FABRICATION OF BINARY QUANTUM SOLIDS FROM COLLOIDAL SEMICONDUCTOR QUANTUM DOTS. Nicholas Edward Schmall. A Thesis

Transcription

1 FABRICATION OF BINARY QUANTUM SOLIDS FROM COLLOIDAL SEMICONDUCTOR QUANTUM DOTS Nicholas Edward Schmall A Thesis Submitted to the Graduate College of Bowling Green State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE August 2009 Committee: Mikhail Zamkov, Advisor Robert Boughton Eric Mandell

2 ii ABSTRACT Mikhail Zamkov, Advisor In this thesis I report on an attempted colloidal synthesis of heterostructured quantum solids comprising of a staggered heterojunction of nearly lattice matched cadmium sulfide and zinc selenide semiconductor quantum dots. I present compelling evidence of photoinduced charge separation between zinc selenide and cadmium sulfide domains, via absorption and photoluminescence spectra, but can not provide conclusive evidence via transmission electron microscopy of the merging of the quantum dots. Also in this thesis I report on a colloidal synthesis of lead selenide, titanium dioxide heterostructures, comprising of small diameter lead selenide nanocrystals, grown onto the surface of titanium dioxide nanorods. The deposition of lead sulfide on titanium dioxide proceeds via formation of sub-2 nm lead selenide islands that can be controllably grown to 5 nm by introducing secondary precursor injections. Evidence of the formation of lead selenide nanocrystal islands on the titanium dioxide rods was determined via the acquisition of transmission electron microscopy images that confirm the statistically distributed formation of lead selenide islands.

3 To my parents, whose love and support through the years made this possible. iii

4 iv ACKNOWLEDGMENTS I would like to sincerely thank my advisor, Dr. Mikhail Zamkov, Department of Physics and Astronomy, Bowling Green State University. I appreciate his willingness to share his knowledge and insight that allowed me to perform my research. His support and understanding made this accomplishment in my life a reality and for that I thank him. I am also grateful for the academic staff of the Department of Physics and Astronomy in developing my understanding and passion for physics. My thanks goes to Dr. Bruno Ullrich for allowing me to carry on some experiments in his laboratory and to Dr. Eric Mandell for his willingness to share his expertise and advice throughout my project. Furthermore I would like to thank my fellow graduate students especially Edward Mandere, Nishshanka Hewa- Kasakarge, Liu Chen and Yinghua Zhang for providing me with assistance and guidance in my research and academic pursuits. I appreciated your willingness to help me whenever I asked. I sincerely thank Sandy Gardner and Diana Tussing for their friendliness and help throughout the years. The financial support for my research offered by the Department of Physics and Astronomy and the Department of Photochemical Science is gratefully appreciated. Finally, I would like to extend heartiest gratitude to my parents, my family and my friends for their endless encouragement and support. This achievement in my life was made possible by all of you, and for that I can not thank you enough.

5 v TABLE OF CONTENTS Page CHAPTER 1. INTRODUCTION Background Aim of my Study CHAPTER 2. FIRST CADMIUM SULFIDE, ZINC SELENIDE PROCEDURE Procedure Data Collection Data Analysis CHAPTER 3. TWICE AS MUCH ZINC SELENIDE PROCEDURE Procedure Data Collection Data Analysis TEM Analysis Conclusion CHAPTER 4. HIGH QUALITY ZINC SELENIDE PROCEDURE Introduction Procedure Analysis Conclusion CHAPTER 5. LEAD SELENIDE ISLANDS ON TITANIUM DIOXIDE RODS Introduction Procedure TEM Images TEM Analysis Conclusion

6 REFERENCES vi

7 vii LIST OF FIGURES Figure Page 1.1 Band Gap in a Quantum Dot[6] Band Gap Energy Between CdS and ZnSe Quantum Dots Cadmium Sulfide Absorption Spectra Zinc Selenide Absorption Spectra Cadmium Sulfide Zinc Selenide Absorption Spectra Cadmium Sulfide Photoluminescence Spectra Zinc Selenide Photoluminescence Spectra Cadmium Sulfide Zinc Selenide Photoluminesence Spectra First Procedure TEM Cadmium Sulfide Absorption Spectra Zinc Selenide Absorption Spectra Cadmium Sulfide Zinc Selenide Absorption Spectra Cadmium Sulfide Photoluminescence Spectra Zinc Selenide Photoluminescence Spectra Cadmium Sulfide Zinc Selenide Photoluminescence Spectra TEM image of CdSZnSe Quantum Solid Close-up TEM image of a CdSZnSe Quantum Solid TEM image of ZnSe Lead Selenide Titanium Oxide TEM PbSe TiO 2 TEM Close-up of Small Diameter PbSe Islands on TiO 2 Rods TEM Close-up of Large Diameter PbSe Island on TiO 2 TEM

