wire among the broken bundles; (b-d) SAED 160 taken on areas marked A, B and C respectively in (a); (e) XRD pattern of the NWs fitting to PbS, Cu, Au and small amount of Cu2S. 5.8 SEM image of NWs prepared by immersing Cu core wires directly into a 161 solution containing 0.1M Na2EDTA and 0.01M Na2S at pH5 for a total of 90 minutes. 5.9 XRD patterns obtained when Cu core wires were immersed in pH5 162 electrolyte bath of 0.1M Na2EDTA and 0.01M Na2S: (a) without CV scanning and (b) with CV scanning. 5.10 Evolution of CV scans for Cu core wires immersing in a pH 5 solution 162 containing 0.1 M Na2EDTA and 0.01M Na2S at 100 mV/s. 5.11 (a) and (b) Representative SEM images of arrays of Cu/PbS core-shell NWs and enlarged view showing the baseball bat structure. (c) TEM and SAED images of a single broken wire. (d) XRD pattern showing clearly Cu and PbS diffraction patterns. xix 163 5.12 SEM image and EDX line scan on bundles of Cu/PbS NWs prepared by 164 constant potential deposition 5.13 SEM image of bundles of Cu/PbS core-shell NWs prepared by constant 164 potential deposition with the corresponding EDX on areas marked as A (near to the base electrode) and B. Samples are placed on a Carbon tape in this analysis. 5.14 (a) Tapping mode AFM images of PbS NWs free standing on a Au electrode, and (b) a typical I-V curve of a single PbS wire measured using the AFM tip. 165 6.1 MnO nanoparticles prepared at manganese acetate-to-OA ratio of 1:3 at 172 300oC. For comparison, simulated XRD patterns from the database are shown as vertical lines 6.2 Schematic showing the procedure in our synthesis of MnO NPs. 6.3 UV-visible absorption spectra of the reaction mixture (a) degassed at 80 174 O C for 30 minutes and subsequently quenched with hexane, and (b) that was heated to 260 OC in nitrogen environment. 6.4 Representative TEM images of faceted MnO nanoparticles produced at 320oC for 60 min using Mn(III) to OA ratio of (a) 1:1; (b) 1:3; (c) 1:5 and (d) 1:6.6. (e) TEM image showing MnO hexapods and fragments produced at 320oC for 15 min at 1: 8 ratio. 179 6.5 A plot showing the relationship between particle size and amount of OA ligand utilized. 180 6.6 Representative TEM images of MnO nanoparticles produced at Mn(III) 183 to OA ratio of (a) 1:1 ratio at 280oC for 60 min, (b) 1:3 ratio at 300oC for 60 min, and (c) 1:3 ratio at 320oC for 30 min. 6.7 HRTEM image of one faceted MnO particle prepared from Mn(III) to OA ratio of 1:3, 320OC and reaction for 30 min. 6.8 (a-b) TEM images of aliquots at Mn(III) to OA ratio 1:6.6 at 320oC 185 withdrawn at (a) 10min and (b) 20min. 6.9 Samples withdrawn at 30 min from reaction of Mn(III) to OA molar ratio 186 of 1:4 at 320OC with different tilted angle relative to x axis: (a) 0o ; (b) 10o; (c) -20 o; (d) -30 o. 173 184 xx 6.10 (a) HRTEM of faceted MnO nanoparticle prepared using acetate:OA 187 molar ratio of 1:3 at 320OC after 1hr of reaction, and (b) its expanded view 6.11 HRTEM images of (a) mulitpod MnO particle prepared using acetate to 189 OA molar ratio of 1:8 at 320OC after 60 min of reaction showing moiré interference patterns, and (b) bipod MnO particle showing the planes of cubic Mn3O4. 6.12 BET isotherm of the as prepared nanocrystalline MnO, showing the 190 adsorption ( ) and desorption ({) of N2 molecules. 6.13 TEM images of MnO nanocrystals: a) as prepared and b) after catalytic reaction. 191 6.14 XRD patterns of the MnO nanocatalyst as prepared, after heat treatment 192 and after catalytic reaction. For comparison, simulated XRD patterns from the database are shown as vertical lines. 6.15 IR spectra of MnO nanocrystals (ai) as prepared and (aii) magnified 193 region between 400 to 1000 cm-1; (bi) after pretreatment at 300oC in N2 and (bii) magnified region between 400 to 1000 cm-1. 6.16 Isothermal TGA curves of as prepared MnO heated under different 195 ambient gases, a) 200°C in air, b) 200°C in N2, c) 300°C in N2 and d) 400°C in N2 6.17 Catalytic activities of bulk MnO (), nanocrystalline MnO pre-treated at 196 200°C (z) and 300°C (S). 6.18 Schematic showing the synthesis of Mn3O4 NPs using the injection 198 method. 6.19 Typical XRD pattern of Mn3O4 particles prepared at different conditions 199 (a) direct heating to180oC with 45mins reaction, (b) injection at 180oC with 3hrs reaction, (c) injection at 140oC with 1hrs reaction. For comparison, simulated XRD patterns from the database are shown as vertical lines. 6.20 IR spectra of (a) OLA-capped Mn3O4 (a) overview and (b) magnified 201 region. 6.21 Representative TEM images of Mn3O4 nanorods produced by injecting 203 0.15g/ml of precursor at 190oC followed by growth at 180OC for (a) 10 min, (b) 60 min and (c) 3hours. xxi 6.22 Representative TEM images of Mn3O4 nanorods produced by (a) 204 injection of 0.05g/ml at 190oC and growth at 180OC for for 30mins ; (bc) injection at 160oC and growth at 150OC of (b) 0.15g/ml for 5mins, (c) 0.05g/ml for 3hrs 6.23 Representative TEM images of (a) Mn3O4 nanorods produced by direct 205 heating to 180oC for 45 min; (b) Mn3O4 particles produced by sequential heating procedure. 6.24 HRTEM image of a typical Mn3O4 nanorod prepared at 180OC after 206 45mins of reaction 6.25 TEM images of Mn3O4 a) as prepared and b) after catalytic reaction 209 6.26 BET plot of the as-prepared nanocrystalline Mn3O4, showing the 209 adsorption ( ) and desorption ({) of N2 molecules. 6.27 XRD patterns of the Mn3O4 nanocatalyst as-prepared, after heat treatment 210 and after catalytic reaction. For comparison, simulated patterns from the database are shown as vertical lines. 6.28 Isothermal TGA curves of as prepared Mn3O4 at a) 250°C and b) 300°C 211 in N2 6.29 Catalytic activities of bulk Mn3O4 (), nanocrystalline Mn3O4 pre- 212 treated at 300°C (S). 6.30 XRD patterns of the α-MnO2 nanocatalysts as-prepared, after heat 216 treatment and after catalytic reaction. Simulated patterns from the database are shown as vertical lines. 6.31 XRD pattern of the ρ-MnO2 nanocatalyst as-prepared, after storing for prolonged period and after catalytic reaction. For comparison, simulated pattern from the database are shown as vertical lines. 217 6.32 Diagram showing the (a) pyrolusite r (a), and ramsdellite R (b), (c) structures, (d) (e): schematic [100] (d) and [001] (e) views of a R–r intergrowth. Rn or rn are a succession of n consecutive slabs R or r, respectively (Diagram taken from Hill et al61). 218 6.33 TEM images of α-MnO2 a) as prepared and b) after reaction. 220 6.34 TEM images of γ-MnO2 a) as prepared, b) after storing for prolonged period and c) after catalytic reaction. 221 xxii 6.35 BET plot of as prepared nanocrystalline (a) α-MnO2 and (b) γ-MnO2, 222 showing the adsorption ( ) and desorption ({) of N2 molecules 6.36 Isothermal TGA curves of as prepared a) γ-MnO2 and b) α-MnO2 at 222 200°C in N2 6.37 Catalytic activities of bulk β-MnO2 ({), nanocrystalline α-MnO2 run 1 223 ( ) and run 2 () and nanocrystalline ρ-MnO2 run 1 (V) and run 2 (T). 6.38 XRD patterns of the α-Mn2O3 nanocatalyst as prepared, after heat 227 treatment and after catalytic reaction. For comparison, simulated patterns from the database are shown as vertical lines. 6.39 TEM images of a) manganese oxalate b) as prepared Mn2O3 and c) 229 Mn2O3 after reaction and d) SEM image of as prepared α-Mn2O3 6.40 (a) BET plot of nanocrystalline α-Mn2O3 after pretreatment, showing the 230 adsorption ( ) and desorption ({) of N2 molecules. (b). TGA profile of α-Mn2O3 at 50°C in N2 6.41 Catalytic activities of bulk Mn2O3 (), nanocrystalline α-Mn2O3 run 1 231 (z) and run 2 (S). 6.42 Reaction rates of nanocrystalline MnO pre-treated at 300°C (), α- 234 Mn2O3 (T) and Mn3O4 pre-treated at 300°C (z), α-MnO2 ( ) and ρMnO2 (V). 6.43 Reaction rates of bulk MnO (), Mn2O3 (z), MnO2 (S) and Mn3O4 234 (T). xxiii List of Symbols Band-gap energy (Eg) Carrier density (N) Circular frequency of the optical field (ω) Closed aperture transmittance (TCA) Complex susceptibility tensor of order i χ (i ) Effective sample length (Leff) Electronic Kerr coefficient (γ in SI units) Fluence (F) Free carrier absorption cross-section (σc) Free-carrier refraction cross-section (σr) Imaginary part (Im) Linear absorption coefficient (α0) Linear index of refraction (n0) Nonlinear refractive index or Electronic or optical Kerr effect (n2 in esu units) Nonlinear scattering coefficient (∝s) One-photon saturation intensity (Is) Permittivity of free space (ε0) Phase of the laser beam (φ) Phase change (Δφ0) Rayleigh range (z0) Real part (Re) Refractive index change (Δn) xxiv Sample position (z) Sample length (L) Speed of light in vacuum (c) Two photon absorption coefficient (β) Wavelength of the laser beam (λ) xxv List of Abbreviations AAO : Anodized alumina oxide AFM : Atomic force microscopy AR :Auger recombination BET : Brunauer, Emmett and Teller CS : Core/shell CTAB : Cetyl trimethyl ammonium bromide CV : Cyclic voltammetry DMF : Dimethylformamide EA : Elemental Analysis ESA : Excited state absorption EDX : Energy dispersive X-ray FCA : Free carrier absorption FCR : Free-carrier refraction FET : Field-effect transistor FTIR : Fourier Transform Infrared FWHM : Full width at half maximum () HRTEM: High Resolution TEM ICP : Inductively Coupled Plasma MEG : Multiple exciton generation MPA : Multi-photon absorption NCs : Nanocrysta NIR : Near infrared NLA : Nonlinear absorption NLO : Nonlinear optics NLR : Nonlinear Refraction NLS : Nonlinear scattering NRs : Nanorods NWs : Nanowires OA : Oleic acid OPV : Oligo(phenylenevinylenes xxvi PANI : Polyaniline PC : Polycarbonate PL : Photoluminescence (PL) PMMA : Poly(methyl methacrylate) PPY : Polypyrrole PS : Polystyrene QD : Quantum dot QY : Quantum yield RSA : Reverse saturable absorption SEM : Scanning Electron Microscopy SA : Saturable absorption (SA) SAED : Selected area electron diffraction SAM : Self-assembly monolayer TA : Transient absorption 2PA : Two photon absorption TEM : Transmission Electron Microscopy TGA : Thermal Gravimetric Analysis TOP : Trioctylphosphine TOPO : Trioctylphosphine oxide UV-VIS : Ultraviolet-visible VOCs : Volatile organic compounds XRD : X-ray Diffraction xxvii Chapter 1 Chapter 1 Introduction Nanocrystals (NCs) display many properties that are both quantitatively and qualitatively different from their respective bulk materials and also the discrete atomic or molecular species from which they are derived. The unique properties such as their electronic, optical, catalytic, and mechanical properties arise from their large surface area to volume ratio, homogeneity (size distribution), highly reactive surfaces and the quantum confinement effect.1-5 In this introduction, we will focus our discussion only on the size-, shape-, and composition-controlled preparation of different NCs and their nonlinear optical (NLO) and catalytic properties which are relevant to the work reported in this thesis. 1.1 Size-controlled preparation of nanomaterials In order to better correlate the properties of NCs to their sizes, the synthesis of NCs with narrow size distribution is an important requirement. Classical studies on crystal growth by La Mer & Dinegar and the separation of nucleation and growth processes have provided a qualitative theoretical framework in the pursuit of monodispersed NCs.6,7 Nucleation occurs when solution becomes supersaturated. Supersaturation can be produced by various methods. Firstly, it can be achieved by directly dissolving the solute at higher temperature followed by decreasing its solubility (increasing its chemical potential) either by lowering the temperature or by adding a miscible non-solvent. 6 Secondly, supersaturation 1 Chapter 1 can be obtained by increasing the reactant monomer concentration product above its thermodynamic product. An example is shown for the reaction in Equation 1.1 and Figure 1.1 .8,9 n(Cd − TOPO ) + n( Se − TOP ) ← ⎯→(CdSe) n TOPOm + (n − m)TOPO + nTOP − −eq(1.1) where TOPO is trioctylphosphine oxide and TOP is trioctylphosphine. Figure 1.1 Sketch of the solubility product [Cd][Se] as a function of temperature. Solid line: thermodynamic curve for the equilibrium between the monomers CdTOPO and Se-TOP and a macroscopic CdSe crystal. Dashed line: the solubility product for the equilibrium between the monomers and the critical nuclei (CdSe)c indicative of supersaturation. The points indicate: nucleation (1), cooling (1–2), and growth of the nuclei at two different temperatures (3 and 3’).9 The situation is most popularly achieved by the rapid injection of the precursors to the reaction mixture at high temperature (hot injection method).10 At such high temperature, the monomers that are formed readily exceed the nucleation concentration threshold C*min as shown in Figure 1.2 and a subsequent short nucleation burst partially relieves the supersaturation. When the concentration of 2 Chapter 1 the monomers decrease below the nucleation threshold, no further nucleation occurs. To achieve a monodispersed size, the nucleation rate must be fast enough such that the concentration drops below C*min rapidly. If the concentration remains between C*max and C*min, growth eventually occurs and will lead to polydisperse sizes. Figure 1.2 The LaMer model for monodispersed particle formation (Cs:solubility; C*min: minimum concentration for nucleation; C*max: maximum concentration for nucleation; I: prenucleation period; II: nucleation period; III: growth period)6,11 The growth stage of NCs has been widely studied by various researchers.7,10,12 Howard Reiss first predicted theoretically that diffusion-limited growth can lead to narrowing of size distributions with time by considering the diffusion area versus size.7 Sugimoto further incorporated the concept of Gibbs-Thomson effect 3 Chapter 1 which becomes significant for small crystallites.11 Finally, using these two effects the model of ‘size-distribution focusing’ was fully developed and verified experimentally by Peng et al.10 Using this model, at any given monomer concentration, there exists a critical size where the dissolution and growth are essentially zero. NCs smaller than the critical size have negative growth rates (dissolve), while larger ones grow at rates dependent strongly on size. At a high monomer concentration, the critical size is small so that all the particles grow by molecular addition. In this situation, smaller particles grow faster than the larger ones because the free energy driving force is larger for smaller particles, thus causing the focusing of the size distribution with time as demonstrated theoretically by Reiss.7 One can thus obtain a monodispersed NCs by quickly quenching the reaction during the focusing stage of the reaction. When the reactant concentrations are depleted and drop below a critical threshold due to particle growth, Ostwald ripening or defocusing will occur. In this situation, the critical size becomes larger than the average size present. Thus the high surface energy of the small NCs promotes their dissolution. The solute redeposited on the larger NCs particles and the size distribution broadens, or defocuses.8,10,13 It is possible to improve the distribution at this stage by the introduction of additional monomers at the growth temperature, thus helping to reduce the critical size and allowing refocusing. 4 Chapter 1 1.2 Shape-controlled synthesis of nanomaterials 1.2.1 Chemical synthesis of nanoparticles in solution Chemical syntheses in solution have become a popular method for the growth of nanoparticles of different shapes and aspect ratios because of substantial understanding of their key determining parameters.8,13-16 The final shape of the nanoparticles is affected by several key factors occurring during the nucleation and growth stages, which include: (1) crystalline phase of initial seeds, (2) intrinsic surface energy of different crystallographic facets and the effect of selective adhesion of capping molecules, (3) monomer concentration effect on the thermodynamic versus kinetic growth regime and (4) oriented attachment of NCs with intrinsic dipole. Depending on whether the nucleating seeds have an isotropic or anisotropic unit cell, different NCs shapes can be formed during the growth phase. For example, hexagonal seeds produced from the precursor of Co(acac)3 by abrupt heating to 200oC lead to the formation of rod and pyramid NCs. However, cubic seeds produced from prolonged heating of an oleylamine-substituted cobalt complex at 135 °C form preferentially cubic shaped NCs.17 On the other hand, MnS forms at high temperature nuclei of rock-salt structure which eventually grow isotropically into cubes. At much lower temperature, MnS nucleates in the hexagonal wurtzite structure which has a crystallographic preferred c-axis thus allowing anisotropic growth along this axis to form long nanowires.18 5 Chapter 1 Kinetic shape control has been accomplished in many semiconductors with hexagonal wurtzite or other anisotropic seeds due to the presence of crystallographic facets with different surface energy.19-21 As the growth rate of a crystal surface depends exponentially on the surface energy, the high energy surface will grow much faster than the low energy surface in the environment of a kinetically controlled high growth rate regime as seen in Figure 1.3(a). The surface energy of a particular surface can be modulated by the selective adhesion of an organic molecule which stabilizes and slows down its growth relative to the other surfaces, thus allowing the growth of rod or disk shaped NCs as seen in Figure 1.3(b). One classic example is the formation of hcp Co nanodisks by rapid decomposition of cobalt carbonyl in the presence of linear amines and TOPO. The alkylamines preferentially adsorb and thus inhibit the growth of the unique (001) face of Co relative to the perpendicular direction thus leading to faster growth along the and directions and the formation of disks. Similar selective adhesion effects on the growth of other anisotropic NCs have been seen in NCs such as PbS multipods.22 Other more complex shapes as seen in Figure 1.3(c) such as arrow-shaped NCs of CdSe and zigzag-shaped crystals of TiO2 can be formed by a sequential elimination of a high-energy facet.13,23,24 6 Chapter 1 Figure 1.3(a) Kinetic shape control at high growth rate. The high-energy surface grows more quickly than low energy surface in a kinetic regime. (b) Kinetic shape control through selective adhesion. (c) Complex shaped NCs of CdSe and TiO2 can be formed by sequential elimination of a high-energy facet. (d) High monomer concentration coupled with presence of two different crystal structure allowed the growth of branched nanostructures like tetrapods.13 7 Chapter 1 Monomer concentration has a direct effect in determining if the nucleating seeds will be in the thermodynamic or kinetic growth regime.14 A classic mechanistic studies of the shape control and evolution of the magic cluster CdSe seeds was provided by Peng et al.14 At low monomer concentration or long reaction time, it becomes more favourable for isotropic thermodynamic growth from the nucleating seeds and these are characterized by the formation of spherical or faceted particles. Higher monomer concentrations lead to kinetically favorable anisotropic growth with different kind of shapes depending on the concentration. When a median monomer concentration is used, rod shaped NCs can transform to the rice shaped NCs due to isotropic growth in a three dimensional growth stage. At higher monomer concentrations, the seeds are in the 1D growth mode and this favors rod shaped NCs. At extremely high monomer concentration, there is sufficient amount of monomers for each seed to fully grow into arms on the four (111) facets of the zinc blende structure of the tetrahedral seeds and yields tetrapod-shaped nanoparticles as seen in Figure 1.3(d). The shapes of the NCs obtained are only metastable in the presence of ligands and other monomers. If the reaction time is prolonged, the rods can eventually grow to become spherical though a 1D-to-2D intraparticle ripening growth stage which occurs when the monomer concentration drops below the concentration needed for one dimensional growth. Another approach to create anisotropic NCs is by the process of ‘oriented attachment’. 25 This process was first discovered by Penn and Banfield26,27 who 8 Chapter 1 showed that long chains of highly ordered titania can emerge from a solution of primary titania nanoparticles and faceted NCs coalesce in such a way as to eliminate two high-energy facets. Typically, the resulting one dimensional product has the same diameter but with its length several times that of the original particles. This process has been generalized for several other systems, such as ZnO28, CdSe29, CdTe30, PbSe31, ZnS32 and ZnTe33. In the formation of CdSe quantum wires using cadmium acetate and selenourea in amines, Peng’s group was able to observe evidence for the formation of these quantum wires through oriented attachment. They detected reversible association of magic-sized clusters in pre-wire aggregates and coexistence of loosely associated nanoclusters and chemically fused fragments of nanowires. In the report by Murray’s group, the importance of dipole moment in the oriented attachment of PbSe nanodots was elucidated. It is possible that such dipole moments are one of the important generalized factors for the successful oriented attachment in the various other systems. 1.2.2 Template synthesis of nanowires Both solution phase and template-assisted synthesis have been widely used to prepare nanowires of different composition.33-37 However, for the preparation of oriented nanowires arrays, template-assisted synthesis is a more convenient approach compared to solution phase synthesis. Two of the most commonly used templates are anodized aluminum oxide (AAO) and polycarbonate (PC). In this brief introduction, we will focus mainly on materials deposited into AAO. As the 9 Chapter 1 nanopores of the AAO are uniform in diameter and hexagonally packed, deposition into the nanopores results in highly ordered and vertically aligned nanowire arrays such as those shown in Fig 1.4. Figure 1.4 Fabrication process of thin film of p-CdTe on highly ordered, singlecrystalline nanopillars of n-CdS.38 Beside the synthesis of single species nanowires using an AAO as a template, hybrid nanowires of different functionalities have been actively pursued recently. Whitesides’s group accomplished the synthesis of core-shell polyaniline-gold (PANI-Au) and segmented (Au-PANI and Ni-Au-PANI) structures electrochemically within AAO membranes.39 They were able to deposit gold as a layer of shell around a preformed PANI nanotube by exploiting the pH dependence of the chemical properties of PANI, and the solubility of the AAO template in basic solution. The control over the length of the gold shells in the core-shell structures was accomplished by adjusting the time and rate of 10 Chapter 1 electrodeposition and the pH of the solution from which the materials were grown. Pure aligned gold nanotubes can be formed using oxygen plasma to remove the PANI from the PANI-Au structures. Segmented Au-PANI and Ni-Au-PANI were accomplished through the self-assembly monolayer (SAM) of thioaniline which helped in the nucleation of PANI on top of metal nanorods and acted as an adhesion layer between the metal and PANI component. This has added stability to the composite during the electrodeposition process when delamination of the film can occur due to shape changes of the electroactive polymers induced by variation in applied potential. A similar method was used by Hendren et al. to form gold nanotubes but using polypyrrole (PPY) as the core on glass substrates by electrodeposition into thin film porous alumina templates.40 The AAO pores were chemically etched after PPY deposition using a sodium hydroxide solution to form spacing between the PPY and the AAO walls that allowed the gold to be deposited as a shell on the PPY core. Plasma etching then removed the PPY core forming a gold nanotube. Capped nanotubes were formed by allowing the gold to overgrow the polypyrrole wires. Extinction spectra for uncapped structures showed strong extinction peaks related to plasmonic resonances and were found to be sensitive to the polarization state and angle of incidence. The p-polarized spectra revealed two peaks which have some similarities to those observed previously in gold nanowires and were attributed to resonance across the nanowires respectively. The capped nanotubes were found to posses an additional extinction peak near 550 nm. Theoretical 11 Chapter 1 calculations using finite element modeling show a high degree of field localization and screening depending on wavelength. Using both model and experimental evidence, they were able to attribute the new peaks of the capped nanotubes to gold nanoparticles of similar size sitting on top of the nanotube. Guo et al have fabricated interesting p-n junction nanowires recently. A sequential electrodeposition template method has been utilized to synthesize CdSpolypyrrole nanowires. The prepared nanowires were found to possess a strong photodependent rectifying effect whereby a maximum rectification ratio of 13 was found at high intensity.41 Homogeneous hybrid organic/inorganic nanorods composed of OPV3 and CdS have been fabricated by a facile template method with the assistance of vacuum.42 Basically, the porous AAO membrane was immersed into the transparent CdSOPV3 mixture in a flask under reduced pressure for about 5-8 min. The CdSOPV3 nanorods were finally embedded into the channels of the AAO membrane after repeating the process three times. The inorganic semiconductor CdS and organic semiconductor OPV3 was found to be homogeneously distributed into the hybrid nanorods. The emission of OPV3 was found to be decreased greatly in the hybrid nanorods. The quenching was believed to arise from efficient charge transfer between the two semiconductors due to the large number of interfaces in the homogenous nanorods. 12 Chapter 1 Gold nanotubes with smooth wall and nanoporous walls were synthesized through a method similar to Whiteside’s method by first growing a PANI and then depositing gold or gold/silver followed by dissolution of PANI and silver using nitric acid and AAO with sodium hydroxide.43 Ultrathin Pt layer was coated on the Au surface through copper under-potential-deposition, followed by the spontaneous replacement with the nobler Pt metal ions. The as generated Pt coated-gold nanotubes have a large surface area and were tested for their electrocatalytic activities towards methanol oxidations referenced against commercially available carbon-supported Pt nanoparticles. The gold nanotubes with nanoporous walls were found to have better electrocatalytic activities towards methanol oxidations due to its higher CO-poisoning tolerance and shorter effective length for electronic transportation. An innovative method to transform single component metal nanowires to multisegment nanowires with multimetal components was developed by the group of Schaak et al.44 They immersed a porous membrane filled with metal nanowires into a tetraethylene glycol metal salt solution which reduces on heating to form metal atoms. Moreover, they also demonstrated that the method could be extended to the formation of metal phosphides segments on the metal nanowires by reacting with trioctylphosphine at high temperature. Their new strategy thus allows the transformation of other alloyed and core-shell NCs from metal NCs, and also a specific section along the nanowires. 13 Chapter 1 1.3 Core-shell quantum dots Three kinds of core-shell (CS) quantum dots (QDs), namely Type-I, Reverse Type-I, and Type-II (Figure 1.5), are differentiated based on the bandgap size and relative position of the valence band and conduction band energy levels of the core and shell components. Figure 1.5 Schematic diagram of the different type of core-shell QDs. The upper and lower edges of the rectangles correspond to the positions of the conductionand valence-band edge of the core (center) and shell materials, respectively.45 Type I CS QDs are formed when a semiconductor with a larger bandgap is grown on the surface of a smaller bandgap core as illustrated in Figure 1.5. The shell in the Type I CS QDs provides surface passivation that can lead to an increase in the fluorescence quantum yield (QY) and stability against photo bleaching. Depending on the actual band offset between the core and shell, the properties of type I CS QDs can be significantly different.45 For example, if the band offset are large for both the conduction and valence band, both the electrons and holes are mostly confined in the core. CdSe/ZnS is the classic example of such a case. In this system, shell growth results in only a small red shift (5–10 nm) of the 14 Chapter 1 excitonic peak in the UV/Vis absorption spectrum and the photoluminescence (PL) wavelength. If the band offset is large for the valence band (holes) and very small for the conduction band (electron), the hole is largely confined in the core while the electron is highly delocalized over the entire CS nanoparticles. Such a band alignment is seen in common heterostructure system of CdSe/CdS. In this system, the red shifts of the UV and PL increase as the thickness of the shell increase and can reach as large as 50 nm.46,47 In the third kind of Type I system typified by CdSe/ZnSe, a large conduction-band offset results in efficient confinement of the electrons in the NC core while the valence band offset is relatively smaller. Lead based NCs have small bandgap and high QY up to 90% and thus are suitable as infrared emiiters. However, their stability towards oxidation is extremely poor and CS of lead based NCs have been synthesized to provide better protection against photooxidation. Air-Stable PbSe/PbS and PbSe/PbSexS1-x CS NCs with QY reaching 45% and 55% have been synthesized by Lifshitz et al. using a sequential injection of the precursors and a single injection procedure.48,49 The successive ion layer adsorption and reaction (SILAR) method has been utilized to synthesize PbSe/PbS with enhanced chemical robustness whereby the PL QY was less ligand dependent and much higher than that of PbSe.50 Core-shell QDs based on PbS core are less common probably due to the high reactivity of the PbS towards aggregations at the high temperatures needed for 15 Chapter 1 shell growth using traditional methods in organic solvent. PbS/CdS was first synthesized using an aqueous method that exhibits both band edge emission and above band edge broad visible white emission.51 More recently, PbSe/CdSe and PbS/CdS were synthesized by a cation exchange method resulting in an improved stability of its optical properties.52 Due to the toxicity issues with cadmium and lead based NCs, synthesis of CS systems using other less toxic elements such as indium has gain rapid interest lately. InP/ZnS CS NCs of high quality have been synthesized recently by various groups giving QY between 40 and 70%. The interest was sparked by the success of Peng’s group in obtaining high quality InP/ZnS CS NCs with QY of 40% through the use of indium carboxylate activated by fatty amine and subsequent growth of ZnS shell in a one pot thermal cycling method using zinc stearate and sulfur.53 Nann et al and Ress et al showed that higher QY of 60% and 70% can be obtained by improving the initial InP QY and using a heating up method respectively.54-56 Type-II CS NCs are characterized by having either the valence-band edge or the conduction band edge of the shell within the bandgap of the core so that staggered energy levels between the core and shell are formed. Photoexcitation of such CS QDs leads to a spatial separation of the hole and the electron in different regions of the CS structure. Bawendi et al. sparked interest in the research field on Type II CS such as CdTe/CdSe and CdSe/ZnTe when he demonstrated the possibility to 16 Chapter 1 reach emission wavelengths of 700-1000 nm which are suitable for biological applications.57 However, the QY of as synthesized CdTe/CdSe is very low due to the spatial separation of carriers and the long radiative lifetimes. An improvement of QY can be achieved by growing a thin shell or another layer such as ZnTe (which have a higher conduction band than CdSe) forming a core-shell-shell system to remove electron traps by containing the electron within the dot. A higher emission QY CdTe/CdSe can be achieved by first synthesizing a better quality core using CdO and TOPTe followed by the sequential addition of shell precursor without precipitation of the core.58 An optimum QY of 38% can be achieved by growing only a CdSe shell thickness of 0.4-0.5 nm. Fluorescence dynamics of Type II CdSe/ZnTe QDs synthesized using CdO showed that the rate of hole transfer decreases with increasing the size of the cores and is independent of ZnTe shell thickness.59 Other examples of type II NCs include ZnTe/CdTe, ZnTe/CdS, CdSe/CdTe/ZnTe and CdS/ZnSe and ZnTe/CdSe.60-62 For the last NCs of ZnTe/CdSe, an interesting anistropic growth of the shell can be seen. By using oleylamine as the selective adhesion agent and alternating injections of precursor at 215oC, it is possible to first grow pyramidal shaped CS NCs, which upon further injections can form tetrapod shaped heterostructures.63 Reverse Type-I systems occurred when a wide-gap semiconductor core is overcoated with a shell of a narrower gap material. Depending on the thickness of the shell, the holes and electrons are partially or completely confined in the shell. ZnSe/CdSe core-shell structures were both experimentally and theoretically 17 Chapter 1 examined by Klimov et al. and was found to exhibit either Type-I or Type-II behaviors, depending on the core radius and the shell thickness.64 Time and spectrally resolved PL measurement on a series of NCs with a fixed core radius and increasing shell thickness provided direct information on the e-h overlap integral and thus showed a continuous transition from Type-I (both electron and hole wave functions are distributed over the entire NC) to Type-II (electron and hole are spatially separated between the shell and the core) and back to Type-I (both electron and hole primarily reside in the shell) localization regimes. 1.4 Nonlinear optical limiting properties For a particular material, optical limiting effect typically results from an accumulation of several processes. The most common nonlinear mechanisms for optical limiting in inorganic nanomaterials are: nonlinear scattering (NLS), freecarrier absorption (FCA), multiphoton absorption (MPA), and nonlinear refraction (NLR). It was reported that several aspects of nanomaterials such as their absorption band, particle size, geometry, aggregation state and host matrix properties have a strong influence on its optical limiting performance.65 Moreover, the wavelength, pulse duration, and repetition rate of the excitation source will also have great influence on the nonlinear response of a particular system. 18 Chapter 1 1.4.1 Nonlinear scattering (NLS) When light interacts with an object with variant refractive index such as particles, bubbles or interfaces between groups of nonexcited and excited molecules, elastic scattering of light can occur.66 The elastic light scattering has been fully described by two theories depending on the sizes of scattering centers. For particles much smaller than the wavelength of light or where the particle is nonabsorbing, the scattering is best described by Rayleigh scattering. When the size of the particles becomes comparable to the wavelength of light or larger, Mie scattering theory has to be used. The light scattering mechanism has serious implications for optical limiting as the scattering can become highly directional or fairly uniform in accordance with the size of the scattering centers. As illustrated in Figure.1.6, Rayleigh scattering by smaller particles tends to be symmetric with respect to forward and backscattering; while the scattered light becomes more forward scattered for larger particles by Mie scattering. Optical limiters based on light scattering normally are more common in liquid solvent as the process can often be reversible or the scattering centers can be easily replaced by diffusion when optical damage occurs. This is less possible in solids as the scattering centers often suffer irreversible decomposition processes and cannot be replaced readily 19 Chapter 1 Fig 1.6 Scattering intensity versus angle from a Rayleigh scatterer. (b) Scattering intensity versus angle from a Mie type scatterer with a size in the order of the wavelength of the incident light. (a) (b) Figure 1.6 (a) Scattering intensity versus angle from a Rayleigh scatterer; (b) scattering intensity versus angle from a Mie type scatterer. For nanomaterials in solutions, optical limiting due to NLS has been found to play an important role in conjunction with the other nonlinear processes. Three main mechanisms have been proposed in the literature to explain the optical limiting due to NLS for a variety of nanomaterials such as carbon black, metal nanoparticles and silica spheres.65 One possible mechanism was proposed by Joudrier et al. in their investigation of a colloidal suspension of silica particles. They proposed a photoinduced refractive index mismatch between the two components of the suspensions at high fluence.67,68 The large refractive index mismatch was found to originate from the interface between the two constituents and was dependent on the polarity of the surrounding medium. A larger optical limiting was observed with a higher solvent polarity, and this was suggested to be related to the nature and strength of the bonds between the silica particles and the surrounding molecules. 