synthesis of nanocatalysts and their applications to

SYNTHESIS OF NANOCATALYSTS AND THEIR APPLICATIONS TO HYDROGEN PRODUCTION BY ELECTROLYSIS OF WATER AND HYDROLYSIS OF CHEMICAL HYDRIDES A THESIS SUBMITTED TO THE

UNIVERSITY OF MUMBAI FOR THE

Ph. D. (SCIENCE) DEGREE IN PHYSICS

SUBMITTED BY

GUPTA SURAJ GULAB SONKALI

UNDER THE GUIDANCE OF

PROF. D. C. KOTHARI AND CO-GUIDANCE OF

DR. NAINESH PATEL

DEPARTMENT OF PHYSICS, UNIVERSITY OF MUMBAI VIDYANAGARI, SANTACRUZ (E) MUMBAI- 400 098, INDIA.

April-2016

Thesis submitted by Mr. Gupta Suraj Gulab Sonkali to the University of Mumbai for degree of Doctor of Philosophy in Physics in the Faculty of Science Title of the thesis

Synthesis of Nanocatalysts and Their Applications to Hydrogen Production by Electrolysis of Water and Hydrolysis of Chemical Hydrides

Name of the candidate

Gupta Suraj Gulab Sonkali

Name of the research guide

Prof. D. C. Kothari, Professor, Department of Physics, University of Mumbai, Vidyanagari, Santacruz (East), Mumbai 400 098.

Name of the research co-guide

Dr. Nainesh Patel Assistant Professor, Department of Physics, University of Mumbai, Vidyanagari, Santacruz (East), Mumbai 400 098.

Place of the research

Department of Physics, University of Mumbai, Vidyanagari, Santacruz (East) Mumbai- 400098.

Number and date of registration

105 – 20/06/2013

Date of submission of synopsis

07/01/2016

Date of submission of thesis

Signature of the candidate

Signature of the guide

Signature of the co-guide

Signature of Head of Department

ii

STATEMENT BY CANDIDATE As required by the University Ordinances 770 and 771, I wish to state that the work embodied in this thesis entitled “Synthesis of Nanocatalysts and Their Applications to Hydrogen Production by Electrolysis of Water and Hydrolysis of Chemical Hydrides” forms my own contribution to the research work carried out under the guidance of Prof. D. C. Kothari and Dr. Nainesh Patel at the Department of Physics, University of Mumbai, Vidyanagari, Santacruz – E, Mumbai – 400098, India. This work has not been submitted for any other degree of this or any other university. Whenever references have been made to previous work of others, it has been clearly indicated as such and included in the Bibliography.

Signature of the Candidate Mr. Gupta Suraj Gulab Sonkali

Certified by:

Signature of the Co-Guide

Signature of the Guide

Dr. Nainesh Patel

Prof. D. C. Kothari

iii

Statement required under the ordinance 771 regarding the new fact This thesis describes the development of new kinds of catalyst that are economical, easy to synthesize and at the same time show high catalytic activity, comparable to that of noble metals for H2 production by hydrolysis of chemical hydrides and electrolysis of water. For hydrolysis of chemical hydrides (Sodium Borohydride and Ammonia Borane), the present study reported various new strategies to improve the catalytic performance of Co-B catalyst. These strategies include supporting Co-B on mesoporous silica substrates (to reduce the agglomeration), using mesoporous Co-B (to increase the surface area) and using nanoparticle assembled coatings (NPACs) (providing high surface area and good stability against aggregation). For water electrolysis, this thesis reports three transition metal borides (Co-B, Co-NiB and Co-Mo-B) as electrocatalytically active materials. Of these, Co-B and Co-Ni-B were found to be active for hydrogen evolution reaction (HER) alone while Co-Mo-B showed bifunctional nature, producing hydrogen and oxygen equally well. The new facts that have emerged from this work are as follows:  Mesoporous Co-B as well as Co-B supported on mesoporous silica, both serve as models for other heterogeneous catalysis reactions, to improve upon the catalytic performance.  Co3O4 nanoparticle assembled coatings (NPACs) serve as an ON/OFF switch for the hydrolysis reaction for easy recovery and reuse.  Co-B electrocatalyst was reported for the first time as a HER active material in neutral media.  Co-Ni-B electrocatalyst was reported to be active for HER in all pH media where the role of Ni was explained using XAS, XPS and DFT studies.  Co-Mo-B was reported as a promising bifunctional transition-metal boride

electrocatalyst in alkaline media.

Signature of the Candidate Mr. Gupta Suraj Gulab Sonkali Certified by:



Dr. Nainesh Patel

Prof. D. C. Kothari

iv

Regarding Authors’ contribution in joint work

The work reported in this thesis has been done under the guidance of Prof. D. C. Kothari and co-guidance of Dr. Nainesh Patel. The experimental work has been carried out by me and the results have been jointly interpreted by us. The results of part of this work have already been published in reputed international journals, some are under communication.

Signature of Candidate Mr. Gupta Suraj Gulab Sonkali

Certified by:



Dr. Nainesh Patel

Prof. D. C. Kothari

v

CERTIFICATE This is to certify that Mr. Gupta Suraj Gulab Sonkali has duly completed his thesis for the degree of Ph.D. of the University of Mumbai and his thesis entitled “Synthesis of Nanocatalysts and Their Applications to Hydrogen Production by Electrolysis of Water and Hydrolysis of Chemical Hydrides” is up to the standards both in respect to its content and literacy presentation for being referred to an examiner. We further certify that the entire work has been done by the candidate under our guidance and that no part of it has been submitted previously for any degree or diploma of any university.

Prof. D. C. Kothari (Guide) Department of Physics, University of Mumbai, Vidyanagari, Santacruz (E), Mumbai–400098, India.

Dr. Nainesh Patel (Co-Guide) Department of Physics, University of Mumbai, Vidyanagari, Santacruz (E), Mumbai–400098, India.

vi

To my beloved Family…

vii

Acknowledgement It gives me immense pleasure to acknowledge the contributions of all, without whom this endeavor would have not been possible. Though the degree will bear just my name, it is a collective effort of many individuals who stood alongside me and shared all the troubles equally. To begin with, I would like to express my heartfelt gratitude to both of my supervisors – Prof. Dushyant Kothari and Asst. Prof. Nainesh Patel. I would like to sincerely assert that none of this work would have been possible without the guidance and unconditional support of these two gentlemen. All the skills and knowledge that I garnered during my doctoral course were inherited from my supervisors. Their constant exertion helped me develop abilities to tackle problems with a positive and logical approach. They both empowered me to explore new realms in my work. One taught me to think of out of the box solutions and other taught me to execute those solutions in the most astonishing fashion. They played various roles in my life over the years even on personal levels and taught me valuable lessons. Prof. Kothari has been a father figure and have always been there in difficult times. He inculcated so many virtues due to his integrity, humility and passion to do good things for the society. Dr. Nainesh has been like an elder brother who always held my back in thick and thin times. I was indeed fortunate to come across a gem of a person like him. His dedication, sincerity and passion towards research infected the same spirit within me that helped me complete my doctoral work on a high. I thank both of them once again for adding that “extra” to my ordinary endeavour. I am deeply indebted to Prof. Anuradha Misra, who is also the head of the department, for her immense support and motivation. Her simple yet charming persona has been inspirational for me in many aspects. She endowed me with unconditional support and help at numerous occasions, despite her busy schedule. In the same breath, I am also grateful to Prof. Antonio Miotello from University of Trento, Italy for his constant support and readily helping my research work in all his capabilities. I am grateful to some of the close associates from my research group without whom this journey would not have been a memorable one. My association with Dr. Rohan Fernandes is not long but grew invariably stronger as I worked beside him. His impeccable research skills, complimented with his infectious smile, induced an viii

extremely positive ambience to work in. I thank him for his kind-heartedness and humility. Another companion who stood firmly beside me since I started my doctoral work is Asha Yadav. She has been an ideal associate and a close friend who selflessly supported me on research as well as personal levels, at all times. I owe a lot of gratitude to her for being so supportive and helpful in all my endeavors and also for her precise scientific suggestions when needed. It gives me great pleasure to thank Assc. Prof. Uday Patil for sharing with me his knowledge, thoughtful vision and humor over the years. I express my sincere thanks to Dr. Ranjana Varma, Mr. Chetan Gurada, Dr. Alpa Dashora and Dr. Kapil Bhatt for their timely suggestions and encouragement. I would also thank my other lab colleagues – Renuka, Abhijeet, Narayan, Manisha, Harshada, Riddhi, Rajesh, Nirmala, Sandeep and Ashutosh for their help and support. My special thanks to other colleagues from the department especially, Mohammad Ghadiyali and Darshan Habale for their support. I thank all the teaching and non-teaching staff of our department, especially Santosh Salunkhe for helping me with all administrative workloads and also, Nitin Baing and Kiran Londhe for doing anything and everything within their reach. On the personal front, many of my close friends have helped me to overcome the hiccups of life. My special word of thanks from the bottom of my heart to all my close friends, especially Digvijay Mishra, Aditya Pandey, Prashant Singh, Jamvant Vishwakarma, Ashutosh Dubey, Pankaj Mishra, Dheeraj Upadhyay, Purushottam Dubey, Dinesh Saroj, Amit Gupta, Mohit Upadhyay, Trupti Manjarekar and Ninad Waman. They provided the motivation whenever I failed and they started the celebrations even on my tiniest of achievements. I thank them all for being there for me whenever I needed them the most. I owe a lot to Aditya, Ashutosh and Ninad for their enlightening scientific discussions as well. Most importantly, none of this would have been a reality without the love, patience and support of my family, to whom I dedicate this dissertation. My parents went out of their depths, to ensure that nothing stops me from achieving my goals. No words will suffice to acknowledge their contribution and no deeds of mine would be enough to repay them. I am deeply indebted to my elder sister and brother for their moral support and love which made it so easier for me to complete my endeavor. I thank all those who were directly or indirectly associated with me during these years and helped me to successfully complete my doctorate.

ix

Abstract There is a serious concern related to increasing energy demands and greenhouse gas emissions. Hydrogen (H2) gas is recognized as a desirable clean fuel for PEM Fuel Cell that converts chemical energy of H2 into electrical energy with no environmentally harmful by-products. H2 can be efficiently produced either by water-splitting in an electrolyzer or by dissociation of chemical hydrides. The rate of H2 production, from either techniques, is a key factor in deciding their feasibility for particular applications. The rate of these reactions and thus that of H2 production can be increased multi-fold by employing suitable catalysts. To achieve this, nanocatalysts are highly preferred owing to their nano-scale architecture which lead to exceptional catalytic performance. Noble metal catalysts are found to be most suitable for such purposes. However, the use of noble metals needs to be minimized owing to their high cost and scarcity. Lowcost transition metal (Co, Ni) borides are considered as good candidates for accelerating the H2 production rate. Different strategies were employed to improve the activity of Co-B catalyst for hydrolysis of chemical hydrides. Co-B supported on different types of mesoporous (MSP) silica substrates showed enhanced H2 production rates for hydrolysis of Ammonia Borane. Using MSP silica substrates, provide a unique opportunity to tune the size of the pores and thus that of the nanoparticles (NPs) enclosed within them. The supported NPs are prevented from agglomeration, thereby improving the activity. Using two different surfactants, two types of MSP Co-B catalysts were developed with significantly higher effective surface area. The mesoporous assembly facilitates easy passage of reactants to the active sites, resulting in enhanced activity for hydrolysis of sodium borohydride (SBH). Co3O4 nanoparticle assembled coatings (NPACs), deposited by pulsed laser deposition (PLD) were also found to serve as efficient catalyst for hydrolysis of SBH. Different parameters were varied during deposition of these films, to obtain the most effective catalyst composition in the films. Co-B was also found to be an excellent electrocatalyst for hydrogen evolution reaction (HER) active in wide pH range (4 - 9). A vast improvement in activity and stability of Co-B electrocatalyst was obtained after introducing other transition metals, specifically Ni and Mo in Co-B. The role played by each element in enhancing the HER rate was studied in details by using numerous characterization tools such as XPS, XAS, BET, TEM, SEM, XRD and theoretical calculations. Co-Mo-B was found to be equally active for hydrogen and oxygen evolutions in alkaline media. x

Contents 1 Introduction

1.1 Energy . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.1 Energy Perspectives . . . . . . . . . . . . . 1.1.2 Environmental impacts . . . . . . . . . . . . 1.2 Hydrogen . . . . . . . . . . . . . . . . . . . . . . . 1.2.1 Hydrogen - Fuel of the future . . . . . . . . 1.2.2 Clean production of Hydrogen . . . . . . . . 1.2.3 Fuel cell . . . . . . . . . . . . . . . . . . . . 1.2.4 Storage and Distribution of hydrogen . . . . 1.3 H2 production from chemical hydrides . . . . . . . . 1.3.1 Sodium borohydride (NaBH4) . . . . . . . . 1.3.2 Ammonia Borane (NH3BH3) . . . . . . . . . 1.4 H2 production from electrochemical water splitting 1.4.1 Technology of water electrolysis . . . . . . . 1.4.2 Hydrogen Evolution Reaction (HER) . . . . 1.4.3 Electrocatalysts for HER . . . . . . . . . . . 1.4.4 Oxygen Evolution Reaction (OER) . . . . . 1.5 Nanocatalysis . . . . . . . . . . . . . . . . . . . . . 1.6 Thesis Overview . . . . . . . . . . . . . . . . . . . .

2 Experimental Techniques

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

2.1 Catalyst preparation techniques . . . . . . . . . . . . . . . . . . 2.1.1 Preparation of Powder catalyst . . . . . . . . . . . . . . 2.1.2 Preparation of mesoporous nanoparticles . . . . . . . . . 2.1.2.1 MCM-41 type mesoporous silica . . . . . . . . . 2.1.2.2 FSM-16 type mesoporous silica . . . . . . . . . 2.1.2.3 SBA-15 type mesoporous silica . . . . . . . . . 2.1.2.4 MSP silica supported Co-B catalyst . . . . . . 2.1.2.5 Mesoporous Co-B catalyst . . . . . . . . . . . . 2.1.3 Thin lm catalyst using Pulsed Laser Deposition (PLD)

. . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . .

1

1 1 3 4 4 5 6 7 7 8 9 10 10 13 14 16 17 18 20

20 20 21 21 21 22 22 22 23

CONTENTS 2.2 Hydrogen Measurement Set up . . . . . . . . . . . . . . . . . . . . 2.2.1 Experimental setup . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Gas Volumetric Method . . . . . . . . . . . . . . . . . . . . 2.3 Electrocatalytic measurement methods . . . . . . . . . . . . . . . . 2.3.1 Preparation of catalyst modied glassy carbon (GC) electrodes 2.3.2 Preparation of electrolyte solutions . . . . . . . . . . . . . . 2.3.3 Electrochemical measurements . . . . . . . . . . . . . . . . .

24 24 24 26 26 26 26

3 Mesoporous Silica supported Co-B for hydrolysis of Ammonia Borane 28

3.1 3.2 3.3 3.4 3.5

Introduction . . . . . . . . . . . . . . . . . . . . . . . . Characterization of mesoporous structures . . . . . . . Mechanism of Co-B loading on dierent pore structures Hydrolysis of Ammonia Borane . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

4 Mesoporous Co-B for hydrolysis of sodium borohydride

4.1 4.2 4.3 4.4 4.5

Introduction . . . . . . . . . . . . . . . . Characterization of MSP Co-B catalysts Mechanism of mesopore formation . . . Hydrolysis of Sodium Borohydride . . . . Conclusion . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

28 30 36 38 44

45

45 46 50 50 55

5 Co3 O4 Nanoparticles Assembled Coatings for hydrolysis of SBH 57

5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Characterization of nanoparticle assembled coatings (NPACs) Hydrolysis of Sodium Borohydride . . . . . . . . . . . . . . . . Eect of variation in O2 pressure . . . . . . . . . . . . . . . . Eect of variation in laser uence . . . . . . . . . . . . . . . . Eect of variation in substrate temperature . . . . . . . . . . Recycling test for NPACs . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

6 Co-B electrocatalyst for hydrogen evolution reaction

6.1 6.2 6.3 6.4 6.5 6.6

Introduction . . . . . . . . . . . . Characterization of Co-B catalyst Electrochemical measurements . . Activity in dierent pH media . . Stability and Reusability . . . . . Conclusion . . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . . . .

. . . . . .

. . . . . . . .

. . . . . .

. . . . . . . .

. . . . . .