8 1 CHAPTER 1 INTRODUCTION 1.1 Background Colloidal semiconductor nanocrystals are emerging as a promising building block for the development of a new generation of low-cost optoelectronic materials with potential applications in solar cells,[1] lasers,[2] biomedical labels[3] and LEDs.four The ability to incorporate a desired combination of properties within these nanocrystals often depends on the existence of synthetic protocols for the conjoining of two or more semiconductor materials into composite nanoscale objects. To achieve this goal, synthetic efforts in this area have been focused on the development of multifunctional heterostructures with architectures that range from symmetric core/shell geometries to more complex shapes including dot-in-arod,[5] barbells,[7] tetrapods[12] and multi-branched structures.[9] Core/shell quantum dots consist of a crystalline core and a layer of another metal. These two metals form a semiconductor that separate a photoexcited electron and hole in different parts of the quantum dot. The core/shell nanocrystals are made with semiconductor materials with a particular alignment of conduction and valence band edges at the interface, which creates a step like potential favoring the localization of one of the carriers in the core of the quantum dot and the other one in the shell. The resulting charge separation leads to a strong dipole moment, indirect band gap radiative emission and a large offset between the emission and absorption spectral profiles.[6]

9 2 Figure 1.1: Band Gap in a Quantum Dot[6] Quantum dots are made by a wet chemistry process involving the nucleation of a particular semiconductor and the addition of a monolayer of another semiconductor, which forms the core/shell structure. The nucleation of the core is controlled by the temperature and the presence of organic ligands. Organic ligiands are chains of phosphates and other organic molecules that lower the surface energy of the core and retard the bonding of the core to other materials. The change in the total surface energy, γ, that accompanies the overall deposition process is given by, γ = σ 1 (σ 2 + γ 1,2 ). Where σ 1 and σ 2 are the surface energies of the respective materials, and γ 1,2 is the solid-solid interfacial energy, which is related to the strain between the two lattices. The growth mode that will be occurring in a given synthesis will be determined by the relationship between these terms. When the material to be added is characterized by a lower surface energy (σ 2 < σ 1 ) and a good lattice fit ( i.e.γ 1,2 is low) with respect to the core, such that γ > 0, then the deposition can occur layer-by-layer with a uniform shell. When the introduced material posses a high surface energy (σ 1 < σ 2 ) and/or it is significantly lattice-mismatched (i.e.,γ 1,2, is high), such that γ < 0, then its deposition can take place only by the formation of island-like formations.[10]

10 3 The lowering of the temperature stops the nucleation of the semiconductor and the organic ligands attach to the core, preventing aggregation and uncontrolled growth. The ligands are removed with heat when the shell material is to be attached. The shell is able to attach to the core if they have similar crystal structures and lattice parameters so that the two materials experience lattice strain. The presence of the shell material also allows for a lowering of the surface energy of the two materials, making the process of core/shell over-coating advantageous.[6] 1.2 Aim of my Study The aim of my research was the design and testing of new types of quantum dots and the production of new types of quantum solids. The first experiment that was undertaken was the fabrication of a new type of semiconductor. My goal was to create an electron-hole recombination between cadmium sulfide and zinc selenide quantum dots. Ordinarily ZnSe and CdS quantum dots exhibit characteristic absorption and flouresencse. The hope was that I could create a superlattice between the two distinct quantum dots and create a new band gap energy between the two. Figure 1.2: Band Gap Energy Between CdS and ZnSe Quantum Dots

11 4 The other type of quantum solid that was attempted to synthesize was the formation of lead selenide islands on titanium dioxide rods. The theory behind this structure is that it is energetically advantageous for the PbSe to grow as spheres on the TiO 2 rods instead of over-coating the rods due to a significant lattice mismatch between the fcc PbSe and anatase TiO 2 crystal phases. The growth proceeds via Volmer-Weber regime, thus producing small islands of PbSe interspersed on the TiO 2 rods.[12]

