The Magnetic and Electrocatalytic Studies of FePt Nanoparticles By Jaemin Kim B.S. Seoul National University, 1999 M.S. Seoul National University, 2003 A Dissertation Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in the Department of Chemistry at Brown University Providence, Rhode Island May 2009 © Copyright 2009 by Jaemin Kim                       iii   This dissertation by Jaemin Kim is accepted in its present form by the Department of Chemistry as satisfying the dissertation requirement for the degree of Doctor of Philosophy. Date_____________________ ___________________________________ Shouheng Sun, Advisor Recommended to the Graduate Council Date_____________________ ___________________________________ Dwight A. Sweigart, Reader Date_____________________ ___________________________________ Eunsuk Kim, Reader Approved by the Graduate Council Date_____________________ ___________________________________ Sheila Bonde, Dean of the Graduate School     iv   VITA Jaemin Kim was born on the 3rd day of November, 1975 in Jinju, Korea. He attended Seoul National University from 1995 to 1999, graduating with a B. Sc. in Chemistry. He was admitted to the same university to earn his M. Sc. in 1999. During his degree course, he joined Korean Army in 2000, and spent 26 months serving his nation. In 2002, after finishing his service, he continued on his path of studying Molecular Dynamics Simulation and obtained M. Sc. in Theoretical Chemistry in 2003. In 2004, he was accepted in the graduate program of the Department of Chemistry at Brown University. He started his research towards a Ph.D. in nanomaterials under the tutelage of Professor Shouheng Sun in the spring of 2005. During his time at the Sun Laboratory of Nanoscale Materials, he co-published 9 papers in the field of nanomagnetism and nanocatalysis, focusing on FePt nanostructures.     v   Publications at BROWN Journals 1. M. Chen, J. Kim, J. P. Liu, H. Fan, S. Sun, “Synthesis of FePt Nanocubes and Their Oriented Self-assembly”, J. Am. Chem. Soc., 2006, 128, 7132. 2. W. Chen, J. Kim, S. Sun, S. Chen, “Electro-oxidation of Formic Acid Catalyzed by FePt Nanoparticles”, Phys. Chem. Chem. Phys., 2006, 8, 2779. 3. W. Chen, J. Kim, S. Sun, S. Chen, “Langmuir-Blodgett Thin Films of Fe20Pt80 Nanoparticles for the Electrocatalytic Oxidation of Formic Acid”, J. Phys. Chem. C, 2007, 111, 13452. 4. W. Chen, J. Kim, S. Sun, S. Chen, “Composition Effects of FePt Alloy Nanoparticles on the Electro-oxidation of Formic Acid”, Langmuir, 2007, 23, 11303. 5. C. Wang, H. Daimon, Y. Lee, J. Kim, S. Sun, “Synthesis of Monodisperse Pt nanocubes and Their Enhanced Catalysis for Oxygen Reduction”, J. Am. Chem. Soc., 2007, 129, 6974. 6. C. Wang, Y. Hou, J. Kim, S. Sun, “A General Strategy for Synthesizing FePt Nanowires and Nanorods”, Angew. Chem. Int. Ed., 2007, 46, 6333. 7. W. Chen, J. Kim, S. Sun, S. Chen, “Electrocatalytic Reduction of Oxygen by FePt Alloy Nanoparticles”, J. Phys. Chem. C, 2008, 112, 3891.     vi   8. J. Kim, C. Rong, Y. Lee, J. P. Liu, S. Sun, “From Core/Shell Structured FePt/Fe3O4/MgO to Ferromagnetic FePt Nanoparticles”, Chem. Mater., 2008, 20, 7242. 9. J. Kim, C. Rong, J. P. Liu, S. Sun, “Dispersible Ferromagnetic FePt Nanoparticles”, Adv. Mater., 2009, 21, 906. 10. J. Kim, S. Sun, “Nanostructure-Controlled Catalysis of FePt Nanoparticles for Oxygen Reduction”, to be submitted. 11. J. Kim, S. Sun, “Fct-FePt Nanoparticles as a Highly Active Catalyst for Methanol Oxidation”, in preparation. 12. J. Kim, D. Li, S. Sun, “Electrocatalytic Activities of FePt Nanowires and Nanorods on Multiwalled Carbon Nanotubes for Oxygen Reduction”, in preparation. Book Chapter Chemical Synthesis, Self-Assembly and Applications of Magnetic Nanoparticles, Annual Reviews of Nano Research, the World Scientific Publishing Co., 2008, submitted.     vii   AKNOWLEDGEMENTS There are so many people I must say ‘Thank You’. First of all, I should mention my advisor, Professor Shouheng Sun. He gladly accepted me into his group, even though I had no previous research experience in “Nano Science”. He has taught me how to come up with ideas regarding research and helped me even with mundane lab work, like the setting up of bench apparatus. His endless passion towards science has made me a better researcher. Professor Sun, you are my role model for the pursuit of my future dream. I could have never accomplished my work if I did not meet so many experts during my Ph.D. course. Professor William Jr. Risen’s keen insight into the nature of science stimulated me to study solid state chemistry, which was essential to understand the structure and morphology change of the crystals. Thank you, Professor Risen. Professor Dwight A. Sweigart’s excellent lecture made me study organometallic chemistry. His humorous and devoted lectures encouraged me to understand transition metal chemistry fully. Thank you, Professor Sweigart. I also should say thank you, Professor J. Ping Liu. He helped me with SQUID measurement in the ONR/MURI project. Thank you, Professor Shaowei Chen - my knowledge of fuel cell catalysis comes from our collaboration. I would like to say thank you, Professor Eunsuk Kim. She gave me lots of advice, about topics on science and beyond. Dr. Sang Bok Kim, your unending knowledge in chemistry was very helpful for the past 10 years. I have no doubt that you will mature into a great scientist and add to the world’s knowledge in chemistry.     viii   I am so lucky to have worked with my hardworking group members. Dr. Jin Xie and Dr. Sheng Peng, thank you guys. We, as Sun Lab’s first group members, have helped with each other from the very beginning, going back to setting up our lab. I wish you all the best and pray that you have great lives. Thank you, Dr. Yanglong Hou and Dr. Zhichuan Xu. We worked together for the nanomagnetics project and you two were good collaborators in my research. I should say thank you, Dr. Chao Wang and Dr. Chenjie Xu for the productive time spent on research and discussion. Thank you, Tetsunori Koda. Your wide knowledge of magnetism helped me a lot. I also would like to say thank you, Youngmin Lee and Vismadeb Mazumder. They helped me a lot in my electrocatalytic research. Yi Liu, Don Ho, Dongguo Li, Xiaolian Sun, Ryan Chan, Max Mankin: thank you guys and I wish you the best of luck in your research. Dr. Nathan Kohler, Dr. Kai Cheng, Dr. Natalie Huls, Dr. Lise-Mary Lacroix, Dr. U-Hwang Lee, Dr. Hongwang Zhang and Haruo Imagawa,: thank you for your input in my research and for the helpful discussion. I love you, Mom and Dad. Your endless and devoted love for me makes me tide over every difficulty. Till now, I have been trying to obtain a Doctorate of Philosophy in “chemistry”, but now I am departing on a voyage to obtain a Doctorate of Philosophy in “life” from you – a journey that will last the rest of my “life”. You are the best. You are my heroes. I am blessed to have born into a family with such an amazing, brother and sister. You have always been my strong supporters and have managed to cheer me up, no matter the circumstance. I love you, Youngmin and Sehna.     ix   Words cannot express my debt towards Mikyoung, for her love, companionship and unyielding support. I could not have finished all my work without your encouragement during the trying years of the PhD. You never let me get bogged down by the difficulties of research. The little success I have achieved would not have ‘material’ized without you. I promise I will always be with you. I love you, my other half.     x   Dedicated to my family and my soul mate, Mikyoung     xi   Abstract of “The Magnetic and Electrocatalytic Studies of FePt Nanoparticles” by Jaemin Kim, Ph.D., Brown University, May 2009. Surfactants-Mediated Method is adapted in the chemical synthesis of FePt nanoparticles. The size and structure of FePt nanoparticles can be simply tuned by controlling the surfactants/solvent ratio and reaction temperature. These simply-tuned FePt nanostructures show spherical, nanocubes and nanowires/nanorods. As-synthesized near equiatomic FePt nanoparticles show chemically disordered face centered cubic (fcc) structure, which is superparamagnetic. Through high temperature heat treatment, the fcc- structured FePt nanoparticles can be changed into chemically ordered face centered tetragonal (fct) structure, which is ferromagnetic. This fct-structured FePt nanoparticles show high magnetocrystalline anisotropy constant K, reaching up to 107 J/m3 along easy axis [001] direction, that is one of the largest among the known hard magnetic materials. To block sintering problem during annealing process, MgO is coated on fcc-FePt nanoparticles, which is high temperature endurable robust material and easy removable by dilute acid after thermal treatment. Phase changed fct-FePt nanoparticles can be dispersed in hexane solution by aid of hexadecanethiol and oleic acid without severe agglomeration. All these series of progress may offer the fabrication of ordered nanomagnet arrays with controlled magnetic alignment, which is an important goal to achieve high density information storage and high performance permanent magnets. Pt-based metal alloys can be also used as electrocatalysts in fuel cells. They catalyze hydrogen oxidation at the anode and oxygen reduction at the cathode, especially     xii   in Proton Exchange Membrane (PEM) fuel cells. Platinum catalysts without any secondary metal element, however, are very expensive, which hinders the development of large scale fuel cell applications. To reduce the costs of fuel cells and improve the performance, a nanosized FePt electrocatalysts have been prepared and evaluated for anodic and cathodic electron-transfer reactions. FePt nanoparticles as electocatalysts show good performance in both oxygen reduction reactions at cathode and small molecule (methanol and formic acid) oxidation reactions at anode.     xiii   Table of Contents Chapter 1: Introduction of FePt Nanoparticles as a Magnet and a Electrocatalyst ........1 1.1. FePt Nanoparticles as a Magnet ...........................................................................2 1.1.1. Backgrounds ..................................................................................................2 1.1.2. Hard Magnetic FePt Nanoparticles and Their Applications ..........................7 1.1.3. General Chemical Synthesis of FePt Nanoparticles ......................................9 1.2. FePt Nanoparticles as a Electrocatalyst .............................................................13  1.2.1. Backgrounds ................................................................................................13 1.2.2. Electrocatalytic Activity of FePt Nanoparticles ..........................................16 REFERENCES ............................................................................................................18   Chapter 2: Synthesis of FePt Nanocubes and Their Oriented Self-Assembly .................25 2.1. Introduction ........................................................................................................26 2.2. Experimental ......................................................................................................28 2.2.1. Chemicals ....................................................................................................28 2.2.2. Synthesis of 6.9 nm Fe50Pt50 Nanocubes .....................................................28 2.2.3. Composition Control during the Synthesis of FePt Nanocubes ..................29 2.2.4. Self-Assembly of FePt Nanocubes ..............................................................29 2.2.5. Characterization ...........................................................................................30 2.3. Results and Discussion .......................................................................................31     xiv   2.4. Conclusion ..........................................................................................................38  REFERENCES ............................................................................................................39   Chapter 3: A General Strategy for Synthesizing FePt Nanowires and Nanorods .........41 3.1. Introduction ........................................................................................................42 3.2. Experimental ......................................................................................................44 3.2.1. Synthesis ......................................................................................................44 3.2.2. Characterization ...........................................................................................44 3.3. Results and Discussion .......................................................................................46 3.4. Conclusion ..........................................................................................................56  REFERENCES ............................................................................................................57   Chapter 4: Dispersible Ferromagnetic FePt Nanoparticles ...........................................59 4.1. Introduction ........................................................................................................60 4.2. Experimental ......................................................................................................62 4.2.1. Synthesis of 7 nm fcc-Fe51Pt49 Nanoparticles ..............................................62 4.2.2. Synthesis of 7 nm fcc-Fe51Pt49/MgO Nanoparticles .....................................62 4.2.3. Characterization ...........................................................................................63 4.3. Results and Discussion .......................................................................................64 4.4. Conclusion ..........................................................................................................73  REFERENCES ............................................................................................................74     xv   Chapter 5: From Core/Shell Structured FePt/Fe3O4/MgO to Ferromagnetic FePt Nanoparticles ...........................................................77 5.1. Introduction ........................................................................................................78 5.2. Experimental ......................................................................................................80 5.2.1. Chemicals .....................................................................................................80 5.2.2. Synthesis of FePt/Fe3O4 Nanoparticles ........................................................80 5.2.3. Synthesis of FePt/Fe3O4/MgO Nanoparticles ...............................................81 5.2.4. Synthesis of fct-FePt from FePt/Fe3O4/MgO Nanoparticles ........................81  5.2.5. FePt Nanoparticle Characterization .............................................................82 5.3. Results and Discussion .......................................................................................83 5.3.1. Synthesis of FePt/Fe3O4 and FePt/Fe3O4/MgO Nanoparticles .....................83 5.3.2. Reductive Annealing of Pt-Rich fcc-FePt/Fe3O4/MgO Nanoparticles ........87 5.3.3. Structural Analysis on the FePt Nanoparticles ............................................90 5.3.4. Dispersion and Characterization of fct-FePt Nanoparticles ........................94  5.4. Conclusion ..........................................................................................................96  REFERENCES ............................................................................................................97 Chapter 6: Electrocatalytic Reduction of Oxygen by FePt Alloy Nanoparticles ........100 6.1. Introduction ......................................................................................................101 6.2. Experimental ....................................................................................................104 6.2.1. Materials .....................................................................................................104     xvi   6.2.2. Preparation of FexPt100-x Nanoparticles ......................................................104 6.2.3. Electrochemistry ........................................................................................106 6.3. Results and Discussion .....................................................................................108 6.3.1. Cyclic Voltammetry ..................................................................................108 6.3.2. Rotating Disk Votammetry ........................................................................115 6.4. Conclusion ........................................................................................................125  REFERENCES ..........................................................................................................126 Chapter 7: Electro-Oxidation of Formic Acid Catalyzed by FePt Nanoparticles .......131 7.1. Introduction ......................................................................................................132 7.2. Experimental ....................................................................................................135 7.2.1. Materials .....................................................................................................135 7.2.2. Preparation of the Au/FePt Electrode .........................................................138 7.3. Results and Discussion .....................................................................................139 7.3.1. Electrochemical Characterizations of the Electrodes ................................139 7.3.2. Electrocatalytic Activity for Formic Acid Oxidation ................................141 7.3.3. Electrochemical Impedance Studies ..........................................................146 7.4. Conclusion ........................................................................................................157  REFERENCES ..........................................................................................................158     xvii   Chapter 8: Langmuir-Blodgett Thin Films of Fe20Pt80 Nanoparticles for The Electrocatalytic Oxidation of Formic Acid ........................................163 8.1. Introduction ......................................................................................................164 8.2. Experimental ....................................................................................................166 8.2.1. Chemicals ...................................................................................................166 8.2.2. FePt Nanoparticles ......................................................................................166 8.2.3. Preparation of Particle Langmuir-Blodgett Thin Films ..............................167 8.2.4. Atomic Force Microscopy (AFM) .............................................................168 8.2.5. Electrochemistry ........................................................................................168  8.3. Results and Discussion .....................................................................................169 8.3.1. Langmuir-Blodgett Thin Films of FePt Nanoparticles ..............................169 8.3.2. Electrochemistry of FePt Functionalized Electrodes ................................174 8.3.3. Formic Acid Oxidation ..............................................................................175 8.3.4. Electrochemical Impedance Studies ..........................................................184 8.4. Conclusion ........................................................................................................192  REFERENCES ..........................................................................................................194 Chapter 9: Composition Effects of FePt Alloy Nanoparticles on the Electro-Oxidation of Formic Acid .................................................198 9.1. Introduction ......................................................................................................199 9.2. Experimental ....................................................................................................201     xviii   9.2.1. Chemicals ...................................................................................................201 9.2.2. Nanoparticle Preparation ............................................................................201 9.2.3. Preparation of the FePt/Au Electrode .........................................................203 9.2.4. Electrochemistry ........................................................................................203  9.3. Results and Discussion .....................................................................................204 9.3.1. Characterization of FexPt100-x Nanoparticles .............................................204 9.3.2. Cyclic Voltammetry of FexPt100-x Nanoparticles .......................................208 9.3.3. Electro-Oxidation of Formic Acid .............................................................210 9.3.4. Electrochemical Impedence Studies ..........................................................217 9.4. Conclusion ........................................................................................................235  REFERENCES ..........................................................................................................236     xix   List of Tables 1.1 ....................................................................................................................................15 6.1 ..................................................................................................................................106 6.2 ..................................................................................................................................110 7.1 ..................................................................................................................................155 9.1 ..................................................................................................................................207     xx   List of Figures 1.1 ......................................................................................................................................3 1.2 ......................................................................................................................................4 1.3 ......................................................................................................................................6 1.4 ......................................................................................................................................7 1.5 ......................................................................................................................................8 1.6 ....................................................................................................................................11 1.7 ....................................................................................................................................12 1.8 ....................................................................................................................................13 2.1 ....................................................................................................................................32 2.2 ....................................................................................................................................34 2.3 ....................................................................................................................................35 2.4 ....................................................................................................................................36 2.5 ....................................................................................................................................37 3.1 ....................................................................................................................................47 3.2 ....................................................................................................................................48 3.3 ....................................................................................................................................50 3.4 ....................................................................................................................................52 3.5 ....................................................................................................................................54 3.6 ....................................................................................................................................56     xxi   4.1 ....................................................................................................................................66 4.2 ....................................................................................................................................67 4.3 ....................................................................................................................................69 4.4 ....................................................................................................................................72 5.1 ....................................................................................................................................85 5.2 ....................................................................................................................................86 5.3 ....................................................................................................................................88 5.4 ....................................................................................................................................89 5.5 ....................................................................................................................................91 5.6 ....................................................................................................................................92 5.7 ....................................................................................................................................93 5.8 ....................................................................................................................................95 6.1 ..................................................................................................................................105 6.2 ..................................................................................................................................107 6.3 ..................................................................................................................................109 6.4 ..................................................................................................................................112 6.5 ..................................................................................................................................114 6.6 ..................................................................................................................................116 6.7 ..................................................................................................................................117 6.8 ..................................................................................................................................119 6.9 ..................................................................................................................................123     xxii   7.1 ..................................................................................................................................136 7.2 ..................................................................................................................................137 7.3 ..................................................................................................................................140 7.4 ..................................................................................................................................145 7.5 ..................................................................................................................................148 7.6 ..................................................................................................................................151 7.7 ..................................................................................................................................153 7.8 ..................................................................................................................................156 8.1 ..................................................................................................................................169 8.2 ..................................................................................................................................171 8.3 ..................................................................................................................................172 8.4 ..................................................................................................................................173 8.5 ..................................................................................................................................175 8.6 ..................................................................................................................................176 8.7 ..................................................................................................................................179 8.8 ..................................................................................................................................183 8.9 ..................................................................................................................................186 8.10 ................................................................................................................................188 8.11 ................................................................................................................................190 8.12 ................................................................................................................................191 9.1 ..................................................................................................................................202     xxiii   9.2 ..................................................................................................................................205 9.3 ..................................................................................................................................206 9.4 ..................................................................................................................................209 9.5 ..................................................................................................................................210 9.6 ..................................................................................................................................213 9.7 ..................................................................................................................................215 9.8 ..................................................................................................................................219 9.9 ..................................................................................................................................220 9.10 ................................................................................................................................221 9.11 ................................................................................................................................223 9.12 ................................................................................................................................224 9.13 ................................................................................................................................225 9.14 ................................................................................................................................226 9.15 ................................................................................................................................227 9.16 ................................................................................................................................228 9.17 ................................................................................................................................229 9.18 ................................................................................................................................230 9.19 ................................................................................................................................231 9.20 ................................................................................................................................232 9.21 ................................................................................................................................234     Chapter 1 Introduction of FePt Nanoparticles as a Magnet and a Electrocatalyst 2   1.1. FePt Nanoparticles as a Magnet 1.1.1. Backgrounds Nanoparticles are typically defined as solids less than 100 nm in all three dimensions. Most often they are made to be spherical having diameters on the order of 10 nm or less. At these length scales, a large fraction of the atoms of the particle are at or near the surface providing them with unique and distinguished properties comparing to bulk materials.