57 59 63 65 69 70 72 73

74

74 75 78 80 83 84

CONTENTS 7 Co-Ni-B electrocatalyst for hydrogen evolution reaction

7.1 7.2 7.3 7.4 7.5 7.6 7.7

Introduction . . . . . . . . . . . . . . Characterization of Co-Ni-B catalyst Electrochemical measurements . . . . Activity in dierent pH media . . . . Stability and Reusability . . . . . . . Eect of heat treatment . . . . . . . Conclusion . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

85

. 85 . 86 . 92 . 96 . 99 . 100 . 101

8 Co-Mo-B electrocatalyst for Hydrogen and Oxygen Evolution Reactions 102

8.1 Introduction . . . . . . . . . . . . . . . . . . 8.2 Characterization of Co-Mo-B catalyst . . . . 8.3 Electrochemical measurements . . . . . . . . 8.3.1 Hydrogen Evolution Reaction . . . . 8.3.2 Electrochemical Surface Area (ESA) 8.3.3 Oxygen Evolution Reaction . . . . . 8.4 Overall Water splitting using Co-Mo-B . . . 8.5 Conclusion . . . . . . . . . . . . . . . . . . .

9 Conclusions and Future Outlook

9.1 Hydrolysis of chemical hydrides 9.1.1 Conclusions . . . . . . . 9.1.2 Future prospects . . . . 9.2 Electrocatalytic water splitting 9.2.1 Conclusions . . . . . . . 9.2.2 Future prospects . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . . . .

. . . . . .

. . . . . . . .

. . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . .

. . . . . .

. . . . . . . .

. . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. 102 . 103 . 109 . 109 . 113 . 114 . 117 . 119 120

. 121 . 121 . 122 . 122 . 122 . 124

A Co-P-B lms for hydrogen evolution from alkaline water electrolysis 134 B Algorithm for data acqusition software to measure produced H2 138 C Calculations for Turn Over Frequency (TOF)

139

D Comparison of HER performance from literature

141

Bibliography

143

Chapter 1 Introduction 1.1 Energy 1.1.1 Energy Perspectives Energy has been one of the primary needs of human beings since ages. We need energy for almost everything that we do in our day-to-day lives. Whether we notice it or not, we are dependent on abundant and uninterrupted supply of energy. The social and economic development of a country is also measured based on its energy consumption. Increase in world population, development in technology and increased standard of living demands increase in the requirement of energy. Energy insecurity and rising prices of conventional energy sources are also major threats to economic and political stability. These factors lead to world population transition, migration, hunger, environmental problem, deteriorating health, diseases, terrorism and even wars[1]. As seen in Figure 1.1, the world marketed energy consumption would grow by 63% from 2004 to 2030, with the world energy use rising from 447 Btu (quardrillion British thermal units) in 2004 to 702 Btu in 2035, if the current laws and policies remain unchanged [2]. In a developing nation like India, this concern grows higher as we are looking to enter the league of developed nations. Figure 1.2 shows the graph of Human development index (HDI) and its relation to per capita electricity consumption (kWh/person/year) by all nations. On a closer look, we see that all the developed nations with HDI more than 0.7 have very high per capita electricity consumptions (> 1000 kWh/person/year) while India and other developing nations with HDI of 0.6 and less, consume very small amount of electricity per capita. If these developing nations are to turn into developed ones, they will need to increase their per capita electricity consumption many folds. This implies excessive utilization of available energy resources. 1

Chapter 1: Introduction

Figure 1.1: World marketed energy consumption in Btu. The world energy systems are largely based on combustion of fossil fuels, i.e. petroleum, natural gas and coal. Coal is the second most used source of energy. It lls much of the growing energy demands of those developing countries, where energy-intensive industrial production is growing rapidly and large coal reserves exist with limited reserves of other energy sources. The world coal reserves are estimated at 909 billion tons, reecting a current reserves-to-production ratio of 129 years [2]. Crude Oil is used as primary source of energy and thus is constantly in high demand. As a dense energy source, it powers majority of vehicles and thus is one of the world's most important commodities. According to Oil & Gas Journal (January 2010), world oil reserves were estimated to be 1,354 billion barrels [3] while the worldwide reserves-to-production ratio is estimated at 45 years [4]. The third major source of energy is Natural gas that is mostly used for household purposes. LPG (Liqueed Petroleum Gas) is a mixture of hydrocarbon gases that primarily contains propane, and mostly used for cooking and heating. CNG (Compressed Natural Gas) mostly contains Methane, a substitute for LPG. As it is cleaner, it is also used as fuel in CNG-powered vehicles. The natural gas reserves are estimated at 6,609 trillion cubic feet, giving us the reserves-to-production ratio of 60 years. These forms of energy are rapidly depleting, i.e., in the next 45 and 60 years, we will run out of both oil and natural gas respectively, while coal will last only for the next 129 years. In this scenario, nding an alternate source of energy becomes an urgent need. If countries do not increasingly reduce their reliance upon fossil fuels, there will be severe economic tragedies in the near future. 2


Figure 1.2:

Relation between human development index (HDI) and per capita electricity

consumption.

1.1.2 Environmental impacts The age of petroleum led to the ascendance of oil producing countries. Today we realize, however, that we have been left with a legacy of greenhouse gases, which are the by-products of 250 years of power generation using fossil fuels. Burning of gasoline and coal releases hazardous gases like CO2, CO, SO and SO2, N O2 into the atmosphere giving rise to serious environmental and health issues. Smog is formed when N O2 reacts with volatile organic compounds in the atmosphere increasing respiratory problems in humans. The emission of CO2 has increased dramatically in the last century. Infact between 2007 and 2008, CO2 emissions from the combustion of coal increased by 3% [5]. In Figure 1.3, we see the rise in global mean CO2 concentration over the last century and the variation in global average temperature. Global climate change is one of the most severe global environmental consequences. CO2 emissions are classied as one of the main driving forces behind global warming today. The total temperature increase from 1980 to 2005 is close to 1 ºC. The data shows that, globally, the last decade has been the warmest ever recorded, which co-relates to the drastic increase in CO2 concentrations.

3


Figure 1.3: Variation in global temperature and CO

2

concentration over the last century.

It becomes imperative to meet the global energy needs by such means that do not contribute to the greenhouse eect. Fossil fuels have had a crucial role in modern society, but since they are non-renewable and dangerous, we have to reduce our dependence on them and explore alternative energy sources. These new alternative sources of energy must have two desirable characteristics: 1. They should be renewable forms of energy so that future generations do not inherit an energy vacuum. 2. They must not produce greenhouse or pollutant gases so that current and future generations do not inherit an uninhabitable planet. Recognizing these problems, countries around the world are relying more and more on renewable energies, not only for economic but also for the environmental benets.

1.2 Hydrogen 1.2.1 Hydrogen - Fuel of the future A lot of alternatives t the list of renewable energy resources - wind energy, solar energy, nuclear energy, etc. In addition to these, use of hydrogen as an energy carrier has attracted major attention from researchers all round the globe. In fact, the next great energy epoch is just around the corner as the world is gearing up to enter the third energy epoch - The Hydrogen Age. 4


The production of hydrogen can reduce our dependence on imported oil and natural gas. There is a growing awareness that hydrogen energy may help avoid the next potential energy crisis, which will be driven by an increasing dependence on the existing energy resources. An important qualier is the fact that hydrogen generated using renewable energy is environmentally friendly. Therefore, hydrogen generated using renewable energy will soon join solar electricity to form the foundation of a sustainable modern energy system. On Earth, hydrogen is one of the most abundant elements, but less than 1% is present in molecular gas form. Hydrogen can be produced in many ways, i.e. from hydrocarbons, electrolysis of water, and processes driven by sunlight to name a few. Hydrogen is environmently friendly and has the highest heating value per mass of all the chemicals [6]. Unlike electricity, hydrogen can be stored for relatively longer periods of time. For the past century, the use of gasoline and diesel fuels to power our vehicles has been extremely wasteful but they are still perhaps the best forms of energy in terms of energy density for a given material and ease of refueling and storage. The current designs of the internal combustion engine using gasoline (petrol) as a fuel are only 18 - 20% ecient in producing usable power [7], the rest of the energy produced is wasted in the form of heat-loss, friction, etc. Hydrogen gas is considered thermally more ecient than gasoline primarily because it burns better in air and permits the use of higher compression ratio. It is now widely accepted that the rst and the most far-reaching eect of hydrogen technologies on civilization will be in transportation. Hydrogen already has been used as a fuel for spacecraft but soon it will be the fuel for aircraft, buses, trains and most visibly - cars. This is likely to be accompanied by the emergence of new ancillary industries that involve hydrogen production, storage, transportation, and utilization. Most projections suggest that these developments will take place over the next 3-10 years. Already, there are prototype hydrogenpowered cars and buses on the road. It is only a matter of time before gasoline stations begin to adapt their infrastructure to hydrogen distribution. It is possible that, within ∼15 years, many if not most gasoline stations will be converted to hydrogen stations.

1.2.2 Clean production of Hydrogen If the produced H2 comes from natural gas, coal, petroleum or biomass, it will produce carbon dioxide (CO2) as a by-product, which is a greenhouse gas. This CO2 must be utilized, eliminated, or sequestrated somehow in order to prevent its escape into the atmosphere. If this is not done, then the environmental gain obtained by the use of hydrogen is greatly reduced. Secondly, one must burn the 5


hydrogen. Despite the widespread perception that hydrogen is a clean-burning fuel, it is not quite correct. This is true when hydrogen is mixed with pure oxygen during combustion. However, essentially all combustion processes use air, which includes ∼20% oxygen and ∼80% nitrogen. Therefore, while the by-product of the reaction of H2 + O2 H2O is harmless water, there is also the under-appreciated reaction N2 + O2 NOx, which yields a greenhouse gas as the reaction product. However, this pollution is substantially lower than that produced by gasoline combustion engines. Renewable hydrogen can be produced in several ways: water electrolysis, biomass conversion, solar conversion and from chemical hydrides. Of these listed methods, biomass-to-hydrogen conversion is complex because of the technical details of the conversion processes [8], whereas the eciency of solar based processes are extremely low for practical uses [9]. Water and chemical hydrides, on the other hand, are preferred sources to produce H2 at room temperature without any emission of harmful gases with high conversion eciency. H2 produced by these two methods are highly pure, which can be stored for later use or transformed directly into electricity, using fuel cells.

1.2.3 Fuel cell The fuel cell oers great promise as a pollution free power source for various applications ranging from mobile phones to automobile [10, 11]. Hydrogen fuel cells will more than likely eventually replace the internal combustion engine entirely. As they do not have any moving parts, they are also very quiet and reliable. Lithium ion batteries that power laptops, cell phones and other electronic devices, deteriorate with time and use; whereas a fuel cell potentially has an unlimited life time, for as long as the fuel is supplied, the fuel cell will continue to generate power. The power generation process in a fuel cell is analogous to that of a battery, i.e. converting chemical energy to electrical energy. A fuel cell combines hydrogen and oxygen to produce electricity, with water and heat as its by-product. Thus, it is a so-called zero emission engine. There are various types of fuel cells available, each with its own advantages, limitations and potential applications. They are dierentiated by the kind of electrolyte, chemical reaction, fuel, catalyst and temperature required for their functioning. Polymer Electrolyte Membrane Fuel Cell (PEMFC), Direct Methanol Fuel Cell (DMFC), Alkaline Fuel Cell (AFC), Phosphorous Acid Fuel Cell (PAFC), Solid oxide Fuel Cell (SOFC) to name a few. A typical fuel cell can produce between 0.5 and 0.9 V of DC electricity. To produce the required amount of voltage, single fuel cells are stacked together to form a fuel cell-stack. 6


1.2.4 Storage and Distribution of hydrogen The hydrogen produced by water electrolysis or from chemical hydrides is pure, clean and free from carbon and sulphur impurities. Even after production of pure hydrogen, one of the main obstacles to its widespread use as an energy carrier, are related to the unsolved problems concerning its storage and distribution. For successful application of hydrogen in stationary and automobile applications, hydrogen storage is important. Secondly, the volumetric and gravimetric density is also a critical issue [6]. The most studied hydrogen storage systems for storage and transport are high pressure tanks, liquid hydrogen [12], hydrogen stored in metal hydrides and chemical hydrides. A classical high pressure tank is made of inexpensive steel and can hold up to 300 bar, and are normally lled up to 200 bar. Thus, storing 4 Kg of hydrogen, one requires 225 liters. Carbon-ber reinforced composite materials are also being considered, that can hold up to 600 bar. These type of tanks need special inner coating to avoid high pressure hydrogen reaction. Other way to store hydrogen is in the form of liquid. The mass per volume of hydrogen can be increased by condensing it to liquid. Liquid hydrogen (LH2) storage is usually preferred as it has a very impressive safety record. The hydrogen is typically liqueed at the production site in large quantities (10-30 tons per day) and then trucked crosscountry in 11,000 gal LH2 tankers with no boil-o losses. Metal hydrides are also considered for hydrogen storage, but have kinetic problems and require high temperatures. Chemical hydrides like Sodium borohydride (NaBH4) [13, 14], Ammonia Borane (NH3BH3) [15, 16], Lithium borohydride (LiBH4) [17], Potassium borohydride (KBH4), etc. with high hydrogen storage capabilities have drawn a lot of attention in recent years as a hydrogen carrier. Chemical hydrides are considered as a breakthrough for its safe hydrogen storage capabilities. NaBH4 and NH3BH3 show high hydrogen storage of 10.6 wt% and 18 wt% respectively [14, 15], making them attractive candidates to supply pure hydrogen for fuel cell to be used for portable applications.

1.3 H2 production from chemical hydrides As mentioned earlier, chemical hydrides appear as capable and promising solution to overcome the issues related to hydrogen safety and handling and to be competitive with respect to conventional fuel. Ammonia borane (NH3BH3) and Sodium borohydride (NaBH4) with high gravimetric and volumetric hydrogen storage ability are the most prospective candidates to supply pure hydrogen for portable and on-board applications at room temperature [18, 19]. 7


1.3.1 Sodium borohydride (NaBH4) NaBH4 is a crystalline white thermally stable substance. It is a powerful reducing agent and is extensively used in wastewater processing and paper bleaching [20]. NaBH4 in presence of appropriate catalysts, hydrolyze to produce clean hydrogen gas. Aqueous NaBH4 seems to be an ideal hydrogen source for the following reasons [21]:

NaBH solution is non-ammable and is stable in air for months. It has a H storage capability of 10.8 wt %. H produced is clean, without any impurities. H generation can be controlled in the presence of selected catalysts. The reaction product, Sodium borate (NaBO ), obtained after de-hydrogenation 4

2

2

2

2

of NaBH4, is environmentally clean and can be recycled to generate the reactant.

H

is generated by water-based hydrolysis reaction of NaBH4, during which half of the hydrogen is produced from the water solvent. 2

In the 1950s, Schlesinger et. al. [22] rst observed that NaBH4 in the presence of appropriate catalysts, at room temperature, could be hydrolysed to H2 gas and Sodium borate, NaBO2 (see Eq. 1.1). (1.1) NaBH4 self hydrolyses, even in the absence of a catalyst, in a solution with pH < 9. The rate of decomposition of the aqueous borohydride solution is expressed by its half-life as a function of pH and temperature [23]. N aBH4 + 2H2 O → 4H2 + N aBO2 + Heat(300kJ)

(1.2) where, t1/2 is indicated in minutes and T in Kelvin. Sodium hydroxide (NaOH) is added to NaBH4 solution to increase the shelf-life of the solution, by making the solution strongly alkaline [24]. Schlesinger et. al. [22] studied eects of many acids for the hydrolysis of NaBH4 (like inorganic protic acids, oxalic acid, phosphorus pentoxide, aluminium chloride). These acids could hydrolyse the NaBH4 solution, but the reaction could not be controlled. In 1985, Kaufman et. al. studied the eect of some metals (Cu, Ni, and Co) and their salts on the overall hydrogen production [25]. Suda et al. [21] examined the eect of uorinated metal hydrides catalyst on the hydrolysis of alkaline NaBH4. Kojima et al. [26] studied the eect log(t1/2 ) = pH − (0.034T − 1.92)

8


of Pt metal coated on metal oxide. Amendola et al. [20] reported the application of Ru-catalyzed hydrolysis of aqueous BH4− solution as hydrogen generator for PEMFC. Precious metals like Pt [27] and Pd supported on carbon [28], Pt-Ru supported on metal oxide [29], Ru on polymer [30], Ru nanoclusters [31] have been extensively studied as catalysts and have shown excellent catalytic activities. These kinds of catalysts are quite expensive and cannot be considered due to their high cost. Some metals, such as Raney Ni and Co, and even nickel and cobalt borides [32, 33], are also used to accelerate the hydrolysis reaction of NaBH4.