20 Chapter 1 The second mechanism proposed is the formation of solvent bubbles. Belousova et al suggested that NLO limiting in an aqueous suspension of carbon particles at high energy densities was due to photoinduced scattering.69,70 The carbon particles absorb the intense laser radiation, raising its temperature and lead to heating up and eventually vaporizing the surrounding solvent to form bubbles of vapour shell around the particle. The intense light was scattered efficiently around the expanding bubble of vapour shell due to the resulting large refractive index discontinuity on the vapour–solvent interface.66 At moderate fluence, bubbles are more likely to form in the presence of nanosecond lasers compared to picosecond or femntosecond lasers because the nucleation time of the bubble is reported to be typically in the order of nanoseconds or longer.66 The third possible mechanism is sublimation or evaporation of the material. Mansour et al. studied the NLS by carbon black suspensions and found that the particles are rapidly heated by strong linear absorption, vaporized and ionized, leading to the formation of rapidly expanding microplasmas that strongly scatter the laser.71 Metal nanoparticles can also act as good optical limiters by forming scattering centers through the second or third mechanisms.72 It was found that the wavelength dependence for the optical limiting process can be correlated to the linear absorption of the gold nanoparticles such that the threshold fluence for nonlinearlity is smaller below 530 nm and increase towards the red. The two types of scattering centers have different timescale for the limiting process. At high fluence using picosecond pulses, the large energy absorbed by metal clusters 21 Chapter 1 causes a fast expansion and vaporization of the metal particle itself in the subnanosecond range. The expanded microplasms thus act as the third type of scattering center. At lower fluence using nanosecond excitations for the smaller size clusters, the formation of bubbles by energy transfer from the cluster to the solvent takes place within tens of nanoseconds. Several general factors on the scattering induced optical limiting have been identified such as the size, shape and extent of aggregation of the nanoparticles;65,73,74 boiling point, surface tension and viscosity of the solvent which affects the depletion of the nanomaterials within the focal volume and lastly the repetition rate of the laser.75 1.4.2 Free carrier absorption (FCA) In semiconductors, single photon absorption or two-photon absorption (2PA) can generate carriers (electrons or holes) in the conduction or valence band respectively. At sufficiently high laser intensities, these carriers can further absorb additional photons and be promoted with the assistance of phonons to states higher (lower) in the conduction (valence) band. This phonon-assisted absorption of the free carriers is thus named free-carrier absorption (FCA).66 An equation incorporating FCA into the nonlinear Z scan measurement can relate the fluence F to vary with the sample position (z) as: ασ dF = −α 0 F − 0 c F 2 - - - eq(1.2) . dz 2h ω where α0 is the linear absorption coefficient, σc is the FCA cross-section, and ħω is incident photon energy used to produce an electron-hole pair. 22 Chapter 1 Equation 1.2 is exactly analogous to the equation describing 2PA loss with the fluence replacing the irradiance and α 0σ c replacing 2PA coefficient (β). Thus 2hω microscopically, FCA can be considered as the limit of 2PA with a resonant intermediate state.76 The FCA-induced optical limiting response is thus dependent on the incidence fluence (energy per unit area ) instead of intensity dependent which results in a device response that is independent of the incident pulse duration.66 However, the independence of the incident pulse duration only occurs if the pulse duration is shorter than the diffusion and recombination processes of the free carriers. As the nonradiative recombination of multi-excitons generated at sufficiently high laser fluence in CdSe can be as short as 1 ps, short pulse duration lasers such as femtosecond lasers are necessary for the device to be pulse width independent.77 FCA occurs in both solid state films and suspensions, covering broad temporal and wavelength ranges. Like many other nonlinear optical limiters, FCA can often coexist with NLS and TPA since the generation of free carriers can also arise from a 2PA process 65 1.4.3 Multi-photon absorption (MPA) A multi-photon process is described as an instantaneous nonlinearity which occurs through the simultaneous absorption of two or more photons via virtual (or real) states in a medium.65,66 The losses from 2PA occur in solids when the photon energy, hϖ , is larger than one-half the band-gap energy, Eg. The equation 23 Chapter 1 describing 2PA of a beam of irradiance I as a function of depth z in a material can be described by a propagation equation in ‘Beer–Lambert’ format: dI = −α 0 I − βI 2 ......eq(1.3) dz where α 0 , in units of m−1, is the linear absorption coefficient and β, in units of mW−1, is the 2PA coefficient. β can be related to imaginary part of the third order complex susceptibility tensor χ ( 3) by the equation: β= 3ω 2ε 0 c n0 2 2 Im[χ ( 3) ] − − − eq(1.4) where ω is the circular frequency of the optical field, ε0 is the permittivity of free space, n0 is the linear index of refraction, and c is the speed of light in vacuum. It is reported that the linear absorption α 0 for ħω < Eg can come from defects, impurities or band tailing and can often be ignored in good quality materials.76 Thus when α 0 = 0, we can obtain the solution for the transmission intensity: I ( L) = I0 ........eq(1.5) 1 + I 0 βL where L is the sample length. Thus, the transmission intensity decreases as the incident intensity increases, resulting in an optical limiting phenomenon. This optical limiting is strongly dependent on the 2PA coefficient, the incident intensity and the propagation length L. As intensity is essentially the energy density divided by the pulse duration, shorter incident pulses such as femtosecond lasers may be used to obtain large optical limiting with 2PA. It is found that several factors such as the concentration and size of the nanoparticles will 24 Chapter 1 significantly affect the optical limiting performances of MPA through their 2PA coefficients or the limiting threshold.78,79 For example, the two-photon absorption coefficient β was found to increase linearly with concentration in the case of CdS in DMF probed using a 532 nm nanosecond laser.78 Similar to FCA, MPA can induce optical limiting in materials that are in both solid state and solution. MPA induced optical limiting effects have been observed in many metal and semiconductor nanomaterials such as CdS, CdSe, CdTe, ZnS and ZnSe.78,80-83 MPA often exists with other nonlinear mechanisms such as NLS and nonlinear refraction (NLR) and can help to improve its overall optical limiting performance. 1.4.4 Nonlinear Refraction (NLR) NLR always accompanies NLA and thus the corresponding refractive index can be changed by the variety of mechanisms discussed above for absorption. The types of nonlinear refractive mechanisms include optical Kerr effect, excited state or free-carrier generation, reorientation, electrostrictive, and thermal nonlinear refraction as well as cascaded second-order nonlinearities. The bound-electronic NLR (optical Kerr effect) is defined as n2 (esu units) by restricting n2 to only the ultrafast electronic response. 76 While free carrier refraction (FCR) has its analogue in atomic and molecular systems, the NLR comes from the redistribution of population among levels. For example, in solids this redistribution generates free carriers which block further transitions (Drude band blocking) and the refractive index is changed.84 25 Chapter 1 We first devote our discussion to the optical Kerr effect. We can describe the 2PA coefficient β in terms of Im χ ( 3) in Equation 1.4 while the corresponding real part of χ ( 3) (Re χ ( 3) ) can be related to the ultrafast NLR as follows: Re χ (3) = 2n0 ε 0 cγ .....eq(1.6) 85 2 where γ is the electronic Kerr coefficient in SI units. The same is true for cascaded processes of NLR associated with carrier generation by either linear or 2PA. The induced phase distortion imposed on a laser beam by NLR is related to the refractive index change, Δn, by: dφ Δn2π ....eq(7) = λ dz where φ is the phase and λ the wavelength of the laser beam. NLR in the sample manifests itself as beam broadening or narrowing in the far field, thus changing the fraction of light passing through the aperture as the sample position is changed. Therefore the aperture transmittance is a function of the sample position z. In a closed aperture Z-scan technique, as we move the sample from far before the focal plane to near to the focal plane, it results in a change from no self lensing to lensing in the sample. If the sample has a negative nonlinearity, the lensing results in collimation of the laser beam with corresponding increase in the aperture transmittance. As we move the sample beyond the focal plane, the negative lensing results in beam divergence and a 26 Chapter 1 corresponding decrease in the aperture transmittance.66 Thus this increase in transmittance followed by a decrease in transmittance (peak-valley) denotes a negative NLR (self defocusing), whereas if we observe a valley-peak it suggests a positive nonlinearity (self focusing) in the sample as illustrated in Figure 1.7.84 Figure 1.7 Schematic representation of z-scan results for a negative refractive nonlinearity (dashed curve) and a positive refractive nonlinearity (dotted curve). Both curves have been corrected for absorption. The solid curve shows the result of removing the aperture from the measurement apparatus and collecting all the transmitted light, thus isolating the nonlinear absorption.66 Incorporating FCR to the nonlinear refraction, we can write Δn = n2 I + σ r N .............................eq(1.8) dN αI = .....................................eq(1.9) dt hω ασ r Δn(t ) = n2 I (t ) + hω t ∫ I (t ' )dt '.....eq(1.10) −∞ 27 Chapter 1 where σr is the FCR cross section and N is the carrier density.76 If the carriers are generated from 2PA the equation becomes: dN β I 2 = ..................................eq(1.11) dt 2hω βσ r t 2 Δn(t ) = n2 I (t ) + I (t ' )dt '.....eq(1.12) 2hω −∫∞ When we compare the relative contributions of the accumulative and instantaneous nonlinearities to the overall NLR, we can see that for the case of 2PA-generated carriers, the accumulative refractive nonlinearity is effectively fifth order, while the Kerr nonlinearity is third order. This will then imply that the carrier related nonlinearity will dominate the Kerr nonlinearity above some excitation level. Lastly, devices that utilize nonlinear absorption for optical limiting might damage the limiter at high intensity due to the strong absorption. In comparison, devices with NLR of either self-focusing or defocusing work by refracting light away from the sensor, thus the limiter would still function at higher intensity. 1.5 Ultrafast electron and lattice dynamics Under laser excitation, electron and lattice excitation and relaxation processes in a direct gap semiconductor have been classified by Sundaram et al. into four distinct regimes: carrier excitation, thermalization, carrier removal, thermal and structural effects.86 For our short discussion below, we will only focus on the first 28 Chapter 1 three regimes and limit the scope to mainly QDs. It is important to note that many of the processes do not occur sequentially but may overlap in timescales. 1.5.1 Carrier Excitation Depending on the bandgap of the semiconductor and the laser excitation wavelength, single or multiphoton absorption (MPA) can occur as seen in Figure 1.8a. Single photon absorption is dominant when the band gap is smaller than the photon energy while MPA becomes more prominent when the band gap becomes larger than the photon energy or when the single photon absorption is inhibited by state filling or Pauli-blocking. FCA can also occur often with the assistance of phonons (Figure 1.8b) after the electron is already excited to the conduction bands by either single or multiphoton absorption. In certain materials, when the electron is excited to levels much higher than the bandgap and exceeding a certain threshold, impact ionization or multiple exciton generation (MEG) as seen in Figure 1.8c will occur generating additional excited carriers. For example, it was reported that the threshold for strong MEG for PbSe is 3Eg.87 29 Chapter 1 Figure 1.8 (a) Single photon and multiphoton excitation, (b) free carrier absorption (Intraband absorption); (c) multiexciton generation (Impact ionization); (d) carrier-carrier scattering (e) carrier–phonon intravalley scattering; (f) carrier–phonon intervalley scattering, (g) Auger-like electron-hole cooling; (h) surface trapped electron cooling via intermediate trap states; (i) electron cooling via resonant high-frequency vibration coupling; (j) radiative recombination; (k) Auger recombination; (l) trapping to surface/defect states, and (m) diffusion of excited carriers. 30 Chapter 1 1.5.2 Thermalization After excitation with a photon of energy in excess of the band gap by kT, the initial distribution of electrons (or holes) is athermal and is thermalized by carrier–carrier and carrier–phonon scattering which occurred concurrently.86,88 Carrier–carrier scattering as seen in Figure 1.8(d) involves an electrostatic interaction between two carriers and does not change the total energy of the excited carrier system or the number of carriers. It leads within 10 fs to a quasiequilibrium distribution exhibiting temperatures above that of the lattice. In a carrier–phonon scattering process, free carriers lose or gain energy and momentum by emission or absorption of a phonon. For bulk semiconductors, the carrier–phonon scattering process is very efficient and can occur within the same conduction or valence band valley (intravalley scattering; Figure 1.8e) or transfer to a different valley (intervalley scattering, Figure 1.8f). The carrier–phonon scattering typically takes hundreds of femntosecond to several picoseconds before the carriers and the lattice can reach thermal equilibrium. In quantum confined semiconductors, carrier cooling by phonon emission was originally proposed to occur less efficiently (“phonon bottleneck”) as the optical phonon energy is typically only ∼30 meV. This is much smaller than the minimum energy spacing of 100 meV between intraband levels in typical semiconductor NCs.89 However, investigations found that the carrier relaxation remains extremely fast for many QDs at sub-picosecond timescale. A non-phonon carrier-carrier 31 Chapter 1 scattering mechanism that has been proven to be very successful in explaining the breaking of the phonon bottleneck dilemma is the Auger-type e-h energy transfer mechanism (Figure 1.8g). This involves the transfer of the electron excess energy via strong coulomb coupling to the more massive hole, which subsequently relaxes rapidly through its dense spectrum of valence band states.90-92 Subsequently, picosecond timescale relaxation rates were observed when attempts were made to decouple the electron and hole. Mechanism such as intermediate trap states (Figure 1.8h) 93-96 and molecular vibrations (Figure 1.8i)97,98 were proposed. However, Califano et al. recently suggested that based on theoretical calculations Auger cooling can be used to explain the wide ranges of the electron relaxation rate without the need to invoke other new mechanisms.99 A breakthrough in electron cooling that extends the timescale to nanoseconds was accomplished by Pandey A et al. They synthesized a unique core-shell architecture that addresses the fast intraband relaxation induced by various mechanisms as described in Figure 1.8g-h.98 Pandey A et al found that by changing the shell compositions and thickness with identical core, they can change the cooling time from 6 ps to longer than 1 ns. When the QD has a composition of CdSe/ZnS1ML/ZnSe10ML with amine or carboxylate ligands, the exposed ZnSe surface leads to electron trapping and fast cooling via intermediate trap states. With a CdS outer layer (CdSe/ZnS1ML/ZnSe4ML/CdS1ML), electron trapping was reduced but hole-trapping was also prevented. Slow electron cooling, >1 ns, was observed with a CdSe outer layer (CdSe/ZnS1ML/ZnSe5ML/CdSe1ML) 32 Chapter 1 that removed electron traps and with hole trapping dodecanethiol ligands. The slow electron cooling was suggested to arise from an increased decoupling of the electron and hole with the thick ZnSe shell, electron traps that are removed from the CdSe layer and smaller energy transfer to the ligand vibrations of dodecanethiol due to their low IR absorbance. 1.5.3 Carrier removal At thermal equilibrium, the carriers and the lattice have the same temperature but an excess of free carriers is still present within the excitation region. The excess carriers are removed either by recombination of electrons and holes or by diffusion out of the excitation region. Radiative recombination between the excited electron and the hole leads to the loss of the excess carrier energy through the emission of a photon (fluorescence, Figure 1.8j). Efficient radiative recombination only occurs when other competitive nonradiative recombination processes have been minimized. Non-radiative recombination processes include Auger recombination, defect and surface trapping. During Auger recombination, an electron and a hole recombine and the excess energy excites a third charge carrier which can be either an electron or a hole (Figure 1.8k). Like other recombination mechanisms, Auger recombination decreases the carrier density. The Auger recombination rate for an intrinsic bulk semiconductor has been reported to depend on the cube of the carrier density.100 The lifetime τm of an mexciton state is also reported to be dependent on the size of the NC radii as R3.101 33 Chapter 1 In defect and surface recombination, the excess energy is given to a defect or surface state as seen in Figure 1.8l. Upon surface or defect trapping, the electron can recombine through either the nonradiative or radiative recombination mentioned earlier. Carrier diffusion removes carriers from the region of the sample where they are originally excited (Figure 1.8m) but does not decrease the total number of free carriers in the material. 1.5.4 Transient absorption (TA) Femtosecond transient absorption (TA) spectroscopy measures the change in transmission of a weak probe beam induced by an intense photoexcitation from the pump. Photoexcitation of the NCs by the pump across the NCs bandgap results in the occupation of the first excitonic state and a partial bleach of absorption of a probe pulse due to state-filling as well as a small red-shift due to a Stark effect as seen in Figure 1.9. If we probe the samples at different wavelengths, one can observe various nonlinear effects such as bleaching or excited state absorption. For example, if we probe at the first exciton energy, one observes a photoinduced bleach. However, if we probe at a larger wavelength that corresponds to the first intraband transition resonance, the probe photons induce 1S-1P transitions and thereby measure a photoinduced increase in absorption. 34 Chapter 1 Figure 1.9 Typical photoinduced absorption changes in a NC sample which allow the exciton population dynamics to be obtained.88 1.6 Nanoparticles in catalysis Nanostructured materials have shown tremendous potential as low cost alternatives for a variety of catalytic applications such as electrocatalysts for fuel cells and metal/air batteries102; catalysts for CO combustion103 and catalysts for oxidation of aromatic hydrocarbons104. Significant research is ongoing to improve the catalytic activity to produce catalysts that are both cheap and efficient. For example, self assembly of metal oxide nanoparticles to form ordered structures has led to an improvement to the catalytic selectivity and conversion yield.105 Porous metal oxide nanoparticles can be doped with metal ions such as Mo and V to form highly active catalysts for catalytic degradation of organic dye.106 35 Chapter 1 The gas phase reaction of CO combustion is a simple, yet extremely important process. For instance, hazardous CO can be formed by incomplete combustion of fuels and as a result contributes to air pollution. It is therefore important to effectively oxidize CO through catalytic combustion to form less harmful CO2 and H2O for environmental benefits. CO is also considered as a contaminant for pure H2 required for fuel cell development.107 The same catalysts for CO combustion are also useful for the abatement of volatile organic compounds like formaldehyde.108 The current catalysts for the combustion of CO and VOCs involve expensive noble metals such as platinum, palladium and rhodium.109 Hence, it is of great interest and need to develop cheaper alternative catalysts that can be effective at lower temperatures for this common oxidation process. Nanostructured materials are paramount in catalysis. Nanoparticles of Fe2O3110, CeO2111, Au112 and Pd113 have shown very promising results. Significant efforts have also been carried out on MnOx because of the advantages such as low cost, low toxicity, high chemical stability and high catalytic activity. For example, γ-MnO2 nanoparticles have shown tremendous potential especially when coupled with small amount of noble metal. These can be used for the elimination of VOC at ambient temperature, for their redox properties and the facile formation of radical species on the metal surface.114 Similarly, nanostructured α-Mn2O3 materials are excellent supports for anchoring gold nanoparticles to achieve high activities for low-temperature CO oxidation.115 Addition of a metal such as Ag to α-MnO2 leads to a strong interaction between 36 Chapter 1 them that can be correlated to the increased catalytic activity for the oxidation of CO.116 1.7 Objective and scope of thesis This project sets out to synthesize and investigate several nanomaterials with different properties that can be tuned by changing the particle size, shape or composition. The properties of interest include steady state PL, optical limiting, carrier dynamics, electrical conductivity and catalysis. Among the many important luminescent materials, PbS is particularly attractive due to its promising applications as near-infrared (NIR) emitting biological probe, electro-luminescent devices, nonlinear optics and solar cells.117-120 Bulk PbS has a narrow bandgap of ~0.4eV and a Bohr exciton radius of ~18 nm,121 and hence PbS nanoparticles with size smaller than the 18 nm are expected to show strong confinement effects. Thus, by controlling the size of the PbS QDs, its band edge emission can be widely tuned from the visible region to NIR region. In Chapter 3, we explored various synthesis methods for the formation of single layer and multi layer polystyrene (PS) or poly(methylmethacrylate) (PMMA) thin films uniformly embedded with CdS or PbS nanoparticles. The absorption and PL properties of the QD-films were studied in relation to the direction of excitation. PL enhancement by UV irradiation was found to be important for CdS QDs embedded in polymer films. The thickness of the PbS polymer films affected significantly the overall PL in the composite multilayer films. 37 Chapter 1 In the literature, efficient optical limiting was observed in PbS QDs stabilized in polymers, glasses and zeolites 78,119,122,123 However, most of these studies were conducted on PbS QDs synthesized in situ in different types of thin films while little was known about PbS NCs that are pre-synthesized before being dispersed in organic solvents or polymer films. Moreover, the sizes of the PbS studied vary greatly from different investigators with few attempts to study the size effect on their nonlinear optical (NLO) responses. Thus, we are motivated to carry out more detailed investigation on their NLO properties using the Z-scan technique. The size-dependent FCA, free-carrier refraction (FCR), optical Kerr nonlinearity and nonlinear scattering in PbS QDs with sizes between 4.5 nm to 11 nm dispersed in either hexane or in PS polymer films are investigated. In Chapter 4, we synthesized different sizes of core-shell PbS/CdS using a cation exchange method and studied the changes in their absorption and luminescence. Blue shift in the absorption and PL peaks was observed with changes in the PL QY. The nonlinear properties of core-shell QDs in hexane or polymer films were studied by both Z-scan technique and transient absorption using femntosecond pump-probe technique and compared to the results for pure PbS obtained in Chapter 3. The effect of free surface ligand and thickness of the polymer film on their nonlinear optical properties were studied. Moreover, we successfully prepared different thickness of QD-polymer films using a new imprinting-thermal cross-linking method. We found that the FCA cross-section and the nonlinear 38 Chapter 1 refractive index of the different thickness of QD-polymer films are the same within experimental error. Transient absorption studies of the pure PbS and coreshell PbS/CdS QDs were accomplished. In particular, the influence of the excitation intensity, pump wavelength and probe delay time on the transient differential transmittance spectra and relaxation kinetics of the PbS and core-shell PbS/CdS QDs were studied in detail. Chapter 5 is on the formation of PbS nanowires using potentiostatic or cyclic electrochemical deposition on a commercial AAO template. PbS nanowires have shown some potentials to sensitize the polymer and provide a better transport of the carriers for photovoltaic applications.34 Thus it is important to study the I-V characteristics of PbS nanowires to evaluate their potential integration into electronic devices. Moreover, different segmented and core-shell metal/lead sulfide nanowires were fabricated using a step-wise pore widening deposition method and a study of the mechanism for their formation was also carried out. In Chapter 6, we synthesized a series of different phases of nanocrystalline MnOx to study their applications as catalyst for the removal of CO. We explored in depth the crystal growth of MnO and Mn3O4 NCs. MnO was grown by a one pot heating up method that involves a nucleation burst at high temperature generating monodispersed cubic and rhombohedra NCs. Several important factors like temperature, ligand concentration and reaction time found to influence the final particle size or shape. Mn3O4 NCs were prepared by both a fast heating method 39 Chapter 1 and an injection method. Monomer concentration, temperature and reaction time were found to be important in influencing the varied shapes synthesized ranging from cubic, spherical NCs, long nanorods and rice shaped NCs. α-MnO2, γ-MnO2 and α-Mn2O3 NCs were also synthesized. The phase transformations of the various MnOx NCs during storage and during the CO oxidation process were found to be important factors in influencing its catalytic activity. The surface area was found to be less important in influencing the overall catalytic activity due to the competing factor of higher absorption strength of reactant gas molecules. 1.8 References (1) Dabbousi, B. O.; Bawendi, M. G.; Onitsuka, O.; Rubner, M. F. Appl. Phys. Lett. 1995, 66, 1316. (2) Murray, C. B.; Kagan, C. R.; Bawendi, M. G. Science 1995, 270, 1335. (3) Alivisatos, A. P. J. Phys. Chem. 1996, 100, 13226. (4) Tian, Z. R.; Tong, W.; Wang, J. Y.; Duan, N. G.; Krishnan, V. V.; Suib, S. L. Science 1997, 276, 926. (5) Lu, L.; Sui, M. L.; Lu, K. Science 2000, 287, 1463. (6) La Mer, V. K.; Dinegar, R. H. J. Am. Chem. Soc. 1950, 72, 4837. (7) Reiss, H. J. Chem. Phys. 1951, 19, 482. (8) Burda, C.; Chen, X. B.; Narayanan, R.; El-Sayed, M. A. Chem.Rev. 2005, 105, 1025. (9) Donega, C. D.; Liljeroth, P.; Vanmaekelbergh, D. Small 2005, 1, 1152. (10) Peng, X. G.; Wickham, J.; Alivisatos, A. P. J. Am. Chem. Soc.1998, 120, 5343. 40 Chapter 1 (11) Sugimoto, T. Adv. Colloid Interface Sci. 1987, 28, 65. (12) Bullen, C. R.; Mulvaney, P. Nano. Lett. 2004, 4, 2303. (13) Yin, Y.; Alivisatos, A. P. Nature 2005, 437, 664. (14) Peng, X. G. Adv. Mater. 2003, 15, 459. (15) Jun, Y. W.; Lee, J. H.; Choi, J. S.; Cheon, J. J.Phys.Chem. B 2005, 109, 14795. (16) Kwon, S. G.; Hyeon, T. Acc. Chem. Res. 2008, 41, 1696. (17) Seo, W. S.; Shim, J. H.; Oh, S. J.; Lee, E. K.; Hur, N. H.; Park, J. T. J. Am. Chem. Soc. 2005, 127, 6188. (18) Jun, Y. W.; Jung, Y. Y.; Cheon, J. J. Am. Chem. Soc. 2002, 124, 615. (19) Peng, X. G.; Manna, L.; Yang, W. D.; Wickham, J.; Scher, E.; Kadavanich, A.; Alivisatos, A. P. Nature 2000, 404, 59. (20) Manna, L.; Milliron, D. J.; Meisel, A.; Scher, E. C.; Alivisatos, A. P. Nat. Mater. 2003, 2, 382. (21) Seo, J. W.; Jun, Y. W.; Ko, S. J.; Cheon, J. J. Phys.Chem. B 2005, 109, 5389. (22) Lee, S. M.; Jun, Y. W.; Cho, S. N.; Cheon, J. J. Am. Chem. Soc. 2002, 124, 11244. (23) Manna, L.; Scher, E. C.; Alivisatos, A. P. J. Am. Chem. Soc. 2000, 122, 12700. (24) Jun, Y. W.; Casula, M. F.; Sim, J. H.; Kim, S. Y.; Cheon, J.; Alivisatos, A. P. J. Am. Chem. Soc. 2003, 125, 15981. (25) Alivisatos, A. P. Science 2000, 289, 736. (26) Penn, R. L.; Banfield, J. F. Am. Miner. 1998, 83, 1077. (27) Penn, R. L.; Banfield, J. F. Geochim. Cosmochim. Acta 1999, 63,1549. (28) Pacholski, C.; Kornowski, A.; Weller, H. Angew. Chem., Int. Ed. 2002, 41, 1188. 41 Chapter 1 (29) Pradhan, N.; Xu, H. F.; Peng, X. G. Nano. Lett. 2006, 6, 720. (30) Tang, Z. Y.; Kotov, N. A.; Giersig, M. Science 2002, 297, 237. (31) Cho, K. S.; Talapin, D. V.; Gaschler, W.; Murray, C. B. J. Am. Chem. Soc. 2005, 127, 7140. (32) Yu, J. H.; Joo, J.; Park, H. M.; Baik, S. I.; Kim, Y. W.; Kim, S. C.; Hyeon, T. J. Am. Chem. Soc. 2005, 127, 5662. (33) Yong, K. T.; Sahoo, Y.; Zeng, H.; Swihart, M. T.; Minter, J. R.; Prasad, P. N. Chem. Mater. 2007, 19, 4108. (34) Yong, K. T.; Sahoo, Y.; Choudhury, K. R.; Swihart, M. T.; Minter, J. R.; Prasad, P. N. Chem. Mater. 2006, 18, 5965. (35) Lee, J.; Govorov, A. O.; Dulka, J.; Kotov, N. A. Nano. Lett. 2004, 4, 2323. (36) Yu, Y.; Jin, C. H.; Wang, R. H.; Chen, Q.; Peng, L. M. J. Phys. Chem. B 2005, 109, 18772. (37) Lee, K. B.; Park, S.; Mirkin, C. A. Angew. Chem., Int. Ed. 2004, 43, 3048. (38) Fan, Z. Y.; Razavi, H.; Do, J. W.; Moriwaki, A.; Ergen, O.; Chueh, Y. L.; Leu, P. W.; Ho, J. C.; Takahashi, T.; Reichertz, L. A.; Neale, S.; Yu, K.; Wu, M.; Ager, J. W.; Javey, A. Nat. Mater. 2009, 8, 648. (39) Lahav, M.; Weiss, E. A.; Xu, Q. B.; Whitesides, G. M. Nano. Lett.2006, 6, 2166. (40) Hendren, W. R.; Murphy, A.; Evans, P.; O'Connor, D.; Wurtz, G. A.; Zayats, A. V.; Atkinson, R.; Pollard, R. J. J. Phys.-Condes. Matter 2008, 20, 5. (41) Guo, Y. B.; Tang, Q. X.; Liu, H. B.; Zhang, Y. J.; Li, Y. L.; Hu, W. P.; Wang, S.; Zhu, D. B. J. Am. Chem. Soc. 2008, 130, 9198. (42) Guo, Y. B.; Li, Y. L.; Xu, J. J.; Liu, X. F.; Xu, J. L.; Lv, J.; Huang, C. S.; Zhu, M.; Cui, S.; Jiang, L.; Liu, H. B.; Wang, S. J. Phys. Chem. C 2008, 112, 8223. (43) Shin, T. Y.; Yoo, S. H.; Park, S. Chem. Mater. 2008, 20, 5682. 42 Chapter 1 (44) Anderson, M. E.; Buck, M. R.; Sines, I. T.; Oyler, K. D.; Schaak, R. E. J. Am. Chem. Soc. 2008, 130, 14042. (45) Reiss, P.; Protiere, M.; Li, L. Small 2009, 5, 154. (46) Peng, X. G.; Schlamp, M. C.; Kadavanich, A. V.; Alivisatos, A. P. J. Am. Chem. Soc. 1997, 119, 7019. (47) Li, J. J.; Wang, Y. A.; Guo, W. Z.; Keay, J. C.; Mishima, T. D.; Johnson, M. B.; Peng, X. G. J. Am. Chem. Soc. 2003, 125, 12567. (48) Brumer, M.; Kigel, A.; Amirav, L.; Sashchiuk, A.; Solomesch, O.; Tessler, N.; Lifshitz, E. Adv. Funct. Mater.2005, 15, 1111. (49) Lifshitz, E.; Brumer, M.; Kigel, A.; Sashchiuk, A.; Bashouti, M.; Sirota, M.; Galun, E.; Burshtein, Z.; Le Quang, A. Q.; LedouxRak, I.; Zyss, J. J.Phys.Chem. B 2006, 110, 25356. (50) Koktysh, D. S.; McBride, J. R.; Dixit, S. K.; Feldman, L. C.; Rosenthal, S. J. Nanotechnology 2007, 18, 4. (51) Fernee, M. J.; Watt, A.; Warner, J.; Cooper, S.; Heckenberg, N.; Rubinsztein-Dunlop, H. Nanotechnology 2003, 14, 991. (52) Pietryga, J. M.; Werder, D. J.; Williams, D. J.; Casson, J. L.; Schaller, R. D.; Klimov, V. I.; Hollingsworth, J. A. J. Am. Chem. Soc. 2008, 130, 4879. (53) Xie, R.; Battaglia, D.; Peng, X. J. Am. Chem. Soc. 2007, 129, 15432. (54) Li, L.; Protiere, M.; Reiss, P. Chem. Mater. 2008, 20, 2621. (55) Li, L.; Reiss, P. J. Am. Chem. Soc. 2008, 130, 11588 (56) Xu, S.; Ziegler, J.; Nann, T. J. Mater. Chem.2008, 18, 2653. (57) Kim, S.; Fisher, B.; Eisler, H. J.; Bawendi, M. J. Am. Chem. Soc. 2003, 125, 11466. (58) Yu, K.; Zaman, B.; Romanova, S.; Wang, D. S.; Ripmeester, J. A. Small 2005, 1, 332. . 43 Chapter 1 (59) Chen, C. Y.; Cheng, C. T.; Yu, J. K.; Pu, S. C.; Cheng, Y. M.; Chou, P. T.; Chou, Y. H.; Chiu, H. T. J. Phys. Chem. B 2004, 108, 10687. (60) Cheng, C. T.; Chen, C. Y.; Lai, C. W.; Liu, W. H.; Pu, S. C.; Chou, P. T.; Chou, Y. H.; Chiu, H. T. J. Mater. Chem.2005, 15, 3409 (61) Xie, R. G.; Zhong, X. H.; Basche, T. Adv. Mater. 2005, 17, 2741. (62) Ivanov, S. A.; Piryatinski, A.; Nanda, J.; Tretiak, S.; Zavadil, K. R.; Wallace, W. O.; Werder, D.; Klimov, V. I. J. Am. Chem. Soc. 2007, 129, 11708. (63) Xie, R. G.; Kolb, U.; Basche, T. Small 2006, 2, 1454. (64) Balet, L. P.; Ivanov, S. A.; Piryatinski, A.; Achermann, M.; Klimov, V. I. Nano. Lett.2004, 4, 1485. (65) Wang, J.; Blau, W. J. J. Opt. A: Pure Appl. Opt 2009, 11, 024001. (66) Tutt, L. W.; Boggess, T. F. Prog. Quantum Electron. 1993, 17, 299. (67) Joudrier, V.; Bourdon, P.; Hache, F.; Flytzanis, C. Appl. Phys, B: Lasers Opt 1998, 67, 627. (68) Joudrier, V.; Bourdon, P.; Hache, F.; Flytzanis, C. Appl. Phys, B: Lasers Opt 2000, 70, 105. (69) Belousova, I. M.; Belousov, V. P.; Danilov, O. B.; Mironova, N. G.; Murav'eva, T. D.; Ryl'kov, V. V.; Skobelev, A. G.; Yur'ev, M. S.; Ponomarev, A. N. J. Opt. Technol. 2004, 71, 130. (70) Belousova, I. M.; Mironova, N. G.; Scobelev, A. G.; Yur'ev, M. S. Opt. Commun. 2004, 235, 445. (71) Mansour, K.; Soileau, M. J.; Vanstryland, E. W. J. Opt. Soc. Am. B-Opt. Phys. 1992, 9, 1100. (72) Francois, L.; Mostafavi, M.; Belloni, J.; Delaire, J. A. Phys. Chem. Chem. Phys. 2001, 3, 4965. (73) Francois, L.; Mostafavi, M.; Belloni, J.; Delouis, J. F.; Delaire, J.; Feneyrou, P. J. Phys. Chem. B 2000, 104, 6133. . 44 Chapter 1 (74) West, R.; Wang, Y.; Goodson, T. J. Phys. Chem. B 2003, 107, 3419. (75) Hernandez, F. E.; Shensky, W.; Cohanoschi, I.; Hagan, D. J.; Van Stryland, E. W. Appl. Optics 2002, 41, 1103. (76) Vanstryland, E. W.; Sheikbahae, M.; Said, A. A.; Hagan, D. J. Prog. Cryst. Growth Charact. Mater. 1993, 27, 279 (77) Qu, Y. L.; Jia, W.; Zheng, Y. G.; Ying, J. Y. Appl. Phys. Lett. 2007, 90, 133112. (78) Venkatram, N.; Kumar, R. S. S.; Rao, D. N. J . Appl. Phys.2006, 100, 074309. (79) He, G. S.; Yong, K. T.; Zheng, Q. D.; Sahoo, Y.; Baev, A.; Ryasnyanskiy, A. I.; Prasad, P. N. Opt. Express 2007, 15, 12818. (80) Nikesh, V. V.; Dharmadhikari, A.; Ono, H.; Nozaki, S.; Kumar, G. R.; Mahamuni, S. Appl. Phys. Lett. 2004, 84, 4602. (81) Padilha, L. A.; Fu, J.; Hagan, D. J.; Van Stryland, E. W.; Cesar, C. L.; Barbosa, L. C.; Cruz, C. H. B. Opt. Express 2005, 13, 6460. (82) Venkatram, N.; Sathyavathi, R.; Rao, D. N. Opt. Express 2007, 15, 12258. (83) Lad, A. D.; Kiran, P. P.; More, D.; Kumar, G. R.; Mahamuni, S. Appl. Phys. Lett. 2008, 92, 043126. (84) Said, A. A.; Sheikbahae, M.; Hagan, D. J.; Wei, T. H.; Wang, J.; Young, J.; Vanstryland, E. W. J. Opt. Soc. Am. B-Opt. Phys. 1992, 9, 405. (85) Sheikbahae, M.; Said, A. A.; Wei, T. H.; Hagan, D. J.; Vanstryland, E. W. IEEE J. Quantum Electron. 1990, 26, 760. (86) Sundaram, S. K.; Mazur, E. Nat. Mater. 2002, 1, 217. (87) Ellingson, R. J.; Beard, M. C.; Johnson, J. C.; Yu, P. R.; Micic, O. I.; Nozik, A. J.; Shabaev, A.; Efros, A. L. Nano Lett. 2005, 5, 865. (88) Beard, M. C.; Ellingson, R. J. Laser & Photonics Rev. 2008, 2, 377. . 45 Chapter 1 (89) Benisty, H.; Sotomayortorres, C. M.; Weisbuch, C. Phys. Rev. B 1991, 44, 10945. (90) Efros, A. L.; Kharchenko, V. A.; Rosen, M. Solid State Commun. 1995, 93, 281. (91) Hendry, E.; Koeberg, M.; Wang, F.; Zhang, H.; Donega, C. D.; Vanmaekelbergh, D.; Bonn, M. Phys. Rev. Lett. 2006, 96, 057408. (92) Wang, L. W.; Califano, M.; Zunger, A.; Franceschetti, A. Phys. Rev. Lett. 2003, 91, 056404. (93) Guyot-Sionnest, P.; Shim, M.; Matranga, C.; Hines, M. Phys. Rev. B 1999, 60, R2181. (94) Klimov, V. I.; Mikhailovsky, A. A.; McBranch, D. W.; Leatherdale, C.A.; Bawendi, M. G. Phys. Rev. B 2000, 61, R13349. (95) Schroeter, D. F.; Griffiths, D. J.; Sercel, P. C. Phys. Rev. B 1996, 54, 1486. (96) Sercel, P. C. Phys. Rev. B 1995, 51, 14532. (97) Guyot-Sionnest, P.; Wehrenberg, B.; Yu, D. J. Chem. Phys. 2005, 123, 074709. (98) Pandey, A.; Guyot-Sionnest, P. Science 2008, 322, 929. (99) Califano, M. J.Phys.Chem. C 2008, 112, 8570. (100) Metzger, W. K.; Wanlass, M. W.; Ellingson, R. J.; Ahrenkiel, R. K.; Carapella, J. J. Appl. Phys. Lett. 2001, 79, 3272. (101) Klimov, V. I.; Mikhailovsky, A. A.; McBranch, D. W.; Leatherdale, C. A.; Bawendi, M. G. Science 2000, 287, 1011. (102) Cheng, F. Y.; Shen, J.; Ji, W. Q.; Tao, Z. L.; Chen, J. ACS Appl. Mater. Interfaces 2009, 1, 460. (103) Liang, S. H.; Bulgan, F. T. G.; Zong, R. L.; Zhu, Y. F. J. Phys. Chem. C 2008, 112, 5307. (104) Jin, L.; Chen, C. H.; Crisostomo, V. M. B.; Xu, L. P.; Son, Y. C.; Suib, S. L. Appl. Catal. A-Gen. 2009, 355, 169. 46 Chapter 1 (105) Yuan, J. K.; Laubernds, K.; Zhang, Q. H.; Suib, S. L. J. Am. Chem. Soc. 2003, 125, 4966. (106) Sriskandakumar, T.; Opembe, N.; Chen, C. H.; Morey, A.; King'ondu, C.; Suib, S. L. J. Phys. Chem. A 2009, 113, 1523. (107) Han, Y. F.; Kahlich, M. J.; Kinne, M.; Behm, R. J. Appl.Catal., B 2004, 50, 209. (108) Lamaita, L.; Peluso, M. A.; Sambeth, J. E.; Thomas, H. J. . Appl.Catal., B 2005, 61, 114. (109) Faticanti, M.; Cioffi, N.; De Rossi, S.; Ditaranto, N.; Porta, P.; Appl.Catal., B Sabbatini, L.; Bleve-Zacheo, T. 2005, 60, 73. (110) Li, P.; Miser, D. E.; Rabiei, S.; Yadav, R. T.; Hajaligol, M. R Appl.Catal., B 2003, 43, 151. (111) Chen, C. L.; Weng, H. S. . Appl.Catal., B 2005, 55, 115. (112) Choudhary, T. V.; Sivadinarayana, C.; Chusuei, C. C.; Datye, A. K.; Fackler, J. P.; Goodman, D. W J. Catal.. 2002, 207, 247. (113) Hwang, C. B.; Fu, Y. S.; Lu, Y. L.; Jang, S. W.; Chou, P. T.; Wang, C. R. C.; Yu, S. J. J. Catal. 2000, 195, 336. (114) Sinha, A. K.; Suzuki, K.; Takahara, M.; Azuma, H.; Nonaka, T.; Fukumoto, K. Angew. Chem., Int. Ed. 2007, 46, 2891. (115) Chen, W. S.; Kocal, J. A.; Brandvold, T. A.; Bricker, M. L.; Bare, S. R.; Broach, R. W.; Greenlay, N.; Popp, K.; Walenga, J. T.; Yang, S. S.; Low, J. J. Catal. Today 2009, 140, 157. (116) Xu, R.; Wang, X.; Wang, D. S.; Zhou, K. B.; Li, Y. D. J. Catal. 2006, 237, 426. (117) Choi, J. H.; Chen, K. H.; Strano, M. S. J. Am. Chem. Soc. 2006, 128, 15584. (118) Bakueva, L.; Musikhin, S.; Hines, M. A.; Chang, T. W. F.; Tzolov, M.; Scholes, G. D.; Sargent, E. H. Appl. Phys. Lett. 2003, 82, 2895. (119) Kim, H. S.; Lee, M. H.; Jeong, N. C.; Lee, S. M.; Rhee, B. K.; Yoon, K. B. J. Am. Chem. Soc. 2006, 128, 15070. 47 Chapter 1 (120) Johnston, K. W.; Pattantyus-Abraham, A. G.; Clifford, J. P.; Myrskog, S. H.; MacNeil, D. D.; Levina, L.; Sargent, E. H. Appl. Phys. Lett. 2008, 92, 151115. (121) Elim, H. I.; Ji, W.; Yang, J.; Lee, J. Y. Appl. Phys. Lett. 2008, 92, 251106. (122) Kurian, P. A.; Vijayan, C.; Sandeep, C. S. S.; Philip, R.; Sathiyamoorthy, K. Nanotechnology 2007, 18, 075708. (123) Asunskis, D. J.; Bolotin, I. L.; Hanley, L. J. Phys. Chem. C 2008, 112, 9555. 