[1] Equally, these properties change with their size or shape. The change in the properties at this length scale is not a result of scaling factor. It results from different causes in different materials. In semiconductors, it results from the further confinement of the electronic motion to a length scale characterizing the electronic motion in bulk semiconducting material. As noble metals are reduced in size to tens of nanometers, a new very strong absorption is observed resulting from the collective oscillation of the electrons in the conduction band from one surface of the particle to the other, which is called surface plasmon absorption.[2] In magnetic materials, the most important characteristic is the process of magnetization reversal. Whether the material requires very little field to reverse the magnetization (soft magnet) or maintains a single magnetization direction to very high magnetic fields (hard magnet), the magnetization reversal determines the performance of the material. Bulk ferromagnetic materials are generally polycrystalline with each grain consisting of thousands of magnetic domains separated by boundaries call domain walls. Figure 1.1a as the simplest case shows multidomain particle with 90° domain walls. The magnetic domain walls have significant width, generally in the tens to hundreds of 3   nanometers. As the size of a bulk ferromagnet is reduced until it is reached to the domain wall width, particles consist of a single domain that thermodynamically cannot support the formation of a domain wall (Figure 1.1b). This requires magnetization reversal by rotation of the magnetization into the applied field direction.[1] Figure 1.1. Non-interacting particles as function of particle size, indicating magnetization reversal mechanism regimes at isothermal temperature. 4   Again, the magnetic properties depends on the material type and size, which can be described as the relaxation of the magnetization orientation of each particle by time τ = τ0eKV/2kT, in which τ is the relaxation time at one orientation, K is the particle’s anisotropy constant (intrinsic property), V is the particle volume, k is Boltzmann’s constant, and T is temperature.[3,4] The term KV measures the energy barrier between two orientation. As the size of the particle decreases to a level where KV becomes comparable to the thermal energy kT, its magnetization starts to fluctuate from one direction to another. As a result, at this T the overall magnetic moment of this particle is randomized to zero, and the particle is said to be superparamagnetic.[4] (Figure 1.2)[5] Figure 1.2. Nanoscale transition of magnetic nanoparticles from ferromagnetism to superparamagnetism: energy diagram of magnetic nanoparticles with different magnetic spin alignment, showing ferromagnetism in a large particle (top) and superparamagentism in a small nanoparticle (bottom). 5   The magnetic behavior of a group of magnetic nanoparticles can be better described by a hysteresis loop that measures the change of magnetic moment (M) over the strength of an applied magnetic field (H).[4] As shown in Figure 1.5A, in the absence of an external field (center point), the magnetization of each particle points in different directions and the overall magnetic moment is zero. When an external magnetic field is applied, magnetic interaction between the particles and the field aligns the magnetization of the particles along the field direction. When the field is strong enough, all particles are aligned in the field direction and the particles are said to besaturated; the corresponding moment is called the saturation moment (Ms). Reducing the strength of the field leads to randomization of the magnetization and a smaller magnetic moment. When the external field drops to zero, the ferromagnetic particles retain a considerable degree of magnetization with a net measurable moment: the remnant magnetic moment (Mr). (This serves as a basis for magnetic memory devices). To demagnetize the particles, the external field must be reversed and increased to a value where the total moment is zero. This value is called the coercivity (Hc). The ferromagnetic nanoparticle materials having a large coercivity can be candidates for high density magnetic date storage and high performance permanent magnet owing to their long magnetic relaxation time.[6,7] If the particles are superparamagnetic, the magnetization of each particle undergoes thermal fluctuation. As long as the field is removed, the overall moment is randomized to zero, leaving no remnant magnetic moment (Figure 1.5B). This superparamagnetism of magnetic nanoparticles are very useful for biomedical applications, as they are not subject to strong magnetic interactions in a dispersion and are stable in physiological conditions.[8–10] 6   Figure 1.3. Schematic illustration of the hysteresis loops of a group of magnetic nanoparticles that are A) ferromagnetic and B) superparamagnetic. Ms: saturation magnetic moment; Mr: remnant magnetic moment, Hc: coercivity. 7   1.1.2. Hard Magnetic FePt Nanoparticles and Their Applications Magnetic iron-platinum (FePt) nanoparticles made from solution phase chemical syntheses have shown great potentials for high performance permanent magnet,[11-13] high density data storage,[7,14,15] and highly efficient biomedicine applications.[4,9,16] Their magnetic properties can be tuned not only by nanoparticle sizes, but also by Fe, Pt composition and Fe, Pt atomic arrangement in the FePt alloy structure.[17] Depending on the Fe to Pt elemental ratio, these alloys can display chemically disordered face centered cubic (fcc) phase (A1, Fm3m) or chemically ordered phases, such as (L12, Pm3m) for Fe3Pt, face centered tetragonal (fct) phase (L10, P4/mmm) for FePt and (L12, Pm3m) for Pt3Fe (Figure 1.4).[18-20,29] Figure 1.4. Structure models of FePt (L10) and Fe3Pt (L12) and the corresponding projections of the structures along different zone axis. 8   These structure variations have dramatic effects on the magnetic properties of the alloys. For example, the Fe3Pt material is paramagnetic[21] and the Pt3Fe is antiferromagnetic. Various experimental results have revealed that the L10 type structure can be formed in FexPt1-x, with x ranging from 0.35 to 0.60.[22] The fcc-structured FePt has a small coercivity and is magnetically soft. The fully ordered fct-structured FePt can be viewed alternating atomic layers of Fe and Pt stacked along the [001] direction in c- axis in Figure 1.5B[4] and show a large coercivity and magnetically hard based on large anisotropy constant K, which can reach as high as 107 Jm-3.[14] This large magnetic anisotropy K is originated from spin-orbit coupling and the hybridization between Fe 3d and Pt 5d electrons.[23-26] Due to the high magnetocrystalline anisotropy and robust chemical stability, fct-FePt nanoparticles are particularly interesting as models for nanomagnetism study[27,28] and as building blocks for constructing single nanoparticle information storage media[7,14,15]. Figure 1.5. Schematic illustration of the unit cell of (A) chemically disordered fcc and (B) chemically ordered fct FePt. 9   1.1.3. General Chemical Syntheses of FePt nanoparticles fcc-FePt nanoparticles can be synthesized in two ways in solution phase: (1) thermal decomposition of Fe(CO)5 and reduction of Pt(acac)2, and (2) co-reduction of metal salts. (1) Thermal decomposition of Fe(CO)5 and reduction of Pt(acac)2 with 1,2- alkanediol in high boiling point-organic solvent is a common synthesis route[14,30]. Fe(CO)5 is thermally unstable and subject to decomposition at high temperature to carbon monoxide and Fe. Pt(acac)2 is readily reduced by a mild reducing agent, 1,2-alkanediol, to Pt. A small group of Fe and Pt atoms combine to form [Fe–Pt] clusters that act as nuclei. The growth proceeds as more Fe–Pt species deposit around the nuclei, forming FePt nanoparticles. Oleic acid and oleylamine (or similar long chain carboxylic acids and primary amines) are used for FePt nanoparticle surface passivation and particle stabilization.[4] In a typical synthesis,[14] the mixture of Pt(acac)2, 1,2-hexadecanediol in dioctyl ether is heated to 100 °C under a gentle flow of N2 to remove oxygen and moisture. Oleic acid and oleylamine are added to the mixture, and, after 10 min, the N2 outlet is closed and the reaction system protected under a blanket of N2. Fe(CO)5 is then introduced into the mixture, which is then further heated to reflux at 297 °C for 30 min. Fine tuning of the sizes of the FePt nanoparticles from 2 to 9 nm is obtained by control of reaction parameters such as reaction time, temperature, heating rate, and the ratio of precursors/surfactants. (2) The use of a diol or polyalcohol to reduce metal salts to metal particles is referred to as the polyol process.[31] By mixing and heating both an iron salt and a platinum salt with the reducing agent, monodisperse FePt nanoparticles can be produced. For example[32], FeCl2 and Pt(acac)2 can be reduced by LiBEt3H and diol to produce high quality FePt nanoparticles. 10   As-synthesized fcc-FePt nanoparticles can be changed into fct structures through high temperature treatment, usually >550 °C[29,33,34] under inert gas or hydrogen mixture gas, and they can be characterized by X-ray diffraction (XRD) patterns (Figure 1.6). As- synthesized particles exhibit the chemically disordered fcc structure. Annealing induces the Fe and Pt atoms to rearrange into the long-range chemically ordered fct structure, as indicated by the (111) peak shifts and evolution of the (001) and (110) peaks. At annealing temperatures below 500 °C, only partial chemical ordering is observed. The chemical ordering can be increased by annealing at higher temperatures or by increasing the annealing time[14]. Previous thermal annealing experiments have shown that, for stoichiometrical bulk FePt alloy, the A1 to L10 transformation temperature is 1300 °C, [22,29] while for nanoscale FePt particles, this temperature is lowered to within 500-700 °C, depending on FePt stoichiometry and particle sizes (Figure 1.7).[34] The chapters from 2 to 5 will be discussed about the control of shape and structure of FePt nanoparticles and their applications as a magnet. 11   Figure 1.6. (Left) XRD patterns (A) of as-synthesized 4 nm Fe52Pt48 particle assemblies and a series of similar assemblies annealed under atmospheric N2 gas for 30 min at tmperatures of (B) 450 °C, (C) 500 °C, (D) 550 °C, and (E) 600 °C. The indexing is based on tabulated fct FePt reflections. The diffraction patterns were collected with a Siemens D-500 diffractometer with CuKα radiation (wavelength λ = 1.54056 Å). (Right) HRSEM image of a ~180 nm-thick, 4 nm Fe52Pt48 nanocrystal assembly annealed at 560 °C for 30 min under 1 atm of N2 gas. 12   Figure 1.7. Equilibrium phase diagram of the Fe-Pt system. 13   1.2. FePt Nanoparticles as a Electrocatalyst 1.2.1. Backgrounds Fuel cells are attractive alternatives to combustion engines for electrical-power generation because of their very high efficiencies and low pollution levels. Like a combustion engine, a fuel cell uses some sort of chemical fuel as its energy source, but like a battery, the chemical energy is directly converted to electrical energy, without a messy and inefficient combustion step. The components in a fuel cell that make this direct electro chemical conversion possible are an ion-conducting electrolyte, a cathode, and an anode, as shown schematically in Figure 1.8.[36] Figure 1.8. Principle of fuel cell operation. x- is a mobile anion. z+ is a mobile cation. 14   In the simplest example, a fuel such as hydrogen (H2) is brought into the anode compartment and oxygen (O2) is brought into the cathode compartment. There is an overall chemical driving force for the oxygen and the hydrogen to react to produce water. In the fuel cell, however, this simple chemical reaction is prevented by the electrolyte that separates the fuel (H2) from the oxidant (O2). The electrolyte serves as a barrier to gas diffusion, but it will let ions, in this example O2- (oxide ions), migrate across it. In order for the reaction between hydrogen and oxygen to occur, the oxygen atoms must somehow pick up electrons at the cathode and give off electrons at the anode. The reactions are then: Cathode : ½ O2 + 2e- Æ O2- Anode : H2 + O2- Æ H2O + 2e- Overall : ½ O2 + H2 Æ H2O In order for the “half-cell” reactions at the anode and cathode to be possible, there must be some external path by which electrons move, and it is precisely this electron motion that provides usable electricity from the fuel cell.[36] Several types of fuel cells according to the type of elctrolyte have been developed over the past few decades as seen in Table 1.1. For example, in PEMFC, the reactions are Cathode : O2 + 4e- + 4H+ Æ 2H2O Anode : 2H2 Æ 4H+ + 4e- Overall : O2 + 2H2 Æ 2H2O. 15   Table 1.1. Fuel Cell Types Type PEM AFC PAFC MCFC SOFC °C 90-110 100-250 150-220 500-700 700-1000 Fuel H2 + H2O H2 H2 HC + CO HC + CO Electrolyte Nafion KOH H3PO4 Na2CO3 Y-ZrO2 Ion H3O+ ↓ OH- ↑ H+ ↓ CO32- ↑ O2- ↑ Oxidant O2 O2 + H2O O2 O2 + CO2 O2 PEM : Proton Exchange Membrane AFC : Alkali Fuel Cell PAFC : Phosphoric Acid Fuel Cell MCFC : Molten Carbonate Fuel Cell SOFC : Solid Oxide Fuel Cell HC : Hydrocarbon 16   1.2.2. Electrocatalytic Activity of FePt Nanoparticles Generally Pt is used for both anode and cathode electrodes as the most active electrocatalyst regardless of fuel cell types. They catalyze hydrogen oxidation at the anode and oxygen reduction at the cathode, especially in PEMFC.[37] However, Pt is very expensive and it hinders the development of large scale fuel cell applications. To reduce cost, nanoparticles of platinum on a carbon support have been developed and still the development of stable nano-structures is ongoing.[38] In the mean time, to improve the performance and to reduce the costs of fuel cells, a wide variety of Pt-based electrocatalysts have been prepared and evaluated for anodic and cathodic electron- transfer reactions.[39-42] Oxygen reduction reactions (ORR) represent a critical cathodic process in fuel cells. In the past decades, various cathode catalysts, such as single crystals of noble metals,[42,43] single crystals modified with non-noble metals,[44-46] Pt-free catalysts,[47-56] and Pt-based metal alloys, have been tested for oxygen reduction. Of these, Pt alloys (e.g., NiPt, CoPt, FePt, CrPt, etc.),[57-65] especially nanosized Pt alloys,[59,66-68] have demonstrated substantially greater activities than others. By incorporating a second metal into the Pt catalysts, the alloying process results in a favorable Pt−Pt distance for the dissociative adsorption of O2 because the base transition-metal atoms are typically smaller than Pt, leading to enhanced catalytic activity for oxygen electroreduction.[69,70] It is usually believed that, in order to enhance the ORR activity of Pt-based alloy catalysts, adsorption of OH on the second metal must be facilitated, whereas the adsorption on the Pt sites should be diminished, since high coverage of adsorbed OH species on the Pt surface has been proved to inhibit the oxygen reduction reactions.[42,69,71,72] In previous 17   X-ray absorption near-edge structure studies, binary alloys such as CrPt and FePt have been found to exhibit these favorable chemisorption characteristics of OH species.[69] In small molecule (methanol or formic acid) fuel cell reactions at anode electrode, one of the major problems is the poisoning of the electrocatalysts by CO-like intermediate species. Also, to improve the performance in fuel oxidation, the combination of Pt and other transition metals has been examined extensively as effective catalysts for methanol and formic acid electro-oxidation, as Pt-based alloy catalysts typically display enhanced catalytic activity toward methanol and formic acid oxidation that has been attributed to the so-called bifunctional and/or electronic (ligand) effect mechanism.[73-83] For Pt-based binary alloy catalysts, Pt and the second metal play different roles in the oxidation catalysis. According to the so-called bifunctional mechanism, the role of the second metal is to dissociate water to form adsorbed OH species, which then react with CO adsorbed on the Pt surface to generate CO2.[84,85] This suggests that for alloy nanoparticle catalysts, in addition to particle dimensions, the composition of the bimetallic particles also acts as a very important parameter to manipulate its catalytic activity. Such composition effects have been observed for CO and methanol electro- oxidation with PtRu alloy catalysts.[77,86-92] Related these works with FePt nanoparticles as active electrocatalyst will refer to from chapter 6 to 9. 18   REFERENCES 1. M. A. Willard, L. K. Kurihara, E. E. Carpenter, S. Calvin, V. G. Harris, Int. Mater. Rev., 2004, 49, 125. 2. C. Burda, X. Chen, R. Narayanan, M. A. EI-Sayed. Chem. Rev., 2005, 105, 1025. 3. A. H. Morrish, The physical Principles of Magnetism, Wiley, New York, 1965. 4. S. Sun, Adv. Mater., 2006, 18, 393. 5. Y.-W. Jun, J.-W. Seo, J. Cheon, Accounts of Chemical Research, 2008, 41, 179. 6. D. Weller, A. Moser, IEEE. Trans. Mag., 1999, 35, 4423. 7. D. Weller, M. F. Doerner, Annu. Rev. Mater. Sci., 2000, 30, 611. 8. A. Jordan, R. Scholz, P. Wust, H. Fähling, R. Felix, J. Magn. Magn. Mater., 1999, 201, 413. 9. Q. A. Pankhurst, J. Connolly, S. K. Jones, J. Dobson, J. Phys. D: Appl. Phys., 2003, 36, R167. 10. T. Neuberger, B. Schöpf, H. Hofmann, M. Hofmann, B. von Rechenber, J. Magn. Magn. Mater., 2005, 293, 483. 11. E. F. Kneller, R. Hawig, IEEE Trans. Magn., 1991, 27, 3588. 12. R. Skomski, J. M. D. Coey, Phys. Rev. B, 1993, 48, 15812. 13. H. Zeng, J. Li, Z. L. Wang, J. P. Liu, S. Sun, Nature, 2002, 420, 395. 14. S. Sun, C. B. Murray, D. Weller, L. Folks, A. Moser, Science, 2000, 287, 1989. 15. A. Moser, K. Takano, D. T. Margulies, M. Albrecht, Y. Sonobe, Y. Ikeda, S. Sun, E. E. Fullerton, J. Phys. D: Appl. Phys., 2002, 35, R157. 19   16. J. Gao, G. Liang, J. S. Cheung,Y. Pan, Y. Kuang, F. Zhao, B. Zhang, X. Zhang, E. X. Wu, B. Xu, J. Am. Chem. Soc., 2008, 130, 11828. 17. J. Kim, C. Rong, Y. Lee, J. P. Liu, S. Sun, Chem. Mater., 2008, 20, 7242. 18. L. J. Cabri, D. R. Owens, J. H. G. Laflamme, Can. Mineral., 1973, 12, 21. 19. L. J. Cabri, C. E. Feather, Can. Mineral., 1975, 13, 117. 20. J. Li, Z. L. Wang, H. Zeng, S. Sun, J. P. Liu, Appl. Phys. Lett., 2003, 32, 3743. 21. J. P. Liu, C. P. Luo, Y. Liu, D. Sellmyer, J. Appl. Phys. Lett., 1998, 72, 483. 22. M. Hansen, Constitution of Binary Alloys, McGraw-Hill: New York, 1958. 23. T. Burkert, O. Eriksson, S. I. Simak, A. V. Ruban, B. Sanyal, L. Nordstrom, J. M. Wills, Phys. Rev. B: Condens. Matter Mater. Phys., 2005, 71, 134411. 24. G. Brown, B. Kraczek, A. Janotti, T. C. Schulthess, G. M. Stocks, D. D. Johnson, Phys. Rev. B. Condens. Matter Mater. Phys., 2003, 68, 052405. 25. O. Kitakami, S. Okamoto, N. Kikuchi, Y. Shimada, Jan. J. Appl. Phys. Part 2, 2003, 42, 455. 26. J. B. Staunton, S. Ostanin, S. S. S. Razee, B. L. Gyorffy, L. Szunyogh, B. Ginatempo, E. Bruno, Phys. Rev. Lett., 2004, 93, 257204. 27. M. Ulmeanu, C. Antoniak, U. Wiedwald, M. Farle, Z. Frait, S. Sun, Phys. Rev. B, 2004, 69, 054417. 28. C. Antoniak, J. Lindner, M. Spasova, D. Sudfeld, M. Acet, M. Farle, K. Fauth, U. Wiedwald, H. G. Boyen, P. Ziemann, F. Wilhelm, A. Rogalev, S. Sun, Phys. Rev. Lett., 2006, 97, 117201. 29. Z. R. Dai, S. Sun, Z. L. Wang, Nano Lett., 2001, 1, 443. 20   30. S. Sun, E. E. Fullerton, D. Weller, C. B. Murray, IEEE Trans. Magn., 2001, 37, 1239. 31. F. Fievet, J. P. Lagier, M. Figlarz, MRS Bull., 1989, 14, 29. 32. S. Sun, S. Anders, T. Thomson, J. E. E. Baglin, M. F. Toney, H. F. Hamann, C. B. Murray, B. D. Terris, J. Phys. Chem. B, 2003, 107, 5419. 33. K. E. Elkins, T. S. Vedantam, J. P. Liu, H. Zeng, S. Sun, Y. Ding, Z. L. Wang, Nano Lett., 2003, 3, 1647. 34. T. S. Vedantam, J. P. Liu, H. Zeng, S. Sun, J. Appl. Phys., 2003, 93, 7184. 35. O. Gutfleisch, J. Lyubina, K.-H. Muller, L. Schultz, Adv. Eng. Mater., 2005, 7, 208. 36. http://addis.caltech.edu/publications/haile_engenious.pdf at Caltech.  37. C. Wang, H. Daimon, T. Onodera, T. Koda, S. Sun, Angew. Chem. Int. Ed., 2008, 47, 3588. 38. N. P. Brandon, S. Skinner, B. C. H. Steele, Annu. Rev. Mater. Res., 2003, 33, 183. 39. R. Parsons, T. Vandernoot, J. Electroanal. Chem., 1988, 257, 9. 40. D. Cameron, R. Holliday, D. Thompson, J. Power Sources, 2003, 118, 298. 41. N. M. Markovic, P. N. Ross, Surf. Sci. Rep., 2002, 45, 121. 42. J. X. Wang, N. M. Markovic, R. R. Adzic, J. Phys. Chem. B, 2004, 108, 4127. 43. H. Inoue, S. R. Brankovic, J. X. Wang, R. R. Adzic, Electrochim. Acta, 2002, 47, 3777. 44. J. Zhang, Y. Mo, M. B. Vukmirovic, R. Klie, K. Sasaki, R. R. Adzic, J. Phys. Chem. B, 2004, 108, 10955. 45. M. H. Shao, T. Huang, P. Liu, J. Zhang, K. Sasaki, M. B. Vukmirovic, R. R. Adzic, 21   Langmuir, 2006, 22, 10409. 46. S. Mukerjee, S. Srinivasan, J. Electroanal. Chem., 1993, 357, 201. 47. Y. Shimizu, K. Uemura, H. Matsuda, N. Miura, N. Yamazoe, J. Electrochem. Soc., 1990, 137, 3430. 48. J. Prakash, H. Joachin, Electrochim. Acta, 2000, 45, 2289. 49. A. Vanderputten, A. Elzing, W. Visscher, E. Barendrecht, J. Electroanal. Chem., 1986, 205, 233. 50. A. Vanderputten, A. Elzing, W. Visscher, E. Barendrecht, J. Chem. Soc., Chem. Commun., 1986, 477. 51. K. Wiesener, Electrochim. Acta, 1986, 31, 1073. 52. S. Gupta, D. Tryk, I. Bae, W. Aldred, E. Yeager, J. Appl. Electrochem., 1989, 19, 19. 53. G. Wei, J. S. Wainright, R. F. Savinell, J. New Mater. Electrochem. Systems, 2000, 3, 121. 54. A. L. Bouwkamp-Wijnoltz, W. Visscher, J. A. R. van Veen, S. C. Tang, Electrochim. Acta, 1999, 45, 379. 55. K. Sawai, D. Uda, Chem. Lett., 2006, 35, 1166. 56. K. Sawai, D. Uda, K. Shirai, Electrochemistry, 2007, 75, 163. 57. T. Toda, H. Igarashi, H. Uchida, M. Watanabe, J. Electrochem. Soc., 1999, 146, 3750. 58. V. Stamenkovic, T. J. Schmidt, P. N. Ross, N. M. Markovic, J. Phys. Chem. B, 2002, 106, 11970. 22   59. U. A. Paulus, A. Wokaun, G. G. Scherer, T. J. Schmidt, V. Stamenkovic, V. Radmilovic, N. M. Markovic, P. N. Ross, J. Phys. Chem. B, 2002, 106, 4181 60. U. A. Paulus, A. Wokaun, G. G. Scherer, T. J. Schmidt, V. Stamenkovic, N. M. Markovic, P. N. Ross, Electrochim. Acta, 2002, 47, 3787. 61. V. Stamenkovic, T. J. Schmidt, P. N. Ross, N. M. Markovic, J. Electroanal. Chem., 2003, 554, 191. 62. G. Tamizhmani, G. A. Capuano, J. Electrochem. Soc., 1994, 141, 968. 63. F. Colmati, E. Antolini, E. R. Gonzalez, J. Power Sources, 2006, 157, 98. 64. B. C. Beard, P. N. Ross, J. Electrochem. Soc., 1986, 133, 1839. 65. V. R. Stamenkovic, B. S. Mun, M. Arenz, K. J. J. Mayrhofer, C. A. Lucas, G. F. Wang, P. N. Ross, N. M. Markovic, Nat. Mater., 2007, 6, 241. 66. H. Yang, W. Vogel, C. Lamy, N. Alonso-Vante, J. Phys. Chem. B, 2004, 108, 11024. 67. M. K. Min, J. H. Cho, K. W. Cho, H. Kim, Electrochim. Acta, 2000, 45, 4211. 68. J. Zhang, F. H. B. Lima, M. H. Shao, K. Sasaki, J. X. Wang, J. Hanson, R. R. Adzic, J. Phys. Chem. B, 2005, 109, 22701. 69. G. A. Camara, E. A. Ticianelli, S. Mukerjee, S. J. Lee, J. McBreen, J. Electrochem. Soc., 2002, 149, A748. 70. V. Jalan, E. J. Taylor, J. Electrochem. Soc., 1983, 130, 2299. 71. J. L. Zhang, M. B. Vukmirovic, K. Sasaki, A. U. Nilekar, M. Mavrikakis, R. R. Adzic, J. Am. Chem. Soc., 2005, 127, 12480. 72. P. B. Balbuena, D. Altomare, N. Vadlamani, S. Bingi, L. A. Agapito, J. M. 23   Seminario, J. Phys. Chem. A, 2004, 108, 6378. 73. H. S. Wang, C. Wingender, H. Baltruschat, M. Lopez, M. T. Reetz, J. Electroanal. Chem., 2001, 509, 163. 74. K. W. Park, J. H. Choi, B. K. Kwon, S. A. Lee, Y. E. Sung, H. Y. Ha, S. A. Hong, H. Kim, A. Wieckowski, J. Phys. Chem. B, 2002, 106, 1869. 75. F. Colmati, E. Antolini, E. R. Gonzalez, Electrochim. Acta, 2005, 50, 549. 76. E. Antolini, J. R. C. Salgado, E. R. Gonzalez, J. Electroanal. Chem., 2005, 580, 145. 77. S. Motoo, M. Watanabe, J. Electroanal. Chem., 1976, 69, 429. 78. C. Roychowdhury, F. Matsumoto, V. B. Zeldovich, S. C. Warren, P. F. Mutolo, M. Ballesteros, U. Wiesner, H. D. Abruña, F. DiSalvo, J. Chem. Mater., 2006, 18, 3365. 79. R. S. Jayashree, J. S. Spendelow, J. Yeom, C. Rastogi, M. A. Shannon, P. J. A. Kenis, Electrochim. Acta, 2005, 50, 4674. 80. W. F. Lin, M. S. Zei, M. Eiswirth, G. Ertl, T. Iwasita, W. Vielstich, J. Phys. Chem. B, 1999, 103, 6968. 81. T. Frelink, W. Visscher, J. A. R. vanVeen, Langmuir, 1996, 12, 3702. 82. T. Page, R. Johnson, J. Hormes, S. Noding, B. Rambabu, J. Electroanal. Chem., 2000, 485, 34. 83. L. B. Lai, D. H. Chen, T. C. Huang, J. Mater. Chem., 2001, 11, 1491. 84. T. Iwasita, Electrochim. Acta, 2002, 47, 3663. 85. Y. Y. Tong, H. S. Kim, P. K. Babu, P. Waszczuk, A. Wieckowski, E. Oldfield, J. 24   Am. Chem. Soc., 2002, 124, 468. 86. Y. Wang, E. R. Fachini, G. Cruz, Y. Zhu, Y. Ishikawa, J. A. Colucci, C. R. Cabrera, J. Electrochem. Soc., 2001, 148, C222. 87. M. Watanabe, M. Uchida, S. Motoo, J. Electroanal. Chem., 1987, 229, 395. 88. H. N. Dinh, X. M. Ren, F. H. Garzon, P. Zelenay, S. Gottesfeld, J. Electroanal. Chem., 2000, 491, 222. 89. L. Dubau, C. Coutanceau, E. Garnier, J. M. Leger, C. Lamy, J. Appl. Electrochem., 2003, 33, 419. 90. T. Iwasita, H. Hoster, A. John-Anacker, W. F. Lin, W. Vielstich, Langmuir, 2000, 16, 522. 91. H. Hoster, T. Iwasita, H. Baumgartner, W. Vielstich, Phys. Chem. Chem. Phys., 2001, 3, 337. 92. Z. G. Shao, F. Y. Zhu, W. F. Lin, P. A. Christensen, H. M. Zhang, J. Power Sources, 2006, 161, 813. Chapter 2 Synthesis of FePt Nanocubes and Their Oriented Self- Assembly 26   2.1. Introduction Fabrication of ordered nanomagnet arrays with controlled magnetic alignment is an important goal in achieving high density information storage[1] and high performance permanent magnets.[2] For a uniaxial anisotropic nanomagnet, magnetic alignment infers that magnetization of the magnet lies only along the “easy direction”. In ferromagnetic hysteresis behaviors, such an alignment corresponds to a hysteresis loop with high squareness, which leads to the remanent magnetization value close to the saturation one. Magnets with this magnetic behavior can have distinct magnetization reversals and retain maximum magnetic energy for data storage and permanent magnetic applications. Among all available hard magnetic nanomaterials, face centered tetragonal (fct) structured FePt nanocrystals are especially interesting because they have uniaxial anisotropy along the [001] direction and their magnetocrystalline anisotropy constant, K, which measures the ease of magnetization reversal along the [001] direction, can reach as high as 107 J/m3,[3] a value that is one of the largest among the known hard magnetic materials. Such hard magnetic FePt nanocrystals have been made by high temperature (250−300 °C) solution phase synthesis, followed by nanocrystal self-assembly and high temperature (usually >550 °C) thermal annealing.[4] As synthesized, the FePt nanoparticles possess face centered cubic (fcc) structure and are superparamagnetic at room temperature. Thermal annealing converts the fcc FePt to fct FePt, yielding nanocrystalline magnets with coercivity reaching as high as 30 kOe.[5] Despite these synthetic progresses, there is still no evidence that magnetic orientation in FePt 27   nanocrystal arrays can be established unless a physical deposition process is used,[6] which often has difficulty in controlling nanomagnet sizes and morphologies. In this chapter, we report a chemical synthesis of FePt nanocubes and their self- assembly into FePt nanocrystal superlattice arrays with controlled texture and magnetic alignment. We recently reported that the shape of the MnFe2O4 nanoparticles could induce texture in self-assembled superlattice arrays.[7] In an assembly containing cube- like nanoparticles, the array shows (100) texture, while in a polyhedron-like nanoparticle assembly, (110) texture was obtained. These early experiments indicate that structural alignment of each particle can be achieved if the shape of the nanoparticles is controlled. In searching for shape-controlled synthesis of FePt nanoparticles, we found that simultaneous decomposition of Fe(CO)5 and reduction of Pt(acac)2 without polyol as a reducing agent yielded FePt nanoparticles with sizes tunable from 3 to 9 nm.[8] Our further synthetic studies revealed that, by controlling the addition sequence of the stabilizers (oleic acid and oleylamine) and Fe/Pt ratio in the precursors, FePt nanocubes could be prepared at relatively low temperature (205 °C). This communication describes the synthesis of FePt nanocubes and self-assembly of these cubes into textured superlattice arrays. The work demonstrates that shape-controlled synthesis and self- assembly can offer a simple solution to fabrication of magnetically aligned FePt nanocrystal arrays that are promising for single nanoparticle recording and for high performance exchange-spring nanocomposite magnet applications. 28   2.2. Experimental 2.2.1. Chemicals The synthesis was carried out using standard airless procedures and commercially available reagents. Absolute ethanol and hexane were used as received. Benzyl ether (99%), octadecene, oleic acid (90%), oleylamine (>70%), and iron pentacarbonyl, Fe(CO)5 were purchased from Aldrich Chemical Company. Platinum (II) acetylacetonate was from Strem Chemicals, Inc. 2.2.2. Synthesis of 6.9 nm Fe50Pt50 Nanocubes 0.5 mmol of Pt(acac)2 and 10 mL of benzyl ether and 5 mL of octadecene were mixed in a 125 mL three-neck flask and magnetically stirred. The mixture was heated under a flow of nitrogen (~5 mL/min) to 60 °C to ensure the dissolution of Pt(acac)2. The mixture solution was further heated to 120 °C and the flask was kept under a blanket of nitrogen. Fe(CO)5 (2.0 mmol) and oleic acid (5 mmol) were introduced and after 5 minutes, oleylamine (5 mmol) was added. The mixture solution was then heated to 205 °C and incubated at this temperature for two hours. The black dispersion was then cooled down to room temperature. Fe50Pt50 nanocubes were precipitated out and washed twice with ethanol. The precipitates were re-dispersed in 10 mL of hexane in the presence of small amount (~0.05 mL) of oleic acid and oleylamine. 29   2.2.3. Composition Control during the Synthesis of FePt Nanocubes There are at least four factors that affect the composition of final FePt nanoparticles: (1) the ratio of Fe(CO)5/Pt(acac)2, (2) reaction temperature, (3) the ratio of oleic acid/oleylamine, and (4) addition sequence of the ligands (for the equal molar of oleic acid and oleylamine). Reaction of 2:1 molar ratio of Fe(CO)5 to Pt(acac)2 in benzeyl ether with 1:1 molar ratio of oleic acid to oleylamine at a reaction temperature of 300 °C generally resulted in near equiatomic FePt nanoparticles, while reaction of the same reactants at 205 °C led to Pt-rich FexPt1-x (x range from 0.25 to 0.38, depending on addition sequence of oleic and oleylamine) nanoparticles. In current synthesis, we achieved the control of the FePt nanocubes in near equiatomic ratio by increasing the Fe:Pt ratio to 4:1 in the presence of a portion of octadecene. 2.2.4. Self-Assembly of FePt Nanocubes In order to obtain self-assembled FePt nanocubes arrays, the FePt nanocube dispersion was diluted 10-20 times with a mixture of hexane and octane (v/v = 1/1), and one drop of the diluted dispersion was dropped onto a carboncoated Cu TEM grid which was put on a 3 mm × 3 mm Si wafer. Nanocubes were selfassembled on the TEM grid after slow evaporation of solvent in about 30 minutes. Samples for XRD, EDX, and SQUID were obtained by dropping the above solution on a 6 mm × 12 mm Si(100) wafer. The as-deposited samples were annealed under Ar at 675 °C for 1 hour. 30   2.2.5. Characterization The nanocubes were imaged using a Philips CM 12 TEM (120 kV). The structure of the cubes was characterized using HRTEM and selected area electron diffraction (SAED) on a JEOL 2010 TEM (200 kV). Quantitative elemental analyses of the cubes were carried out with spatially resolved electron diffraction spectrum (EDS) or Energy Disperse spectroscopy using a LEO 1560 Scanning electron microscope equipped with the EDS capability. X-ray diffraction patterns of the particle assemblies were collected on a Siemens D-500 diffractometer under CoKα radiation (λ = 1.788965 Å). Magnetic studies were carried out using a MPMS2 Quantum Design SQUID magnetometer with fields up to 7 T and temperatures from 10 K to 300 K. 31   2.3. Results and Discussion The 6.9 nm Fe50Pt50 nanocubes were synthesized by mixing oleic acid and Fe(CO)5 with benzyl ether/octadecene solution of Pt(acac)2 and heating the mixture to 120 °C for about 5 min before oleylamine was added, and the mixture was heated at 205 °C for 2 h. The key to the success of the nanocube synthesis at 205 °C, rather than 300 °C the temperature normally used for FePt nanoparticle synthesis is that more Fe(CO)5 should be introduced to the reaction system. The shape is controlled by adding oleic acid first during the reaction. The nanocubes are likely formed from the growth of the cubic Pt-rich nuclei generated during the initial stage of the reaction, as the −COOH does not have a strong tendency to bind to Pt. It is known that the surface energy of crystallographic planes of a fcc Pt crystal generally follow the trend of (111) < (100).