1.3.2 Ammonia Borane (NH3BH3) Recently, Ammonia borane (NH3BH3, AB) has attracted increasing attention as an ecient hydrogen storage material because of its high hydrogen content. AB can produce hydrogen by either hydrolysis or thermolysis [34] and the end product, which is non-toxic and environmentally safe, can be recycled to regenerate AB [58]. While thermolysis of AB requires moderate temperature, the catalytic hydrolysis can generate sustained amount of hydrogen even at room temperature (see Eq. 1.3). (1.3) Along with NaBH4, AB is also considered as a promising candidate to supply pure hydrogen for portable and on-board applications at room temperature [18, 19] because of the following properties: N H3 BH3 + 2H2 O → N H4+ + BO2− + 3H2

AB has H storage capacity of 19.6 wt%. It has low molecular weight of 30.7 g/mol. High gravimetric and volumetric hydrogen storage ability. AB is highly stable and soluble in water and does not need alkaline solution 2

like in the case of NaBH4.

The end product, which is non toxic and environmentally safe, can be recycled.

The hydrolysis reaction rate can be eectively increased by using several inorganic and organic acids but the reaction usually becomes uncontrollable. On the other hand, solid state catalysts such as precious and transition metals and their salts are found to be very ecient in accelerating the hydrolysis reaction in a controllable manner. Noble catalysts like Pt, Rh and Ru supported on Al2O3 [35], K2Pt6Pt [35] and nanoclusters of Pd(0) [36], Ru(0) [37] and Rh(0) [38] have been utilized 9


in the past to enhance the hydrogen production rates. However, these catalysts do not seem viable for industrial applications considering their cost and availability. Transition metals such as Co supported on Al2O3, SiO2 and C [39]; Co(0), Ni(0) and Fe(0) nanoclusters [40, 41], Ni based alloy [42], and Ni-SiO2 nanospheres [43] are generally used to accelerate the hydrolysis reaction of NH3BH3.

1.4 H2 production from electrochemical water splitting Electrolysis is a process in which a direct electric current is passed between two electrodes through an ionic substance that is either molten or dissolved in a suitable solvent (electrolyte), to perform a non-spontaneous chemical reaction to separate the reaction products. The history of water electrolysis dates back to as early as the rst industrial revolution, when, in the year 1800, Nicholson and Carlise accidentally discovered the ability to split water using a voltaic pile. The laws of electrolysis determined by Faraday were reported in 1834. Few years after publishing their work, electrolysis had been used by Humphry Davy to isolate sodium, potassium, calcium, strontium, barium, magnesium, and lithium. By the beginning of 20th century, more than 400 industrial water electrolyzers were in operation, and in 1939 the rst large-scale water electrolysis plant with a capacity of 10,000 m3 H2 h−1 went into operation. 20 years later, in 1949 the rst pressurized industrial electrolyzer by Zdansky/Lonza was built [44]. Currently, the development of a new water-electrolysis technology based on the use of polymer-electrolyte membranes and the optimization and reconstruction of the old (classical) alkaline water electrolyzers that have been in use for long time, is the subject of intensive research across all nations.

1.4.1 Technology of water electrolysis The device used to perform electrolysis is called an electrolyzer. An electrolyzer is made by the interconnection of several elementary electrolysis cells. Electrolysis cells are characterized by their electrolyte type. There are three types of electrolysis cells: alkaline, proton exchange membrane (PEM) and solid oxide. Alkaline electrolysis uses a liquid electrolyte consisting of highly concentrated potassium or sodium hydroxide. PEM electrolysis is based on the use of a solid conducting polymer that conducts ions when hydrated with water. Solid oxide electrolyte cells split water in the form of steam into pure H2 and O2 at high temperatures. Apart from these, there are also microbial electrolysis cells, that generate hydrogen or methane from organic materials by applying an electric current, operating 10


in neutral pH conditions. We will see in detail the working mechanism of two well developed electrolysis cells, operating at room temperature - alkaline and PEM electrolyzers. Alkaline electrolyzer:

Industrial plants prefer using alkaline media for electrolysis, as corrosion is more easily controlled and cheaper construction materials can be used than in acidic electrolysis technology [44]. Aqueous potassium hydroxide (KOH) has traditionally been used as an electrolyte in alkaline water electrolyzers, due to its high ionic conductivity; a 20-30 wt% solution has been used because of the optimal (maximum) ionic conductivity and remarkable corrosion resistance of stainless steel in this concentration range. Typical operating temperatures and pressures of such electrolyzers are 70-100 °C and 1-30 bars, respectively [45]. The principle of operation of alkaline water electrolysis is shown schematically in Figure 1.4. The major disadvantages of the alkaline process are high energy consumption, low specic production, low eciency (up to 65%), voluminous systems and safety issues related to the use of caustic electrolytes [46]. The following reactions occur at each electrode in an alkaline system: Cathode : 2H2 O + 2e− → H2 + 2OH − Anode : 2OH − → 1/2O2 + H2 O + 2e−

T otal reaction : H2 O → H2 + 1/2O2

Figure 1.4: Schematic of an alkaline water electrolyzer. 11


PEM electrolyzer:

In 1960s, General Electric developed [47] the rst water electrolyzer based on the concept of a solid polymer electrolyte. This concept was idealized by Grubb [48], who used a solid sulfonated polystyrene membrane as an electrolyte. This concept was also referred to as proton exchange membrane or polymer electrolyte membrane (both with the acronym PEM) water electrolysis. The polymer electrolyte membrane (Naon, Fumapem) is responsible for providing high proton conductivity, low gas crossover, compact system design and high pressure operation. The low membrane thickness (∼20-300 mm) is one of the many advantages of the solid polymer electrolyte. The principle of operation of PEM electrolyzer is shown schematically in Figure 1.5.

Figure 1.5: Schematic of a PEM water electrolyzer. The following reactions occur at each electrode in an acidic system: Cathode : 2H + + 2e− → H2 Anode : H2 O → 2H + + 1/2O2 + 2e−

T otal reaction : H2 O → H2 + 1/2O2

PEM electrolyzers can operate at very high current densities, capable of achieving values above 2 A/cm2. They face problems related to higher operational pressures, such as cross-permeation phenomenon which increases with pressure [49]. Also, the corrosive acidic regime provided by the proton exchange membrane requires the use of distinct materials. These materials must not only resist the harsh 12


corrosive low pH condition (pH ∼ 2), but also sustain the high applied over voltage (∼2 V), especially at high current densities. This demands the use of scarce, expensive materials and components such as noble catalysts (Pt, Ir, Ru, etc.). Replacement of these catalysts by inexpensive, non-toxic, earth-abundant elements, is the subject matter of a lot of on-going research.

1.4.2 Hydrogen Evolution Reaction (HER) Understanding the principles of the HER mechanisms and kinetics is crucial for the development of new electrocatalysts. This reaction is also of great importance in fundamental and technological electrochemistry and has been widely studied using a broad range of solution conditions and electrode materials. During water electrolysis, the HER on a metallic electrode (M), in acidic media, proceeds according to the following three-reaction-step mechanism [50]: M + H + + e− ↔ M − Hads

(V olmer Reaction)

M − Hads + H + + e− ↔ H2 + M

M − Hads + M − Hads ↔ H2 + 2M

(Heyrovsky Reaction)

(T af el Reaction)

(1.4) (1.5) (1.6)

The rst reaction (Volmer reaction) is electroreduction of the proton with the formation of hydrogen adsorbed on the electrode surface, M-Hads. This is followed either by the electrochemical (Heyrovsky reaction) or chemical (Tafel reaction) desorption of hydrogen. The adsorbed hydrogen atom (M-Hads) plays a signicant role in the HER mechanism, governing both its thermodynamics and kinetics. Both mechanisms, Volmer-Heyrovsky and Volmer-Tafel, require the formation and then cleavage of the M-Hads bond. Hence, the reaction rate for the HER is determined by the strength of the hydrogen-metal adsorption bond. Consequently, the maximum rate of hydrogen evolution will occur at intermediate values of MHads bond strength, resulting in a behavior characterized by the "volcano curve", shown in Figure 1.6 [51]. Unfortunately, it is clear from the gure that only noble metals (e.g. Pt, Rh, Re, Pd, Ir) express high HER activity, while less noble pure metals show much lower activity [51]. This is due to the fact that Pt- and Ir-group of metals have almost lled d-orbital, thus making the M-Hads bond of an intermediate strength. Therefore, the basis for the design of a cheap non-noble HER catalyst is designing a material that would express d-electronic density similar to Pt-group metals and in that way at least approach the electrochemical activity in the HER similar to 13


the noble metals.

Figure 1.6: Volcano curve for HER.

1.4.3 Electrocatalysts for HER Three characteristics play a crucial role in selecting catalytically active materials for the HER: (1) intrinsic electrocatalytic activity of the material, (2) active surface area per unit volume and (3) catalyst stability [46]. In water electrolysis for hydrogen production, the cathode is required to have low hydrogen overpotential in order to minimize energy consumption. Over the last decade, a lot of active research has been focused on structural and compositional engineering of electrocatalysts to preferentially expose active sites and/or to promote site-specic activities. Nanostructured HER electrocatalysts have been made in a variety of forms and dimensions, with greatly improved electrochemical performance compared to their bulk counterparts. Many eorts have been made to explain the characteristics of hydrogen overpotential of individual metals using various physical and electronic parameters such as the atomic number [52], the work function [53], the bond strength of metals [54], the heat of adsorption of hydrogen on metals [51] and the electro negativity [55]. It has been concluded from many studies that the intrinsic catalytic activity for the HER can be related to the electronic structure of metals. As seen from the Volcano plot, platinum group metals (PGMs, including Pt, Ru, Rh, Ir and Pd) are the best known HER electrocatalysts located close to the apex. Amongst these, Pt has been the most popular choice, and frequently used as a benchmark for other HER electrocatalysts. However, the largescale applications of these catalysts are inhibited due to their high cost and low abundance. These issues were tackled using two dierent strategies. The rst involves the use 14


of microstructured or nanostructured electrocatalysts with large surface to volume ratio. As, electrocatalysis is a surface process, ecient utilization of surface catalytic atoms can signicantly reduce the amount of catalyst loading. To limit the amount of catalyst loading, Chen et al. [56] proposed to use monolayer Pt supported on low-cost materials as alternatives to PGM electrocatalysts. WC was identied as one of the support materials, where loading Pt monolayer on bulk WC leads to comparable activity to that of bulk Pt, with a signicant reduction in Pt loading and cost. This method was further extended to other metal overlayers (Pd and Au) and supports (Mo2C) to obtain electrocatalysts with excellent HER activity and stability in acidic environments. The second strategy is to make alloys of PGM with other metals so as to increase their site-specic activity, thereby allowing for lower catalyst loading to be used [57]. Among a number of promising candidates identied, the surface alloy formed from Pt and Bi was found to be the most eective. DFT calculations suggested that ΔGH for Bi-Pt [57] was ∼0.04 eV closer to the thermoneutral voltage compared to Pt, indicating that its HER activity must be comparable to, or even better than pure Pt. The strategies dicussed above seem relevant to minimize the precious metal loading, however, the ultimate way out is to replace them completely. Industrially, Ni-based electrodes are commonly employed as cathodes in alkaline electrolyzers for the production of hydrogen owing to their low cost and corrosion resistance at high pH values [58]. However, Ni suers from insucient electrocatalytic activity and more importantly, progressive deactivation upon continuous alkaline electrolysis due to the formation of reversible nickel hydride species [59]. On the other hand, Cobalt has emerged as an interesting non-noble metal for its catalytic power towards water splitting. A lot of past eorts have been devoted to making homogenous molecular catalysts based on Co complexes for HER and OER [60, 61]. To be used on a practical scale, these molecular catalysts need to be grafted onto electrode materials operating with fully aqueous electrolytes, which is still a challenging task. Moreover, these catalysts usually suer from issues like large overpotential and/or low stability. Consequently, the recent research interest has been focused on developing Co-based heterogeneous catalysts as possible replacements for PGM. In a recent work, Chen, Tour and colleagues [62] showed that individual cobalt atoms disperse on the graphene substrate, instead of forming discrete nanoparticles, through coordination to the nitrogen atoms in the graphene, depicting excellent electrochemical performance. The intrinsic activity (TOF) per cobalt atom was found to be similar or even higher than that of most non-noble metal catalysts, the overpotentials were very low and the catalyst was durable in both acidic and basic media. Not only in isolated form, but Co compounds such as 15


phosphides, sulphides, chalcogenides and borides [63, 64, 65], have also turned out to be excellent HER catalysts over the past 5 years. These electrocatalysts are not only active in acidic or basic media but have shown superior performance in neutral media as well. Asefa et.al. [66] reported Cobalt embedded in nitrogen-rich CNTs (Co-NRCNTs) with high HER activity and great electrochemical stability under the whole pH range. Co@NC was reported as a stable electrocatalyst for HER in a wide pH range, along with excellent performance for OER in alkaline medium [67]. A recent work by the Cui group investigated the HER performance of rst-row transition metal dichalcogenides (ME2 , M = Fe, Co, Ni; E = S, Se) in acidic media [68]. In 2013, Sun et al. [63] reported electrodeposited cobaltsulde (Co-S) lm active in neutral aqueous solutions and was also resistant to seawater. Following this report, other forms of CoS2 were reported such as, CoS2 nanowires (NWs) [69], CoS2/RGO-CNT [70] and so on. Ni0.33Co0.67S2 NW [69] was also reported as highly active for HER in both neutral and alkaline media. In 2014, Popczun et. al. conrmed CoP to be highly ecient and acid-stable HER electrocatalyst [71]. Huang et. al. described the HER activity of Co2P nanorods in acidic and basic solutions [72]. As a exible integrated 3D cathode, CoP/CC [73] showed excellent HER activity over the entire pH range. As far as transition metal borides are concerned, Lasia and Los [74] reported amorphous Ni2B electrocatalyst for HER in alkaline medium. Subsequently, there were more reports on electrodeposited Ni2B [75] and doped Ni2B [76] catalysts for alkaline water electrolysis. Since then, the interest in metal boride electrocatalysts almost vanished until Vrubel and Hu reported MoB [77] as HER active material in both acidic and basic conditions. Following these work, a part of this thesis is dedicated to present Co-B, Co-Ni-B and Co-Mo-B as HER active materials in various pH media.

1.4.4 Oxygen Evolution Reaction (OER) For complete water splitting, two half cell reactions take place simultaneously: hydrogen evolution reaction (HER) and oxygen evolution reaction (OER). To split water, a voltage of at least 1.23 V must be applied to provide the thermodynamic driving force. However, present-day industrial electrolyzers operate at voltages of 1.8 2 V, a whopping 40% more voltage than the ideal case (1.23 V). A large fraction of this energy is spent to overcome the hurdle of oxygen evolution reaction (OER). Even after a decade long research on OER materials, the quest for a catalyst that can slash down the overpotential considerably, is still in pursuit. OER is a rather sluggish process involving multi-electron transfer steps. On transition metal oxides, the most widely accepted mechanism for oxygen evolution 16


was proposed by Bockris [78]. As per this mechanism, under anodic potential, the rst step involves adsorption of OH− ions and their discharge to form an adsorbed hydroxide layer 1.7. In the next step, the adsorbed OH reacts with OH− to form an oxide layer (MO) releasing H2O and an electron 1.8. This step is followed by a hydroxylation reaction where the oxide layer reacts with OH− to form OOH species on the surface1.9. This OOH (oxide-hydroxide) species reacts with OH− to form adsorbed O2 and H2O with the release of an electron 1.10. The adsorbed O2 then undergoes desorption in the last step 1.11. M + OH − → M OH + e-

(1.7)

M OH + OH − → M O + H2 O + e−

(1.8)

M O + OH − → M OOH + e−

(1.9)

M OOH + OH − → M O2 + H2 O + e−

(1.10)

(1.11) For ecient water splitting, it becomes imperative to develop a single catalyst, made up of transition metals, to catalyse HER and OER equally well. M O2 → M + O2

1.5 Nanocatalysis The eld of nanocatalysis is not as new as it looks and actually, its concept is known since the 1950s when the term nanotechnology was not even coined. Nanocatalysis merges the advantages of both homogenous and heterogeneous catalyses, while ignoring their respective drawbacks. In homogeneous catalysis, the reactants and the catalyst are present in the same phase, ensuring high catalytic activity and selectivity. However, the practical application of homogeneous catalysis is limited owing to the diculties in separating the catalyst from the product after completion of the reaction. In heterogeneous catalysis, the reactants and the catalyst reside in dierent phases, thereby alleviating the separation of products and catalyst. A major drawback of traditional heterogeneous catalysis when compared to homogeneous catalysis is the reduced surface area that is accessible to reactant molecules, thereby limiting the catalytic activities and also leading to an unnecessarily high consumption of expensive catalyst materials. A possible solution to 17


this problem is to increase the surface to volume ratio by decreasing the size of the catalytically active material. A high surface to volume ratio can be achieved by synthesizing specically engineered catalysts on the nanoscale. Heterogeneous nanocatalyts play a crucial role in both the H2 production techniques under consideration. In the process of hydrolysis of chemical hydrides, the rate of H2 production is enhanced signicantly when mediated by a nanocatalyst. The high H2 production rates are crucial when used for on-board applications. For water electrolysis, usually high overpotentials are required to split water using conventional electrodes. The role of nanocatalysts here is to lower the activation barrier for water splitting, thereby reducing the overpotential requirements. Also, nano-size domain imparts unique characteristics to the catalysts making them viable for multiple reactions.