48 Chapter 2 Chapter 2 Experimental 2.1 Chemical reagents used Table 2.1 Chemical and solvents used in the work described in this thesis; their purity and sources. Purity ≥ 90% 99% 99% Formula CH2=CH(CH2)15CH3 C5H12O C6H5COC(OCH3)2C6H5 Source Aldrich Aldrich Acros 98% C10H20O5Si Aldrich AR grade 99% 99% OC(CH3)2. Merck Al2O3 C2H8N2O4.H2O Aldrich Fluka AnalaR Fluka 97% 98% (NH4)2S2O8 (CH3)2(CN)CN=NC(CH3)2 CN [C6H5C(O)]2O2 Cd(CH3COO)2·2H2O Cadmium Chloride 98% CdCl2 Copper Acetate Monohydrate Copper Sulphate pentahydrate Cetyl trimethyl ammonium bromide (CTAB) Chloroform 99.9% Cu(CH3COO)2·H2O 99% CuSO4·5 H2O Fischer Scientific Alfa Aesar Merck - C16H33NCH3 Br Aldrich HPLC grade 80% 95% 99.9% - CHCl3 Tedia C6H4(HC=CH2)2 CH3(CH2)9SH CH3CH3OH commercial Aldrich Fluka MERCK Technic 99.3% CH3(CH2)5CH3 JT Baker Chemical 1-Octadecene 1-Pentanol 2,2-Dimethoxy-2phenylacetophenone (DMPA) 3-(Methacryloyloxy)propyl] trimethoxysilane (γ-MPS) Acetone Calcined alumina Ammonium oxalate monohydrate Ammonium persulfate Azobisisobutyronitrile (AIBN) Benzoyl peroxide Cadmium Acetate Dihydrate Divinylbenzene Decanethiol Ethanol Gold plating solution OROTEMP RTU24 n-heptane 99% 98% Fluka Aldrich 49 Chapter 2 Hexadecylamine (HDA) Hexane IR125 dye Isopropanol Lead Acetate Dihydrate Lead Chloride Manganese (III) Acetate Dihydrate Manganese Acetate(II) Tetrahydrate Manganese (II) Nitrate Tetrahydrate Methanol Methyl methacrylate Oleic acid 99% HPLC grade 80% 99.5 99% 98% 98% CH3(CH2)15NH2 CH3(CH2)4CH3 Fluka TEDIA C43H47N2NaO6S2 (CH3)2CHOH Pb(CH3COO)2·2H2O PbCl2 Mn(CH3COO)3·2H2O Aldrich Tedia FLuka Alfa Aesar Merck 99% Mn(CH3COO)2·4H2O Fluka 97% Mn(NO3)2·4H2O Fluka 99.9% CH3OH 99% 90% CH2=C(CH3)CO2CH3 Fisher Scientific Merck Aldrich Oleylamine (OLA) 70% Phenyl Ether Potassium permanganate Sodium Carbonate Decahydrate Sodium salt, ethylenediaminetetraacetic acid(Na2EDTA) Sodium Hydroxide 98% 99% 99% CH3(CH2)7CH=CH(CH2)7 COOH CH3(CH2)7CH=CH(CH2)8 NH2 O(C6H5)2 KMnO4 Na2CO3·10H2O 99.5% C10H14N2Na2O8 JT Baker 99% NaOH 98% 99%, 99% 97% 90% HPLC grade 90% Na2S C6H5CH=CH2 S H2SO4 C6H5COSH C6H5CH3 CHEMICO N Acros Fluka Chemicon Merck Fluka Fisher Scientific Fluka Sodium Sulfide Styrene Sulfur Sulfuric acid Thiobenzoic Acid Toluene Trioctylphosphine (TOP) [CH3(CH2)7]3P Aldrich Fluka GCE Dumont 50 Chapter 2 2.2 Synthesis procedures 2.2.1 Preparation of nanosized PbS We adapted the literature method with slight modification for the synthesis of PbS QDs reported in Chapter 3.1 Briefly, 9.6 mmol of PbCl2 was mixed in 16 mL of oleylamine in a 50 mL 3-neck flask. The solution was vigorously stirred and heated to 100oC and degassed for 5 minutes before it was heated and stabilized at the reaction temperature (110 oC or 120 oC ) for 30 minutes. A red solution was formed by dissolving 0.0451g of sulfur in 12 mL of oleylamine in a 2-neck flask by heating to 80oC. 8 mL of the resulting red solution was withdrawn and swiftly injected into the 3-neck flask containing the PbCl2. The solution was allowed to grow at different temperature (100 oC, 120 oC and 210 oC) and different reaction time from 1 minutes up to 1 hour depending on the size of the QD required. At the desired time, 4 mL of the solution was withdrawn and quenched by injecting into 4 mL of cold hexane. The flask was further cooled by immersing into an ice bath. Ethanol was added to the solution and centrifuged to precipitate the PbS nanoparticles along with some PbCl2. The precipitate was re-dispersed in hexane and then added with twice the amount of oleic acid. The PbCl2 was removed by re-dispersing the precipitate in hexane and centrifuging to obtain a dark red supernatant. The dark red supernatant solution was then precipitated with ethanol to obtain the PbS precipitate. This precipitate was dried to form PbS nanoparticles 51 Chapter 2 powder, which were dissolved in hexane to concentration 5×10-3 M for all our optical measurements. 2.2.2 Preparation of styrene pre-polymer Styrene monomer was vacuum distilled to remove any inhibitor added by the manufacturer for storage purpose. 1 weight % of benzoyl peroxide was added into the distilled monomer and the mixture was stirred under nitrogen for 5 minutes. The mixture was then heated to 80°C to initiate the polymerization reaction. The temperature was held for 3 hours until a clear viscous pre-polymer was obtained. 2.2.3 Preparation of PbS/PS nanocomposites The PbS nanoparticles prepared in Section 2.2.1 were dispersed into decanethiol by sonicating the mixture for 5 minutes. Pre-polymer from Section 2.2.2 was added to the solution and the mixture was stirred for a further 10 minutes before 3% wt of 2,2-dimethoxy-2-phenylacetophenone (DMPA) photoinitiator was added. Several drops of the mixture was drop-casted onto a glass plate or a small aluminum tray and cured under UV lamp (356nm, 12W) for 6 hours. 2.2.4 Preparation of CdS/PMMA nanocomposites The preparation of cadmium (II) thiobenzoate (CdTB) precursor is as follows. Briefly, 2.9 g of Na2CO3.10H2O was dissolved in 20 mL of water. 2.4 mL of 52 Chapter 2 thiobenzoic acid was added to the sodium carbonate solution and stirred for 30 minutes. 2.3 g of CdCl2 in 20 mL of water was added dropwisely into the stirring mixture and allowed to react for 1 h. The cream color precipitate formed was filtered, washed thoroughly with water, ethanol and acetone, and dried overnight in vacuum. All procedures for the preparation of CdS QDs were carried out using standard techniques under a nitrogen atmosphere. Hexadecylamine (HDA) was dried and degassed at 120oC before use. A degassed solution of CdTB (0.04 g) in trioctylphosphine (TOP) (0.2 mL) was injected into a hot solution of 0.395g of HDA at 120oC. After 10 minutes, the reaction solution was cooled to room temperature and the product was precipitated by adding ethanol. The precipitate was centrifuged, washed thoroughly with ethanol, and dried in vacuum for 10 minutes. No size sorting was needed for all the samples. CdS QDs synthesized by our precursor method can be incorporated into a PMMA film by the polymerization of the CdS-MMA monomer mixture. Briefly, 1.5 mL of distlled MMA was added to 8mg of CdS QDs and 3.5 mg of azobisisobutyronitrile (AIBN). The mixture was sonicated to form a clear yellowish solution. The solution was heated under nitrogen to 80oC and reacted for an additional 45-50 minutes to form a viscous yellow solution. CdS/PMMA films can next be formed by drop casting the slightly viscous solution onto a cleaned glass slide and further subjected to UV illumination using an UV lamp at 365 nm for 8 hours. 53 Chapter 2 2.2.5 Preparation of heterogenous layered nanocomposites Using the methods given in Sections 2.2.3 and 2.2.4, layered nanocomposites were prepared as shown in Figure 2.1. The first layer of composite consisted of PbS QDs which absorb in 1200 nm region, dispersed in PS and cast on a glass slide and cured. After the first layer was harden, the second layer of styrene prepolymer without any PbS was cast on top and cured for a further 6 hours under UV lamp. The third layer of composite which consisted of PbS particles which absorb in 1600 nm range was then laid onto the pure PS layer. In order to make a comparison with mixed composites, a single layered composite was prepared by sonicating two sizes of PbS QDs together in decanethiol before adding prepolymer and initiator followed by curing under UV lamp. Figure 2.1 Schematic diagram of layered PbS-PS nanocomposites studied in Chapter 3. 54 Chapter 2 Heterogeneous nanocomposites containing layers of different types of QDs were also prepared as shown in Figure 2.2. The first layer of composite consisted of PbS which absorb in 1200 nm region, dispersed in PS and cast onto a glass slide and cured. Thick layer of PbS/PS was prepared using the same method as in Section 2.2.3, while much thinner PbS/PS layer was prepared by adding toluene to the prepolymer mixture. After the first layer was hardened, the second layer of CdS in PMMA was formed by the same drop casting method for the formation of single layer CdS/PMMA. The composite was then subjected to UV illumination for complete polymerization of the CdS layer. Excitation source (Top) CdS/PMMA PbS/PS Excitation source (Bottom) Figure 2.2 Schematic diagram of layered PbS/PS-CdS/PMMA nanocomposites studied in Chapter 3. 55 Chapter 2 2.2.6 Imprinting-thermal cross-linking QD-polymer film This methodology was specially developed for the preparation of micron thickness composite films. Firstly, polished glass plates A and B were soaked in concentrated sulfuric acid for 5 minutes and rinsed extensively with deionized water. To enhance the adhesion between crosslinked polymers and the glass plate A, surface modification (Methacryloyloxy)propyl] was conducted. trimethoxysilane A (γ-MPS, coupling 98%, agent, [3- Aldrich) was hydrolyzed in methanol/water (80:20 by volume) under a pH value of 4. Glass plate A was then immersed in this solution for 10 minutes. After treatment, glass plate A was rinsed in water, isopropanol and acetone sequentially before use. A thermally cross-linkable recipe, containing styrene (99%, Fluka) as monomer, divinylbenzene (80%, mixture of isomers, Aldrich) as cross-linker and benzyl peroxide (Aldrich) as thermal initiator in the molar ratio of 60:13:1, was mixed and stirred over 6 hours. Filtration through a 0.2 μm filter was performed and then the QDs were dispersed into the mixture by sonication to form the precursor. A small amount of this precursor was added to the surface of the glass plate A, and the glass plate B was brought into contact. The precursor was allowed to spread between the two plates and then the imprinting was carried out on a 4-inch imprinter (Obducat Inc.) at a temperature of 110 oC for a period of 10 minutes under a pressure of 4 MPa. The glass plate B was carefully separated from A, yielding an even QD polymer composite film on glass plate A. Post-cure treatment under vacuum was carried out at 110 oC for a period of one hour. 56 Chapter 2 Multilayer films were also deposited sequentially using the above method with a baking period of a few hours before the second layer is applied to ensure the earlier layer is stable. 2.2.7 Preparation of PbS/CdS core-shell nanoparticles The preparation of PbS/CdS core-shell QDs was adapted with slight modification from a previously reported method.2 A cadmium oleate stock solution was prepared by heating up a 100ml three neck flask with 9.6mmol of Cd(Ac)2.2H2O, 7.6ml of oleic acid and 20ml of phenyl ether(or 30ml of 1-octadecene) to 150oC. The solution is then cooled down to 60oC (80oC) and degassed for 2hours at this temperature before heating up to 155oC. For the formation of 5.0nm PbS/CdS( Sample A in Table 2.2), the total yield from the formation of 5nm PbS QD reaction is dispersed in 12ml of toluene and added to another three neck flask and flushed with nitrogen for 30 minutes before heating up to 100oC using an oil bath. The cadmium oleate stock solution was cannula transferred at 155oC to the toluene solution which was at 70oC. Upon transfer the temperature of the solution raises rapidly and growth was continued at 100oC for 20 hours. Aliquots were removed at 1 hour and quenched by injecting into a 3ml of hexane. At 20 hours of reaction the solution is allowed to cool to 50oC whereby 8ml of hexane is added to the solution. Equal volume of methanol is added twice to extract away the byproducts and the solution is finally 57 Chapter 2 precipitated with a methanol/ethanol mixture. The ppt is redispersed in hexane and washed with the methanol/ethanol mixture thrice to remove the excess oleic acid ligand. Sample B-D is prepared using the same method as sample A but only a quarter of the yield from the formation of 6nm and 7.3nm PbS QD reaction is used as described in Table 2.2. Table 2.2 Sizes of core-shell QDs and the corresponding reaction time Sample Size of PbS/CdS( nm) Reaction time(hours) A 5.0 20 B 5.9 2 C 6.1 20 D 7.2 1 To study the effect of the oleic acid ligand on the nonlinear scattering of the QDs, the 5.0nm PbS/CdS is prepared as normal. The crude solution is split into two tubes. The QDs with excess oleic acid is prepared by just extraction with methanol once, precipitated with ethanol and redispersed in hexane and precipitated with ethanol. The QDs with excess oleic acid removed are prepared by the same procedure as normal with twice methanol extraction and four times of methanol/ethanol washing. 58 Chapter 2 2.2.8 Electrochemical preparation of PbS and composite nanowires A commercial AAO membrane (Anodisc 47, Whatman Co., thickness ≈ 60 μm, pore density ≈ 109 pores/cm2) was used. SEM analysis revealed that the pore sizes are 100 and 200 nm respectively on the two sides of the membrane. A thin layer of ~300 nm gold film was sputtered onto the 100 nm-pore side of the AAO and this served as the working electrode by contact with a copper foil. Electrodeposition of PbS NWs was performed using potentiostat Autolab PGSTAT 30. In the three electrode cell, Ag/AgCl (3 M KCl) was used as reference electrode and a platinum rod as the counter electrode. The electrodeposition bath consisted of 0.01 M Pb(CH3COO)2, 0.1 M Na2EDTA and 0.01M Na2S in ultrapure water. Crystalline PbS NWs can be deposited in AAO by using either a constant potential method or cyclic voltammetry (CV) scanning. For the constant potential method, potentials ranging between -0.7 to -0.8V were utilized for the deposition. CV scanning was performed by cycling the potential between -0.4V and -0.95V. Scanning rates from 0.05V/s to 1V/s were utilized to investigate their effect on the length of the NWs produced. Core-shell Cu/PbS NWs were synthesized sequentially by first depositing Cu NWs into the AAO template for a total of 4C using 0.5M CuSO4 bath at pH = 1. The template was thoroughly washed with water before immersing in 6% 59 Chapter 2 phosphoric acid for 30 minutes to widen the annular gaps surrounding the core Cu NWs (i.e. “Pore Widening” procedure).3 This is then followed by several washing and immersion in deionized water for 5-10 minutes to ensure all the acid within the pores has been removed. PbS NWs were next electrodeposited into the annular gaps to form a shell layer surrounding the Cu core NWs by using the same procedures given above. Au/PbS NWs were synthesized in a similar manner except that 4C of gold instead of Cu were first electrodeposited at -1V using a commercial OROTEMP 24 RTU gold solution. The AAO template could be dissolved by soaking into 0.5M NaOH for 45 minutes to give freestanding NWs on a gold base. In order to isolate the NWs from the base, the sample has to be sonicated in ethanol for a few minutes. The ethanol solution containing the free NWs was dropped onto either copper or nickel grids for TEM characterization. 2.2.9 Preparation of MnO Nanocrystals MnO nanocrystallites were prepared via the decomposition of manganese (III) acetate (Mn(CO2CH3)3) using a method adapted from previous studies.4,5 A one pot solution of 1 mmol of manganese (III) acetate dihydrate (Mn(Ac)3.2H2O), 3 mmol (0.96ml) of oleic acid (OA, C18H34O2) and 6 mL of 1-octadecene (ODE, CH3(CH2)15CH=CH2) was degassed at 80oC for one hour in a three-necked roundbottom flask, purging with N2 three times. The resulting dark brown solution was 60 Chapter 2 then heated to 300°C under N2 environment. Upon observing a sudden change in colour from clear orange/yellow to dirty green, the solution was heated for an additional hour for annealing. The reaction solution was then cooled to room temperature, hexane (ca. 4 mL) was added and the product was precipitated by the addition of twice the amount of acetone. The nanocrystals were collected by centrifugation and washed with acetone/ethanol for several times. After vacuum drying, black powders were obtained. The effect of OA concentration and temperature on the growth process will be discussed in Sections 6.1.2 and 6.1.3. 2.2.10 Preparation of Mn3O4 nanorods Manganese (II) nitrate tetrahydrate was weighed in a drybag filled with N2 environment as it is highly sensitive to moisture. For the direct heating method, a solution of the nitrate and oleylamine (in mole ratio of 1:8) was stirred and heated to 180°C under N2 environment at a rate of approximately 8-9OC/minute. During the heating process, the color of the mixture changed from orange brown to beige and finally dark reddish brown. After 45 minutes from the start of the heating, the reaction was cooled and centrifuged. The precipitate was washed with toluene once, precipitated with acetone and centrifuged. Hexane was added to the resulting precipitate forming a dark reddish brown solution. The small amount of precipitate that cannot be dissolved was discarded and the dark reddish brown solution left was precipitated with acetone. The final precipitate was dried in vacuum overnight and dark brown powder was obtained. 61 Chapter 2 For the injection method, 0.6 g of manganese (II) nitrate tetrahydrate was weighed in a drybag with N2 environment into a two necked flask and 4 mL of oleylamine was then added to the flask to form a suspension. The suspension was heated to 70oC but with occasional sonication to ensure a homogeneous suspension was formed during the heating. 3 mL of the suspension was injected into a preheated three neck flask at 190OC filled with 3.2 mL of oleylamine and subsequently the flask was maintained at 180oC. Upon injection, the solution turned dark brown in two minutes indicating the formation of Mn3O4 rods. The solution was heated for 1-3 hours to study the effect of heating time on the shape of the rod before the reaction was cooled (Results discussed in Section 6.2.1). The dark brown solution was topped up with 2-3 mL of hexane, precipitated with acetone and centrifuged. Hexane was added to the resulting precipitate and centrifuged. The small amount of precipitate was discarded and the dark reddish brown solution left was precipitated with acetone again. The final precipitate was dried in vacuum overnight and dark brown powder was obtained. 2.2.11 Preparation of MnO2 Nanocrystals MnO2 nanocrystallites were synthesized by two pathways: reduction and oxidation. Reduction method With reference to a previous study6, a solution of potassium permanganate (KMnO4) and sulfuric acid (H2SO4) in molar ratio of 1:20 in deionized water was 62 Chapter 2 set to reflux at 100°C under N2 environment in a three-necked round-bottom flask. After 3 h, the mixture was cooled to ~80°C. The product was washed thoroughly with ethanol and deionized water. The mixture was centrifuged and the precipitate was washed with ethanol. This washing process was repeated thrice. The final precipitate was dried in vacuum overnight. Black powder was obtained and characterized. Oxidation method With adaptation from a previous study7, a solution of manganese (II) acetate tetrahydrate (Mn(CO2CH3)2) and ammonium persulfate ((NH4)2S2O8) in molar ratio of 1:1 in deionized water was set to reflux at 100°C at atmosphere for 3 h in a three-necked round-bottom flask. The washing process was the same as above. Black powder was obtained and characterized. 2.2.12 Preparation of nanosized Mn2O3 Mn2O3 nanocrystallites were prepared by thermal decomposition of manganese oxalate, which was prepared by the reverse micellar method with reference to a previous study.8 A colourless solution of n-heptane, 1-pentanol, cetyl trimethyl ammonium bromide (CTAB) and ammonium oxalate monohydrate was added to another colourless solution of n-heptane, isopentane, CTAB and manganese (II) acetate tetrahydrate. The resulting cloudy solution was left to stir overnight at 63 Chapter 2 room temperature and atmospheric pressure. The product was precipitated with chloroform/methanol (1:1) mixture. The mixture was centrifuged and the precipitate was washed with chloroform/methanol once and dried in vacuum for 4–5 h. The white flakes obtained were heated in an oven at 120°C for 1 h. The resulting beige flakes were heated in a furnace in air at 500°C for 6 h. A dark brown powder was obtained and characterized. 2.3 Characterization techniques 2.3.1 Transmission Electron Microscopy (TEM) and High Resolution TEM (HRTEM) TEM was performed on a JEOL JEM 2010F field emission electron microscope with an acceleration voltage of 200 kV. HRTEM and EDX were performed on a JEOL 3010 Electron Microscope operating at an acceleration voltage of 300 kV. The nanoparticles were dispersed in hexane or toluene, dripped onto a 200 mesh carbon-coated copper grid and dried in vacuum before analysis. 2.3.2 Scanning Electron Microscopy (SEM) SEM images were obtained using a JEOL JSM 6701-F microscope operating at 10μA and 5.0kV which also has an EDX component for elemental analysis. 64 Chapter 2 2.3.3 Powder X-ray Diffraction (XRD) XRD patterns were obtained using Siemens D5005 X-ray Diffractometer with Cu Kα radiation (λ = 0.15406 nm). The purified nanocrystals were crushed into fine powder and placed directly on the sample holder or onto double-sided tapes that were mounted onto the sample holder for samples of low yields. 2.3.4 Elemental Analysis (EA) Carbon, hydrogen, nitrogen and sulfur (CHNS) analysis was performed on an Elementar Vario Micro Cube. Inductively Coupled Plasma (ICP) analysis was carried out using a PerkinElmer Optima 5300 DV ICP-OES system 2.3.5 Ultraviolet-visible (UV-VIS-NIR) Absorption Spectroscopy Absorption spectra were recorded on a Shimadzu UV-3600 UV-VIS-NIR spectrophotometer using pure hexane or TCE as reference. Solid measurements were recorded using an integrating sphere attachment (ISR-3100) with photomultiplier and PbS detector. 65 Chapter 2 2.3.6 Steady-state Photoluminescence Spectroscopy (PL) Steady-state PL spectra were collected with a Jobin Yvon Fluorolog-3 modular spectrofluorometer system coupled with a 450 W Xe lamp. The PL spectra for IR emission region up to 1600 nm were recorded using a Horiba Jobin Yyon FluoroLog®-3 with an iHR320 attachment equipped with lock-in amplifier and liquid nitrogen cooled InGaAs photodiode detector. Solid state measurements were performed using the same instruments with a solid sample holder. Determination of PL quantum yield (QY) The PL QY of PbS and PbS/CdS core/shell nanoparticles were determined with a protocol as described below. Firstly, PbS and PbS/CdS core/shell nanoparticles were dissolved in hexane and their absorption spectra were recorded in 1 cm cuvettes. The absorbance values at the excitation wavelength used for PL measurement, i.e. 715 nm, were recorded. Next, PL spectra of the same solutions were measured at the above excitation wavelength. The integrated PL intensity, i.e. the area of PL emission peak, was calculated from the spectrum. The above two steps were repeated for four solutions with increasing concentrations, while the absorbance of all solutions were adjusted below 0.1 in order to minimize reabsorption effect. A plot of the integrated PL intensity versus absorbance followed a straight line passing through zero are presented in appendices A and B. 66 Chapter 2 IR125 was used as the standard in our QY measurement. It has been reported to have a QY of 0.04 in methanol solution. The absorption and PL emission spectra (excited at 715 nm) of IR125 solutions in methanol were recorded in the same way as mentioned above. The gradient GradST (ST stands for “standard”) was obtained from this plot. The QY of our nanoparticles in hexane was then calculated using the following equation: Grad X η X Φ X = Φ ST ( )( ) Grad ST η ST 2 2 ( Equation 2.1) where the subscripts ST and X denote IR125 and our samples respectively, Φ is the PL QY and η the refractive index of the corresponding solvents (1.329 for methanol and 1.3749 for hexane). 2.3.7 Profile meter The thickness of QD-polymer films on glass substrate studied in Chapter 3 is measured by an Alpha-step 500 surface profiler 67 Chapter 2 2.38 Thermal Gravimetric Analysis (TGA) TGA was performed on a SDT 2960 Simultaneous DTA-TGA analyzer. Approximately 10 mg of the sample was measured under a flow of nitrogen gas (flow rate 100 mL/minute) at a heating rate of 20 degree/minute. 2.3.9 Fourier Transform Infrared (FT-IR) FT-IR spectra were obtained using Varian 3100 FT-IR Excalibur Series spectrometer. Samples were grinded with KBr and pressed into disks. Spectra were recorded at room temperature in 32 scan cycles that were signal-averaged at a resolution of 4 cm-1. 2.3.10 Atomic force Microscopy AFM images were obtained with tapping mode using a Dimension 3100 Scanning Probe Microscope ( Digital Instrument Veeco Metrology Group). Conductive AFM measurements were conducted with the Dimension 3100 Scanning Probe Microscope by using a silicon probe with Cr/Pt conductive coating at a resonant frequency of 13 kHz scanning from either -2 to 2 V or -5 to 5 V. 68 Chapter 2 2.4 Nonlinear and transient optical studies 2.4.1 Z-scan technique The nonlinear-optical (NLO) properties of the PbS and PbS/CdS QDs dispersed in hexane were investigated by femtosecond (fs) Z-scan at wavelength of 780 nm, ~300 fs laser pulses at a 1 kHz repetition rate as shown in Figure 2.3 . The laser pulses were generated by a mode-locked Ti:Sapphire laser (Quantronix, IMRA), which seeded a Ti:Sapphire regenerative amplifier (Quantronix, Titan). The laser pulses were focused onto a 1 mm-thick quartz cuvette (or ~10-micron-thick film) that contained the nanoparticles, with a minimum beam waist of 30 μm. The linear transmittance of all the solutions (or films) was adjusted to 70% (or 80%) at 780 nm. The incident and transmitted laser powers were monitored as the cuvette (or film coated glass slide) was moved along the propagation direction of the laser pulses. An aperture was introduced before the detector for the collection of closed-aperture Z-scan curves. One more detector was introduced for the collection of nonlinear scattering at angle of 10° to the propagation direction as shown in Figure 2.4. 69 Chapter 2 Figure 2.3 Z-Scan setup: L1, L2 - Lenses, BS - Beam Splitter, S - Sample, PD – Photo detector. Figure 2.4 Nonlinear scattering setup: L - Lens, S - Sample, PD - Photodiode. θ angle at which scattered light was collected. 70 Chapter 2 2.4.2 Transient absorption studies Femtosecond pump-probe spectroscopy enables us to follow in real time vibrational motions coupled to electronic transitions. In a pump-probe experiment, the output pulse train from an ultrafast laser is divided into two beams: the sample is excited by one pulse train (pump) and the changes it induces in the sample are probed by the second pulse train (probe), which is suitably delayed with respect to the pump. Some properties related to the probe (reflectivity, absorption, luminescence, Raman scattering) is then monitored to investigate the changes produced by the pump in the sample. In our case we have introduced SHG (second harmonic generation) crystal in between the pump path. Our experiments (discussed in Chapter 4) were performed with both 390nm and 780nm pump wavelength and 780nm probe wavelength using setup as shown in Figure 2.5. Figure 2.5 Pump-probe setup: L- Lens, S - Sample, PD - Photodiode. M - mirror, BS - beam splitter, SHG - second harmonic generator. 71 Chapter 2 Transient absorption data were also collected by taking white light (super continuum generated by sapphire rod) as the probe beam. Transient absorption measurement were performed with a regeneratively amplified, mode-locked femtosecond Ti:sapphire laser system. Briefly, pulses of 50fs duration with 3nJ/pulse energy at a repetition rate of 82 MHz were generated and amplified in a Ti:sapphire regenerative amplifier using chirped-pulse amplification. The final output pulses obtained were typically 100 fs with energy of 0.5mJ/pulse, center at 800 nm at 1 kHz and were divided into pump beam and probe beam which is much weak than pump beam. The pump beam was frequency doubled in a BBO crystal to generate 400 nm pulses with 200uJ/pulse, which was used to excite the sample contained in a quartz cell. The probe beam was used to generate white light through a sapphire plate, from which a single wavelength can be selected using a monochrometer and used as the probe. The probe beam was split again into a signal and reference, which were detected by silicon photodiodes. Pulse-to pulse fluctuation of the laser beam was eliminated by dividing the signal by the reference for each laser pulse. The optical delay between pump and probe pulses was controlled by an optical delay line based on a translation stage. The pump beam was focused with a 30 cm focal length lens and overlapped with probe beam which was focused with a 10 cm focal length. The spot sizes of pump and probe beams were 300-400 um and 100 um, respectively. Behind the sample, the pump beam was blocked while the probe beam was collected by a photodiode connecting to a lock-in amplifier. The pump power was attenuated by neutral filters. 72 Chapter 2 2.5 CO oxidation catalytic studies The CO oxidation of MnOx nanoparticles was investigated in Chapter 6. All activity measurements were performed in a micro-plug-flow reactor (Figure 2.6) under atmospheric pressure with a catalyst bed of 50 mg and an on-line gas chromatography (Shimadzu GC-14B) analyzer. Before each experiment, the dried nanocatalyst was pre-treated by heating in argon gas at various temperatures as shown in Table 2.3 for 1 h at an Ar flow rate of 20 ml/minute. The temperaturescreening experiment was performed in the range of 35-250°C, with the reaction gas mixture (5% CO, 5% O2, and 90% He) flow at 30 ml/min. The product CO2 was detected by an online GC. Conversions were calculated using peak areas of product gas CO2 as follows:2.08 x ACO2 (2.08 x ACO2) + (2.8 x ACO) ACO2 - area under CO2 peak ACO - area under CO peak Conversion (%) = Figure 2.6 Schematic diagram of the set up for catalytic activity measurement. 73 Chapter 2 Table 2.3 Pre-treatment temperatures of nanocatalysts. Temperature / °C 200 / 300 200 200 50 250 / 300 MnOx MnO α-MnO2 ρ-MnO2 Mn2O3 Mn3O4 Differential flow measurements were performed to obtain the intrinsic reaction rates and apparent activation energies. The procedure is the same as above but the conversion was kept below 20% by diluting the MnOx nanocatalyst with calcined Al2O3 (Aldrich, 99+%) in the ratio of 1:10. A total of 50 mg of catalyst bed was maintained. The temperature range for this experiment was 150-300°C, depending on the sample. Absolute reaction rate of CO conversion was calculated as follows:- rCO = c&CO,in ⋅ X CO ⋅ V&gas mMe [mol· s-1 gMe-1] mMe - mass of MnOx in the reactor bed V& - total molar flow rate X CO - conversion of CO based on CO2 formation c&CO - concentration of CO in gas mixture In order to find out the total surface area of the nano-catalysts, the Brunauer, Emmett and Teller (BET) analysis was performed. The total BET surface area was measured using Quantachrome Autosorb 6B analyzer. A volumetric method was adopted to make multipoint measurements of the equilibrium volume of N2 gas adsorbed at different pressures. 74 Chapter 2 2.6 References (1) Cademartiri, L.; Montanari, E.; Calestani, G.; Migliori, A.; Guagliardi, A.; Ozin, G. A. J. Am. Chem. Soc 2006, 128, 10337. (2) Pietryga, J. M.; Werder, D. J.; Williams, D. J.; Casson, J. L.; Schaller, R. D.; Klimov, V. I.; Hollingsworth, J. A. J. Am. Chem. Soc 2008, 130, 4879. (3) Liu, C. M. L., P.Y.; Liang, E.P.; Sow; C.H.; Chin, W.S. Submitted 2009. (4) Yin, M.; O'Brien, S. J. Am. Chem. Soc. 2003, 125, 10180. (5) Jana, N. R.; Chen, Y. F.; Peng, X. G. Chem. Mat. 2004, 16, 3931. (6) Chen, Y.; Liu, C.; Li, F.; Cheng, H. M. J. Alloys Compd. 2005, 397, 282. (7) Wang, X.; Li, Y. D. J. Am. Chem. Soc. 2002, 124, 2880. (8) Ahmad, T.; Ramanujachary, K. V.; Lofland, S. E.; Ganguli, A. K. J. Mater. Chem. 2004, 14, 3406. 75 Chapter 3 Chapter 3 Preparation of CdS and PbS polymer composite films and the study of their luminescence and nonlinear optical properties Hybrid nanocomposites with interesting optical and nonlinear properties can be obtained by the incorporation of inorganic quantum dots (QDs) into different polymer matrices.1,2 Two important aspects must be considered during fabrication in order to achieve optically active nanocomposites: firstly, the polymer or host matrix must retain its transparency within the wavelength region of interest; secondly, the quantum yield of the inorganic QDs should not be significantly quenched. Among the many QDs/polymer systems studied, PbS has been incorporated into polymers by several different methods. The in situ synthesis of the nanocrystals at either room temperature or at elevated temperatures in the presence of polymers has been carried out by several researchers.3,4 A slight variation of these methods is the simultaneous polymerization of polymer such as acrylamide and the thermal decomposition of a lead–dithiooxamide complex to form in situ the PbS/polyacrylamide nanocomposite.5 On the other hand, highly luminescence PbS nanocrystals synthesized can be dispersed in different polymers forming PbS/polymer films with high PL quantum yield in NIR region by properly tuning the polymer and concentration.2 In this Chapter, we prepared PbS/polystyrene (PS) and CdS/polymethylmethacrylate (PMMA) composite by dispersing preformed 76 Chapter 3 nanocrystals in a suitable monomer mixture followed by polymerization to form thin films. The advantage of this latter method over the in situ method is a better control of the size of PbS or CdS nanocrystals, and consequently their bright PL or large NLO properties, as shown in this Chapter. In the past decade, significant research has been performed on the study of NLO properties such as nonlinear absorption and nonlinear refraction of different QDs so as to develop better optical limiting and faster optical switching devices.6-11 For example, large optical limiting can be observed in CdS nanoparticles dispersed in dimethylformamide (DMF) due to strong two-photon absorption and nonlinear scattering (NLS), 12 while efficient optical limiting can also be observed in PbS QDs stabilized in a polyvinyl alcohol glue due to free carrier absorption (FCA). 3,6 Moreover, some research studies have also been carried out on the NLO properties of polymers, glasses and zeolites doped with PbS.4,13 Large excitedstate absorption and self-defocusing have been reported for PbS nanocrystals in zeolites at 532nm excitation.13 Most of these reported studies, however, were conducted on PbS QDs in different types of thin films while little was known about the PbS nanocrystals synthesized and dispersed in organic solvents. Furthermore, most of the above-mentioned reports aimed at understanding the two-photon absorption process, but there were few reports on the effects of excess free carriers induced by one-photon absorption. FCA was found to be enhanced by orders of magnitude in smaller 77 Chapter 3 nanocrystals compared with its bulk counterparts.14 Only a few reports are available on the theoretical explanation of size-dependent FCA of quantum wells,15 but neither theoretical nor experimental report is found on the size dependent FCA of semiconductor QDs. Moreover, the sizes of the PbS studied vary greatly from different investigators with few attempts to study the size effect on their NLO responses. Thus, in Section 3.3, we used the Z-scan technique to perform a systematic investigation into the size-dependent FCA, free-carrier refraction (FCR), optical Kerr nonlinearity and NLS in PbS QDs with sizes between 4.5 nm to 11 nm, either dispersed in hexane or embedded in PS polymer composite films. By using Z-scan theories, nonlinear parameters such as the FCA cross-section (σc), FCR coefficient (σr)16,17, optical-Kerr refractive index, and NLS coefficient (∝s) are unambiguously determined and their dependence on size is quantified. 3.1 The PbS/PS and CdS/PMMA composite films 3.1.1 PbS/PS composite films PbS QDs is synthesized by a previously reported method18 which involves the injection of S dissolved in oleylamine into a hot solution of PbCl2 in oleyamine at 120oC. More details of the synthesis procedure are given in Chapter 2. Typical XRD pattern of the PbS QDs synthesized is shown in Figure 3.1. All of the diffraction peaks could be indexed to the PbS face-centered cubic (fcc) phase (JCPDS 5-0592). The TEM images of three slightly different sizes of PbS QDs 78 Chapter 3 prepared are shown in Figure 3.2. This series of particles are prepared by varying the reaction time at 120OC. The average particle size obtained with reaction time of 1, 30 and 120 minutes was found to be 5.2 ±0.6, 6.1± 0.5 and 6.4± 0.6nm respectively. The size dispersity remained fairly constant even after 2hrs of reaction, suggesting that Osward ripening was not operative probably because of the large 10:1 excess of lead chloride compared to elemental sulfur used. Figure 3.1 Typical XRD pattern of a sample of PbS QDs indexed to JCPDS 050592. 79 Chapter 3 (a) (a) (b) (c) Figure 3.2 TEM images of PbS nanoparticles obtained at synthesis time of (a) 1, (b) 30 and (c) 120 minutes with averaged sizes estimated at 5.2 ± 0.6, 6.1 ± 0.5 and 6.4 ± 0.6 nm respectively. The absorption spectra of a series of PbS QDs withdrawn from the reaction mixture, washed and redispersed in hexane at 120OC are shown in Figure 3.3. The peaks red-shift as expected when the particles grow in size with the increase in 80 Chapter 3 reaction time. Artifacts of small spikes can be seen at 1400nm in the spectra due to excess ethanol in the solution. P b S 1 m in P S 1 .5 m in Arbitary Absorbance P S 2 m in P b S 1 5 m in P b S 3 0 m in P b S 6 0 m in P b S 1 2 0 m in e th a n o l 1000 1200 1400 1600 W a v e le n g th (n m ) Figure 3.3 Absorption spectra of PbS nanoparticles withdrawn and isolated at different reaction times from 1 min to 120 min & re-dispersed in hexane. (Also attached at the bottom is the spectrum of ethanol in hexane). We next attempted to prepare homogeneous films of PbS in PS. Our successful attempt utilizes the bulk polymerization of styrene pre-polymers, of which the steps are illustrated schematically in Figure 3.4. The styrene pre-polymers are first prepared by thermal initiated free radical polymerization of styrene monomers until a certain viscosity is reached. We found that viscosity is an important 81 Chapter 3 parameter for successful preparation of uniform QD/polymer films. Thus, the preprepared PbS QDs are mixed with the prepolymer and a photoinitiator, sonicated to give a homogeneous mixture, drop-cast onto a glass slide and cured under UV irradiation for 8 hours. Figure 3.4 A schematic showing the steps involved for the preparation of homogeneous PbS/PS nanocomposite. In our experiments, we discovered that addition of small amount of compatibilizer such as decanethiol can aid the homogenization of the prepolymer mixture and avoid segregation of the nanoparticles. A slightly brown tinted transparent film is obtained when decanethiol is added. SEM images of PbS/PS films prepared with and without the compatibilizer are compared in Figures 3.5 and 3.6. Large pinholes of sizes greater than 1μm can be seen in the cross-sectional surfaces of the film prepared without decanethiol. We believed these pinholes are aggregated PbS nanoparticles that fall off when we cut the films. A much smoother surface and homogeneous film are obtained when decanethiol is added prior to UVcuring. 