[9] In a kinetic growth process, the Fe-rich species prefer to deposit on the (100) plane, leading to the formation of a cube. If oleylamine was added first, sphere-like FePt nanoparticles were separated. This indicates that the amine reacts with Pt, forming a stable Pt−NH2− complex and hindering the nucleation process. The resulting nuclei may contain more Fe in this case, producing an amorphous spherical structure. The uniform FePt nanocubes or spherical FePt nanoparticles must be derived from atomic diffusion between a Pt-rich core and an Fe-rich shell in a process that is similar to what has been described in a previous publication.[8] Transmission electron microscopic (TEM) image of the 6.9 nm Fe50Pt50 nanocubes is given in Figure 2.1A. Figure 2.1B is the high-resolution TEM (HRTEM) image of a single Fe50Pt50 nanocube. It shows the lattice fringes with an inter-fringe distance of 32   0.192 nm, close to the lattice spacing of the {100} planes at 0.1908 nm in the fcc structured FePt. Fast Fourier transformation (FFT) of the single particle (Figure 2.1C) reveals a 4-fold symmetry, consistent with the fcc structure projected from the [001] direction. These indicate that the {100} lattice fringes are parallel to the edges of the cube. Figure 2.1. TEM bright field images of (A) 6.9 nm Fe50Pt50 nanocubes; (B) HRTEM of a single FePt nanocube; (C) FFT of the cube in (B). 33   Controlled evaporation of the carrier solvent from the hexane dispersion (~2 mg/mL) of the nanocubes led to a Fe50Pt50 nanocube superlattice array, as shown in Figure 2.2. FFT of the image reveals that the assembly has the cubic packing with a view from the [001] direction showing a 4-fold symmetry as seen in Figure 2.3. The image from a 40° tilting angle shows that the bottom layer of the FePt nanocube array is lined with the top layer (Figure 2.4). This assembly pattern is energetically favored as it gives the maximum van der Waals interaction energy arising from face−face interactions in short distance of the cube assembly.[10] The interparticle distance is around 4−5 nm, close the simple thickness addition of the cube coating layer (2−2.5 nm, the length of oleate or oleylamine). Selected area electron diffraction (SAED) of the assembly in Figure 2.2A exhibits four bright (200) spots that are linked by a 4-fold symmetry, as shown in Figure 2.2B. The (111) diffraction ring is very weak in this diffraction pattern. These indicate that the assembly in Figure 2.2A is (100) textured. The textured cubic assembly is further revealed by a small angle diffraction of the assembly (Figure 2.2C). X-ray diffraction (XRD) of the self-assembled FePt nanocubes on a Si(100) substrate also shows a strong (200) reflection (Figure 2.3B). This is markedly different from that of a 3D randomly oriented FePt nanoparticle assembly with a strong (111) peak,[8] indicating that each nanocube in the assembly has a preferred crystal orientation with {100} planes parallel to the substrate. 34   Figure 2.2. (A) TEM image of a multilayer assembly of 6.9 nm Fe50Pt50 nanocubes; and (B) SAED of the assembly in (A), and (C) small angle diffraction of the assembly in (A). 35   Figure 2.3. (A) TEM image of the 6.9 nm Fe50Pt50 nanocube assembly (A1: FFT of the assembly; A2: SAED of the assembly); (B) XRD of the 6.9 nm Fe50Pt50 nanocube assembly. 36   Figure 2.4. TEM images of a multilayer 6.9 nm Fe50Pt50 nanocube assembly with and without the sample grid tilting. 37   Thermal annealing of the FePt nanocube superlattice induces FePt structure transformation from fcc to fct. The XRD pattern of the annealed assembly (675 °C under Ar for 1 h) has two strong (001) and (200) peaks, as shown in Figure 2.5A. The narrowed peaks indicate the particle growth during the annealing process. However, the peak intensity is different from that of the spherical FePt nanoparticle assembly, which shows only one strong (111) peak.[8] This indicates that (001) and (200) planes in the thermally annealed FePt nanocube array are now parallel and perpendicular to the substrate. Figure 2.5B is the room temperature hysteresis loop of the annealed FePt nanocube assembly with coercivity at 22 kOe. The loop is exactly the same in both parallel and perpendicular directions of the assembly, confirming what is concluded from the XRD analysis in Figure 2.5A. Figure 2.5. (A) XRD of thermally annealed FePt nanocube assembly on a Si surface, and (B) room temperature in-plane hysteresis loop of the FePt nanocube assembly in (A). 38   2.4. Conclusion In conclusion, we have reported that FePt nanocubes can be produced at 205 °C by controlling Fe/Pt ratio in the precursors and addition sequence of oleic acid and oleylamine. Self-assembly of these FePt nanocubes creates a (100) textured array. Thermal annealing induces the internal particle structure change and transforms the nanocube assembly from superparamagnetic to ferromagnetic. The work demonstrates that it is possible to use particle shape to control assembly texture and further on magnetic alignment. When this is achieved, shape-controlled synthesis and self-assembly will evolve as a convenient approach to magnetically aligned FePt nanocrystal arrays for single particle recording and for maximization of energy product in an exchange-spring nanocomposite system. 39   REFERENCES 1. (a) D. Weller, A. Moser, IEEE Trans. Magn., 1999, 35, 4423. (b) D. Weller, M. E. Doerner, Annu. Rev. Mater. Sci., 2000, 30, 611. (c) A. Moser, K. Takano, D. T. Margulies, M. Albrecht, Y. Sonobe, Y. Ikeda, S. Sun, E. E. Fullerton, J. Phys. D: Appl. Phys., 2002, 35, R157. 2. (a) E. F. Kneller, R. Hawig, IEEE Trans. Magn., 1991, 27, 3588. (b) R. Skomski, J. M. D. Coey, Phys. Rev. B, 1993, 48, 15812. (c) T. Schrefl, H. Kronmüller, J. Fidler, J. Magn. Magn. Mater., 1993, 127, L273. (d) H. Zeng, J. Li, Z. L. Wang, J. P. Liu, S. Sun, Nature, 2002, 420, 395. 3. R. F. C. Farrow, D. Weller, R. F. Marks, M. F. Toney, A. Cebollada, G. R. Harp, J. Appl. Phys., 1996, 79, 5967. 4. (a) S. Sun, C. B. Murray, D. Weller, L. Folks, A. Moser, Science, 2000, 287, 1989. (b) S. Sun, E. E. Fullerton, D. Weller, C. B. Murray, IEEE Trans. Magn., 2001, 37, 1239. (c) S. Sun, Adv. Mater., 2006, 18, 393. 5. K. Elkins, D. Li, N. Poudyal, V. Nandwana, Z. Jin, K. Chen, J. P. Liu, J. Phys. D: Appl. Phys., 2005, 38, 2306. 6. (a) J. U. Thiele, L. Folks, M. F. Toney, D. K. Weller, J. Appl. Phys., 1998, 84, 5686. (b) B. Bian, K. Sato, Y. Hirotsu, A. Makino, Appl. Phys. Lett., 1999, 75, 3686. (c) K. Kang, Z. G. Zhang, C. Papusoi, T. Suzuki, Appl. Phys. Lett., 2003, 82, 3284. (d) T. Shima, K. Takanashi, Y. K. Takahashi, K. Hono, Appl. Phys. Lett., 2004, 85, 2571. (e) W. K. Shen, J. H. Judy, J. P. Wang, J. Appl. Phys., 2005, 97, 10H301/1. 40   7. H. Zeng, P. M. Rice, S. X. Wang, S. Sun, J. Am. Chem. Soc., 2004, 126, 11458. 8. M. Chen, J. P. Liu, S. Sun, J. Am. Chem. Soc., 2004, 126, 8394. 9. Z. L. Wang, J. Phys. Chem. B, 2000, 104, 1153. 10. (a) B. A. Korgel, S. Fullam, S. Connolly, D. Fitzmaurice, J. Phys. Chem. B, 1998, 102, 8379. (b) S. Yamamuro, K. Sumiyama, Chem. Phys. Lett., 2006, 418, 166.   Chapter 3 A General Strategy for Synthesizing FePt Nanowires and Nanorods 42   3.1. Introduction Synthesis of FePt nanoparticles with controlled shape and magnetic alignment has become an important goal in developing nanocrystal arrays for applications in information storage,[1] permanent-magnet nanocomposites,[2] and catalysis.[3] FePt alloys are chemically stable owing to the spin-orbit coupling and the hybridization between Fe 3d and Pt 5d states,[4] and their magnetic properties can be tuned by simply controlling the atomic ratio of Fe and Pt in the alloy structure. Recent syntheses have shown that spherical FePt nanoparticles are readily made by simultaneous reduction of platinum acetylacetonate ([Pt(acac)2]) and thermal decomposition of iron pentacarbonyl ([Fe(CO)5]).[1a], [5] Thermal annealing results in hard magnetic FePt nanoparticle assemblies with coercivity reaching 30 kOe.[6] These small FePt nanoparticles are also very active in formic acid oxidation under fuel cell reaction conditions.[7] Despite these synthetic progresses, aligning these nanoparticles magnetically has constantly been a problem, and the magnetic easy axes of the nanocrystals in the assemblies are randomly oriented in three dimensions. Previous work on the synthesis and self-assembly of FePt nanocubes suggests that elongated nanocrystals may be used to achieve texture and magnetic alignment.[8] This controlled alignment of FePt nanoparticles is essential for the fabrication of single-particle recording media with ultrahigh density, magnetic nanocomposites with maximum energy product, and magnetotransport devices with optimum magnetoresistivity. In this chapeter, we report a general strategy for synthesizing FePt nanowires (NWs) and nanorods (NRs). We refer to the one-dimensional nanostructures with a 43   length of 100 nm or longer as NWs, and those below 100 nm as NRs. The diameters of both nanostructures are controlled to be 2-3 nm. The nanostructures were synthesized by reduction of [Pt(acac)2] and thermal decomposition of [Fe(CO)5] in a mixture of oleylamine (OAm) and octadecene (ODE) at 160 °C with the length readily tuneable. Our synthesis is fundamentally different from the very recent report on the preparation of FePt nanorods, for which the reaction was performed in oleic acid and oleylamine in a closed autoclave reaction system without stirring,[9] and offers much better control of both the dimensions and composition of the NWs/NRs. Owing to the structure confinement in the elongated shapes, these NWs and NRs show partial structural and magnetic alignment in thermally annealed self-assemblies. This study indicates that well-controlled NWs or NRs are likely the future choice for controlling texture and magnetic alignment in self- assembled nanomagnet arrays to support high-density magnetic information and as building blocks for fabricating highly sensitive magnetotransport devices. 44   3.2. Experimental 3.2.1. Synthesis Representative synthesis of FePt NWs and NRs: For the synthesis of 200-nm NWs, oleylamine (20 mL) was mixed with [Pt(acac)2] (0.2 g) at room temperature. Under a gentle nitrogen flow, the mixture was heated to 60 °C to form a light yellow solution. The solution was then heated to 120 °C in less than 5 min and kept at this temperature for 30 min. The color of the solution changed to dark yellow. [Fe(CO)5] (0.15 mL) was injected into the hot solution. The temperature was then raised to 160 °C. After 30 min, the solution was cooled down to room temperature by removing the heating mantle from the reaction flask. The NWs were separated by adding hexane (10 mL) and ethanol (50 mL) and subsequent centrifugation (6000 rpm). The product that was obtained was dispersed in hexane (10 mL). Other NWs/NRs were synthesized under similar reaction conditions but with different volume ratios of oleylamine/octadecene. For example, 20-nm FePt NRs were made from 10 mL oleylamine and 10 mL octadecene. 3.2.2. Characterization The FePt NWs/NRs were characterized with a transmission electron microscope (TEM, Philips EM420 at 120 kV and JEOL 2010 at 200 kV). The Fe and Pt compositions were measured by energy-dispersive spectroscopy (EDS). X-ray powder diffraction patterns of the samples were collected on a Bruker AXS D8-Advanced diffractometer 45   with CuKα radiation (λ=1.5418 Å). Magnetic properties were measured with a Lakeshore 7404 high sensitivity vibrating-sample magnetometer (VSM) with fields up to 1.5 T at room temperature. The nanoparticles were deposited from their hexane dispersions either on an amorphous carbon-coated copper grid for TEM image analyses or on a Si substrate for XRD and magnetic studies. 46   3.3. Results and Discussion The length control of the FePt NWs/NRs was realized by tuning the volume ratio of OAm/ODE, reaching from over 200 nm for NWs down to 20 nm for NRs. For example, FePt NWs with a length of over 200 nm were made when only OAm was used as both surfactant and solvent, while an OAm/ODE ratio of 3:1 gave FePt NWs of length 100 nm, and a 1:1 volume ratio of OAm/ODE led to FePt NRs of length 20 nm. Notably, using a greater proportion of ODE (OAm/ODE 1:3) led to the formation of spherical FePt nanoparticles of diameter 3 nm (Figure 3.1). With the amount of [Pt(acac)2] fixed, the compositions of these FePt nanostructures were controlled by varying the amount of [Fe(CO)5] added to the reaction mixture and were measured by energy-dispersive spectroscopy (EDS). For example, for the FePt NWs of length 200 nm, using 0.15 mL [Fe(CO)5] led to about 55 % Fe in the final product, while using 0.1 mL [Fe(CO)5] yielded the product with about 45 % Fe. Note that CoPt NWs can also be made by reduction of [Pt(acac)2] and decomposition of [Co2(CO)8] under similar reaction conditions (Figure 3.2). 47   Figure 3.1. TEM Image of the spherical FePt nanoparticles synthesized from 5 ml oleylamine and 15 ml ODE. 48   Figure 3.2. TEM image of the Co55Pt45 NWs synthesized from reduction of Pt(acac)2 and thermal decomposition of Co2(CO)8. 49   Transmission electron microscope (TEM) images of the representative NWs and NRs are given in Figure 3.3. The images in Figure 3.3 a-c show NWs of length 200 nm and NRs of length 50 nm and 20 nm. The diameter of these NWs and NRs is about 2- 3 nm. Figure 3.3 is a high-resolution TEM (HRTEM) image of two single NWs of empirical formula Fe55Pt45. In one NW, the lattice fringes are oriented approximately 55° from the wire-growth direction. The interfringe distance was measured to be 0.214 nm, which is close to the lattice spacing of the (111) planes (0.22 nm) in the face-centered cubic (fcc) FePt structure. This result indicates that the [100] direction is parallel (or perpendicular) to the wire-growth direction, which is further confirmed by the image showing the lattice fringe in the second NW with clearly discernible (100) planes (0.198 nm interfringe spacing). 50   Figure 3.3. a-c) TEM images of Fe55Pt45 NWs and NRs with a length of 200 nm (a), 50 nm (b), and 20 nm (c). d) HRTEM image of portions of two single 50-nm Fe55Pt45 NWs. 51   The synthesis and TEM analyses imply that OAm, a common organic surfactant, induces the one-dimensional growth of FePt under the current synthetic conditions. It is likely that OAm self-organizes into an elongated reverse-micelle-like structure within which the FePt nuclei are formed. This type of formation is similar to what has been proposed in the synthesis of Au NRs in the presence of cetyltrimethylammonium bromide (CTAB).[10] The elongated nuclei result in the different OAm packing densities on different surfaces, as indicated by (1), (2), and (3) in Figure 3.4a. In area (1), the molecules are well-organized and addition of FePt in this direction is more difficult owing to the presence of the hydrophobic barrier. Area (2) has less densely packed OAm and facilitates the growth of FePt along this direction and the formation of NWs or NRs. Area (3) is the most readily accessible place for the addition of FePt, leading to the fast growth of FePt and the rounded end of the NRs/NWs. The growth was monitored by taking aliquot reaction mixtures from oleylamine at 120 °C after different reaction times, and the product was quickly precipitated out and redispersed into hexane for TEM analyses. In Figure 3.4b, c, TEM images are shown of the product obtained from the reaction mixtures. It can be concluded that small, thin rodlike structures are initially present in the reaction medium and grow quickly into NWs/NRs. The fact that the growth in the [100] direction is controlled by the OAm/ODE ratio indicates that more OAm results in longer micellar structure and the formation of NWs, while dilution of OAm with ODE reduces the size of the structure, yielding NRs. 52   Figure 3.4. a) Schematic illustration of the growth of FePt NWs/NRs; b, c) TEM images of the FePt NRs/NWs obtained from the reaction in oleylamine at 120 °C for 2 min and 5 min, respectively. 53   Controlled evaporation of the carrier solvent from the hexane dispersion of the 50-nm Fe55Pt45 NRs led to an Fe55Pt45 NR array with the NRs parallel to each other (Figure 3.5a). This assembly pattern is energetically favored as it gives the maximum van der Waals interaction energy arising from face-face interactions.[11] X-ray diffraction (XRD) of the as-synthesized NWs or NRs shows a fcc structure of FePt (Figure 3.5b). The broad peak originates from the small diameter of the NWs or NRs (2-3 nm). Thermal annealing in an argon atmosphere at 750 °C transforms the chemically disordered fcc structure of FePt to a chemically ordered face-centered tetragonal (fct) FePt structure, as shown by the XRD pattern in Figure 3.5b. More interestingly, the XRD pattern of the annealed 200-nm Fe55Pt45 NW assembly shows a much stronger (001) peak than the (111) peak or any other peaks, indicating a partial structural alignment with the (001) planes parallel to the substrate. The in-plane magnetic hysteresis loop of the assembly shows the better squareness with the coercivity reaching 9.5 kOe (Figure 3.5c). However, the annealed assembly of 20-nm Fe55Pt45 NRs (Figure 3.5c) shows the reduced alignment as evidenced by the intensity drop of the (001) peak. Furthermore, the XRD diffraction peaks for both NWs and NRs reveal a sharp decrease in the diffraction-peak width. This result indicates that both NWs and NRs are thermally unstable under the current annealing conditions, leading to the fusion of the nanostructures into larger aggregates. The NWs may be able to retain a larger portion of the elongated nanostructures and show better texture than the NRs do after the annealing process (Figure 3.6). Our preliminary test in NR stabilization by a MgO matrix revealed that the aggregation problem could be solved. 54   Figure 3.5. a) TEM image of self-assembled 50-nm FePt NRs; b) XRD pattern of the as- synthesized (as-syn; before thermal annealing) NWs, and 200-nm NWs and 20-nm NRs annealed in Ar at 750 °C for 1 h; c) magnetic hysteresis loop of the 200-nm Fe55Pt45 NWs annealed in Ar at 750 °C for 1 h. emu=electromagnetic unit. 55   Figure 3.6. a) Schematic illustration of the morphology change of FePt NWs/NRs after thermal annealing; and b) TEM image of the FePt NRs and spherical nanoparticles obtained from the annealing of the as-synthesized NRs. Annealing condition: 500 °C for 1 hour in Ar. 56   3.4. Conclusion In conclusion, we report that controlled reduction of [Pt(acac)2] and decomposition of [Fe(CO)5] in a mixture of oleylamine and octadecene leads to a facile synthesis of FePt NWs and NRs with diameters of 2-3 nm. The length of the NWs/NRs is tunable from over 200 nm down to 20 nm by simply controlling the volume ratio of oleylamine/octadecene. The synthesis can also be extended to the preparation of CoPt NWs and NRs. With the structure confinement, these NWs or NRs may serve as unique building blocks for fabricating magnetically aligned nanomagnet arrays to support high- density magnetic information and to achieve highly sensitive magnetoresistive detection. 57   REFERENCES 1. (a) S. Sun, C. B. Murray, D. Weller, L. Folks, A. Moser, Science, 2000, 287, 1989. (b) D. Weller, M. F. Doerner, Annu. Rev. Mater. Sci., 2000, 30, 611. (c) A. Moser, K. Takano, D. T. Margulies, M. Albrecht, Y. Sonobe, Y. Ikeda, S. Sun, E. E. Fullerton, J. Phys. D, 2002, 35, R 157. (d) I. R. McFadyen, E. E. Fullerton, M. J. Carey, MRS Bull., 2006, 31, 379. (e) R. F. Service, Science, 2006, 314, 1868. (f) S. Sun, Adv. Mater., 2006, 18, 393. (g) Y. F. Xu, M. L. Yan, D. J. Sellmyer, J. Nanosci. Nanotechnol., 2007, 7, 206. 2. (a) E. F. Kneller, R. Hawig, IEEE Trans. Magn., 1991, 27, 3588. (b) R. Skomski, J. M. D. Coey, Phys. Rev. B, 1993, 48, 15812. (c) T. Schrefl, H. Kronmüller, J. Fidler, J. Magn. Magn. Mater., 1993, 127, L273. (d) H. Zeng, J. Li, J. P. Liu, Z. L. Wang, S. Sun, Nature, 2002, 420, 395. (e) S. D. Bader, Rev. Mod. Phys., 2006, 78, 1. 3. (a) T. Toda, H. Igarashi, H. Uchida, M. Watanabe, J. Electrochem. Soc., 1999, 146, 3750. (b) J. Jiang, A. Kucernak, J. Electroanal. Chem., 2002, 520, 64. (c) E. Antolini, J. R. C. Salgado, E. R. Gonzalez, J. Power Sources, 2006, 160, 957. 4. (a) C. Antoniak, J. Lindner, M. Spasova, D. Sudfeld, M. Acet, M. Farle, K. Fauth, U. Wiedwald, H.-G. Boven, P. Ziemann, F. Wilhelm, A. Rogalev, S. Sun, Phys. Rev. Lett., 2006, 97, 117201. (b) J. B. Staunton, S. Ostanin, S. S. A. Razee, B. L. Gyorffy, L. Szunyogh, B. Ginatempo, E. Bruno, Phys. Rev. Lett., 2004, 93, 257204. (c) O. Kitakami, S. Okamoto, N. Kikuchi, Y. Shimada, Jpn. J. Appl. Phys., 2003, 58   42, L455. (d) G. Brown, B. Kraczek, A. Janotti, T. C. Schulthess, G. M. Stocks, D. D. Johnson, Phys. Rev. B, 2003, 68, 052405. (e) T. Burkert, O. Eriksson, S. I. Simak, A. V. Ruban, B. Sanyal, L. Nordström, J. M. Wills, Phys. Rev. B, 2005, 71, 134411. 5. (a) S. Sun, E. E. Fullerton, D. Weller, C. B. Murray, IEEE Trans. Magn., 2001, 37, 1239. (b) M. Chen, J. P. Liu, S. Sun, J. Am. Chem. Soc., 2004, 126, 8394. (c) S. Momose, H. Kodama, T. Uzumaki, A. Tanaka, Jpn. J. Appl. Phys., 2005, 44, 1147. 6. K. Elkins, D. Li, N. Poudyal, V. Nandwana, Z. Q. Jin, K. H. Chen, J. P. Liu, J. Phys. D, 2005, 38, 2306. 7. W. Chen, J. Kim, S. Sun, S. Chen, Phys. Chem. Chem. Phys., 2006, 8, 2779. 8. M. Chen, J. Kim, J. P. Liu, H. Fan, S. Sun, J. Am. Chem. Soc., 2006, 128, 7132. 9. M. Chen, T. Pica, Y. Jiang, P. Li, K. Yano, J. P. Liu, A. K. Datye, H. Fan, J. Am. Chem. Soc., 2007, 129, 6348. 10. (a) J. Gao, C. M. Bender, C. J. Murphy, Langmuir, 2003, 19, 9065. (b) L. Gou, C. J. Murphy, Chem. Mater., 2005, 17, 3668. (c) I. Pastoriza-Santos, J. Pérez-Juste, L. M. Liz-Marzán, Chem. Mater., 2006, 18, 2465. (d) P. Zijlstra, C. Bullen, J. W. M. Chon, M. Gu, J. Phys. Chem. B, 2006, 110, 19315. 11. (a) C. Burda, X. Chen, R. Narayanan, M. A. El-Sayed, Chem. Rev., 2005, 105, 1025. (b) S. Yamamuro, K. Sumiyama, Chem. Phys. Lett., 2006, 418, 166.   Chapter 4 Dispersible Ferromagnetic FePt Nanoparticles 60   4.1. Introduction The synthesis of ferromagnetic (FM) FePt nanoparticles (NPs) has attracted much attention in magnetic NP research.[1] Such NPs, with controlled magnetic properties, have great potential for ultrahigh-density information-storage[2–4] and high-performance permanent-magnet applications.[5–9] Monodisperse FePt NPs have been produced by solution-phase chemical decomposition of iron pentacarbonyl (Fe(CO)5) and reduction of platinum acetylacetonate (Pt(acac)2),[10–13] or by co-reduction of an iron and a platinum salt.[14–19] As synthesized, the FePt NPs have a chemically disordered face-centered cubic (fcc) structure, and are superparamagnetic at room temperature. FM FePt NPs with a face-centered tetragonal (fct) structure can only be obtained by thermal annealing of fcc- FePt NPs at high temperatures (usually higher than 550 °C). This high-temperature annealing renders FM FePt NPs undispersible, and unsuitable for controlled assembly and for single-particle magnetism studies. To make dispersible FM FePt NPs, one can coat the as-synthesized FePt NPs with SiO2 before thermal annealing, and remove SiO2 after the fct FePt is formed.[20–23] Alternatively, the as-synthesized FePt NPs can be ground with a large excess of NaCl before thermal annealing – the fine grains of NaCl protect the FePt NPs from sintering at high-temperature-annealing conditions, and NaCl can be dissolved with water.[24–27] Despite these efforts, stabilizing a FM FePt-NP dispersion is still challenging. In this chapter, we report a new synthesis and stabilization strategy, as outlined in Scheme 4.1, to produce a FM FePt-NP dispersion in a hydrocarbon solvent. During the preparation, we first coat the as-synthesized fcc-FePt NPs with MgO – a basic oxide that 61   is thermally stable, with a melting point of 2000 °C, and that can react with HCl to form MgCl2 and water. MgO protects the fcc-FePt NPs from aggregation at annealing temperatures up to 800 °C, and the fcc FePt is converted to fct FePt at above 700 °C. Dissolving MgO with 0.5 M HCl in the presence of a hexane solution of hexadecanethiol (HDT) and oleic acid (OA) leads to well-dispersed FM fct-FePt NPs in hexane. The fct- FePt NPs show coercivity up to 1 T at room temperature, and a magnetization of 56.4 emu per gram of FePt. Scheme 4.1. Synthesis of fct-FePt NPs from fcc-FePt/MgO NPs. 62   4.2. Experimental 4.2.1. Synthesis of 7 nm fcc-Fe51Pt49 Nanoparticles Pt(acac)2 (0.5 mmol), oleic acid (4 mmol), and oleylamine (4 mmol) were added into a 50 mL four-neck flask that contained 10 mL of octadec-1-ene under gentle argon gas flow. The flask was heated to 120 °C at a heating rate of 6 °C min-1. The flask was maintained for 13 min at this temperature, to ensure the dissolution of Pt(acac)2. Under a blanket of argon gas, 0.20 mL of Fe(CO)5 was added. The solution was then heated to 240 °C at a heating rate of 3 °C min-1, and kept at this temperature for 1 h. The heating source was then removed, and the solution was cooled to room temperature, after which the solution was exposed to air. A black product was precipitated by adding 40 mL of ethanol, and separated by centrifugation. The dark yellow supernatant was discarded. The NPs were dispersed in 15mL of hexane, and precipitated out by adding 20 mL of ethanol followed by centrifugation. The dispersion/precipitation procedure was repeated three times. Finally, the 7 nm fcc-FePt NPs were re-dispersed in 10 mL of hexane. 4.2.2. Synthesis of fcc-Fe51Pt49/MgO Nanoparticles Mg(acac)2 (2 mmol), tetradecane-1,2-diol (4 mmol), oleic acid (4 mmol), and oleylamine (4 mmol) were mixed with 20 mL of benzyl ether and heated to 80 °C. The 7 nm FePt NPs (~100 mg) dispersed in 5mL of hexane were rapidly added into the reaction solution, within ~2 s. The solution was heated to 120 °C to remove hexane. Under nitrogen gas, the solution was heated to reflux (298 °C), and kept at this temperature for 1 63   h. The fcc-FePt/MgO NPs were separated and purified by using hexane, ethanol, and centrifugation. Finally, the particles were kept in 5 mL hexane. 4.2.3. Characterization The size, morphology, and structure of the FePt NPs were characterized using a Philips EM 420 (120 kV) and a JEOL 2010 (200 kV). X-ray powder diffraction patterns were recorded using a Bruker AXS D8-Advanced diffractometer with CuKα radiation (λ=1.5418 Å). Magnetic studies were performed using a Quantum Design Superconducting Quantum Interface Device (SQUID) with a field up to 70 kOe. NP compositions were analyzed by Oxford energy-disperse X-ray spectroscopy and inductively coupled plasma–atomic emission spectroscopy. 64   4.3. Results and Discussion The 7 nm fcc-Fe51Pt49 NPs were synthesized according to a previously published method.[12] During the synthesis, 0.5 mmol of Pt(acac)2, 4 mmol of OA, 4 mmol of oleylamine (OAm), and 10 mL of octadec-1-ene were mixed with 0.20 mL of Fe(CO)5 and heated to 240 °C for 1 h. Figure 4.1a and 1b shows the transmission electron microscopy (TEM) images of two-dimensional assemblies of the monodisperse 7 nm fcc- FePt NPs. While the monolayer assembly in Figure 4.1a illustrates a typical hexagonal close packing, the double-layer assembly in Figure 4.1b shows an anomalous array, where the second layer sits on the two-fold saddle points. This is similar to what is reported for Au NP assemblies.[28] MgO was coated on the Fe51Pt49 NPs by decomposition of Mg(acac)2 in the presence of tetradecane-1,2-diol, OA, and OAm in benzyl ether, and heated to 298 °C. Figure 4.1c is the TEM image of the 7 nm Fe51Pt49/MgO NPs with a MgO coating over the FePt surface. Thermal annealing of the fcc-FePt/MgO and fcc-FePt NPs was performed under Ar + 5% H2 at various temperatures. For the fcc-FePt/MgO NP assemblies, TEM analyses indicated that for annealing below 800 °C for less than 4 h, there was no obvious FePt morphology change in the FePt/MgO structure, and MgO formed a flower-like pattern, as shown in Figure 4.1d. However, upon annealing at 800 °C for over 6 h, these FePt/MgO NPs sintered. X-ray diffraction (XRD) analyses of the annealed FePt/MgO assemblies revealed that an fcc-to-fct structure transformation in the FePt NPs is not readily characterized until the annealing temperature reaches above 700 °C. Figure 4.2a is the XRD pattern of the fct-FePt/MgO assembly obtained from the annealing of the fcc- 65   FePt/MgO at 750 °C for 6 h. This structure-transformation temperature is much higher than the 550 °C needed for the as-synthesized fcc-FePt NPs, as previously reported.[10,11] Figure 4.2b is the XRD pattern of the as-synthesized fcc-FePt NPs annealed at 750 °C for 6 h. Comparing Figure 4.2a and b, we can see that: i) the broad diffraction pattern from the FePt/MgO NPs indicates that the FePt in the FePt/MgO structure does not experience grain growth during the high-temperature annealing process; ii) FePt in FePt/MgO adopts a fct-like structure, even though the fully ordered fct structure cannot be confirmed; iii) the diffraction peaks from the FePt/MgO appear at lower angles than the corresponding peaks from the FePt NPs, indicating that the nanostructured FePt in FePt/MgO has a slightly larger crystal-lattice spacing than bulk fct-FePt, which is most likely a result of the imperfect fcc-to-fct transformation and/or FePt NP surface effects; iv) the fcc-to-fct transition is hindered by the limit of Fe and Pt mobility within the constrained MgO structure. 66   Figure 4.1. TEM images of a) 7 nm fcc-FePt NPs, b) abnormal assembly of the 7 nm fcc- FePt, c) as-synthesized fcc-FePt/MgO NPs, and d) fct-FePt/MgO NPs obtained from thermal annealing of the fcc-FePt/MgO NPs under Ar + 5% H2 at 750 °C for 6 h. Scale bar = 20 nm. 67   Figure 4.2. XRD patterns of a) fct-FePt/MgO NP and b) fct-FePt NP assemblies obtained from the thermal annealing of the fcc-FePt/MgO and the fcc-FePt NP assemblies at 750 °C for 6 h in the Ar + 5% H2 atmosphere. 68   Magnetic measurements show that the as-synthesized fcc-FePt and fcc-FePt/MgO NPs are superparamagnetic at room temperature. After thermal annealing at 650 °C under Ar + 5% H2, the uncoated FePt assembly became FM, and the room temperature coercivity reached 2 T. Under the same annealing conditions, the FePt/MgO NPs show FM properties at low temperatures of 5 K and 100 K, but are superparamagnetic/weakly FM at room temperature, as shown in Figure 4.3a. Clearly, the annealing at 650 °C for 6 h did not completely convert the fcc-FePt into fct-FePt in the FePt/MgO structure. This is also evidenced from the XRD studies. The fct-FePt/MgO was obtained by annealing the fcc-FePt/MgO at 750 °C for 6 h. Figure 4.3b shows the hysteresis loops of the fct- FePt/MgO NPs measured at 5 K, 100 K, and 300 K. We can see that these fct-FePt/MgO NPs are FM, with Hc reaching 1.8 T (at 5 K), 1.6 T (at 100 K), and 1 T (at 300 K), respectively. By comparing the structural transformation and magnetic-property change between the uncoated FePt and the coated FePt/MgO, we can conclude that: i) MgO in FePt/MgO protects FePt from sintering at annealing temperatures up to 800 °C; ii) in a constrained MgO structure, where atom mobility is limited because of the robust MgO coating, the fcc-to-fct conversion is still possible, but is much more difficult, and requires a high annealing temperature (150 °C higher than that for the 7 nm FePt NPs in this work) and longer annealing time. 69   Figure 4.3. Hysteresis loops of the nanocrystalline FePt/MgO NPs annealed at a) 650 °C, and b) 750 °C for 6 h under Ar + 5 % H2. 70   The fct-FePt/MgO NPs prepared from the thermal annealing of fcc-FePt/MgO NPs are hardly dispersed in any solvent. Although MgO in the FePt/MgO structure can be removed by washing with dilute HCl (0.5 M), the bare fct-FePt NPs quickly aggregate. Adding poly(vinyl pyrrolidone) (PVP) to the aqueous solution during the MgO removal process did not yield any efficient on FePt-NP stabilization. To protect the fct-FePt NPs from aggregation upon MgO removal, we attempted extraction from their aqueous phase into an organic phase by a phase-transfer process. We added the fct-FePt/MgO NPs to a mixture of aqueous 0.5 M HCl and surfactant-containing hexane, to extract the fct-FePt NPs into hexane with their surface protected by the surfactant. We tested several different surfactant combinations, which included OA/OAm/hexane, HDT/hexane, and HDT/OA/hexane. We found that OA/OAm could not stabilize the fct-FePt NPs under the current extraction conditions. HDT alone offered only temporary stabilization—the NP dispersion became unstable and aggregated after 2 h. The HDT/OA produced the most efficient protection to the fct-FePt NPs. The stabilization difference between HDT and HDT/OA seems to indicate that thiol (SH) reacts with Pt, while COOH binds to Fe, on the surface of the FePt NP, and the double bond in OA is essential for the stabilization, as the saturated hydrocarbon chain tends to fold on the NP surface, which leads to strong NP interactions and aggregation. Figure 4.4a outlines the chemistry of transferring the fct-FePt/MgO NPs from their aqueous suspension to the hexane dispersion. In the figure, the bottom layer represents the aqueous solution of 0.5 M HCl for MgO removal, and the upper layer is the hexane phase, with the dispersed fct-FePt NPs coated with HDT/OA. Figure 4.4b is a photograph that shows the fct-FePt NP hexane dispersion at the top, and the colorless aqueous 71   solution at the bottom, after the FePt extraction by HDT/OA. Figure 4.4c is a TEM image of the fct-FePt NPs obtained from their hexane dispersion. By comparing with Figure 4.1a, we can see that there is only a slight morphology change before and after the annealing/MgO removal process. Figure 4.4d is the high-resolution (HR)TEM image of a single fct-FePt NP, where the (111) lattice fringe is measured to be 0.24 nm, which is larger than the 0.22 nm of the bulk fct-FePt. This is consistent with the lower angle position of the (111) diffraction peak observed in the XRD pattern shown in Figure 4.2a. Note that after MgO removal, Fe46Pt54 NPs were obtained, which indicates a small reduction in Fe composition during the acid washing and transfer process. This Fe loss is likely caused by HCl etching. Despite this loss, the fct-FePt NPs obtained from the hexane dispersion are still FM, with a coercivity that reaches near 1 T and magnetization at 56.4 emu per gram of FePt (the data was normalized by weighing the FePt after the removal of the HDT/OA surfactant at 800 °C under argon. As the thiol and the –COOH groups cannot be removed completely from the FePt surface, and the hydrocarbon chain in the surfactant may also form carbon deposits around each NP, the actual magnetization value should be slightly higher.). 