1.6 Thesis Overview This thesis reports the development of new kinds of catalysts that are economical, easy to synthesise and at the same time show high catalytic activity, comparable to that of noble metals for H2 production by hydrolysis of chemical hydrides and electrolysis of water. All the reported catalysts consist non-noble elements (Co, Ni, Mo, etc.) which can be easily upscaled to meet the present and future energy demands. The thesis covers two major objectives: 1. To present dierent strategies employed in the development of Co-B based catalysts for H2 production by hydrolysis of chemical hydrides. 2. To establish Co-B and its alloys (CoNiB, CoMoB) as active materials for electrochemical water splitting. The present chapter sheds light on the immediate need for clean energy. We saw a glimpse of the important role that H2 fuel will play in our bid to meet the future energy demands. The dierent techniques to produce clean and renewable H2 were dicussed in brief with impetus on hydrolysis of chemical hydrides and electrolysis of water. The importance of nanocatalysis towards successful implementation of these techniques on a mass-scale was also discussed. The development of various nanocatalysts for ecient H2 production by both the means were listed to highlight the importance of Co-based materials. Chapter 2 describes the techniques used for preparation of the catalysts and the experimental setup that was used to carry out all the hydrogen generation measurements. Chapter 3 focuses on Co-B NPs supported on three dierent types of mesoporous (MSP) silica i.e. SBA-15, MCM-41 and FSM-16. Catalytic performance of these supported NPs for H2 production by hydrolysis of ammonia 18


borane was discussed. Based on several characterization results, a connection was established between pore structures of the three dierent mesoporous silica and dispersion of Co-B NPs on each of them. In chapter 4, we discuss the catalytic activity of MSP Co-B catalysts for H2 production by hydrolysis of alkaline sodium borohydride solution. Two types of MSP Co-B catalysts were prepared using two dierent surfactants - a cationic and a non-ionic. Enhanced catalytic activity of MSP Co-B was observed in both cases which was correlated to their high surface area and typical pore structures. In chapter 5, we focus on the use of inexpensive Co3O4 catalyst NPACs synthesized on glass substrates, by using PLD technique, for H2 production from hydrolysis of sodium borohydride. The Co3O4 NPACs showed enhanced H2 generation rate as compared to the bulk Co3O4 powder, the causes of which were investigated and reported in the chapter. In Chapter 6, we present, for the rst time, Cobalt-Boride (Co-B) as an electrocatalyst material for hydrogen evolution reaction (HER). We discuss the outstanding HER activity that was recorded with Co-B catalyst in wide range of pH values. From experimental and theoretical results, it was found that a reverse electron transfer takes place from B to Co, providing favorable charge conduction during water reduction along with high resistance against deactivation. Chapter 7 presents Co-Ni-B nanocatalyst, prepared by a facile reduction method, to be highly active for hydrogen evolution reaction (HER) in various pH media. By XPS, EXAFS, XANES and DFT calculations, it was observed that the presence of Ni causes B enrichment on the surface of Co-Ni-B and also provides excess electrons to Co sites, making them more active than Co-B. In chapter 8, we discuss the application of Co-Mo-B nanocatalyst as a bifunctional material to catalyze hydrogen and oxygen evolution reactions, in alkaline media. Co-Mo-B was also found to be active for HER in neutral media. From the morphological, physicochemical and electrochemical analysis, it was explained that addition of a small amount Mo leads to formation of well dispersed NPs with an increment in the specic surface area by 2 times and electrochemical surface area (ESA) by 1.6 times, when compared to Co-B, providing possible explanations for enhancement in activity.

19

Chapter 2 Experimental Techniques 2.1 Catalyst preparation techniques 2.1.1 Preparation of Powder catalyst Co-B powder catalyst was synthesised by chemical reduction method. In a typical procedure, an appropriate amount of cobalt chloride, CoCl2 (Sigma-Aldrich) is reduced using a strong reducing agent like sodium borohydride [NaBH4] (which also acts as a boron source) at room temperature (see eq.2.1 ). The molar ratio of NaBH4 over CoCl2 was taken as 3:1 to ensure complete reduction of CoCl2. NaBH4 being a strong reducing agent causes a highly exothermic reaction turning the aqueous solution black instantaneously with lots of eervescence. The mixture was stirred vigorously using a magnetic stirrer until all the hydrogen was liberated from the solution. (2.1) Once the eervescence ceases, the nal solution was centrifuged to separate the remnant black powder which was further cleaned with double distilled water (DDW) and ethanol to remove unwanted ions (Cl− , Na+) and other impurities from the powder formed. The cleaned powder was then vacuum dried at room temperature to obtain the nal nanocatalyst. Co-Ni-B and Co-Mo-B powder catalysts were prepared in a similar manner. An aqueous mixture solution of cobalt chloride (CoCl2) and nickel chloride (NiCl2) / sodium molybdate (Na2MoO4) is prepared. The solution is then reduced using NaBH4 under vigorous stirring at room temperature. The same procedure as for Co-B was used for washing and drying the powder. The molar ratio in the nal powder can be adjusted by varying the concentration of Co and M (= Ni or Mo) salts in the starting solution. 2CoCl2 + N aBH4 → Co2 B + N aCl + 12H2

20

Chapter 2: Experimental Techniques

Co3O4 powder was prepared by the co-precipitation (CP) method. Dilute ammonia solution (25%) was added as a precipitator to the homogeneous 0.04 M aqueous solution of CoCl2 under continuous stirring. After stirring at room temperature, the obtained cobalt hydroxide (Co(OH)2) precipitate was ltered and subsequently washed with distilled water and ethanol for several times. Finally, the Co(OH)2 precipitate was vacuum dried and later thermally treated at 400 °C for 2 h to obtain the Co3O4 powder.

2.1.2 Preparation of mesoporous nanoparticles 2.1.2.1 MCM-41 type mesoporous silica

This type of mesoporous (MSP) silica particle was synthesized by using n-cetyltrimethyl ammonium bromide (CTAB) as the surfactant template to assemble mesopores on the surface. CTAB (0.25 g) was rst dissolved in 120 ml of deionized water and stirred for 30 min. NaOH (2 M, 0.875 ml), as a catalyst, is then added to the above mixture, followed by adjusting the temperature to 353 K. Dropwise addition of tetraethylorthosilicate (TEOS, 99.999%) (1.25 ml) was carried out to the surfactant solution and resulting mixture was stirred at 353 K for 2 h to obtain white silica precipitate. The solid powder obtained by centrifuging was washed thoroughly with distilled water and dried in vacuum condition at ambient temperature. After drying, silica powder (0.7 g) was reuxed in a solution of 70 ml of methanol and 3.5 ml of HCl (35%) for 24 h in order to remove surfactant template (CTAB) from the surface. The white powder was separated by centrifuging the solution and later washed with distilled water and methanol followed by drying in vacuum. The obtained powder was nally treated at 398 K for 2 h to remove the OH− group from the pore surface. 2.1.2.2 FSM-16 type mesoporous silica

Kanemite (NaHSi2O5) and CTAB were used as precursor and template, respectively, to synthesize FSM-16 type MSP silica. To synthesize kanemite, 100 ml (0.528 M) sodium hydroxide was added to 50 g of sodium silicate solution (SiO2/Na2O = 2.21) to adjust SiO2/Na2O = 2. The mixture was then stirred for 3 h at ambient temperatures. The excess water was removed by drying the solution at 393 K for 15 h. After crushing, the dried sodium silicate powder was calcinated at 973 K for 6 h in air to obtain white kanemite foam. Crushed kanemite powder (2 g) was dispersed in 20 ml of distilled water and stirred for 3 h. The resulting suspension was ltered out to obtain wet kanemite paste. This paste was mixed into 32 ml of CTAB (0.125 M) solution and then stirred for 3 h at 343 K. At this stage, pH 21


of the dispersion was adjusted to 8.5 by adding HCl solution (2 M). The resulting mixture was stirred for another 3 h at 343 K. After cooling, ltered solid powder was washed with distilled water and then dried in air at ambient temperature. The as-synthesized silica powder was calcinated at 873 K for 6 h to remove the surfactant from the silica material to obtain FSM-16 MSP silica. 2.1.2.3 SBA-15 type mesoporous silica

Pluronic (P123) was used as the surfactant template to fabricate SBA type MSP silica. The synthesis was carried out without using hydrothermal conditions. P123 (4 g) was rst dissolved in 150 ml of HCl solution (2 M) under constant stirring at 308 K for 2 h. Later TEOS (8.5 g) was added to the above mixture and stirred for 20 h at 308 K. The resulting solution mixture was aged at 353 K for 48 h. The solid powder was separated by ltration and followed by washing with distilled water and ethanol. After drying at room temperature, the white powder was calcinated at 773 K for 6 h to burnout template molecules from the silica material. Non-Porous Silica (NPS) particles were also prepared by hydrolysis and condensation of TEOS in ethanol, and in the presence of ammonia (NH3) as catalyst. The molar ratio of TEOS:NH3:H2O:ethanol was kept about 1:1:10:30. 2.1.2.4 MSP silica supported Co-B catalyst

Co-B catalyst was loaded on the three dierent types of MSP silica (MCM-41, SBA-15 and FSM-16) by impregnationreduction method. 300 mg of all types of silica particles were immersed in the 4.5 ml of aqueous cobalt chloride solution (0.5 M). To have better particle dispersion the mixture was ultra-sonicated for 10 min and later left undisturbed for 24 h. After impregnation, the samples were ltered to remove the excess CoCl2 solution and dried in vacuum condition. The dried mixture was reduced by addition of 4.5 ml of aqueous NaBH4 solution (1 M) and later stirred until the bubble generation was ceased. The gray powder formed during the reaction was separated from the solution by sedimentation and ltration and later washed several times with distilled water and ethanol. In the end the catalyst was dried in vacuum under ambient condition. 2.1.2.5 Mesoporous Co-B catalyst

Mesoporous Co-B catalyst was synthesized by two dierent methods, employing two dierent types of surfactants - cationic (CTAB) and non-ionic (P123). 0.362 g of surfactant CTAB (or P123) was added to aqueous solution (0.05 M) of cobalt chloride (CoCl2) and stirred for 60 min at around 318 K. After cooling at room temperature, sodium borohydride (NaBH4) (0.24 g), as a reducing agent, was 22


added to the mixture under continuous stirring. When bubble generation ceased, the remnant solution was ltered and the resulting black powder was washed extensively with distilled water for several times followed by ethanol for two times. Later, the catalyst powder was transferred into 100 ml of ethanol and the combined mixture was reuxed at 353 K for 24 h. The reuxed solution was then collected and washed with ethanol for three times followed by drying in vacuum.

2.1.3 Thin lm catalyst using Pulsed Laser Deposition (PLD) The experimental set-up is schematically represented in Figure 2.1. The energy source used in the experiments is a KrF ( = 248 nm) excimer laser (LPX220i Lambda Physik) capable of operating at frequencies between 1 - 200Hz. The pulse width is about 20 ns (FWHM) and the maximum pulse energy is 450 mJ with maximum average power 80W. The pulse energy can be lowered by dropping the pumping high voltage. The beam dimensions is 8 mm x 23 mm (vertical x horizontal) and beam divergence 1 mrad x 3 mrad (V x H).

Figure 2.1: 3D view of the PLD apparatus. The laser beam passes through a slit where it is cleaned of haloes. It is deected by 90º by means of a dielectric mirror and then focused by a 40 cm or 30 cm focal length lens, which is mounted on a slide permitting it to be translated along the beam direction from the focusing position (where the beam is focused onto the target) away by about 15 cm. The slide is used to change the energy uence on the target. The focused beam nally enters the treatment chamber through a fused silica window and impinges on the target at 45°. To obtain Co3O4 nanoparticle assembled coatings (NPACs), laser ablation was carried out by using a pure metallic Co target under oxygen gas atmosphere. Prior to the deposition, 23


the PLD chamber was evacuated up to a base pressure of 10−6 mbar. NPACs were deposited on Si and glass substrates by maintaining target to substrate distance at 5.5 cm, with substrate position parallel to the target, or normal to the ow velocity of the plume expansion. Several samples were deposited by varying O2 gas pressure (3 Ö 10−3, 8 Ö 10−3, 4.5 Ö 10−2 and 8 Ö 10−2 mbar), substrate temperature (25, 200, 250 and 300º C) and dierent laser uences (1, 3, 5 and 7 J/cm2). Out of the three parameters, two were kept constant and the third was varied during each deposition of the coating. Weight of the catalyst lms was evaluated by measuring the weight of the glass substrate, before and after the deposition.

2.2 Hydrogen Measurement Set up 2.2.1 Experimental setup Figure. 2.2 shows a schematic of the H2 measurement setup. The reaction chamber is a thick-walled glass vessel designed to accommodate the catalysts in the form of powder as well as thin lms. The reaction chamber is closed by a glass lid provided with an O-ring to avoid any leakage. The lid has 3 necks (24/29) and is closed using a SS clamp to make the joint completely air-tight. The three necks serve the following purposes gas outlet, thermometer pocket, solution insertion. The gas outlet is connected to an Erlenmayer ask (1000 ml) lled with water through a gas tube. Another tube connects the ask to a graduated cylinder (1000 ml) placed on an electronic precision balance (Contech CT-6001, readout 0.1 g) which records the change in volume of water in the cylinder. The ow of water from ask to the cylinder is controlled by a valve placed between them which isolate the two systems. The balance is coupled to a computer using RS-232 interface and an indigenous novel algorithm helps in acquiring the data in real time and save it for post-processing. 2.2.2 Gas Volumetric Method Before beginning the experiment, the valve (8) shown in Figure 2.2 is always kept closed. The catalyst to be used (powder or lm) along with a magnetic stirring bead is placed in the reaction chamber with 190 ml of double distilled water (DDW) and the lid is clamped to make it air-tight. Once the reaction chamber is sealed, the valve is opened so that water ows from the ask into the cylinder to compensate for the dierence of pressure in the two systems. Once the system is stabilized and no more water ows into the cylinder, the setup is ready to start 24


the reaction.

Figure 2.2: Schematic diagram of the experimental setup for hydrogen gas measurement using gas-volumetric method. 1: Water bath; 2: Reaction chamber; 3: Syringe for insertion of hydride solution; 4: thermometer; 5: Erlenmayer ask; 6: Lab jack; 7: Electronic balance; 8: ball valve; 9: measuring cylinder; 10: RS 232 coupled acquisition software.

In a typical experiment, 10 ml of an alkaline stabilized solution of chemical hydride (desired molarity) is inserted into the reaction chamber through a syringe via a rubber septum. As soon as the solution comes in contact wih the catalyst, hydrogen production starts and the pressure inside the reaction chamber increases. The produced hydrogen gas displaces the water in the ask to ow in to the cylinder. Continuous rise in volume of water is proportional to the amount of H2 gas produced. This change in volume (V) is continuously acquired by the software as a function of time (t). As density (ρ) of H2 gas is known (0.0898 g/lit), the volume of displaced water can be used to determine the mass (m) of H2 gas produced; (2.2) From the mass obtained, number of moles (N) of H2 gas evolved can be determined by; m=ρ∗V

N=

25

m M

(2.3)


Where, M = molar mass of H2 (2.016 g/mol). Knowing the initial molarity of hydride solution, the H2 generation yield can be calculated; H2 generation Y ield =

N ∗ 100% Expected moles

(2.4)

The rate of H2 production can also be calculated using; RateH2 =

1 dV (t) ∗ dt weight of catalyst used (g)

(2.5)

in units of ml/min/g of catalyst.