12 5 CHAPTER 2 FIRST CADMIUM SULFIDE, ZINC SELENIDE PROCEDURE 2.1 Procedure The CdSZnSe quantum dots were prepared using a wet chemistry, bottom up technique. The first step was the production of the cadmium sulfide quantum dots. The preparation of the cadmium was carried out by mixing 0.1 mmol ( grams) of cadmium oxide (CdO) with 4 ml of 1-octadecene (ODE) and 0.3 grams (0.2 ml) of oleic acid (OA) in a three-neck flask. The three-neck flask was heated to 275 degrees Celsius under argon flow while stirring. The sulfur was prepared by mixing grams of sulfur with 1.5 ml of ODE in a three neck flask and heated to 200 degrees Celsius under argon flow while stirring with a magnetic stirrer. Once the solutions had reached the appropriate temperatures, the sulfur solution was injected into the three-neck flask containing the CdO. The temperature was lowered to 250 degrees Celsius and maintained for five minutes to allow for nanocrystal growth. Samples of the CdS solution were taken at one minute and at four minutes after the injection of the sulfur. The samples were prepared by withdrawing 0.1 ml of the solution and combining it with 5 ml of chloroform. The samples were used to check the photoluminescence of the CdS. The photoluminescence of the samples was checked by shining a ultraviolet light on them and simply observing their fluorescence. The samples fluorescence increasing in intensity and becoming blue is a sign that the synthesis is proceeding as expected. Once the samples indicated successful synthesis, the mixture in the three-neck flask was cooled by raising the flask from the heating mantle. Once the mixture had cooled to sixty degrees Celsius, it was injected with 8 ml of hexane to prevent aggregation. Purification of the nanocrystals was carried out by adding approximately 2 ml of methanol and hexane to the final solution. Any unreacted material would fall into the methanol layer and the upper layer would contain the quantum dots. The top layer was extracted and cleaned one more time using the methanol/hexane technique.

13 6 The top layer of the second cleaning was then centrifuged for 15 minutes at 3500 rpm. The clear layer was then extracted and kept for future use. A 0.1mL sample of the final, cleaned sample was added to a vial containing 5 ml of hexane for the purpose of testing the samples absorption and photoluminescence spectra. A distinct absorption spectra peak near 400nm ensures that the quantum dots were properly prepared. The next step was the preparation of the ZnSe quantum dots. The preparation of the zinc was carried out by mixing 0.6 mmol ( grams) of zinc stearate and 9.5 ml (3.2 grams) of 1-octadecene (ODE) in a three neck flask and heated under argon flow, while stirring to 300 degrees Celsius. The selenium was prepared by placing grams of selenium in a single-neck flask with 1.8 ml of trioclylphohine (TOP) and sonicated under argon flow. Once the zinc solution had reached 300 degrees Celsius the selenium solution was injected. The temperature dropped to 270 degrees Celsius but was raised to 280 degrees and maintained for fifteen minutes to facilitate the growth of the nanocrystals. Samples were taken at four, eight, twelve and fifteen minutes after the selenium was injected. The samples were prepared by taking a 0.1 ml of the solution and combining it with 5 ml of chloroform in a glass vial. After the samples indicated proper synthesis, the solution was cooled to sixty degrees Celsius. Nine ml of hexane were added to the mixture to prevent aggregation of the quantum dots. The samples were cleaned twice by adding equal parts hexane and methanol and then centrifuged. Before merging, the CdS and ZnSe had to be prepared with ligands before synthesis could take place. Six ml of the ZnSe quantum dots in hexane solution was mixed with 0.5 grams of ODA and 1.25 ml of ODE in a three-neck flask and was heated to 220 degrees Celsius for thirty minutes under vacuum to degas the solution. Six ml of the CdS quantum dot solution was mixed with 2 ml of ODE and heated to 100 degrees Celsius under argon flow, while stirring. Synthesis was achieved by adding the CdS mixture in 0.2 ml increments to the ZnSe every three minutes. 0.1 ml samples of the CdSZnSe solution were taken every