72   Figure 4.4. a) Schematic illustration of the fct-FePt NP transfer from the 0.5 M HCl aqueous phase to the hexane phase. b) Photograph showing the fct-FePt NP transfer from the aqueous phase to the hexane phase (the fct-FePt NPs are obtained from the FePt/MgO NPs with magnetic properties, shown in Figure 4.3b). c) TEM image of the fct-FePt NPs from b). d) HRTEM image of a single fct-FePt NP with interfringe spacing at 0.24 nm. 73   4.4. Conclusion In summary, we have reported that a dispersion of FM fct-FePt NPs can be prepared from thermal annealing of the core/shell-structured fcc-FePt/MgO NPs. MgO in FePt/MgO protects the FePt NPs from sintering at high annealing temperatures. In a constrained MgO structure, where atom mobility is limited because of the robust MgO coating, the conversion from fcc- into fct-FePt is possible, but is more difficult compared with the non-MgO-coated FePt NPs, which require a high annealing temperature (150 °C higher). The fct-FePt NPs obtained from the annealing of the fcc-FePt NPs at 750 °C for 6 h show a room temperature coercivity of 1 T and magnetization of 56.4 emu per gram of FePt. MgO in the fct-FePt/MgO can be removed by acid (HCl) washing, and the fct- FePt NPs are protected by a surfactant combination of hexadecanethiol and oleic acid, which forms a stable hexane dispersion. Such fct-FePt NPs dispersed in hexane should serve as ideal building blocks for constructing FM nanostructures and for information/energy storage applications. 74   REFERENCES 1. S. Sun, Adv. Mater., 2006, 18, 393. 2. D. Weller, M. F. Doerner, Annu. Rev. Mater. Sci., 2000, 30, 611. 3. A. Moser, K. Takano, D. T. Margulies, M. Albrecht, Y. Sonobe, Y. Ikeda, S. Sun, E. E. Fullerton, J. Phys. D. Appl. Phys., 2002, 35, R157. 4. I. R. McFadyen, E. E. Fullerton, M. J. Carey, MRS Bull., 2006, 31, 379. 5. E. F. Kneller, R. Hawig, IEEE Trans. Magn., 1991, 27, 3588. 6. R. Skomski, J. M. D. Coey, Phys. Rev. B, 1993, 48, 15812. 7. T. Schrefl, H. Kronmüller, J. Fidler, J. Magn. Mag. Mater., 1993, 127, L273. 8. H. Zeng, J. Li, J. P. Liu, Z. L. Wang, S. Sun, Nature, 2002, 420, 395. 9. S. D. Bader, Rev. Mod. Phys., 2006, 78, 1. 10. S. Sun, C. B. Murray, D. Weller, L. Folks, A. Moser, Science, 2000, 287, 1989. 11. S. Sun, E. E. Fullerton, D. Weller, C. B. Murray, IEEE Trans. Magn., 2001, 37, 1239. 12. M. Chen, J. P. Liu, S. Sun, J. Am. Chem. Soc., 2004, 126, 8394. 13. S. Momose, H. Kodama, T. Uzumaki, A. Tanaka, Jpn. J. Appl. Phys., 2005, 44, 1147. 14. S. Sun, S. Anders, T. Thomson, J. E. E. Baglin, M. F. Toney, H. F. Hamann, C. B. Murray, B. D. Terris, J. Phys. Chem. B, 2003, 107, 5419. 15. K. E. Elkins, T. S. Vedantam, J. P. Liu, H. Zeng, S. Sun, Z. L. Wang, Y. Ding, Nano Lett., 2003, 3, 1647. 75   16. C. Liu, X. Wu, T. Klemmer, N. Shukla, X. Yang, D. Weller, A. G. Roy, M. Tanase, D. Laughlin, J. Phys. Chem. B, 2004, 108, 6121. 17. B. Jeyadevan, A. Hobo, K. Urakawa, C. N. Chinnasamy, K. Shinoda, K. Tohji, J. Appl. Phys., 2003, 93, 7574. 18. T. Iwaki, Y. Kakihara, T. Toda, M. Abdullah, K. Okuyama, J. Appl. Phys., 2003, 94, 6807. 19. Y. Kitamoto, R. Minami, Y. Shibata, T. Chikata, S. Kato, IEEE. Trans. Magn., 2005, 41, 3880. 20. S. Yamamoto, Y. Morimoto, T. Ono, M. Takano, Appl. Phys. Lett., 2005, 87, 032503. 21. S. Yamamoto, Y. Morimoto, Y. Tamada, Y. K. Takahashi, K. Hono, T. Ono, M. Takano, Chem. Mater., 2006, 18, 5385. 22. Y. Tamada, Y. Morimoto, S. Yamamoto, M. Takano, S. Nasu, T. Ono, J. Magn. Magn. Mater., 2007, 310, 2381. 23. Y. Tamada, S. Yamamoto, M. Takano, S. Nasu, T. Ono, Appl. Phys. Lett., 2007, 90, 162 509. 24. K. Elkins, D. Li, N. Poudyal, V. Nandwana, Z. Jin, K. Chen, J. P. Liu, J. Phys. D. Appl. Phys., 2005, 38, 2306. 25. D. R. Li, N. Poudyal, V. Nandwana, Z. Q. Jin, K. Elkins, J. P. Liu, J. Appl. Phys., 2006, 99, 08E 911. 26. B. A. Jones, J. D. Dutson, K. O'Grady, B. J. Hickey, D. R. Li, N. Poudyal, J. P. Liu, IEEE Trans. Magn., 2006, 42, 3066. 76   27. C. B. Rong, N. Poudyal, G. S. Chaubey, V. Nandwana, Y. Liu, Y. Q. Wu, M. J. Kramer, M. E. Kozlov, R. H. Baughman, J. P. Liu, J. Appl. Phys., 2008, 103, 07E 131. 28. D. Zanchet, M. S. Moreno, D. Ugarte, Phys. Rev. Lett., 1999, 82, 5277.     Chapter 5 From Core/Shell Structured FePt/Fe3O4/MgO to Ferromagnetic FePt Nanoparticles 78   5.1. Introduction Magnetic iron−platinum (FePt) nanoparticles (NPs) made from solution phase chemical syntheses have shown great potential for high performance permanent magnet,[1-3] high density data storage,[4-6] and highly efficient biomedicine applications.[7- 9] Their magnetic properties can be tuned not only by NP sizes but also by Fe, Pt composition and Fe, Pt atomic arrangement in the FePt alloy structure. Due to the high magnetocrystalline anisotropy and robust chemical stability, face-centered tetragonal (fct) FePt NPs are particularly interesting as models for nanomagnetism study[10,11] and as building blocks for constructing single NP information storage media.[4-6] The fct-FePt NPs are commonly made by annealing face centered cubic (fcc) FePt NPs at high temperatures, usually >550 °C.[12-14] But this high temperature annealing also results in various degrees of NP aggregation and sintering, deteriorating the quality of NP arrays. To avoid this aggregation/sintering problem, FePt NPs are either coated with SiO2[15,16] or MgO[17] or grinded with a large excess of NaCl[18,19] before high temperature annealing is applied. The controlled MgO coating is especially interesting as it not only protects FePt NPs from sintering at temperatures up to 800 °C[17] but also provides a model system for studying magnetic tunneling between FePt NPs with potentially much higher magneto- resistance than the thin film structure.[20] Further, the MgO shell can be removed readily by a dilute acid washing, giving either fcc-FePt or fct-FePt NPs with minimum surface contamination for catalytic applications.[21-23] In this chapter, we report an improved approach to fct-FePt NPs and their hexane dispersion via reductive annealing of FePt/Fe3O4/MgO followed by MgO removal. In the 79   FePt/MgO NP system reported previously,[17] the robust MgO coating is found to limit the mobility of the Fe, Pt atoms in the FePt structure, making the fcc to fct conversion so difficult that even at 750 °C for 6 h, the FePt structure is not fully fct-ordered. In the current approach, the Pt-rich fcc-FePt/Fe3O4 NPs were first made and coated with a thin layer of MgO. The FePt/Fe3O4/MgO NPs were then reduced by Ar + H2 (5%). Due to the interfacial diffusion between Pt and Fe, well-ordered fct-FePt NPs were obtained at 650 °C for 6 h, 100 °C lower than that for the formation of fct-FePt from the fcc- FePt/MgO NPs. In the FePt/Fe3O4/MgO structure, the Fe3O4 shell thickness was used to control the final FePt composition. For example, the 12 nm fct-Fe52Pt48 NPs were synthesized by annealing 7 nm/2.5 nm fcc-FePt/Fe3O4/MgO NPs. Their coercivity reached 3.2 T at 5 K and 2 T at 300 K. They could be stabilized by hexadecanethiol and oleic acid and redispersed in hexane by the MgO removal and NP extraction process.[17] The FePt NPs prepared in this work are suitable for various nanomagnetic and catalytic applications. 80   5.2. Experimental 5.2.1. Chemicals Iron pentacarbonyl (98%), oleic acid (OA) (90%), oleylamine (OAm) (>70%), 1- octadecene (90%), and 1,2-tetradecanediol (98%) were purchased from Aldrich. Platinum(II) acetylacetonate, Pt(acac)2 (98%), and magnesium(II) acetylacetonate, Mg(acac)2 (98%), were purchased from Strem. Hydrochloric acid was purchased from EMD chemicals. All syntheses were carried out under a standard airless condition using the Schlenk line. The commercially available reagents were used without further purification. 5.2.2. Synthesis of FePt/Fe3O4 Nanoparticles To get ~12 nm FePt/Fe3O4 NPs by one-pot synthesis, 0.5 mmol of Pt(acac)2, 4 mmol of OA, and 4 mmol of OAm were mixed with 10 mL of 1-octadecene under gentle argon gas flow. The mixture was heated up to 120 °C, and 0.20 mL of Fe(CO)5 was added under the blanket of argon gas. The solution was heated to 240 °C for 30 min to ensure decomposition of Fe(CO)5 and growth of FePt NPs and followed by refluxing and short time air oxidation at 320 °C for 30 min. The heating source was removed, and the solution was cooled down to room temperature. The black product was precipitated by adding 40 mL of absolute ethanol and separated by centrifugation. The dark yellow supernatant was discarded, and the precipitate was dispersed in 15 mL of hexane. By adding 20 mL of ethanol, the NPs were precipitated out again and centrifuged. This 81   procedure was repeated for two more times. Finally, the product (~200 mg) was redispersed in 10 mL of hexane in the presence of 0.05 mL of each of OA and OAm. 5.2.3. Synthesis of FePt/Fe3O4/MgO Nanoparticles FePt/Fe3O4/MgO NPs were synthesized by the following procedure. First, 2 mmol of Mg(acac)2, 4 mmol of 1,2-tetradecanediol, and 4 mmol of OA and 4 mmol of OAm were put into the four-neck flask containing 20 mL of benzyl ether. Then, the vigorously stirred solution was heated to 80 °C and kept for 5 min to dissolve the mixture completely under a gentle nitrogen gas flow. A total of ~100 mg of the as-synthesized FePt/Fe3O4 NPs dispersed in 5 mL of hexane was added quickly into the flask within ~2 s. The solution was heated to 120 °C and kept at this temperature for 20 min to ensure hexane was evaporated. Under the blanket of nitrogen gas, the solution was heated to 298 °C and was refluxed for 1 h. Finally, the heating source was removed and the solution was cooled down to room temperature. The FePt/Fe3O4/MgO NPs were separated and purified by using hexane and ethanol and centrifugation as described in the synthesis of FePt/Fe3O4 NPs. Finally, the NPs were kept in 10 mL of hexane. 5.2.4. Synthesis of fct-FePt from FePt/Fe3O4/MgO Nanoparticles To make ferromagnetic FePt NPs, the FePt/Fe3O4/MgO NPs powder was put in a porcelain combustion boat and annealed at temperatures from 550 to 800 °C and the optimum annealing condition was found to be 650 °C for 6 h. After cooled down, the magnetic powder was characterized. To make FePt NP hexane dispersion, the annealed FePt/MgO powder was transferred into a vial. The diluted HCl solution (<10% vol) and 82   hexane containing both HDT and OA were added into the vial. The suspension was sonicated for ~10 s and shaken for ~10 min. The FePt NPs stabilized by HDT and OA were extracted and dispersed in hexane. 5.2.5. FePt Nanoparticle Characterization The samples for transmission electron microscopy (TEM) images were prepared by depositing the hexane dispersion of the NPs on the amorphous carbon-coated copper grids. The NP samples for high temperature annealing were prepared on carbon type-A copper grid. TEM images were acquiredon a Philips EM 420 at 120 kV and HRTEM images were obtained on a JEOL 2010 at 200 kV. The Fe and Pt composition of the FePt NPs were measured by Oxford energy-disperse X-ray spectroscopy. The powder X-ray diffraction (XRD) patterns of the samples were collected on Bruker AXS D8-Advance diffractometer with CuKα radiation (λ = 1.5418 Å). Magnetic studies were performed on a Quantum Design Superconducting Quantum Interface Device (SQUID) with a field up to 70 kOe. The samples for diffraction and magnetic measurements were deposited on Si substrates. 83   5.3. Results and Discussion 5.3.1. Synthesis of FePt/Fe3O4 and FePt/Fe3O4/MgO Nanoparticles Pt-rich FePt/Fe3O4 NPs were synthesized from one-pot reaction of Fe(CO)5, Pt(acac)2, OA, and OAm in 1-octadecene. The mechanism for the formation of these Pt- rich FePt/Fe3O4 NPs under the current reaction conditions is similar to what has been proposed for the formation of fcc-FePt/Fe3O4,[24] but the process is controlled so that there is no significant diffusion of Fe into Pt in the reaction condition, as shown in Scheme 5.1. The Pt-rich FePt NPs are formed from the simultaneous reduction of Pt(acac)2 and partial decomposition of Fe(CO)5 at temperature < 240 °C. At a higher reaction temperature, more Fe atoms coat over the existing Pt-rich FePt NPs, forming Pt- rich FePt/Fe NPs that are further oxidized to Pt-rich FePt/Fe3O4 NPs. The amount of Fe(CO)5 is optimized so that the ratio of Fe/Pt is close to 1/1. MgO is coated over the FePt-Fe3O4 NP surface via the direct thermal decomposition of Mg(acac)2 in the condition that is similar to what has been used for the synthesis of Fe3O4 NPs.[25,26] Scheme 5.1. Synthesis of Pt-Rich FePt/Fe3O4/MgO NPs and fct-FePt/MgO NPs 84   Figure 5.1A shows the TEM image of the as-synthesized FePt/Fe3O4 NPs with average total diameter of 12 nm. The core FePt NPs are in cube-like shape. Figure 5.1B is the HRTEM image of one representative FePt/Fe3O4 NP. It can be seen that the FePt is surrounded by a polycrystalline Fe3O4 shell. The lattice fringes of the core FePt NP are not clearly identified due to the image interference by the Fe3O4 shell. The elemental analysis of the core/shell structure shows Fe:Pt = 52:48 (Figure 5.2). Figure 5.1C, D shows the TEM and HRTEM images of the as-synthesized FePt/Fe3O4/MgO NPs. The MgO shell and Fe3O4 shell are not clearly distinguished due to the light electron density and the close lattice spacings of these oxides. The measured lattices with a spacing 0.212 nm could be from either MgO (200) (lattice spacing 0.211 nm) or Fe3O4 (400) (lattice spacing 0.209 nm). 85   Figure 5.1. (A) TEM and (B) HRTEM images of the as-synthesized FePt/Fe3O4 NPs; (C) TEM and (D) HRTEM images of the as-synthesized FePt/Fe3O4/MgO NPs. 86   Figure 5.2. EDS analysis of the FePt/Fe3O4 NPs, showing an overall Fe, Pt composition at Fe:Pt = 52:48. 87   5.3.2. Reductive Annealing of the Pt-Rich fcc-FePt/Fe3O4/MgO NPs The Pt-rich fcc-FePt/Fe3O4/MgO NPs were transformed into ferromagnetic fct- FePt/MgO NPs upon reductive annealing under Ar + H2 (5%) at high temperatures for 6 h. Hydrogen reduces Fe3O4 to Fe, releasing H2O and causing defects in oxygen sites. Such defects may promote interdiffusion between Fe and Pt-rich FePt matrix, facilitating the formation of fct-FePt. This easy fct structure formation may be compared with what is observed in the ternary MFePt NPs in which different M’s are doped into the FePt matrix for decreasing the fcc to fct conversion temperature.[28-36] MgO, as a coating material, is thermally robust due to its high melting point (over 2000 °C)[37,38] and stays intact at the high temperature annealing conditions, effectively protecting the FePt NPs from sintering (Figure 5.3), as shown in Scheme 5.1. Figure 5.4A, B is the FePt NPs obtained from the annealing of the FePt/Fe3O4/MgO NPs at 650 °C for 6 h under Ar + H2 (5%). The FePt NPs (~12 nm) in MgO shell have no obvious overall size change compared to the as-synthesized FePt/Fe3O4 NPs. The HRTEM image of a single FePt/MgO NP shows that MgO is better crystallized around the FePt NP. The FePt NP lattices are not readily seen due to the MgO shell interference. 88   Figure 5.3. TEM images of FePt/Fe3O4/MgO NPs after reductive annealing for 6 hrs under Ar + H2 (5%) at (left) 600 °C, (right) 650 °C. 89   Figure 5.4. (A) TEM and (B) HRTEM (B) images of the FePt/MgO NPs obtained from the reductive annealing of the FePt/Fe3O4/MgO NPs at 650 °C for 6 h under Ar + H2 (5%). 90   5.3.3. Structural Analysis on the FePt Nanoparticles X-ray diffraction (XRD) patterns in Figure 5.5 show the as-synthesized FePt/Fe3O4 (Figure 5.5A) and FePt/Fe3O4/MgO NPs before (Figure 5.5B) and after the thermal annealing (Figure 5.5C, D) at different temperatures. The diffraction peaks from the as- synthesized Fe3O4 are weak due likely to its low crystallinity or thin coating. The chemically disordered fcc-FePt NPs show the typical (111) and (200) peaks in Figure 5.5A. The peaks around 2θ = 43° and 62° in Figure 5.5B come from MgO. After annealing at 600 °C, the (111) and (200) peaks from the FePt NPs (Figure 5.5C) are shifted to the position that are ~1° higher than that for the fcc-FePt, but the diffraction pattern does not show a clear fct phase in the FePt structure after this annealing. As the annealing temperature is increased to 650 °C, the (001) and (110) peaks emerge from the diffraction pattern (Figure 5.5D), indicating the formation of the fct-FePt at 650 °C. This annealing temperature is 100 °C lower than that (750 °C) needed for the fcc to fct conversion in the fcc-FePt/MgO NPs,[17] implying that diffusion between Fe and Pt-rich FePt facilitates the formation of fct-FePt. Magnetic hysteresis loops in Figure 5.6A,B show that the FePt NPs obtained from the annealing of the fcc-FePt/Fe3O4/MgO NPs at 600 and 650 °C for 6 h under Ar + H2 (5%) are ferromagnetic. But the coercivity values of the FePt NPs from the 600 °C annealing are at 1.6 T (5 K), 1.2 T (100 K), and 0.4 T (300 K), much smaller than the FePt NPs from the 650 °C annealing at 3.2 T (5 K), 3.0 T (100 K), and 2.0 T (300 K). More importantly, the magnetization values for the 650 °C annealed sample show little temperature dependence (Figure 5.6B). This indicates that the fct structure with large magnetocrystalline anisotropy in the 650 °C annealed FePt NPs is better formed. As a 91   comparison, the FePt/MgO NPs annealed under the same 650 °C annealing conditions are superparamagnetic at room temperature (Figure 5.7).[17] Figure 5.5. XRD patterns of (A) the as-synthesized FePt/Fe3O4 NPs and (B) the FePt/Fe3O4/MgO NPs before annealing and the FePt/MgO NPs obtained from annealing the MgO coated NPs at (C) 600 °C and (D) 650 °C for 6 h under Ar + H2 (5%). 92   Figure 5.6. Hysteresis loops of the FePt/MgO NPs obtained after the thermal annealing of the FePt/Fe3O4/MgO NPs at (A) 600 °C and (B) 650 °C for 6 h under Ar + H2 (5%). 93   Figure 5.7. TEM image (left) and hysteresis loops (right) of the FePt/MgO NPs annealed at 650 °C for 6 hrs under Ar + H2 (5%). Composition ratio Fe : Pt = 51% : 49%. 94   5.3.4. Dispersion and Characterization of fct-FePt Nanoparticles The MgO shell in the fct-FePt/MgO NPs can be easily removed by diluted HCl (<10 vol %) washing and hexane extraction in the presence of HDT and OA, as reported before.[17] The TEM image of the fct-FePt NPs (Figure 5.8A) shows the isolated NPs that have little morphology change compared to the NPs seen in Figure 5.4. Figure 5.8B is the selected area electron diffraction (SAED) pattern of the fct-FePt NPs shown in Figure 5.8A. The ring patterns are from the fct-structured FePt NPs with the (001), (111), and (200) rings clearly visible in the inner circle. Figure 5.8C, D is two representative HRTEM images of the fct-FePt NPs. In Figure 5.8C, the lattice fringe spacing is measured to be 0.383 nm, corresponding to the (001) planes (0.384 nm spacing) in the fct-FePt. The FePt NP in Figure 5.8D has a lattice spacing of 0.224 nm, close to that of the (111) planes (0.222 nm spacing) in the fct-FePt. Magnetic measurements of these NPs show the similar hysteresis behavior to that in Figure 5.6B. 95   Figure 5.8. (A) TEM image of the fct-FePt NPs from their hexane dispersion; (B) SAED of the fct-FePt NP assembly in (A); and HRTEM images of the fct-FePt NPs with the lattices seen from (C) the (001) and (D) the (111) planes in fct-FePt. 96   5.4. Conclusion This chapter demonstrates an improved synthesis of ferromagnetic fct-FePt NPs and their dispersion from the Pt-rich fcc-FePt/Fe3O4/MgO NPs. The Pt-rich fcc- FePt/Fe3O4 NPs are made by a one-pot reaction of Pt(acac)2 with Fe(CO)5 in the presence of oleic acid and oleylamine and are coated with a layer of MgO via the thermal decomposition of Mg(acac)2. The robust MgO coating prevents FePt NPs from sintering at high temperature reductive annealing conditions, and the core/shell structured FePt/Fe3O4 facilitates the fcc to fct conversion. The fct-FePt NPs show the coercivity values at 3.2 T at 5 K and 2 T at 300 K. These hard magnetic NPs can be stabilized by hexadecanethiol/oleic acid and dispersed in hexane. The synthetic strategy developed here is not limited to the FePt NP system but can be extended to the synthesis of other hard magnetic NPs of CoPt and SmCo5 as well. With the size, composition, structure, and stability control, these ferromagnetic NPs can act as the key components in 2D or 3D magnetic NP superlattice arrays for nanomagnetism studies and for ultrahigh density magnetic information storage. The FePt NPs made by removing MgO in the absence of surfactant should be chemically active for various catalytic applications. 97   REFERENCES 1. E. F. Kneller, R. Hawig, IEEE Trans. Magn., 1991, 27, 3588. 2. R. Skomski, J. M. D. Coey, Phys. Rev. B, 1993, 48, 15812. 3. H. Zeng, J. Li, Z. L. Wang, J. P. Liu, S. Sun, Nature, 2002, 420, 395. 4. D. Weller, M. F. Doerner, Annu. Rev. Mater. Sci., 2000, 30, 611. 5. S. Sun, C. B. Murray, D. Weller, L. Folks, A. Moser, Science, 2000, 287, 1989. 6. A. Moser, K. Takano, D. T. Margulies, M. Albrecht, Y. Sonobe, Y. Ikeda, S. Sun, E. E. Fullerton, J. Phys. D: Appl. Phys., 2002, 35, R157. 7. Q. A. Pankhurst, J. Connolly, S. K. Jones, J. Dobson, J. J. Phys. D: Appl. Phys., 2003, 36, R167. 8. S. Sun, Adv. Mater., 2006, 18, 393. 9. J. Gao, G. Liang, J. S. Cheung, Y. Pan, Y. Kuang, F. Zhao,B. Zhang, X. Zhang, E. X. Wu, B. Xu, J. Am. Chem. Soc., 2008, 130, 11828. 10. M. Ulmeanu, C. Antoniak, U. Wiedwald, M. Farle, Z. Frait, S. Sun, Phys. Rev. B, 2004, 69, 054417. 11. C. Antoniak, J. Lindner, M. Spasova, D. Sudfeld, M. Acet, M. Farle, K. Fauth, U. Wiedwald, H.-G. Boyen, P. Ziemann, F. Wilhelm, A. Rogalev, S. Sun, Phys. Rev. Lett., 2006, 97, 117201. 12. Z. R. Dai, S. Sun, Z. L. Wang, Nano Lett., 2001, 1, 443. 13. K. E. Elkins, T. S. Vedantam, J. P. Liu, H. Zeng, S. Sun, Y. Ding, Z. L. Wang, Nano Lett., 2003, 3, 1647. 98   14. T. S. Vedantam, J. P. Liu, H. Zeng, S. Sun, J. Appl. Phys., 2003, 93, 7184. 15. S. Yamamoto, Y. Morimoto, T. Ono, M. Takano, Appl. Phys. Lett., 2005, 87, 032503. 16. Y. Tamada, S. Yamamoto, M. Takano, S. Nasu, T. Ono, Appl. Phys. Lett., 2007, 90, 162509. 17. J. Kim, C. Rong, J. P. Liu, S. Sun, Adv. Mater., 2009, 21, 906. 18. D. Li, N. Poudyal, V. Nandwana, Z. Jin, K. Elkins, J. P. Liu, J. Appl. Phys., 2006, 99, 08E911. 19. C. Rong, N. Poudyal, G. S. Chaubey, V. Nandwana, Y. Liu, Y. Q. Wu, M. J. Kramer, M. E. Kozlov, R. H. Baughman, J. P. Liu, J. Appl. Phys., 2008, 103, 07E131. 20. S. S. P. Parkin, C. Kaiser, A. Panchula, P. M. Rice, B. Hughes, M. Samant, S. Yang, Nat. Mater., 2004, 3, 862. 21. W. Chen, J. Kim, L. Xu, S. Sun, S. Chen, J. Phys. Chem. C, 2007, 111, 13452. 22. W. Chen, J. Kim, L. Xu, S. Sun, S. Chen, Langmuir, 2007, 23, 11303. 23. W. Chen, J. Kim, L. Xu, S. Sun, S. Chen, J. Phys. Chem. C, 2008, 112, 3891. 24. M. Chen, J. P. Liu, S. Sun, J. Am. Chem. Soc., 2004, 126, 8394. 25. S. Sun, H. Zeng, J. Am. Chem. Soc., 2002, 124, 8204. 26. H. Zeng, P. M. Rice, S. X. Wang, S. Sun, J. Am. Chem. Soc., 2004, 126, 11458. 27. T. Maeda, T. Kai, A. Kikitsu, T. Nagase, J. Akiyama, Appl. Phys. Lett., 2002, 80, 2147. 28. T. Kai, T. Maeda, A. Kikitsu, J. Akiyama, T. Nagase, T. Kishi, J. Appl. Phys., 2004, 99   95, 609. 29. X. Sun, S. Kang, J. W. Harrell, D. E. Nikles, Z. R. Dai, J. Li, Z. L. Wang, J. Appl. Phys., 2003, 93, 7337. 30. Y. K. Takahashi, M. Ohnuma, K. Hono, J. Magn. Magn. Mater., 2002, 246, 259. 31. S. S. Kang, D. E. Nikles, J. W. Harrell, J. Appl. Phys., 2003, 93, 7178. 32. S. S. Kang, Z. Jia, D. E. Nikles, J. W. Harrell, IEEE Trans. Magn., 2003, 39, 2753. 33. Z. Jia, S. S. Kang, D. E. Nikles, J. W. Harrell, IEEE Trans. Magn., 2005, 41, 3385. 34. Q. Yan, T. Kim, A. Purkayastha, P. G. Ganesan, M. Shima, G. Ramanath, G. Adv. Mater., 2005, 17, 2233. 35. J. W. Harrell, D. E. Nikles, S. S. Kang, X. C. Sun, Z. Jia, S. Shi, J. Lawson, G. B. Thompson, C. Sricastava, N. V. Seetala, Scr. Mater., 2005, 53, 411. 36. A. Tomou, I. Panagiotopoulos, D. Gournis, B. Kooi, J. Appl. Phys., 2007, 102, 023910. 37. J. I. Gersten, F. W. Smith, The Physics and Chemistry of Materials; Wiley- Interscience: New York, 2001. Chapter 6 Electrocatalytic Reduction of Oxygen by FePt Alloy Nanoparticles 101   6.1. Introduction In recent years, fuel cells such as direct methanol fuel cells (DMFCs) and polymer electrolyte membrane fuel cells (PEMFCs) are attracting much attention due to their potential applications as effective power sources with high-energy efficiency and low air pollution.[1-3] To improve the performance and to reduce the costs of fuel cells, a wide variety of electrocatalysts have been prepared and evaluated for anodic and cathodic electron-transfer reactions.[4-7] Oxygen electroreduction represents a critical cathodic process in fuel cells. In the past decades, various cathode catalysts, such as single crystals of noble metals,[8,9] single crystals modified with non-noble metals,[10-12] Pt-free catalysts,[13-22] and Pt-based metal alloys, have been tested for oxygen reduction. Of these, Pt alloys (e.g., NiPt, CoPt, FePt, CrPt, etc.),[23-31] especially nanosized Pt alloys,[25,32-34] have demonstrated substantially greater activities than others. By incorporating a second metal into the Pt catalysts, the alloying process results in a favorable Pt−Pt distance for the dissociative adsorption of O2 because the base transition-metal atoms are typically smaller than Pt, leading to enhanced catalytic activity for oxygen electroreduction.[35,36] Oxygen reduction reactions (ORR) on the cathode surface have complicated reaction pathways.[6,37] The most efficient pathway is the so-called direct four-electron reduction of adsorbed oxygen to water (the subscript “ad” denotes surface adsorbed states). O2 Æ O2(ad) (1a) O2(ad) + 4H+ + 4e- Æ 2H2O (1b) 102   Another less efficient pathway involves two steps in which oxygen is first reduced to hydrogen peroxide intermediates and then electrochemically reduced to water. O2 Æ O2(ad) (2a) O2(ad) + 2H+ + 2e- Æ H2O2(ad) (2b) H2O2(ad) + 2H+ + 2e- Æ 2H2O (2c) Although considerable efforts have been devoted to the studies of cathode electrocatalyts, in most cases a measurable O2 reduction current is detected only at potentials that are much more negative than the standard potential of +1.229 V (vs the normal hydrogen electrode, NHE). The high overpotentials arise primarily because of the sluggish electron-transfer kinetics. In recent years, numerous mechanistic studies have been carried out that are geared toward a fundamental understanding of the reaction dynamics for catalytic activity improvement and optimization.[24,35,38-42] It is usually believed that, in order to enhance the ORR activity of Pt-based alloy catalysts, adsorption of OH on the second metal must be facilitated, whereas the adsorption on the Pt sites should be diminished, since high coverage of adsorbed OH species on the Pt surface has been proved to inhibit the oxygen reduction reactions.[8,35,39,43] In previous X-ray absorption near-edge structure studies, binary alloys such as CrPt and FePt have been found to exhibit these favorable chemisorption characteristics of OH species.[35] Of these, FePt bulk alloys have been studied extensively both experimentally and theoretically.[23,31,35,38,40,44-46] They are typically prepared by the sputtering method and exhibit higher electrocatalytic activities for ORR than the pure Pt metal[23] as well as other binary alloys such as NiPt, MnPt, and CoPt.[23,35] In these studies, the alloy composition has been found to be one of the important parameters that dictate the 103   electrocatalytic performance.[23] Yet, so far no ORR studies have been carried out with nanosized FePt alloy particles. It is well-known that nanoparticles exhibit a substantially higher density of active atomic steps, edges, and kinks that may result in higher catalytic activities than the bulk counterparts.[47] Thus, in this paper, a series of FexPt100-x nanoparticles (x = 63, 58, 54, 42, 15, and 0) were prepared by a chemical reduction method and their electrocatalytic activities for ORR were examined and compared by cyclic and rotating disk voltammetry (RDV). Because of the similarity of the particle size and dispersity, the core size effects[45,48-50] on the catalytic activity were minimized, and the results indicated that FePt nanoparticles with the Fe/Pt atomic ratio around 1:1 appeared to exhibit the optimal composition for ORR among the series of nanoparticle catalysts in the present study. 104   6.2. Experimental 6.2.1. Materials Perchloric acid (HClO4, Fisher, 99.999%) was used as received. Water was supplied by a Barnstead Nanopure water system (18.3 MΩ cm). Ultrapure N2 and O2 were used for the deaeration of the electrolytes and oxygen reduction reaction, respectively. 6.2.2. Preparation of FexPt100-x Nanoparticles The synthesis and characterization of the FexPt100-x nanoparticles stabilized by oleylamine and oleic acid have been described previously.[51,52] In a typical reaction, under a gentle flow of N2, Pt(acetylacetonate)2, 1,2-hexadecanediol, dioctylether were mixed at room temperature and heated to 100 °C. Oleic acid, oleyl amine, and different amounts of Fe(CO)5 were added and the mixture was heated to reflux (297 °C) for 30 min. The heat source was removed, and the reaction mixture was cooled down to room temperature. The black product was precipitated by adding ethanol and separated by centrifugation. The supernatant was discarded and the black precipitate was dispersed in hexane in the presence of oleic acid and oleylamine. Then, ethanol was added to the dispersion and the suspension was centrifuged again. The precipitate was redispersed in hexane. From transmission electron microscopy (TEM) measurements, the particles were found to exhibit an average core diameter of 2−3 nm with a very narrow size distribution 105   (a representative TEM micrograph for the Pt particles was shown in Figure 6.1.). The particle size and composition were summarized in Table 6.1. Figure 6.1. Representative TEM micrograph of Pt nanoparticles. The scale bar is 20 nm. The average core size is 2.78 ± 0.95 nm. TEM images of other FePt particles have been reported earlier.[53] 106   Table 6.1. Core Size and Elemental Composition of the FexPt100-x Nanoparticles53 core compositiona Pt Fe15Pt85 Fe42Pt58 Fe54Pt46 Fe58Pt42 Fe63Pt37 core size (nm)b 2.78 ± 0.95 2.47 ± 0.56 2.35 ± 0.18 2.52 ± 0.81 2.48 ± 0.38 2.42 ± 0.61 a Particle core compositions were evaluated by energy-dispersive X-ray analysis. b Particle core sizes were derived from TEM measurements. 6.2.3. Electrochemistry A glassy carbon (GC) disk electrode (Bioanalytical Systems, 3.0 mm in diameter) was first polished with alumina slurries (0.05 μm) and then cleaned by sonication in 0.1 M HNO3, H2SO4, and Nanopure water for 10 min successively. Eight μL of the FexPt100-x nanoparticles dissolved in CH2Cl2 (1.0 mg/mL) was then dropcast onto the clean GC electrode surface by a Hamilton microliter syringe (the resulting electrodes were denoted as FexPt100-x/GC). The particle films were dried by a gentle nitrogen stream for 2 min. The organic protecting ligands were then removed by oxidation in an ultraviolet ozone (UVO) chamber (Jelight Company, Inc., Model 42) for about 15 min (the effectiveness of UVO removal of the organic layers was demonstrated by voltammetric measurements in Figure 6.2).