2.3 Electrocatalytic measurement methods 2.3.1 Preparation of catalyst modied glassy carbon (GC) electrodes A 3 mm glassy carbon electrode was polished to remove any form of contaminants from the surface. 5 mg of the catalyst to be loaded (Co-B, Co-Ni-B or Co-MoB) was dispersed in 1 ml ethanol, with 10 µL of 5% Naon, under continuous sonication, to obtain a homogenous catalyst ink. 30 μL of this ink (in steps of 10 μL ) was drop-casted onto the GC surface and dried under ambient environment to obtain a uniform catalyst lm with geometric surface area of 0.07 cm2 and a mass loading of ∼2.1 mg/cm2. 2.3.2 Preparation of electrolyte solutions HClO4 (0.1 M) and NaOH (1 M) were used as the electrolytes for all the electrochemical measurements at extreme pH values of 1 and 14 respectively. For neutral pH (7), 0.5 M potassium phosphate buer solution (KPi) was prepared by mixing K2HPO4 and KH2PO4 in appropriate concentrations. 0.5 M KH2PO4 and 0.4 M K2HPO4 solutions were used as electrolytes for the measurements at pH = 4.4 and pH = 9.2 respectively. Prior to their use, all the electrolyte solutions were purged with pure N2 gas for about 20 min to remove dissolved oxygen. 2.3.3 Electrochemical measurements All the electrochemical measurements were performed with a potentiostat-galvanostat system (PGSTAT 30) from Autolab equipped with electrochemical impedance spectroscopy (EIS). The electrochemical cell used was a conventional three elec26


trode design with catalyst modied GC electrode as the working electrode, a saturated calomel electrode as the reference and a Pt sheet (0.5 cm2) as the counter electrode. The reference electrode was calibrated with respect to reversible hydrogen electrode (RHE). The calomel electrode was immersed in a cell lled with H2 saturated electrolyte where one Pt sheet served as the counter while another Pt sheet as the working electrode. Cyclic volammetry (CV) measurements at a scan rate of 1 mV/s were recorded and average of the two potentials where current becomes zero was considered as the thermodynamic potential for HER. The measured potentials were later converted to RHE by adding a value of 0.241 + (0.05916 x pH). In 0.1 M HClO4, E(RHE) = E(SCE) + 0.301; in 0.5 M KPi, E(RHE) = E(SCE) + 0.655 and in 1 M NaOH, E(RHE) = E(SCE) + 1.068. The actual polarization measurements were performed at a sweep rate of 5 mV/s under continuous stirring (∼ 900 rpm) to avoid accumulation of gas bubbles over the GC electrode. The series resistance (Rs) values were determined using impedance measurement data to compensate for iR losses (Rs = 5 Ω, 3Ω and 1 Ω for pH = 1, 7 and 14 respectively). Tafel slope and exchange current density values were obtained by linear tting the plot of log (i) versus overpotential (η). Turnover frequency (TOF) value was determined using the procedure reported in Appendix A. BET technique was used to establish the actual surface area of the catalyst used for electrochemical measurements. To determine the electrochemical surface area of the catalyst used, CV scans were recorded in a solution of 0.1 M Na2SO4 (pH 7) in the potential window of 100 mV from the open circuit potential (OCP), on either sides. This potential window was chosen so as to ensure that there are no Faradaic processes occuring and the currents are mainly due to charging of the double layer. Long-term stability was examined in potentiostatic mode by maintaining the potential at a constant value and measuring the resultant current density for 40-45 hours. Reusability behavior of the catalysts were tested by conducting cyclic voltametric sweep for 1000 cycles in suitable range with a scan rate of 100 mV/s.

27

Chapter 3 Mesoporous silica supported Co-B: A study on inuence of pore morphology on hydrolysis of Ammonia Borane 3.1 Introduction Hydrolysis of chemical hydrides is one of the best ways to produce pure H2 at room temperature, even if regeneration of spent hydrides would require energy that could be taken from renewable energy sources. With high gravimetric storage (19.6 wt%) capacity, Ammonia Borane (NH3BH3, AB) is the most attractive source for the supply of pure H2 via hydrolysis reaction for portable and on-board applications with fuel cell at room temperature [79, 80]. However, catalyst is the key to control the H2 generation rate during the hydrolysis reaction. Transition metals such as Co and Ni, when fabricated on nano scale with high surface area, can exhibit catalytic performance comparable to that of noble metals. Among the low cost materials, Cobalt Boride (Co-B) showed an exceptional catalytic activity mainly owing to its unique properties with high concentration of coordinative unsaturated sites, and its chemical stability [81, 82]. In addition, Co-B can be produced by simple chemical reduction of cobalt salt. However, due to the exothermic nature of the reduction reaction and ferromagnetic nature of the material, the produced Co-B NPs agglomerate to reduce the eective surface area which in turn hampers the catalytic activity considerably. Several routes were adopted in the past to avoid agglomeration such as by doping with transition metals [83, 84, 85], by using organic templates [86] or by supporting the catalyst on high surface area materials such as rough carbon 28

Chapter 3: Mesoporous Silica supported Co-B for hydrolysis of Ammonia Borane

[87, 88]. However, the preparation of catalyst NPs of desired size, which allows tuning of catalytic activity, still remains a challenge. Thus, it is of paramount importance in developing a system that can provide a degree of freedom to control the size of the catalyst NPs during preparation as well as to maintain this size during the catalytic reaction and also at elevated temperatures. Thus, due to all these reasons, catalyst particles supported over porous materials such as alumina, silica, carbon, and zeolites seems to be a better option to deliver sustainable solution with improved catalytic activity. In such supported catalyst, NPs acquire the size of pores which can be easily tuned using various templates and synthetic pathways. When conned into the pore channels, catalyst NPs attain high resistance against agglomeration as they are isolated from each other in the solid matrix. Microporous materials (pore size < 2 nm) such as zeolite can trap and produce smaller NPs but the diusion of reactant in the small pores to the internal catalytically active sites is restricted [89]. Thus, mesoporous silica materials with highly ordered structure are the most appropriate candidate as support material for the catalyst NPs. Mesoporous silica of various type (SBA-15, FSM-16, MCM-41, HMS, etc.) possess high specic surface area, large pore volume, highly ordered pore structure, narrow size distribution and most importantly the pore size can be tuned from 2 to 50 nm with minimum eort [90, 91, 92]. Their large and interconnected channels allow easy passage for the reactant and product and also oer shape related selectivity. The pore walls of mesoporous silica hold very high strength at elevated temperatures and in corrosive environment to withstand against the disintegration of the pore structure. Several kinds of metal NPs, mainly precious metals such as Pd [89], Pt [93], Ru [94], Au [95], and Ag [96] were supported in the pore channels of mesoporous silica in order to acquire desired size and minimize the usage of these metals. Ni-B [97] and Co-B [98] catalyst supported over mesoporous silica have been eectively used for the hydrogenation reaction with enhanced activity and selectivity. These supported metal NPs not only provide high catalytic activity and selectivity but also show high resistance and stability for long term usage. The present chapter is focused on the synthesis of Co-B NPs within the pores of mesoporous (MSP) silica in order to achieve both control and maintain uniform particle size having high degree of dispersion and resistance against deactivation due to sintering at elevated temperatures. For this purpose, Co-B catalyst NPs were supported over three types of mesoporous silica (MCM-41, FSM-16, SBA15) of dierent pore size and texture by impregnation reduction method. Ncetyltrimethyl ammonium bromide (CTAB) was used as surfactant template to synthesize MCM-41 and FSM-16 type of MSP silica while Pluronic (P123) was used for SBA-15. Non-Porous Silica (NPS) particles were also prepared and used 29


for comparative studies. Catalytic performance of these supported NPs was tested for the H2 production by hydrolysis of AB. Based on several characterization results, the connection between pore structures of the mesoporous silica and the size, location and dispersion of Co-B NPs was established which are correlated to the obtained catalytic variation in the hydrolysis of AB.

3.2 Characterization of mesoporous structures The three types of MSP silica used here possess peculiar pore morphology. To understand the dierences in their morphology, structural characterization of the silica supports in small-angle and wide-angle was performed by conventional X-ray diraction (XRD) using the Cu Kα radiation ( = 1.5414 Å ) in Bragg-Brentano (2Θ) conguration. The BET surface area of the unsupported and Co-B supported MSP silica particles was determined by nitrogen adsorption at 77 K after degassing at 473 K for 2 h. Pore size, pore wall thickness, Co-B particle size and pore structure were examined using a transmission electron microscope (energy of 200 keV). The amount of Co-B loading in mesoporous silica was established by energydispersive spectroscopy analysis equipped with scanning electron microscope.

X-ray Diraction (XRD): Small-angle XRD (SAXRD) patterns of NPS, MCM-41, FSM-16 and SBA-15 silica particles are reported in Fig. 3.1. No evidence of any peak in the pattern of NPS particles clearly indicates the absence of porosity while a single peak is observed at two values of 2.23 and 2.61 for MCM-41 and FSM-16 silica, respectively, which is indexed as (1 0 0) reection. On the other hand, SBA-15 type silica prepared by non-ionic block co-polymer surfactant displays three well-resolved diraction peaks at 1.05, 1.68, and 1.92 which can be indexed as (1 0 0), (1 1 0) and (2 0 0) reections associated with p6mm hexagonal symmetry [99]. All these peaks in the mesoporous silica are assigned to the regular array of hexagonal pore structure. The presence of higher order reections demonstrate that SBA-15 silica acquire highly ordered hexagonal array of pores with uniform pore size as compared to MCM-41 and FSM-16 silica. XRD pattern in wide-range for unsupported Co-B catalyst and supported over NPS, MCM-41, FSM-16 and SBA-15 silica particles are reported in Fig. 3.2. The broad peak at around 2Θ = 45 assigned to the amorphous state of Co-B alloy is observed for unsupported and supported over NPS. On the contrary, no peak of Co-B is observed for the catalyst supported over all the three mesoporous silica particles, thus indicating that the catalyst species were highly dispersed in the 30


pores of the support material. The diraction pattern indicates short-range order and long-range disorder of Co-B alloy and both features are expected to enhance the catalytic activity.

Figure 3.1:

Small-angle XRD pattern of NPS, MCM-41, FSM-16 and SBA-15 type silica

materials.

Figure 3.2:

Wide-angle XRD pattern of unsupported Co-B powder and that supported on NPS, MCM-41, FSM-16 and SBA-15 type silica materials.

31


Brunauer, Emmett and Teller (BET) surface area measurements: Nitrogen adsorption desorption isotherms of all the three mesoporous silica supports (MCM-41, FSM-16 and SBA-15) before and after Co-B loading are presented in Fig. 3.3. The inset of these gures show the pore size distribution calculated using BJH method. The physico-chemical parameters such as BET surface area, average pore diameter and pore volume obtained from the isotherm are summarized in Table 3.1. All these mesoporous silica showed Type IV adsorption desorption isotherm according to the IUPAC classication [100]. This kind of shape is typical characteristic of mesoporous material with tubular pores [101]. The Type IV isotherm generally signals three regions. Initially at low P/P0, the at region is related to the monolayer formation on the silica outer surface and on the pore walls. As the P/P0 increases, due to the spontaneous lling of the mesopores by the capillary condensation, a certain step or inection is observed in the volume of gas adsorbed. Later, volume increases gradually with the P/P0 and is mainly attributed to the multilayer adsorption on the outer surface of the particle. Finally, signicant rise in N2 adsorption at P/P0 > 0.9 is caused by the lling of macropores formed by the gaps between the silica particles. The sharpness of the inection indicates the uniformity of the mesopore size distribution. Thus, Fig. 3.3 shows that the mesopores of SBA-15 and FSM-16 attain narrow size distribution as compared to MCM-41 silica. The inection region is extended in P/P0 range from 0.05 to 0.30 (Fig. 3.3a) for MCM-41 silica with the pore size distribution in the range of 2 8 nm (inset of Fig. 3.3a). Another important characteristic of the mesopores can be obtained from the shape of adsorption desorption hysteresis loop of the isotherm which are correlated with the texture of the adsorbent. SBA-15 silica showed H1 type hysteresis loop corresponding to ordered porous material with cylindrical pores open at both ends and interconnected pore structure [102]. H4 type hysteresis loop is demonstrated by the FSM-16 type silica indicating a mixture of microporous (pore size less than 2 nm) and mesoporous structure [100]. On the contrary, distinct hysteresis loop is not observed in case of MCM-41 type silica. Though pore volume of all the three mesoporous silica is nearly same, the average pore size decreases in order of SBA-15 (6.1 nm) > MCM-41 (3.4 nm) > FSM-16 (2.5 nm). Pore wall thickness (Table 3.2) was calculated by subtracting the pore diameter from the spacing between the regular array of pores (a0) (obtained from the formula {a0 = 1.154 Ö d100}, where d100 is estimated from the rst peak in SAXRD (Fig. 3.1)). The wall thickness of SBA-15 is two times thicker than that of MCM-41 and FSM-16, thus showing the strength and the robustness of pore structure of 32


SBA-15 type silica.

Figure 3.3:

Nitrogen adsorption desorption isotherms of (a) MCM-41, (b) FSM-16 and (c) SBA-15 type mesoporous silica supports with and without Co-B catalyst loading. Inset of the shows the pore size distribution curves of the corresponding mesoporous silica supports.

gures

33


Samples

Co-B powder NPS Co-B/NPS MCM-41 Co-B/MCM-41 FSM-16 CoB/FSM-16 SBA-15 CoB/SBA-15

BET surface area (m2 /g)

Average Pore Diameter (nm)

20 31.5 11 970 335 958 630 627 455

Pore Volume (cm3 /g)

3.40 3.10 2.52 2.50 6.15 6.14

0.86 0.35 0.85 0.62 0.83 0.64

Table 3.1:

Physico-chemical properties of non-porous and three dierent mesoporous silica supports (MCM-41, FSM-16, and SBA-15) with and without Co-B catalyst loading.

The BET surface area of SBA-15 is lower than that of MCM-41 and FSM16 type due to larger pore size and pore wall thickness. Incorporation of Co-B catalyst to SBA-15 silica did not inuence the shape of the isotherm. In contrast, incorporation of Co-B to MCM-41 silica caused signicant variation in the lineshape of the isotherms, where inection characteristic completely disappeared. While in case of FSM-16, the inection is still present after loading Co-B but the sharpness of the step is decreased along with fading of hysteresis loop. The average pore diameter remains more or less intact for all the three mesoporous materials even after incorporation of Co-B (Table 3.1). However, the pore volume and BET surface area decreases on Co-B loading with more prominent eect in MCM-41 as compared to SBA-15 and FSM-16. The above results clearly suggests that Co-B particles are located inside the pores of SBA-15 silica by keeping the pore structure intact while for MCM-41, catalyst particles either completely ll the pores or lie outside on the face of pores thus making the pores inaccessible to nitrogen. In case of FSM-16, the micropores are completely blocked by the Co-B as indicated by the disappearance of the hysteresis loop. Samples

d(100) (nm)

a0 (nm)

MCM-41 FSM-16 SBA-15

4.03 3.38 8.40

4.65 3.91 9.70

Table 3.2:

Pore Wall Thickness (nm)

1.25 1.40 3.55

Pore wall thickness and spacing between the two regular arrays of pore channel of mesoporous silica supports calculated from the SAXRD and N2 absorption desorption isotherms.

34


Transmission Electron Microscopy (TEM): From the above nitrogen adsorption results, it is very dicult to visualize the location of the Co-B particles whether they lie inside the channels of pores or they are covering the pores from external surface. Thus, for better understanding, TEM images were recorded for all the samples. Unsupported Co-B powder is mainly composed of spherical particles with size in the range of 30 40 nm (Fig. 3.4). However, due to the exothermic nature of the reduction reaction and the high surface energy involved, these particles are mostly present in agglomerated state to acquire low specic surface area. In addition, ferromagnetic nature of the Co-B particles can also assist in agglomeration of NPs. TEM images show that NPS particles prepared by Stober method are perfectly spherical with narrow size distribution in the range of 150 160 nm (Fig. 3.5a). Co-B particles supported on these NPS showed similar size (30 40 nm) and morphology to that of unsupported Co-B catalyst with slightly less agglomeration (Fig. 3.5b).