14 7 five minutes to check fluorescence and for future testing. After the entire CdS solution was added the CdSZnSe solution was cooled to sixty degrees Celsius and 8 ml of hexane was added to the solution to prevent aggregation. Purification of the nano crystals was carried out by adding approximately 2 ml of methanol and hexane to the final solution. Any unreacted material would fall into the methanol layer and the upper layer would contain the quantum dots. The top layer was extracted and cleaned one more time using the methanol/hexane technique. The top layer of the second cleaning was then centrifuged for 15 minutes at 3500 rpm. The clear layer was then extracted and kept for future use. 2.2 Data Collection The acquisition of the absorption spectra was carried out by using Cary 50 software and a spectrophotometer. During this process approximately 4mL of hexane was placed in quartz vial and then placed in the specrophotometer where a beam of light was directed into the hexane. The software then obtained an absorption spectra of hexane and used this as a baseline. Then approximatly 4 ml of the hexane solution with the quantum dots was placed in the quartz vial and placed in the spectophotometer where an absorption spectra was acquired. This software then compared how much light was absorbed by the hexane and the quantum dots and then corrected the spectra to show only the absorbtion spectra of the quantum dots. On the following graphs the y-axis represents the (T rasmittedlight/absorpedlight) and the x-axis represents the wavelength of the incident light. The photoluminescence spectra was recorded using the chemistry department s FL3-11 fluorescence spectrometer by placing approximately 4 ml of the final solutions into a quartz vial and placing it in the spectrometer. In this process approximatly 4mL of hexane was placed in a quartz vial and a beam of light was directed into the vial, where the software recorded the flourescence of the hexane and used this as a baseline. Then 4mL of the hexane/quantum dot solution was placed in the spectrometer where a beam of light was

15 8 directed into it whereby only the flourescence spectra of the quantum dots was acquired. High-resolution transmission electron microscopy measurements were carried out by depositing a drop of the nanocrystal/hexane solution onto a forvar-coated copper grid and letting it air dry. The samples were then taken to the University of Michigan by my advisor, Dr. Mikhail Zamkov, who acquired the images. 2.3 Data Analysis Confirmation of successfully merging CdS and ZnSe and creating an electron hole recombination between CdS and ZnSe can be confirmed several ways. The first way to confirm successful merging was by confirming that the CdS/ZnSe absorption spectra had been shifted to a higher wavelength than the CdS and ZnSe precursors absorption spectra. Generally there exists a distinct peak around 400 nm in the absorption profile of both CdS and ZnSe. The peak is present for both CdS and ZnSe. These sharp peaks indicate that the synthesis of the quantum dots was successful. As can be seen from the following graphs the CdS/ZnSe quantum solid has an absorption spectra that is shifted more towards the red region then the CdS and the ZnSe spectra. The CdS and ZnSe have peaks slightly higher than 400 nm, while the final solution has an absorption peak slightly higher.[6]

16 9 Figure 2.1: Cadmium Sulfide Absorption Spectra Figure 2.2: Zinc Selenide Absorption Spectra

17 10 Figure 2.3: Cadmium Sulfide Zinc Selenide Absorption Spectra Another sign of successful merging is indicated by the red shifting of the photoluminescence spectra. As can be seen from the following graphs, the photoluminescence of the CdS peaks at around 450 nm and ZnSe quantum dots peaks at around 410 nm. The final sample of CdS/ZnSe flouresces at close to 550 nm, thus giving the most definitive evidence that there exists a new band gap energy. Figure 2.4: Cadmium Sulfide Photoluminescence Spectra

18 11 Figure 2.5: Zinc Selenide Photoluminescence Spectra Figure 2.6: Cadmium Sulfide Zinc Selenide Photoluminesence Spectra

19 12 This shifting of the final sample s fluorescence seems to indicate that the CdS/ZnSe has been red shifted from the precursors and that there exists some type of electron-hole recombination other then just CdS and ZnSe fluorescing individually. The red shifting of the absorption and photoluminescence spectra are indications of the successful merging of the CdS and ZnSe but the only proof would be through a clear transmission electron miscroscopy (TEM) image of the final CdS/ZnSe sample. Figure 2.7: First Procedure TEM The TEM image of the CdS/ZnSe quantum solid was inconclusive. No indication of merging of the CdS and the ZnSe could be inferred from these images. In fact it seems to indicate a great deal of unmerged material. A powder X-ray diffraction pattern was undertaken with a powder sample of CdS/ZnSe using Scintag XDS-2000 X-Ray Powder Diffratometer at the University of Toledo, and the results indicated the presence of unbound

20 13 CdS. This is the material primarily seen in the TEM images. It was thought that the ZnSe was lost in the initial purification of the ZnSe quantum dots. After production any unreacted or densier materials were seperated from the ZnSe quantum dots in the methanol layer. It is assumed that the clumped ZnSe quantum dots were in this layer, which was eventually thrown out. To resolve this problem and to hopefully obtain a clearer TEM picture of the final product it was decided to double the amount of ZnSe quantum dots that were used for the merging.