[52-54] The particle film was then rinsed with excessive ethanol and nanopure water to remove loosely bound particles and remaining organic deposits before being immersed into an electrolyte solution for electrochemical studies. All electrochemical experiments were performed in a single-compartment glass cell using a standard three-electrode 107   configuration. A Ag/AgCl (in 3 M NaCl, aq) (Bioanalytical Systems, MF-2052) and a Pt coil were used as the reference and counter electrodes, respectively. All electrode potentials in the present study were referred to this Ag/AgCl reference electrode. Cyclic voltammetry (CV) and RDV were carried out by using a computer-controlled Bioanalytical Systems (BAS) Electrochemical Analyzer (Model 100B). Oxygen reduction reactions were examined by first bubbling the electrolyte solution with ultrahigh purity oxygen for at least 15 min and then blanketing the solution with an oxygen atmosphere during the entire experimental procedure. All electrochemical experiments were carried out at room temperature. Figure 6.2. Cyclic voltammograms of the glassy carbon electrode (GC) (red line), Pt nanoparticles modified GC electrode (Pt/GC) before (green line) and after (black line) UVO treatment in 0.1 M HClO4. Potential scan rate 0.1 V/s. 108   6.3. Results and Discussion 6.3.1. Cyclic Voltammetry Figure 6.3 shows the cyclic voltammograms of the FexPt100-x/GC electrodes (x = 63, 58, 54, 42, 15, and 0) in 0.1 M HClO4 deaerated by ultrahigh purity nitrogen. The voltammetric features are all very similar to those of a polycrystalline Pt electrode:[55,56] the adsorption and desorption of hydrogen within the potential range of −0.25 to 0.0 V, the double-layer capacitance region between 0.0 and +0.7 V, the formation of Pt oxides at potentials more positive than +0.7 V, and the reduction of Pt oxides in the cathodic potential scan. Such voltammetric features have been observed with other Pt-based alloys such as CoPt, NiPt, and FePt, and they are ascribed to the formation of a “Pt skin” on the catalyst surface.[23-25,52-54,57,58] This Pt surface enrichment is caused by Pt dissolution from the alloys and then redeposition and rearrangement on the catalyst surface during potential cycling in acidic electrolytes. From these voltammetric measurements (Figure 6.3), the active (Pt) surface areas were then quantitatively evaluated on the basis of the charge for the oxidation of surface-adsorbed hydrogen by assuming that hydrogen desorption yields 210 μC per cm2 of the Pt surface area. Table 6.2 summarizes the surface areas of the six FexPt100-x/GC electrodes. It can be seen that the effective (Pt) surface area increases with increasing Pt content in the particles, suggesting the formation of a thicker/larger Pt skin on particles with a higher Pt concentration. Note that although the Pt alloys display voltammetric features consistent with those of (polycrystalline) Pt, the catalytic properties of the Pt skins might differ drastically from those of a pure Pt surface, which has been ascribed to the electronic 109   effect of the intermetallic bonding of the base metal-rich inner layers with the surface Pt atoms.[24,25,31,59,60] Figure 6.3. Cyclic voltammograms of the FexPt100-x/GC electrodes (x = 63, 58, 54, 42, 15, and 0) after UVO treatment in 0.1 M HClO4. The electrolyte solution was deaerated by ultrahigh-purity nitrogen for 15 min and protected by a nitrogen atmosphere during the entire experimental procedure. Current density was calculated by normalizing the voltammetric currents to the surface area of the glassy carbon electrode. Potential scan rate was 0.1 V/s. 110   Table 6.2. Kinetic Parameters for O2 Reduction at FexPt100-x Nanoparticle- Functionalized GC Electrodes FexPt100-x Fe63Pt37 Fe58Pt42 Fe54Pt46 Fe42Pt58 Fe15Pt85 Pt Pt surface area (cm2)a 0.016 0.026 0.034 0.037 0.040 0.051 ORR onset potential (V vs Ag/AgCl)b 0.49 0.50 0.56 0.61 0.57 0.46 no. electron transfer (n)c 3.69 4.06 3.60 3.60 3.83 3.90 reaction rate constant at +0.32 V 3.70 × 10-3 1.47 × 10-3 0.90 × 10-3 6.92 × 10-3 6.93 × 10-3 7.53 × 10-3 (k, cm/s)d a On the basis of the charge for the oxidation of hydrogen adsorbed onto the Pt surfaces (Figure 6.3). b Estimated from Figures 6.4 and 6.8. c The numbers of electron transferred for ORR were calculated from eq 4b. d The reaction rate constants were derived by using eq 4c. 111   Figure 6.4 shows the cyclic voltammograms for the FexPt100-x/GC electrodes in an O2 saturated 0.1 M HClO4 aqueous solution at the potential scan rate of 0.1 V/s. Note that the currents have been normalized to the effective Pt surface areas estimated from Figure 6.3 (Table 6.2). The observed cathodic current is ascribed to O2 reduction on the particle surfaces since the GC substrate is electrochemically inert to O2 reduction. The peak current density increases linearly with the square root of potential scan rate (not shown), suggesting that the O2 reduction is under diffusion control. It can be seen from Figure 6.4 that the current density and onset potential of O2 reduction vary with the composition of the particles. For instance, for the pure Pt nanoparticles, the onset potential for O2 reduction is about +0.46 V, and a current peak can be observed at −0.035 V; whereas for Fe42Pt58, the onset potential shifts positively to +0.61 V, and a peak appears at ca. +0.23 V. Results for other particles fall in the intermediate range (Table 6.2). These observations suggest that the FePt nanoalloy catalysts with a “Pt-skin” actually behave more favorably in oxygen reduction than the pure Pt particles. For comparison, by using PdPt alloy particles,[61] He and Crooks observed a similar variation of the onset potential with the elemental composition of the nanoparticles, which ranged from ca. +0.45 to +0.65 V (vs Ag/AgCl); and the most positive onset potential was found with the Pt-rich Pd17Pt83 particles. In contrast, in another study using ordered arrays of Pt and CoPt nanoparticles,[62] Kumar and Zou found that the ORR onset potentials remained virtually invariant at +0.45 V (vs Ag/AgCl), regardless of the particle elemental composition. The discrepancy most likely arises from the different structures of the nanoparticle catalysts that are prepared by different synthetic protocols. 112   Figure 6.4. Cyclic voltammograms of the FexPt100-x/GC electrodes (same as those in Figure 6.3) in 0.1 M HClO4 saturated with oxygen. Current density was calculated by normalizing the voltammetric currents to the effective platinum surface areas which were estimated from Figure 6.1. Potential scan rate was 0.1 V/s. 113   The compositional effects of the FexPt100-x nanoparticles on the catalytic activity can be further illustrated by the ORR peak current density, which is summarized in Figure 6.5. It can be seen that the O2 reduction current density first increases and then decreases with the Pt content in FexPt100-x nanoparticles; and the FePt particles with the Fe:Pt atomic ratio around 1:1 exhibit the largest current density for O2 reduction. For instance, for the Fe54Pt46 and Fe42Pt58 particles, the peak current density is 2.97 and 3.15 mA/cm2, respectively; whereas it decreases markedly to 0.60 mA/cm2 for the Fe63Pt37 particles and 0.85 mA/cm2 for the Pt particles. This is in good agreement with previous studies of O2 reduction on FePt bulk alloys prepared by the sputtering method[23] but is markedly different from the observation with PdPt alloy nanoparticles,[61] where the maximum activity for ORR occurs at a Pd:Pt ratio of 1:5. It has been proposed that the addition of an early transition metal (M) to Pt changes the geometric (Pt−Pt bond distance and coordination number) and/or electronic structures (Pt−OH bond energy) of Pt.[24,27] Consequently, for M−Pt catalysts, the catalytic properties for ORR are strongly dependent on the type and concentration of M in the (sub)surface atomic layers. 114   Figure 6.5. Peak current density of O2 reduction as a function of the FexPt100-x nanoparticle composition. Data were obtained from Figure 6.4. 115   6.3.2. Rotating Disk Voltammetry To further examine the electrocatalytic activity, the kinetics of oxygen reduction at the FexPt100-x nanoparticles was also examined as a function of the catalyst composition by RDV. The rotating disk electrodes (RDEs) were the same as those used in the cyclic voltammetric studies (Figures 6.3 and 6.4). Figures 6.6A and 6.7A show a series of RDE voltammograms of oxygen reduction at the Pt/GC and Fe42Pt58/GC electrodes, respectively, at different rotation rates in a 0.1 M HClO4 solution saturated with O2 (DC ramp 20 mV/s). Again, the currents were all normalized to the effective Pt surface areas as summarized in Table 6.2. Note that, between +0.30 and +0.55 V, the cathodic currents were under mixed kinetic diffusion control, and at more negative potentials, the oxygen reduction was limited by diffusion, as reflected by the linearity of the Koutecky−Levich plots in Figures 6.6B and 6.7B. On both electrodes (Figure 6.6A and 6.7A), the currents in the hydrogen adsorption/desorption potential region (i.e., below −0.05 V) exhibited a slight decrease with increasingly negative potential. Similar results have also been observed previously with Pt single crystal and Pt alloy electrodes,[25,63] which are ascribed to the blocking of the Pt sites by hydrogen adsorption leading to impeded dissociation of the O−O bond and hence peroxide production (i.e., incomplete reduction of oxygen, eq 2b). 116   Figure 6.6. (A) RDE voltammograms for the Pt/GC electrode in oxygen-saturated 0.1 M HClO4 aqueous solution at different rotation rates (shown as figure legends). Current density was calculated by normalizing the voltammetric currents to the effective platinum surface areas which were estimated from Figure 6.3. DC ramp 20 mV/s. (B) The corresponding Koutecky−Levich plots (J-1 vs ω-0.5) at different electrode potentials. Lines are the linear regressions. 117   Figure 6.7. (A) RDE voltammograms for the Fe42Pt58/GC electrode in oxygen-saturated 0.1 M HClO4 aqueous solution at different rotation rates (shown as figure legends). Current density was calculated by normalizing the voltammetric currents to the effective platinum surface areas, which were estimated from Figure 6.3. DC ramp 20 mV/s. (B) The corresponding Koutecky−Levich plots (J-1 vs ω-0.5) at different electrode potentials. Lines are the linear regressions. 118   Additionally, at the same electrode potentials and rotation rates, the current density is the largest with the Fe42Pt58/GC electrode among the series of electrocatalysts under study. Figure 6.8 depicts the representative RDE voltammograms (at 1600 rpm) for the six electrodes. For instance, at the electrode potential of +0.20 V, the current density for the six electrodes was −0.56 mA/cm2 (Fe63Pt37), −1.84 mA/cm2 (Fe58Pt42), −2.20 mA/cm2 (Fe54Pt46), −3.91 mA/cm2 (Fe42Pt58), −3.55 mA/cm2 (Fe15Pt85), and −1.64 mA/cm2 (Pt), respectively. Overall, the variation of the current density with the particle composition is very similar to that observed voltammetrically (Figure 6.5). The same conclusion can be reached by further comparison of the onset potential (Figure 6.8 and Table 6.2) where the Fe42Pt58 particles again exhibit the most positive onset potential for ORR. 119   Figure 6.8. Representative RDE voltammograms of the six electrodes in oxygen-saturated 0.1 M HClO4 aqueous solution. Rotation rate was 1600 rpm, and DC ramp was 20 mV/s. Other experimental conditions were the same as those in Figure 6.4. 120   It should be noted that the compositional effect on the catalytic performance as observed in the above voltammetric and RDE measurements can be rationalized by the oxygen reduction mechanism. In this, the first electron-transfer process for the adsorbed oxygen molecules represents the rate-determining step.[23,25] Pt(O2(ad)) + e- Æ Pt(O2(ad)-) (3) For FePt nanoparticles, O2 is adsorbed onto the Pt surface sites in a linear or bridge- bonded configuration, where electrons are donated from the filled O2 orbitals to the empty orbitals of Pt surface atoms by σ overlap. At the same time, Pt back-donates electrons from the filled d orbitals to the empty O2 antibonding orbitals (π*). That is, O2 is adsorbed onto the Pt surface by strong σ- and π-bonding interactions. Previous studies[23,35] have showed that Pt alloys by the addition of a second metal (e.g., Fe, Co, Ni, etc.) exhibit an increase of the surface d-vacancy, which facilitates electron donation from O2 to surface Pt and hence strong interactions between Pt and O2. The resulting enhancement of oxygen adsorption and weakening of the O−O bond lead to fast scission of the bond and consequently enhanced ORR activities. However, with increasing Fe content in the FePt nanoparticles, excessive d-vacancy might actually diminish the back- donation of electrons from Pt to oxygen and thus reduce the activity for ORR.[23] As a result, too much or too little Fe in the alloy particles will weaken the ORR catalytic activity, i.e., the ORR current is anticipated to exhibit peak-shape dependence on the particle composition as observed above (e.g., Figure 6.5). Moreover, in comparison with previous ORR studies using FePt bulk alloys, the nanoparticles in the present studies exhibit much larger current density. For example, at 121   the rotation rate of 1600 rpm, the limiting current density was about −3.0 mA/cm2 with bulk alloys,[23] whereas it was ca. −5.4 mA/cm2 for the Fe42Pt58 particles (Figure 6.7A). Further insights into the ORR dynamics were then obtained from the analyses of the Koutecky−Levich plots (J-1 vs ω-1/2). Figures 6.6B and 6.7B depict some representative plots at various electrode potentials for the Pt/GC and Fe42Pt58/GC electrodes, respectively. It can be seen that the slopes of the linear regressions remain approximately constant over the potential range of +0.02 to +0.32 V, indicating consistent numbers of electron transfer for ORR at different electrode potentials. The linearity and parallelism of the plots are usually taken as a strong indication of a first- order reaction with respect to dissolved O2, where the observed current can be expressed as where J is the measured current density, JK and JD are the kinetic and diffusion limiting current density, respectively, ω is the electrode rotation rate, n is the overall number of electron transfer, F is the Faraday constant, CO is the bulk concentration of O2 dissolved in the electrolyte, DO is the diffusion coefficient for O2, and ν is the kinematic viscosity of the electrolyte. So the plots of J-1 vs ω-1/2 are anticipated to yield straight lines with the intercept corresponding to JK (eq 4c) and the slopes reflecting the so-called B factors (eq 4b). From the latter, the numbers of electron transfer for ORR can then be estimated, by using the literature data for CO = 1.26 × 10-3 M,[64] DO = 1.93 × 10-5 cm2/s, and ν = 1.009 122   10-2 cm2/s.[65,66] Table 6.2 summarizes the results for the six electrodes where the number of electron transfer was all close to 4, suggesting that oxygen reduction on the FexPt100-x nanoparticle surface proceed by the efficient four-electron reaction pathway (eq 1). From the y-axis intercepts of the linear regressions of the Koutecky−Levich plots (Figures 6.6B and 6.7B), the kinetic- limiting current density (JK) can be quantitatively evaluated. Figure 6.9 shows the corresponding Tafel plots for the six electrodes, where the overall shape is clearly similar despite the different elemental composition of the nanoparticle catalysts. This again implies that the oxygen reduction follows the same mechanism on these six different electrocatalysts. Additionally, in each Tafel plot, there exist two linear regions with distinctly different slopes, with a transition potential at ca. +0.18 V. Such a variation of the slope in the Tafel plot has been attributable to the variation of the adsorption (surface coverage) of the reaction intermediates (OHad) and specifically adsorbed electrolyte anions with electrode potentials that consequently affects the adsorption of the O2 molecules.[23,24,61] 123   Figure 6.9. Tafel plots of the kinetic-limiting current density (JK) for the six electrodes. Symbols are experimental data obtained from the linear regressions of the Koutecky−Levich plots as exemplified in Figures 6.4B and 6.5B. Lines are the corresponding linear regressions. Horizontal dotted line denotes the transition potential separating the two linear regions of different slopes. 124   Additionally, on the basis of eq 4c, the reaction rate constants (k) can be assessed in a quantitative fashion. Table 6.2 lists the ORR rate constants at the different FexPt100- x/GC electrodes (at +0.32 V vs Ag/AgCl), which are all of the order of 10-3 cm/s, very comparable to the values observed with other catalysts such as Pt−NbPOx supported on multiwalled carbon nanotubes,[67] and quinone-[68] or anthraquinone-modified[69] GC electrodes. Additionally, it can be seen that the ORR rate constant increases with increasing Pt content in the FexPt100-x nanocatalysts with a drastic jump at x = 42, and at higher Pt contents, the increment becomes very small. Such a distinction is also manifested in Figure 6.9 where the Tafel plots are essentially divided into two groups. At the same electrode potentials, the kinetic currents at x ≤ 42 (Fe42Pt58, Fe15Pt85, and Pt) are substantially greater than those at x ≥ 54 (Fe63Pt37, Fe58Pt42, and Fe54Pt46), again, signifying that the Fe42Pt58 particles may represent the optimal composition for ORR catalysis among the series of electrocatalysts under study. 125   6.4. Conclusion The catalytic activities of the FexPt100-x nanoparticles for oxygen reduction were examined and compared by using CV and RDV in an acidic electrolyte. It was found that the formation of a “Pt skin” on the alloy particle surface rendered them effective ORR catalysts as compared to the pure Pt particles. In addition, the catalytic activity for oxygen reduction was strongly dependent on the composition of the FexPt100-x particles. Among the series of catalysts in the present study, Fe42Pt58 particles exhibited the best performance for O2 reduction with the most positive onset potential and maximum reduction current density. The variation of the ORR activity with particle composition was explained by the influence of the Fe content on the electronic structure of Pt active sites and consequently the Pt−O bonding interactions. The direct four-electron reduction of adsorbed oxygen was derived from the kinetic studies using the rotating disk electrode technique. The reaction rate constant of O2 reduction on FexPt100-x alloy nanoparticles was found to exhibit a drastic increase when x decreased to 42, and further increase of the Pt content in the nanoparticle electrocatalysts only led to a very small increment. These studies suggest that the Fe42Pt58 particles might represent an optimal composition among the series of nanoparticle catalysts under the present study for oxygen reduction. 126   REFERENCES 1. S. Wasmus, A. J. Kuver, Electroanal. Chem., 1999, 461, 14. 2. R. A. Lemons, J. Power Sources, 1990, 29, 251. 3. B. C. H. Steele, A. Heinzel, Nature, 2001, 414, 345. 4. R. Parsons, T. Vandernoot, J. Electroanal. Chem., 1988, 257, 9. 5. D. Cameron, R. Holliday, D. Thompson, J. Power Sources, 2003, 118, 298. 6. N. M. Markovic, P. N. Ross, Surf. Sci. Rep., 2002, 45, 121. 7. J. Zhang, K. Sasaki, E. Sutter, R. R. Adzic, Science, 2007, 315, 220. 8. J. X. Wang, N. M. Markovic, R. R. Adzic, J. Phys. Chem. B, 2004, 108, 4127. 9. H. Inoue, S. R. Brankovic, J. X. Wang, R. R. Adzic, Electrochim. Acta, 2002, 47, 3777. 10. J. Zhang, Y. Mo, M. B. Vukmirovic, R. Klie, K. Sasaki, R. R. Adzic, J. Phys. Chem. B, 2004, 108, 10955. 11. M. H. Shao, T. Huang, P. Liu, J. Zhang, K. Sasaki, M. B. Vukmirovic, R. R. Adzic, Langmuir, 2006, 22, 10409. 12. S. Mukerjee, S. Srinivasan, J. Electroanal. Chem., 1993, 357, 201. 13. Y. Shimizu, K. Uemura, H. Matsuda, N. Miura, N. Yamazoe, J. Electrochem. Soc., 1990, 137, 3430. 14. J. Prakash, H. Joachin, Electrochim. Acta, 2000, 45, 2289. 15. A. Vanderputten, A. Elzing, W. Visscher, E. Barendrecht, J. Electroanal. Chem., 1986, 205, 233. 127   16. A. Vanderputten, A. Elzing, W. Visscher, E. Barendrecht, J. Chem. Soc., Chem. Commun., 1986, 477. 17. K. Wiesener, Electrochim. Acta, 1986, 31, 1073. 18. S. Gupta, D. Tryk, I. Bae, W. Aldred, E. Yeager, J. Appl. Electrochem., 1989, 19, 19. 19. G. Wei, J. S. Wainright, R. F. Savinell, J. New Mater. Electrochem. Systems, 2000, 3, 121. 20. A. L. Bouwkamp-Wijnoltz, W. Visscher, J. A. R. van Veen, S. C. Tang, Electrochim. Acta, 1999, 45, 379. 21. K. Sawai, D. Uda, Chem. Lett., 2006, 35, 1166. 22. K. Sawai, D. Uda, K. Shirai, Electrochemistry, 2007, 75, 163. 23. T. Toda, H. Igarashi, H. Uchida, M. Watanabe, J. Electrochem. Soc., 1999, 146, 3750. 24. V. Stamenkovic, T. J. Schmidt, P. N. Ross, N. M. Markovic, J. Phys. Chem. B, 2002, 106, 11970. 25. U. A. Paulus, A. Wokaun, G. G. Scherer, T. J. Schmidt, V. Stamenkovic, V. Radmilovic, N. M. Markovic, P. N. Ross, J. Phys. Chem. B, 2002, 106, 4181. 26. U. A. Paulus, A. Wokaun, G. G. Scherer, T. J. Schmidt, V. Stamenkovic, N. M. Markovic, P. N. Ross, Electrochim. Acta, 2002, 47, 3787. 27. V. Stamenkovic, T. J. Schmidt, P. N. Ross, N. M. Markovic, J. Electroanal. Chem., 2003, 554, 191. 28. G. Tamizhmani, G. A. Capuano, J. Electrochem. Soc., 1994, 141, 968. 128   29. F. Colmati, E. Antolini, E. R. Gonzalez, J. Power Sources, 2006, 157, 98. 30. B. C. Beard, P. N. Ross, J. Electrochem. Soc., 1986, 133, 1839. 31. V. R. Stamenkovic, B. S. Mun, M. Arenz, K. J. J. Mayrhofer, C. A. Lucas, G. F. Wang, P. N. Ross, N. M. Markovic, Nat. Mater., 2007, 6, 241. 32. H. Yang, W. Vogel, C. Lamy, N. Alonso-Vante, J. Phys. Chem. B, 2004, 108, 11024. 33. M. K. Min, J. H. Cho, K. W. Cho, H. Kim, Electrochim. Acta, 2000, 45, 4211. 34. J. Zhang, F. H. B. Lima, M. H. Shao, K. Sasaki, J. X. Wang, J. Hanson, R. R. Adzic, J. Phys. Chem. B, 2005, 109, 22701. 35. G. A. Camara, E. A. Ticianelli, S. Mukerjee, S. J. Lee, J. McBreen, J. Electrochem. Soc., 2002, 149, A748. 36. V. Jalan, E. J. Taylor, J. Electrochem. Soc., 1983, 130, 2299. 37. E. Yeager, Electrochim. Acta, 1984, 29, 1527. 38. A. B. Anderson, J. Roques, S. Mukerjee, V. S. Murthi, N. M. Markovic, V. Stamenkovic, J. Phys. Chem. B, 2005, 109, 1198. 39. J. L. Zhang, M. B. Vukmirovic, K. Sasaki, A. U. Nilekar, M. Mavrikakis, R. R. Adzic, J. Am. Chem. Soc., 2005, 127, 12480. 40. Y. Xu, A. V. Ruban, M. Mavrikakis, J. Am. Chem. Soc., 2004, 126, 4717. 41. Y. X. Wang, P. B. Balbuena, J. Phys. Chem. B, 2005, 109, 18902. 42. J. K. Norskov, J. Rossmeisl, A. Logadottir, L. Lindqvist, J. R. Kitchin, T. Bligaard, H. Jonsson, J. Phys. Chem. B, 2004, 108, 17886. 43. P. B. Balbuena, D. Altomare, N. Vadlamani, S. Bingi, L. A. Agapito, J. M. 129   Seminario, J. Phys. Chem. A, 2004, 108, 6378. 44. V. S. Murthi, R. C. Urian, S. Mukerjee, J. Phys. Chem. B, 2004, 108, 11011. 45. J. T. Hwang, J. S. Chung, Electrochim. Acta, 1993, 38, 2715. 46. V. R. Stamenkovic, B. S. Mun, K. J. J. Mayrhofer, P. N. Ross, N. M. Markovic, J. Am. Chem. Soc., 2006, 128, 8813. 47. N. Tian, Z.-Y. Zhou, S.-G. Sun, Y. Ding, Z. L. Wang, Science, 2007, 316, 732. 48. J. A. Poirier, G. E. Stoner, J. Electrochem. Soc., 1994, 141, 425. 49. Y. Takasu, N. Ohashi, X. G. Zhang, Y. Murakami, H. Minagawa, S. Sato, K. Yahikozawa, Electrochim. Acta, 1996, 41, 2595. 50. M. Peuckert, T. Yoneda, R. A. D. Betta, M. Boudart, J. Electrochem. Soc., 1986, 133, 944. 51. S. H. Sun, C. B. Murray, D. Weller, L. Folks, A. Moser, Science, 2000, 287, 1989. 52. W. Chen, J. Kim, S. Sun, S. W. Chen, Phys. Chem. Chem. Phys., 2006, 8, 2779. 53. W. Chen, J. Kim, S. Sun, S. W. Chen, Langmuir, 2007, 23, 11303. 54. W. Chen, J. Kim, L.-P. Xu, S. Sun, S. W. Chen, J. Phys. Chem. C, 2007, 111, 13452. 55. W. Chen, S. G. Sun, Z. Y. Zhou, S. P. Chen, J. Phys. Chem. B, 2003, 107, 9808. 56. Y. E. Seidel, R. Lindstrom, Z. Jusys, J. Cai, U. Wiedwald, P. Ziemann, R. J. Behm, Langmuir, 2007, 23, 5795. 57. M. Watanabe, K. Tsurumi, T. Mizukami, T. Nakamura, P. Stonehart, J. Electrochem. Soc., 1994, 141, 2659. 58. H. Igarashi, T. Fujino, Y. M. Zhu, H. Uchida, M. Watanabe, Phys. Chem. Chem. 130   Phys., 2001, 3, 306. 59. U. Bardi, B. C. Beard, P. N. Ross, J. Catal., 1990, 124, 22. 60. U. Bardi, A. Atrei, E. Zanazzi, G. Rovida, P. N. Ross, Vacuum, 1990, 41, 437. 61. H. C. Ye, R. M. Crooks, J. Am. Chem. Soc., 2007, 129, 3627. 62. S. Kumar, S. Z. Zou, Electrochem. Commun., 2006, 8, 1151. 63. N. Markovic, H. Gasteiger, P. N. Ross, J. Electrochem. Soc., 1997, 144, 1591. 64. A. Schumpe, I. Adler, W. D. Deckwer, Biotechnol. Bioeng., 1978, 20, 145. 65. N. A. Anastasijevic, Z. M. Dimitrijevic, R. R. Adzic, Electrochim. Acta, 1986, 31, 1125. 66. N. M. Markovic, H. A. Gasteiger, B. N. Grgur, P. N. Ross, J. Electroanal. Chem., 1999, 467, 157. 67. J.-S. Huang, X.-G. Zhang, J.-M. Luo, Y. Sun, J. Yang, J. Solid State Electrochem., 2008, 12, 113. 68. K. Vaik, A. Sarapuu, K. Tammeveski, F. Mirkhalaf, D. J. Schiffrin, J. Electroanal. Chem., 2004, 564, 159. 69. A. Sarapuu, K. Vaik, D. J. Schiffrin, K. Tammeveski, J. Electroanal. Chem., 2003, 541, 23.   Chapter 7 Electro-Oxidation of Formic Acid Catalyzed by FePt Nanoparticles 132   7.1. Introduction There has been an increasing interest in the development of efficient fuel cells due to the need for alternative energy sources with high energy density, low operating temperature and low environmental pollution. For fuel cells, aqueous solutions of formic acid and methanol represent two potential, attractive energy sources because of the ease of handling, transportation and storage in comparison to those of gaseous or liquid hydrogen.[1] Previous studies have showed that formic acid can be oxidized at less positive potentials than methanol and that crossover of formic acid through the polymer membrane is lower than that of methanol.[2-4] Thus, a great deal of research effort has been focused on the electrochemical properties of formic acid, which serves as an important model system for studying electrochemical oxidation of small organic molecules.[5-13] Fuels based on the oxidation of small organic molecules require electrocatalysts to achieve the current density needed for commercial fuel cell applications. For instance, single crystal and polycrystalline platinum, palladium, rhodium and gold electrodes have been used extensively as catalysts for the electro-oxidation of formic acid.[14-16] Among these metal catalysts, platinum shows the highest catalytic activity for electro-oxidation of methanol and formic acid. In order to enhance the oxidation efficiency and reduce costs, Pt-based alloy catalysts with various transition metals such as Pt–Ru,[17-20] Pt– Ni,[21,22] Pt–Sn,[23,24] Pt–Co,[25,26] Pt–Pb,[27] Pt–Bi,[28,29] Pt–Pd,[30,31] Pt–Ti[32] and Pt–Cr[33] have been prepared and studied as possible catalysts for the electro-oxidation of small organic molecules. It is widely accepted that formic acid is oxidized to CO2 via the so- 133   called dual-pathway mechanism, which involves a reactive intermediate (direct path) or adsorbed CO as a poisoning intermediate species (indirect path). In this complicated reaction, carboxylic acid species (HCOO) is generally proposed as the reactive intermediate and adsorbed CO is assigned as the poisoning species. These results have been identified by in situ Surface-Enhanced IR Absorption Spectroscopy (SEIRAS).[34-38] It is well-known that when pure platinum is used as the catalyst, it will be rapidly poisoned by the adsorption of CO produced during the oxidation of HCOOH. However, many investigations have shown that some Pt-based alloy catalysts exhibit enhanced tolerance of CO and, consequently, improved electrocatalytic activities compared to those with platinum alone. For instance, two mechanisms have been proposed to account for the promotional effect of Pt–Ru alloy catalysts. One is the so-called bifunctional mechanism,[39-41] in which the role of ruthenium is to dissociate water to form adsorbed OH species, which then reacts with adsorbed CO to generate CO2. Another explanation is the electronic ligand-effect mechanism, i.e. the electronic properties of platinum are modified by Pt–Ru orbital overlaps so that the binding strength of CO adsorbed on Pt is weakened, leading to the enhancement of electrocatalytic activities for formic acid electro-oxidation.[17,42] Due to the high proportion of surface to bulk atoms, the surface area and the reactivity of nanostructured metal materials are significantly higher than those of the corresponding bulk metals, rendering them ideal candidates in catalytic applications. In fact, currently, many studies of the electrocatalytic oxidation of methanol and formic acid are focused on Pt and Pt-base alloy nanoparticles.[13,32,33,35] For instance, Watanabe et al.[43-45] prepared Fe–Pt alloy thin films using the magnetron sputtering deposition 134   method and found that such alloy films exhibited high CO-tolerance toward H2 oxidation or O2 reduction. Such materials will be candidates for CO-tolerant alloyed catalysts in fuel cells. However, to the best of our knowledge, there have been no studies on methanol or formic acid oxidation catalyzed by FePt alloy nanoparticles. We used monodispersed FePt nanoparticles as catalysts for the electro-oxidation of formic acid. These FePt nanoparticles were prepared by a simultaneous decomposition of iron pentacarbonyl, Fe(CO)5, and reduction of platinum acetylacetonate, Pt(acac)2, in the presence of oleic acid and oleylamine.[46] FePt nanoparticles with an average diameter of ~3 nm were deposited onto a gold electrode surface (denoted as Au/FePt electrode) and were found to be efficient in the electro-catalytic oxidation of formic acid, with a high tolerance of CO poisoning. The onset potential and current density for the HCOOH oxidation, as evidenced by voltammetric and electrochemical impedance spectroscopic studies, demonstrate that the FePt nanoparticles may be used as a powerful catalyst for the electro-oxidation of formic acid fuel. 135   7.2. Experimental 7.2.1. Materials Perchloric acid (HClO4, Fisher, 99.999%) and methanol (CH3OH, ACROS, 99.999%) were used as received. Water was supplied by a Barnstead Nanopure water system (18.3 M ). All solutions were deaerated by bubbling ultra-high-purity N2 for 20 min and protected with a nitrogen atmosphere during the entire experimental procedure. The FePt nanoparticles were prepared according to a previous publication, where the composition of the particles was controlled at Fe20Pt80 to ensure particle stability in strong acid media.[46] Briefly, under a gentle flow of N2, Pt(acetylacetonate)2 (0.5 mmol), 1,2-hexadecanediol (1.5 mmol) and dioctyl ether or benzyl ether (20 mL) were mixed at room temperature and heated to 100 °C. Oleic acid (0.5 mmol), oleylamine (0.5 mmol) and Fe(CO)5 (1.0 mmol) were added and the mixture was heated to reflux (297 °C) for 30 min. (Note: N2 was kept flowing through the reaction system to ensure the Pt-rich Fe20Pt80 nanoparticles were formed. This was different from the previous synthesis in which the reaction was run under a blanket of N2.) The heat source was removed and the reaction mixture was cooled down to room temperature, at which point the reaction system was opened to the ambient environment. The black product was precipitated by adding ethanol (40 mL) and separated by centrifugation. The supernatant was discarded and the black precipitate was dispersed in hexane (25 mL) in the presence of oleic acid (0.05 mL) and oleylamine (0.05 mL). Then, ethanol (20 mL) was added to the dispersion and the suspension was centrifuged again. The precipitation was re-dispersed by hexane. The average particle core diameter was estimated to be 3 nm based on transmission 136   electron microscopy measurements (Figure 7.1). From the X-ray diffraction (XRD) pattern of the FePt nanoparticles (Figure 7.2), the alloy structure of the resulting FePt particles can be clearly identified. Figure 7.1. TEM image of 3 nm Fe20Pt80 nanoparticles. The sample was prepared by drying a hexane dispersion of the particles on an amorphous carbon-coated copper grid. The particles were imaged using a Philips TEM 420 (120 kV). 137   Figure 7.2. X-ray diffraction pattern of the FePt nanoparticle assembly on a glass substrate. The pattern was collected on a Bruker AXS D8 Advance diffractometer under CuKα radiation (λ = 1.5405Å).  138   7.2.2. Preparation of the FePt/Au Electrode A polycrystalline gold disk electrode (sealed in glass tubing) was firstly polished with alumina slurries (0.05 µm) and then cleansed by sonication in 0.1 M HNO3, H2SO4 and Nanopure water for 10 min, successively. FePt nanoparticles (10 µL) dissolved in hexane (0.9 mg mL–1) was then deposited onto the Au electrode surface by a Hamilton microlitre syringe. The particle film was dried by a gentle nitrogen flow for ca. 2 min. The surface coverage of this particle assembly was estimated to be ca. 9 layers by assuming a fully intercalated nanoparticle assembly. The organic protecting ligands were then removed by oxidation in an ultraviolet ozone (UVO) chamber (Jelight Company, Inc., Model 42) for 15 min. The particle film was then rinsed with excessive Nanopure water and ethanol to remove remaining organic deposits. 7.2.3. Electrochemistry Voltammetric measurements were carried out with a CHI 440 electrochemical workstation. The Au/FePt electrode was used as the working electrode. An Ag/AgCl wire and a Pt coil were used as the reference and counter electrodes, respectively. All electrode potentials in the present study will be referred to this Ag/AgCl quasi-reference. Electrochemical impedance spectroscopy (EIS) measurements were carried out using an EG&G PARC Potentiostat/Galvanostat (model 283) and a Frequency Response Detector (model 1025). The impedance spectra were recorded between 100 kHz and 10 mHz, with the amplitude (rms value) of the ac signal being 10 mV. 139   7.3. Results and Discussion 7.3.1. Electrochemical Characterizations of the Electrodes Figure 7.3 shows the cyclic voltammograms of the naked Au electrode and the FePt particles-modified Au electrode before and after UVO treatment in 0.1 M HClO4, at a potential sweep rate of 0.1 V s–1. For the naked Au electrode, the Au oxidation current at potentials above +0.8 V can be clearly seen. In the cathodic potential sweep, there is a voltammetric peak at +0.54 V corresponding to the reduction of the Au oxide. Such voltammetric responses have been observed previously at polycrystalline Au electrode surfaces.