82 Chapter 3 Figure 3.5 SEM images of PbS/PS nanocomposite film prepared without adding decanthiol compatibilizer at (a) 1600X and (b) 4300X magnifications. Figure 3.6 SEM images of PS/PbS nanocomposite film prepared with decanethiol at (a) 65000X and (b) 8000X and (c) 850X magnifications. 83 Chapter 3 Using the above method, we have successfully prepared homogeneous PbS/PS composite films and studied their NIR luminescence properties. It has been reported previously that luminescence is related to film morphology and phase segregation.2 The absorption and NIR luminescence of the composite films containing different sizes of PbS are shown in Figure 3.7(a-b), with the peak positions tabulated in Table 3.1. We can see that the NIR luminescence is Stoke shifted by up to 80 nm in the film. (a) 5.2nm 5.6nm 6.0nm 6.2nm 6.4nm Polystyrene 5.2nm 5.6nm 6.0nm 6.2nm 6.4nm Polystyrene (b) 800 1000 900 1000 1100 1200 1300 1400 1500 1600 1200 1400 1600 1800 Wavelength (nm) Wavelength (nm) Figure 3.7 Absorption (a) and PL (b) spectra of PbS/PS composite films containing different sized nanoparticles. Excitation wavelength = 532 nm Table 3.1 Absorption and emission wavelength maxima of PbS nanoparticles in solvent and in PbS/PS composite films. Photoluminescence Absorption maximum (nm) of PbS/PS film (nm) of PbS/PS film Reaction time (min) Average sizes (nm) Absorption maximum (nm) of PbS QDs 120 60 6.4 6.2 1530 1500 1490 1430 1540 1510 15 1.5 1 6.0 5.6 5.2 1450 1329 1250 1400 1308 1200 1480 1341 1280 84 Chapter 3 3.1.2 CdS/PMMA composite films CdS nanoparticles are synthesized by our reported precursor method19 which involves the thermal decomposition of cadmium thiobenzoate (CdTB) in hexadecylamine at high temperature.20 The detailed procedure is given in Chapter 2. The TEM and UV absorption spectrum of the CdS QDs prepared are shown in Figure 3.8(a) and (b) respectively. The synthesized CdS QDs are of slightly elongated shapes, with typical sizes of 4.2 ± 0.7 nm in width and 6.4 ± 1.4 nm in length. The first excitonic peak of CdS QDs is observed at 461 nm and found to be very broad probably due to its large size distribution. 2.0 1.5 (b) Absorbance (a) 1.0 0.5 0.0 300 400 500 600 700 800 900 wavelength/nm Figure 3.8 Typical (a) TEM image and (b) absorption spectrum of CdS nanoparticles prepared at 120oC. The CdS QDs can be incorporated into a PMMA film via the thermal polymerization of the MMA monomer mixture. Unlike the PS films discussed above, the formation of homogeneous CdS QDs in PMMA is relatively easy (1 hour) without the need of prepolymer and compatibilizer. We have found that 85 Chapter 3 exposure of these CdS/PMMA films to UV irradiation results in an enhancement of emission as shown in Figure 3.9. After ~8 hours of UV radiation, an enhancement factor up to 21 times can be obtained for the bandgap emission. This is much higher than the corresponding enhancement for the defect emission (Table 3.2). UV-induced crosslinking and photochemical reactions are the two most possible mechanisms occurring in the system.21 Typically, UV-induced crosslinking of the individual PMMA polymer chains can result in a more efficient polymer wrapping, thus providing a much better passivation layer around the nanoparticle surfaces. The CdS surfaces may also undergo photochemical reactions with surrounding oxygen or water to form CdSO4 or Cd(OH)2 species. These surface reaction products can passivate the surface traps on the particles and led to an increase in quantum efficiency. It is also possible that oxidation leads to a decrease in the CdS particle size, thus resulting in a blue shift of the PL Intensity(Arbitary units) peak maximum as shown in Table 3.2. 400 CdS/PMMA after 8h uv irradiation "as prepared CdS/PMMA" 500 600 wavelength/nm 700 Figure 3.9 PL spectra of CdS/PMMA composite films before and after uv irradiation with excitation at 375nm. (The two vertical lines are visual guides to show the blue shifting of the luminescence upon uv irradiation). 86 Chapter 3 Table 3.2 CdS/PMMA bandgap emission, PL shift and the luminescent enhancement with excitation at 375 nm. Sample Bandgap Bandgap Bandgap Enhancement Enhancement PL PL PL peak factor of factor of ( Before ( After blue bandgap defect UV UV shift emission emission Irradiation Irradiation /nm intensity intensity ) /nm ) /nm CdS/PMMA 492 484 8 21 12 3.2 Layered Nanocomposite films Next, we prepared layered PbS/PS composite films comprising different sizes of PbS QDs by casting the layers sequentially from the respective solutions. As shown schematically in Figure 3.10, the first layer (cast right onto the glass substrate) comprising smaller bandgap 6 nm PbS/PS, while the second (top) layer comprising larger bandgap 5.2 nm PbS/PS. Our objective is to compare the emission properties of the resultant composites when irradiated from the top and bottom faces. The SEM cross-sectional side view of such a layered structure is shown in Figure 3.11. The magnified view showed that only small holes smaller than 1 μm are observed at some regions thus showing that our films are well formed with little segregation of the PbS. 87 Chapter 3 Glass Figure 3.10 A schematic diagram showing the formation of layered PbS/PS nanocomposite. The bottom layer comprises 6 nm PbS/PS, while the top layer comprising 5.2 nm PbS/PS. The centre PS was added as a barrier layer. (a) (b) Figure 3.11 Cross-sectional side-view SEM images of the layered PbS/PS nanocomposite at (a) 50X and (b) 16000X magnifications. The UV absorption and PL emission spectra of the PbS/PS layered composite films are presented in Figure 3.12. The PL emission spectra are measured by excitation at 532 nm through the top and the bottom orientations respectively. 88 Chapter 3 Thus, as shown in Figure 3.12(b), illumination through the top surface excites preferentially the larger-gap QDs in the top layer and hence the associated luminescence at 1250 nm increases in intensity relative to that at 1600 nm. On the other hand, illumination from the bottom (substrate side) excites preferentially the smaller-gap QDs, hence the intensity of the two peaks reverses. As a comparison, we also investigated another composite film whereby the two sized QDs were mixed in a single PS layer. As expected, the two peaks give similar intensity in this case. In addition, the peak maximum at ~1250 nm (i.e. smaller QDs) shifted slightly towards longer wavelength. A similar red shift of the PL peak was observed previously in a CdSe QD solid having a large size distribution and the shift has been attributed to energy transfer from the smaller particles to the larger particles.22 When the two sized PbS QDs were mixed in a single PS layer, the probability of these QDs being very close to each other is increased and allows the energy transfer to occur. 0.50 0.45 1.0 0.40 Normalized Intensity Intensity (a.u.) Layered Com posite (top surface) M ixture Com posite Layered Com posite (bottom surface) PbS/PS/PbS (Top) PbS/PS/PbS (Bottom) Polystyrene 0.35 0.30 (a) 0.25 0.20 1000 0.8 0.6 0.4 (b) 0.2 1100 1200 1300 1400 Wavelength (nm) 1500 1600 0.0 800 1000 1200 1400 1600 W avelength/nm Figure 3.12 Typical (a) absorption and (b) PL spectra of layered PbS/PS nanocomposite irradiated from top and bottom (refer to Fig. 3.10 for orientation) surface. Also included in (b) is the PL spectrum of a film comprising of mixed QDs of the two sizes. Excitation wavelength = 532 nm. 89 Chapter 3 Composite films comprising layers of CdS/PMMA and PbS/PS can also be prepared by the drop-casting method. Briefly, the PbS-PS prepolymer mixture is first drop casted onto the glass slide and polymerized under uv light. Subsequently, the next layer of CdS/PMMA film is drop casted and polymerized forming the layered composites as seen in Figure 3.13(a). We have found that the thickness of the PbS/PS layer will affect the overall PL profile of the composite measured from the bottom face at excitation wavelength of 375 nm. As shown in Figure 3.13(c) and (d), when the thickness of PbS/PS film (measured using a profilometer) increased from 10μm to 150μm, the CdS bandgap and defect emissions attenuated almost completely. Two possible reasons can lead to the luminescence quenching observed above. Firstly, as the excitation wavelength used to excite the CdS film lies within the wavelength region that PbS/PS film absorbs, the excitation radiation will be strongly absorbed by the thicker PbS/PS layer. This thus reduces the intensity of light reaching the CdS/PMMA layer and leads to attenuation of the CdS luminescence. Secondly, the CdS luminescence is likely be reabsorbed when passing through the PbS layer leading to a quenching along the thickness of the PbS/PS layer. 90 Chapter 3 top surface bottom surface Excitation (a) Intensity(a.u) (b) CdS/PMMA(top) PbS/PS (bot) Glass 900 1000 1100 1200 1300 1400 1500 1600 W a ve le ng th /n m 600000 Excitation top surface before uv irradiation bottom surface before uv irradiation top surface after uv irradiation bottom surface after uv irradiation 500000 Intensity(a.u) 400000 (c) 300000 200000 100000 0 400 500 600 700 W avelength/nm 100000 top surface before uv irradiation bottom surface before uv irradiation top surface after uv irradiation bottom surface after uv irradiation 90000 80000 Intensity 70000 60000 (d) 50000 40000 30000 20000 10000 0 400 500 600 700 W avelength/nm Figure 3.13(a) Schematic diagram of layered PbS/PS-CdS/PMMA nanocomposites. (b) NIR PL spectra of layered PbS/PS-CdS/PMMA nanocomposite excited at 532nm, (c-d) PL spectra of layered PbS/PSCdS/PMMA nanocomposite excited at 375 nm before and after uv irradiation with PbS/PS of 10 μm and 150 μm thickness respectively. 91 Chapter 3 When the layered PbS/PS-CdS/PMMA nanocomposite is excited from the top orientation, however, the CdS luminescence is found to be similar regardless of the thickness of the PbS/PS layer (Figure 3.13d). The NIR luminescence at 1280 nm is also found to be unaffected by the CdS layer (Figure 3.13b). In this case, the excitation wavelength (532nm) is too high for the CdS absorbance regime. When we compared the CdS bandgap and defect emissions in Figure 3.13(c-d), we can see that uv irradiation has led to a greater enhancement of the defect emissions relative to that of the bandgap emission. The exact reason for this defect peak is unknown at this stage, but it may suggest that our layered films are not homogeneous or clustering of QDs may have occurred to some extent. While we were successful in preparing QDs/polymer films which showed strong luminescence by the drop cast method as discussed above, we were not completely successful when trying to prepare thinner films. Thus, an alternative method was developed by using an imprinting thermal polymerization. Detailed procedures and optimized parameters of this method are given in Chapter 2. Using this method, we can routinely prepare composite films as thin as 6μm with the QDs well dispersed into the polymer matrix, unlike the segregation problems faced using the drop cast method. We investigated the PL properties of such a composite CdS/PS-PbS/PS film using excitation wavelength at 375 nm. As shown in Figure 3.14, a single PL peak is observed at 478 nm due to the bandgap emission of CdS from the CdS/PS layer. 92 Chapter 3 Similarly, the CdS PL was quenched when excited from the bottom layer due to absorption by the PbS layer. 100000 90000 80000 Relative intensity 70000 (a) 60000 50000 40000 30000 20000 (b) 10000 0 400 500 600 700 Wavelength/nm Figure 3.14 PL of multilayer CdS/PS- PbS/PS nanocomposite films prepared by imprinting method with excitation at 375 nm from the (a) top surface and (b) bottom surface. 3.3 Nonlinear optical properties of PbS in hexane and in polymer films The imprinting method developed above opens up the possibility to form highly luminescent films of μm thickness suitable for nonlinear optical (NLO) measurements. Since our composites were prepared from pre-formed QDs, we can thus perform comparative NLO studies of monodispersed PbS QDs dispersed in a polymer thin film and in a suitable solvent. We compared the size dependence NLO behaviour, and attempted to obtain some useful quantitative NLO parameters. 93 Chapter 3 TEM images in Figure 3.15(a-d) shows relatively monodispersed PbS QDs having an average size of 4.6 ± 0.5, 5.3 ± 0.5, 6.0 ± 0.6 and 11 ± 2 nm (a) (b) 40 20 0 4 5 Sizes (nm) Frequency Frequency respectively that are used for the study of NLO properties below. 40 20 0 6 4 5 6 Sizes (nm) 100nm 60 (d) 40 20 0 5 6 40 20 0 6 8 10 12 14 16 18 Sizes (nm) 7 Sizes (nm) 50nm Frequency Frequency (c) 20nm 100nm Figure 3.15 TEM micrographs of PbS QDs with average sizes of: (a) 4.6 ± 0.5 nm, (b) 5.3 ± 0.5 nm, (c) 6.0 ± 0.6 nm, and (d) 11.0 ± 2.0 nm. Inserts give histograms of the respective size distributions. In Figure 3.16, the UV-VIS-NIR absorption spectra of these four samples are 100nm measured in solution. The first and second excitonic peaks for the 4.6 nm, 5.3 nm and 6 nm PbS QDs are well resolved showing that good size distributions have been achieved. The first excitonic absorption peak blue-shifts as the particle size 94 Chapter 3 decreases due to quantum confinement effect as summarized in Table 3.3. The lack of a distinct peak for the 11 nm QDs may be attributed to its broader size distribution and weak quantum confinement. This latter sample is also more difficult to disperse homogeneously for luminescence measurement. The NIR bandgap emissions of the other three samples measured in hexane are displayed in Figure 3.17(a), with the peaks red-shifting with sizes as expected. The small Stoke shift tabulated in Table 3.3 suggests that the luminescence is due to bandgap excitonic recombination and can be attributed to efficient passivation of the nanocrystals surface by oleylamine ligands, similar to that observed previously.18 4.6 nm 5.3 nm 6 nm 11 nm Absorbance 0.8 0.4 0.0 600 800 1000 1200 1400 1600 1800 Wavelength (nm) Figure 3.16 Typical UV-VIS-NIR absorption spectra of the PbS QDs in solution. The spectra are labeled with the corresponding sample sizes determined from Figure 3.15. 95 Chapter 3 Table 3.3 A summary of the average sizes of PbS QDs estimated from TEM, the respective first excitonic position and bandgap luminescence measured in hexane. A B C D Intensity(normalized) 1.0 0.8 0.6 Average sizes from TEM (nm) 4.6 ± 0.5 5.3 ± 0.5 6.0 ± 0.6 11.0 ± 2.0 First excitonic position (nm) 1220 1290 1473 N.A 1.6 (a) 4.6nm 5.3nm 6.0nm 0.4 0.2 0.0 900 1000 1100 1200 1300 1400 1500 Wavelength (nm) Intensity (normalized) Sample 1.4 1.2 1.0 (b) Bandgap Luminescence (nm) 1269 1323 1511 N.A 4.6nm PbS in polystyrene 5.3nm PbS in polystyrene 6.0nm PbS in polystyrene pure polystyrene 0.8 0.6 0.4 0.2 0.0 1000 1100 1200 1300 1400 1500 Wavelength (nm) Figure 3.17 Typical NIR luminescent spectra of the PbS QDs in (a) hexane, and (b) polystyrene composite film. Excitation wavelength = 532 nm. By mixing the PbS QDs in a thermally cross-linkable recipe, followed by imprinting cum polymerization, uniform and transparent PbS/PS composite films of a few μm thicknesses are prepared on glass substrates. The normalized NIR spectra of these homogeneous PbS/PS films are shown in Figure 3.17(b). Slight broadening of the luminescence is observed as expected and the strong NIR signals confirm that the fast polymerization method has successfully retained the original luminescence properties of the PbS QDs. The abrupt drop at wavelength higher than 1520 nm is due to the spectral limit of the detector. 96 Chapter 3 Optical limiting property in inorganic and hybrid nanomaterials typically arises from a collective response of some of the following nonlinear mechanisms: nonlinear refraction, NLS, FCA , multiphoton absorption and reverse saturable absorption (RSA).23 Thus, the NLO optical properties of our PbS nanoparticles were characterized with Z-scan method. Open aperture Z-scan curves of the QDs in hexane and PS films respectively is shown in the Figure 3.18(a) and (b). It is noted that the absorptive nonlinearities of these QDs show different behavior in different matrices. All the sizes of PbS QDs display RSA at all the intensities in solutions. In the PS films, on the other hand, all the samples except the smallest size QDs (4.6 nm, showing saturable absorption at lower intensities and RSA at high intensities) show RSA behavior. As stated before, the RSA behaviour could result from FCA and/or NLS mechanisms. Thus, these size-dependent Z-scan curves show interesting behaviours - at lower intensities, there is a larger nonlinear absorption signal (or the change in the transmittance at the focus) for the smaller size QDs than the bigger size QDs; while at higher intensities, the reversed behavior is observed. This implies that more than one nonlinear phenomenon is manifesting themselves in these samples. 97 1.2 1.0 0.8 0.6 1.0 2 2 I00 = 85 GW/cm 4.6 nm 5.3 nm 6.0 nm 11.0nm The. fit 2 0.8 0.6 0.4 1.0 1.2 I00 = 22 GW/cm I00 = 126 GW/cm 0.8 0.6 1.0 2 I00 = 141 GW/cm 0.8 0.6 (a) 0.4 -3 -2 -1 0 1 2 3 Normalized Transmittance Normalized Transmittance Chapter 3 2 I00 = 50 GW/cm 1.0 0.8 2 I00 = 100 GW/cm 1.0 0.8 0.6 2 I00 = 125 GW/cm 1.0 0.8 0.6 0.4 -2 Z (cm) (b) -1 0 4.6 nm 5.3 nm 6.0 nm the. fit 1 2 Z (cm) Figure 3.18 Open-aperture Z-scans at varying laser intensities for different sized PbS QDs in (a) hexane and (b) PS film. Solid lines represent the theoretical fits. In order to investigate this additional nonlinear effect, we performed intensitydependent NLS experiments on these QDs. Figure 3.19(c) confirms that, at higher intensities, the bigger size QDs show more NLS than the smaller ones, and NLS is the dominant phenomena for bigger QDs at higher intensities. 98 Scattering Intensity (arb. units) Chapter 3 Angle = 10 1.0 o 4.6 n m 5.3 n m 6.0 n m 11.0n m fit th ird o rd er p o ly. 0.5 0.0 10 100 In ten sity (G W /cm 2 ) Figure 3.19 Nonlinear scattering measurements of different sized QD collected at 10o. Combined losses of NLS and FCA lead to lower nonlinear transmission at the focus for bigger size QDs at higher intensities. Since our pumping photon energy is much greater than the band gap, FCA is speculated to be the dominant process. However at higher intensities, NLS becomes significant for bigger QDs. NLS contribution is negligible in case of film samples, nevertheless, below the damage thresholds (~125 GW/cm2). As such, in the Z-scan theory, both FCA16 and NLS12 should be taken into consideration as follows: 99 Chapter 3 dI = − α 0 I − σ c N1 I − α s I dz ∂N 1 α 0 I = hω ∂t where eqn(3.1) eqn(3.2) α s = g s [Δn]2 Δ n = Δnl + Δn nl = Δnl + Δnth I 1+ τ th tp Here α0 is the linear absorption coefficient, N1 is the intensity-dependent carrier density, σc is the FCA cross-section, αS is the effective scattering coefficient, I is the peak intensity, gs is a parameter which is independent of intensities but depends only on the size, shape, concentration of particles and wavelength of light, Δnl is the difference in the linear refractive index of PbS particle and hexane, and Δnnl is the difference in the nonlinear refractive indices of PbS particles and hexane. As Δnnl is a function of the intensity inside the medium, it can be derived further in the following manner. The relaxation of light-induced changes in the refractive indices of the components can be taken into account by introducing a Debye-type equation for femtosecond pulses as Δnnl= ΔnthI/(1+(τth/tp)). Here, Δnth is the difference between the nonlinear thermal refractive index of PbS particles and hexane, τth (~ms) is the relaxation time of the thermal nonlinearity and tp (~1 ms) is the time period between two laser pulses. We can integrate both equations (3.1) and (3.2) numerically over time, length, and along the radial direction by using Runge-Kutta fourth order method. The best fits (solid lines in Figure 3.18) between the numerical solutions and the Z-scan data generates the nonlinear parameters as tabulated in Table 3.4. 100 Chapter 3 Table 3.4: Fitted FCA cross-section (σc), nonlinear refractive index (n2), FCR cross-section (σr), and scattering coefficient (αS) of the PbS QDs in solution. (All values listed are within 10% error). Size, R (nm) 4.5 nm 5.3 nm 6.0 nm 11 nm σc × 1019 (cm2) 61.3 30.2 23.4 8.2 n2 × 106 (cm2/GW) -4.2 -3.0 -2.0 -0.5 σr × 1022 (cm3) -1.5 -1.4 -0.4 -0.3 αs(cm-1) 2.1 2.4 3.9 7.1 For the smallest QDs in the case of the composite films, we noted that saturable absorption (SA) occurs at lower intensity, while RSA dominates at higher intensity. In order to include this SA phenomenon to the theoretical calculation, we consider the saturation equation. This saturation is mostly due to one-photon bleaching and we may modify the one-photon absorption coefficient to α0s24,25. α 0s = α0 eqn(3.3) 1+ I / Is where Is is the one-photon saturation intensity. Saturation intensities of bigger particles are not measurable due to higher saturation intensities of bigger particles. All the coefficients used in the fittings are tabulated in Table 3.5. Table 3.5 Fitted FCA cross-section (σc), nonlinear refractive index (n2), FCR cross-section (σr) of the PbS QDs in PS films (All values listed are within 10% error). Size, R (nm) 4.5 nm 5.3 nm 6.0 nm σc × 1019 (cm2) 204 159 110 n2 × 106 (cm2/GW) -5.4 -2.5 -1.8 σr × 1022 (cm3) -1.4 -0.7 -0.2 101 Chapter 3 Figure 3.20 display typical closed aperture Z-scans curves of PbS QDs. All the samples, in film and solution, show negative nonlinear refraction, indicating selfdefocusing effect. It is well known that thermal nonlinearities also contribute to negative nonlinear refraction, and high single-photon absorption cross-sections may be responsible for these thermal nonlinearities. Thus, to minimize thermal effects, all our closed-aperture Z-scan curves were collected at lower intensities. Figure 3.20 Closed-aperture Z-scan curves of different sized PbS QDs in (a) hexane and (b) PS film with theoretical fits. Free-carriers induced by one-photon absorption also contribute significantly to the phase changes of the wave in the sample. Including the free-carrier refraction (FCR) contribution, the nonlinear phase equation is given by: 16 102 Chapter 3 dΔϕ 0 2π (n2 I + σ r N1 ) = KK eqn(3.4) λ dz 2πLeff (n2 I + σ r N1 ) KK eqn(3.5) Δϕ 0 = λ TCA = 1 − 4Δϕ 0 ( z / z 0 ) KK eqn(3.6) [1 + ( z / z 0 ) 2 ][9 + ( z / z 0 ) 2 ] Here, n2 is the nonlinear refractive index (or optical Kerr nonlinearity), σr is the FCR coefficient, Δφ0 is the phase change, z is the sample position, z0 is the Rayleigh range, TCA is the closed aperture transmittance and Leff is the effective sample length.8 It is known that FCR is a self-defocusing phenomena added to the overall nonlinear refraction.16 Both the n2 and σr values of the PbS QDs are obtained from the best simulations to the experimental data and listed in Table 3.4 and 3.5. PbS QDs with the smaller sizes show higher FCR and optical Kerr nonlinearity. It is observed that the FCA cross-sections of the films are greater than the ones in the solution. The FCA cross-sections of smaller QDs in PS films show at least three-times enhancement compared with the particles in hexane as shown in Figure 3.21. . The FCA cross-sections of PbS QDs in both hexane and PS films are linearly dependent on 1/R2, where R is the size of nanoparticles. Our results are consistent with size-dependent theoretical simulations on quantum wells and wires.15 103 2 25 18 (FCA cross-section)x10 (cm ) Chapter 3 20 in Hexane in PS film 15 10 5 0 0.01 0.02 0.03 2 0.04 0.05 -2 1/R (nm ) Figure 3.21 Size dependent FCA cross-section of PbS QDs in hexane and in PS film and its linear fits with 15% error bars. Surface structure and ligand have been previously suggested to have strong influence on the nonlinear-optical properties of PbS nanoparticles. Oleic acid capped PbS nanocrystals synthesized by the popular Hines and Scholes method showed previously to possess only saturable absorption at high intensity. . This was suggested to be due to quenching at the interface between the adsorbed oleic acid ligand and the nanocrystal surface which had been shown to have an excess of sulfur at PbS nanocrystal surfaces.4,26 The synthesis method we follow for PbS QDs has been previously shown to be Pb rich and passivated with Cl- and olelyamine.18 More theoretical and experimental work need to be done to establish if the difference in surface structure or ligand can explain our successful attempt in obtaining RSA from the PbS QDs in the solution phase. A previous study of a series of sizes of PbS nanocrystals grown in polymer solutions did not 104 Chapter 3 correlate their nonlinear absorption to FCA but instead to two-photon absorption.4 These author suggested that the absence of size dependence was due to their similar surface chemistry. Unlike their observation with nanosecond laser pulses, we believe that, by using a femtosecond laser, the size-dependent nonlinear absorption of PbS QDs cannot be smeared by thermal effects. 3.4 Summary In summary, PbS QDs of varying sizes were synthesized via chemical method and by varying reaction temperature and time. Single layer drop-casted PbS/PS composite films were successfully prepared using decanethiol as a compatibilizer. Using PbS QDs of different sizes, layered nanocomposites were successfully prepared and revealed interesting PL properties dependent on the orientation of excitation. The luminescence of drop casted single layer CdS/PMMA was improved by UV illumination probably through improved passivation of surface reaction products. Multilayer CdS/PMMA-PbS/PS was similarly prepared by drop casted method and the PL was found to depend on the thickness of the PbS layer and the orientation of the excitation source. Micrometer thickness PbS/PS thin films were successfully prepared by a novel imprinting-thermal cross-linking method. The PbS particle sizes and NIR luminescence properties were preserved in the final films. The NLO responses of these films were studied with femtosecond Z-scan technique and compared to the 105 Chapter 3 solution phase properties. The size-dependent free-carrier absorption, free-carrier refraction, and optical Kerr nonlinearity were determined at laser excitation wavelength of 780 nm. In addition, nonlinear scattering was also found to play an important role in the solution study at high excitations. 3.5 References (1) Du, H.; Xu, G. Q.; Chin, W. S.; Huang, L.; Ji, W. Chem. Mater. 2002, 14, 4473. (2) Chang, T. W. F.; Maria, A.; Cyr, P. W.; Sukhovatkin, V.; Levina, L.; Sargent, E. H. Synth. Met. 2005, 148, 257. (3) Kurian, P. A.; Vijayan, C.; Sandeep, C. S. S.; Philip, R.; Sathiyamoorthy, K. Nanotechnology 2007, 18, 075708. (4) Asunskis, D. J.; Bolotin, I. L.; Hanley, L. J. Phys. Chem. C 2008, 112, 9555. (5) Nair, P. S.; Radhakrishnan, T.; Revaprasadu, N.; Kolawole, G. A.; Luyt, A. S.; Djokovic, V. Appl. Phys, A: Mater. Sci. Process. 2005, 81, 835. (6) Venkatram, N.; Kumar, R. S. S.; Rao, D. N. J . Appl. Phys. 2006, 100, 074309. (7) He, J.; Ji, W.; Ma, G. H.; Tang, S. H.; Elim, H. I.; Sun, W. X.; Zhang, Z. H.; Chin, W. S. J . Appl. Phys. 2004, 95, 6381. (8) Venkatram, N.; Sathyavathi, R.; Rao, D. N. Opt. Express 2007, 15, 12258. (9) Vasa, P.; Ayyub, P.; Singh, B. P. Appl. Phys. Lett.2005, 87, 063104. (10) Jia, W. L.; Douglas, E. P.; Guo, F. G.; Sun, W. F. Appl. Phys. Lett. 2004, 85, 6326. 106 Chapter 3 (11) Nikesh, V. V.; Dharmadhikari, A.; Ono, H.; Nozaki, S.; Kumar, G. R.; Mahamuni, S. Appl. Phys. Lett. 2004, 84, 4602. (12) Venkatram, N.; Rao, D. N.; Akundi, M. A. Opt. Express 2005, 13, 867. (13) Kim, H. S.; Lee, M. H.; Jeong, N. C.; Lee, S. M.; Rhee, B. K.; Yoon, K. B. J. Am. Chem. Soc. 2006, 128, 15070. (14) Kekatpure, R. D.; Brongersma, M. L. Nano Letters 2008, 8, 3787. (15) Kubakaddi, S. S.; Mulimani, B. G. J. Phys. C: Solid State Phys. 1985, 18, 6647. (16) Li, H. P.; Kam, C. H.; Lam, Y. L.; Ji, W. Opt. Commun. 2001, 190, 351. (17) Bindra, K. S.; Singh, C. P.; Oak, S. M. Opt. Commun. 2007, 271, 248. (18) Cademartiri, L.; Bertolotti, J.; Sapienza, R.; Wiersma, D. S.; von Freymann, G.; Ozin, G. A. J. Phys. Chem. B 2006, 110, 671. (19) Lim, W. P.; Zhang, Z.; Low, H. Y.; Chin, W. S. Angew. Chem., Int. Ed.2004, 43, 5685. (20) Lim, W. P. Architectural Control of Metal Sulfide Nanocrystals and Polymer Composites, National University of Singapore, 2006, pp147- 165. (21) Bol, A. A.; Meijerink, A. J. Phys. Chem. B 2001, 105, 10203. (22) Kagan, C. R.; Murray, C. B.; Bawendi, M. G. Phys. Rev. B 1996, 54, 8633. (23) Wang, J.; Blau, W. J. J. Opt. A: Pure Appl. Opt. 2009, 11, 024001. (24) Samoc, M.; Samoc, A.; Luther-Davies, B.; Reisch, H.; Scherf, U. Opt. Lett. 1998, 23, 1295. (25) Elim, H. I.; Ji, W.; Ng, M. T.; Vittal, J. J. Appl. Phys. Lett. 2007, 90, 033106. (26) Lobo, A.; Moller, T.; Nagel, M.; Borchert, H.; Hickey, S. G.; Weller, H. J. Phys. Chem. B 2005, 109, 17422. 107 Chapter 4 Chapter 4 Synthesis of PbS/CdS core-shell QDs and their nonlinear optical properties Nonlinear optical (NLO) properties have been shown to be enhanced in several heterostructured nanocrystals (NCs) such as core-shell QDs as compared to their core semiconductor counterparts.1-8 For example, core-shell CdS/Ag2S and ternary CdxAg1−xS have enhanced nonlinear absorption in comparison with the CdS nanoparticles due to contribution from free carrier absorption (FCA) and nonlinear scattering (NLS).1,2 Wang et al. also showed that the chitosan CdSe/ZnS core-shell QDs possess a higher optical nonlinearity of about 200% (third-order susceptibility) as compared to chitosan CdSe QDs alone, due to the larger nonlinear refraction from the electronic Kerr effect.4 Metallic shell was also found to have a significant effect on the NLO properties of the core QDs. Chitosan CdS/Ag core-shell QDs were observed to have 400 and 200 times enhancement of their effective nonlinear absorption coefficient and nonlinear refraction index over that of Chitosan CdS QDs. 3 The enhancement in the CdS/Ag core-shell QDs was attributed to the large local field and the strong surface plasmon resonant absorption of the Ag shell. Lastly, the nonlinear refractive index and two photon absorption (2PA) coefficient of Mn-doped ZnSe QDs with the same core size increased and reached a saturation value as the overcoating ZnSe layer increases. The size dependence of the optical nonlinearity was found to fit well with a model incorporating the influence of a dielectric local field effect which changes with different shell thicknesses.5 108 Chapter 4 Carrier relaxation dynamics in semiconductors have been widely studied in II-VI semiconductors9-13 but slightly less so for IV-VI semiconductors, especially in the area of bi-excitons and Auger recombination (AR). Wu et al. utilized transient absorption to study the nonlinear relaxation processes in monodisperse PbS nanocrystals of sizes between 7 to 16nm.14 They found that all the sizes studied have similar decay profiles but showed excitation intensity dependence in their decay. The fast component of the decay was observed to increase faster than the slow component with the increase of the pump intensity and was attributed to exciton–exciton annihilation.10,15 This finding was in contrast to the work by Patel et al. who observed excitation intensity independence in PbS NPs with a broader size distribution.16 More recently, transient absorption studies on the carrier dynamics of PbS QDs at the first and second exciton energies were reported by Istrate et al. and fitted to a model incorporating intraband and interband relaxation processes.17 They studied the Auger recombination process which occurs when more than one electron-hole pair has been excited in a dot. By utilizing the Poisson statistics models developed for multiparticle Auger rates in CdSe and further limiting the fundamental level occupancy to 8 fold degeneracy for PbS, they were able to account for the observed saturation of its absorption bleaching.12,17,18 Synthesis of core-shell QDs of PbS/CdS have been accomplished previously in an aqueous solution.19 Recently, however, higher quality PbS/CdS QDs have been synthesized by a cationic exchange method.20,21 However, the NLO and electron 109 Chapter 4 dynamics properties of such QDs which might be important for applications like nonlinear photonics, photovolatics and photoconductivity have not yet been studied. Thus in this chapter, we first explored the synthesis and characterization of core-shell PbS/CdS QDs by adapting an existing synthesis method (details given in Chapter 2).20 As the formation of core-shell structures have been shown to enhance the NLO properties of the core QDs, we investigate the NLO properties of these PbS/CdS QDs by using both Z-scan technique and femntosecond transient absorption. We compare these results with the pure PbS QDs reported in the previous chapter. 4.1 Synthesis and characterization of core-shell PbS/CdS PbS/CdS core-shell QDs were prepared via a cationic exchange method by mixing cadmium oleate in octadecene (ODE) with a hot toluene solution of preformed PbS QDs for different duration of reaction time. We prepared three coreshell samples from three sizes of PbS QDs as depicted in Table 4.1. The TEM micrographs of the three PbS QDs and their corresponding PbS/CdS QDs are shown in Figure 4.1. From the TEM, the average diameters of the PbS QDs and their corresponding PbS/CdS QDs after the cationic exchange were found to be almost the same (within the error bar). This suggests that no Ostwald ripening has occurred. For the convenience of our discussion in the following, all the samples will be labeled with their average diameters rounded up to whole number determined from TEM. 110 Chapter 4 40 (b) 20 0 Frequency Frequency (a) 40 20 0 4.0 4.5 5.0 5.5 6.0 Sizes (nm) 20nm Sizes (nm) 20nm (d) 20 Frequency 40 Frequency (c) 4.0 4.5 5.0 5.5 6.0 6.5 40 20 0 4.5 5.0 5.5 6.0 6.5 7.0 0 Sizes (nm) 4.5 5.0 5.5 6.0 6.5 7.0 Sizes (nm) 20nm 20nm Frequency (e) 20 (f) 15 10 5 0 6 7 8 20 10 0 9 Sizes (nm) 20nm Frequency 30 25 6.0 6.5 7.0 7.5 8.0 8.5 9.0 Sizes (nm) 20nm Figure 4.1. TEM of PbS QDs and the corresponding core-shell QDs: (a-b) 5 nm PbS and 5 nm PbS/CdS, (c-d) 6nm PbS and 6 nm PbS/CdS, and (e-f) 7 nm PbS and 7 nm PbS/CdS. 111 Chapter 4 Table 4.1 Average sizes of PbS QDs and PbS/CdS core-shell QDs obtained at the optimum reaction time. Sample Label Average diameter of PbS (nm) Average diameter of PbS/CdS (nm) Reaction time (hrs) 5 nm QDs 5.1 ± 0.5 5.0 ± 0.4 20 6 nm QDs 6.0 ± 0.6 5.9 ± 0.4 2 7 nm QDs 7.3 ± 0.8 7.2 ± 0.7 1 Typical XRD patterns of the three sizes of PbS QDs and their corresponding coreshell PbS/CdS QDs synthesized are shown in Figure 4.2. All of the diffraction peaks for the PbS QDs could be indexed to the PbS face-centered cubic (fcc) phase (JCPDS 5-0592). We can see that for all the cationic exchanged core-shell QDs, the (100), (200) and (220) peaks have been significantly broadened compared to those of the pure PbS QDs. Two possible reasons might have contributed to the broadening of the XRD peaks. Firstly, since the cationic exchange led to core-shell PbS/CdS QDs with the same overall size, it suggested a sacrificial replacement of the outer layer of PbS giving a smaller PbS core size. As the diffraction peak width has been widely known to be inversely proportional to the crystalline size using the Debye-Scherrer formula, a smaller PbS size will lead to a broaden peak. Secondly, as the diffraction peaks (100), (200) and (220) peaks of cubic CdS (JCPDS 80-019) are very near to those of the PbS fcc phase, the diffraction peaks may overlap thus leading to an overall broadened peak. 112 Relative intensity Chapter 4 1 1 1 1 1 8 6 4 2 0 8 6 4 2 0 0 0 0 0 0 0 0 0 0 7 n m P b S 7 n m P b S /C d S c u b ic P b S s ta n d a r d c u b ic C d S s ta n d a r d (c ) 2 0 3 0 4 0 5 0 2 θ 1 2 0 Relative intensity 7 0 8 0 6 n m P b S 6 n m P b S /C d S c u b ic P b S s ta n d a r d c u b ic C d S s ta n d a r d (b ) 1 4 0 6 0 1 0 0 8 0 6 0 4 0 2 0 2 0 1 2 0 4 0 5 0 2 θ 6 0 7 0 8 0 5 n m P b S 5 n m P b S /C d S c u b ic P b S s ta n d a r d c u b ic C d S s ta n d a r d (a ) (220) (111) 1 4 0 Relative intensity 3 0 (200) 0 1 0 0 8 0 6 0 4 0 2 0 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 2 θ Figure 4.2. XRD patterns of (a) 5 nm PbS and 5 nm PbS/CdS; (b) 6 nm PbS and 6 nm PbS/CdS; (c) 7 nm PbS and 7 nm PbS/CdS. Standard patterns of cubic PbS and CdS (JCPDS 5-0592 and JCPDS 800019) are marked for comparison. 113 Chapter 4 We attempt to estimate the dimension of both the PbS core and the CdS shell from the atomic ratios of Pb to Cd obtained from elemental analysis (Table 4.2) using Equation 4.1: 4/3πr 3 ρ PbS MWCdS molePb VPbS ρ PbS MWCdS = = ...eqn(4.1) (VR − VPbS )( ρ CdS MWPbS ) molecd 4/3π ( R 3 - r 3 ) ρ CdS MWPbS where r is the radius of the core PbS QD (nm), ρPbS is the density of bulk PbS (kgm-3), MWCdS is the molecular weight of CdS, R = radius of the core-shell QD (nm), ρCdS is the density of bulk CdS (kgm-3), molePb/moleCd is the atomic ratio of Pb to Cd obtained from elemental analysis. In using this simplified approach, we have made the assumptions that the shell thickness is uniform throughout the nanoparticles and stoichiometric PbS and CdS are formed. These assumptions are not totally valid as there is evidence of inhomogeneous reaction as indicated by the broadening of absorption and PL peaks (vide infra).22 Nonetheless, this estimation clearly confirms the sacrificial replacement of PbS core particles and also accounts for the observed blue shifts of the absorption and PL peaks (vide infra, Table 4.3). 