Figure 3.4: SEM image of bare Co-B powder MCM-41 silica particles mainly acquire irregular spherical shape with size in the range of 80 120 nm (Fig. 3.5c). The surface of these particles is composed of regular hexagonal arrays of mesopores with uniform pore size as shown in the TEM image of MCM- 41 (Inset of Fig. 3.5c). The distance that is repeated between the pores is measured around 4.5 nm, which is in perfect agreement with the spacing value (a0) (Table 3.1) obtained from the peak in the SAXRD pattern. The pore size and wall thickness were measured around 3.2 3.5 nm and 1 1.3 nm, respectively. These values are consistent to those obtained by SAXRD and BET measurement (Table 3.2). Crumpled paper like structure was observed for FSM-16 silica (Fig. 3.5e). TEM viewed down in the direction of the pore axis reveals a hexagonally ordered mesoporous structure with regular arrangement of pores of uniform size. The pores are seen to be arranged in the patches composed of regular rows on the silica sheet with the spacing of 4 nm between them. This value along with 35


pore size (2.3 2.5 nm) and pore wall (1.2 1.4 nm) measured by TEM is in good agreement with that obtained by the structural and N2 adsorption data (Tables 3.1 and 3.2). Like FSM-16 and MCM- 41, SBA-15 type silica also shows well ordered hexagonal arrays of 2D mesoporous channels, especially along the direction of the pore axis (Fig. 3.5g) or in the direction perpendicular to the pore axis (Fig. 3.5i). However, the pores here are well distinct with bigger size (6.2 nm) and walls are much thicker (3.5 nm) than the other two mesoporous materials. Even though all the three mesoporous silica have similar pore structures, they accommodate Co-B particles in a quite dierent manner. As observed by the TEM micrograph, the Co-B particles are located on the outer surface of MCM-41 (Fig. 3.5d) and FSM-16 (Fig. 3.5f) silica while the catalyst particles are well placed inside the channel of the SBA-15 type silica (Fig. 3.5h and j). This was concluded on the basis that the porous structure of FSM-16 and MCM-41 is not visible while for SBA-15, the hexagonal array structure is well maintained after Co-B loading. The Co-B particles are well conned in the pores of SBA-15 acquiring the size of pores (∼6 nm) (Fig. 3.5h). Along the channel, the size of Co-B slightly increases to around 10 nm (Fig. 3.5j). In case of FSM-16 and MCM-41, the Co-B particles are well dispersed on the surface having broad distribution of size in the range from 3 to 30 nm. However, most of them (90%) have size lower than 15 nm. Due to the irregular shape of particles and large thickness of the support, the determination of the exact particle size is hindered. Roughly calculated average particle size is about ∼10 nm and ∼12 nm for MCM41 and FSM-16, respectively. Particle size greater than the pore size conrms that Co-B particles are located on the surface of the MCM-41 and FSM-16 type silica with some portion of the particle anchored into the pores. SAED pattern for Co-B particles supported on all three mesoporous silica exhibits diuse diraction rings thereby conrming the amorphous nature of Co-B particle as observed in the XRD pattern.

3.3 Mechanism of Co-B loading on dierent pore structures The morphological analysis clearly shows that the particles do not directly acquire the size of the support pores and depending on the texture of the mesopores, the particles are located on the pores. The dierent morphologies obtained on dierent mesoporous silica are attributed to the impregnation reduction process. During impregnation, CoCl2 solution lls most of the pores of mesoporous silica by capillary action. 36


Figure 3.5: Bright eld TEM micrograph of bare (a) NPS, (c) MCM-41, (e) FSM-16, (g) and

(i) SBA-15 type silica supports while (b), (d), (f), (h) and (j) shows micrographs of corresponding supports with Co-B catalyst loading.

Generally, during the reduction process by NaBH4, Co-B particles are formed by release of H2 gas. Thus, in case of SBA-15, due to the interconnected pore 37


assembly H2 can leave the interior of the SBA-15 easily. In FSM-16 and MCM-41, the pores are not connected and thus, in this case, H2 can be released only from the pore face which is blocked by the Co-B particles. Thus, due to the pressure exerted by the H2 gas, the Co-B particles are pushed out on the external surface of MCM-41 and FSM-16. The other possibility is that due to the limitation of the preparation method, Co-B particle size cannot be reduced lower than the size of mesopores of MCM-41 and FSM-16 to accommodate it. However, the average size of Co-B particles is around 10 12 nm for MCM-41 and FSM-16 silica with a size distribution from 3 to 30 nm. This is attributed to the fact that CoCl2 not only lls the pores but is also additionally adsorbed on the surface of the support. Thus, when CoCl2 in the pores is reduced by NaBH4 solution, it forms Co-B particles which act as nucleation site for Co-B particles formed on the support surface to grow further. Since large numbers of nucleation sites are available in the form of regularly arranged pores, most of the particles do not grow larger than 10 12 nm. However, some of the pore channels on the surface are not impregnated with CoCl2 solution; thus, Co-B particles formed around these pores may agglomerate to form big particles having size of about 30 nm on the outer surface similar to the case of NPS. High degree of dispersion of these nanoparticles is obtained due to the presence of large number of nucleation sites in the form of mesopores in MCM-41 and FSM-16.

3.4 Hydrolysis of Ammonia Borane The catalytic activity of Co-B particles supported over various silica (non-porous and mesoporous) was tested for H2 production by hydrolysis of AB. H2 generation yield was measured, as a function of time, during the hydrolysis of AB solution (0.025 M) in presence of Co-B catalyst loaded onto NPS, MCM-41, FSM-16 and SBA-15 type silica and Co-B unsupported powder at 298 K (Fig. 3.6). The amount of Co-B (10 mg) loading was kept same for all the catalyst powders. The catalyst weight was thus selected to maintain NaBH4 to catalyst molar ratio around 50. Active nature of all the catalyst powders was conrmed by the fact that H2 is instantaneously produced as soon as they come in contact with the AB solution. Irrespective of type, Co-B supported on mesoporous silica denitely shows higher catalytic activity as compared to Co-B supported on NPS (73 min) and unsupported Co-B catalyst (75 min), and has the ability to complete the reaction within considerably less time (15 min). Among all the mesoporous materials, Co-B supported on the SBA-15 type silica (15 min) was the fastest to complete the reaction: the time taken was two times lesser than that observed for MCM-41 (29 min) and FSM-16 (30 min) supported 38


Co-B catalyst. Expected amount of H2 (H2 / NH3BH3 = 3) was produced by Co-B when supported on mesoporous silica of dierent types, while only 85% of H2 is produced by unsupported Co-B and that supported on NPS. Negligible activity for H2 generation is obtained by using the bare mesoporous and non-porous silica supports without Co-B loading. The H2 generation yield values reported in Fig. 3.6 was perfectly tted by single exponential function for unsupported and NPS supported Co-B catalyst powders. This indicates that hydrolysis reaction proceeds following the rst order kinetics with respect to AB concentration. On the contrary, linear function was required to t the H2 production data for Co-B supported on all the mesoporous silica materials thus proving zero order kinetic reaction with respect to AB concentration. The maximum H2 generation rate (Rmax) achieved by Co-B supported on SBA-15 silica (∼1900 ml/min/g of Co-B catalyst) is 4.2 and 5.3 times higher than that obtained by NPS supported Co-B (∼480 ml/min/g of Co-B catalyst) and unsupported Co-B powder catalyst (∼360 ml/min/g of Co-B catalyst). For mesoporous silica supports, Co-B supported on SBA-15 showed highest H2 generation rate which is about 1.5 times higher than that measured with MCM- 41 (∼1150 ml/min/g of Co-B catalyst) and FSM-16 (∼1200 ml/min/g of Co-B catalyst). The obtained H2 generation rate value with SBA- 15 supported Co-B catalyst is denitely better than that of Co-B nanospindles (1293 ml/min/g of catalyst) [103], and amorphous Co- B catalyst used under ultrasonic hydrolysis reaction (395 ml/min/g of catalyst) [104], and comparable to transition metal doped Co-B[83].

Figure 3.6:

Hydrogen generation yield, as a function of reaction time, obtained by hydrolysis of AB (0.025 M) in the presence of unsupported Co-B catalyst powder, and that supported over NPS, MCM-41, FSM-16, and SBA-15 type silica supports.

39


The catalytic performance of Co-B catalyst supported on NPS is almost similar to that of unsupported Co-B catalyst, a result that may be mainly attributed to the similar morphological characteristic observed in both the cases. Higher activity of Co-B supported on mesoporous silica is ascribed to the presence of smaller sized (6 12 nm) Co-B particles with proper dispersion on the surface. This scenario provides large number of under-coordinated Co active atoms for the interaction with reactant to produce expected amount of H2 with much higher rate than that of unsupported and NPS supported Co-B catalyst. By anchoring Co-B particles over the mesopores, agglomeration could be avoided, thus ensuring high eective surface area during the course of catalysis reaction. This is the main reason for the observed zero order reaction with respect to AB concentration for mesoporous silica supported Co-B catalyst. For mesoporous silica, SBA-15 showed the best catalytic activity which is mainly attributed to the Co-B particle connement in the mesopores. Due to the connement, Co-B particles acquires the size of the pores (∼6 nm) which is smaller than the size of Co-B particles supported over MCM-41 (∼10 nm) and FSM-16 (∼12 nm) silica. As observed from the TEM image, the pore size of SBA-15 is highly uniform. Thus, the Co-B particles also acquire a very narrow size distribution, unlike MCM-41 and FSM-16 supported Co-B particles having size distribution in the range of 3 30 nm. The pores in SBA-15 are open from both ends with internal interconnectivity to provide easy passage for reactant and product solutions as well as for the produced H2 gas. Connement of Co-B particle in the pores also eliminates any possibility of agglomeration during the reaction course and high temperature treatments. This shows that not only the size of the pores but also its texturing aects the location, size and dispersion of the Co-B catalyst particle. The H2 production rate obtained with the catalyst supported over MCM-41 and FSM-16 is same, mostly due to the similar size (∼10 to 12 nm) and location of the Co-B particles on support surface anchored with the pores.

Activation Energy: In order to conrm the eectiveness of Co-B catalyst supported on SBA-15, activation energy barrier was evaluated by varying AB solution temperature. The H2 generation yield, as a function of time, was measured at dierent solution temperatures by hydrolysis of AB (0.025 M) solution using Co-B catalyst supported over MCM-41, FSM-16 and SBA-15 type mesoporous silica as reported in Fig. 4.5. The evaluation of activation energies of rate limiting step is carried out from Arrhenius plot (inset of Fig. 4.5) of the H2 generation rate. Co-B catalyst supported on SBA-15 (43 ± 2 kJ/mol) displays signicantly lower energy barrier value in 40


comparison to Co-B supported on MCM-41 (51 ± 3 kJ/mol) and FSM-16 (58 ± 3 kJ/mol).

41


Figure 3.7:

Hydrogen generation yield, as a function of reaction time, obtained by hydrolysis of AB (0.025 M) at 4 dierent solution temperatures in the presence of MCM-41, FSM-16, and SBA-15 type silica supported Co-B catalyst. Inset shows the Arrhenius plot of the H2 generation rates for each support.

The favorable activation energy again provides an evidence that connement of Co-B particles in the pores with interconnectivity is well suited for the catalytic hydrolysis reaction. In general, the obtained activation energies are lower than that obtained with Rh (67 kJ mol−1) nanoclusters [105], K2PtCl6 (86 kJ mol−1) [106], Ni-Ag (51.5 kJ mol−1) [107], Co/α-Al2O3 (62 kJ mol−1) [108], and comparable to Co-Mo-B/Ni foam (44 kJ mol−1) [109], Ru (47 kJ mol−1) [110], and Co(0) nanoclusters (46 kJ mol−1) [111].

Eect of heat treatment: Generally, unsupported Co-B particles agglomerate when heated at elevated temperatures in anaerobic conditions to form big crystallites [112]. This agglomeration initiates above the treatment temperature of 473 K and the related kinetics reach maximum at 773 K to form micron sized crystals of metallic Co [113]. This modication strongly hinders the catalytic activity of the Co-B catalyst. The heat treatment at 873 K completely deteriorates the catalytic eciency of the Co-B catalyst [84]. The stability of Co-B particles conned in the pores and robustness of the pore structure was tested by heat treating the Co-B catalyst supported over 42


SBA-15 silica at elevated temperatures (673, 773, and 873 K) in Ar atmosphere for 2 hours. The H2 generation yield, as a function of time, obtained by hydrolysis of AB (0.025 M) solution using these heat treated Co-B supported over SBA-15 silica catalysts are reported in Fig. 3.8. H2 generation rate increases when SBA15 supported catalyst was treated at 673 K and just 10 minutes are required to complete the reaction.

Figure 3.8:

Hydrogen generation yield, as a function of reaction time, obtained by hydrolysis of AB (0.025 M) in the presence of untreated Co-B catalyst supported over SBA-15 type silica and that heat treated in Ar atmosphere for 2 h at 673, 773, and 873 K.

However, further increase in treatment temperature to 773 K causes slight decrement in the catalytic activity but it is still higher than the untreated catalyst. At 873 K, the initial induction time was increased slightly to 2 min but later, during the reaction course, the H2 generation rate increases and the reaction is completed in same time as that of untreated sample. This result indicates that, while unsupported Co-B particles get completely deteriorated by heat treating at 873 K, when conned in the pores of SBA-15 they maintain their catalytic activity similar to that of untreated catalyst. Unsupported amorphous Co-B powder undergoes crystallization when heated above 473 K in anaerobic condition to form Co2B phase which is very active phase for hydrolysis of chemical hydrides [109]. But above 673 K, the agglomeration of Co-B particles increases to form big clusters which decompose to form metallic Co crystallites of micron size. This 43


process, especially the decomposition of Co-B to Co metal crystallite, mainly due to the high degree of agglomeration, strongly decreases the catalytic activity of the Co-B alloy catalyst. For SBA-15 supported Co-B, the particles lie within the pores where Co2B active phase is formed in the temperature range of 473 673 K which is responsible for the improvement of the catalytic performance of the supported catalyst at elevated temperatures (673 and 773 K). At higher temperatures, the connement of Co-B particles in the pores possibly avoids the agglomeration of the neighboring particles thus hindering the decomposition of Co2B phase to form inactive metallic Co phase. Thus, even at 873 K, high H2 generation rate is maintained with slight induction time (2 min).

3.5 Conclusion In this chapter, H2 production by hydrolysis of AB was studied by using Co-B NPs catalyst supported by impregnation reduction method over three kinds of mesoporous silica (MCM-41, FSM-16, and SBA-15) of dierent pore size and texture. TEM images and N2 adsorption desorption isotherm revealed that size, dispersion degree and location of Co-B NPs are aected by the pore texturing of the support. Co-B supported over SBA-15 silica was found to be most active catalyst that produces the expected amount of H2 gas from hydrolysis of AB with H2 generation rate of about 1900 ml/min/g of Co-B catalyst, a value that is denitely higher than that measured with MCM-41 (∼1150 ml/min/g of Co-B catalyst) and FSM-16 (∼1200 ml/min/g of Co-B catalyst). The higher eciency of Co-B supported on SBA-15 is explained on the basis of the geometrical connement of Co-B particles within the pores with Co-B NPs that have average size of about 6 nm and uniform size distribution while exhibiting higher degree of dispersion. The eective activation energy of the rate limiting process in hydrolysis operated by Co-B NPs with interconnected pores of SBA-15 results lower than that established with MCM-41 and FSM-16 supported catalyst. Finally, the thicker pore walls of SBA-15 support avoids agglomeration of the Co-B NPs thus providing high stability at elevated temperatures (873 K) as opposed to what occurs with unsupported Co-B NPs catalyst.

44

Chapter 4 Mesoporous Co-B nanocatalyst for ecient hydrogen production by hydrolysis of sodium borohydride 4.1 Introduction In previous chapter, we discussed the role of mesoporous (MSP) silica as template material for Co-B catalyst in enhancing the rate of hydrogen production by hydrolysis of ammonia borane. Supporting the catalyst on MSP silica provides a convenient way of increasing the active surface area but the quantity of catalyst that can be supported is very small as compared to that of the template itself. Thus, it becomes important to identify routes to increase the active surface area of the Co-B catalyst itself without supporting them on other substrates or introducing other elements. This can be achieved by forming nanostructures on the Co-B catalyst surface. Tong et al. [114] and Li et al. [115] reported the synthesis of amorphous Co-B catalyst with mesoporous structures which provide high surface area for improved hydrogenation. In another work by the same authors, Co-B owers with mesoporous structures were synthesized and tested for hydrogen production from hydrolysis of KBH4[116]. In this chapter, we discuss the catalytic activity of MSP Co-B catalysts for H2 production by hydrolysis of alkaline Sodium Borohydride (NaBH4, SBH) solution. Amongst the various chemical hydrides, SBH has been widely accepted as a relevant hydrogen source owing to its high storage capacity (10.9 wt%). SBH has also been reported as a potential fuel for direct fuel cells [117]. Importantly, the reaction product of SBH is borax which is environmentally safe and can be recycled [118]. Herein, we developed two types of MSP Co-B catalysts using two dierent surfactants - a cationic (cetyl trimethyl ammonium bromide, CTAB) and 45

Chapter 4: Mesoporous Co-B for hydrolysis of sodium borohydride

a non-ionic (Pluronic, P123). H2 production rates of the two MSP Co-B catalysts, named Co-B/CTAB and Co-B/P123, were compared with non-porous Co-B powder for the same reaction. Enhanced catalytic activity of MSP Co-B was observed in both cases which can be correlated to the high surface area and the typical pore structures. Activation energy barriers and stability of the mesoporous catalysts were also investigated for hydrolysis reaction.