[47] However, upon the deposition of FePt particles onto the Au electrode surface, the voltammetric features observed above become substantially suppressed. This may be ascribed to the hydrophobic nature of the organic protecting-layers on the particles that render the Au electrode surface inaccessible by electrolyte ions. In addition, the essentially featureless response also suggests that the FePt particles remain electrochemically inactive within this potential window. However, after the organic protecting ligands were removed by the UVO treatment, the voltammetric response exhibited a drastic variation (Figure 7.3). Firstly, at low potentials, there is a pair of broad current peaks between +0.08 V and –0.3 V which can be attributed to the adsorption– desorption of hydrogen on the metallic Pt surface. Secondly, a reduction current peak can be seen at +0.21 V in the cathodic potential sweep that may be assigned to the reduction of platinum oxides formed at more positive potentials during the anodic potential scan. The observed CV feature is very similar to that for polycrystalline Pt electrodes.[43,45,47] Igarashi et al. attributed the CV resemblance of Pt-based alloys with polycrystalline Pt to 140   the formation of a Pt skin layer after the electrochemical process.[43-45] Such an effect may also explain the voltammetric behaviors observed above for the FePt nanoparticles. The electrocatalytic activity of this functionalized electrode was then examined for formic acid oxidation. Figure 7.3. Cyclic voltammograms of the naked Au electrode ( ), FePt particles modified Au electrode before ( ) and after ( ) UVO treatment in 0.1 M HClO4. Potential scan rate = 0.1 V s–1. Electrode surface area = 0.119 mm2. 141   7.3.2. Electrocatalytic Activity for Formic Acid Oxidation It is well-known that the second metal in Pt-based alloy catalysts can promote the electro-oxidation of small organic molecules. On the basis of previous results of the reaction kinetics in methanol or formic acid electro-oxidation at binary metal electrodes,[48-50] a similar effect might be suggested for FePt nanoparticles from the CV investigation in Figure 7.3. The reaction mechanism for formic acid oxidation on Pt- based alloy (MPt) nanoparticle surfaces is proposed as below: For the indirect oxidation path MPt + HCOOH MPt-(HCOOH)ad (1) MPt-(HCOOH)ad MPt-(CO)ad+ H2O (2) MPt + H2O MPt-(OH)ad+ H++ e– (3) MPt-(CO)ad + MPt-(OH)ad CO2+ H++ e– (4) For the direct oxidation path MPt-(OH)ad+ HCOOH CO2+ H2O + H++ e– (5) The first step entails the adsorption of HCOOH onto the surface of the nanocatalysts. These HCOOH molecules then undergo rapid dissociation into water and CO, and the latter binds strongly to the catalyst surface (i.e., the poisoning effect, step 2). The CO molecules can be further oxidized into CO2 (step 4) by reacting with the hydroxyl species generated by water electrolysis on the catalyst surface (step 3). It is suspected that the electro-oxidation of HCOOH on the FePt particle surface also follows 142   this mechanism (vide infra). By contrast, in the direct oxidation path (step 5), electro- oxidation of HCOOH will be initiated by surface-adsorbed hydroxyl species (step 3) into CO2 and H2O. Figure 7.4 presents the steady-state cyclic voltammograms of the Au/FePt electrode in 0.1 M HCOOH and 0.1 M HClO4. It can be seen that the anodic current is substantially greater than that in 0.1 M HClO4 alone (dotted line, which is identical to the solid curve in Figure 7.3), suggesting that the voltammetric features are arising from the electro-oxidation of HCOOH. This observation is in good agreement with those using Pt and other Pt-based alloy electrodes.[38-51] There are three anodic peaks at about +0.2, +0.51 and +1.06 V in the anodic scan. In the cathodic scan, a very large peak at +0.18 V is observed. It is known that formic acid is either electro-oxidized directly to CO2 by dehydrogenation (step 5), or dissociates spontaneously to produce CO which then becomes oxidized to CO2 (steps 1-4). The poisoning CO species are usually formed within the hydrogen region as well as in the double-layer region. It can be seen from Figure 7.4 that the hydrogen adsorption–desorption currents are significantly inhibited at the Au/FePt electrode in the presence of HCOOH, indicating the surface active sites have been blocked noticeably by adsorbed CO species. Since the CO formation from formic acid does not generate Faradaic current (step 2), the anodic peaks in Figure 7.4 are most probably arising from the oxidation of formic acid or CO. Of these, the anodic current peak at +0.2 V can be attributed to the oxidation of HCOOH to CO2 on surface active sites that have not been poisoned by CO (direct path in step 5). The anodic current peak at +0.51 V may arise from the oxidation of the adsorbed CO and formic acid as a consequence of the release of surface active sites by CO removal (step 4). With further 143   increase in electrode potentials, platinum oxides begin to form and the electrode becomes inactive. At higher potentials, some catalytically active surface oxides can be formed, leading to the anodic current peaks at +1.06 V. In the subsequent cathodic scan, only one voltammetric peak can be seen at +0.18 V, with a significantly greater peak current that can be attributed to the direct oxidation of formic acid, through an active intermediate, into CO2 (the direct path, vide ante). In this, it should be noted that, in the cathodic scan, formic acid begins to be oxidized only when the potential moves to about +0.36 V. This phenomenon can be attributed to the effect of metal surface oxides of different valence states on formic acid oxidation. For instance, previous studies[52-54] have demonstrated that at high oxidation states, some Pt surface oxides (i.e., Pt(OH)3 and PtO2) formed at high potentials may actually be poisoning species. Thus, formic acid can be oxidized only when these surface oxides are reduced at low potentials, so that the electrode surface active sites are restored. The enhanced peak current of formic acid oxidation observed in the reverse cathodic scan, as compared to that in the anodic scan, can then be ascribed to the fact that at these potentials, the surface-adsorbed CO species would have been oxidized to CO2. Consequently, the HCOOH catalytic reaction actually follows the direct path. Usually, onset potential and current density are the two important parameters to compare the activities of electrocatalysts for the electro-oxidation of formic acid or methanol. For instance, Pt and PtRu nanoparticle electrodes exhibited onset potentials of +0.10 and +0.16 (vs. Ag/AgCl in saturated NaCl) for the electro-oxidation of formic acid, respectively.[28] It can be seen from Figure 7.4 that for formic acid oxidation on the FePt nanoparticle electrode, the onset potential is –0.17 V (vs. Ag/AgCl quasi-reference), 144   which is about +0.13 V (vs. Ag/AgCl in saturated NaCl) by using the Au oxidation peak as the calibration point. In comparison to the performance of Pt and PtRu nanoparticles,[28] the FePt nanoparticle-functionalized electrode exhibits comparable onset potential for formic acid oxidation. These results also agree well with the excellent CO tolerance of the FePt alloy.[43-45] However, from the current intensity of formic acid oxidation in the positive and negative scans, it can be seen that the electrode is heavily poisoned by adsorbed CO under the present experimental conditions. Actually, we found that the electrocatalytic activity of the electrode was sensitive to the thickness of the FePt particle film. For instance, for FePt monolayers (by Langmuir–Blodgett deposition), the electrocatalytic activities toward formic acid oxidation were dramatically enhanced. Under the present conditions, the CO adsorbed on the inner FePt particle layers may be difficult to remove by electro-oxidation. Such properties are currently under investigation and the results will be reported in due course. 145   Figure 7.4. Steady-state cyclic voltammograms of the Au/FePt electrode in 0.1 M HCOOH + 0.1 M HClO4 ( ) and in 0.1 M HClO4 ( ). Potential scan rate 0.1 V s–1. 146   7.3.3. Electrochemical Impedance Studies Further studies of the electro-oxidation of HCOOH at the Au/FePt electrode were carried out with electrochemical impedance measurements. Figure 7.5 shows the Nyquist complex-plane impedance spectra of the Au/FePt electrode in 0.1 M HCOOH and 0.1 M HClO4 at various electrode potentials. In the top panel, at E = –0.3 V, the impedance spectrum shows a large arc, with the diameter significantly greater than those at more positive potentials, which can be attributed to the slow reaction rate of formic acid oxidation. It is most probable that the presence of resistive and capacitive components in the equivalent circuit arises from the double-layer effects. From the CV measurements in Figure 7.4, it can be seen that at –0.3 V, formic acid dissociates spontaneously to form CO (step 2), which adsorbed readily on the Au/FePt electrode surface. Thus, the slow reaction kinetics of formic acid oxidation as inferred from the impedance measurements can be ascribed to poisoning by intermediate CO, which blocks continuing adsorption and dehydrogenation of HCOOH on the electrode surface. With a further increase of the electrode potential up to +0.1 V, the impedance spectra exhibit a drastic variation: (i) in addition to the arc in the first quadrant (at high frequency), a smaller one starts to emerge in the fourth quadrant (at low frequency); and (ii) the diameter of both arcs decreases sharply with increasing electrode potential. Such pseudo-inductive behavior has also been observed in methanol electro-oxidation.[55,56] Initially, the reaction sites on the electrode surface are occupied by adsorbed CO generated from formic acid dehydrogenation (step 2). At higher potentials, the weakly-bound CO will be oxidized, leading to the recovery of the surface reaction sites where electro-oxidation of formic acid can then take place. It should be recognized that the low-frequency intersect of the impedance spectra with the x 147   axis (i.e., charge-transfer resistance) also decreases with increasing electrode potentials, signifying enhanced reaction kinetics of the overall electro-oxidation of formic acid.[57,58] These observations are also in agreement with the voltammetric results in Figure 7.4, where the broad anodic peak at ca. +0.2 V is ascribed to the direct oxidation of formic acid to CO2 (vide ante). 148   Figure 7.5. Complex-plane electrochemical impedance plots (Nyquist plots) of the Au/FePt electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials, which are given in the figure legends. The solid lines show some representative fits to the experimental data by the equivalent circuit in Figure 7.5: –0.3 V (top); +0.5 V (middle); and +0.65 V (bottom). 149   Figure 7.5 (middle panel) depicts the impedance plots at potentials between +0.3 V and +0.5 V. An interesting feature of the impedance plot can be observed in this potential range; when the potential is more positive than +0.3 V, negative Faradaic impedance can be observed that are drastically different from the normal Nyquist plots under other potentials (e.g., top and bottom panels). Namely, the impedance spectra now show up in the second quadrant instead of the conventional first one. Again, similar behaviors have also been observed during electro-oxidation of methanol and formic acid on Pt and other Pt-based alloy electrodes.[29,55,56] Such a rapid transition from positive to negative Faradaic impedance suggests the presence of an inductive component.[59] Inductive behavior is often observed in systems involving adsorbed intermediates or metal surface corrosion.[55, 56, 59, 60] Here, the negative Faradaic impedance can be explained by the formation of chemisorbed hydroxyl species within this potential range (step 3), which competes for surface adsorption sites against the poisoning intermediates (CO) and, at the same time, enhances their oxidative removal from the electrode surface (step 4). It is worth noting from the CV measurements in Figure 7.4 that between the potentials of +0.3 and +0.5 V, adsorbed CO begins to be oxidized, leading to high activities of the electrode surface for CO oxidation in this potential range. Usually, the hydroxyl species is considered as the oxygen-donating species for adsorbed CO. Thus, the impedance results agree well with those of CV measurements; and both experimental results indicate that the formation of chemisorbed hydroxyl significantly enhances the oxidation of surface-adsorbed CO. In addition, from Figure 7.5 (bottom panel), it can be seen that at potentials more positive than +0.6 V, the impedance plots return to normal behaviors and the diameter of 150   the arc firstly decreases (from +0.6 to +0.7 V) and then increases (+0.7 to +0.9 V) with increasing potentials. The increase of the arc diameter above +0.7 V is probably due to the formation of Pt surface oxides, which leads to the increase of the charge-transfer resistance for formic acid oxidation. Figure 7.6 shows the corresponding Bode plots of the Au/FePt electrode in 0.1 M HCOOH + 0.1 M HClO4 within different potential ranges (indicated in the figure legends). The kinetic process of the electrode reaction can also be evaluated from the variation of the effective phase angle with electrode potentials. It can be seen from the top panel that there exists a maximum phase angle (somewhat less than –90°, as anticipated from a purely capacitive element) at a characteristic frequency (f1) for all Bode plots. This frequency, and hence the corresponding electrochemical reaction rate, increases with electrode potentials,[55,58] as it represents the time constant for the overall electrochemical reaction. When the potential is more positive than –0.25 V, negative phase angles start to appear at low frequencies, signifying that the reaction kinetics changes from resistive behaviors to pseudo-inductive behaviors.[55,58] Additionally, the frequency (f2) at zero phase angle also increases with increasing electrode potential; again, indicative of enhanced reaction kinetics as mentioned above. 151   Figure 7.6. Bode plots of the electrochemical impedance of the Au/FePt electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials, which are given in the figure legends. 152   In the potential range from +0.3 V to +0.5 V (middle panel), however, an abrupt jump between the positive and negative values of the phase angle was observed. This interesting phenomenon corresponds to the transition to negative faradaic impedance as shown in the Nyquist plots in Figure 7.5 (middle panel). It can be ascribed to the difference of potential dependency between the dehydrogenation reaction of formic acid (step 2) and the electro-oxidation of adsorbed CO (step 4). The reaction rates of these two processes are typically reflected by a maximum in the high and low frequency region, respectively. As mentioned above, generally, the latter (step 4) is a slower process (and therefore the rate-determining step) and more sensitive to electrode potentials than the former (though both frequencies increase with increasing electrode potentials). At sufficiently high electrode potentials, the reaction rate of step 4 starts to be comparable to that of step 2, leading to the abrupt change of the phase angle. Such behavior is very similar to that observed in methanol electro-oxidation.[55] When the potential is more positive than +0.6 V, surface-adsorbed CO is removed almost completely and there is only one positive maximum of phase angle in the Bode plots (bottom panel). From the above CV and impedance results (Figure 7.5 and 7.6), the electrode reaction for formic acid oxidation can be derived in different potential regions. The equivalent circuit, shown in Figure 7.7(A), can be used to fit the above impedance data at potentials more negative than +0.3 V or more positive than +0.5 V. Two representative fits are shown in the top (–0.3 V) and bottom (+0.65 V) panels of Figure 7.5. It can be seen that both fits are excellent. Here, RS is the solution resistance, CPE (constant-phase element) and RCT are the capacitance (which represents the double layer capacitance) and charge transfer resistance, respectively. It has often been observed that the impedance 153   spectrum of a solid electrode may be distorted as a consequence of the roughness of the catalytic layer or a current constriction effect.[55,57,58,61,62] In the present study, the depressed semicircles in the complex-plane impedance plots can be ascribed to the high surface roughness of the electrode modified with FePt nanoparticles. On such a porous solid electrode, the double layer impedance often exhibits CPE characteristics instead of behaving like a pure capacitor.     Figure 7.7. Equivalent circuits for the electro-oxidation of formic acid at the Au/FePt electrode in varied potential regimes: (A) at potentials more negative than +0.3 V or more positive than +0.5 V; and (B) at potentials between +0.3 V and +0.5 V.  154   In the potential range between +0.3 V and +0.5 V, the adsorbed CO begins to oxidize thanks to the formation of chemisorbed hydroxyl species. Thus, a component corresponding to this reaction should be included in the equivalent circuit. The impedance data in this potential range were then fitted using the equivalent circuits shown in Figure 7.7(B),[63] where Co and Ro represent the reaction capacitance and resistance arising from the oxidation of adsorbed CO on the Au/FePt electrode surface. A representative fit using such a circuit is depicted in the middle panel of Figure 7.5 (+0.5 V). Table 7.1 summarizes the fitting results of RS, RCT, CPE, n, Co and Ro at different potentials by using the equivalent circuits in Figure 7.7, where n is a parameter for CPE, and at n = 1, the CPE can be considered as a capacitor. From Table 7.1, it can be seen that the values of RS (solution resistance), CPE, and n are virtually invariant within the entire potential range under study (–0.3 V to +0.9 V). The fact that n 0.9 at all electrode potentials indicates that the CPE in this study is close to pure capacitance. However, it is interesting to note that the charge transfer resistance (RCT) exhibits a clear dependence on electrode potentials, which is depicted in Figure 7.8. At E = –0.3 V, RCT is more than 2 M ; whereas at a slightly more positive potential, E = –0.25 V, RCT decreases sharply by a factor of five. At even more positive potentials (–0.2 to +0.1 V), RCT remains positive and exhibits only a slow decrease with electrode potential. However, at +0.3 V≤ E≤+0.5 V, RCT becomes negative and decreases drastically with increasing electrode potential. This can be ascribed to the inductive behavior arising from the electro-oxidation of surface adsorbed CO species, as speculated above (Figure 7.5 and 7.6). Further increase 155   of the electrode potential to +0.6 V leads to the recovery of positive RCT, which shows a weak dependence on electrode potential (+0.6 V≤ E≤+0.9 V). Table 7.1. Fitting parameters of the electrochemical impedance for Au/FePt electrode at various potentialsa E/V RS/ RCT/k CPE/µF n Co/µF Ro/k –0.3 180.1 2233.0 2.58 0.89 — — –0.25 260.7 450.3 2.44 0.93 — — –0.2 263.2 193.9 2.30 0.94 — — –0.1 184.0 138.1 1.39 0.96 — — 0 255.4 233.0 2.23 0.92 — — +0.1 164.9 100.5 4.54 0.79 — — +0.3 174.8 –118.5 2.72 0.88 5.20 0.14 +0.4 263.7 –373.3 2.37 0.91 3.73 0.45 +0.5 177.8 –1122.0 2.28 0.91 0.97 1.39 +0.6 172.6 626.4 3.07 0.89 — — +0.65 271.6 398.5 2.83 0.89 — — +0.7 176.8 379.6 2.52 0.90 — — +0.8 176.0 215.9 2.01 0.93 — — +0.9 177.9 641.1 1.97 0.93 — — a Experimental data were measured in 0.1 M HCOOH + 0.1 M HClO4 (Figure 7.5) and fitted by using the equivalent circuits shown in Figure 7.7. 156   Figure 7.8. Dependence of the charge-transfer resistance (RCT) on electrode potentials for the electro-oxidation of formic acid at the Au/FePt electrode by fitting the experimental data (Figure 7.5) with the equivalent circuits in Figure 7.7. Symbols are experimental data and the line is for eye-guiding only. 157   7.4. Conclusion In this paper, the electro-oxidation of formic acid at FePt alloy nanoparticle surfaces was studied by electrochemical voltammetry and impedance spectroscopy. The Au electrode modified with FePt alloy nanoparticles after UVO treatment was successfully used as an electrocatalyst for the oxidation of formic acid in an acid electrolyte. The FePt nanoparticle-functionalized electrode exhibited comparable onset potential for formic acid oxidation. Thus, FePt nanoparticles will be an excellent electrocatalytic candidate in fuel cell applications. Voltammetric and EIS studies showed that the formic acid oxidation was affected by reaction intermediates adsorbed on the electrode surface. In this study, EIS was used in the investigation of the reaction kinetics and mechanism of electro-oxidation of formic acid. The variation of the reaction mechanism in different potential regions was attributed to the formation of different intermediates on the electrode surface. With the increase of electrode potential, it was observed that the kinetic behavior evolved from resistive to pseudo-inductive and then to inductive characteristics. At low potentials, formic acid dissociated spontaneously to produce the CO intermediate, which adsorbed readily onto the electrode surface. At more positive electrode potentials, chemisorbed hydroxyl was formed, which enhanced the oxidative removal of the adsorbed CO intermediate. These results will be of fundamental importance in understanding the electrochemical mechanism for liquid organic fuel oxidation at different electrode potentials, and hence enhanced performance of fuel cell catalysts. 158   REFERENCES 1. C. Lamy, A. Lima, V. LeRhun, F. Delime, C. Coutanceau, J. M. Leger, J. Power Sources, 2002, 105, 283. 2. J. Willsau, J. Heitbaum, Electrochim. Acta, 1986, 31, 943. 3. X. Wang, J. M. Hu, I. M. Hsing, J. Electroanal. Chem., 2004, 562, 73. 4. Y. W. Rhee, S. Y. Ha, R. I. Masel, J. Power Sources, 2003, 117, 35. 5. R. Parsons, T. Vandernoot, J. Electroanal. Chem., 1988, 257, 9. 6. G. Q. Lu, A. Crown, A. Wieckowski, J. Phys. Chem. B, 1999, 103, 9700. 7. M. C. Zhao, C. Rice, R. I. Masel, P. Waszczuk, A. Wieckowski, J. Electrochem. Soc., 2004, 151, A131. 8. V. M. Jovanovic, D. Tripkovic, A. Tripkovic, A. Kowal, J. Stoch, Electrochem. Commun., 2005, 7, 1039. 9. H. Okamoto, W. Kon, Y. Mukouyama, J. Phys. Chem. B, 2004, 108, 4432. 10. J. D. Lovic, A. V. Tripkovic, S. L. J. Gojkovic, K. D. Popovic, D. V. Tripkovic, P. Olszewski, A. Kowal, J. Electroanal. Chem., 2005, 581, 294. 11. M. Gholamian, A. Q. Contractor, J. Electroanal. Chem., 1990, 289, 69. 12. T. Iwasita, X. H. Xia, E. Herrero, H. D. Liess, Langmuir, 1996, 12, 4260. 13. J. Jiang, A. Kucernak, J. Electroanal. Chem., 2002, 520, 64. 14. A. Capon, R. Parsons, J. Electroanal. Chem. Interfacial Electrochem., 1973, 44, 1. 15. S. Wasmus, D. A. Tryk, W. Vielstich, J. Electroanal. Chem., 1994, 377, 205. 16. A. Capon, R. Parsons, J. Electroanal. Chem. Interfacial Electrochem., 1973, 44, 159   239. 17. T. Frelink, W. Visscher, J. A. R. vanVeen, Langmuir, 1996, 12, 3702. 18. W. F. Lin, M. S. Zei, M. Eiswirth, G. Ertl, T. Iwasita, W. Vielstich, J. Phys. Chem. B, 1999, 103, 6968. 19. T. Iwasita, H. Hoster, A. John-Anacker, W. F. Lin, W. Vielstich, Langmuir, 2000, 16, 522. 20. H. S. Wang, C. Wingender, H. Baltruschat, M. Lopez, M. T. Reetz, J. Electroanal. Chem., 2001, 509, 163. 21. K. W. Park, J. H. Choi, B. K. Kwon, S. A. Lee, Y. E. Sung, H. Y. Ha, S. A. Hong, H. Kim, A. Wieckowski, J. Phys. Chem. B, 2002, 106, 1869. 22. J. F. Drillet, A. Ee, J. Friedemann, R. Kotz, B. Schnyder, V. M. Schmidt, Electrochim. Acta, 2002, 47, 1983. 23. Y. Morimoto, E. B. Yeager, J. Electroanal. Chem., 1998, 444, 95. 24. F. Colmati, E. Antolini, E. R. Gonzalez, Electrochim. Acta, 2005, 50, 5496. 25. T. Page, R. Johnson, J. Hormes, S. Noding, B. Rambabu, J. Electroanal. Chem., 2000, 485, 34. 26. E. Antolini, J. R. C. Salgado, E. R. Gonzalez, J. Electroanal. Chem., 2005, 580, 145. 27. S. Motoo, M. Watanabe, J. Electroanal. Chem., 1976, 69, 429. 28. C. Roychowdhury, F. Matsumoto, P. F. Mutolo, H. D. Abruña, F. J. DiSalvo, Chem. Mater., 2005, 17, 5871. 29. J. Y. Lee, P. Strasser, M. Eiswirth, G. Ertl, Electrochim. Acta, 2001, 47, 501. 160   30. M. Baldauf, D. M. Kolb, J. Phys. Chem., 1996, 100, 11375. 31. R. S. Jayashree, J. S. Spendelow, J. Yeom, C. Rastogi, M. A. Shannon, P. J. A. Kenis, Electrochim. Acta, 2005, 50, 4674. 32. L. B. Lai, D. H. Chen, T. C. Huang, J. Mater. Chem., 2001, 11, 1491. 33. H. Yang, N. Alonso-Vante, J. M. Leger, C. Lamy, J. Phys. Chem. B, 2004, 108, 1938. 34. Y. X. Chen, A. Miki, S. Ye, H. Sakai, M. Osawa, J. Am. Chem. Soc., 2003, 125, 3680. 35. S. Park, Y. Xie, M. J. Weaver, Langmuir, 2002, 18, 5792. 36. S. C. Chang, L. W. H. Leung, M. J. Weaver, J. Phys. Chem., 1990, 94, 6013. 37. T. Yajima, H. Uchida, M. Watanabe, J. Phys. Chem. B, 2004, 108, 2654. 38. H. Shiroishi, Y. Ayato, K. Kunimatsu, T. Okada, J. Electroanal. Chem., 2005, 581, 132. 39. H. A. Gasteiger, N. Markovic, P. N. Ross, E. J. Cairns, J. Electrochem. Soc., 1994, 141, 1795. 40. H. N. Dinh, X. M. Ren, F. H. Garzon, P. Zelenay, S. Gottesfeld, J. Electroanal. Chem., 2000, 491, 222. 41. H. F. Oetjen, V. M. Schmidt, U. Stimming, F. Trila, J. Electrochem. Soc., 1996, 143, 3838. 42. J. B. Goodenough, A. Hamnett, B. J. Kennedy, R. Manoharan, S. A. Weeks, J. Electroanal. Chem., 1988, 240, 133. 43. M. Watanabe, Y. M. Zhu, H. Uchida, J. Phys. Chem. B, 2000, 104, 1762. 161   44. H. Igarashi, T. Fujino, Y. M. Zhu, H. Uchida, M. Watanabe, Phys. Chem. Chem. Phys., 2001, 3, 306. 45. T. Toda, H. Igarashi, H. Uchida, M. Watanabe, J. Electrochem. Soc., 1999, 146, 3750. 46. S. H. Sun, C. B. Murray, D. Weller, L. Folks, A. Moser, Science, 2000, 287, 1989. 47. R. Woods, in Electroanalytical Chemistry, ed. A. J. Bard, Marcel Dekker, Inc., New York, Vol. 9, 1976. 48. H. A. Gasteiger, N. Markovic, P. N. Ross, E. J. Cairns, J. Phys. Chem., 1993, 97, 12020. 49. M. T. M. Koper, J. J. Lukkien, A. P. J. Jansen, R. A. van Santen, J. Phys. Chem. B, 1999, 103, 5522. 50. T. Iwasita, Electrochim. Acta, 2002, 47, 3663. 51. P. Waszczuk, T. M. Barnard, C. Rice, R. I. Masel, A. Wieckowski, Electrochem. Commun., 2002, 4, 599. 52. N. H. Li, S. G. Sun, S. P. Chen, J. Electroanal. Chem., 1997, 430, 57. 53. N. H. Li, S. G. Sun, J. Electroanal. Chem., 1997, 436, 65. 54. N. H. Li, S. G. Sun, J. Electroanal. Chem., 1998, 448, 5. 55. I. M. Hsing, X. Wang, Y. J. Leng, J. Electrochem. Soc., 2002, 149, A615. 56. A. C. Chen, D. J. La Russa, B. Miller, Langmuir, 2004, 20, 9695. 57. Y. C. Liu, X. P. Qiu, W. T. Zhu, G. S. Wu, J. Power Sources, 2003, 114, 10. 58. W. Sugimoto, K. Aoyama, T. Kawaguchi, Y. Murakami, Y. Takasu, J. Electroanal. Chem., 2005, 576, 215. 162   59. G. Markovich, C. P. Collier, J. R. Heath, Phys. Rev. Lett., 1998, 80, 3807. 60. D. D. Macdonald, Electrochim. Acta, 1990, 35, 1509. 61. J. T. Muller, P. M. Urban, W. F. Holderich, J. Power Sources, 1999, 84, 157. 62. A. Maritan, F. Toigo, Electrochim. Acta, 1990, 35, 141. Chapter 8 Langmuir-Blodgett Thin Films of Fe20Pt80 Nanoparticles for the Electrocatalytic Oxidation of Formic Acid 164   8.1. Introduction Extensive efforts have been devoted to fuel cell research in the past few decades, as fuel cells are considered as the promising candidates of power sources in portable electronic devices and transportation (electrical) vehicles.[1,2] Yet, despite substantial improvements of fuel cell performance in recent years, some challenging obstacles remain.[2-5] Of these, optimization of the electrocatalysts represents a critical step. Noble metals (mainly Pt) have often been employed as the catalysts in fuel cells.[6] Recent studies have shown that Pt-based alloy catalysts[7-15] exhibit greater catalytic activities for the electro-oxidation of small organic fuels such as methanol and formic acid due to the so-called bifunctional mechanism[16-18] or electronic (ligand) effect.[7,19] In order to further reduce the costs of catalysts, either less expensive non-noble metal catalysts have to be used or the loading of the noble metals on the electrode surface has to be reduced, and at the same time, the catalytic performance and utilization efficiency of electrocatalysts remain uncompromised. To this end, it is essential and important to have a mechanistic understanding of the electron-transfer dynamics involved, by carrying out a systematic study with variation of the catalyst structures and assemblies. For Pt and Pt-based alloy nanostructured catalysts,[20-22] the metal particles were typically studied by dropcasting the colloidal solution onto an electrode surface forming a thick film. However, by using this method, it is very difficult to control the thickness and ordering of the films, as such dropcast films usually are not uniform with extensive structural defects. Since it has been well-known that the surface layers of the particle films play a critical role in the electro-oxidation of formic acid and other fuels, [3,23,24] it is 165   of fundamental and technological significance to examine the effects of particle loading and ordering on the catalytic activities. This can be readily achieved by the Langmuir−Blodgett (LB) technique. Using formic acid as a model fuel, we have carried out an earlier study with a dropcast thick film of Fe20Pt80 nanoparticles (average diameter ~ 3 nm) and found that the particles exhibited very good electrocatalytic activities for formic acid electro- oxidation,[25] as compared to Pt and other metal alloys. This was motivated, in part, by the fact that aqueous solutions of formic acid have been found to be an attractive potential fuel because of the ease of their handling, transportation, and storage in comparison to hydrogen.[26] In this paper, we carried out further studies to examine the effect of the FePt particle film thickness on the electrocatalytic activities in the oxidation of formic acid. The particle films, from a single monolayer to multilayers, were prepared by the LB technique. The voltammetric and electrochemical impedance measurements showed that the catalytic behaviors of FePt particles were strongly dependent on the assembly thickness, suggestive of a minimal and optimal loading of the nanoparticle catalysts for the fuel cell reactions. 166   8.2. Experimental 8.2.1. Chemicals Perchloric acid (HClO4, Fisher, 99.999%) and formic acid (HCOOH, ACROS, 99%) were used as received. Water was supplied by a Barnstead Nanopure water system (18.3 MΩ·cm). All solutions were deaerated by bubbling ultrahigh-purity N2 for 20 min and protected with a nitrogen atmosphere during the entire experimental procedure. 8.2.2. FePt Nanoparticles The FePt nanoparticles were synthesized according to the procedure described previously,[25] but the composition of the particles was controlled at Fe20Pt80 to ensure particle stability in strong acid media. Briefly, under a gentle flow of N2, Pt(acetylacetonate)2 (0.5 mmol), 1,2-hexadecanediol (1.5 mmol), and dioctylether or benzyl ether (20 mL) were mixed at room temperature and heated to 100 °C. Oleic acid (0.5 mmol), oleyl amine (0.5 mmol), and Fe(CO)5 (1.0 mmol) were added, and the mixture was heated to reflux (297 °C) for 30 min. (Note: N2 was kept flowing through the reaction system to ensure the Pt-rich Fe20Pt80 nanoparticles were formed. This was different from the previous synthesis in which the reaction was run under a blanket of N2.) The heat source was removed, and the reaction mixture was cooled down to room temperature. At this point, the reaction system was opened to the ambient environment. The black product was precipitated by adding ethanol (40 mL) and separated by centrifugation. The supernatant was discarded, and the black precipitate was dispersed in 167   hexane (25 mL) in the presence of oleic acid (0.05 mL) and oleylamine (0.05 mL). Then, ethanol (20 mL) was added to the dispersion and the suspension was centrifuged again. The precipitation was redispersed by hexane. From transmission electron microscopic (TEM) measurements, the particle core size was estimated to be 3.98 ± 0.73 nm (a representative TEM micrograph was included in the Supporting Information). 8.2.3. Preparation of Particle Langmuir−Blodgett Thin Films. The experimental setup has been described in detail previously.[27-29] In a typical experiment, 100 μL of a FePt particle solution (1 mg/mL in CH2Cl2) was spread dropwise by using a Hamilton microliter syringe onto the water surface of a Langmuir−Blodgett trough (NIMA 611D). At least 20 min was allowed for solvent evaporation prior to the first compression and between compression cycles. Compression speed was set at 10 cm2/min. Prior to particle deposition, a gold film electrode was first treated in an ultraviolet ozone (UVO) chamber (Jelight Company, Inc., model 42) for about 15 min and then coated with a self-assembled monolayer of n-butanethiol in order to render the electrode surface hydrophobic. The FePt particles were then transferred onto the gold film electrode surface by the LB technique at a dipping speed of 1 mm/min at controlled surface pressures. Three particle thin films were prepared at the same surface pressure, consisting of one, two, and four monolayers of particles, which were hence referred to as the LB1, LB2, and LB4 electrodes, respectively. 