114 Chapter 4 Table 4.2 Elemental ratios of the different elements from ICP-OES, estimated sizes of PbS core and CdS shell from Eqn 4.1, and the first excitonic peak of the different PbS/CdS core-shell QDs. Diameter of PbS/CdS QDs determined from TEM (nm) 5.0 5.9 7.2 ICP-OES Wt% (atomic ratio) S Pb Cd 10.77 (4.62) 11.94 (2.58) 11.84 (3.13) 15.05 (1) 29.93 (1) 24.45 (1) 16.03 (1.96) 29.76 (1.84) 11.57 (1.15) Estimated Estimated Absorption PbS core CdS shell peak (nm) size (nm) size (nm) 3.5 0.75 902 4.2 0.85 1173 5.9 0.65 1530 The sacrificial replacement mechanism can be followed by isolating the core-shell particles after 1 hour of cationic exchange. In Figure 4.3(a), the NIR absorption spectrum of the 5 nm PbS core particles was compared with those of its core-shell QDs isolated after 1 and 20 hours reaction and redispersed in hexane. Similar to earlier reports,23-25 three main excitonic peaks were observed for the core particles at 1242 nm (1Se-1Sh), 900 nm (1Pe-1Ph) and 500-600 nm (1De-1Dh) respectively. For the core-shell QDs, the first excitonic (1Se-1Sh) peak blue shifted from 1242 nm to 1001 nm and 902 nm, for products isolated at 1 and 20 hours respectively. It is noted that the lower wavelength excitonic peaks have disappeared in the core-shell QDs and the peak broadens significantly with the formation of the CdS shell. There are two plausible explanations for the broadening and disappearance of the excitonic peaks: Firstly, the formation of a CdS shell has allowed the exciton in the core to be delocalized; secondly, the cationic exchange could be anisotropic and lead to the formation of polydispersed PbS core particles. In 115 Chapter 4 Figure 4.3(a), rising absorbance at wavelengths smaller than 700 nm due to the absorption of CdS can also be seen as the thickness of the CdS shell layer was increased. The main NIR luminescence of the PbS QDs was also found to be broader and shifted to shorter wavelength after the cationic displacement as shown in Figure 4.3(b). The Stoke shift was found to increase significantly from 58 nm to 165 nm after the cationic exchange as shown in Table 4.3. It is possible that a significant amount of traps are introduced at the interface between the CdS and PbS layers, thus resulting in the large shift. In addition to the main peak, a smaller peak at 1200-1300 nm was observed at 1 hour of reaction time and was found to further diminish when the reaction time was increased to 20 hours. This peak could be due to unreacted PbS core QDs, or defect levels introduced during the exchange. The increased broadness of the major NIR PL suggested that a larger distribution of PbS cores have been formed. Table 4.3. A comparison of the NIR absorption PL peak positions, Stoke shifts and QY for the various PbS QDs and PbS/CdS core-shell QDs prepared. Sample 5 nm PbS 5 nm PbS/CdS 5 nm PbS/CdS 6 nm PbS 6 nm PbS/CdS 7 nm PbS 7 nm PbS/CdS Duration of cationic exchange (hours) 0 Absorption peak (nm) PL peak (nm) Stoke Shift (nm) QY(%) 1242 1300 58 40-60 1 1001 1133 132 - 20 902 1067 165 27 0 1454 1492 38 - 2 1176 1290 114 24 0 1690 - - - 1 1530 >1520 - - 116 Chapter 4 1.0 0.9 5 nm PbS 5 nm PbS/CdS(1h) 5 nm PbS/CdS(20h) 0.8 Absorbance 0.7 0.6 (a) 0.5 0.4 0.3 0.2 0.1 0.0 400 200000 180000 600 800 1000 Wavelength/nm 1200 5 nm PbS 5 nm PbS/CdS(1h) 5 nm PbS/CdS(20h) 160000 (b) Intensity 140000 120000 100000 80000 60000 40000 20000 0 800 1000 1200 1400 Wavelength/nm 1600 Figure 4.3. (a) Absorption and (b) PL spectra of 5 nm PbS QDs (black line) and core-shell 5 nm PbS/CdS QDs with 1hrs (red line) or 20hrs of (blue line) of cationic exchange reaction. Excitation wavelength for PL = 532 nm. 117 Chapter 4 Similarly, products were isolated after 1 or 2 hours of cationic exchange for PbS QDs of sizes 6 nm and 7 nm. In all the samples, similar blue shifts were observed in the NIR absorption and PL peaks (Table 4.3 and Figure 4.4). The abrupt drop at wavelengths higher than 1520 nm is due to the spectral wavelength limit of the detector. We also studied the quantum yield (QY) of the PbS core particles and the corresponding PbS/CdS core-shell QDs at excitation wavelength of 715 nm by referencing to a standard dye IR125. QY of 24% was obtained for the 6 nm PbS/CdS QDs but we were unable to measure the QY for the 6 nm PbS due to its emission reaching the detector limit. The QY for the 5 nm PbS was found to be around 40-60%, while a reduction to 27% was observed for the corresponding PbS/CdS core-shell particles. The decrease in QY can either be attributed to the presence of traps at the interface between the CdS and PbS layers or to the possibility of an electron transfer from the PbS to CdS shell. As the conduction band offset between PbS and CdS is small, an electron transfer process will cause a larger separation between the electron and hole, thus leading to a reduction in their radiative recombination.26-29 118 Absorption Intensity( arb units) Chapter 4 6 nm P bS 6 nm P bS/C dS 780nm (ai) 600 900 1200 1500 7 nm P bS 7 nm P bS/C dS 1800 600 900 1200 1800 1500 (aii) W avelength/nm PL Intensity (Arbitary units) (b i) 1100 1200 1300 1400 1500 1600 1300 1400 1500 1600 (b ii) 1100 1200 W a v e le n g th /n m Figure 4.4 (a) absorption and (b) PL spectra of (i) 6nm PbS QDs (dotted line) and its core-shell PbS/CdS QDs (straight line), (ii) 7 nm PbS QDs (dotted line) and its coreshell PbS/CdS QDs (straight line). Excitation wavelength for PL = 532 nm. 119 Chapter 4 HRTEM analysis was attempted to differentiate the PbS core from the CdS shell as shown in Figure 4.5. However, as the d-spacing for cubic CdS is rather similar to cubic PbS, no distinct interface can be resolved, except it is clear that an unresolved shell is surrounding the resolved lattice planes of the core particle. A d-spacing of 0.293 nm determined can be indexed to (200) planes of cubic PbS, which is its dominant lattice planes. Figure 4.5 HRTEM of a 6 nm PbS/CdS core-shell QDs showing crystalline PbS core with (200) planes and an unresolved shell layer. In Chapter 3, we have shown that highly uniform thin films of PbS well dispersed in PS polymer can be fabricated using an imprinting thermal polymerization method. Thus, we prepared thin PS composite films of PbS/CdS core-shell QDs using oleic 120 Chapter 4 acid ligands in a similar way. The absorption and PL spectra of three different thicknesses of such films produced with 5 nm PbS/CdS QDs are shown in Figure 4.6. The QDs showed a broad absorption peak at 902 nm which is similar to that in hexane, indicating no aggregration or growth of the QDs upon impregnation into the PS polymer. However, the PL peak position was found to be slightly red shifted from 1070 nm in hexane to 1107 nm in the film. This could arise from energy transfer from the smaller particles to the larger particles similar to what was observed in Section 3.2. Similar observations and deductions can be made from PS films of different thicknesses prepared with 6 nm PbS/CdS QDs that give a luminescence peak at 1320 nm in hexane (Figure 4.7). 18.3 μm 6.0 μm 4.5 μm Absorbance 0.4 0.3 (a) 80000 Increasing film thickness 70000 18.3 μm 6.0 μm 4.5 μm 60000 50000 Intensity 0.5 (b) 40000 0.2 increasing film thickness 30000 0.1 20000 10000 0.0 400 600 800 1000 1200 1400 Wavelength/nm 0 800 1000 1200 1400 Wavelength/nm Figure 4.6 (a) Absorption and (b) PL spectra of composite films of 5 nm PbS/CdS core-shell QDs dispersed in polystyrene with 4.5 μm, 6 μm and 18.3 μm film thickness respectively. 121 Chapter 4 1.0 0.8 6.6 μm 2.9 μm 100000 80000 Intensity Absorbance 120000 6.6 μm 2.9 μm 0.6 (a) 0.4 40000 0.2 0.0 300 (b) 60000 20000 0 600 900 1200 Wavelength/nm 1500 1000 1200 1400 Wavelength/nm 1600 Figure 4.7 (a) Absorption and (b) PL spectra of composite films of 6 nm PbS/CdS core-shell QDs dispersed in polystyrene with 2.9 μm and 6.6 μm film thickness respectively. 4.2 Z-scan study for PbS/CdS QDs in hexane and PS polymer films The NLO properties of the PbS/CdS core-shell nanoparticles were characterized with Z-scan method. Open aperture Z-scan curves of the PbS/CdS QDs in hexane are shown in Figure 4.8. It is noted that the absorptive nonlinearities of the QDs strongly depend on core size. All the sizes of PbS/CdS QDs in solution display reverse saturable absorption (RSA) at all the intensities. This RSA behavior could result from excited state absorption (ESA), free-carrier absorption (FCA) and/or nonlinear scattering (NLS) mechanisms. Similar to what was observed for the PbS core particles in Chapter 3, NLS is also a dominant phenomenon at higher intensities along with FCA. As such, in the Z-scan theory, both FCA30 and NLS31 should be taken into consideration and Equations (3.1) and (3.2) in Chapter 3 can be used to obtain the best 122 Chapter 4 fits (solid lines in Figure 4.8) between the numerical solutions and the Z-scan data to Normalized Transmittance 1.0 0.8 0.6 0.4 -2 (a) -1 2 65 GW/cm 2 130 GW/cm 0 Z (cm) Normalized Transmittance Normalized Transmittance generate the nonlinear parameters as tabulated in Table 4.4. 1 2 1.0 0.8 (b) 0.6 2 65 GW/cm 2 130 GW/cm 0.4 -2 -1 0 Z (cm) 1 2 1.0 0.8 (c) 0.6 0.4 -2 2 65 GW/cm 2 130 GW/cm -1 0 Z (cm) 1 2 Figure 4.8 Open aperture Z scan curves at different irradiance for different sizes of PbS/CdS core-shell QDs in hexane: (a) 5 nm, (b) 6 nm, and (c) 7 nm. Figure 4.9 displays typical closed aperture Z-scan curves of PbS/CdS QDs in hexane. All the QD sizes in film and solution show negative nonlinear refraction, indicating self-defocusing effect. It is also well-known that thermal nonlinearities also contribute to negative nonlinear refraction; and high single-photon absorption cross-sections may be responsible for these thermal nonlinearities. In order to minimize thermal effects, all our closed-aperture Z-scan curves were collected at lower intensities. Free-carriers 123 Chapter 4 induced by one-photon absorption also contribute significantly to the phase changes of the wave in the sample. Including the free-carrier refraction (FCR) contribution, the nonlinear phase equation is given by Equations (3.4), (3.5) and (3.6) in Chapter 3.30 Both the n2 and σr values of the PbS/CdS core-shell QDs are obtained from the best 1.2 2 30 GW/cm 1.0 (a) 0.8 -2 -1 0 1 2 Normalized Transmittance Normalized Transmittance simulations to the experimental data and listed in Table 4.4. 1.0 (b) 0.8 -2 Z (cm) Normalized Transmittance 2 30 GW/cm 1.2 -1 0 1 2 Z (cm) 1.2 I00= 30GW/cm2 (c) 1.0 0.8 -2 -1 0 1 2 Z (cm) Figure 4.9 Close aperture Z scan curves at different irradiance for different sizes of PbS/CdS core-shell QDs in hexane: (a) 5 nm, (b) 6 nm and (c) 7 nm. Our synthesized PbS/CdS samples showed higher RSA with a decrease in the PbS core sizes from 7 nm to 5 nm. Moreover, when we compared the optical limiting performance of our 5 nm and 6 nm core-shell QDs with the PbS QDs (Chapter 3), we can see that the obtained FCA cross-section for the core-shell QDs are significantly 124 Chapter 4 larger than that for PbS of comparable core sizes. Many reasons can account for the improved nonlinear absorption in the core-shell QDs such as improved electronic passivation of the QD surface, improved FCA due to a smaller PbS core size, contribution from CdS 2PA absorption and photo-induced charge separation.6,7,32 Previously, ZnSe/ZnS core-shell QDs were found to have enhanced intrinsic threephoton absorption (3PA) or 2PA coefficient as compared to ZnSe QDs due to effective surface passivation and localization of the charge carriers in the heterostructure.6,7 We have shown previously that the FCA coefficient increases with the decrease in PbS nanoparticle sizes leading to a greater RSA in the Z scan curves. A similar trend is also observed in our core-shell QDs. Moreover, at excitation wavelength of 780 nm, the CdS shell can undergo MPA at higher excitation intensity due to the much larger band gap of CdS. It is interesting to note that the nonlinear refractive index and FCR cross-section for the 6 nm PbS/CdS QDs is the highest among the samples measured. No theoretical explanation for this occurrence can be given at the moment. From our experiments, we observe that the optical limiting behaviour might be influenced by the surface of the PbS QDs or the thickness of the QD-polymer film. Thus, we would be exploring the effect of surface treatment and film thickness in the following sections. 125 Chapter 4 Table 4.4 Fitted FCA cross-section (σc), nonlinear refractive index (n2), FCR crosssection (σr), and scattering coefficient (αS) of the PbS/CdS core-shell QDs in solution compared to PbS QDs studied in Chapter 3. Estimated σc (cm2) n2×106 σr × 1022 Scattering × 1018 (cm2/GW) (cm3) αs(cm-1) 3.5/0.75 14.7 -4.6 -1.0 3.1 4.2/0.85 9.8 -9.2 -2.4 3.9 5.9/0.65 3.3 -0.8 -0.5 4.5 4.6 nm PbS 6.1 -4.2 -1.5 2.1 5.3 nm PbS 3.0 -3 -1.4 2.4 Sample Size (nm) Core-shell 5 nm PbS/CdS 6 nm PbS/CdS 7 nm PbS/CdS 4.2.1 Effect of surface treatment on nonlinear scattering During our experiments, we found that washing of the samples with ethanol has a consistent effect on its optical limiting performance. Thus, we performed a control experiment whereby the reaction products were split into two tubes with one sample washed twice (Sample I) while the other sample washed at least 4 times (Sample II). This control experiment was repeated twice by synthesizing two different batches of the core-shell QDs. The Z scan curves in Figure 4.10 showed that treatment of the QDs have significant influence on their NLO properties. When we compare samples that have most of the free oleic acid ligand washed away (Sample II) with samples that may have free ligands around (Sample I), we can see that optical limiting response is 126 Chapter 4 dependent on the irradiance. At lower irradiance, we can see that both samples display similar response but at a certain critical irradiance of 80 GW/cm2, a much larger optical limiting response can be observed in the sample with excess ligand. We can attribute the large increase to the difference in nonlinear scattering. Nonlinear scattering in nanomaterials have been attributed to three mechanisms: photoinduced scattering due to solvent bubbles formation33,34, sublimation or evaporation of the material35,36 and photoinduced refractive index mismatch.37 Typically, solvent bubbles are enhanced using nanosecond lasers as the bubbles typically take nanoseconds to nucleate.38 Evaporation of the nanomaterials such as metal clusters can occur at high irradiance using a shorter duration laser such as picosecond lasers.36,39 For our PbS/CdS QDs coated with oleic acid ligand (Sample I), we cannot conclude at the moment which mechanism is operative or if both are operative. However, as our QDs show large linear absorption at the wavelength of excitation, we suggest that at high irradiance the QDs can be rapidly heated up and that transfer its energy to the surrounding oleic acid ligands, inducing chemical reaction to form lead or cadmium oleate. The lead or cadmium oleate forms rapidly expanding micro plasmas which strongly scatter the lasers.35,36 This process might be more favorable when free oleic acid is present in the solution, thus explaining the larger scattering effect in Sample I as tabulated in Table 4.5. 127 Normalized Transmittance Chapter 4 2 2 I00= 50GW/cm I00= 80GW/cm 1.0 0.8 0.6 Sample I Sample II 0.4 -2 -1 0 -2 2 1 -1 0 1 2 Z (cm) Figure 4.10 Open aperture Z scan curves at different irradiance for 5.0 nm PbS/CdS QDs: Sample I (▪) and Sample II (◦)in solution at two different intensities. Table 4.5 Sizes of QDs and scattering coefficient (αS) of the PbS/CdS QDs in Sample I and II. Sample Scattering αs(cm-1) 5 nm PbS/CdS (Sample I) 4.5 5 nm PbS/CdS (Sample II) 1.8 128 Chapter 4 4.2.2 Influence of film thickness on optical limiting We also studied the nonlinear absorption and refraction of different thickness of PbS/CdS QDs in PS polymer films using two different sizes of the core-shell QDs. The open and close aperture Z-scan curves in Figures 4.11 and 4.12 showed that the nonlinear absorption and refraction response increase with increasing film thickness for both sizes of core-shell QDs. In Table 4.6, we can see that the derived FCA crosssection σc and the nonlinear refractive index n2 values are (within experimental error) independent of the film thickness. Any error probably arises from the measurement of the film thickness using the profilometer due to slight non-uniformity of the polymer film prepared. We can see that the free carrier coefficient σc and the nonlinear refractive index n2 values for the films of smaller QDs are larger than its corresponding solution values (14.7 x 10-18 cm2 and -4.6 x 10-6 cm2/GW). We suspect even larger values should be possible if we increase the concentration of the QDs in the monomer mixture used in the polymerization process. Comparing the two series of polymer films, we can see that the smaller QDs (5 nm) have a larger free carrier coefficient σc value but a smaller nonlinear refractive index n2 value compared to the larger QDs (6 nm). This is similar to the trend observed in hexane. 129 Chapter 4 1.2 Normalized Transmittance Normalized Transmittance 2 1.0 0.8 2 20 GW/cm 2 50 GW/cm 2 80 GW/cm 2 100 GW/cm (ai) 0.6 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 50 GW/cm 1.0 (bi) 0.8 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 Z (cm) Z (cm) Normalized Transmittance Normalized Transmittance 1.2 1.0 0.8 2 20 GW/cm 2 50 GW/cm 2 80 GW/cm 2 100 GW/cm 0.6 (aii) -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2 50 GW/cm 1.0 (bii) 0.8 -1.5 -1.0 -0.5 Z (cm) 0.0 0.5 1.0 1.5 Z (cm) 1.0 0.8 (aiii) 2 20 GW/cm 2 50 GW/cm 2 80 GW/cm 2 100 GW/cm 0.6 -1.5 -1.0 -0.5 0.0 Z (cm) 0.5 1.0 1.5 Normalized Transmittance Normalized Transmittance 1.2 2 50 GW/cm 1.0 (biii) 0.8 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 Z (cm) Figure 4.11 (a) Open aperture and (b) close aperture Z scan curves at different irradiance for 5 nm PbS/CdS QDs in PS polymer films of different thickness (i-iii): 4.5 μm, 6 μm and 18.3 μm. 130 Chapter 4 Normalized Transmittance Normalized Transmittance 1.2 1.0 0.8 (ai) 2 20 GW/cm 2 50 GW/cm 2 80 GW/cm 2 100 GW/cm 0.6 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2 50 GW/cm 1.0 (bi) 0.8 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 Z (cm) Z (cm) 1.0 0.8 (aii) 2 20 GW/cm 2 50 GW/cm 2 80 GW/cm 2 100 GW/cm 0.6 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 Normalized Transmittance Normalized Transmittance 1.2 2 50 GW/cm 1.0 (bii) 0.8 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 Z (cm) Z (cm) Figure 4.12 (a) Open aperture and (b) close aperture Z scan curves at different irradiance for 6 nm PbS/CdS QDs in PS polymer films of different thickness (i-ii): 2.9 μm and 6.6 μm. Table 4.6 Size, thickness of films, fitted FCA cross-section (σc), nonlinear refractive index (n2), and FCR cross-section (σr) of the prepared PbS/CdS QDs in PS films. Thickness n2×104 σc (cm2) × Sample σr × 1022 (cm3) 2 18 (cm /GW) (μm) 10 5 nm PbS/CdS 18.3 15.6 -0.44 -1.0 5 nm PbS/CdS 6 15.4 -0.46 -1.1 5 nm PbS/CdS 4.5 12.2 -0.5 -1.1 6 nm PbS/CdS 6.6 6 -2.2 -0.6 6 nm PbS/CdS 2.9 6.1 -1.9 -0.65 131 Chapter 4 4.3 Femntosecond pump-probe transient absorption study of PbS and PbS/CdS QDs Initial studies of relaxation dynamics of the first exciton of PbS synthesized in glasses40,41 revealed the dominance of carrier trapping states on the observed absorption bleaching and induced absorption. Relaxation dynamics of PbS of different sizes synthesized in organic media were subsequently explored by various researchers.14,17,42,43 Ji et al. first explored the transient absorption of 12 nm PbS and composite PbS-Au4 nanocrystals films at two different wavelengths at 700 nm42 and 780 nm43 respectively. At 700 nm, reverse saturable absorption (RSA) was observed and was suggested to be due to dominance of free-carrier (or intraband) absorption 44,45. The relaxation of the photoexcited electron in PbS was suggested to be dominated by two relaxation processes: the fast (3.5 ps) and the slow (29 ps) decay components.42 However, at laser wavelength of 780 nm, an intensity dependent enhancement of the saturable absorption (SA) can be observed in the film of PbS–Au4 as compared to that of PbS films.43 At high laser irradiance of 78 GW/cm2 the PbS–Au4 nanohybrid composites demonstrated stronger photobleaching than the PbS QDs due to resonant energy transfer from PbS QDs to Au nanoparticles by Auger recombination (AR). 4.3.1 Transient Absorption of PbS QDs Before studying the carrier dynamics of the core-shell PbS/CdS QDs, we first studied the corresponding carrier dynamics of the PbS core QDs. As seen in Figure 4.13, the prepared 4.6 nm PbS QDs showed bleaching of the probe pulse at pump irradiance of 132 Chapter 4 45 GW/cm2 with a 780 nm pump pulse. The excitation wavelength of 780 nm is near to the blue tail of the 1Pe-1Ph exciton peak of the 4.6 nm QDs. Such a bleaching of the second exciton absorption arises due to state filling (occupation of the excitonic state by pump photoexcitation across the NC bandgap).18 Bleaching (absorption saturation) have been investigated previously on PbS QDs dispersed in toluene with an absorption peak at 1480 nm by probing at the first and second exciton positions.17 It was found that the absorption saturation was much weaker at the second exciton compared to the first exciton. It was also noticed that at probe wavelength shorter than the second exciton, a photoinduced increase in absorption can be observed. As our pump intensity increases from 45 to 90 GW/cm2, we can see that the bleaching becomes a weakly induced absorption at longer probe delay times. In contrast, the 5.3 nm, 6 nm and 11 nm PbS QDs were found to have only induced absorption at the pump and probe wavelength of 780 nm. Induced absorption in QDs has been attributed to a variety of different mechanisms, e.g. transitions to biexciton states carrier (excited state or intraband) absorption states50, transient carrier induced Stark effects 40,44,49 18,51,52 46-48 , free , absorption from surface-trap and trapped carrier induced shift of the biexcitonic transitions49. The relaxation of the carriers can be fitted to a double exponential decay. It is found that, with a decrease in the sizes of the QDs, a shortening of both the fast and slow components is observed. The slow dynamics in QDs is typically attributed to an electron-hole recombination for deep trap state16, while the fast dynamics has been 133 Chapter 4 commonly attributed to two types of models: AR 12,53 or exciton-exciton annihilation.10 Previously, the fast component relaxation rate showed a 1/r3 dependence on the dot size for CdSe, PbSe and PbS QDs, and was explained by the assumption of carrier transition from the QDs to a certain volume VS out of the QDs Normalized ΔT/T surface.12,44,54 1.0 0.5 0.0 1.0 0.5 0.0 2 I00= 45 GW/cm 2 I00= 90 GW/cm 1.0 0.5 0.0 2 I00= 150 GW/cm 0 30 60 Time delay (ps) Figure 4.13 Pump-probe signal of 4.6 nm PbS QDs excited at 780 nm with different pump intensity and probed at 780 nm. 134 Chapter 4 1.2 0.6 Normalized -ΔT/T 11 nm t1= 100ps t2 =8 ps 0.3 0.0 -0.3 0 70 6 nm t1= 60ps 0.6 t2 = 3 ps 0.3 0.0 0 30 60 1.2 1.2 2 2 I00= 90 GW/cm 0.9 t2 = 2.5 ps 0.3 0.0 0 20 0.6 90 GW/cm 4.6 nm t1= 15ps 0.3 t2 = 1.3 ps 0.9 5.3 nm t1=30ps 0.6 -0.3 2 I00= 90 GW/cm 0.9 -0.3 140 Normalized ΔT/T Normalized -ΔT/T 0.9 Normalized -ΔT/T 1.2 2 I00= 75 GW/cm 0.0 -0.3 40 0 30 60 Time delay (ps) Figure 4.14 Pump-probe signal and life times of PbS QDs, pump at 780 nm probe at 780 nm. 4.3.2 Transient Absorption of core-shell PbS/CdS QDs At the probe wavelength of 780 nm (1.59 eV), bleaching can be observed in the 5 nm PbS/CdS core-shell QDs while induced absorption is observed in the 6 nm PbS/CdS QDs as seen in Figure 4.15. For the 5 nm PbS/CdS sample with the first exciton state (1Se-1Sh) at 900 nm, the pump and probe wavelength of 780 nm is at the blue tail of the first exciton. For the 6 nm PbS/CdS sample with the first exciton peak at 1176 nm, both the pump and probe energies are slightly larger than the 1Pe-1Ph exciton energies. As the energy of the 1.59 eV pump is not large enough to excite the 5 nm PbS/CdS to the 1Pe excited state of the core PbS we observe an overall bleaching of the first exciton absorption of the probe pulse. In the 6 nm PbS/CdS sample where the 1.59 eV 135 Chapter 4 pump can excite the QDs to the 1Pe excited state, induced absorption seems to be the dominant process (Fig 4.16). The pulse duration of our pump/probe is faster than the typical intraband relaxation of PbS of 1-2 ps.55 If the electron is excited to the 1Pe excited state, several processes can occur before the intraband relaxation process from 1Pe to 1Se (Fig 4.16a) is completed. Firstly, the hot electrons can absorb a photon from the probe pulse to higher excited states (excited state absorption, ESA, Fig 4.16b). Secondly, the probe pulse can excite another electron across the bandgap forming another electron-hole pair in addition to the pump generated exciton (biexciton formation, Fig 4.16c).Thirdly, the electron can be trapped to shallow trap states (Fig 4.16d) where both ESA and biexciton formation can also occur. After intraband relaxation is completed, the exciton can undergo interband relaxation through either radiative recombination (Fig 4.16e) or AR (Fig 4.16f). Bleaching of the 1Pe-1Ph exciton state can still occur but previous investigation showed that it is much weaker than at the first exciton.17 However, since we observed a reduced transmission even at low pump intensity, the overall induced absorption process is larger than any weak bleaching. 136 Chapter 4 1.0 5 20 GW/cm2 0.8 (a) 3 −ΔT/T ΔT/T 0.6 30 GW/cm2 70 GW/cm2 4 0.4 0.2 (b) 2 1 0 0.0 -1 -0.2 0 80 Delay (ps) -40 -20 0 20 40 60 80 100 Delay (ps) Figure 4.15. Pump-probe signal of PbS/CdS core-shell QDs of different sizes: (a) 5 nm (b) 6 nm, excited at 780 nm with different pump intensity and probed at 780 nm. Figure 4.16. Schematic diagram of pump generated absorption and subsequent relaxation processes. After excitation by the pump, the exciton relaxes by (a) intraband relaxation, (b) excited state absorption, (c) biexciton formation, (d) trapping to surface/defect states, (e) radiative recombination and (f) Auger recombination (AR). We fitted the relaxation of the charge carriers of our samples to a single exponential decay at lower intensity, while a double exponential decay becomes prominent at higher intensity in Table 4.7. It is found that with increasing pump intensity, shortening of the fast component is observed. Such a shortening for the fast component has been previously observed for both CdSe and PbS samples and have been attributed 137 Chapter 4 to the formation of multi-exciton and subsequent AR or exciton-exciton annihilation.10,12 In AR, nonradiative electron-hole recombination leads to excitation of an electron or a hole and possibly lead to electron ejection.12,53 In exciton-exciton annihilation, high excitation laser intensity for the pump pulses produces multiple excitons per particle that may interact with each other, e.g. one exciton recombines and transfers its energy to the other. Exciton-exciton annihilation typically only occurs at high pump laser intensities when the trap states are saturated as the rate of trapping are typically faster and thus competes with the exciton-exciton annihilation.10. Exciton-exciton annihilation was also found to be dependent on the surface quality of the PbS QDs. A more monodisperse PbS QDs which has a higher surface quality (less trap states) similar to what we have synthesized would show a lower threshold for exciton–exciton annihilation14 compared to a less monodispersed PbS QDs16 due to the greater ease to saturate the lower density of surface traps. It has been well documented that the fast component (biexciton lifetime) scales linearly with the volume of the NCs.12 Comparing our smallest core-shell QDs sample of 5.2 nm to PbS core particles of 4.6nm and 5.3 nm, the fast component life-time was found to remain very fast at around 4 ps (Table 4.7). 138 Chapter 4 Table 4.7. Intensity of the pump; fast and slow lifetime relaxation dynamics of different core-shell PbS/CdS QDs and referenced PbS QDs. Sample 2 Irradiance (GW/cm ) t1 (ps) t2 (ps) 30 3.6 ± 0.8 - 30 6 ± 0.3 - 70 2.8 ± 0.3 255 ± 323 45 2.5 35.7 90 1.3 15 90 2.5 30 5 nm PbS/CdS 6 nm PbS/CdS 4.6nm PbS 5.3nm PbS We also investigated the transient absorption of our samples using a 400 nm pump and probing them at 600 nm. The PbS QDs showed different relaxation of its carriers excited to high energy levels as compared to the PbS/CdS QDs when the pump intensity is increased from 8 nJ to 50 nJ (Figure 4.17). For the PbS QDs at low excitation intensity, the carriers only showed photoinduced absorption. However, as the pump intensity increases, the initial excited state absorption becomes bleaching at longer delay time. We can see from the inset in Figure 4.17a that the switch from excited state absorption to bleaching occurs at a lower probe delay time as the pump power increases. However, for both of the core-shell QDs, the excited state absorption observed is retained with increase in pump intensity within the range investigated. Dementjev et al. observed that the high energy exciton state of PbS in silicate glass 139 Chapter 4 showed only excited state absorption even at longer delay time. They suggested that this is due to the exciton not relaxing to the lowest exciton state in their case.40 We can see clearly that our PbS QDs is different from their observation. This also suggested that in the case for PbS/CdS, the carriers might bypass the first exciton level when undergoing recombination. The relaxation of the carriers in Figure 4.17(a-c) can be fitted to a double exponential function: I probe = I 0 + Ae − t τf + Be − t τs and the resulting fast (τf) and slow (τs) time constants are plotted in Figure 4.18. We can see that both the fast and slow relaxation components decrease in time as the pump intensity increases. The decrease in relaxation time with increase in pump intensity has been previously suggested to be due to AR of multi-excitons.12 140 Chapter 4 0.2 Pump:400nm; Probe:600nm 5.1nm PbS Normalized ΔT/T 0.0 0.0 -0.2 -0.2 -0.4 (a) -0.4 -0.6 8nJ 14nJ 23nJ 30nJ 40nJ 50nJ -0.6 -0.8 -0.8 -1.0 0 -1.0 0 200 50 400 100 600 800 150 1000 1200 Delay Time (ps) 5.1nm PbS/CdS Normalized ΔT/T 0.0 Pump:400nm; Probe:600nm -0.2 -0.4 0 8 nJ 15nJ 24nJ 32nJ 40nJ 50nJ -0.2 -0.6 (b) -0.4 -0.6 -0.8 -0.8 -1 -1.0 0 0 5 200 400 10 15 600 20 800 1000 1200 Delay Time (ps) Normalized ΔT/T 0.0 5.9nm PbS/CdS Pump:400nm; Probe:600nm -0.2 0.0 -0.4 -0.2 -0.6 (c) -0.4 8nJ 14nJ 23nJ 30nJ 40nJ 50nJ -0.6 -0.8 -0.8 -1.0 -1.0 0 0 200 20 400 40 600 60 80 800 100 120 1000 140 1200 Delay Time (ps) Figure 4.17 Relaxation dynamics of (a) 5 nm PbS; (b) 5 nm PbS/CdS and (c) 6 nm PbS/CdS QDs in hexane at 600 nm probe wavelength at different pump intensity with 400 nm pump. 141 Chapter 4 8 5.0nm PbS/CdS 5.1nm PbS 5.9nm PbS/CdS τf (ps) 6 4 (a) 2 0 0 20 40 60 Excitation intensity (nJ/Pulse) 240 τs (ps) 160 5.0nm PbS/CdS 5.1nm PbS 5.9nm PbS/CdS (b) 80 0 20 40 60 Excitation intensity (nJ/Pulse) Figure 4.18 (a) Fast (τf) component and (b) slow (τf) component of relaxation kinetics of 5 nm PbS, 5 nm PbS/CdS and 6 nm PbS/CdS QDs at 600 nm probe wavelength in hexane at different pump excitation intensity using 400 nm pump. In order to further investigate why the excited state absorption is retained in the coreshell QDs, we studied the transient absorption at two different probe delay time of 0 ps 142 Chapter 4 and 500 ps as shown in Figure 4.19. Initially at zero probe delay time, we can see that the PbS core QDs and two core-shell PbS/CdS QDs in hexane showed excited state absorption induced by the probe beam throughout all the probe wavelengths with the largest change at around 600 nm. However, at 500 ps probe delay time, the pure PbS displayed weak exciton bleaching in contrast to the excited state absorption observed in the two core-shell PbS/CdS QDs. At 0 ps delay time, bleaching of these high energy states is much weaker because of the larger number of closely spaced energy levels and thus the excited state absorption can compete more efficiently leading to an overall induced absorption observed. However, at 500 ps delay time whereby intraband relaxation of the PbS carriers have almost been completed (intraband relaxation of the PbS and PbSe carriers have been previously determined to be ~ 3 ps) 49,56, most of the carriers would have relaxed to either the lowest 8 fold degenerate 1Se–1Sh exciton state or are trapped in surface states. This allowed the saturation of the excitonic or trapped states at high pump intensity to compete more efficiently with the excited state absorption leading to an overall bleaching of the absorption of the probe pulse. The presence of excited state absorption in the core-shell QDs at 500 ps delay time (which is much longer than the ∼100 ps timescale 57 reported for the AR of biexciton states in 4.7 nm PbSe) further confirms that induced absorption is likely a combination of several mechanisms such as excited state absorption in addition to the biexcitonic transitions observed at high pump intensity. The coating of the PbS QDs by the CdS shell might have prevented the saturation of the band edge excitonic or trapped states through a possible electron transfer from the high energy levels in PbS to the CdS 143 Chapter 4 shell. The electron transfer is facilitated in quantized PbS as it has a small conduction band offset with CdS.19 Upon excitation to the high energy excited states in the PbS core of the PbS/CdS core-shell QDs, the carrier can transfer to the CdS shell and recombine, thus bypassing the first exciton state. PS172, 400nmex 50nJ PS173, 400nmex 50uW 0.00 0ps 500ps 0.00 -0.02 0ps 500ps ΔT/T ΔT/T (a) -0.01 (b) -0.04 -0.06 -0.02 -0.08 500 600 700 800 500 600 700 800 Wavelength /nm Wavelength /nm PS135, 400nm ex 50uW 0.00 ΔT/T 0ps 500ps (c) -0.02 -0.04 500 600 700 800 Wavelength /nm Figure 4.19 Transient differential transmittance spectra of (a) 5 nm PbS; (b) 5 nm PbS/CdS and (c) 6 nm PbS/CdS measured at different delay times under 50 nJ/pulse excitation intensity with 400 nm pump. 144 Chapter 4 4.4 Summary We have successfully prepared a series of different sized core-shell PbS/CdS QDs and studied their NLO properties using both Z-scan technique and femntosecond transient absorption. Z scan revealed that the core-shell PbS/CdS QDs have larger FCA crosssection compared to PbS of comparable core sizes and that nonlinear scattering was enhanced in the presence of excess oleic acid. Different thickness of QD-PS polymer films prepared by imprinting-thermal cross-linking method showed an expected independence of the FCA cross-section and nonlinear refractive index on the thickness suggesting a uniform distribution of the QDs in the PS film. Transient absorption studies on PbS core QDs revealed either bleaching or induced absorption depending on the size of the QDs. The lifetime of both the fast and slow components are found to be dependent on the size of the QDs. Transient absorption studies at 780 nm pump and probe wavelength on core-shell PbS/CdS QDs showed similar trends as PbS QDs with the smaller core-shell QDs showing bleaching while larger QDs showing excited state absorption. Shortening of the fast relaxation component at higher intensity revealed the dominance of bi-exciton recombination through AR or exciton-exciton annihilation on the overall relaxation processes in both PbS and core-shell PbS/CdS. Excitation of the carriers by using a 400 nm pump into high energy states revealed different relaxation dynamics by the PbS QDs and PbS/CdS QDs depending on the excitation intensity and probe delay time. 145 Chapter 4 4.5 References (1) Han, M. Y.; Huang, W.; Chew, C. H.; Gan, L. M.; Zhang, X. J.; Ji, W. J. Phys. Chem. B 1998, 102, 1884. (2) Jia, W. L.; Douglas, E. P.; Guo, F. G.; Sun, W. F. Appl. Phys. Lett. 2004, 85, 6326. (3) Gong, H. M.; Wang, X. H.; Du, Y. M.; Wang, Q. Q. J. Chem. Phys. 2006, 125, 024707. (4) Wang, X. H.; Du, Y. M.; Ding, S.; Wang, Q. Q.; Xiong, G. G.; Xie, M.; Shen, X. C.; Pang, D. W. J. Phys. Chem. B 2006, 110, 1566. (5) Gan, C. L.; Xiao, M.; Battaglia, D.; Pradhan, N.; Peng, X. G. Appl. Phys. Lett. 2007, 91, 201103. (6) Lad, A. D.; Kiran, P. P.; Kumar, G. R.; Mahamuni, S. Appl. Phys. Lett. 2007, 90, 133113. (7) Lad, A. D.; Kiran, P. P.; More, D.; Kumar, G. R.; Mahamuni, S. Appl. Phys. Lett. 2008, 92, 043126. (8) Nag, A.; Kumar, A.; Kiran, P. P.; Chakraborty, S.; Kumar, G. R.; Sarma, D. D. J. Phys. Chem. C 2008, 112, 8229. (9) Klimov, V. I.; McBranch, D. W. Phys. Rev. Lett. 1998, 80, 4028. (10) Roberti, T. W.; Cherepy, N. J.; Zhang, J. Z. J. Chem. Phys. 1998, 108, 2143. (11) Guyot-Sionnest, P.; Shim, M.; Matranga, C.; Hines, M. Phys. Rev. B 1999, 60, R2181. (12) Klimov, V. I.; Mikhailovsky, A. A.; McBranch, D. W.; Leatherdale, C. A.; Bawendi, M. G. Science 2000, 287, 1011. (13) Burda, C.; Link, S.; Mohamed, M.; El-Sayed, M. J. Phys. Chem. B 2001, 105, 12286. (14) Wu, F. X.; Yu, J. H.; Joo, J.; Hyeon, T.; Zhang, J. Z. Opt. Mater. 2007, 29, 858. (15) Burda, C.; Link, S.; Mohamed, M. B.; El-Sayed, M. J. Chem. Phys. 2002, 116, 3828. 146 Chapter 4 (16) Patel, A. A.; Wu, F.; Zhang, J. Z.; Torres-Martinez, C. L.; Mehra, R. K.; Yang, Y.; Risbud, S. H. J. Phys. Chem. B 2000, 104, 11598. (17) Istrate, E.; Hoogland, S.; Sukhovatkin, V.; Levina, L.; Myrskog, S.; Smith, P. W. E.; Sargent, E. H. J. Phys. Chem. B 2008, 112, 2757. (18) Klimov, V. I. J. Phys. Chem. B 2000, 104, 6112. (19) Fernee, M. J.; Watt, A.; Warner, J.; Cooper, S.; Heckenberg, N.; Rubinsztein-Dunlop, H. Nanotechnology 2003, 14, 991. (20) Pietryga, J. M.; Werder, D. J.; Williams, D. J.; Casson, J. L.; Schaller, R. D.; Klimov, V. I.; Hollingsworth, J. A. J. Am. Chem. Soc. 2008, 130, 4879. (21) Liu, T. Y.; Li, M. J.; Ouyang, J. Y.; Zaman, M. B.; Wang, R. B.; Wu, X. H.; Yeh, C. S.; Lin, Q.; Yang, B.; Yu, K. J. Phys. Chem C 2009, 113, 2301. (22) Lambert, K.; De Geyter, B.; Moreels, I.; Hens, Z. Chem. Mat. 2009, 21, 778. (23) Kang, I.; Wise, F. W. J. Opt. Soc. Am. B-Opt. Phys. 1997, 14, 1632. (24) Tudury, G. E.; Marquezini, M. V.; Ferreira, L. G.; Barbosa, L. C.; Cesar, C. L. Phys. Rev. B 2000, 62, 7357. (25) Nanda, K. K.; Kruis, F. E.; Fissan, H.; Behera, S. N. J. Appl. Phys. 2004, 95, 5035. (26) Kim, S.; Fisher, B.; Eisler, H. J.; Bawendi, M. J. Am. Chem. Soc. 2003, 125, 11466. (27) Balet, L. P.; Ivanov, S. A.; Piryatinski, A.; Achermann, M.; Klimov, V. I. Nano. Lett. 2004, 4, 1485. (28) Chen, C. Y.; Cheng, C. T.; Yu, J. K.; Pu, S. C.; Cheng, Y. M.; Chou, P. T.; Chou, Y. H.; Chiu, H. T. J. Phys. Chem B 2004, 108, 10687. (29) Reiss, P.; Protiere, M.; Li, L. Small 2009, 5, 154. (30) Li, H. P.; Kam, C. H.; Lam, Y. L.; Ji, W. Opt. Commun. 2001, 190, 351. (31) Venkatram, N.; Rao, D. N.; Akundi, M. A. Opt. Express 2005, 13, 867. 147 Chapter 4 (32) Sun, Y. P.; Riggs, J. E.; Rollins, H. W.; Guduru, R. J. Phys. Chem. B 1999, 103, 77. (33) Belousova, I. M.; Belousov, V. P.; Danilov, O. B.; Mironova, N. G.; Murav'eva, T. D.; Ryl'kov, V. V.; Skobelev, A. G.; Yur'ev, M. S.; Ponomarev, A. N. J. Opt. Technol. 2004, 71, 130. (34) Belousova, I. M.; Mironova, N. G.; Scobelev, A. G.; Yur'ev, M. S. Opt. Commun. 