21 14 CHAPTER 3 TWICE AS MUCH ZINC SELENIDE PROCEDURE 3.1 Procedure As mentioned in the previous chapter the aim of this procedure was to obtain a higher quality TEM image of the CdSZnSe quantum solid. It was hoped that by preparing twice as much ZnSe in the synthesis of ZnSe and CdS that there would be less unbonded CdS quantum dots, which would lead to a clear TEM image of CdS/ZnSe. The previous procedure for the synthesis of the CdS was replicated and the only change was for the ZnSe precursor. In this procedure grams of zinc stearate and 19.0 ml (6.4 grams) of octadecane (ODA) were added to a three-neck flask and then heated under argon flow to 300 degrees Celsius while stirring. The selenium was prepared by adding grams of selenium powder with 3.6 ml of trioclylphohine in a single neck flask and sonicating. The selenium solution was injected into the zinc solution once the zinc solution reached 300 degrees Celsius. Once the selenium was injected the temperature dropped to 270 degrees Celsius and the solution was allowed to recover to 280 degrees Celsius for approximately 15 minutes to allow for nanocrystal growth. Samples were taken at four, eight, twelve and fifteen minutes after the Se was injected. The samples were prepared by withdrawing 0.1 ml of the ZnSe solution and adding it with 5 ml of chloroform in a glass vial. After fifteen minutes the solution was stopped and cooled to 60 degrees Celsius and 18 ml of hexane was added to prevent aggregation of the quantum dots. Purification was carried out using the previous method of 1:1 methanol and hexane two times and then the solution was centrifuged. The final sample was prepared by adding 0.1 ml of the final ZnSe/CdS quantum dot solution to 5 ml of hexane as in the first procedure. The synthesis of ZnSe/CdS was carried out by mixing 12 ml of the ZnSe quantum dot solution with 1.0 g of ODA and 2.5 ml of ODE in a three-neck flask and it was heated to 220 degrees Celsius for thirty minutes under vacuum, in order to degas the solution. The

22 15 degassing was needed to eliminate all oxygen in the solution. Six ml of the CdS solution was mixed with 2 ml of ODE and heated to 100 degrees Celsius under argon flow. The synthesis was carried out by adding 0.2 ml increments of the CdS to the ZnSe solution every five minutes. Samples were taken every five minutes after the first CdS injection. After all the CdS was injected the ZnSe/CdS mixture was cooled to 60 degrees and 8 ml of hexane was added. Again the cleaning of the quantum dots was carried out by the same 1:1 methanol and hexane cleaning procedure. After cleaning the final solution was centrifuged. 3.2 Data Collection Absorption and photoluminescence spectra were acquired using the same procedure as described in section 2.2. Dr. Zamkov acquired the TEM image at the University of Michigan. 3.3 Data Analysis below. The absorption spectra for the final Cds, ZnSe and CdS/ZnSe samples are displayed Figure 3.1: Cadmium Sulfide Absorption Spectra

23 16 Figure 3.2: Zinc Selenide Absorption Spectra Figure 3.3: Cadmium Sulfide Zinc Selenide Absorption Spectra

24 17 These three graphs indicate that the ZnSe/CdS sample has been clearly red shifted from the ZnSe and CdS precursors. This indicates that there is a new band gap energy that has been created between the semiconductors. It does not however indicate that the band gap exists between the ZnSe and CdS quantum dots, just that there exists some new band gap. The graphs of the three sample s photoluminescence spectra are displayed below. Figure 3.4: Cadmium Sulfide Photoluminescence Spectra Figure 3.5: Zinc Selenide Photoluminescence Spectra

25 18 Figure 3.6: Cadmium Sulfide Zinc Selenide Photoluminescence Spectra Again these graphs indicate that the fluorescence has been red shifted and that clearly there exists a new band gap energy that is emitting light at a different wavelength than the precursors. Again this does not confirm that the band gap energy exists between a ZnSe and CdS quantum dots. 3.4 TEM Analysis The TEM images obtained from the final ZnSe/CdS solution are displayed at the end of this section. Once again the TEM image cannot confirm that a new quantum solid between a ZnSe quantum dot and a CdS quantum dot exists. The image does seem to indicate that a new structure has been comprised of heterogeneous materials. The only way to prove that the two types of quantum dots have merged is a clear picture of one sphere merged with another sphere. The TEM does not show a clear image of this and therefore does not provide clear evidence of the bonding of a cadmium sulfide quantum dot with a zinc selenide quantum dot. The lattice spacing of the material are very similar making its hard to determine if the respective materials have merged. It is difficult to determine if this image represents both