4.2 Characterization of MSP Co-B catalysts X-ray Diraction (XRD) Figure 4.1 shows the XRD patterns of Co-B/CTAB, Co-B/P123 and non-porous Co-B. The gure shows only one broad peak at 2Θ = 45.5◦ for all the samples. This indicates the typical amorphous nature of the catalysts with short-range order and long-range disorder in Co-B alloy. Both these features are expected to enhance the catalytic activity [119].

Figure 4.1: XRD pattern of Co-B/CTAB, Co-B/P123, and nonporous Co-B catalyst.

BET surface area measurement: The enhancement in surface area for both types of MSP Co-B was established from their BET surface area and corresponding adsorption-desorption isotherms as 46


shown in Figure 4.2. The obtained isotherms resemble Type IV of the adsorptiondesorption isotherm with hysteresis loop as per the IUPAC classication [120, 121]. This type of isotherm is a characteristic of mesoporous materials where gases get condensed in the tiny capillary pores of the adsorbent at pressures lower than the saturation pressure of the gas. The initial at region of the curve observed for relative pressures (P/P0) < 0.4, indicates the monolayer formation on the pore wall and on the outer surface. As the pressure rises further, capillary condensation takes place within the mesopores. Finally, the signicant rise in N2 adsorption at P/P0 > 0.9 is caused by the lling of macropores formed by the gaps between the catalyst particles.

Figure 4.2:

Nitrogen adsorption-desorption isotherms of Co-B/CTAB, Co-B/P123 and nonporous Co-B catalyst.

The hysteresis eect in the isotherms arises due to condensation of the adsorbed layer within the capillaries resulting in a lag in desorption, thus strongly indicating the presence of mesopores. Non-porous Co-B does not portray any hysteresis eect except at higher pressure (P/P0 > 0.9) which is due to the condensation in the macropores formed between the catalyst particles. The shape of 47


Catalyst

BET surface area (m2 /g)

Non-porous Co-B Co-B/CTAB Co-B/P123

∼24.6 ∼114.1 ∼85.5

Table 4.1:

Average pore diameter a (nm)

∼12 ∼16

Pore volume b (cm3 /g)

∼0.37 ∼0.32

Physico-chemical properties of Co-B/CTAB, Co-B/P123 and non-porous Co-B

catalyst.

hysteresis loop is indicative of the pore structure (shape and texture). Based on the IUPAC classication, it is concluded that both the catalysts exhibit H3 type of hysteresis loop. Isotherms with type H3 hysteresis do not exhibit any limiting adsorption at high P/P0. This behavior is caused by the existence of non-rigid aggregates of plate-like particles or assembly of slit-shaped pores with wide pore size distribution. To collect more information on the microporous properties of the catalyst, t-plot analysis was performed which revealed the absence of micropores on the catalyst surface. The physico-chemical parameters such as BET surface area, average pore diameter and pore volume obtained from the isotherm are summarized in Table 4.1. Irrespective of the type of surfactant used, MSP Co-B showed signicantly higher surface area than non-porous Co-B catalyst (∼24 m2/g). Average pore diameter of the mesopores were determined using BET method and it was found that Co-B/CTAB (12 nm) has smaller pore width as compared to Co-B/P123 (16 nm). The mesopores obtained from CTAB and P123 are reported to exhibit much smaller pore sizes of around 6 - 7 nm in the case of mesoporous silica [87]. In the present case, the pore sizes are dierent as there can be variations in the pore size depending upon the interaction between the precursor and surfactant ions which decides how strongly or loosely they will be coupled to each other. Owing to their small pore sizes, Co-B/CTAB catalyst also exhibits higher surface area (BET) (∼114 m2/g) than Co-B/P123 (∼84.5 m2/g). Pore volume, obtained by BJH adsorption method, for Co-B/CTAB (0.37 cm3/g) is also marginally higher than that of Co-B/P123 (0.32 cm3/g). Thus, all the physico-chemical parameters along with the hysteresis nature of the isotherm are conclusive of the presence of slitlike pores on the catalyst surface, with Co-B/CTAB exhibiting smaller pore size, higher surface area and higher pore volume as compared to Co-B/P123 catalyst.

48


Figure 4.3:

SEM image of (a) non-porous Co-B, (b) Co-B/CTAB, (c) Co-B/P123 and (d) bright eld TEM micrograph of Co-B/P123.

Electron Microscopy: The morphology of the as prepared catalysts was investigated by SEM and presented in Figure 4.3(a-c) above. Non-porous Co-B powder is composed of agglomerated spherical particles with the size in the range of 30 - 40 nm (Figure 4.3a). Exothermic nature of the reduction reaction is mainly responsible for the particle agglomeration to acquire low specic surface area (24 m2/g). Co-B/CTAB catalyst shows ower like particles of the size in the range of 120 - 170 nm (Figure 4.3b). This ower shape particle contains nano-petals on the surface with a width of 50 - 100 nm and a thickness of few nanometers. On the other hand, sphere like nodules with highly porous microstructure on the surface are observed for CoB/P123 catalyst (Figure 4.3c). TEM analysis also conrmed the microstructure of Co-B/P123 samples where worm-like porous structure with size in mesoporous range is visible on the surface of sphere like nodules (Figure 4.3d). Microstructural 49


features for MSP Co-B are in good agreement with N2 isotherms where slit-like pores are attributed to the worm-like pore structure of Co-B/P123. In the case of Co-B/CTAB, particles with ower microstructure, having two dimension platelike nano-petals with a gap in the range of mesopores, are responsible for the H3 type hysteresis loop.

4.3 Mechanism of mesopore formation In a solvent, beyond a critical concentration, surfactant forms micellar structure in a way such that the hydrophilic head-groups shield the hydrophobic tails. The charge present on the head-group of a surfactant molecule is mainly responsible for the interaction with charged precursor species to condensate on the micelles. In the case of non-ionic surfactants (like P123), the interaction between the neutral head-group (N0) and the precursor species (Co2+) is mainly due to hydrogen bonding and the assembly (N0Co2+) lies at the micellar interface [122]. NaBH4 reduces Co2+ to form Co-B alloy covering the micellar structures. These surfactant micelles are later removed, leaving behind wide mesoporous structures with large pores. On the other hand, surfactants like CTAB with cationic head-group (S+) usually interact with the like charged precursor species (P+) by means of electrostatic interactions mediated by counter-ions (X−), such as a base, to form an assembly (S+X−P+) around the micelles [122, 123]. This assembly which is driven by electrostatic interaction results in well-ordered, compact mesoporous structures. However, in the present scenario, for Co-B/CTAB, there was no intermediate counter-ion to induce electrostatic interactions between cationic head-group of CTAB and Co2+. Thus Co-B particles, formed after reduction, agglomerate to form clusters with micelles on the surface which are later removed during reuxing to form petal like nano-structure on the surface resulting in high surface area.

4.4 Hydrolysis of Sodium Borohydride The catalytic activity of the synthesized nanocatalysts was tested for hydrogen production by hydrolysis of NaBH4. The amount of hydrogen evolved during the hydrolysis of NaBH4 (0.025 M) was measured and the H2 generation yield was represented as a function of time in Figure 4.4 in the presence of Co-B/CTAB, Co-B/P123, and non-porous Co-B. The amount of Co-B (15 mg) loading was kept same for all the catalyst powders. From the gure, it is strongly evident that irrespective of the type of surfactant used, mesoporous Co-B catalysts are highly active for H2 production than non-porous Co-B catalyst. Though all the three catalysts could produce expected amount of H2, it can be seen that the 50


time taken for completion of the reaction (100% H2 yield) is signicantly lower for MSP Co-B catalysts as compared to non-porous Co-B. Among the MSP Co-B, Co-B/P123 shows better catalytic activity and took shorter time interval than Co-B/CTAB.

Figure 4.4:

Hydrogen generation yield, as a function of reaction time, obtained by hydrolysis of NaBH4 (0.025 M) in presence of Co-B/CTAB, Co-B/P123 and non-porous Co-B catalysts.

During hydrolysis reaction, both the concentrations of the catalyst and of NaOH remain constant but the concentration of NaBH4 decreases with time as hydrogen gas is evolved. This change in concentration of NaBH4 can give us an estimate of the reaction order with respect to NaBH4. For zero-order reaction, the H2 production volume shows a linear dependence as a function of time as given by, d[H2 ] = 4k0 dt

(4.1)

[H2 ](t) = [H2 ]max (1 − e−k1 t ) = 4[[BH4− ]0 (1 − e−k1 t )]

(4.2)

where k0 is the rate constant of the zero-order reaction. In case of rst order reactions, the H2 production volume as a function of time has an exponential dependence as given below,

51


where [BH4−]0 is the initial molar concentration of sodium borohydride in the solution and k1 is the overall rate constant of the reaction. Experimentally obtained data of H2 production volume as a function of time can be reproduced using equations 4.1 and 4.2 where the reaction order and the rate constant are kept as variables to be determined by best tting procedures. Single exponential function (equation 4.2) was required to perfectly t the H2 generation yield values reported in Figure 4.4 for nonporous Co-B catalyst powder indicating the rst order kinetics with respect to NaBH4 concentration. On the contrary, linear function (equation 4.1) was required to t the H2 production data for both the MSP Co-B thus proving zero order kinetic reaction with respect to NaBH4 concentration. The maximum H2 generation rate (Rmax) achieved by CoB/P123 (∼3350 ml/min/g of Co-B catalyst) is the highest and is 4 and 1.5 times higher than that obtained by non-porous Co-B (∼850 ml/min/g of Co-B catalyst) and Co-B/CTAB (∼2200ml/min/g of Co-B catalyst) catalyst, respectively. The obtained value of H2 generation rate with Co-B/P123 catalyst is also comparable to that of 1.5 wt% Pt-LiCoO2 catalyst (3100 ml/min/g of catalyst) [124] whereas both Co-B/P123 and Co-B/CTAB mesoporous catalysts show much higher rate than Raney Ni50 - Co50 (700 ml/min/g of catalyst) [125], Ru catalyst (1600 ml/min/g of catalyst) [126], 10 wt% Pt-Ru-LiCoO2 (1200 ml/min/g of catalyst) [127] and 5 wt% Ru - C catalyst (700 ml/min/g of catalyst) [128]. The high H2 generation rate by both the types of MSP Co-B can be attributed to the presence of slit-like pores which provide high eective surface area for increased surface interaction of the catalyst with the reactants. Among the MSP Co-B nanocatalysts, Co-B/P123 showed better catalytic activity and higher hydrogen generation rate than Co-B/CTAB. This can be understood on the basis that Co-B/P123 exhibits larger pore width which provides easy access of reactants within the pores and at the same time oers no blockage to the release of evolved H2. In the case of Co-B/CTAB, nano-petal like structures are present but the surface pores have smaller width which makes it more dicult for the reactants to enter the pores, thereby restricting usage of available reaction sites on the surface. The uniformity of pores might also play a major role in enhancing the catalytic activity of Co-B/P123.

Activation Energy: To determine the eectiveness of the catalysts, the activation energy barrier was measured from Arrhenius plot obtained from the H2 generation yield measurements at dierent solution temperatures (Figure 4.5). 52


Figure 4.5:

Hydrogen generation yield, as a function of reaction time, obtained by hydrolysis of NaBH4 (0.025 M) at 4 dierent solution temperatures in presence of MSP Co-B/CTAB and Co-B/P123 catalysts. Inset shows the Arrhenius plot of the H2 generation rates.

The slope of the Arrhenius plot (inset of Figure 4.5) gives the value of the activation energy barrier of the rate limiting step for the hydrolysis reaction. The activation energy measured for Co-B/P123 catalyst (40 kJ mol−1) is marginally lower than that measured for Co-B/CTAB (44 kJ mol−1) and non-porous Co-B (45 kJ mol−1) [129]. In general, the obtained activation energy for the Co-B/P123 catalyst is lower than that obtained with Raney Co (53.7 kJ mol−1) [125], carbon supported Co-B (57.8 kJ mol−1) [130], structured Co2B (45 kJ mol−1) [131], Co-B thin lm (44.47 kJ mol−1) [132] and Co-B nanoparticles (42.7 kJ mol−1) [133]. The values are also lower than that obtained with Ru catalyst (47 kJ mol−1) [117] and dierent bulk metal catalysts such as, Co (75 kJ mol−1), Ni (71 kJ mol−1), 53


and Raney nickel (63 kJ mol−1) [100].

Eect of heat treatment: The stability of the mesoporous catalysts was investigated by heat treatment at various elevated temperatures (573, 673, and 773 K) in Ar atmosphere for 2 h.

Figure 4.6:

Hydrogen generation yield, as a function of reaction time, obtained by hydrolysis of NaBH4 (0.025 M) in presence of untreated MSP Co-B/CTAB and Co-B/P123 catalysts, and heat treated in Ar atmosphere for 2 h at 573,673, and 773 K.

The H2 generation yield, as a function of time, obtained by hydrolysis of NaBH4 (0.025 M) solution using these heat treated MSP Co-B catalysts are reported in Figure 4.6. The gure shows that the catalytic activity deteriorates after heat 54


treatment at all high temperatures for Co-B/P123 catalyst. While in the case of Co-B/CTAB catalyst, the activity remains more or less stable up to 673 K but degrades as temperature increases to 773 K. The decline in activity can be attributed to the waning of porous structures with the increase in temperature as conrmed by the SEM images of heat treated MSP Co-B (Figure 4.7). This shows the unstable nature of the pore structures of Co-B at elevated temperatures. Further studies will be carried out to understand the waning mechanism of the pores in various conditions, especially during the hydrolysis reaction, when exposed to ambient atmosphere for long period of time, heat treated in dierent surrounding atmosphere, etc.

Figure 4.7:

SEM images of heat treated MSP Co-B/CTAB and Co-B/P123 catalysts in Ar atmosphere for 2 h at 773 K.

4.5 Conclusion Mesoporous Co-B nanocatalysts were prepared by the reduction of cobalt chloride by NaBH4 in the presence of cationic (CTAB) and non-ionic (P123) surfactant templates. The XRD analysis of Co-B/CTAB, Co-B/P123 and non-porous Co-B showed the amorphous nature of the catalysts with short-range order and longrange disorder of Co-B alloy. Non-porous Co-B powder was found to be composed of agglomerated spherical particles with size in the range of 30 - 40 nm while Co-B/CTAB catalyst showed ower like particles with size in the range of 120 - 170 nm. Sphere like nodules with highly porous microstructure on the surface were observed for Co-B/P123 catalyst. Nitrogen adsorption-desorption isotherms were studied for all the catalyst powders and the obtained isotherms resembled type IV of the adsorption-desorption isotherm with hysteresis loop being characteristic of mesoporous materials with slit-like pores on the catalyst surface which provides high eective surface area. These catalysts were tested for H2 production 55


by hydrolysis of NaBH4. The Co-B/P123 catalyst showed the highest hydrogen generation rate owing to the presence of wide uniform pores which facilitated easier interaction of the reactants to release hydrogen. The lack of stability in pore structure at elevated temperatures was observed for both the mesoporous Co-B catalyst.