2004, 235, 445. (35) Mansour, K.; Soileau, M. J.; Vanstryland, E. W. J. Opt. Soc. Am. B-Opt. Phys. 1992, 9, 1100. (36) Francois, L.; Mostafavi, M.; Belloni, J.; Delaire, J. A. Phys. Chem. Chem. Phys. 2001, 3, 4965. (37) Joudrier, V.; Bourdon, P.; Hache, F.; Flytzanis, C. Appl. Phys. B 2000, 70, 105. (38) Tutt, L. W.; Boggess, T. F. Prog. Quantum Electron. 1993, 17, 299. (39) Wang, J.; Blau, W. J. J. Opt. A: Pure Appl. Opt. 2009, 11, 024001. (40) Dementjev, A.; Gulbinas, V.; Valkunas, L.; Raaben, H. Phys. Status Solidi B 2004, 241, 945. (41) Savitski, V. G.; Malyarevich, A. M.; Demchuk, M. I.; Yumashev, K. V.; Raaben, H.; Zhilin, A. A. J. Opt. Soc. Am. B-Opt. Phys. 2005, 22, 1660. (42) Yang, J.; Elim, H. I.; Zhang, Q. B.; Lee, J. Y.; Ji, W. J. Am. Chem. Soc. 2006, 128, 11921. (43) Elim, H. I.; Ji, W.; Yang, J.; Lee, J. Y. Appl. Phys. Lett. 2008, 92. (44) Malyarevich, A. M.; Gaponenko, M. S.; Yumashev, K. V.; Lagatsky, A. A.; Sibbett, W.; Zhilin, A. A.; Lipovskii, A. A. J. Appl. Phys.2006, 100. (45) Qu, Y. L.; Jia, W.; Zheng, Y. G.; Ying, J. Y. Appl. Phys. Lett. 2007, 90, 133112. (46) Ellingson, R. J.; Beard, M. C.; Johnson, J. C.; Yu, P. R.; Micic, O. I.; Nozik, A. J.; Shabaev, A.; Efros, A. L. Nano. Lett. 2005, 5, 865. (47) Hwang, Y. N.; Je, K. C.; Kim, D.; Park, S. H. Phys. Rev. B 2001, 64, 4. 148 Chapter 4 (48) Sewall, S. L.; Cooney, R. R.; Anderson, K. E. H.; Dias, E. A.; Sagar, D. M.; Kambhampati, P. J. Chem. Phys. 2008, 129, 8. (49) Dementjev, A.; Gulbinas, V. Opt. Mater. 2009, 31, 647. (50) Brelle, M. C.; Zhang, J. Z.; Nguyen, L.; Mehra, R. K. J. Phys. Chem. A 1999, 103, 10194. (51) Klimov, V. I.; McBranch, D. W.; Leatherdale, C. A.; Bawendi, M. G. Phys. Rev. B 1999, 60, 13740. (52) Colvin, V. L.; Alivisatos, A. P. J. Chem. Phys. 1992, 97, 730. (53) Wang, L. W.; Califano, M.; Zunger, A.; Franceschetti, A. Phys. Rev. Lett. 2003, 91, 056404. (54) Okuno, T.; Masumoto, Y.; Ikezawa, M.; Ogawa, T.; Lipovskii, A. A. Appl. Phys. Lett. 2000, 77, 504. (55) Wundke, K.; Potting, S.; Auxier, J.; Schulzgen, A.; Peyghambarian, N.; Borrelli, N. F. Appl. Phys. Lett. 2000, 76, 10. (56) Harbold, J. M.; Du, H.; Krauss, T. D.; Cho, K. S.; Murray, C. B.; Wise, F. W. Phys. Rev. B 2005, 72, 6. (57) Beard, M. C.; Ellingson, R. J. Laser Photonics Rev 2008, 2, 377. 149 Chapter 5 Chapter 5 Fabrication of PbS and Metal/PbS Core-shell Nanowires via Electrochemical Methods Lead chalcogenide nanocrystals have demonstrated tremendous potentials in a variety of applications, including near infrared LEDs, photorefractive materials, photovoltaic cells and biological applications.1-6 Multiple exciton generation in these small bandgap materials has the potential to vastly improve the photovoltaic efficiency in nanoparticle-based photovoltatics.7-9 However, the short exciton lifetime in nanocrystals has placed practical challenges in the extraction of its photogenerated charges before the carriers reunite through Auger recombination.10 One solution to this is to use elongated lead chalcogenides or metal/lead chalcogenide core/shell nanowires (NWs) that allow the extraction of the photogenerated charges from the axial direction of the NWs. Much effort has been dedicated recently to synthesize lead chalcogenide NWs. Branched PbSe NWs have been synthesized by using solution-liquid-solid method with Au/Bi core/shell nanoparticles while PbS NWs were synthesized using Au nanoparticles with dual source starting materials or using Bi nanoparticles with a single source precursor.10-12 Synthesis of lead chacolgenide NWs can also be accomplished through vapour-liquid-solid methods.13-17 In particular, hydrogen free chemical vapor deposition (CVD) synthesis of PbS and hyperbranched PbSe,13,14,17 as well as the hydrogen assisted growth of PbS and PbSe have been 150 Chapter 5 accomplished.15,16 Oriented attachment process was another alternative method that has proven successful in synthesizing PbSe NWs.18,19 Ultra narrow PbS NWs and nanorods (NRs) have also been synthesized using a single source precursor metal xanthate with a suitable ligand. Most interestingly, these PbS NRs of 1.7 nm diameter were found to demonstrate intense fluorescence in the strong confinement regime.20,21 PbS NW architecture can also be obtained through thermal decomposition of the precursor L-cysteine Pb at high temperatures.22 Talapin et al. first accomplished the growth of PbSe/PbS core-shell NWs and PbS/Au NW-nanocrystals. 23 More recently, PbSexS1-x alloys, and PbSe/PbS and PbSe/PbTe core–shell NWs were prepared by solution phase synthesis at moderate temperature through minimizing competitive processes such as ripening and formation of pure lead chalcogenide nanoparticles.24 Electrodeposition of PbS thin films has been performed previously by several different methods. Cathodic deposition of PbS in acidic media using thiosulfate was carried out by several authors.25-27 Anodic deposition of PbS has also been carried out on a lead rod at alkaline pH using Na2S.28 A modified ECALE Method based on the under potential co-deposition of Pb and S was carried out using a solution bath containing EDTA, Pb2+, and S2- on Au(111) single crystals.29 Cyclic deposition of PbS thin films has also been carried out using the same bath by Saloniemi, H. et al on SnO2 and Au substrates.30,31 Electrodepositon of PbS within the anodized alumina (AAO) template is rare and only recently was PbS successfully deposited in AAO using an aqueous solution of DMSO containing 151 Chapter 5 PbCl2 and S precursors.32 Electrodeposition of PbS into well defined pores of AAO has several advantages. Firstly, well-aligned NRs or NWs freely standing on an electrode can be achieved. This allows ready measurements using conductive atomic force microscope to obtain current-voltage (I-V) characteristics and resistivity of the NWs. Secondly, these well-aligned arrays on electrode are useful as the active area in photoelectrochemical cells. Thus, in this paper we report the preparation of free-standing PbS NWs arrays adapting the cyclic process by Saloniemi et al. In particular, we coupled the deposition with a porewidening procedure recently developed33 to make composite metal core (Cu, Au)/PbS shell NWs that are useful for future photoelectrochemical studies. 5.1 Deposition of pure PbS NWs Typical SEM image of freestanding PbS NWs prepared by CV method is shown in Fig 5.1(a). Well-crystallized NWs with average length of 0.75 µm can be deposited by cycling the potential between -0.4 and -0.95 at a scanning rate of 100mV/s. XRD pattern in Fig 5.1(b) gives diffraction peaks that match well with the standard cubic phase of PbS (JCPDS 05-0592) and the Au electrode. The SAED pattern of the sample in Fig 5.1(c) suggests that the NWs are polycrystalline in nature. The TEM and HRTEM images of the PbS nanowires in Fig 5.1(d-e) showed clear lattice planes with d-spacing of 0.345 nm that are close to the planes of PbS. EDX collected on a typical area of NWs showed the presence of Pb and S at a ratio of Pb:S = 1:1.1 (see Fig 5.2). Variations in scan 152 Chapter 5 rate and number of scans confirm that the length of PbS NWs can be linearly controlled by the total time of CV deposition (see Table 5.1). Au (a) Intensity (b) Au Au Au PbS standard JCPDS 05-592 20 (c) 30 40 50 60 2θ (degrees) (d ) 70 80 (e) Figure 5.1. (a) Representative SEM image of the PbS NWs grown by CV deposition at 0.1V/s for 500 scans; (b) XRD pattern indexed to JCPDS 05-0592; (c-e) respectively the SAED, TEM and HRTEM images of the PbS NWs produced. Spectrum7 Pb S Pb Pb 0 2 4 6 Full Scale 668 cts Cursor: 0.013 keV (791 cts) Pb 8 Pb 10 keV Figure 5.2. EDX spectrum revealing that the NWs are composed of Pb and S. Cu peak present is due to the copper grid used for TEM analysis. 153 Chapter 5 Table 5.1. Average length of PbS NWs obtained using different total CV deposition time. Scan rate(V/s) No of scans Total deposition time (sec) Mean length/µm 1 1 500 550 0.26 ± 0.04 2 1 2600 2860 0.61 ± 0.08 3 0.1 250 2750 0.57 ± 0.05 4 0.1 500 5500 0.75 ± 0.13 5 0.05 500 11000 1.15 ± 0.17 The growth process of PbS NWs within the pores of AAO can be investigated by comparing CV profiles obtained at different stages of the deposition. Since resistance is increasing due to the growing NWs, the cathodic current is expected to decrease as shown in Figure 5.3. In addition, it was found that the cathodic peak shifted swiftly from -0.86V to -0.76V during the first few cycles and stabilizes at -0.76V even after 500 scans. The shift of the cathodic peak suggests an irreversible process such as a direct chemical reaction between PbEDTA2- and HS- similar to that observed previously.31 2 Scan Scan Scan Scan (a) Current/mA 1 1 2 10 500 0 -1 -2 -0.9 Voltage/V -0.6 -0.3 Figure 5.3. Typical CV scans obtained at different stages of the deposition of PbS NWs at 0.1V/s between -0.4V and -0.95V. 154 Chapter 5 Formation of the PbS NWs is likely to proceed via the reduction of PbEDTA2- to Pb0 during the cathodic scan, as illustrated in Eqn. (1) below.30 At anodic potentials, two possible reactions can occur. Firstly, Pb0 may oxidize through reaction with hydrogen sulfide to form PbS as in Eqn. (2). It is also possible that S colloids are formed in the solution from the oxidation of H2S at pH 5, and these S colloidal particles adsorb on the electrode surface to react with Pb as in Eqn. (3). PbEDTA2- (aq) + H2O +2e- ↔Pb0 (s) + HEDTA3- (aq) +OH- (aq) Eqn. (1) Pb0 (s) + H2S (aq) +2OH- (aq) ↔ PbS (s) + 2H2O+2e- Eqn. (2) Pb0 (s) + S (coll) ↔ PbS (s) Eqn. (3) 5.2 Deposition of metal (Au, Cu)/PbS core-shell NWs Next, we attempt to prepare metal/PbS core-shell NWs using the pore widening procedure illustrated in Figure 5.4.33 Such procedure can widen the AAO channel walls to give an annular gap around the core wires for deposition of subsequent shell. It has been found that, due to “point effect”, the deposition of shell will be more pronounced on the tip rather than the root of the core metal, thus resulting in a “baseball bat” structure as shown in Fig. 5.4(c). 155 Chapter 5 (a) (b) (c) Figure 5.4. Schematic showing the deposition of core-shell NWs: (a) Core metal wire is deposited into AAO channel, (b) pore widening to create the annular gaps around the core wires, (c) PbS shell is then deposited onto the metal core wires. The synthesis of Au/PbS core-shell NWs was first attempted. Figure 5.5 shows the Au/PbS NWs prepared having a baseball bat structure as shown in (a). EDX data in Fig. 5.5(b) confirms that the top segment of the NWs is PbS. XRD analysis showed that, in addition to Au, the peaks match the cubic phase of PbS but with much enhanced (200) peak. A preferred (200) orientation has also been reported in the constant potential deposition of PbS film on Au(111) single crystal, although this was not suggested in the CV deposition of PbS on SnO2.29,31 Epitaxial electrodeposition of PbS on Au(100) single crystal also showed preferred (200) orientation.27 It was suggested by Vertegel et al. that the lattice mismatch between Au and PbS can be reduced from 45.5% to a mere +2.9% due to a Au(100)/(√2 x √2R45°)-PbS (100) coincidence lattice formations. It is interesting to note, however, that our Au core wires are polycrystalline in nature. Thus, we could not ascertain why a stronger PbS (200) orientation is observed in this case. 156 Chapter 5 003 3600 3200 (a) (a) 2400 Au S AuPb Pb 2000 S C ounts 2800 1600 (b) Pb Au 1200 C 800 Pb Au Au Au Pb Au Pb Au 400 0 0.00 1.50 3.00 4.50 6.00 7.50 9.00 10.50 12.00 13.50 20 30 (c) 40 50 PbS Au PbS PbS Au PbS PbS Au PbS PbS Au PbS Counts PbS keV 60 70 80 2θ (degrees) Figure 5.5. (a) Representative SEM image of Au/PbS core-shell NWs prepared; (b) EDX collected near the top segment of the NWs; (c) XRD pattern fitted to the standard PbS (JCPDS 05-0592). Fabrication of Cu/PbS core-shell NWs using the procedure above, however, encounters slight complications. It was found that the PbS core-shell NWs prepared using CV deposition readily break into segments as shown in Figure 5.6(a). While NWs remain intact in the area marked as I, area marked as II shows remains that have their top segments broken off. On the other hand, bundles of 157 Chapter 5 free NWs, believed to be the broken top segments from areas such as II, were found on the substrate as shown in Fig 5.6(b). EDX analysis tabulated in Table 5.2 revealed that the NWs in area I have elemental ratio of Pb:Cu:S = 1:1.4:1.6; whereas the remains in area II have Cu:S ratio of 9:1, with no Pb signal detected. (a) (b) I II Figure 5.6. Representative SEM images of (a) Cu/PbS NWs prepared by CV deposition, (b) bundles of NWs detected breaking off from the base electrode. Regions marked in (a) indicate (I) the intact NWs arrays and (II) the remains of broken bundles, respectively. Table 5.2. Relative elemental ratio of Pb, Cu and S detected from EDX analysis on areas I and II in Figure 5.6(a) and areas A, B and C in Figure 5.7(a). Elemental ratio detected from EDX (%) Element Area I in Area II in Area A in Area B in Area C in Fig. 5.6(a) Fig. 5.6(a) Fig. 5.7(a) Fig. 5.7(a) Fig. 5.7(a) Pb 25 0 36 28 16 Cu 35 90 30 35 46 S 40 10 34 37 38 158 Chapter 5 The broken bundles of NWs were further investigated by TEM and SAED analysis as shown in Figure 5.7(a-d). SAED pattern taken from region A suggests that single crystalline PbS covers almost 1µm of the broken wire (Fig. 5.7(b)). The PbS deposited becomes polycrystalline in nature as we move down to regions B and C (but we cannot discount the fact that diffraction spots of Cu may interfere in these SAED patterns). As compared in Table 2, EDX analysis on region A of the single wire gives stoichiometric 1:1 ratio of S to Pb. This ratio increases along the wire, to almost 2:1 at the broken ends (region C). Concurrently, an increase in the Cu to Pb ratio was also noted towards the end of the NWs. While sulfur colloids adsorption is a typical problem in the synthesis of PbS in acidic solution,26,30 XRD pattern of the NWs in Fig. 6(e) suggests the formation of chalcocite-M Cu2S in addition to cubic PbS. Thus, it seems that sulfurization of the core Cu wire had occurred during the CV deposition, with well-crystalline and stiochiometric PbS segments only obtained at the top part of the core-shell NWs. 159 Chapter 5 (a) C u 2 S (106,-136) Au Au Intensity C u 2 S (630) Cu (e) (d) Cu PbS,Cu C Au B (c) Au A (b) P bS standard JC P D S 05-592 C u 2 S C halcocite-M standard 33-49 20 25 30 35 40 45 50 55 60 65 70 75 80 2 θ ( degrees) Figure 5.7. (a) TEM image on a single wire among the broken bundles; (b-d) SAED taken on areas marked A, B and C respectively in (a); (e) XRD pattern of the NWs fitting to PbS, Cu, Au and small amount of Cu2S. Further control experiments were performed to study the chemical or electrochemical reaction responsible for the formation of Cu2S in our system. Thus, in one experiment, pre-prepared Cu core wires in pore-widened AAO channels were immersed directly into the pH5 electrolyte bath of Na2EDTA and Na2S for 90 minutes. SEM image (Fig. 5.8) showed that portions of the resultant NWs are roughened, giving fragile intersections along the Cu core wires. XRD analysis confirms the presence of both the Cu and Cu2S diffraction patterns (Fig. 5.9). In the literature, monoclinic Cu2S was reported when a 2 μm Cu film was reacted with a mixture of air and H2S (1:1) for a total of 10 hours.34 Thus, we 160 Chapter 5 believe that partial transformation of our Cu core wires to Cu2S has occurred due to chemical reaction with H2S, which is the dominant species of Na2S at acidic pH 5. Figure 5.8. SEM image of NWs prepared by immersing Cu core wires directly into a solution containing 0.1M Na2EDTA and 0.01M Na2S at pH5 for a total of 90 minutes. In another experiment, we investigated the evolution of CV scans with the Cu core wires immersing in the deposition bath. As shown in Figure 5.10, we can see the eventual appearance of an anodic peak near -0.6V. This peak may correspond to the oxidation of S2- to S, as observed in the literature for the underpotential deposition of S.29 This observation suggests that oxidation of S2- to S is likely to play a role in the transformation of Cu to Cu2S during CV scanning. The transformation of the Cu core wires to Cu2S was found to be more significant during CV scanning, as deduced from the lower intensity of the Cu(111) peak to those of Cu2S(630) and Cu2S(106,-136) in the XRD patterns (Figure 5.9). A similar transformation of Cu NWs to monoclinic Cu2S was reported by using anodic polarization in HS- solutions.35 However, there are at least two different 161 Chapter 5 features in our system compared to the reported system. Firstly, the potential range used in our CV scanning is insufficient for the oxidation of Cu to Cu+ or Cu2+; and secondly, the dominant species of Na2S is H2S at pH 5 whereas the dominant species are HS- in the pH12 system reported. C u (1 1 1 ) (a ) C u /C u 2 S n o c yc lic s c a n n in g Relative intensity C u 2 S (6 3 0 ) C u 2 S (1 0 6 ,-1 3 6 ) (b ) C u /C u 2 S w ith c yc lic s c a n n in g C u s ta n d a rd 0 4 -8 3 6 C u 2 S C h a lc o c ite -M s ta n d a rd 3 3 -4 20 25 30 35 40 45 50 55 60 65 70 75 80 2 θ (d e g re e s ) Figure 5.9. XRD patterns obtained when Cu core wires were immersed in pH5 electrolyte bath of 0.1M Na2EDTA and 0.01M Na2S: (a) without CV scanning and (b) with CV scanning. 0.003 anodic peak Current/A 0.002 scan1 scan50 scan200 scan500 0.001 0.000 -0.001 -0.002 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 Voltage/V Figure 5.10. Evolution of CV scans for Cu core wires immersing in a pH 5 solution containing 0.1 M Na2EDTA and 0.01M Na2S at 100 mV/s. 162 Chapter 5 More regular structures of Cu/PbS NWs could be obtained when the PbS shell wire was deposited using the constant potential procedure. Typical SEM and TEM images (Figure 5.11) show two segments of the NWs - a rough bottom and a denser top segment with larger diameter, resembling the baseball bat structure. SAED analysis on the top segment (area A) indicated that the wire is polycrystalline and most of the spots can be indexed to cubic PbS. EDX line scan and analysis (Fig 5.12 and 5.13) confirmed that the lower segment of the NWs consist of only Cu while the upper segments has peaks that can be assigned to Cu, S and Pb. XRD analysis in Figure 5.11(d) suggested major peaks belonging to both PbS and Cu with only very small amount of Cu2S. (a) (b Au (c) PbS,Cu PbS (d) PbS Counts A Cu Au PbS Au Au PbS PbS PbS Cu PbS PbS B Cu2S Chalcocite-M standard 33-490 20 25 30 35 40 45 50 55 60 65 70 75 2θ (degrees) Figure 5.11. (a) and (b) Representative SEM images of arrays of Cu/PbS coreshell NWs and enlarged view showing the baseball bat structure. (c) TEM and SAED images of a single broken wire. (d) XRD pattern showing clearly Cu and PbS diffraction patterns. 163 80 Chapter 5 IM G 1 0 Intensity 200 1. 0 µm 0.00 Distance 3.24 µm Figure 5.12. SEM image and EDX line scan on bundles of Cu/PbS NWs prepared by constant potential deposition. 007 1000 300 200 CuKa CuKb AuMz OKa 400 A AuLa CKa 500 CuLl Counts 600 AuLl 700 AuMr 800 AuMa CuLa 900 100 0 A 007 B 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 keV 500 006 200 CuKa B 150 100 PbLb 250 PbLa 1.0 1.0 µm µm Counts 300 CuKb 350 PbMz SKa PbMb PbMa PbMr SKb 400 CKa OKa CuLa CuLl 450 50 0 0.00 1.50 3.00 4.50 6.00 7.50 9.00 10.50 12.00 13.50 keV Figure 5.13. SEM image of bundles of Cu/PbS core-shell NWs prepared by constant potential deposition with the corresponding EDX on areas marked as A (near to the base electrode) and B. Samples are placed on a Carbon tape in this analysis. 164 15.00 Chapter 5 5.3 Conductive AFM measurements We attempted to investigate the I-V characteristics of the synthesized NWs in order to evaluate their potential integration into electronic devices. As the PbS NWs are free standing on the gold base, the I-V curve could be readily measured using the tip of a conductive AFM. Typical tapping mode AFM images of these well-aligned NWs are shown in Figure 5.14(a). The resistance of one wire can be obtained from the gradient of the linear region of I-V curve as shown in Figure 5.14(b). (a) 300 200 Current/nA 100 (b) 0 -100 -200 -300 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 Applied Voltage/V Figure 5.14. (a) Tapping mode AFM images of PbS NWs free standing on a Au electrode, and (b) a typical I-V curve of a single PbS wire measured using the AFM tip. By measuring the I-V responses of at least 25 different PbS wires, an average resistivity of 2.3 ± 0.5 Ωcm was obtained for the PbS NWs prepared, using average length = 1.15 μm and diameter = 201 nm determined respectively from 165 Chapter 5 SEM images. This resistivity value is comparable to the 1.7 Ωcm obtained from the 63 nm diameter PbS NWs field-effect transistor (FET) device from Mokari et al.24 5.4 Summary PbS NWs were successfully fabricated by both potentiostatic and cyclic voltametric deposition using a commercial AAO template. Different core-shell and segmented metal/PbS nanowires were also fabricated by using a step-wise pore widening procedure coupled with either CV or constant potential deposition. We have also investigated the mechanism of the formation of composite Cu/PbS NWs using control experiments. Conductive AFM analysis on the PbS NWs give an average resistivity of 2.3±0.5 Ωcm. 5.5 References (1) Bakueva, L.; Musikhin, S.; Hines, M. A.; Chang, T. W. F.; Tzolov, M.; Scholes, G. D.; Sargent, E. H. Appl. Phys. Lett. 2003, 82, 2895. (2) Steckel, J. S.; Coe-Sullivan, S.; Bulovic, V.; Bawendi, M. G. Adv. Mater. 2003, 15, 1862. (3) Choudhury, K. R.; Sahoo, Y.; Jang, S. J.; Prasad, P. N. Adv. Funct. Mater. 2005, 15, 751. (4) Chang, T. W. F.; Maria, A.; Cyr, P. W.; Sukhovatkin, V.; Levina, L.; Sargent, E. H. Syn. Met. 2005, 148, 257. 166 Chapter 5 (5) McDonald, S. A.; Konstantatos, G.; Zhang, S. G.; Cyr, P. W.; Klem, E. J. D.; Levina, L.; Sargent, E. H. Nat. Mater. 2005, 4, 138. (6) Konstantatos, G.; Howard, I.; Fischer, A.; Hoogland, S.; Clifford, J.; Klem, E.; Levina, L.; Sargent, E. H. Nature 2006, 442, 180. (7) Schaller, R. D.; Klimov, V. I. Phys. Rev. Lett. 2004, 92, 186601. (8) Schaller, R. D.; Sykora, M.; Pietryga, J. M.; Klimov, V. I. Nano. Lett. 2006, 6, 424. (9) Luther, J. M.; Beard, M. C.; Song, Q.; Law, M.; Ellingson, R. J.; Nozik, A. J. Nano. Lett. 2007, 7, 1779. (10) Yong, K. T.; Sahoo, Y.; Choudhury, K. R.; Swihart, M. T.; Minter, J. R.; Prasad, P. N. Chem. Mater. 2006, 18, 5965. (11) Hull, K. L.; Grebinski, J. W.; Kosel, T. H.; Kuno, M. Chem. Mater. 2005, 17, 4416. (12) Sun, J. W.; Buhro, W. E. Angew. Chem., Int. Ed. 2008, 47, 3215. (13) Ge, J. P.; Wang, J.; Zhang, H. X.; Wang, X.; Peng, Q.; Li, Y. D. Chem. Eur.J. 2005, 11, 1889. (14) Afzaal, M.; O'Brien, P. J. Mater. Chem. 2006, 16, 1113. (15) Bierman, M. J.; Lau, Y. K. A.; Jin, S. Nano. Lett. 2007, 7, 2907. (16) Fardy, M.; Hochbaum, A. I.; Goldberger, J.; Zhang, M. M.; Yang, P. D. Adv. Mater. 2007, 19, 3047. (17) Zhu, J.; Peng, H. L.; Chan, C. K.; Jarausch, K.; Zhang, X. F.; Cui, Y. Nano. Lett. 2007, 7, 1095. (18) Cho, K. S.; Talapin, D. V.; Gaschler, W.; Murray, C. B. J. Am. Chem. Soc. 2005, 127, 7140. (19) Talapin, D. V.; Black, C. T.; Kagan, C. R.; Shevchenko, E. V.; Afzali, A.; Murray, C. B. J. Phys. Chem. C 2007, 111, 13244. (20) Patla, I.; Acharya, S.; Zeiri, L.; Israelachvili, J.; Efrima, S.; Golan, Y. Nano. Lett. 2007, 7, 1459. (21) Acharya, S.; Gautam, U. K.; Sasaki, T.; Bando, Y.; Golan, Y.; Ariga, K. J. Am. Chem. Soc. 2008, 130, 4594. (22) Shen, X. F.; Yan, X. P. J. Mater. Chem. 2008, 18, 4631. (23) Talapin, D. V.; Yu, H.; Shevchenko, E. V.; Lobo, A.; Murray, C. B. J. Phys. Chem. C 2007, 111, 14049. 167 Chapter 5 (24) Mokari, T.; Habas, S. E.; Zhang, M. J.; Yang, P. D. Angew. Chem., Int. Ed.2008, 47, 5605. (25) Takahashi, M.; Ohshima, Y.; Nagata, K.; Furuta, S. J. Electroanal. Chem. 1993, 359, 281. (26) Sharon, M.; Ramaiah, K. S.; Kumar, M.; Neumann-Spallart, M.; LevyClement, C. J. Electroanal. Chem. 1997, 436, 49. (27) Vertegel, A. A.; Shumsky, M. G.; Switzer, J. A. Angew. Chem., Int. Ed.1999, 38, 3169. (28) Scharifker, B.; Ferreira, Z.; Mozota, J. Electrochim. Acta 1985, 30, 677. (29) Oznuluer, T.; Erdogan, I.; Sisman, I.; Demir, U. Chem. Mater. 2005, 17, 935. (30) Saloniemi, H.; Ritala, M.; Leskela, M.; Lappalainen, R. J. Electrochem. Soc. 1999, 146, 2522. (31) Saloniemi, H.; Kemell, M.; Ritala, M.; Leskela, M. Thin Solid Films 2001, 386, 32. (32) Wu, C.; Shi, J. B.; Chen, C. J.; Lin, J. Y. Mater. Lett. 2006, 60, 3618. (33) Liu, C. M. L., P.Y.; Liang, E.P.; Sow; C.H.; Chin, W.S. Submitted 2009. (34) Wu, Q. B.; Ren, S.; Deng, S. Z.; Chen, J.; Xu, N. S. J. Vac. Sci. Technol. B, 2004, 22, 1282. (35) Liang, C. H.; Terabe, K.; Hasegawa, T.; Aono, M. Solid State Ionics 2006, 177, 2527. 168 Chapter 6 Chapter 6 Synthesis of MnO, Mn3O4, Mn2O3 and MnO2 nanocrystals and their catalytic studies Manganese oxides have generated much interest because of their promising applications as electrode materials, soft magnetic materials and catalysts. Transition metal oxides such as MnOx often possess good catalytic properties either on their own or as supported, making them suitable candidates as cheaper alternative catalysts that can be effective at lower temperatures for the oxidation of CO, NOx and other volatile organic compounds.1,2 Catalytic properties of unsupported MnOx nanoparticles however, have not been widely studied. The main objective of this chapter is to explore effective ways to synthesize nanosized MnOx, and to study their catalytic properties as unsupported MnOx for the oxidation of CO - one of the most important model reactions in pollution control. In this chapter, nanocrystalline MnO, MnO2, Mn2O3 and Mn3O4 were synthesized and investigated. These nanocatalysts were characterized using XRD, TEM, BET surface area analysis, FT-IR and TGA. Their catalytic activities towards CO oxidation are investigated using a micro-plug-flow reactor connecting to an online GC instrument. The effects of different particle sizes, different synthesis reagents and stability of nanocrystalline MnOx over time as well as their catalytic performances are studied and correlated. We organized our discussions in the following according to the different types of manganese oxides 169 Chapter 6 concerned, with a brief literature review of each oxide given at the beginning of each section. The detailed preparation methodologies of the various MnOx are described in Chapter 2. The influence of some important parameters such as reaction temperature, time, and relative concentration of the reagents will be further elaborated below. 6.1 Synthesis of MnO nanocrystals and their catalytic studies Large MnO octahedral nanocrystals and MnO@C core-shell composite nanoparticles have been studied for their electrocatalytic activities in oxygen reduction in aqueous basic medium.3 MnO nanoclusters have also been shown to possess ferromagnetic properties in contrast to the antiferromaganetism of bulk MnO.4 Furthermore, investigation of MnO nanomaterials with hierarchically spherical superstructures synthesized from manganese(III) acetylacetonate revealed interestingly that both an antiferromagnetic transition temperature at 121 K, similar to that of bulk, and a low temperature ferromagnetic ordering were present.5 MnO has also been utilized as a hard nonmagnetic shell for magnetic materials like FePt and CoFe2O4 leading to the enhancement of the magnetic properties of the core.6,7 Another interesting study on a doubly inverted core-shell system consisting of antiferromagnetic MnO nanoparticles with ferrimagnetic Mn3O4 shells displayed an unusually large magnetization above its TC produced by the uncompensated spins on the surface of the MnO particles.8,9 Non-hydrolytic methods have become one of the most popular methods to prepare a variety of MnO and Mn3O4 nanocrystals because of the ability to obtain highly 170 Chapter 6 monodisperse particles of various shapes and sizes without the problems of hydroxylated surfaces commonly encountered in hydrolytic synthesis. Decomposition of manganese cupferronate was investigated in the presence of TOPO under solvothermal conditions to generate MnO.10 O’Brien et al 11 utilized Mn(ac)2 in oleic acid (OA) and trioctylamine to synthesize monodisperse 7-20 nm MnO nanocrystals; while Park et al.12 synthesized Mn3O4 and MnO from Mn(acac)2 in oleylamine. More recently, O’Brien also prepared single-crystals oxides of Fe, Mn, Co or Ni by the decomposition of their metal acetylacetonate in hexadecylamine followed by overcoating the surface with amphiphilic polyelectrolytes to ensure their complete phase transfer into the aqueous medium.13 The noncoordinating solvent approach first highlighted by Peng et al. in the synthesis of colloidal II-VI semiconductor nanocrystals has been further extended by both Peng et al. and Hyeon et al. to the synthesis of a variety of metal oxides.14-16 The method proves to be an effective method to balance the nucleation and growth of nanocrystals since varying the concentration or chain length of the ligands for the monomers allows control of the nanocrystal sizes. In particular, Peng et al. presented a simple and general strategy for the size and shape control growth of magnetic (Cr, Mn, Fe, Co, Ni) oxide nanocrystals based on the thermolysis of different metal carboxylates in noncoordinating solvents; while Hyeon et al. developed a multigram synthesis of monodisperse transition metal (Fe, Mn, Co) oxide nanocrystals using metal oleates.15,16 It was proposed 171 Chapter 6 that the synthesis of such monodisperse particles was a result of the separation of the nucleation and the growth steps due to their different temperaturedependency.16 There is still relatively fewer reports on anisotropic manganese oxide in the organic media. Synthesis of MnO multipods was vigorously explored by three different groups.17-19 Synthesis of one dimensional Mn3O4 and MnO was accomplished by Cheon et al. and Hyeon et al. using metal chloride in mixed solvent of oleylamine and oleic acid, and manganese carbonyl in mixed solvent of oleylamine and trioctylphosphine respectively.20,21 In this section, we prepared MnO nanoparticles via the decomposition of manganese(III) acetate in a noncoordinating solvent (ODE). Details of the experimental procedure have been given in Chapter 2. 6.1.1 Synthesis of MnO nanoparticles Typical XRD patterns of the prepared MnO nanoparticles (NPs) are shown in Figure 6.1. All diffraction peaks can be fitted to the cubic rock salt structure of MnO (JCPDS 07-0230; Fm3m, a = 4.445 Å). Figure 6.1. MnO nanoparticles prepared at manganese acetate-to-OA ratio of 1:3 at 300oC. For comparison, simulated XRD patterns from the database are shown as vertical lines 172 Chapter 6 Since our precursor manganese(III) acetate is in +3 oxidation state, the mechanism of formation of MnO in our synthesis must account for the change in the oxidation state of the manganese. We depict in Figure 6.2 a schematic of our preparation procedure for better understanding of the following discussions. Figure 6.2. Schematic showing the procedure in our synthesis of MnO NPs. Thus, we monitor the reaction using uv-vis spectroscopy. When the spectrum of the degassed reaction mixture is compared to that heated to 260OC (Figure 6.3), a disappearance of the 485 nm peak is detected. It has been reported that Mn(III) species give an absorption band at 460 nm due to a d-d transition (with assignment of 5B1g + 6Eg when D4h symmetry is assumed) while Mn(II) is known to show no appreciable absorption in this region.22 173 Chapter 6 2 .5 2 .0 Absorbance 1 .5 (a ) 1 .0 (b ) 0 .5 0 .0 300 400 500 600 700 800 900 W a v e le n g th /n m Figure 6.3. UV-visible absorption spectra of the reaction mixture (a) degassed at 80 OC for 30 minutes and subsequently quenched with hexane, and (b) that was heated to 260 OC in nitrogen environment. It has been suggested that reduction of Mn(III) in n-butanoic acid at 125oC proceeded via two competitive pathways of either oxidative decarboxylation or alkyl oxidation:23 Oxidative decarboxylation : Mn III + CH 3 (CH 2 ) n COOH ← ⎯→ Mn III ( - O 2 C(CH 2 ) n CH 3 ) + H + Mn III ( - O 2 C(CH 2 ) n CH 3 ) ⎯ ⎯→ Mn II + ⋅ (CH 2 ) n CH 3 + CO 2 eq(1) eq(2) Alkyl oxidation : Mn III + CH 3 (CH 2 ) n COOH ⎯ ⎯→ ⋅ CH 2 (CH 2 ) n COOH + Mn II + H + eq (3) As oleic acid is a primary acid, it is expected that both pathway will contribute to the reduction leading to Mn(II) species as the common products. When we use a 174 Chapter 6 manganese(III) acetate to OA ratio that is less than 3, the amount of oleic acid is insufficient to completely replace the acetate ions. Thus, the acetate ions is likely to be also involved in the reduction of Mn(III). Thus, it is conceivable that, based on the disappearance of the absorbance peak, all the Mn(III) acetate/Mn(III) oleate species have been transformed to the Mn(II) acetate/ Mn(II) oleate species during heating to high temperatures. Formation of MnO nanocrystals would subsequently occur through the decarboxylation of the manganese(II) acetate/oleate species to form manganese(II) oxide with acetone and CO2 as the byproducts, similar to that reported by O’Brien.11 A similar mechanism was proposed by Hyeon et al. who described the formation of Fe3O4 from iron(III) oleate in 1-octadecene solution.16 They proposed a two-step formation whereby initial nucleation occurs at 200– 240 °C triggered by the dissociation of one oleate ligand from the Fe(oleate)3 precursor by CO2 elimination and subsequent growth through dissociation of the remaining two oleate ligands at ~300 °C. In the following sections, we discuss the effect of the various experimental parameters in controlling the size and shape of MnO NPs produced as summarized in Table 6.1. In this particular system, control of particle size and shape can be relatively easy by three primary factors, namely the OA concentration, the temperature and the reaction time. 175 Chapter 6 Table 6.1. The average size and shape of MnO produced using different reaction conditions. Average size from TEM[nm] b Morpholog y 30 Reaction temperatur e [oC] 280 22.6±1.4 22 60 280 21.0±2.0 1:1.0 22 60 320 13±3.4 d 1:3.0 31 60 300 e 1:3.0 28 30 320 f 1:3.0 28 60 320 g 1:5.0 62 60 320 36.8±2.4 (edge length) 19.7±1.7 (edge length) 18.9±2.0 (side width)∼21. 2 (Diagonal length) 27.7± 2.2 spherical with some minor rods faceted with some minor rods spherical with some minor rods octahedron h 1:6.6 98 10 320 28.4±2.0 (Cubic) i j 1:6.6 1:6.6 98 98 20 60 320 320 31.8± 1.5 33.1± 2.2 Exp t [Precursor ] /[OA] Time Interva l [min]a Reaction time[min ] a 1:1.0 22 b 1:1.0 c Octahedron faceted (Truncated octahedron TO) Mostly faceted Mostly faceted Octahedron Mostly faceted k 1:8.0 126 15 320 Hexapod N.A. and its related fragments a the time interval needed to observe visible colour change of the reaction mixture signaling onset of decomposition. b For faceted shape, size is measured from edge to edge. For octahedron shape, size is measured between two opposite apexes. 176 Chapter 6 6.1.2 Size control: Oleic acid concentration The concentration of OA has a strong influence on the temperature and time interval before decomposition is visible (colour changes from clear orange/yellow to dirty green). By increasing the OA concentration from 1 to 8 equivalents, the time interval before visual colour change was found to be lengthened from 22 to 126 minutes (Table 6.1). The minimum reaction temperature whereby MnO NPs can be formed was also found to increase with increasing OA concentration, i.e. from 280OC, 300OC to 320OC for 1, 3 and 5 equivalents of OA respectively. These temperature and time dependences of the reaction are similar to the formation of iron oxide from iron acetate reported by O’Brien and Murray, who suggested that decomposition of different intermediates occurs in the reaction solution.24 Thus, when the oleic acid is in surplus as in 5-8.5 equivalents, formation of manganese(III) oleate and its subsequent transformation to manganese(II) oleate is dominant. When the oleic acid is insufficient to completely replace the acetate ion, both oleate and acetate ions contribute to the nucleation and growth of the NPs. The lengthening of time needed to observe a visual color change when OA is in excess can also be accounted for using the effect of monomer activity.25 Complexation of oleate ion to the manganese ion decreases the active monomer concentration, leading to a reduction in the rate of nucleation. This will lead to a longer reaction time and a higher reaction temperature needed before the concentration reaches the supersaturation concentration for nucleation. 177 Chapter 6 Complexation with the OA ligand also affects the final particle size because the ligand complexes with the precursor/monomer and thus hinder nucleation, leading to a smaller number of initial nuclei during nucleation15,26 Hence, a control of the nanocrystal size is easily obtained by changing the OA concentration. By heating at 320OC while changing the manganese(III) acetate to OA ratio from 1:1, 1:3, 1:5 to 1:6.6, relatively monodisperse nanoparticles were obtained as seen in Figure 6.4. The estimated sizes are respectively: 13.0±3.4, 18.9±2.0 and 27.7± 2.2 and 33.1± 2.2 nm as shown in Table 6.1 (samples c, f, g and j) and a plot showing the correlation between average sizes and OA concentration is presented in Figure 6.5. As the amount of oleic ligand is not enough to completely replace the acetate ligand for the 1:1 ratio, both acetate and oleic acid might contribute to the nucleation and subsequent growth. However, as acetate is a poor capping ligand as compared to oleic acid, we can see that beside the relatively larger particles, some smaller particles and rodlike particles can also be observed. We can see that some of the rods are aggregations of small particles and oriented attachment is a likely process under these surfactant-limited conditions. 178 Chapter 6 (b) (a) (d) (c) (e) Figure 6.4. Representative TEM images of faceted MnO nanoparticles produced at 320oC for 60 min using Mn(III) to OA ratio of (a) 1:1; (b) 1:3; (c) 1:5 and (d) 1:6.6. (e) TEM image showing MnO hexapods and fragments produced at 320oC for 15 min at 1: 8 ratio. 179 Chapter 6 40 Particle siz e (n m ) 35 30 25 20 15 10 5 0 0 1 2 3 4 5 6 7 Equivalence of Oleic acid Figure 6.5. A plot showing the relationship between particle size and amount of OA ligand utilized. 