26 19 CdS and ZnSe quantum dots or if a new unique formation has occured. It is also unclear if there is any material underneath this particular quantum solid that would indicate the addition of any precursory semiconductors bonded to this structure. There could exist a type of pancaking or over-coating of a semiconducotor underneath this quantum solid that could account for the presence of a new band gap energy. Although there are methods for determining the elemental structure of the quantum solid, such as electron energy loss spectroscopy and energy filtered transmission electron microscopy, these methods were not available for use in this experiment, and thus, no conclusive interpretaion of the structure could be made. It was thought that the problem in the formation of the ZnSe/CdS lay in the formation of the ZnSe due to earlier problems with synthesizing it in the laboratory. To gauge the quality of the zinc selinide precursor a 0.1 ml sample of ZnSe was placed on a TEM grid and taken to University off Michigan for imaging. Figure 3.9 is a TEM image that clearly shows a clumping of ZnSe. It was thought that the reason for this clumping was that ODA was too weak of a ligand and that it was not preventing the binding of the ZnSe to itself. If CdS were introduced to a solution of ZnSe that was significantly aggregated then it would seem to significantly lower the amount of CdS bonded to ZnSe and that could be the reason that for the difficulty in obtaining a clear TEM image. 3.5 Conclusion This procedure clearly indicated that some new type of quantum solid had been created. The red shifting of both the absorption and photoluminescence indicates that a new band gap had been created and that the absorption and photoluminescence spectra were not from individual CdS and ZnSe quantum dots. The absorption and photoluminescence spectra do not indicate that the band gap exists between synthesized CdS and ZnSe. Again the only

27 20 proof of the synthesizing of ZnSe and CdS comes from the TEM image. And again the TEM does not indicate that the synthesis between a CdS quantum dot and a ZnSe quantum dot has taken place. It was imperative for this project that a clear TEM image of the two quantum dots with merged lattice spacing be obtained. The evidence of the clumping of the ZnSe indicated that a new procedure to synthesis ZnSe with stronger ligands was needed.

28 Figure 3.7: TEM image of CdSZnSe Quantum Solid 21

29 Figure 3.8: Close-up TEM image of a CdSZnSe Quantum Solid 22

30 Figure 3.9: TEM image of ZnSe 23

31 Introduction CHAPTER 4 HIGH QUALITY ZINC SELENIDE PROCEDURE The goal of this procedure was to make a high quality zinc selenium quantum dot. The goal was to eliminate a significant amount of the ODA in the production of the ZnSe quantum dots. One of the reasons that the ODA needed to be replaced was that a large amount of unreacted ODA was in the final solution. Despite the amount of cleaning attempted, the ODA was still present and this could have been interferring with obtaining a quality TEm image. Another reason for wanting to eliminate the amount of ODA is that it is a weak ligand and could be affecting the final synthesis of the ZnSe, which could ultimately be affecting the synthesis of the final quantum solid. If a ligand is too weak it can retard the boning of the ZnSe core-shell structure with CdS. The ligands do not stay in place and prevent the aggregation of the ZnSe quantum dots. Once the ZnSe quantum dots have clumped together they are either discarded during the purification process or they are far less likely to bond with the CdS quantum dots. A new procedure was devised that had far less ODA and more of the stronger bonding ligand, ODE. 4.2 Procedure The first step for this procedure was to mix g of zinc stearate with g of ODA and 4 ml of ODE in a three-neck flask under argon flow with a stirrer until it reached 300 degrees Celsius. At the same time in a single neck flask g of selenium is dissolved in 0.25 ml of TBP by sonicating. Once the selenium was dissolved, 0.75 ml of ODE was added. Once the zinc solution reached 300 degrees Celsius the selenium mixture was injected. After two minutes the reaction was stopped and allowed to cool to sixty degrees Celsius. Once the mixture was cooled, eight ml of chloroform was added and then it was centrifuged. Samples were also taken at 30 seconds and 90 seconds after the selenium was injected.