168   8.2.4. Atomic Force Microscopy (AFM) The morphology of the LB films was studied by tapping-mode AFM. The AFM images were acquired with a PicoLE SPM instrument (Molecular Imaging Inc.). The scan speed was 1.8 lines/s. All the AFM images were flattened with Molecular Imaging software. For the measurements of the particle layer thickness, the AFM images were first divided into a 20 × 20 matrix and line analyses were carried out of 20 evenly spaced sections in both the vertical and horizontal directions, from which the height information at each crossing point was obtained. The average thickness (and the standard deviation) of the particle films was then estimated. 8.2.5. Electrochemistry Voltammetric measurements were carried out with a CHI 440 electrochemical workstation. The FePt particles-coated Au film electrodes were used as the working electrode. Prior to any electrochemical measurements, the electrode was subject to UVO treatment to remove the organic coating around the FePt nanoparticles. A Ag/AgCl wire and a Pt coil were used as the reference and counter electrodes, respectively. All electrode potentials in the present study were referred to this Ag/AgCl quasi-reference. Electrochemical impedance spectroscopy (EIS) measurements were carried out using an EG&G PARC potentiostat/galvanostat (model 283) and frequency response detector (model 1025). The impedance spectra were recorded between 100 kHz and 10 mHz with the amplitude (rms value) of the ac signal 10 mV. 169   8.3. Results and Discussion 8.3.1. Langmuir−Blodgett Thin Films of FePt Nanoparticles Figure 8.1 shows the Langmuir isotherm of a FePt particle monolayer at the air|water interface, which is very similar to those observed with other metal and semiconductor nanoparticles,27-29 with a takeoff area of 290 cm2. The particles were then transferred onto the electrode surface by the LB technique at the surface pressure of 13.5 mN/m. From the isotherm, the corresponding particle density on the water surface can be estimated to be 0.52 μg/cm2. Thus, for an ideal deposition, the particle surface density on the electrode surface should be 0.52 (LB1), 1.04 (LB2), and 2.08 μg/cm2 (LB4), respectively. Figure 8.1. Langmuir isotherm of a Fe20Pt80 particle monolayer at the air/water interface: particle concentration, 1 mg/mL in CH2Cl2; amount spread, 100 μL; compression speed, 10 cm2/min. 170   Tapping-mode AFM imaging was then carried out to examine the surface morphology and coverage of the deposited particle thin films. Figure 8.2 shows the corresponding topographical images and line scan profiles of the three LB thin films of nanoparticles. It can be seen that the resulting particle assemblies are all very closely packed, signifying very efficient transfer of the particle layers from the air|water interface onto the electrode surface. Yet, line scans across the AFM images reveal that the surface roughness of the particle LB thin films increases with increasing thickness of the particle assemblies. For instance, the average thickness of the particle layers increases from ca. 6.5 ± 1.5 nm (roughness ratio 23%) in the LB1 sample (panel a) to about 12 ± 2.3 nm (19%) and 20 ± 7.5 nm (38%) in the LB2 (panel b) and LB4 (panel c) samples, respectively. This deviates slightly from the expected 1:2:4 for an ideal deposition. There are at least two plausible explanations to these observations. The first was possibly due to the less efficient transfer of particles for a thicker film, as the AFM image (Figure 2c) shows some clustering of the particles. Measurements of the height of the bright features indicate that they are not of isolated particles but rather a cluster of particles. Additionally, in solid state, it is well-known that intercalation of the surface protecting ligands of organically capped nanoparticles occurs.[30] Thus, when a second layer of particles is deposited onto the electrode surface, it is likely that ligand intercalation will also occur between the layers of particles. Yet the intercalation is probably not uniform across the entire layer, leading to enhanced roughness of the particle films. The resulting surface roughness will propagate with the deposition of more layers of particles. However, it should be noted that the surface roughness observed with the LB films here is 171   significantly lower than that of nanoparticle films prepared by the dropcast method which exhibit extensive clustering of nanoparticles (roughness ratio >30%, Figure 8.3).   Figure 8.2. Tapping-mode AFM topographical images and representative line scan profiles of the Fe20Pt80 nanoparticle LB films: (a) LB1, (b) LB2, (c) LB4. Scan size is 1500 nm × 1500 nm, and scan rate is 1.8 lines/s.  172   Figure 8.3. AFM images of dropcast films of FePt nanoparticles. The amount of particles deposited varied from 0.2 μg (a), 0.5 μg (b), and 2.5 μg (c). The average particle film thickness is found to be 13.1 ± 4.2 nm (a), 21.1 ± 9.79 nm (b), and 39.1 ± 12.3 nm (c), respectively. Incomparison to the thickness of the LB layers, the number of layers of these dropcast films can be estimated to be about 2, 4, and 8 respectively. 173   Also, it should be noted that the thickness measured by tapping-mode AFM in the LB1 particle film reflects the physical diameter of the nanoparticles, which consists of the metal core plus two protecting ligands. Considering the average particle core size of 3.98 nm (from TEM measurements, Figure 8.4) and the fully extended chain length of an oleic acid or oleylamine around 2.0 nm (from Hyperchem), the results are reasonable. Figure 8.4. TEM image of the Fe0.2Pt0.8 nanoparticles stabilized by oleylamine and oleic acid. Inset shows the histogram of the particles core size. 174   8.3.2. Electrochemistry of FePt Functionalized Electrodes In the previous studies of FePt alloy electrodes, it has been found that during the electrochemical scans, there will be a Pt skin formed on the electrode surface due to the dissolution of Fe on the surface of FePt alloys.[31,32] Thus, the resulting steady-state cyclic voltammograms usually behave similarly to those of pure Pt. Figure 8.5 shows the stead- state cyclic voltammograms of the three FePt particle films (LB1, LB2, and LB4) in 0.1 M HClO4, which are all similar to those of other Pt-based alloys.[25] (Note that the voltammetric features from the Au film substrate were markedly suppressed, signifying efficient deposition of the particle layers on the electrode surface. Additional cyclic voltammograms are showed in Figure 8.6) At low potentials, there are a pair of broad current peaks between −0.50 and −0.27 V which are attributed to the adsorption−desorption of hydrogen on the particle (Pt) surface (at more negative potentials, extensive hydrogen evolution occurred). This feature becomes better-defined with thicker particle films. In addition, in the negative potential scan, there is a reduction current peak around +0.11 V which can be ascribed to the reduction of platinum oxides formed at more positive potentials during the positive potential scan. It should be noted that the current of hydrogen adsorption−desorption and the reduction current of platinum oxides increase with increasing particle layer thickness. On the basis of the charge integration within the hydrogen adsorption−desorption potential region, the actual surface (Pt) area of the three particle films can be evaluated, which is 0.36 (LB1), 0.52 (LB2), and 1.19 cm2 (LB4), respectively. Note that the ratio of these areas, 1:1.4:3.3, is close to the thickness ratio calculated from the AFM measurements (Figure 8.2), 1:1.9:3.1, which suggests that the active surface area does increase almost proportionally with the surface 175   density of nanoparticles, and thus the majority of the particles are electrochemically accessible even with thicker particle layers. In contrast, for nanoparticle dropcast films of comparable thickness, the effective surface area is much smaller, primarily because of the clustering of the nanoparticles that limits the accessibility of the catalyst surface (Figure 8.3). Figure 8.5. Cyclic voltammograms of the Au film electrode coated with the three different particle LB layers in 0.1 M HClO4: potential scan rate, 0.1 V/s. The currents are all normalized to the Au film surface area (0.66 cm2). 176   Figure 8.6. Cyclic voltammograms in 0.1 M HClO4 at potential scan rate of 0.1 V/s at three different electrodes: naked gold, gold with butanethiol self-assembled monolayer exposed to ozone, and the Au-LB1 electrode. Two points warrant attention here. First, the voltammetric responses in 0.1 M HClO4 at the naked Au electrode and the Au/C4SH SAM exposed to ozone are very consistent, signifying the effective removal of the organic coating layer by UVO treatment. Second, the large double-layer charging current observed at the LB1 electrode at negative potentials may be ascribed to the partial exposure of the gold film to the solution due to incomplete coating of the Au surface by the nanoparticles. 177   8.3.3. Formic Acid Oxidation In the last decades, the mechanism of formic acid electro-oxidation on Pt and Pt- based alloys has been investigated extensively, and it is generally accepted that formic acid is oxidized to CO2 via the so-called dual-pathway mechanism.[33-35] For Pt-based alloys (PtM), formic acid can be oxidized to CO2 by the direct pathway, HCOOH + PtM Æ CO2 + 2H+ + 2e- (1) Alternatively, in the indirect pathway, HCOOH first reacts to form poisonous CO intermediate, which is then oxidized to CO2, PtM + HCOOH Æ PtM-(HCOOH)ad (2) PtM-(HCOOH)ad Æ PtM-(CO)ad + H2O (3) PtM + H2O Æ PtM-(OH)ad + H+ + e- (4) PtM-(CO)ad + PtM-(OH)ad Æ CO2 + H+ + e- (5) In both cases, the overall reaction is HCOOH Æ CO2 + 2H+ + 2e- (6). It has been found in our previous study[25] that from the onset potential and current densities, dropcast films of Fe20Pt80 nanoparticles exhibit very good electrocatalytic activities for HCOOH oxidation as compared to Pt and other metal alloys.[33-35] However, from the voltammetric currents of HCOOH oxidation in the positive and negative scans, the particles are found to be heavily poisoned by adsorbed CO species. Figure 8.7 depicts the steady-state cyclic voltammograms of the gold film electrode loaded with the three FePt particle films in 0.1 M HCOOH and 0.1 M HClO4. It can be seen that the voltammetric features exhibit a rather substantial variation with the nanoparticle layer thickness. For the LB1 and LB2 thin films, in the positive potential 178   scan, an oxidation current peak can be observed at about +0.42 and +0.46 V, respectively; and in the negative potential scan, an oxidation peak at almost the same potential is observed. However, the cathodic current density is somewhat smaller than the anodic one. This can be attributed to the effect of the formation of CO poisoning intermediate and its adsorption onto the catalyst surface. In the positive potential scan, when the electrode potential is scanned from −0.3 to +0.4 V, the oxidation of particle- bound CO starts to occur, leading to the recovery of the catalyst active sites on which direct electro-oxidation of formic acid can now take place. That is, the anodic currents measured reflect the combined contributions from the electro-oxidation of both CO intermediates and formic acid, whereas in the negative potential scan, the currents mostly arise from the formic acid oxidation alone. 179   Figure 8.7. Cyclic voltammograms of the Au electrode coated with the LB1, LB2, and LB4 Fe20Pt80 particle thin films in 0.1 M HCOOH + 0.1 M HClO4: potential scan rate, 0.02 V/s; gold film surface area, 0.66 cm2. The currents are all normalized to the respective Pt surface areas that are evaluated from Figure 8.5. 180   For the LB4 nanoparticle film, the cyclic voltammetry (CV) responses are drastically different. There are two current peaks in the positive potential scan, one with a higher intensity at about +0.07 V and the other with a weaker current density at +0.36 V; and in the negative potential scan, a very large oxidation current peak can be seen at about +0.07 V. Similar voltammetric features have also been observed in the studies of electro-oxidation of HCOOH on Pt and other Pt-based alloy electrodes.[36,37] It is well- known that during the electro-oxidation of methanol, formic acid, and other small organic molecules, the so-called dual-pathway mechanism exists which involves a reactive intermediate (direct path, eq 1) and adsorbed CO as a poisoning intermediate species (indirect path, eqs 2−5). For formic acid oxidation at the LB4 electrode, the two oxidation peaks observed in the positive potential scan can be accounted for by the effects of CO poisonous species, because the CO adsorption occurs within the hydrogen region and the double-layer region, and consequently parts of the FePt particle surface become inactive for formic acid oxidation (the poisoning effect). Here the major peak at +0.07 V can be attributable to the direct oxidation of formic acid into CO2 at active surface sites which have not been poisoned by CO adsorption (eq 1), whereas the minor peak at +0.36 V is from the oxidation of the adsorbed CO and of formic acid as a consequence of the release of surface-active sites by CO stripping (eqs 2−5). With further increase in electrode potentials, the adsorbed CO will be removed. So the enhanced peak current of formic acid oxidation observed in the reverse negative potential scan, as compared with that in the positive scan, can be explained by the recovery of all the surface-active sites. 181   However, at the LB1 and LB2 thin film electrodes, there is only one oxidation peak in both positive and negative potential scans at substantially more positive potentials. The discrepancy in the voltammetric responses may be explained by the different CO poisoning on these particle thin film electrodes. For the LB1 and LB2 thin films, the particle surfaces appear to be easily poisoned by CO adsorption. So the direct oxidation of formic acid is blocked before CO is removed by electro-oxidation. In contrast, at the LB4 electrode, not all surface-active sites are poisoned by CO adsorption; thus, the direct oxidation of formic acid can occur, which is reflected by the anodic peak at a less positive potential (+0.07 V). Moreover, for LB1 and LB2, the peak current in the negative potential scan is smaller than that in the positive potential scan, which is different from that for LB4. Similar phenomena have also been observed in the electrocatalytic oxidation of HCOOH and CH3OH.[38-44] For instance, in the study of formic acid oxidation on a Au(111) surface modified with Pd layers, it was found that the peak current in the negative potential scan was smaller than that in the positive potential scan when only a single monolayer of Pd was deposited on the Au(111) surface.[12] Such voltammetric features were ascribed to the rapid CO accumulation on the electrode surface.[44] In the present study, the difference of the voltammetric behaviors may also be accounted for by CO accumulation on the LB1 and LB2 electrode surface leading to heavier poisoning than at the LB4 surface. We would like to point out that in our previous studies[25] with a dropcast thick film of FePt nanoparticles (about nine layers), the catalyst layers were much more extensively poisoned by adsorbed CO species. From the current ratio of the anodic and cathodic peaks for formic acid direct oxidation, the fraction of the catalyst surface that 182   was blocked by CO adsorption was estimated to be 90%.[25] In the present study with the LB4 electrode, this fraction was found to decrease to 42% (Figure 8.7). These results indicate that there exists a minimal (and optimal) loading of the electrocatalysts in the electro-oxidation of formic acid with improved resistance to CO poisoning. In addition, on the basis of the current density of the formic acid oxidation, the electrodes functionalized with particle LB layers (Figure 8.7) exhibited much higher catalytic activities than those with dropcast films of the same particles of similar layer thickness (Figure 8.8), suggesting the important role of particle arrangements in the determination of their electrocatalytic performance. 183   Figure 8.8. Cyclic voltammograms of the dropcast film-functionalized gold electrodes in 0.1 MHClO4 + 0.1 M HCOOH. Scan rate 0.02 V/s. The thickness of the particle layers was characterized by AFM measurements (Figure 8.3): 2 layers (Dropcast 1), 4 layers (Dropcast 2), and 8 layers (Dropcast 3). Yet, based on the integration of the hydrogen adsorption/desorption currents in 0.1 M HClO4, the effective surface areas are found to be 0.03, 0.14, and 0.15 cm2 respectively, significantly smaller that those of the LB films of comparable thickness (Figure 8.5), most probably because of the clustering of the particles that limits the particle accessibility (Figure 8.3). In addition, the overall current density is smaller than that found with the electrode modified with LB thin films of similar thickness (Figure 8.7), suggesting the significance of particle arrangements in the determination of their electrocatalytic activities. 184   8.3.4. Electrochemical Impedance Studies Electrochemical impedance spectroscopy has been a powerful and sensitive technique to study electrochemical kinetics, for instance, in the investigations of the electro-oxidation process of small organic molecules in fuel cells.[45-48] From the voltammetric studies presented above, it can be seen that the loading of the FePt nanoparticle catalysts strongly affects the electrochemical reaction dynamics of formic acid oxidation, which were further examined by EIS measurements, as presented below. Figure 8.9 shows the Nyquist complex-plane impedance spectra of the electrodes with one, two, and four FePt particle layers in 0.1 M HCOOH and 0.1 M HClO4 at varied electrode potentials from −0.3 to +0.9 V (shown as figure legends). It can be seen from Figure 8.9a that for the LB1 electrode, at negative potentials (−0.2 to 0.0 V) the diameter of the arcs (indicating the presence of resistive and capacitive components) exhibits a small increase with increasing electrode potential. This may be ascribed to the formation of intermediate poisoning (CO) species in the potential range that impedes the electro- oxidation of formic acid. With further increase of the electrode potential from 0.0 to +0.3 V, the diameter of the arc in Figure 8.9a decreases because the electro-oxidation and hence removal of the adsorbed poisoning species starts to occur. At more positive electrode potentials (+0.55 to +0.65 V), the arc bends from the positive to negative direction of the x-axis (left panel of Figure 8.9a), signifying the electro-oxidation process of adsorbed CO, and hence the electrode exhibits pseudoinductive behaviors. Further increase of electrode potentials saw the impedance spectra return to the normal behaviors; and the arc diameter decreases with electrode potentials, indicative of enhanced electron- transfer kinetics. 185   Figure 8.9b depicts the impedance spectra with the LB2 particle film in the same electrolyte solution. We can see that at the potentials less than +0.5 V, the impedances plots show almost the same features as those of the LB1 particle film. However, at the potentials between +0.55 and +0.65 V, the impedance starts to appear in the second quadrant instead of the conventional first one. Such negative faradaic impedance is often observed in systems containing adsorbed intermediates and metal surface corrosion,[47-49] which suggests the presence of an inductive component. For the present system, the presence of negative faradic impedance (the inductive component) may be due to the oxidative removal of the adsorbed CO intermediate because of the formation of chemisorbed hydroxyl species in the potential range. From the voltammetric response of HCOOH oxidation at the LB2 nanoparticle film (Figure 8.7), it can be seen that within the potential range of +0.55 to +0.65 V, the adsorbed intermediates (CO) become oxidized. The impedance results agree well with the above voltammetric studies. When the potential is more positive than +0.65 V, the impedance plots return to the normal Nyquist spectra in the first quadrant. In accord with the impedance change, the voltammetric response of the LB2 film in Figure 8.7 shows that at potentials greater than +0.65 V, the adsorbed intermediate should be removed by electro-oxidation and platinum oxides begin to be formed. For the LB4 nanoparticle film (Figure 8.9c), in comparison with the impedance spectra of the LB2 electrode, the diminishment of the diameter of the negative arcs from +0.55 to +0.65 V implies that more inductive component and less CO poisoning are present in the electrochemical system, consistent with the voltammetric results presented earlier (Figure 8.7). It should be noted that the impedance results observed here (Figure 186   8.9c) are very similar to those with dropcast thick films of the same particles[25] except for a small difference of the potentials at which the negative impedance starts to appear. Figure 8.9. Complex-plane electrochemical impedance plots (Nyquist plots) of the Au film electrode functionalized with varied LB layers of the Fe20Pt80 nanoparticles: (a) LB1, (b) LB2, and (c) LB4 in 0.1 M HCOOH + 0.1 M HClO4 at different electrode potentials which are shown in the figure legends. Solid lines are representative curve fits by the equivalent circuits shown in Figure 8.12. 187   Figure 8.10 depicts the corresponding Bode plots of the LB1, LB2, and LB4 thin film electrodes. It can be seen that there is a maximum phase angle at a characteristic frequency (fmax) in all the three Bode plots (insets in each Bode plot display the variation of this characteristic frequency with electrode potentials). This characteristic frequency usually represents the time constant of the electrochemical reaction.[46,47] From Figure 8.10, it can be seen that fmax displays no drastic change with electrode potentials, similar to that observed earlier with FePt dropcast multilayers.[25] Of note is that for the LB2 and LB4 films, an abrupt change of sign of the phase angle was observed in the range of from +0.55 to +0.65 V, which was attributed to the oxidation of adsorbed CO and the resulting inductive behaviors of the electrodes (i.e., negative faradaic impedance).     188   Figure 8.10. Three-dimensional Bode plots of the electrochemical impedance phase angle of the Au film electrode functionalized with different LB layers of the Fe20Pt80 nanoparticles: (a) LB1, (b) LB2, and (c) LB4 in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. Insets show the corresponding peak frequency at varied electrode potentials. 189   On the basis of the above voltammetric and impedance measurements, the equivalent circuit for multilayer films of FePt particle[25] was then used to model the impedance responses of the LB1, LB2, and LB4 thin films in this study (Figure 8.11). Figure 8.11a depicts the equivalent circuit for the electrodes that exhibit normal impedance behaviors, where RS represents the solution resistance, CPE (constant-phase element) and RCT are the double-layer capacitance and charge-transfer resistance, respectively. For negative impedance, the equivalent circuit is shown in Figure 8.11b, where C0 and R0 represent the capacitance and resistance of the electro-oxidation of adsorbed CO intermediates. Several representative fits (solid lines) at both negative and positive electrode potentials were shown in each of the Nyquist plots in Figure 8.9. From the fitting, it can be seen that the charge-transfer resistance (RCT) exhibits a very substantial dependence on the electrode potential for the three LB films (Figure 8.12), and more importantly, the charge-transfer resistance for formic acid electro-oxidation decreases with increasing thickness of the particle layers, LB1 > LB2 ≈ LB4, again in good agreement with the above voltammetric and impedance measurements. Furthermore, the charge-transfer resistances observed here with the particle LB thin films were all drastically smaller than those observed earlier[25] with a dropcast thick film of the same particles (approximately nine layers), which implies that the ordering of the electrocatalysts on the electrode surface might also play an important role in the optimization of the overall catalytic activities, as suggested in the aforementioned voltammetric studies. 190   Figure 8.11. Equivalent circuits for the electro-oxidation of formic acid on the Fe20Pt80 LB thin film electrodes: (a) for normal impedance and (b) for negative impedance shown in Figures 8.9 and 8.10. 191   Figure 8.12. Dependence of charge-transfer resistance (RCT) on electrode potentials for the electro-oxidation of formic acid at the three particle thin film electrodes. 192   8.4. Conclusion In this report, Fe20Pt80 particle layers of varied thickness were prepared by the LB technique and the corresponding electrocatalytic activities for the oxidation of formic acid in an acidic electrolyte were examined by voltammetry and EIS. The voltammetric responses exhibited rather drastic variation with the particle layer thickness. For thinner particle films (LB1 and LB2), the direct electro-oxidation of formic acid was impeded significantly by the adsorption of poisonous CO species (heavy CO poisoning), whereas at thicker films (LB4), the catalyst surface was only partially poisoned leading to the appearance of two oxidation peaks in the positive potential scan, similar to those with dropcast thick films of the same particles. Importantly, the current density for formic acid oxidation was found to be markedly larger than that with dropcast films of comparable particle layer thickness, signifying the critical role of particle surface accessibility and/ordering in the electrocatalytic performance. Consistent results were also observed in the impedance studies, where negative impedance in the Nyquist plots and the abrupt change of sign of the phase angle in the Bode plots were seen with all three particle layers, suggesting the oxidation of surface- adsorbed CO species and the appearance of an inductive electrode. Yet the charge- transfer resistance for formic acid electro-oxidation was found to decrease with increasing film thickness, though was all smaller than that observed with dropcast thick films of the same particles. These results suggest that there exists a minimal and optimal loading of the electrocatalysts in the electro-oxidation of formic acid, where the ordering 193   and loading of the electrocatalysts on the electrode surface might also play an important role. 194   REFERENCES 1. R. A. Lemons, J. Power Sources, 1990, 29, 251. 2. B. C. H. Steele, A. Heinzel, Nature, 2001, 414, 345. 3. A. Lewera, W. P. Zhou, C. Vericat, J. H. Chung, R. Haasch, A. Wieckowski, P. S. Bagus, Electrochim. Acta, 2006, 51, 3950. 4. B. Gurau, R. Viswanathan, R. X. Liu, T. J. Lafrenz, K. L. Ley, E. S. Smotkin, E. Reddington, A. Sapienza, B. C. Chan, T. E. Mallouk, S. Sarangapani, J. Phys. Chem. B, 1998, 102, 9997. 5. R. Dillon, S. Srinivasan, A. S. Arico, V. Antonucci, J. Power Sources, 2004, 127, 112. 6. N. Tian, Z.-Y. Zhou, S.-G. Sun, Y. Ding, Z. L. Wang, Science, 2007, 316, 732. 7. T. Frelink, W. Visscher, J. A. R. vanVeen, Langmuir, 1996, 12, 3702. 8. T. Iwasita, H. Hoster, A. John-Anacker, W. F. Lin, W. Vielstich, Langmuir, 2000, 16, 522. 9. S. Motoo, M. Watanabe, J. Electroanal. Chem., 1976, 69, 429. 10. R. S. Jayashree, J. S. Spendelow, J. Yeom, C. Rastogi, M. A. Shannon, P. J. A. Kenis, Electrochim. Acta, 2005, 50, 4674. 11. H. Yang, N. Alonso-Vante, J. M. Leger, C. Lamy, J. Phys. Chem. B, 2004, 108, 1938. 12. M. Baldauf, D. M. Kolb, J. Phys. Chem., 1996, 100, 11375. 13. T. Page, R. Johnson, J. Hormes, S. Noding, B. Rambabu, J. Electroanal. Chem., 195   2000, 485, 34. 14. K. W. Park, J. H. Choi, B. K. Kwon, S. A. Lee, Y. E. Sung, H. Y. Ha, S. A. Hong, H. Kim, A. Wieckowski, J. Phys. Chem. B, 2002, 106, 1869. 15. W. F. Lin, M. S. Zei, M. Eiswirth, G. Ertl, T. Iwasita, W. Vielstich, J. Phys. Chem. B, 1999, 103, 6968. 16. H. A. Gasteiger, N. Markovic, P. N. Ross, E. J. Cairns, J. Electrochem. Soc., 1994, 141, 1795. 17. H. N. Dinh, X. M. Ren, F. H. Garzon, P. Zelenay, S. Gottesfeld, J. Electroanal. Chem., 2000, 491, 222. 18. H. F. Oetjen, V. M. Schmidt, U. Stimming, F. Trila, J. Electrochem. Soc., 1996, 143, 3838. 19. J. B. Goodenough, A. Hamnett, B. J. Kennedy, R. Manoharan, S. A. Weeks, J. Electroanal. Chem., 1988, 240, 133. 20. T. Iwasita, X. H. Xia, E. Herrero, H. D. Liess, Langmuir, 1996, 12, 4260. 21. S. Park, Y. Xie, M. J. Weaver, Langmuir, 2002, 18, 5792. 22. L. B. Lai, D. H. Chen, T. C. Huang, J. Mater. Chem., 2001, 11, 1491. 23. S. T. Kuk, A. Wieckowski, J. Power Sources, 2005, 141, 1. 24. F. Gloaguen, J. M. Leger, C. Lamy, J. Appl. Electrochem., 1997, 27, 1052. 25. W. Chen, J. Kim, S. Sun, S. W. Chen, Phys. Chem. Chem. Phys., 2006, 8, 2779. 26. C. Lamy, A. Lima, V. LeRhun, F. Delime, C. Coutanceau, J. M. Leger, J. Power Sources, 2002, 105, 283. 27. I. A. Greene, F. X. Wu, J. Z. Zhang, S. W. Chen, J. Phys. Chem. B, 2003, 107, 196   5733. 28. S. W. Chen, Anal. Chim. Acta, 2003, 496, 29. 29. S. Pradhan, S. W. Chen, S. Z. Wang, J. Zou, S. M. Kauzlarich, A. Y. Louie, Langmuir, 2006, 22, 787. 30. A. C. Templeton, M. P. Wuelfing, R. W. Murray, Acc. Chem. Res., 2000, 33, 27. 31. M. Watanabe, Y. M. Zhu, H. Uchida, J. Phys. Chem. B, 2000, 104, 1762. 32. T. Toda, H. Igarashi, H. Uchida, M. Watanabe, J. Electrochem. Soc., 1999, 146, 3750. 33. X. H. Xia, T. Iwasita, J. Electrochem. Soc., 1993, 140, 2559. 34. N. M. Markovic, P. N. Ross, Surf. Sci. Rep., 2002, 45, 121. 35. C. Rice, S. Ha, R. I. Masel, A. Wieckowski, J. Power Sources, 2003, 115, 229. 36. H. Shiroishi, Y. Ayato, K. Kunimatsu, T. Okada, J. Electroanal. Chem., 2005, 581, 132. 37. P. Waszczuk, T. M. Barnard, C. Rice, R. I. Masel, A. Wieckowski, Electrochem. Commun., 2002, 4, 599. 38. Z. F. Liu, E. T. Ada, M. Shamsuzzoha, G. B. Thompson, D. E. Nikles, Chem. Mater., 2006, 18, 4946. 39. Z. G. Shao, F. Y. Zhu, W. F. Lin, P. A. Christensen, H. M. Zhang, J. Power Sources, 2006, 161, 813. 40. V. Selvaraj, M. Alagar, I. Hamerton, J. Power Sources, 2006, 160, 940. 41. C. Y. Kwong, W. C. H. Choy, A. B. Djurisic, P. C. Chui, K. W. Cheng, W. K. Chan, Nanotechnology, 2004, 15, 1156. 197   42. L. J. Zhang, Z. Y. Wang, D. G. Xia, J. Alloys Compd., 2006, 426, 268. 43. A. V. Tripkovic, K. D. Popovic, R. M. Stevanovic, R. Socha, A. Kowal, Electrochem. Commun., 2006, 8, 1492. 44. D. Aberdam, R. Durand, R. Faure, F. Gloaguen, J. L. Hazemann, E. Herrero, A. Kabbabi, O. Ulrich, J. Electroanal. Chem., 1995, 398, 43. 45. Y. C. Liu, X. P. Qiu, W. T. Zhu, G. S. Wu, J. Power Sources, 2003, 114, 10. 46. W. Sugimoto, K. Aoyama, T. Kawaguchi, Y. Murakami, Y. Takasu, J. Electroanal. Chem., 2005, 576, 215. 47. I. M. Hsing, X. Wang, Y. J. Leng, J. Electrochem. Soc., 2002, 149, A615. 48. A. C. Chen, D. J. La Russa, B. Miller, Langmuir, 2004, 20, 9695. 49. G. Markovich, C. P. Collier, J. R. Heath, Phys. Rev. Lett., 1998, 80, 3807.   Chapter 9 Composition Effects of FePt Alloy Nanoparticles on the Electro-Oxidation of Formic Acid 199   9.1. Introduction Fuel cells have been hailed as an important power source in the future because of the high energy conversion efficiency and low environmental pollution.[1] Up to now, one of the major problems in small molecule (e.g., methanol or formic acid) fuel cells is the poisoning of the electrocatalysts by CO formed during the incomplete oxidation of the organic fuels. To improve the performance in fuel oxidation, the combination of Pt and other transition metals has been examined extensively as effective catalysts for methanol and formic acid electro-oxidation, as Pt-based alloy catalysts typically display enhanced catalytic activity toward methanol and formic acid oxidation that has been attributed to the so-called bifunctional and/or electronic (ligand) effect mechanism.[2-12] In previous studies of metal nanoparticle catalysts, it has been found that the electrocatalytic activity depends strongly on the particle dimensions and surface morphology because of the variation of the density of active sites such as atomic steps, edges, and kinks.[13-16] For example, a recent report demonstrates that platinum n0anoparticles with high-index facets exhibit unusually high catalytic activities in the electro-oxidation of small organic fuels.[16] In another study,[13] the effects of the size of palladium nanoparticles on the electrocatalytic oxidation of formic acid were examined, and it was found that the smallest Pd nanoparticles (9 and 11 nm) exhibited the best catalytic activity within the size range of 9−40 nm. For Pt-based binary alloy catalysts, Pt and the second metal play different roles in the oxidation catalysis. According to the so- called bifunctional mechanism, the role of the second metal is to dissociate water to form adsorbed OH species, which then react with CO adsorbed on the Pt surface to generate 200   CO2.[17,18] This suggests that for alloy nanoparticle catalysts, in addition to particle dimensions, the composition of the bimetallic particles also acts as a very important parameter to manipulate its catalytic activity. Such composition effects have been observed for CO and methanol electro-oxidation with PtRu alloy catalysts.