6.1.3 Size control: Temperature Temperature is also another important factor affecting the final particle size. At the same Mn : OA ratio of either 1:1 or 1:3, a decrease in the size of the nanoparticles can be observed as seen in Table 6.1 (samples b, c and d, f) with increasing temperature from 280 or 300 to 320oC respectively. In a noncoordinating solvent system, the nucleation step plays a significant role in the control of the particle size. Thus the smaller final size observed suggests that higher temperature affects the kinetics of the nucleation step. It has been reported by Mulvaney who studied the nucleation of CdSe in hot ODE that both the initial and final sizes of CdSe particles are smaller at higher temperatures.26 They suggested that the nucleation and particle growth processes were decoupled and 180 Chapter 6 the activation temperature for nucleation is relatively higher. Thus, higher temperatures would produce more nuclei resulting in overall smaller particle sizes.26 It should be noted that the former argument hinges on the assumption that higher temperatures do not accelerate coagulation. Thus, it is likely that a similar phenomenon led to the decrease in size at higher temperatures in our system. Although higher temperature allows the formation of smaller particles, processes such as Ostwald ripening might also be accelerated. Hence, when the Mn: OA ratio reached 1:1, the size distribution was found to be larger at higher temperature of 320oC (Figure 6.3a) as compared to 280oC (Figure 6.5a) because of the insufficient capping leading to Ostwald ripening. 181 Chapter 6 6.14 Shape control The shape of the nanocrystals obtained can also be varied by changing the ligand concentration, the reaction time and the growth temperature. The effect of ligand concentration on the particle shape can be divided into two regimes. At low OA concentration, typical shapes ranging from spherical, truncated cubes and faceted particles can be obtained depending on the reaction time and the reaction temperature employed for the reaction. However, at high ligand concentration when the nucleation becomes very much hindered, formation of bipods, hexapods and their corresponding fragments were observed. Thus, at 280OC and very low ligand concentration, the common product changes from spherical at 30 min to more faceted after 60 min of reaction (Figure 6.6a). At 320OC, highly faceted nanocrystals seem to be the common product obtained after 60 min of reaction time (Figure 6.4b-d). However, one can see that the amount of particles that are more cubic-like is higher when the concentration of the ligand is increased. Cubic shape/ rhombohedra nanocrystals (which are likely the 2D projection of octahedron that will be discussed later) can be readily obtained using Mn(III) to OA ratio of 1: 3 when the growth reaction is carried out at 300OC for 60 min (Figure 6.6b) or when the growth time is only 30 min at 320OC (Figure 6.6c). 182 Chapter 6 (a) (b) (c) Figure 6.6. Representative TEM images of MnO nanoparticles produced at Mn(III) to OA ratio of (a) 1:1 ratio at 280oC for 60 min, (b) 1:3 ratio at 300oC for 60 min, and (c) 1:3 ratio at 320oC for 30 min. When the nanocrystals are aged at 320OC for prolong period of time, the cubes seem to transform to faceted nanocrystals. By comparing the nanocrystals obtained at 30 min and 60 min at 320OC (Figure 6.4b, Figure 6.6c and Table 6.1(e-f)), one can observe that the nanocrystal sizes are quite similar although the shapes of the particles are different. One possible explanation is an intraparticle ripening process that is accelerated with an increase in temperature.15 Thus, at 300OC and Mn(III): OA ratio of 1:3, the particles are still stable as the cubic-like shape after 60 min of reaction but when the reaction temperature is increased to 320oC, the truncated cubic shape is slowly transformed to the faceted particle shape with time. The HRTEM image of a single 30 nm MnO cubic like particle prepared from Mn(III) to OA ratio of 1:3 after 1 hr reaction shows high crystallinity as seen in Figure 6.7a. The lattice spacing of 0.25 nm corresponds to the d spacing between 183 Chapter 6 adjacent (111) crystallographic planes of fcc rock salt MnO nanocrystals. It is known that, typically for a fcc cubic structure, the (100) faces are of relatively lower energy than the (111) surfaces.27 Thus, growth is often faster along the direction leading to cubic shape particles that are typically bounded by four (100) faces. The dominance of (111) lattice on the crystal face found in Fig 6.6a suggests that this cube is actually octahedron particles bound by (111) faces oriented along or .28 Moreover, aliquots withdrawn at 10 and 20 min of reaction for the 1:6.6 ratio as show in Figure(6.8a-b) show clearly that a mixture of cubic and octahedron shapes can be observed readily in one grid. Figure 6.7. HRTEM image of one faceted MnO particle prepared from Mn(III) to OA ratio of 1:3, 320OC and reaction for 30 min. 184 Chapter 6 (a) (b) Figure 6.8. (a-b) TEM images of aliquots at Mn(III) to OA ratio 1:6.6 at 320oC withdrawn at (a) 10min and (b) 20min. 185 Chapter 6 In Figure 6.9, we have performed TEM analysis on samples tilted to different angle with respect to the electron beam. We can see that faceted and spherical particles have transformed to particles that resemble cubes/rhombohedral at larger tilt angles. This further confirms that the cubic shaped particles observed could well be octahedrons. (a) (c) (b) (d) Figure 6.9. Samples withdrawn at 30 min from reaction of Mn(III) to OA molar ratio of 1:4 at 320OC with different tilted angle relative to x axis: (a) 0o ; (b) -10o; (c) -20 o; (d) -30 o. The HRTEM image of a faceted MnO particle prepared at 320OC after 1hr of reaction is presented in Figure 6.10a. Two distinctive lattice spacing of 0.221 nm and 0.256 nm corresponding to d spacings between adjacent (200) and (111) crystallographic planes of MnO nanocrystals can be observed. The appearance of (200) facets in the faceted particles suggests that the octahedron, upon further 186 Chapter 6 annealing for an additional 30 min, becomes more truncated due to either intraparticle ripening or Ostawald ripening.28 Wang et al. has described the geometrical shapes of cubooctahedral nanocrystals as a function of a ratio, R, being the ratio of growth rate along the to that of the .28 A value of R = 1.73 can be used to describe an octahedron while R = 0.87 can be used to describe the cubooctahedron. We can see from our TEM and HRTEM evidences that particles at the early stage of the reaction are octahedron in shape which thus corresponds to a high R value close to 1.73. As the reaction proceeds, growth in the direction leads to the development of the lower energetic (100) surfaces. As the (100) surfaces develop the octahedron becomes more truncated giving truncated octahedral shapes with a lower R value (0.87 < R < 1.73). (a) (b) Figure 6.10. (a) HRTEM of faceted MnO nanoparticle prepared using acetate:OA molar ratio of 1:3 at 320OC after 1hr of reaction, and (b) its expanded view. 187 Chapter 6 Formations of manganese bipods and hexapods have been recently studied vigorously by three different groups.17-19 Zitoun et al. and Zhong et al. attributed the formation to the oriented attachment17,18 while Ould-Ely et al. proposed an alternative solvothermal growth/etching process occurring in conjunction with microstructural defects. In our present system, the multipods as seen in Figure 6.6e are more dominant when the ligand concentration utilized is very high but the precise mechanism cannot be easily elucidated. Although we also observe chain-like aggregates that suggest oriented attachment such as those seen in Fig 6.9b, Rusakova et al. suggested the chain-like aggregates can be formed from solvothermal etching that facilitates a 2D growth mechanism from the corners of a nanocube.19 Our HRTEM images of the multipods as shown in Figure 6.11b also reveal the presence of planes with d spacing of 0.21 nm which can possibly be attributed to the planes of cubic Mn3O4. Moreover, we can also observe Moiré interference patterns as seen in Figure 6.11a similar to that observed in the findings of Rusakova et al. whereby MnO has been partially oxidized to cubic Mn3O4.29 However, our XRD analysis do not reveal a distinctive presence of Mn3O4 which suggest that any Mn3O4 formed are likely to be in a small quantity.29 188 Chapter 6 Figure 6.11. HRTEM images of (a) mulitpod MnO particle prepared using acetate to OA molar ratio of 1:8 at 320OC after 60 min of reaction showing moiré interference patterns, and (b) bipod MnO particle showing the planes of cubic Mn3O4. 6.1.5 Catalytic activity of MnO The catalytic properties of unsupported MnOx compared to their bulk counterparts for the oxidation of CO have not been studied previously. We proceed to test the catalytic properties after the successful preparation of the different manganese oxide nanocrystals in this Chapter. The catalytic activity measurements were performed in a micro-plug-flow reactor as detailed in Section 2.5. Before each experiment, the dried sample was pre-treated by heating in argon gas at a suitable temperature between 50-300°C. 189 Chapter 6 Monodispersed cubic MnO nanoparticles with average diameter of 36 nm were chosen for the catalytic studies. The total surface area of the as prepared sample was estimated from a BET plot as shown in Figure 6.12. The value of 20 m2/g determined is much lower as compared to the other porous catalysts such as zeolite, which often have surface areas in the order of hundreds. However, clear hysteresis behaviour was observed for MnO nanocrystals, suggesting that the sample was of porous nature. Adsorption increased more significantly at higher P/Po values with pore condensation, and this gave rise to the hysteresis as desorption was higher than the adsorption. Figure 6.12. BET isotherm of the as prepared nanocrystalline MnO, showing the adsorption ( ) and desorption ({) of N2 molecules. TEM and XRD analysis were performed to monitor the sample before and after testing. As compared in Figure 6.13, post-reaction TEM indicated some degree of 190 Chapter 6 degradation in the shape of the nanoparticles. The loss of capping ligands also led to clustering of the particles, although the particle sizes remained as 30-40 nm. Figure 6.13. TEM images of MnO nanocrystals: a) as prepared and b) after catalytic reaction. XRD (Figure 6.14), on the other hand, suggested a phase conversion has occurred when the as prepared samples were subjected to pre-treatment. From the XRD patterns recorded after pre-treatment at 300°C and post-reaction, Mn3O4 (JCPDS: 24-0734) was identified instead. Ramesh et al. have performed a temperatureprogrammed oxidation experiment and found that bulk MnO was oxidized at around 165°C.30 Even though the pre-treatment was conducted in Ar atmosphere, we suspected trace amounts of O2 could have been present and oxidized the MnO sample. 191 Chapter 6 Figure 6.14. XRD patterns of the MnO nanocatalyst as prepared, after heat treatment and after catalytic reaction. For comparison, simulated XRD patterns from the database are shown as vertical lines. IR spectral analysis could also allow us to differentiate between different manganese oxide since MnO has been reported to display two shoulders at 462 and 560 cm-1.31 Similarly, our as prepared sample showed broad peak with two shoulders at 476 and 596 cm-1 as observed in Figure 6.15 and Table 6.2.After pretreatment in nitrogen at 300oC, the two shoulder peaks have shifted to values that are identical to those obtained when the MnO are heated in air. This suggested the partial oxidation of the MnO catalyst during the pretreatment phase. It is known that oleate anion (C17H33COO-) display asymmetric and symmetric carboxylate anion (COO-) stretching bands at ∼1565 and 1434 cm-1 respectively.32 192 Chapter 6 Thus, the peaks at 1560 and 1431 cm-1 observed in our MnO samples in Figure 6.12 and Table 6.2 suggested that the oleate anion was strongly bonded to the Transmittance MnO surface despite repeated washings and heatings. (bi) (ai) 3500 3000 2500 2000 1500 1000 -1 W avenum ber(cm ) 500 Transmittance (b ii) (a ii) 1000 900 800 700 600 -1 W a v e n u m b e r(c m ) 500 400 Figure 6.15. IR spectra of MnO nanocrystals (ai) as prepared and (aii) magnified region between 400 to 1000 cm-1; (bi) after pretreatment at 300oC in N2 and (bii) magnified region between 400 to 1000 cm-1. 193 Chapter 6 Table 6.2. Main peaks in the IR spectrum of MnO (a) as prepared; (b) heated in air; (c) after pretreatment at 200oC in N2 and (d) after pretreatment at 300oC in N2. Sample (a) MnO as prepared MnO (air) Mn-O vibration 476 (w, sp) (b) 490 (vs, sp) (c) MnO 494 (200°C) (vs, sp) (d) MnO 488 (300°C) (vs, sp) Notation: w = wide; sp = ……. 596 (w, sh) 606 (vs, sp) 606 (vs, sp) 606 (vs, sp) C-H stretching (aliphatic) 1431 1560 (s, sp) (s, sp) 1539 (w, b) 1558 (w, b) 1550 (w, b) 1655 (w, b) Isothermal TGA profiles showing weight changes of the MnO samples under a set temperature and ambient are compared in Figure 6.16. In purified air at 200°C, a weight gain was observed following a weight loss. The initial weight loss might be attributed to the loss of moisture and organic ligands, whereas the weight gain was probably due to oxidation of MnO. Such a weight gain was not noticeable under nitrogen ambient and three temperatures (200, 300 and 400°C) were tried out in order to determine the effectiveness in the removal of the organic ligands. As suggested by the FTIR spectroscopic analysis above (Table 6.2), trace amounts of the organic ligand persistently remained in all samples. Thus, in order not to adversely affect the nanostructured morphology, the pre-treatment temperature was kept at or below 300°C for the subsequent catalytic experiments. 194 Chapter 6 Figure 6.16. Isothermal TGA curves of as prepared MnO heated under different ambient gases, a) 200°C in air, b) 200°C in N2, c) 300°C in N2 and d) 400°C in N2 Figure 6.17 shows the catalytic activities of nanocrystalline MnO pre-treated in nitrogen at 200°C and 300°C, as compared to that of bulk MnO. For both nanocatalysts, the light-off temperature (defined as the temperature to attain a 10% conversion) was observed at 180°C and 250°C respectively. A comparatively lower light-off point generally implies higher activity. By the end of the experiments, both nanocatalysts showed an almost 100% conversion of CO. Bulk MnO, on the other hand, had its light-off at 240°C and only achieved 13% conversion by 250°C. Thus, the nanocrystalline MnO has shown improved catalytic behavior towards oxidation of CO as compared to bulk MnO. 195 Chapter 6 Figure 6.17. Catalytic activities of bulk MnO (), nanocrystalline MnO pretreated at 200°C (z) and 300°C (S). It was found that the light-off temperature was delayed to 250°C after pretreatment at 300oC. While pre-treatment at higher temperatures was expected to remove the organic ligands more effectively, several adverse effects, e.g. oxidation, sintering or carbon coating might also occur.33 Previous mechanistic studies on bulk MnO have suggested that its catalytic operation proceeds by the Mars-van Krevelen mechanism.34,35,36 This mechanism occurs in two steps: 1) lattice oxygen of the manganese oxide oxidizes an adsorbed CO molecule, leaving behind an oxygen vacancy and a partially reduced metal ion, and 2) re-oxidation of the metal ion with gaseous oxygen to restore to original state. However, as our samples after pretreatment have undergone partial oxidation, the higher activities 196 Chapter 6 observed after the light-off temperature in our case can most likely be attributed to a mixed oxide or that of Mn3O4. 6.2 Synthesis of Mn3O4 nanocrystals and their catalytic studies Alivisatos was the first to generate Mn3O4 nanoparticles with a reasonable dispersity using metal cupferron complexes.37 Next, Zhang and Yan et al. discovered that simple formate was also an effective precursor to synthesize Mn3O4 nanoparticles in oleylamine solvent.38 A self-assembled three-dimensional arrays of monodispersed Mn3O4 nanoparticles encapsulated within a thin shell of MnO2 were synthesized by a one-step thermal decomposition of [Mn(acac)2] in oleylamine in air. This sample showed interesting behavior similar to a spin-glass material, a possible consequence of the strong interaction between the Mn3O4 core and the MnO2 shell.39 Syntheses of Mn3O4 nanowires and nanorods have been accomplished by several methods including calcinations of precursor powders prepared by inverse microemulsion, thermal oxidation of pure manganese powders and aqueous chemical precipitation.40-42 Thermal evaporation of MnCl2 was also found to be an effective method to produce mixture of MnO and Mn3O4 nanowires.43 Recently, vacuum calcinations of a precursor prepared by hydrothermal reactions have also produced nanorods with diameters of about 100 nm and lengths up to 15–20 μm.44 197 Chapter 6 In this section, we prepared Mn3O4 nanoparticles from the decomposition of manganese nitrate tetrahydrate in oleylamine. Different procedures as detailed in Section 2.2.10 were followed, i.e. direct heating method or injection method. A schematic showing the injection method is presented in Figure 6.18. Figure 6.18. Schematic showing the synthesis of Mn3O4 NPs using the injection method. 6.2.1 Synthesis of Mn3O4 nanocrystals Typical XRD patterns of the Mn3O4 nanoparticles prepared using different methods and conditions are shown in Figure 6.19. All diffraction peaks can be indexed to the tetragonal Mn3O4 structure (JSPDS 24-0734). 198 Chapter 6 o (c)140 C injection 1hr o (b)180 C injection 3hr o (a)180 C direct heating 45mins tetragonal Mn3O4 20 30 40 50 60 70 80 2θ (degree) Figure 6.19. Typical XRD pattern of Mn3O4 particles prepared at different conditions (a) direct heating to180oC with 45mins reaction, (b) injection at 180oC with 3hrs reaction, (c) injection at 140oC with 1hrs reaction. For comparison, simulated XRD patterns from the database are shown as vertical lines. IR spectra shown in Figure 6.20 and Table 6.3 indicated two strong peaks at 623 and 521 cm-1 for the Mn3O4 nanoparticles prepared. The presence of three strong and well-resolved peaks at around 420, 499 and 616 cm-1 in bulk Mn3O4 has been ascribed to the co-existence of Mn2+ and Mn3+ ions in the spinel lattice, thus 199 Chapter 6 facilitating vibration coupling.31. It is also noted that a very strong peak at 1385cm-1 is detected in our samples, which can be assigned to the asymmetric NO3 stretching close to the value in crystalline nitrates such as Co(NO3)2.45 A similar peak was also observed to evolve in the IR spectra when aqueous Mn(NO3)2 was heated in air.46 Interestingly, for the Mn3O4 sample prepared at 140oC for 1 hr, a splitting of the asymmetric NO3 stretching can be observed similar to that in Co(NO3)2. Ehrhardt et al. suggested that the splitting could be caused by a lowered NO3− symmetry in the crystal lattice (site symmetry or crystal field splitting).45 When Mn3O4 was annealed at 180oC for 3hrs, the asymmetric NO3− stretching was found to be very weak. Table 6.3. Major peaks observed in the IR spectrum of Mn3O4 prepared with different methods. Sample Mn-O vibration C-H stretching asymmetric (aliphatic) NO3 stretching 623 521 ~2922; ~2853 1385(vs) Mn3O4 (vs, sp) (vs, sp) (direct heating 180oC 45mins) 623 517 ~2922; ~2853 1385(vs) Mn3O4 (vs, sp) (vs, sp) ~1353(s) (injection 140oC 1hr) Mn3O4 631 527 ~2922; ~2853 1385(w) (OLA(vs, sp) (vs, sp) injection 180oC 3hrs) Key: vs, very strong; s, strong; m, medium; w, weak; b, broad; sp, sharp; sh, shoulder 200 Chapter 6 o Transmission abritary unit 180 C 45mins o 140 C 1hrs o 180 C 3hrs (a) 4000 100 Transmittance 90 3000 2000 -1 1000 Wavenumber(cm ) 0 (b) 80 70 60 50 40 1000 900 800 700 600 -1 500 Wavenumber cm 400 Figure 6.20. IR spectra of (a) OLA-capped Mn3O4 (a) overview and (b) magnified region. 201 Chapter 6 Shape controlled synthesis of different nanocrystals had been studied previously by numerous authors.25,47,48 Various factors, e.g. the crystalline phase of the nucleating seeds, selective adsorption of capping ligands onto particular crystallographic facets, kinetic vs thermodynamic growth, etc. have been identified to play significant roles in determining the final particle shape.47,48 In Table 6.4, the average sizes and morphologies of the Mn3O4 nanoparticles obtained under different reaction conditions are tabulated. Table 6.4. The particle sizes and morphologies of Mn3O4 obtained at different reaction conditions. Expt [Precursor] Reaction /[Amine] time[min] Reaction T [oC] Diameter [nm] Length [nm] Aspect Morphology ratio a Injection 1 1:10.6 10 180 7.1±0.2 46.0±8.7 6.5 nanorod b Injection 2 Injection 3 Injection 4 Injection 5 Injection 6 1:10.6 30 180 7.1±0.3 47.3±7.2 6.7 nanorod 1:10.6 60 180 7.6±0.4 44.9±9.0 5.9 nanorod 1:10.6 180 180 10.4±1.2 37.0±9.7 3.5 Fat nanorod 1:31.8 30 180 8.1±1.3 24.2±4.9 3.1 nanorod 1:31.8 180 150 Nanorods and polydehrons g Injection 7 1:10.6 180 150 h Fast heating Sequential heating 1:8 45 180 Spherical particles and cubes nanorod 1:8 45 120 (20mins) 180 (25mins) c d e f i 7.9±0.2 50.0±12.2 6.3 Spherical particles and cubes 202 Chapter 6 We first monitored the size and shape evolution of Mn3O4 nanorods formed via injection of 0.15g/ml of manganese nitrate precursor at 190OC with growth at 180 O C by analyzing products withdrawn at different reaction time (Figure 6.21). The nanorods formed after 10 min of reaction was found to be quite monodisperse with diameter of 7.1 ± 0.2 nm and length of 46.0 ± 8.7 nm. The diameter of the nanorods increased slightly to 7.6±0.4 nm and the length shortened to 44.9±9.0 nm after 60 min of reaction. After 3 hrs of reaction, the diameter had grown to 10.4 ± 1.2 nm while the average length had further shortened to 37.0 ± 9.7 nm. This corresponded to an overall decrease of the aspect ratio from 6.5 to 3.5. Figure 6.21. Representative TEM images of Mn3O4 nanorods produced by injecting 0.15g/ml of precursor at 190oC followed by growth at 180OC for (a) 10 min, (b) 60 min and (c) 3hours. We have found that the average length of the nanorods can also be adjusted by changing the precursor concentration used for injection. By decreasing the concentration from 0.15g/ml to 0.05g/ml, nanorods with diameter of 8.1±1.3 nm 203 Chapter 6 and length of 24.2±4.9 nm (i.e. aspect ratio of 3.1) can be obtained after 30 min of reaction at 180oC (Figure 6.22a). (a) (b) (c) Figure 6.22. Representative TEM images of Mn3O4 nanorods produced by (a) injection of 0.05g/ml at 190oC and growth at 180OC for for 30mins ; (b-c) injection at 160oC and growth at 150OC of (b) 0.15g/ml for 5mins, (c) 0.05g/ml for 3hrs. In addition, temperature of the injection and growth also has a significant effect on the particle shape and size. By using 0.15g/ml and temperature of injection at 160OC followed by growth at 150 OC, a mixture of small nanorods, nanoparticles and larger polyhedrons are obtained (Figure 6.22b) after 5 min of reaction. If we reduce the injection concentration to 0.05g/ml, on the other hand, a mixture of small nanospheres and larger nanocubes can be observed (Figure 6.22c) after three hours of reaction. Mn3O4 nanorods can also be obtained by using the direct heating method whereby the reaction mixture was heated directly from room temperature to 180OC at 9oC/min (Figure 6.23a). As the reaction mixture was heated from room 204 Chapter 6 temperature, the orange-brown mixture started to turn dark brown after 110OC and became dark-reddish brown by 160OC. If the reaction mixture was subjected to sequential heating instead, i.e. by heating to 120OC for 20 min (including heating up) followed by 180OC for 25 min, only a mixture of small spherical particles and larger cubic particles were observed as seen in Figure 6.23b. In the following paragraphs, we attempt to explain the changes in size and shape of the particles as observed above. Figure 6.23. Representative TEM images of (a) Mn3O4 nanorods produced by direct heating to 180oC for 45 min; (b) Mn3O4 particles produced by sequential heating procedure. The HRTEM image of a typical Mn3O4 nanorod prepared at 180OC after 45 min of reaction is presented in Figure 6.24. Lattice spacing of 0.237 nm corresponding to the d spacing between adjacent (004) planes in Mn3O4 nanocrystals with tetragonal symmetry can be observed. In a study of the kinetic 1D growth mode of CdSe, it was proposed that monomers are consumed by the rapid growth of the 205 Chapter 6 higher energy (001) crystal facet, thus resulting in the direction being the long axis of the quantum rods.25 It has also been previously reported that the (004) surfaces in tetragonal Mn3O4 NCs have a higher energy compared to other crystal surfaces.21 Thus, in our present system, the higher energy (004) crystal facet compared to the other lower energy surfaces might have caused the much faster growth along the direction. Figure 6.24. HRTEM image of a typical Mn3O4 nanorod prepared at 180OC after 45mins of reaction We have noted from Figure 6.21b, Figure 6.22a and Table 6.4(b,e) that monomer concentration is important for the growth of 1D nanorods. A large drop in aspect ratio was detected when the precursor concentration was reduced. Higher precursor concentration under the same injection condition would generate a higher monomer concentration, and this produced longer nanorods. 206 Chapter 6 The decreasing aspect ratio of the nanorods with increasing reaction time has been previously explained by Peng to be a consequence of the change in growth mode of the nanocrystals.25 Initially when the monomer concentration is high, the nanocrystals are in the 1D kinetic growth mode and large aspect ratio nanorods can be observed after 10 min of reaction. As the reaction progresses with reaction time, a corresponding decrease in the monomer concentration also occurs. The growth mode possibly reaches the 3D growth mode between 10 min to 1 hour of reaction since the diameter of the nanorods has increased but the overall aspect ratio remains similar at 6.7 and 5.9 after 30 min and 60 min of reaction respectively. The nanorods obtained in our studies, however, had a rather polydispersed length thus making it difficult to conclusively distinguish the 1D growth stage from the 3D growth stage. When the monomer concentration dropped even further below the level needed for 1D growth after 2-3 hours, significant change in the aspect ratio and length of the nanorods suggested that the reaction has reached the 1D to 2D intraparticle ripening growth stage. It is interesting that by using either direct heating or sequential heating, different nanocrystal shape and sizes can be obtained. This suggests a certain induction period may be operative, whereby the manganese nitrate precursors has been converted to the active monomer/species that accumulates until the nucleation threshold is exceeded. Similar induction period have been observed in previous studies of iron oxide nanocrystal synthesis.49 Hence, if we utilize a fast heating method, the monomer concentration will rise rapidly to that required for the kinetic 1D growth. In the sequential heating method, on the other hand, a gradual 207 Chapter 6 and slower monomer generation occurs thus favouring a slower thermodynamic 3D growth. This observation is similar to those observed previously for iron oxide nanospheres.24,49 The same explanation can also account for the shift from 1D to 3D growth mode when we reduce the temperature of injection from 190oC to 160oC using the same precursor concentration because the generation of the active monomer is slower at lower temperature. Manganese nitrate tetrahydrate is known to decompose under both air and nitrogen in a series of steps at different temperature forming different MnOx phases46,50: Mn( NO3 ) 2 ⋅ 4 H 2 O ⎯175 ⎯⎯ ⎯→ MnO1.2 ( NO3 ) 0.8 o → Mn ( NO3 ) 2 ⎯⎯ C 208o C ⎯220 ⎯⎯ ⎯→ α − Mn2 O3 ⎯780 ⎯⎯ o → MnO2 ⎯⎯ o → Mn3 O4 C C 560 o C In our present system, the manganese nitrate tetrahydrate was able to dissolve\disperse in oleylamine forming possibly an unknown Mn(NO3)2oleylamine complex. XRD and IR analysis clearly showed that the oxide formed in the system after 1 hour of reaction at both 140oC and 180oC was tetragonal Mn3O4 instead of either MnO2 or α-Mn2O3. The presence of asymmetric NO3 stretching in the IR but absence of any distinct manganese nitrate in the XRD suggested that the anions could be either adsorbed on the nanorods surfaces, compensating for the charge balance, or have formed an amorphous layer on the nanorods.51 208 Chapter 6 6.2.2 Catalytic activity of nanocrystalline Mn3O4 The Mn3O4 nanocatalysts that we used for the CO oxidative catalytic experiments were the nanorods formed from direct heating at 180oC. The average length of the Mn3O4 is 48.6 ± 12.0 nm with a uniform diameter of 7.9 ± 0.2 nm (Figure 6.25a). TEM image of the post-reaction sample (Figure 6.25b) indicated some sintering of the oxide due to the removal of the capping groups upon heating. Nevertheless, the dimensions of the observable nanorods were approximately the same as those before reaction. Similar to nanocrystalline MnO, there was some degradation to the smoothness of the nanorods surfaces. Figure 6.25. TEM images of Mn3O4 a) as prepared and b) after catalytic reaction. BET analysis of this Mn3O4 nanocatalyst gave surface area of 89 m2/g (Figure 6.26), showing hysteresis behaviour typical of porous materials. 209 Chapter 6 Figure 6.26. BET plot of the as-prepared nanocrystalline Mn3O4, showing the adsorption ( ) and desorption ({) of N2 molecules. XRD patterns (Figure 6.27) confirmed the identity of Mn3O4 (JCPDS: 24-0734) after pretreament and after catalytic reaction at 300 oC). Figure 6.27. XRD patterns of the Mn3O4 nanocatalyst as-prepared, after heat treatment and after catalytic reaction. For comparison, simulated patterns from the database are shown as vertical lines. 210 Chapter 6 After heating isothermally at 250°C in N2 for more than 1 h, it was observed from the TGA profile (Figure 6.28a) that there was still a continuing weight loss. This could mean that the oleylamine capping groups were still being removed gradually at this temperature. Another TGA profile at 300°C (Figure 6.28b) was obtained. This time, the weight loss was sharp and remained almost constant with a negligible increase over the isothermal period. Hence, it was concluded that 300°C is the more suitable temperature for purging of the capping groups during the pre-treatment process. Figure 6.28. Isothermal TGA curves of as prepared Mn3O4 at a) 250°C and b) 300°C in N2 The catalytic activities of the Mn3O4 NCs pre-treated at 300°C were determined as shown in Figure 6.29. A light-off temperature of 234°C and conversion of almost 100% at 300°C were obtained and compared well with those of the bulk catalyst. It is noted that the overall catalytic profile of Mn3O4 nanocatalyst is rather similar to that of MnO pre-treated at 300°C (Figure 6.17). As discussed earlier, nanocrystalline MnO could be oxidized during the catalytic experiment to 211 Chapter 6 a mixed oxide with Mn3O4. However, there are also other factors such as morphology and surface structure differences between the two nanocatalyst samples. 100 Conversion (%) 80 60 40 20 0 50 100 150 200 o 250 300 Temperature ( C) Figure 6.29. Catalytic activities of bulk Mn3O4 (), nanocrystalline Mn3O4 pretreated at 300°C (S). The catalytic activity of Mn3O4 was expected to be high if we consider the Marsvan Krevelen mechanism. In order to be catalytically active, the catalyst had to contain a redox couple and also high lattice oxygen mobility. Mn3O4 fulfilled both conditions. Although the onset of high activity occurred at slightly high temperature, the high conversion achieved at 300°C was still a promising result. The sudden sharp increase in the catalytic conversion could be explained by considering the adsorption of gaseous molecules on the nanocrystalline oxide surfaces. The free surface energy of these NCs was likely to be much higher than 212 Chapter 6 the bulk form. As a result, the initial adsorption of gaseous molecules could have a stabilizing effect that reduces the surface energy. Therefore, at lower temperatures, adsorption occurred continuously on the nanocatalyst surfaces, causing the nanocrystalline oxide to show an even lower initial activity than the bulk oxide. However, the high surface area of the nanocrystalline oxide eventually led to a high build-up of the concentration of adsorbed molecules. After a certain critical temperature, these molecules gained enough energy for chemical reaction and desorption, resulting in the sudden sharp increase in catalytic activity. 6.3 Catalytic studies of nanocrystalline MnO2 Mesoporous gamma manganese oxide/gold nanoparticle composites have been reported to be efficient catalyst for the elimination of VOC as well as NOx and SO2 at ambient temperatures due to the facile redox properties and formation of radical species on the gold surface.52 Recently, a series of ∝-, β-, γ-, and δ- MnO2 nanorods were synthesized by hydrothermal method and their activities for CO oxidation were found to decrease in the order of ∝- ≈δ > γ- >β- MnO2. This trend was found to be correlated with the different channel structures present in the different crystal phases of the MnO2 nanorods.53 Synthesis of different crystal phases and structures of MnO2 was, however, more easily achieved by using aqueous phase methods. For example, various α-MnO2 hierarchical, urchin-like structures were prepared by Xie et al. via a homogeneous 213 Chapter 6 catalytic route.54 Wang et al. synthesized α-MnO2 nanowires using a sol-gel solution of manganese (II) acetate and AAO templates.55 There was also hydrothermal synthesis, such as the work of Li et al., who studied preparation of α-MnO2 using aqueous solution of MnSO4 and (NH4)2S2O8,56 while Cheng et al. used KMnO4 and H2SO4.57 More recently, different hierarchical 3D architectures of MnO2 have also been synthesized by different groups. Suib et al. synthesized using a hydrothermal route 3D hierarchical nanoarchitecture of ε-MnO2 that showed interesting semiconducting and magnetic properties.58 Portehault et al. utilized a low temperature one pot aqueous precipitation method to synthesize unique hierarchical 3D architectures consisting of uniform core–corona birnessite layered manganese oxide particles that was free of any organic additives.59 All of the above methods suffer some disadvantages in one way or the other. Even via a catalystic reaction, preparation of some of these manganese oxides still require a few days. Templating process can be troublesome, while hydrothermal treatment requires high temperature and pressure. In our investigation, we adapted two different low temperature methods to produce the nanowires of MnO2. In the first method, α-MnO2 was obtained through the reduction of KMnO4 by H2SO4. The chemical reactions may be postulated as follows: MnO4- + 8H+ + 5e Æ Mn2+ + 4H2O (1) 2MnO4- + 3Mn2+ + 2H2O Æ 5MnO2 + 4H+ (2) 214 Chapter 6 Gamma phase ρ-MnO2, on the other hand, was obtained through the oxidation of Mn2+ by S2O82-. The chemical reactions may be postulated as follows: Mn2+ + 2H2O Æ MnO2 + 4H+ + 2e (3) S2O82- + 2e Æ 2SO42- (4) XRD pattern of the as-prepared MnO2 obtained from the first reduction method is shown in Figure 6.30, together with those after pre-treatment and catalytic experiments. The diffraction patterns can be assigned to either α-MnO2 (JCPDS: 44-0141) or cryptomelane-M (K2-xMn8O16, JCPDS: 44-1386) in the standard database. Cryptomelane-M is a type of porous manganese oxide, also known as octahedral molecular sieves. It has MnO6 octahedra sharing edges and corners to form 2 x 2 tunnels. These tunnels can be occupied by K+ ions, which may be incorporated during synthesis. It is noted that the intensities of the obtained XRD pattern are slightly different as compared to the reference peaks, indicating that there are preferred directions in the prepared samples. XRD patterns of the nanocatalysts after pretreatment and catalytic reaction do not show any noticeable changes. 215 Chapter 6 Figure 6.30. XRD patterns of the α-MnO2 nanocatalysts as-prepared, after heat treatment and after catalytic reaction. Simulated patterns from the database are shown as vertical lines. On the other hand, the products prepared by the second oxidation method are identified as ρ-MnO2 (JCPDS: 12-0714) from the XRD analysis (Figure 6.31). ρMnO2 can be seen as one variation of γ-MnO2 using the defects model proposed by De Wolff.60 In this model, γ-MnO2 are constructed from chains of MnO6 octahedra with different amounts of intergrowth of 1 x 1 pyrolusite(r) and 1 x 2 ramsdellite(R) tunnel structures as seen in Figure 6.32. Typically, the De Wolff defect in the γ MnO2 can be estimated by using a parameter Pr which is the 216 Chapter 6 probability of an r fault that consists of the occurrence of a single chain of octahedra instead of double chains. Figure 6.31. XRD pattern of the ρ-MnO2 nanocatalyst as-prepared, after storing for prolonged period and after catalytic reaction. For comparison, simulated pattern from the database are shown as vertical lines. 217 Chapter 6 Figure 6.32. Diagram showing the (a) pyrolusite r (a), and ramsdellite R (b), (c) structures, (d) (e): schematic [100] (d) and [001] (e) views of a R–r intergrowth. Rn or rn are a succession of n consecutive slabs R or r, respectively (Diagram taken from Hill et al61). 218 Chapter 6 The γ-MnO2 sample showed some changes in structure after the catalytic reaction. Most of the major peaks remained after reaction except the (201) peak at 2θ = 22.2°, which was greatly diminished. An additional broad peak at ~30o was also detected after reaction. This observation suggests some changes to the oxide during reaction. According to the De Wolff model, the intensity of the (201) peak can change with the parameter Pr. As Pr was increased from 0 to 0.5, the intensity was known to have decreased.60,61 The broad peak at 2θ = 30° might be due to peaks attributed to other phases of MnOx formed during the CO oxidation reaction. The oxidation of CO over the different MnO2 has been suggested previously to undergo the following mechanism:53 CO (ads) + 2MnO2 → CO2 + Mn2O3 (1) CO (ads) + MnO2 + Mn2O3 → CO2 + Mn3O4 (2) 2Mn3O4 + O2 (air) → 2MnO2 + 2Mn2O3 (3) 2Mn2O3 + O2 (air) → MnO2 (4) The γ-MnO2 structure was known to collapse easily during the CO oxidation process, thus slowing down oxidization of the intermediates Mn2O3 and Mn3O4 phases back to MnO2.53 In Figure 6.30, we also analyze the XRD pattern of the oxide after storage for a prolonged period of time. The material remains as γMnO2 upon storage but the intensity of the (201) peak is slightly diminished. As explained above, a decrease in this intensity suggested an increased Pr parameter. This has significant implications on the CO catalytic oxidation as will be described later. 219 Chapter 6 Typical TEM image of the as prepared α-MnO2 (Figure 6.33a) shows long nanowires with diameters of 20-25 nm. The length of the nanowires is approximately 190-230 nm. The post-reaction TEM image (Figure 6.33b) shows that most of the nanowires remain intact, with average diameter of about 20 nm and length ~170-230 nm. This implies that the oxide did not undergo any major morphology or size changes during the reaction. Figure 6.33. TEM images of α-MnO2 a) as prepared and b) after reaction. 