32 Analysis The two samples taken during synthesis did not show any fluorescence. The final sample s absorption was checked but unfortunately it did not show the profile edge or absorption peak that is indicative of correct synthesis. 4.4 Conclusion Many attempts were tried using this method to synthesize so-called high quality zinc selenium but unfortunately all attempts proved unsuccessful. Since synthesizing usable ZnSe seemed so difficult to perfect, it was decided to abandon the project at this time. After many attempts to develop a type II semiconductor out of ZnSe and CdS quantum dots, the conclusion was reached that all available methods had been exhausted and there was not a way to determine without a doubt that synthises, indeed had taken place. Clearly there is evidence of a new band gap being formed between some of the constituent semiconductor, but without clear evidence from a TEM image as to what type of structure has been formed I could not say with any confidence.

33 26 CHAPTER 5 LEAD SELENIDE ISLANDS ON TITANIUM DIOXIDE RODS 5.1 Introduction Titanium dioxide is an important photovoltaic and photocalytic material, which is utilized in dye-sensitized solar cells and hydrogen production is encouraged by its low fabrication costs and minimal environmental hazards. Absorption of solar radiation within TiO 2 generally requires extending its absorption range into the visible and near infrared by introducing an appropriate sensitizer that reduces TiO 2 upon photoexcitation. To date, the most common strategy for the TiO 2 sensitization involves modification of its surface with organic dyes. The use of semiconductor nanocrystals (NC) for this purpose is now being actively explored due to a number of advantages offered by inorganic NCs over organic sensitizers, including resistance to photobleaching and tenability of NC conduction levels. As shown in recent reports, successful modification of TiO 2 with colloidal CdSe,[13] PbS[15] and InAs[14] NCs has lead to heterostructures that exhibit photoinduced charge separation. In these processes, however, deposition of NC onto TiO 2 still relies on organic linkers that are subject to photo-degradation. To avoid this problem several groups have attempted growth of CdS NCs onto TiO 2 films in ionic solutions. While the observation of improved charge transport characteristics in these experiments was encouraging, the quality and size-distribution of fabricated NCs was inferior to those synthesized through colloidal techniques, making it difficult to control relative positions of electron energy levels in a donor-acceptor system. Here is demonstrated a colloidal approach to the synthesis of PbSe/TiO 2 hetero-nanocrystals (HNCs), comprising 2-5 nm PbSe NCs grown on the surface of TiO 2 nanorods (NRs). As a main benefit of colloidal injection techniques, the present approach allows for a controlled adjustment of the dot diameter during synthesis, which is critical for the offset between donor and acceptor conduction band edges. A significant lattice mismatch between fcc PbSe and TiO 2 crystal phases ensures that the growth is characterized by the formation of small

34 27 PbSe NC islands. One benefit associated with such growth is the possibility of several PbSe NCs per single TiO 2 NR, which increases the light absorption cross section. Such growth supports the formation of small diameter PbSe NCs which is critical for the realization of a type II (staggered) heterojunction between PbSe and TiO 2 materials.[11] Figure 5.1: Lead Selenide Titanium Oxide TEM 5.2 Procedure The first step in this procedure was to make the lead and selenium precursors. The lead was prepared by dissolving 0.45 grams of lead oxide with 1.8 ml of oleic acid and 7.45 ml of ODE in a three-neck flask. The flask was heated to 193 degrees under argon and maintained at that temperature for thirty minutes. The solution was cooled to 120 degrees Celsius prior to injection.

35 28 The selenium was prepared by dissolving 0.21 grams of selenium in 2.7 ml of TOP in a single neck flask under argon and sonicating at room temperature. The first step in making the titanium oxide rods was to combine ml of oleic acid and 2.2 ml of oleylamine in a three-neck flask and heating to 120 degrees Celsius under vacuum for thirty minutes in order to degas the solution. After thirty minutes the solution was cooled down to 40 degrees Celsius and switched to argon flow. At this time 0.05 ml of titanium chloride was injected into the flask. The solution was then heated to 290 degrees Celsius and the temperature was held for twenty-five minutes to promote the growth of the TiO 2 rods. After this time period the solution was cooled to 180 degrees and the lead and selenium solutions were injected simultaneously. After the injection the solution was held at 160 degrees or five minutes and then the reaction was stopped by cooling the flask. No samples were taken due to the fact that this particular type of quantum dot is believed to photoluminescence in the ultraviolet range. After the solution cooled to sixty degrees Celsius, 3 ml of anhydrous chloroform was added to prevent aggregation. Then the solution was added to a test tube and enough anhydrous ethanol was added until the mixture became cloudy, approximately three milliliters. This cloudiness indicated that the quantum dots had precipitated out. The solution was then centrifuged for fifteen minutes at which time there was a black and a cloudy layer. The black layer contained unreacted agents and was extracted and disposed of. The cloudy layer was the layer that contained the quantum dots and was diluted with approximately three ml of anhydrous chloroform and kept in a vial that had argon pumped into it. 5.3 TEM Images A 0.1 ml drop of the final solution was placed on a forvar-coated copper grid for TEM images. The TEM images are presented on the following pages.