[6,19-25] It was found that for direct methanol fuel cells (DMFC), the optimum catalyst corresponded to a composition consisting of 1:1 atomic ratio of Ru/Pt.[20,21] Yet, so far, the studies of the influence of the catalyst composition on the electrocatalytic activity have been mostly confined to bulk alloy electrodes. Systematic studies of alloy nanoparticles, however, are still scarce due to the difficulty in the synthesis of alloy nanoparticles of controlled compositions. In our previous studies,[26,27] Fe20Pt80 nanoparticles were found to exhibit very good electrocatalytic activities for formic acid electro-oxidation. In this paper, a series of (almost) monodispersed FexPt100-x alloy nanoparticles with different atomic ratios of Fe to Pt were prepared by a chemical reduction method. The FexPt100-x nanoparticles (average diameter ~2.5 nm) were deposited onto a gold electrode surface (denoted as FexPt100- x/Au), and the electrochemical performance in formic acid oxidation was then examined by voltammetric and electrochemical impedance measurements. The electrocatalytic activities were compared on the basis of (steady-state) current density and tolerance to CO adsorption of formic acid oxidation. It was found that within the present experimental context, the FexPt100-x particles at x ≈ 50 exhibited the optimal composition for formic acid electro-oxidation. 201   9.2. Experimental 9.2.1. Chemicals Perchloric acid (HClO4, Fisher, 99.999%) and formic acid (HCOOH, ACROS, 99%) were used as received. All solvents were obtained from typical commercial sources at their highest purity and used as received as well. Water was supplied by a Barnstead Nanopure water system (18.3 MΩ·cm). 9.2.2. Nanoparticle Preparation The synthesis and characterization of the FexPt100-x alloy nanoparticles stabilized by oleylamine and oleic acid have been described previously.[26,28] In a typical experiment, 20 mL of dioctylether or benzyl ether was mixed with 0.5 mmol of Pt(acac)2 and 1.5 mmol of 1,2-hexadecanediol. Under a gentle N2 flow, the mixture was heated to 100 °C. At this temperature, 0.5 mmol of oleic acid, 0.5 mmol of oleylamine, and 1.0 mmol of Fe(CO)5 (the composition was varied by the amount of Fe(CO)5, from 0.5 to 2.5 mmol) were added under a N2 blanket, and the mixture was heated to reflux (297 °C) for 30 min before it was cooled down to room temperature. The product was precipitated by adding ethanol and separated by centrifugation. The supernatant was discarded, and the precipitate was dispersed in hexane. Then, an excess of ethanol was added to precipitate out the particles, and the suspension was centrifuged again. By repeating the dispersion−precipitation cycle, excess free ligands were removed affording purified particle samples. 202   The size of the FePt nanoparticles was then characterized by transmission electron microscopy (TEM, Philips EM420, 120 kV), and energy-dispersive X-ray analysis (EDX) was carried out to evaluate the particle compositions by using an INCA accessory (from OXFORD instruments) that was attached onto a scanning electron microscope (SEM, Zeiss (LEO) 1530VP FESEM). UV−vis spectroscopic studies were performed with an ATI Unicam UV4 spectrometer using a 1 cm quartz cuvette with a resolution of 2 nm. The particles were dissolved in dichloromethane at a concentration of ca. 0.1 mg/mL. Only featureless Mie characters were observed in the absorption spectra of these alloy nanoparticles (Figure 9.1). Figure 9.1. UV-visible spectra of FexPt100-x nanoparticles (shown as figure legends). All the particle concentrations are 0.1 mg/mL in CH2Cl2. 203   9.2.3. Preparation of the FePt/Au Electrode The procedure has been described previously.[26] Briefly, a polycrystalline gold disk electrode (sealed in a glass tubing) was first polished with alumina slurries (0.05 μm) and then cleansed by sonication in 0.1 M HNO3, H2SO4, and Nanopure water for 10 min successively. A volume of 4 μL of FexPt100-x nanoparticles dissolved in CH2Cl2 (1.2 mg/mL) was then dropcast onto the clean Au electrode surface by a Hamilton microliter syringe. The particle film was dried by a gentle nitrogen flow for ca. 2 min. The organic protecting ligands were then removed by oxidation in an ultraviolet ozone (UVO) chamber (Jelight Company, Inc., model 42) for about 15 min. The particle film was then rinsed with excessive ethanol and Nanopure water to remove loosely bound particles and remaining organic deposits. 9.2.4. Electrochemistry Voltammetric measurements were carried out with a CHI 440 electrochemical workstation. The FexPt100-x/Au electrode was used as the working electrode. A Ag/AgCl wire and a Pt coil were used as the reference and counter electrodes, respectively. All electrode potentials in the present study will be referred to this Ag/AgCl quasi-reference. Electrochemical impedance spectroscopy (EIS) measurements were carried out using an EG&G PARC potentiostat/galvanostat (model 283) and frequency response detector (model 1025). The impedance spectra were recorded between 100 kHz and 10 mHz with the amplitude (rms value) of the ac signal 10 mV. In these voltammetric and impedance measurements, the solutions were deaerated by bubbling ultrahigh-purity N2 for 20 min and protected with a nitrogen atmosphere during the entire experimental procedure. 204   9.3. Results and Discussion 9.3.1. Characterization of FexPt100-x Nanoparticles Figure 9.2 depicts the representative TEM micrographs of the FexPt100-x nanoparticles along with the particle core size histograms as the corresponding insets. It can be seen that the vast majority of the particles exhibit a core diameter close to 2.5 nm with no apparent aggregation, and the core size distributions are all very narrow. The composition of the FexPt100-x nanoparticles was tuned by simply varying the stoichiometic ratio of the Fe and Pt precursors in the synthesis, where the value of x was found to vary from 10 to 63, as estimated from EDX analysis (Figure 9.3). Table 9.1 summarizes the average size, size distribution, and the corresponding composition of the six FePt particle samples shown in Figure 9.2. Because of the uniformity of the particle size and shape, we anticipate that the electrocatalytic activity of these six samples of nanoparticles for formic acid oxidation can then be directly correlated to the particle composition.                 205       Figure 9.2. Representative TEM micrographs of the series of FePt nanoparticles used in this study: (A) Fe10Pt90, (B) Fe15Pt85, (C) Fe42Pt58, (D) Fe54Pt46, (E) Fe58Pt42, and (F) Fe63Pt37. The scale bars are all 20 nm. The inset shows the corresponding particle core size histogram. 206   Figure 9.3. Energy-dispersive x-ray spectra of FexPt100-x nanoparticles: (A) Fe10Pt90; (B) Fe15Pt85; (C) Fe42Pt58; (D) Fe54Pt46; (E) Fe58Pt42; and (F) Fe63Pt37. 207   Table 9.1. Average Core Size, Size Distribution, and Composition of the FexPt100-x Nanoparticles particle A B C D E F core size (nm)a 2.44 ± 0.70 2.47 ± 0.56 2.35 ± 0.18 2.52 ± 0.81 2.48 ± 0.38 2.42 ± 0.61 core compositionb Fe10Pt90 Fe15Pt85 Fe42Pt58 Fe54Pt46 Fe58Pt42 Fe63Pt37 a Particle core sizes were derived from TEM images as exemplified in Figure 9.2. b Particle core compositions were evaluated by energy-dispersive X-ray analysis (EDX spectra are showed in Figure 9.3). 208   9.3.2. Cyclic Voltammetry of FexPt100-x Nanoparticles As discussed in our previous report,[26] in order to render the FePt alloy particles electrochemically active, all the FePt particles modified Au electrodes were treated in UVO for 15 min to remove the organic protecting ligands. Figure 9.4 shows the steady- state cyclic voltammograms (CV) of the resulting FexPt100-x (x = 10, 15, 42, 54, 58, and 63) nanoparticles deposited onto a Au electrode in 0.1 M HClO4 at a potential sweep rate of 0.1 V/s. It should be noted that Pt-based metal alloys usually exhibit the formation of a platinum skin after electrochemical cycling in acidic electrolytes, which is caused by the Pt dissolution from the alloy and then redeposition and rearrangement on the surface.[29- 33] From Figure 9.4, it can be seen that the voltammetric features of all the FexPt100-x/Au electrodes are similar to those at a polycrystalline Pt electrode, suggesting the formation of a Pt skin on the alloy nanoparticle surface after electrochemical activation in 0.1 M HClO4. First, the characteristic features for hydrogen adsorption−desorption at low electrode potentials can be clearly observed. For instance, for the Fe42Pt58/Au electrode, one can see two pairs of well-defined current peaks at −0.43 and −0.31 V for hydrogen adsorption−desorption. Pt oxides are formed in the anodic scan at potentials above +0.3 V. In the reverse scan, a voltammetric peak for the reduction of Pt oxides can be observed at about +0.17 V. Similar responses can be seen with all other alloy particles, except for a broad and featureless hydrogen adsorption−desorption region. Second, by comparing the CV profiles of the FexPt100-x nanoparticles, it can be seen that the peak currents for the reduction of Pt oxides and for hydrogen adsorption−desorption become more prominent for the alloy nanoparticles with a higher Pt content, indicating a larger/thicker Pt skin 209   layer on the particle surface. These electrodes were then employed in the subsequent studies of electro-oxidation of formic acid. Figure 9.4. Cyclic voltammograms of the FexPt100-x/Au electrodes in 0.1 M HClO4. The potential scan rates were 0.1 V/s. 210   9.3.3. Electro-Oxidation of Formic Acid Figure 9.5 depicts the steady-state cyclic voltammograms of formic acid oxidation at the six FexPt100-x/Au electrodes prepared above. Note that the voltammetric currents have been normalized to the actual active surface areas of the respective electrodes which are calculated on the basis of the charge associated with hydrogen desorption (anodic current between −0.5 and −0.2 V in Figure 9.4), assuming that hydrogen desorption yields 210 μC/cm2 of the Pt surface area.[34,35] It can be seen that the overall voltammetric behaviors are strongly dependent on the nanoparticle composition. Figure 9.5. Cyclic voltammograms of the FexPt100-x/Au electrodes in 0.1 M HCOOH + 0.1 M HClO4. The potential scan rates were 0.1 V/s. 211   For the Fe10Pt90 nanoparticles, there are two voltammetric peaks at +0.083 and +0.378 V in the anodic scan, similar to the voltammetric features of HCOOH oxidation at Fe20Pt80 nanoparticle modified electrodes.[26,27] The first oxidation peak (EP1 = +0.083 V) is attributed to the direct oxidation of formic acid into CO2 on the particle surface that is free of adsorbed poisonous intermediates (e.g., CO), i.e., the direct path, whereas the second peak at a more positive potential (EP2 = +0.378 V) is ascribed to the oxidation of CO species adsorbed on the particle surface that arise from the nonfaradaic dissociation of formic acid (the so-called indirect path). Additional current contributions may come from the direct oxidation of formic acid at the recovered particle surface upon CO oxidation. In the cathodic scan, an oxidation peak at +0.14 V is observed which is assigned to the direct oxidation of formic acid with the entire catalyst surface free of CO poisoning. Thus, one can see that the ratio of the current density of the first anodic peak (Ja) to the cathodic peak (Jc) essentially reflects the fraction of the catalyst surface that is not poisoned by CO adsorption and can be used to measure the catalyst tolerance to CO poisoning. For the Fe10Pt90 nanoparticles, the Ja/Jc ratio is 0.29, suggesting that initially about 70% of the catalyst surface was poisoned by CO adsorption. Similar voltammetric behaviors can also be observed with Fe58Pt42, Fe54Pt46, and Fe15Pt85 nanoparticles. The corresponding Ja/Jc ratios are found to be 0.42, 0.47, and 0.52, respectively, indicative of reduced CO poisoning as compared to the Fe10Pt90 particles. In contrast, the Fe42Pt58 and Fe63Pt37 particles exhibit drastically different voltammetric responses (Figure 9.5). For the low-platinum Fe63Pt37 particles, only one anodic peak is observed at +0.40 V, whereas the oxidation peak in the cathodic scan 212   appears at +0.017 V, implying extensive poisoning of the catalyst surface by CO adsorption and, hence, a very low Ja/Jc ratio (0.17). For the Fe42Pt58 nanoparticles, an opposite behavior can be seen (Figure 9.5). Here the voltammetric responses of HCOOH oxidation also show only one oxidation current peak in the anodic and cathodic scan. However, in comparison to those of the Fe10Pt90 particles, the peak potentials shift cathodically to +0.045 and +0.088 V in the anodic and cathodic scan, respectively, suggesting much enhanced electrocatalytic activities. More importantly, the absence of a second anodic peak at a more positive potential indicates little CO generation/adsorption during formic acid electro-oxidation. Such excellent tolerance to CO adsorption is further confirmed by the observation that the current density of the oxidation peak in the anodic scan is almost equal to that in the cathodic scan, with a corresponding Ja/Jc ratio close to 0.99. This implies that the voltammetric currents mostly arise from the direct oxidation of formic acid into CO2. Figure 9.6 summarizes the Ja/Jc ratios of the six FexPt100-x alloy particles under study which clearly exhibit a peak-shaped variation with particle compositions. It is apparent that the Fe42Pt58 particles represent the optimum composition among the group of alloy catalysts. For alloy particles with too high or low a platinum content (e.g., Fe10Pt90, and Fe63Pt37), extensive CO poisoning occurs leading to minimum catalytic activities.         213     Figure 9.6. Ratio of the anodic and cathodic peak currents for the direct oxidation of formic acid at varied nanoparticle composition. The results were obtained from data presented in Figure 9.5. 214   Similar assessments of the electrocatalytic activities can be made by the comparison of the anodic current density as well as the onset potential. For instance, from Figure 9.5 the current densities of the (first) anodic peak of the six FexPt100-x alloy nanoparticles (x = 10, 15, 42, 54, 58, and 63) can be estimated to be 1.94, 1.32, 104.02, 61.18, 42.22, and 0.21 mA/cm2, respectively. Of these, the Fe42Pt58 particles exhibit the highest current density. In addition, the onset potentials for formic acid oxidation can be found at −0.33, −0.37, −0.40, −0.34, −0.39, and −0.25 V, respectively. Again, Fe42Pt58 stands out with the most negative onset potential for formic acid oxidation among the series. To further evaluate the activity and stability of the FexPt100-x catalysts for formic acid oxidation, chronoamperometric analyses were also carried out by stepping the potential from Ei = −0.40 V to Ef = −0.20, +0.05, and +0.20 V in 0.1 M HCOOH + 0.1 M HClO4, which corresponds, respectively, to the onset, peak, and end of the formic acid direct oxidation at the catalytic sites that are not poisoned by CO intermediates, as shown in Figure 9.7A−C. From the chronoamperometric curves at Ef = −0.20 V (Figure 9.7A), one can see that the maximum initial and steady-state oxidation current density was obtained with the Fe42Pt58/Au electrode followed by Fe58Pt42/Au and Fe54Pt46/Au, and the oxidation current density at the Fe15Pt85, Fe10Pt90 and Fe63Pt37 electrodes is much smaller. At Ef = +0.05 V (Figure 9.7B), which corresponds to the voltammetric peak in the CV studies of formic acid oxidation (Figure 9.5), again the Fe42Pt58/Au electrode shows the highest oxidation current density, and the electrodes exhibit the same activity and stability sequence as that at Ef = −0.2 V. For chronoamperometric curves at Ef = +0.20 V (Figure 9.7C), whereas the Fe54Pt46/Au electrode exhibits a higher initial oxidation 215   current density than that on the Fe42Pt58/Au electrode, the current density decays rapidly. In contrast, the Fe42Pt58/Au electrode maintains a rather steady current profile, and at t > 150 s, the current density becomes the largest among the series. These chronoamperometric data indicate that Fe42Pt58 has the best activity and stability for formic acid oxidation among the series of FexPt100-x particles used in the present study, in good agreement with the CV results. Figure 9.7. Chronoamperometric curves in 0.1 M HCOOH + 0.1 M HClO4 for different FexPt100-x/Au electrodes (shown in the figure legends) where the potential was stepped from Ei = −0.4 V to Ef = −0.02 (A), +0.05 (B), and +0.20 V (C). 216   On the basis of the above CV and chronoamperometric evaluations, it can be seen that the catalytic activity decreases in the sequence of Fe42Pt58 > Fe54Pt46 ≈ Fe58Pt42 > Fe15Pt85 > Fe10Pt90 > Fe63Pt37. That is, the FexPt100-x particles at x ≈ 50 show the highest electrocatalytic activity in HCOOH oxidation among the series. For Pt-based bimetallic alloy electrocatalysts, the optimum composition should maximize the adsorption of HCOOH on the Pt active sites and, concurrently, there should be enough surface sites of the second metals which promote the effective oxidative removal of poisonous intermediates (e.g., CO) with adsorbed hydroxyl species.[18,36,37] For the FexPt100-x alloy particles in the present study, the optimum composition appears to correspond to a Fe/Pt atomic ratio of ca. 1:1, which is very similar to the PtRu catalysts used in previous investigations.[20,21] 217   9.3.4. Electrochemical Impedance Studies Electrochemical impedance spectroscopy has been used as a powerful and sensitive technique to study the kinetics of electron-transfer processes. Here we carry out electrochemical impedance studies to examine the electro-oxidation dynamics of formic acid catalyzed by FexPt100-x alloy nanoparticles. Overall the EIS responses are very consistent with the respective voltammetric results as shown in Figures 9.5−7. Figure 9.8 depicts the complex-plane (Nyquist) impedance plots of the Fe10Pt90/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 with electrode potentials varied from −0.50 to +0.80 V (shown as figure legends). It can be seen that, at potentials below +0.00 V, the impedance arcs are located within the first quadrant and the diameter of the arcs increases slightly from −0.50 to −0.20 V and then decreases somewhat when the potential shifts positively from −0.20 to 0 V. The former may be ascribed to the formation and adsorption of CO intermediate species on the electrode surface, and the latter may be related to the onset of electro-oxidation of formic acid. With further increase of the electrode potentials (+0.10 to +0.30 V), the impedance arcs remain in the first quadrant, but the slope of the linear portions exhibit a rather rapid increase suggesting an increasingly dominant diffusion-controlled component because of the potential-enhanced electron-transfer kinetics. This agrees well with the voltammetric studies in Figure 9.5, as this potential region corresponds to the direct oxidation of formic acid at the catalytic sites that are not poisoned by CO adsorption. Additionally, the impedance arc starts to bend from the positive to negative direction of the x-axis at +0.30 V. When the electrode potential reaches +0.35 to ~ +0.40 V, the impedances actually appear in the second quadrant. The negative faradaic impedance suggests the presence of 218   an inductive component. This is attributed to the formation of chemisorbed hydroxyl species in this potential range, which enhances the oxidative removal of the adsorbed CO intermediate, consistent with the voltammetric response (Figure 9.5) where a second anodic peak is observed. At more positive electrode potentials (≥+0.40 V), the impedance arcs return to the first quadrant and the diameter of the arcs decreases with increasing potential, indicative of diminishing charge-transfer resistance. Overall, the impedance results are in very good agreement with voltammetric data presented above (Figure 9.5). Similar behaviors have been observed with the Fe15Pt85 andFe58Pt42 alloy particles in Figures 9.9 and 9.10, respectively, as well as with the Fe20Pt80 nanoparticles[26,27] that was reported previously. 219   Figure 9.8. Complex-plane (Nyquist) impedance plots of the Fe10Pt90/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. The lines are representative simulations based on the equivalent circuits in Figure 9.20 for the impedance data at the electrode potentials of −0.4 and +0.35 V. The inset is the magnification of the impedance spectra in a smaller range. 220   Figure 9.9. Complex-plane electrochemical impedance plots (Nyquist plots) of Fe15Pt85/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. Line is a representative simulation based on the equivalent circuits in Figure 9.20 for the impedance data at the electrode potential of +0.10 V. 221   Figure 9.10. Complex-plane electrochemical impedance plots (Nyquist plots) of Fe58Pt42/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. Lines are representative simulations based on the equivalent circuits in Figure 9.20 for the impedance data at the electrode potentials of +0.10 and +0.2 V. 222   Figure 9.11 shows the impedance spectra of the Fe42Pt58/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at different electrode potentials. First, similarly, the impedance arcs are found mostly in the first quadrant (top panel), and the arc diameter first increases (from −0.50 to −0.30 V) and then decreases (from −0.30 to 0.10 V) with increasing electrode potential. The minimal arc diameter observed at +0.1 V appears to coincide with the voltammetric peak observed in Figure 9.5. With further increase of the electrode potential, the arc diameter increases correspondingly, and interestingly, the low- frequency ends of the arcs start to extend into the fourth quadrant (+0.15 to +0.20 V), indicating the presence of a small pseudoinductive component as a consequence of the oxidative removal of CO adsorbed on the particle surface. However, when the electrode potentials become more positive than +0.30 V (bottom panel), a drastic variation of the impedance spectra can be observed. First, the impedance spectra even show up in the second and third quadrants within the potential range of +0.30 to +0.50 V and then return to the first and fourth quadrants at higher potentials (≥+0.55 V). Such interesting impedance patterns have also been reported in the studies of methanol electro-oxidation at Pt/C thin film electrodes.[38] They are ascribed to the variation of the rate-determining step at different potential positions, namely, the formation of adsorbed CO by dehydrogenation reaction of HCOOH and the electro-oxidation of adsorbed CO through adsorbed OH species. At potentials more negative than +0.3 V, the rate-determining step is considered to be the oxidation of adsorbed CO intermediates, leading to the appearance of the pseudoinductive behavior. At more positive potentials, the electro-oxidation of surface adsorbates will be drastically accelerated and the formation of adsorbed CO by 223   the dehydrogenation reaction of HCOOH now becomes the rate-determining step. This gives rise to the inductive response.[38] Figure 9.11. Complex-plane (Nyquist) impedance plots of the Fe42Pt58/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. The lines are representative simulations based on the equivalent circuits in Figure 9.20 for the impedance data at the electrode potentials of −0.40 and +0.60 V. The insets are the magnification of the impedance spectra in a smaller range. 224   In contrast, for the Fe54Pt46/Au electrode, all the impedance spectra are located in the first quadrant within the entire potential range of −0.50 to +0.80 V, as shown in Figure 9.12. This indicates that there exist only resistive behaviors for HCOOH electro- oxidation at the Fe54Pt46/Au electrode. The lack of (pseudo)inductive characters of the impedance spectra seems to imply a relatively strong adsorption (and retarded oxidation kinetics) of CO on the particles surface, as voltammetric measurements (Figure 9.5) exhibit only a very weak current peak for the oxidation of adsorbed CO intermediates. In general, the impedance is rather comparable to that observed in Figure 9.8 for the Fe10Pt90 particles. Similar features are also observed with the low-platinum Fe63Pt37 nanoparticles (Figure 9.13), except that the overall impedance is substantially larger.                          Figure 9.12. Complex-plane (Nyquist) impedance plots of the Fe54Pt46/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. The line is a representative simulation based on the equivalent circuits in Figure 9.20 for the impedance data at the electrode potential of +0.20 V. 225   Figure 9.13. Complex-plane electrochemical impedance plots (Nyquist plots) of Fe63Pt37/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. Line is a representative simulation based on the equivalent circuits in Figure 9.20 for the impedance data at the electrode potentials of –0.30 V. 226   The corresponding Bode plots of the impedance spectra for the series of FexPt100-x alloy particles are showed in Figures 9.14-19. Overall, the characteristic frequency (fmax) at the maximum phase angle is found to increase with increasing electrode potentials, indicative of potential-enhanced electron-transfer kinetics. Additionally, within the potential range where surface-adsorbed CO is oxidized, an abrupt jump between the positive and negative values of the phase angle can be observed. This interesting behavior is attributed to the (pseudo) inductive characteristics upon the oxidative removal of CO that gives rise to the negative faradaic impedance as shown in Figures 9.8, 9.11, and 9.12. Figure 9.14. Bode plots of the electrochemical impedance of Fe10Pt90/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. 227   Figure 9.15. Bode plots of the electrochemical impedance of Fe15Pt85/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. 228   Figure 9.16. Bode plots of the electrochemical impedance of Fe42Pt58/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. 229   Figure 9.17. Bode plots of the electrochemical impedance of Fe54Pt46/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. 230   Figure 9.18. Bode plots of the electrochemical impedance of Fe58Pt42/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. 231   Figure 9.19. Bode plots of the electrochemical impedance of Fe63Pt37/Au electrode in 0.1 M HCOOH + 0.1 M HClO4 at various electrode potentials. 232   On the basis of the above voltammetric and impedance measurements, two equivalent circuits are proposed in Figure 9.20, parts a and b, which are used to fit the impedance spectra corresponding to normal and negative impedance behaviors, respectively. Here RS represents the solution resistance, CPE (constant-phase element) is the electrode double layer capacitance, RCT is the charge-transfer resistance, and C0 and R0 represent the capacitance and resistance of the electro-oxidation of adsorbed CO intermediates, respectively. Some representative fittings are shown as solid lines in Figures 9.8, 9.11, and 9.12, which show very good agreement with the corresponding experimental data. Figure 9.20. Equivalent circuits for the electro-oxidation of formic acid at FexPt100-x/Au electrodes. 233   Figure 9.21 depicts the variation of the charge-transfer resistance (RCT) with the electrode potential for the series of FexPt100-x alloy nanoparticles. First, negative RCT can be found within the potential range where (pseudo)inductive characters arise from the electro-oxidation of surface-adsorbed CO species. Second, the RCT for formic acid oxidation at the Fe42Pt58/Au electrode appears to reach a minimum (a few hundred ohms) as compared to other alloy particles in the series (of the order of a few thousand ohms), again, confirming that this represents the optimum particle composition for formic acid electro-oxidation. Third, for Fe63Pt37 particles, the charge-transfer resistance (RCT) is at least an order of magnitude greater than those of other particle catalysts, in agreement with the low electrocatalytic activities as shown in the voltammetric and impedance measurements. 234   Figure 9.21. Charge-transfer resistance (RCT) at different electrode potentials. Data are obtained by curve fitting of the impedance spectra (Figures 9.8−13) by the equivalent circuits in Figure 9.20. The inset shows the magnification in a smaller range. 235   9.4. Conclusion In this study, CV, chronoamperometry, and EIS were employed to examine the electrocatalytic activities of a series of FexPt100-x alloy nanoparticles (x = 10, 15, 42, 54, 58, and 63) in the oxidation of formic acid in acid electrolytes. Three parameters in the CV studies were used as the indicators to evaluate the corresponding catalytic performance, including anodic oxidation current density, onset potential, as well as tolerance to CO poisoning. Our measurements suggest that the catalytic activity depends on the composition of the alloy nanoparticles, i.e., the catalytic activity decreases in the sequence of Fe42Pt58 > Fe54Pt46 ≈ Fe58Pt42 > Fe15Pt85 > Fe10Pt90 > Fe63Pt37. Consistent evaluation was observed in chronoamperometric measurements where the Fe42Pt58 particles exhibited the maximum current density and stability. In EIS measurements, for particles with excellent CO tolerance, negative impedance was observed at potentials where CO was removed by electro-oxidation, indicating the presence of an inductive component. However, for the nanoparticles heavily poisoned by CO, only normal impedance profiles were observed. These impedance measurements agree well with the CV results. Within the present experimental context, the FexPt100-x particles at x ≈ 50 (i.e., the atomic ratio of Fe/Pt close to 1) represent the optimum alloy composition that exhibits the maximum anodic current density, most negative onset potential, and strongest tolerance to CO poisoning. 236   REFERENCES 1. R. Dillon, S. Srinivasan, A. S. Arico, V. Antonucci, J. Power Sources, 2004, 127, 112. 2. H. S. Wang, C. Wingender, H. Baltruschat, M. Lopez, M. T. Reetz, J. Electroanal. Chem., 2001, 509, 163. 3. K. W. Park, J. H. Choi, B. K. Kwon, S. A. Lee, Y. E. Sung, H. Y. Ha, S. A. Hong, H. Kim, A. Wieckowski, J. Phys. Chem. B, 2002, 106, 1869. 4. F. Colmati, E. Antolini, E. R. Gonzalez, Electrochim. Acta, 2005, 50, 549. 5. E. Antolini, J. R. C. Salgado, E. R. Gonzalez, J. Electroanal. Chem., 2005, 580, 145. 6. S. Motoo, M. Watanabe, J. Electroanal. Chem., 1976, 69, 429. 7. C. Roychowdhury, F. Matsumoto, V. B. Zeldovich, S. C. Warren, P. F. Mutolo, M. Ballesteros, U. Wiesner, H. D. Abruña, F. DiSalvo, J. Chem. Mater., 2006, 18, 3365. 8. R. S. Jayashree, J. S. Spendelow, J. Yeom, C. Rastogi, M. A. Shannon, P. J. A. Kenis, Electrochim. Acta, 2005, 50, 4674. 9. W. F. Lin, M. S. Zei, M. Eiswirth, G. Ertl, T. Iwasita, W. Vielstich, J. Phys. Chem. B, 1999, 103, 6968. 10. T. Frelink, W. Visscher, J. A. R. vanVeen, Langmuir, 1996, 12, 3702. 11. T. Page, R. Johnson, J. Hormes, S. Noding, B. Rambabu, J. Electroanal. Chem., 2000, 485, 34. 237   12. L. B. Lai, D. H. Chen, T. C. Huang, J. Mater. Chem., 2001, 11, 1491. 13. W. P. Zhou, A. Lewera, R. Larsen, R. I. Masel, P. S. Bagus, A. Wieckowski, J. Phys. Chem. B, 2006, 110, 13393. 14. K. Bergamaski, A. L. N. Pinheiro, E. Teixeira-Neto, F. C. Nart, J. Phys. Chem. B, 2006, 110, 19271. 15. K. J. J. Mayrhofer, B. B. Blizanac, M. Arenz, V. R. Stamenkovic, P. N. Ross, N. M. Markovic, J. Phys. Chem. B, 2005, 109, 14433. 16. N. Tian, Z.-Y. Zhou, S.-G. Sun, Y. Ding, Z. L. Wang, Science, 2007, 316, 732. 17. T. Iwasita, Electrochim. Acta, 2002, 47, 3663. 18. Y. Y. Tong, H. S. Kim, P. K. Babu, P. Waszczuk, A. Wieckowski, E. Oldfield, J. Am. Chem. Soc., 2002, 124, 468. 19. Y. Wang, E. R. Fachini, G. Cruz, Y. Zhu, Y. Ishikawa, J. A. Colucci, C. R. Cabrera, J. Electrochem. Soc., 2001, 148, C222. 20. M. Watanabe, M. Uchida, S. Motoo, J. Electroanal. Chem., 1987, 229, 395. 21. H. N. Dinh, X. M. Ren, F. H. Garzon, P. Zelenay, S. Gottesfeld, J. Electroanal. Chem., 2000, 491, 222. 22. L. Dubau, C. Coutanceau, E. Garnier, J. M. Leger, C. Lamy, J. Appl. Electrochem., 2003, 33, 419. 23. T. Iwasita, H. Hoster, A. John-Anacker, W. F. Lin, W. Vielstich, Langmuir, 2000, 16, 522. 24. H. Hoster, T. Iwasita, H. Baumgartner, W. Vielstich, W. Phys. Chem. Chem. Phys., 2001, 3, 337. 238   25. Z. G. Shao, F. Y. Zhu, W. F. Lin, P. A. Christensen, H. M. Zhang, J. Power Sources, 2006, 161, 813. 26. W. Chen, J. Kim, S. Sun, S. W. Chen, Phys. Chem. Chem. Phys., 2006, 8, 2779. 27. W. Chen, J. Kim, L.-P Xu, S. Sun, S. W. Chen, J. Phys. Chem. C, 2007, 111, 13452. 28. S. Sun, C. B. Murray, D. Weller, L. Folks, A. Moser, Science, 2000, 287, 1989. 29. T. Toda, H. Igarashi, H. Uchida, M. Watanabe, J. Electrochem. Soc., 1999, 146, 3750. 30. M. Watanabe, K. Tsurumi, T. Mizukami, T. Nakamura, P. Stonehart, J. Electrochem. Soc., 1994, 141, 2659. 31. H. Igarashi, T. Fujino, Y. M. Zhu, H. Uchida, M. Watanabe, Phys. Chem. Chem. Phys., 2001, 3, 306. 32. L. J. Wan, T. Moriyama, M. Ito, H. Uchida, M. Watanabe, Chem. Commun., 2002, 58. 33. Y. Xu, A. V. Ruban, M. Mavrikakis, J. Am. Chem. Soc., 2004, 126, 4717. 34. H. C. Ye, R. M. Crooks, J. Am. Chem. Soc., 2007, 129, 3627. 35. V. R. Stamenkovic, B. S. Mun, M. Arenz, K. J. J. Mayrhofer, C. A. Lucas, G. F. Wang, P. N. Ross, N. M. Markovic, Nat. Mater., 2007, 6, 241. 36. M. Watanabe, S. Motoo, J. Electroanal. Chem., 1975, 60, 275. 37. J. Kua, W. A. Goddard, J. Am. Chem. Soc., 1999, 121, 10928. 38. I. M. Hsing, X. Wang, Y. J. Leng, J. Electrochem. Soc., 2002, 149, A615.