220 Chapter 6 Typical TEM image of the as prepared γ-MnO2 in Figure 6.34a shows that it is made up of very long nanofibers with diameter of ~15 nm. The nanofibers coagulate into larger structures of approximately 300-400 nm. Post-reaction TEM image (Figure 6.34b) shows an increased degree of agglomeration, such that the amount of nanofibers observed is greatly reduced as compared to the as-prepared sample. This could pose as a problem because sintering may reduce the catalytic activity of the oxide. TEM analysis on the as prepared oxide after it was stored for a prolonged period of time (Figure 6.34c) also indicates some coagulation of the oxide. This is a common phenomenon for nanosized compounds, which tend to sinter and coagulate to form bigger particles in order to reduce free surface energy. Figure 6.34. TEM images of γ-MnO2 a) as prepared, b) after storing for prolonged period and c) after catalytic reaction. From the BET plots in Figure 6.35, the surface area of α-MnO2 was determined to be 77 m2/g, while that of γ-MnO2 was 71 m2/g. Similarly, isotherms with hysteresis were observed, implying that these manganese oxides are porous in 221 Chapter 6 nature. The TGA profiles in Figure 6.36 show that both types of nanocrystalline MnO2 sustained only one weight loss over a time span of 100 min. This confirms that the catalysts remain stable as MnO2. (a) (b) Figure 6.35. BET plot of as prepared nanocrystalline (a) α-MnO2 and (b) γ-MnO2, showing the adsorption ( ) and desorption ({) of N2 molecules Figure 6.36. Isothermal TGA curves of as prepared a) γ-MnO2 and b) α-MnO2 at 200°C in N2 222 Chapter 6 Nanocrystalline MnO2 was shown to have fairly high activities towards catalyzing the CO oxidation reaction (Figure 6.37), with almost 100% conversion of CO by 250°C (Run 1). The light-off temperatures are 169°C for α-MnO2 and 105°C for γ-MnO2. Comparatively, bulk MnO2 with a surface area of 0.9m2/g only managed a 9% conversion by 250°C. Figure 6.37. Catalytic activities of bulk β-MnO2 ({), nanocrystalline α-MnO2 run 1 ( ) and run 2 () and nanocrystalline ρ-MnO2 run 1 (V) and run 2 (T). The higher conversions of nanocrystalline MnO2, as compared to the bulk oxide, could be attributed to the difference in surface area. The large surface area of the nanoparticles provided a larger number of active sites for catalysis. Bulk MnO2 particles had a wide size distribution, ranging from 60 to 500 nm and surface area of 0.9 m2/g. It is noted, however, that although the surface area of the nanowires was about 70 times larger than the bulk sample, its conversion did not increase as proportionately. Moreover, although the surface areas of the α-MnO2 and γ-MnO2 223 Chapter 6 are quite similar, γ-MnO2 showed much higher activity toward CO oxidation. Hence the activities of these catalysts are probably not only dependent on their high surface areas. In general, it is known that differences in phase, morphology or surface structure can also result in different levels of activity for CO oxidation. In particular for MnOx, studies have suggested that the CO chemisorption process, the strength of the Mn-O bond and the ease of transformation of intermediate oxides Mn2O3 and Mn3O4 into MnO2 are important factors for the catalytic activities of the MnO2 nanorods.53 Since surface structure is crucial for heterogeneous catalysis, it is important to consider the different surfaces of the different types of MnO2. Previously, it was found that α-MnO2 had a higher activity as compared to γMnO2 due to the larger channel dimensions and larger OH-/H2O content in the αMnO2.53 However, a reverse activity was observed in our α-MnO2 and γ-MnO2 catalyst. A possible explanation is that our α-MnO2 is actually cryptomelane-M (K2-xMn8O16) whereby some of the OH-/H2O have interaction with the K+ ions in the channel hence leading to a decreased OH-/H2O amount that can facilitate the CO chemisorption. A more detailed study of the amount of CO adsorption on the MnO2 nanorods with different phase structures via CO- and CO2- temperature programmed desorption (TPD) analysis may be necessary to study the changes going on during the reaction for us to gain more insight. 224 Chapter 6 Although there is a potential for nanocrystalline MnO2 to be an active catalyst, the issue of its stability is another challenge. When attempting to repeat the activity measurements, it was found that the initial high activity was difficult to be reproduced for both oxides. In fact, it was decreasing with time (e.g. Run 2 in Figure 6.37). After eliminating possible experimental errors for examples faulty equipment or gas leakage, it was deduced that the nanocrystalline oxides are probably not stable. When the samples were left unprotected over a prolonged period of time, changes to its morphology, surface area or poisoning of active sites could have occurred. BET surface area analysis gave a value of about 80 m2/g, suggesting that the surface area did not decrease drastically in the aggregated form. However, TEM and XRD analysis discussed above for the γMnO2 sample under prolong storage suggested aggregation and an increased Pr value under prolong storage. This suggested that some of the ramsdellite 1x2 (R) which has a larger channel dimension might have collapsed to the smaller 1 x 1 pyrolusite(r) (Figure 6.32). It has been previously shown that the pure pyrolusite β-MnO2 has the lowest activity for CO oxidation due to its smallest channel structure and easier breaking down of its channel structure during the CO oxidation process.53 Thus, in our samples with aging or drying the decreased activity might be attributed to the greater formation of the less catalytic pyrolusite phase in our γ-MnO2. However, we cannot exclude the possibility of poisoning of active sites by, for example, water vapor, sulfur, chlorine and carbon.62,63 Surface sensitive techniques such as Auger electron spectroscopy or TPD may be required to determine qualitatively the elements present and quantitatively the coverage of 225 Chapter 6 the poisons. Although such impurities may be rid off by oxygen during the reaction, they will still impede the catalytic CO oxidation reaction. 6.4 Catalytic studies of nanocrystalline Mn2O3 Nano-sized Mn2O3 and Mn3O4 supported on mesoporous silica SBA-15 have shown promising results as catalysts for removing carbon monoxide.64 Nanostructured α-Mn2O3 has also been deposited with gold nanoparticles and showed high activities for low temperature CO oxidation in the absence or presence of excess H2 due to its unique redox properties and ability to activate the reactant O2 leading to the generation of highly reactive surface oxygen species.65 Mn2O3 was previously prepared using chemical methods which oxidize MnCl2 or reduce KMnO4.66,67 However, the resulting phase is γ-Mn2O3, which is difficult to characterize as its XRD pattern is rather similar to that of Mn3O4. Other methods involve gas condensation and γ-radiation which require expensive equipment. 68,69 A more distinct phase, α-Mn2O3, can be prepared via thermal decomposition of manganese oxalate, which is obtained via reverse micellar method or by decomposition of manganese carbonate.70,71 In this section, Mn2O3 nanocrystals were prepared by the thermal decomposition of a manganese oxalate precursor.70 Our as prepared α-Mn2O3 (JCPDS: 41-1442) was identified from its XRD pattern as shown in Figure 6.38. A closer look at the 226 Chapter 6 XRD pattern, however, revealed some weak peaks assignable to Mn3O4 which could be present as an intermediate during the preparation from manganese oxalate. After pre-treatment, and after reaction, the XRD patterns showed clearly that the catalysts do not convert to other phases. Moreover, the peaks due to Mn3O4 were also largely diminished. FT-IR results (Table 6.5) also reaffirmed the identity of α-Mn2O3 as they are very close to the literature values.31 Table 6.5 FT-IR results of nanocrystalline Mn2O3 and reference Mn2O3.31 MnOx α-Mn2O3 Reference Mn2O3 525 (s, sp) 523 (vs, sp) Wavenumber / cm-1 576 606 (m, sp) (m, sp) 574 599 (vs, sp) (s, sh) 667 (m, sp) 666 (s) Figure 6.38. XRD patterns of the α-Mn2O3 nanocatalyst as prepared, after heat treatment and after catalytic reaction. For comparison, simulated patterns from the database are shown as vertical lines. 227 Chapter 6 Interesting morphological changes was observed for α-Mn2O3. After heating the prepared manganese oxalate shown in Figure 6.39a, the rod-like precursor was generally maintained in the resultant product. However, in the process of heating, incomplete cleavages were observed within the rods. This resulted in some interlinked nanoparticles with diameters of approximately 40 nm (Figure 6.39b). The SEM images of such morphology showed flat pieces of rectangular rods made up of interlinked particles (Figure 6.39d). However, there were also nanoparticles which were not rod-like. We believe that such morphology is somewhat porous and should have a relatively high surface area. The TEM image of the oxide after catalytic reaction showed serious aggregation of the particles (Figure 6.39c). Although some of the interlinked rod-like morphologies were still observed, they were of the minority. Most of that observed are big aggregates with no distinct shape. This is an undesirable observation, as it implies that the catalyst may not be reusable. 228 Chapter 6 Figure 6.39. TEM images of a) manganese oxalate b) as prepared Mn2O3 and c) Mn2O3 after reaction and d) SEM image of as prepared α-Mn2O3. From the BET isotherm of α-Mn2O3 after reaction (Figure 6.40a), hysteresis was again observed, implying a porous material was obtained. The surface area measured was 13 m2/g. This is higher than that of bulk Mn2O3, which was determined to be 4.0 m2/g. The TGA profile of α-Mn2O3 (Figure 6.40b) showed only a slight change in weight loss. This loss can be attributed mainly to moisture loss from the sample. 229 Chapter 6 Figure 6.40. (a) BET plot of nanocrystalline α-Mn2O3 after pretreatment , showing the adsorption ( ) and desorption ({) of N2 molecules. (b). TGA profile of α-Mn2O3 at 50°C in N2 The catalytic behavior of α-Mn2O3 towards CO oxidation was studied similarly and shown in Figure 6.41. Although the conversion is considered quite high under mild conditions (≤ 300°C), it is still lower as compared to the other three nanocrystalline oxides as described above. One of the reasons could be the serious sintering that had occurred as observed from TEM analysis (Figure 6.39). Another reason could be that the sample was impure, as XRD analysis has suggested the presence of Mn3O4 impurities. As mentioned, Mn3O4 is the more stable form as compared to α-Mn2O3. Hence, in order to achieve equilibrium and reduce free 230 Chapter 6 surface energy, it is very likely that this Mn3O4 resides on the surface of the αMn2O3 through diffusion processes. Although the activity of Mn3O4 is also quite high as we saw in Section 6.2, it is difficult to conclude the effect of the mixed oxides and the state of the catalyst’s surface structure. Figure 6.41. Catalytic activities of bulk Mn2O3 (), nanocrystalline α-Mn2O3 run 1 (z) and run 2 (S). Besides the Mars-van Krevelen mechanism, another widely accepted mechanism of transition metal oxide catalysts is the Langmuir-Hinshelwood mechanism. Reaction is assumed to occur between adsorbed gaseous reactants, in this case, chemisorbed CO and O2. When adsorbed activated complexes are in close proximity, CO2 will be formed and desorb from the surface. Therefore, surface structure is very important if this mechanism takes place. With a possibly mixed oxide on the surface of the catalyst and adsorption sites which in reality have 231 Chapter 6 different adsorption energies for different species, the scenario becomes more complicated. Previous studies on manganese oxides, such as the one by Foley et al. on CO hydrogenation72 and Vannice et al. on N2O decomposition73 usually reported Mn2O3 as more active catalyst when compared to MnO or MnO2. Hence a pure sample of α-Mn2O3 when obtained may show higher conversions. As the pre-treatment temperature can most likely affect its catalytic activity as well, a higher pre-treatment temperature may be necessary to induce a higher activity.74 Furthermore, pre-treating at a higher temperature helps to purge the surface moisture, which could act as a poison to the catalytic reaction. Hence, it could be concluded that being a nanocrystalline oxide with a much higher surface area does not necessarily imply a better catalytic activity. 6.5 Comparison of the catalytic activities of different manganese oxides Thus, nanocrystalline MnO, α-MnO2, γ-MnO2, α-Mn2O3 and Mn3O4 were successfully prepared via the various chemical routes and characterized using XRD, TEM, TGA, FT-IR and BET as discussed above. The crystal growth of MnO and Mn3O4 were studied in greater detail in Sections 6.1 and 6.2 to reveal the importance of different factors such as temperature, ligand concentration, monomer concentration and time of reaction on the synthesis. 232 Chapter 6 The catalytic behaviors of all the prepared manganese oxides towards CO oxidation were studied and the light-of temperatures are summaried in Table 6.6. Some of these nanocrystaliine oxides exhibit lower light-of temperatures as compared to their corresponding bulk counterparts. In most cases, nevertheless, higher conversion is achieved using the nanocrystalline catalysts. Table 6.6. Light-off temperatures of the nanocrystalline and bulk MnOx MnOx Light-off Temperature (°C) Nano Bulk MnO 180 / 250 240 α-MnO2 169 - ρ-MnO2 105 - α-Mn2O3 232 180 Mn3O4 234 / 256 230 In addition to conversion percentages of the nanocatalysts, it is also important to study their chemical kinetics. Thus we attempted to estimate the reaction rates and apparent activation energies of the nanocatalysts. Figure 6.42 and 6.43 show the Arrhenius plots of the nanocrystalline and bulk MnOx respectively, relating reaction rate and the inverse of temperature. From these plots, apparent activation energies can be calculated and the results are summarized in Table 6.7. 233 Chapter 6 Figure 6.42. Reaction rates of nanocrystalline MnO pre-treated at 300°C (), αMn2O3 (T) and Mn3O4 pre-treated at 300°C (z), α-MnO2 ( ) and ρ-MnO2 (V). Figure 6.43. Reaction rates of bulk MnO (), Mn2O3 (z), MnO2 (S) and Mn3O4 (T). 234 Chapter 6 Since the CO oxidation reaction is an exothermic process, there may be several hot spots within the catalyst bed whereby intrinsic heat evolved cannot be dissipated, leading to a temperature rise in the catalyst bed. This implies that the actual temperature of the high activity could be higher than observed. Hence, for catalysts which showed conversions of higher than 20%, they were diluted with alumina so that the reaction rates can be calculated more accurately. Diluting to 20% conversion or less can minimize the errors involved with hot spots. In addition, it usually allows the intrinsic activity of the catalyst to emerge. Table 6.7 Apparent activation energies of nanocrystalline and bulk MnOx Apparent activation energies of nanocrystalline and bulk MnOx. MnOx MnO α-MnO2 ρ-MnO2 α-Mn2O3 Mn3O4 Apparent activation energy (kJ/mol) Nano Bulk 85.5 ± 7.3 128.1 ± 3.1 Bulk β-MnO2 51.7 ± 9.1 42.9 ± 0.0 48.0 ± 0.3 83.2 ± 0.3 54.0 ± 0.0 123.1 ± 0.3 108.0 ± 0.5 Several interacting factors come into play for the Arrhenius rate equation. Reaction rate is related to a pre-exponential constant, activation energy and temperature. In this study, it is also possible to see it as directly related to the conversion percentages. The pre-exponential constant itself encompasses various determining factors, such as active site concentration and collision frequency. The activation energy determines the ease for the reaction to occur. A higher activation energy implies increasing difficulty for reaction to occur. As such, it 235 Chapter 6 can be concluded that having a higher reaction rate or conversion does not imply lower activation energy. Hence, from the Arrhenius plots, a clearer view on the performances of the various nanocrystalline MnOx can be compared. In this study, the general trend of catalytic activity is: ρ-MnO2 > α-MnO2 > Mn3O4 > α-Mn2O3 > MnO. As shown in Table 6.7, Mn3O4 which has a moderate reaction rate has the highest activation energy of 123 kJ/mol. The level of activation energies probably depends on the different reaction route taken by the reactants during catalysis. There are in fact mainly three mechanisms proposed for the catalytic CO oxidation reaction, namely the Eley-Rideal, Langmuir-Hinshelwood and the Mars-van Krevelen mechanism. While we do not know which the main governing mechanism for nanocrystalline MnOx is, it is possible a combination of these mechanisms could be operative. Previously, it was suggested that below 523 K and with unit ratio of O2/CO, CO oxidation proceeds via the Langmuir–Hinshelwood mechanism on bulk Mn2O3 and MnO2 catalysts while Mars-van-Krevelen mechanism is primarily responsible for CO2 formation on the MnO.36 As a general rule of thumb, if the activation energies of oxides in comparison do not differ by an order of two, they are likely to proceed by a similar catalytic mechanism. Hence, since the activation energy of both the bulk and nanosized MnOx do not differ significantly, we expect the nanosized MnOx undergo similar mechanism as their corresponding bulk forms. 236 Chapter 6 In summary, under the same reaction conditions, some of the nanocrystalline MnOx such as MnO and MnO2 indeed showed higher catalytic activity as compared to their bulk counterparts. The higher surface area of nanocrystalline oxides can probably play a role in increasing their catalytic activities. However, this may not always occur, due to the competing factor of higher absorption strength of reactant gas molecules on the nanocatalysts surfaces. In addition, other surface structural differences due to the different phase and morphology present are likely to play equally important roles. Nanosized Mn2O3 and Mn3O4, on the other hand, although active under mild conditions (≤ 300°C), did not perform as well as expected when compared to their bulk counterparts. Although nanosized MnOx are promising catalysts in general, their stability with time remains a problem. Different MnOx also showed different stability at the range of temperatures for the catalytic testing. MnO was oxidized to Mn3O4, ρ-MnO2 was found to transform to a mixture of different MnOx while α-MnO2, Mn3O4 and Mn2O3 were stable. This would call for better experimental control and studies on their reproducibility. 6.6 References (1) Lee, S. J.; Gavriilidis, A.; Pankhurst, Q. A.; Kyek, A.; Wagner, F. E.; Wong, P. C. L.; Yeung, K. L. J. Catal. 2001, 200, 298. (2) Alvarez-Galvan, M. C.; Pawelec, B.; O'Shea, V. A. D.; Fierro, J. L. G.; Arias, P. L. Appl.Catal., B 2004, 51, 83. (3) Shanmugam, S.; Gedanken, A. J. Phys. Chem. B 2006, 110, 24486. 237 Chapter 6 (4) Lee, G. H.; Huh, S. H.; Jeong, J. W.; Choi, B. J.; Kim, S. H.; Ri, H. C. J. Am. Chem. Soc. 2002, 124, 12094. (5) Zhang, H. T.; Chen, X. H. J. Phys. Chem. B 2006, 110, 9442. (6) Masala, O.; Seshadri, R. J. Am. Chem. Soc. 2005, 127, 9354. (7) Kang, S. S.; Miao, G. X.; Shi, S.; Jia, Z.; Nikles, D. E.; Harrell, J. W. J. Am. Chem. Soc. 2006, 128, 1042. (8) Berkowitz, A. E.; Rodriguez, G. F.; Hong, J. I.; An, K.; Hyeon, T.; Agarwal, N.; Smith, D. J.; Fullerton, E. E. “Monodispersed MnO nanoparticles with epitaxial Mn3O4 shells”; 6th International Conference on Fine Particle Magnetism, 2007, Rome, ITALY. (9) Berkowitz, A. E.; Rodriguez, G. F.; Hong, J. I.; An, K.; Hyeon, T.; Agarwal, N.; Smith, D. J.; Fullerton, E. E. Phys. Rev. B 2008, 77, 024403. (10) Ghosh, M.; Biswas, K.; Sundaresan, A.; Rao, C. N. R. J. Mater. Chem. 2006, 16, 106. (11) Yin, M.; O'Brien, S. J. Am. Chem. Soc. 2003, 125, 10180. (12) Seo, W. S.; Jo, H. H.; Lee, K.; Kim, B.; Oh, S. J.; Park, J. T. Angew. Chem., Int. Ed. 2004, 43, 1115. (13) Li, Y.; Afzaal, M.; O'Brien, P. J. Mater. Chem. 2006, 16, 2175. (14) Yu, W. W.; Peng, X. G. Angew. Chem., Int. Ed.2002, 41, 2368. (15) Jana, N. R.; Chen, Y. F.; Peng, X. G. Chem. Mater.. 2004, 16, 3931. (16) Park, J.; An, K. J.; Hwang, Y. S.; Park, J. G.; Noh, H. J.; Kim, J. Y.; Park, J. H.; Hwang, N. M.; Hyeon, T. Nature Mater. 2004, 3, 891. (17) Zitoun, D.; Pinna, N.; Frolet, N.; Belin, C. J. Am. Chem. Soc. 2005, 127, 15034. (18) Zhong, X. H.; Xie, R. G.; Sun, L. T.; Lieberwirth, I.; Knoll, W. J. Phys. Chem. B 2006, 110, 2. 238 Chapter 6 (19) Ould-Ely, T.; Prieto-Centurion, D.; Kumar, A.; Guo, W.; Knowles, W. V.; Asokan, S.; Wong, M. S.; Rusakova, I.; Luttge, A.; Whitmire, K. H. Chem. Mat. 2006, 18, 1821. (20) Park, J.; Kang, E. A.; Bae, C. J.; Park, J. G.; Noh, H. J.; Kim, J. Y.; Park, J. H.; Hyeon, T. J. Phys. Chem. B 2004, 108, 13594. (21) Seo, J. W.; Jun, Y. W.; Ko, S. J.; Cheon, J. J. Phys. Chem. B 2005, 109, 5389. (22) Davis, T. S.; Fackler, J. P.; Weeks, M. J. Inorg. Chem 1968, 1994. (23) Anderson, J. M.; Kochi, J. K. J. Am. Chem. Soc 1970, 92, 2450. (24) Redl, F. X.; Black, C. T.; Papaefthymiou, G. C.; Sandstrom, R. L.; Yin, M.; Zeng, H.; Murray, C. B.; O'Brien, S. P. J. Am. Chem. Soc. 2004, 126, 14583. (25) Peng, Z. A.; Peng, X. G. J. Am. Chem. Soc. 2001, 123, 1389. (26) Bullen, C. R.; Mulvaney, P. Nano. Lett. 2004, 4, 2303. (27) Jun, Y. W.; Lee, J. H.; Choi, J. S.; Cheon, J. J. Phys. Chem. B 2005, 109, 14795. (28) Wang, Z. L. J. Phys. Chem. B 2000, 104, 1153. (29) Rusakova, I.; Ould-Ely, T.; Hofmann, C.; Prieto-Centurion, D.; Levin, C. S.; Halas, N. J.; Luttge, A.; Whitmire, K. H. Chem. Mat. 2007, 19, 1369. (30) Ramesh, K.; Chen, L. W.; Chen, F. X.; Liu, Y.; Wang, Z.; Han, Y. F. Catal. Today 2008, 131, 477. (31) Nohman, A. K. H.; Zaki, M. I.; Mansour, S. A. A.; Fahim, R. B.; Kappenstein, C. Thermochim. Acta 1992, 210, 103. (32) Zhang, Z. H.; Zhong, X. H.; Liu, S. H.; Li, D. F.; Han, M. Y. Angew. Chem., Int. Ed. 005, 44, 3466. (33) Perez-Dieste, V.; Castellini, O. M.; Crain, J. N.; Eriksson, M. A.; Kirakosian, A.; Lin, J. L.; McChesney, J. L.; Himpsel, F. J.; Black, C. T.; Murray, C. B. Appl. Phys. Lett. 2003, 83, 5053. (34) Sheldon, R. A.; van Santen, R. A. Catalytic Oxidation: Principles and Applications; World Scientific Publishing, 1995. 239 Chapter 6 (35) Makwana, V. D.; Son, Y. C.; Howell, A. R.; Suib, S. L J. Catal. 2002, 210, 46. (36) Ramesh, K.; Chen, L. W.; Chen, F. X.; Liu, Y.; Wang, Z.; Han, Y. F.“Re-investigating the CO oxidation mechanism over unsupported MnO, Mn2O3 and MnO2 catalysts”; 4th Asia-Pacific Congress on Catalysis, 2006, Singapore, SINGAPORE (37) Rockenberger, J.; Scher, E. C.; Alivisatos, A. P. J. Am. Chem. Soc. 1999, 121, 11595. (38) Sun, X.; Zhang, Y. W.; Si, R.; Yan, C. H. Small 2005, 1, 1081. (39) Jiao, F.; Harrison, A.; Bruce, P. G. Angew. Chem., Int. Ed. 2007, 46, 3946. (40) Wang, W. Z.; Xu, C. K.; Wang, G. H.; Liu, Y. K.; Zheng, C. L. Adv. Mater. 2002, 14, 837. (41) Chang, Y. Q.; Yu, D. P.; Long, Y.; Xu, J.; Luo, X. H.; Ye, R. C. J. Cryst. Growth 2005, 279, 88. (42) Chen, Z. W.; Lai, J. K. L.; Shek, C. H. Appl. Phys. Lett. 2005, 86, 181911. (43) Na, C. W.; Han, D. S.; Kim, D. S.; Park, J.; Jeon, Y. T.; Lee, G.; Jung, M. H. Appl. Phys. Lett. 2005. 87,142504. (44) Du, J.; Gao, Y. Q.; Chai, L. L.; Zou, G. F.; Li, Y.; Qian, Y. T. Nanotechnology 2006, 17, 4923. (45) Ehrhardt, C.; Gjikaj, M.; Brockner, W. Thermochim. Acta 2005, 432, 36. (46) Gallagher, P. K.; Schrey, F.; Prescot, B. Thermochim. Acta 1971, 2, 405. (47) Lee, S. M.; Cho, S. N.; Cheon, J. Adv. Mater. 2003, 15, 441. (48) Si, R.; Zhang, Y. W.; You, L. P.; Yan, C. H. Angew. Chem., Int. Ed. 2005, 44, 3256. (49) Casula, M. F.; Jun, Y. W.; Zaziski, D. J.; Chan, E. M.; Corrias, A.; Alivisatos, A. P. J. Am. Chem. Soc. 2006, 128, 1675. . 240 Chapter 6 (50) Nohman, A. K. H.; Ismail, H. M.; Hussein, G. A. M. J. Anal. Appl. Pyrolysis 1995, 34, 265. (51) Cademartiri, L.; Bertolotti, J.; Sapienza, R.; Wiersma, D. S.; von Freymann, G.; Ozin, G. A. J. Phys. Chem. B 2006, 110, 671. (52) Sinha, A. K.; Suzuki, K.; Takahara, M.; Azuma, H.; Nonaka, T.; Fukumoto, K. Angew. Chem., Int. Ed. 2007, 46, 2891. (53) Liang, S. H.; Bulgan, F. T. G.; Zong, R. L.; Zhu, Y. F. J. Phys. Chem. C 2008, 112, 5307. (54) Li, Z. Q.; Ding, Y.; Xiong, Y. J.; Xie, Y. Cryst. Growth Des. 2005, 5, 1953. (55) Wang, X. Y.; Huang, W. G.; Sebastian, P. J.; Gamboa, S. J. Power Sources 2005, 140, 211. (56) Wang, X.; Li, Y. D. J. Am. Chem. Soc. 2002, 124, 2880. (57) Chen, Y.; Liu, C.; Li, F.; Cheng, H. M. J. Alloys Compd. 2005, 397, 282. (58) Ding, Y. S.; Shen, X. F.; Gomez, S.; Luo, H.; Aindow, M.; Suib, S. L. Adv. Funct. Mater. 2006, 16, 549. Portehault, D.; Cassaignon, S.; Nassif, N.; Baudrin, E.; Jolivet, J. P. Angew. Chem., Int. Ed. 2008, 47, 6441. (59) (60) De Wolff, P. M. Acta Crystallogr. 1959, 12, 341. (61) Hill, L. I.; Verbaere, A. Journal of Solid State Chem. 2004, 177, 4706. (62) Bonzel, H. P.; R., K. J. Chem. Phys 1973, 59,1641. (63) Xia, G. G.; Yin, Y. G.; Willis, W. S.; Wang, J. Y.; Suib, S. L. J. Catal. 1999, 185, 91. (64) Han, Y. F.; Chen, F. X.; Zhong, Z. Y.; Ramesh, K.; Chen, L. W.; Widjaja, E. J. Phys. Chem. B 2006, 110, 24450. (65) Wang, L. C.; Huang, X. S.; Liu, Q.; Liu, Y. M.; Cao, Y.; He, H. Y.; Fan, K. N.; Zhuang, J. H. J. Catal. 2008, 259, 66. (66) Karthikeyan, J.; Berndt, C. C.; Tikkanen, J.; Wang, J. Y.; King, A. H.; Herman, H. Nanostruct. Mater. 1997, 8, 61. 241 Chapter 6 (67) Gui, Z.; Fan, R.; Chen, X. H.; Wu, Y. C. Inorg. Chem. Commun. 2001, 4, 294. (68) Liu, Y. P.; Qian, Y. T.; Zhang, Y. H.; Zhang, M. W.; Chen, Z. Y.; Yang, L.; Wang, C. S.; Chen, Z. W. Mater. Lett. 1996, 28, 357. (69) Dimesso, L.; Heider, L.; Hahn, H. Solid State Ionics 1999, 123, 39. (70) Ahmad, T.; Ramanujachary, K. V.; Lofland, S. E.; Ganguli, A. K. J. Mater. Chem. 2004, 14, 3406. (71) Lei, S. J.; Tang, K. B.; Fang, Z.; Liu, Q. C.; Zheng, H. G. Mater. Lett. 2006, 60, 53. (72) Te, M.; Foley, H. C. Appl. Catal. A-Gen. 1994, 119, 97. (73) Yamashita, T.; Vannice, A. J. Catal. 1996, 161, 254. (74) Gardner, S. D.; Hoflund, G. B.; Upchurch, B. T.; Schryer, D. R.; Kielin, E. J.; Schryer, J. J. Catal. 1991, 129, 114. 242 Chapter 7 Chapter 7 Conclusions and Outlook In Chapter 3 and 4, we explored the optical and NLO properties of PbS QDs and core-shell PbS/CdS QDs in both hexane and PS films. Core-shell PbS/CdS QDs were produced from a cationic exchange of the pre-prepared PbS QDs at moderate reaction temperature. A series of core-shell QDs were formed from using different PbS core sizes and different cationic exchange time. Broadening of the XRD peaks, blue shifts of the absorption and PL peaks, and the same overall sizes of QDs before and after cationic exchange from TEM measurements suggested that the composite core-shell PbS/CdS QDs were formed via sacrificial replacement of surface PbS QDs. Different methods were explored in the formation of PS thin films uniformly embedded with the QDs. An imprinting-thermal cross-linking method was developed, which can be used to prepare large areas of polymer/QDs composite thin films on glass slides. Except a slight broadening of the peaks, the absorption and PL properties of the QDs were preserved in these films. The nonlinear properties of these QD-films could also be readily studied. In this work, we derived the following conclusions from the NLO Z-scan studies of the QDs: 243 Chapter 7 (1) FCA increased with a decrease in the particle size in both PbS and PbS/CdS QDs and in both hexane and polymer films. The core-shell PbS/CdS was found to have larger FCA cross-section compared to PbS of comparable core sizes. (2) NLS dominated the optical limiting in hexane solvent at high laser intensity. (3) Negative nonlinear refractions indicating self-defocusing effects were observed in all QDs. Nonlinear refractive index (n2) and FCR cross-section (σr) of the PbS QDs increased with a decrease in the particle size. No specific trend can be clearly observed for the n2 and σr of the core-shell QDs with the changes in its size. No theoretical explanation for this occurrence can be given at the moment. (4) The presence of excess surface ligand led to enhancement in the nonlinear scattering and optical limiting performance of the core-shell PbS/CdS QDs. We have also made the following conclusions based on our femntosecond pumpprobe transient absorption studies of the QDs: (1) Bleaching of the first exciton was observed when the pump and probe wavelengths were close to the first exciton peak in the core-shell QD; while excited state absorption became dominant when the pump and probe wavelength was higher than the second exciton peak of the QD. 244 Chapter 7 (2) Excited state absorption was retained at high pump intensity and at long delay time for core-shell QDs but became bleaching for pure PbS QDs at high pump intensity and longer probe delay time. The difference could be attributed to the strong interaction with the CdS shell. In the core-shell system, we suggested that an electron transfer from the PbS core to the CdS shell, and further relaxation and recombination to the ground state through states that bypass the PbS first exciton state has occurred. In the PbS system, the carriers have relaxed to the PbS first exciton state at long probe delay time and thus at high pump intensity saturation of the states led to the observed bleaching. (3) Relaxation lifetimes of both the fast and slow components are revealed to be size dependent with a decrease in lifetime with decreasing size. (4) Relaxation lifetimes of PbS and core-shell PbS/CdS with increasing pump intensity revealed that both the fast and slow component shorten with an increase in pump intensity due to Auger recombination or exciton-exciton annihilation of bi-excitons. The cationic exchange method of forming PbS/CdS core-shell QDs might be extended to the synthesis of other interesting cationic exchanged materials using the appropriate shell precursors. The explorations of the PbS/CdS core-shell QDs for applications like photovoltaic and photoconductivity devices were not studied 245 Chapter 7 in the literature. It would be interesting to see if the presence of shell would impede or improve their overall performances. In Chapter 5, PbS nanowires were successfully fabricated by both potentiostatic and cyclic electrochemical deposition onto AAO template. Using pore-widening technique coupled with either CV or constant potential deposition, well-aligned arrays of core-shell metal/PbS nanowires were prepared readily. Deposition of copper as core wires led to the formation of composite nanowires containing an intermediate layer of copper sulfide. Conductive AFM analysis on the PbS nanowires gave an average resistivity of 2.3±0.5 Ωcm. The nanowires synthesized at present are not as useful by itself. Thus, for future studies, the synthesis of heterostructures that might be more useful for applications such as photovoltaic cells or photodetectors, e.g. the growth of n-type CdS on the p-type PbS nanorods by electrodepostion or chemical methods are desirable. Other heterostructures that could be interesting is the growth of PbS thin films on indium tin oxide (ITO) nano pillars with aluminum as the final metallic contact since recent studies showed that PbS nanocrystals formed efficient Schottky photovoltaic devices with aluminum and indium tin oxide contacts.1 In Chapter 6, the crystal growth of MnO and Mn3O4 were studied in details revealing the importance of factors such as temperature, ligand concentration, monomer concentration and time of reaction. In particular, we were able to 246 Chapter 7 synthesize nanocrystals with shapes ranging from spherical, octahedron, long nanorods, tetrapods to rice shaped nanocrystals. The catalytic studies of the different MnOx, which include temperature screening, intrinsic reaction rates and apparent activation energies estimation, suggested the following conclusions: (1) Surface area did not play the most important role in achieving high catalytic activity especially for nanocatalysts due to the competing factor of higher absorption strength of reactant/product gas molecules onto the surface of the nanocatalysts. (2) The phase and structure of the catalysts played an important role in affecting the catalytic activity. For example, a reverse activity was observed in the α-MnO2 and γ-MnO2 catalysts studied in Sections 6.3. The decreased activity can be attributed to the presence of K+ ions in the channels of the α-MnO2 cryptomelaneM (K2-xMn8O16) phase which led to a decreased amount of OH-/H2O that facilitated the CO chemisorption. In another example, decreased activity of the γMnO2 nanowires that were stored for a prolonged period of time can be attributed to the collapse of the tunnel structures in the γ-MnO2. In order to gain more insights into the process affecting the amount of CO adsorption on the MnO2 nanorods, CO and CO2-TPD analysis should be carried out with different phase structures. 247 Chapter 7 (3) Impurities phases, especially if they reside on the surface, might dominate the activity of the catalysts. In our Mn2O3 system, we observed that the presence of small amount of Mn3O4 drastically reduce the activity of the nanocatalysts. Looking forward, further integration of our as synthesized materials with other new materials would be of interest for new applications. For example, new bifunctional nanomaterials can be formed from a magnetic oxide and a fluorescent semiconductor. Structured materials such as nanowires can be decorated with quantum dots to study their electronic interaction and feasibility for applications such as catalysis and photovoltaic cells. 7.1 References (1) Johnston, K. W.; Pattantyus-Abraham, A. G.; Clifford, J. P.; Myrskog, S. H.; MacNeil, D. D.; Levina, L.; Sargent, E. H. Appl. Phys. Lett. 2008, 92, 151115 248 Integrated PL intensity (arb units) Appendix A Integrated PL intensity versus absorbance for (a) 5.0nm PbS and (b) IR125 dye. Gradx(5nm PbS)=2.53E8 1.8x10 7 9.0x10 6 (a) GradST(IR125)=1.73E7 0.0 (b) 0.00 0.05 0.10 Absorbance Integrated PL intensity (arb units) Appendix B Integrated PL intensity versus absorbance for (a) 5.0nm PbS/CdS and (b) 6.0nm PbS/CdS and (c) IR125 dye. 7 3.0x10 7 2.8x10 7 2.6x10 7 2.4x10 7 2.2x10 7 2.0x10 7 1.8x10 7 1.6x10 7 1.4x10 7 1.2x10 7 1.0x10 6 8.0x10 6 6.0x10 6 4.0x10 6 2.0x10 0.0 Gradx(5nm PbS/CdS)=2.9E8 (a) (b) Gradx(6nm PbS/CdS)=2.3E8 GradST(IR125 dye)=4.1E7 (c) 0.00 0.02 0.04 0.06 0.08 0.10 Absorbance 249 [...]... (e-f) 7 nm PbS and 7 nm PbS/CdS 111 4.2 XRD patterns of (a) 5 nm PbS and 5 nm PbS/CdS; (b) 6 nm PbS and 6 nm PbS/CdS; (c) 7 nm PbS and 7 nm PbS/CdS Standard patterns of cubic PbS and CdS (JCPDS 5-0592 and JCPDS 800019) are marked for comparison 113 4.3 (a) Absorption and (b) PL spectra of 5 nm PbS QDs (black line) and core-shell 5 nm PbS/CdS QDs with 1hrs (red line) or 20hrs of (blue line) of cationic... FT-IR results of nanocrystalline Mn2O3 and reference Mn2O3 227 6.6 Light-off temperatures of the nanocrystalline and bulk MnOx 233 6.7 Apparent activation energies of nanocrystalline and bulk MnOx 235 Apparent activation energies of nanocrystalline and bulk MnOx xiii List of Figures 1.1 Sketch of the solubility product [Cd][Se] as a function of temperature 2 Solid line: thermodynamic curve for the equilibrium... structure and stability of the MnOx x List of Publications Size-Dependent Optical Nonlinearities and Scattering Properties of PbS Nanoparticles , M.S Neo, N Venkatram, G.S Li, W.S Chin, and Ji Wei , J Phys Chem C 2009, 113, 19055 Fabrication of PbS and metal/PbS core/shell and composite nanowires, M.S Neo, R.P Rajiv, C.H Sow, and W.S Chin , Submitted for publication Synthesis of PbS/CdS core-shell QDs and. .. Chin, and Ji Wei, manuscript in preparation xi List of Tables 2.1 Chemical and solvents used in the work described in this thesis; their purity and sources 49 2.2 Sizes of core-shell QDs and the corresponding reaction time 58 2.3 Pre-treatment temperatures of nanocatalysts 74 3.1 Absorption and emission wavelength maxima of PbS nanoparticles in 84 solvent and in PbS/PS composite films 3.2 CdS/PMMA bandgap... TEM images of MnO nanoparticles produced at Mn(III) 183 to OA ratio of (a) 1:1 ratio at 280oC for 60 min, (b) 1:3 ratio at 300oC for 60 min, and (c) 1:3 ratio at 320oC for 30 min 6.7 HRTEM image of one faceted MnO particle prepared from Mn(III) to OA ratio of 1:3, 320OC and reaction for 30 min 6.8 (a-b) TEM images of aliquots at Mn(III) to OA ratio 1:6.6 at 320oC 185 withdrawn at (a) 10min and (b) 20min... injection of 0.05g/ml at 190oC and growth at 180OC for for 30mins ; (bc) injection at 160oC and growth at 150OC of (b) 0.15g/ml for 5mins, (c) 0.05g/ml for 3hrs 6.23 Representative TEM images of (a) Mn3O4 nanorods produced by direct 205 heating to 180oC for 45 min; (b) Mn3O4 particles produced by sequential heating procedure 6.24 HRTEM image of a typical Mn3O4 nanorod prepared at 180OC after 206 45mins of. .. measurements of different sized QD collected at 99 10o 3.20 Closed-aperture Z-scan curves of different sized PbS QDs in (a) hexane and (b) PS film with theoretical fits 102 3.21 Size dependent FCA cross-section of PbS QDsin hexane and in PS film and its linear fits with 15% error bars 104 4.1 TEM of PbS QDs and the corresponding core-shell QDs: (a-b) 5 nm PbS and 5 nm PbS/CdS, (c-d) 6nm PbS and 6 nm PbS/CdS, and. .. (n2), and FCR cross-section (σr) of the prepared PbS/CdS QDs in PS films 131 4.7 Intensity of the pump; fast and slow lifetime relaxation dynamics of different core-shell PbS/CdS QDs and referenced PbS QDs 139 5.1 Average length of PbS NWs obtained using different total CV 154 deposition time 5.2 Relative elemental ratio of Pb, Cu and S detected from EDX analysis 158 on areas I and II in Figure 5.6(a) and. .. without CV scanning and (b) with CV scanning 5.10 Evolution of CV scans for Cu core wires immersing in a pH 5 solution 162 containing 0.1 M Na2EDTA and 0.01M Na2S at 100 mV/s 5.11 (a) and (b) Representative SEM images of arrays of Cu/PbS core-shell NWs and enlarged view showing the baseball bat structure (c) TEM and SAED images of a single broken wire (d) XRD pattern showing clearly Cu and PbS diffraction... results for pure PbS obtained in Chapter 3 We also studied the effect of free surface ligands and thickness of the polymer film on their optical limiting properties The influence of the excitation intensity, pump wavelength and probe delay time on the transient differential transmittance spectra and relaxation kinetics of the PbS and core-shell PbS/CdS QDs were studied in details Chapter 5 presents the formation