36 Figure 5.2: PbSe TiO 2 TEM 29

37 Figure 5.3: Close-up of Small Diameter PbSe Islands on TiO 2 Rods TEM 30

38 Figure 5.4: Close-up of Large Diameter PbSe Island on TiO 2 TEM 31

39 TEM Analysis Transmission electron microscopy analysis of PbSe/TiO 2 HNCs reveals a difference between PbSe shapes forming on the surface of TiO 2 NRs as a result of a single and multiple injections of Pb and Se precursors. The initial injection leads to the formation of several small-diameter PbSe sites per NR, with an average diameter of nm. According to the TEM image of a typical PbSe/TiO 2 structure, PbSe dots appear to be uniformly scattered over the entire NR surface and exhibit a moderate dispersion of sizes. A symmetric placement of PbSe dots on the surface of TiO 2 can be explained in terms of fundamental energy requirements on the deposition of secondary material in hetero-epitaxial growth. Spatially isotropic addition of PbSe monomers onto TiO 2 NRs initially results in the formation of a thin PbSe shell. Subsequent lateral expansion of the shell is associated with the mismatchinduced increase of the interfacial energy, which promotes the collapse of the PbSe layer into segregated islands. 5.5 Conclusion This work demonstrates as all inorganic modification of the TiO 2 surface with semiconductor NCs which should lead to improved light conversion efficiencies in photovoltaic applications. The growth of lead NCs is not limited to TiO 2 NRs and can also be adapted to other nanostructured forms of TiO 2 including porous films and nanotubes, whereby introducing a colloidal route to sensitization of TiO 2 surfaces without organic linkers.

40 33 REFERENCES [1] P. V. Kamat, J. Phys. Chem. C 2008, [2] V. I. Klimov, A. A. Mikhailovsky, S. Xu, A. Malko, J. A. Hollingsworth, C. A. Laetherdale, H. J. Eisler, M. G. Bawendi, Science 2000, 290, 314. [3] I. L. Medintz, H. T. Uyeda, E. R. Goldman, H. Mattoussi, Nature Mat. 2005, 4, 435. [4] S. Coe, W. K. Woo, M. Bawendi, V. Bulovic, Nature 2002, 420, 800. [5] N. N. Hewa-Kasakarage, M. Kirsanova, A. Nemchinov, N. Schmall, P. Z. El-Khoury, A. N. Tarnovsky, M. Zamkov, J. Am. Chem. Soc. 2009, 131, [6] A Nemchinov, M Kirsanova, N. N. Hewa-Kasakarage, M. Zamkov, J. Phys. Chem. C 112, , (2008). [7] F. Shieh, A. E. Saunders, B. A. Korgel, J. Phys. Chem. B 2005, 119, [8] L. Carbone, S. Kudera, E. Carlino, W. J. Parak, C. Giannini, R. Cingolani, L. Manna, J. Am. Chem. Soc. 2006, 128, 748. [9] D. J. Milliron, S. M. Hugher, Y. Cui, L. Manna, J. Li, L. Wang, A. P. Alivisatos, Nature 2004, 430, 190. [10] M. Casavola, R. Buonsanti, G. Caputo, P. D. Cozzoli, Eur: J. Inorg. Chem. 2008, [11] B. L. Wehrenberg, P. Guyot-Sionnest, J. Am. Chem. Soc. 2003, [12] Crystal Growth for Beginners: Fundamentals of Nucleation, Crystal Growth, and Epitaxy; Markov, I. V.; World Scientific: Singapore, Malaysia, [13] I. Mora-Sero, J. Bisquert, Th. Dittrich, A. Belaidi, A. S. Susha, A. L. Rogach, J Phys. Chem. C 2007, 111, [14] P. R. Yun, K. Zhu, A. G. Norman, S. Ferrere, A. J. Frank, A. J. Nozik, J. Phys. Chem. B 2006, 110, [15] B. R. Hyun, Y. W. Zhong, A. C. Bartnik, L. Sun, H. D. Abruna, F. W. Wise, J. D. Goodreau, J. R. Mattews, T. M. Leslie, N. F. Borrelli ACS NANO 2008, 2,