56

Chapter 5 Co3O4 Nanoparticles Assembled Coatings prepared by Pulsed Laser Deposition for enhanced H2 production from hydrolysis of Sodium Borohydride 5.1 Introduction From the previous two chapters, it is now well established that cobalt boride (CoB) is an excellent material for catalytic hydrolysis of boron hydrides owing to its exceptional activity, low-cost, ease of preparation and good stability. However, features like low-surface area, broadly distributed particle size and poor thermal stability against crystallization are some of the major drawbacks of this material. These problems arise during the synthesis of Co-B powder catalyst from cobalt salt, where exothermic nature of reduction reaction reduces Co2+ by BH4− in aqueous solution to produce Co-B particles with high surface energy. These unstable particles agglomerate to form clusters and decrease the surface area. As time progresses, the aggregation increases in the ambient atmosphere as well as during the reaction course, which further decreases the catalytic activity and the possibility to be reused. In the previous chapters, we discussed two novel routes to increase the eective surface area of Co-B. Though these routes show improvement in the performance of the Co-B catalyst, a far more robust and stable catalyst is required in order to implement it on an industrial scale. The best way to resolve the issue of agglomeration is by using precursor of cobalt in an oxidized form. The reduction reaction of cobalt oxide to form Co-B 57

Chapter 5: Co3 O4 Nanoparticles Assembled Coatings for hydrolysis of SBH

proceeds slowly with signicantly lower heat release than in the case of cobalt salt. This slowness reduces the surface energy to avoid agglomeration and leads to higher surface area. Krishnan and co-workers [134] rst observed that cobalt oxide supports (Co3O4 and LiCoO2) get reduced during hydrolysis of NaBH4 by replacing oxygen with B to form highly active Co-B phase. This Co-B not only showed better activity than Co-B synthesized from cobalt salts, but also exhibited similar activity to that of noble metals such as Pt-Ru, Pt/C and Ru/C. Later, the eect of crystallinity and morphology of Co3O4 on the activity was also studied [135]. Cobalt oxide encapsulated in C2N polymer network displayed very high H2 generation rate by hydrolysis of NaBH4 [136]. Simagina et. al. [137] investigated the mechanism of Co3O4 powder catalyst at various stages during the hydrolysis reaction course and established that catalytically active Co2B phase is formed in-situ as the reaction progresses. Unfortunately, all these catalysts were developed in the form of powder where separating the catalyst for their reuse is dicult. On the other hand, catalyst in the form of coatings or thin lm, which can serve as an environmentally friendly green catalyst for easy recovery, reuse and can function as ON/OFF switch for the reaction, has not yet been studied. Mainly nanoparticle assembled coatings (NPACs) are best suited catalyst providing high surface area and good stability against aggregation as a result of immobilized and adherent nature of nanoparticles (NPs) on the suitable substrate. Amongst various synthesis techniques, Pulsed Laser Deposition (PLD) has emerged as a viable method for the production of NPACs, in a single step, with desired properties [138, 139, 85] by simply tuning the deposition parameters. Cobalt NPs embedded in boron matrix thin lm catalyst synthesized by PLD showed similar activity to that of Pt/C for hydrolysis of NaBH4 and NH3BH3 [140]. In a recent work [141], reactive PLD was used to synthesize Co3O4 NPs assembled coating with narrow size distribution. The NPs produced by PLD were found to be more ecient for photocatalysis reaction as compared to NPs synthesized by various other chemical and physical methods [142]. In this chapter, we focus on the use of inexpensive Co3O4 catalyst NPACs synthesized on glass substrates, by using PLD technique, for H2 production from hydrolysis of Sodium Borohydride (SBH). The Co3O4 NPACs showed enhanced H2 generation rate as compared to the bulk Co3O4 powder, the causes of which have been investigated and reported in the present chapter. The laser deposition parameters were varied to study the eect of crystallinity, morphology and chemical states of Co3O4 NPs on the catalytic activity.

58


5.2 Characterization of nanoparticle assembled coatings (NPACs) For comparison with Co3O4 powder, Co3O4 coating was deposited by PLD using laser uence of 3 J/cm2, O2 pressure of 4.5 x 10−2 mbar and a substrate temperature of 250 ºC.

Structural and morphological analysis: Figure 5.1 shows the XRD patterns of Co3O4 powder and NPAC samples. The XRD pattern of Co3O4 powder (Figure 5.1b) prepared by co-precipitation method shows diraction peaks at 31.2º, 36.5º, 38.4º, 44.8º, 55.6º, 59.0º and 65.2º.

Figure 5.1: XRD pattern of: (a) Co O

3 4 NPAC prepared by PLD at 250° C with laser uence of 3J/cm2 and O2 pressure of 4.5 x 10−2 mbar, and (b) Co3 O4 powder prepared by chemical method, and (c) standard JCPDS pattern of Co3 O4 .

These diraction peaks correspond to (111), (200), (311), (222), (400), (422) and (511) reection, respectively, of spinel type cubic structure of Co3O4 with Fd3m space group [142]. XRD pattern of PLD deposited Co3O4 coating was obtained in glancing angle mode (Figure 5.1a), so that the maximum signal is acquired from the coating. In comparison to the powder sample, the NPACs showed all the XRD peaks related to cubic Co3O4 although with low intensity and 59


broad nature. The crystallite size of the Co3O4 powder and coating catalyst was around 45 nm and 5 nm, respectively, as calculated using Scherrer equation from the most intense (200) peak. The SEM image of Co3O4 powder sample (Figure 5.2a) shows particle-like morphology with irregular shape and high degree of agglomeration. The average particle size was measured to be about ∼50 nm which is in good agreement with that of crystallite size obtained from the XRD results. On the other hand, PLD deposited Co3O4 coating surface is composed of randomly arranged spherical particles with narrow size distribution (3 - 10 nm) and the average particle size is less than ∼5 nm (Figure 5.2b).

Figure 5.2:

SEM images of (a) Co3 O4 powder prepared by chemical method and (b) Co3 O4 NPAC prepared by PLD at 250° C with laser uence of 3J/cm2 and O2 pressure of 4.5 x 10−2 mbar.

These particles are well distributed on the substrate surface with very low degree of agglomeration. These NPs are formed by high laser uence induced phase explosion phenomena [85]. Here, the irradiated target material reaches the temperature of ∼0.9Tc (Tc is thermodynamic critical temperature) causing a homogeneous nucleation of vapor bubbles of very high pressure below the target surface. The target surface then explodes to make a rapid transition from superheated liquid to a matrix of vapor and liquid nano-droplets, which leave the target surface. Later, these liquid nano-droplets collide with oxygen atoms and oxidation occurs along with re-solidication during the ight to form solid NPs onto the substrate [143]. In case of larger droplets (few hundreds of nm), only some monolayers get oxidized during the ight to form core - shell structure with Co3O4 as shell and Co as core as conrmed by TEM analysis [143]. Figure 5.3 represents the bright eld TEM image of the PLD deposited Co3O4 coating with size distribution analysis. It is clearly visible that the coating surface is decorated with uniformly dispersed NPs of Co3O4 having a mean diameter of about ∼5 nm with a standard deviation value of 1.1 nm. As seen in HR-TEM image (Figure 60


5.3a), the spacing (2.88 ± 0.05) between the lattice planes conrms that NPs are composed of Co3O4 phase (2.85 nm).

Figure 5.3:

(a) Bright Field TEM image and (b) particle size distribution of Co3 O4 NPAC prepared by PLD at 250 °C with laser uence of 3J/cm2 and O2 pressure of 4.5 x 10−2 mbar. Inset of (a) shows HR-TEM of corresponding sample.

XPS analysis: XPS was used to gain insight on the chemical state and surface composition of Co3O4 catalysts and the results are reported in Figure 5.4. In Co 2p core level (Figure 5.4a), two distinct peaks assigned to cobalt oxide in Co 2p3/2 and Co 2p1/2 states are clearly visible in both powder and NPAC synthesized by chemical route and PLD respectively. The prominent peak of Co 2p3/2 level was de-convoluted into two peaks centered at 779.4 and 780.9 eV attributed to Co3+ 2p3/2 and Co2+ 2p3/2 conguration, respectively. The corresponding peaks were also observed in the other spin orbit component (2p1/2). This conrmed the formation of Co3O4 phase with Co2+ and Co3+ species. The energy dierence between the 2p1/2 and 2p3/2 peaks was measured around 14.9 eV which is characteristic of Co3O4 cubic phase thus further conrming the presence of Co3O4 spinel structure in both the samples. Small signal due to the shake-up peaks of Co3O4 phase were also observed at 789 eV and 804 eV. However, the amount of Co3+ was higher than Co2+ species for the NPs assembled coating as compared to chemically prepared powder where both the species were present in equal amounts. In O1s core level (Figure 5.4b), a peak at about 529.4 eV was observed with a shoulder close to 531.5 eV corresponding to oxygen in the Co3O4 crystal lattice and to OH (hydroxyl) group attached to Co, respectively. These peak positions are in well agreement with that reported for Co3O4 network. The presence of hydroxyl group on the catalyst 61


surface was due to our ex-situ experimental conditions and its content was higher for coating than powder.

Figure 5.4:

XPS spectra with (a) Co 2p and (b) O 1s core levels of Co3 O4 powder prepared by chemical method and Co3 O4 NPAC prepared by PLD at 250° C with laser uence of 3J/cm2 and O2 pressure of 4.5 x 10−2 mbar.

62


5.3 Hydrolysis of Sodium Borohydride Catalytic activity of Co3O4 coating synthesized by PLD was compared with chemically prepared Co3O4 powder for hydrogen production through hydrolysis of NaBH4. H2 generation yield was measured, as a function of time, during the hydrolysis of NaBH4 solution (0.025 M) in the presence of both the Co3O4 catalysts at 30 ºC (Figure 5.5) by using same amount of catalyst. Co3O4 coating prepared by PLD was able to produce H2 instantaneously as soon as it came in contact with NaBH4 solution, while the powder catalyst displayed a small induction period to initiate the hydrolysis reaction. However, Co3O4 coating prepared by PLD exhibited much higher catalytic activity than Co3O4 powder and was able to complete the reaction within 210 min while the Co3O4 powder catalyst was able to produce only 30 % of H2 yield in same time period. Three lms synthesized by using the same parameters of PLD showed almost identical (±5%) hydrogen generation rate in all the experiments, thus establishing the reproducibility of the Co3O4 coatings by PLD technique. During the reaction, the catalyst lms were quite stable in terms of H2 production. Maximum H2 generation rate (HGR) obtained by PLD Co3O4 coating (5010 ml min−1 gcat−1) was about 5 times higher than that obtained by Co3O4 powder (1000 ml min−1 gcat−1). The generation rate initially increased to a maximum value and then decreased with time, as expected for a rst order reaction, with respect to NaBH4 concentration [144] which decreases with the reaction time. The H2 generation rate obtained was higher than the reported values of other Co3O4 based catalysts [145, 146, 147] in the literature (Table 5.1). Nanofoam of Co3O4 [135] was able to produce HGR of 2530 ml min−1 gcat−1, while commercial Co3O4 powder reported in the same ref [135] just produces 830 and 860 ml min−1 gcat−1. Krishnan et al. [134] reported the HGR of 850 ml min−1 gcat−1for Co3O4 catalyst prepared by thermal decomposition. HGR of around 2500 ml min−1 gcat−1 was achieved with highly crystalline Co3O4 powder by Komova et al. [148]. Only cobalt oxide nanoparticles embedded in the C2N polymer structure [136] displayed higher HGR (8903 ml min−1 gcat−1) than in the present case. The dierence in the H2 generation rates for Co3O4 powder and coating catalyst can be explained in terms of the existing surface morphology and, importantly, the NPs size distribution. In coating, NPs with smaller average size (∼5 nm) and narrow size distribution (3 - 10 nm) oer high surface area than those in the powder sample, having an average size of ∼50 nm. Oxide NPs in the coating are highly dispersed on the substrate surface to provide large number of active sites for hydrolysis reaction. Most importantly, the NPs are immobilized and strongly held on the substrate surface, thus even during the hydrolysis reaction the catalytic activity was maintained. 63


Figure 5.5: Hydrogen generation yield as a function of reaction time obtained by hydrolysis of

alkaline NaBH4 (0.025 M) using Co3 O4 powder prepared by chemical method and Co3 O4 NPAC prepared by PLD at 250 °C with laser uence of 3J/cm2 and O2 pressure of 4.5 x 10−2 mbar.

Table 5.1:

A comparison of various Co-based catalysts used for hydrogen production by hydrolysis of aqueous sodium borohydride solution.

On the contrary, NPs in powder catalyst are in agglomerated form which might 64


further coagulate during the hydrolysis reaction thus reducing the H2 production rate. Hence, it was clearly shown that the catalyst NPs in the form of coating oer a better solution in increasing the activity and stability.

5.4 Eect of variation in O2 pressure The catalytic hydrolysis of NaBH4 over cobalt oxide proceeds by replacing O with B to form Co-B active sites. Thus it is important to investigate which oxidation state of Co (Co2+ or Co3+) is preferred for generation of the active phase. Thus, the eect of O2 content in the deposited NPACs was investigated by varying the O2 pressure during the deposition. Cobalt oxide NPACs were prepared in dierent O2 pressures, namely 3 x 10−3, 8 x 10−3, 4.5 x 10−2 and 8 x 10−2 mbar with the substrate temperature of 250° C and the laser uence of 3 J/cm2. The coating deposited with higher O2 pressures of 4.5 x 10−2 and 8 x 10−2 mbar produces expected amount of H2 while that deposited at lower O2 pressures 3 x 10−3 and 8 x 10−3 mbar generates only 24 % and 45% of H2 yield in 4 h, respectively. The H2 generation rate proles as a function of O2 pressure are shown in Figure 5.6. The NPACs deposited at higher O2 pressures displayed high H2 generation rate of 4590 and 5010 ml min−1 gcat−1 for samples prepared using O2 pressures of 4.5 x 10−2 and 8 x 10−2 mbar respectively and the coatings deposited at lower O2 pressures showed only 250 ml min−1 gcat−1.

Figure 5.6:

Maximum H2 generation rate, obtained hydrolysis of alkaline NaBH4 (0.025 M), as a function of dierent O2 pressures used for the deposition of Co3 O4 NPACs by PLD.

XRD pattern in Figure 5.7 reveals major dierences in the structural properties 65


of the coatings deposited at dierent O2 pressures. NPACs deposited at high O2 pressures (4.5 x 10−2 and 8 x 10−2 mbar) show peaks at 31.2º, 36.9º, 39.6º, 44.8º and 54.5º assigned to the face-centered cubic (fcc) spinel structure of Co3O4. Whereas low O2 pressure (3 x 10−3 and 8 x 10−3 mbar) deposited coatings displayed three dierent peaks at 36.5º, 42.4º and 61.5º corresponding to the CoO cubic phase. These results suggest that lack of oxygen during the deposition creates dierent phases in the coating. In order to further conrm this phenomenon, the chemical states of cobalt in these samples were studied using XPS. The Co 2p3/2 core levels of NPACs deposited at lowest (3 x 10−3 mbar) and highest (8 x 10−2 mbar) pressures are presented in Figure 5.8a and 5.8b, respectively. XPS spectra of coating deposited with 4.5 x 10−3 mbar are reported in Figure 5.4 while that deposited at 8 x 10−3 mbar pressure showed similar spectra as that of coating deposited at 3 x 10−3 mbar. The main peak in the spectra of NPAC deposited at 8 x 10−2 mbar (Figure 5.8b) is de-convoluted into two peaks centered at 779.6 and 780.7 eV attributed to Co3+ 2p3/2 and Co2+ 2p3/2 corresponding to Co3O4 phase.

Figure 5.7: XRD pattern of Co O 3

4

NPACs deposited by PLD using dierent O2 pressures.

In the case of low pressure deposited NPACs (Figure 5.8a), the peaks due to Co3+ disappear while another peaks at higher BE of 782.1 eV assigned to Co2+ state in Co(OH)2 phase is visible. The peak at 780.4 eV corresponding to Co2+ oxidation state is maintained in this coating which indicates the presence 66


of CoO phase. The dierence in CoO and Co3O4 phase in XPS spectra can be distinguished on the basis of satellite features of Co 2p spectra where Co2+ state in CoO phase shows observable satellite features at 786 eV. On the other hand, Co3O4 containing mixed oxidation state of Co2+ and Co3+ display very small satellite features due to these states. Indeed, this feature is observed in our sample with the satellite peak clearly visible for the NPACs deposited at low pressure whereas the satellite feature is marginal for high pressure deposition of NPACs. This nding matches well with the XRD results. At high O2 pressures, the ablated elements from the Co target lose most of their energy due to large number of multiple collisions between the ablated species and O2 in the chamber. This provides more time for interactions between the Co species and O2 to form stoichiometric Co3O4 phase on the substrate.

Figure 5.8:

XPS spectra of Co 2p3/2 core level of Co3 O4 NPACs deposited by PLD using O2 pressures of (a) 3 x 10−3 mbar and (b) 8 x 10−2 mbar.

On the other hand, the Co elements ablated at lower O2 pressure are expected to arrive at the substrate without signicant loss of energy due to the limited interaction between the ablated materials and O2: in this case the CoO phase formation is more favorable [144, 149]. It is clearly observed that the formation of these oxides depend upon the growth conditions/reaction kinetics such as the availability of O2 in the chamber and the energetics required for the oxidation process. Considering the above results, the Co3O4 phase possessing higher oxidation 67


state (Co3+ and Co2+) is easily reduced to form active Co-B as compared to CoO that only exhibits the Co2+ state. The surface morphologies of the coatings deposited with dierent O2 pressures was studied by SEM and are reported in Figure 5.9 (SEM image of coating surface deposited at 4.5 x 10−2 mbar is shown in Figure 5.2b). At lowest O2 pressures (3 x 10−3 and 8 x 10−3 mbar), the coating surface is mostly continuous, smooth and dense (Figure 5.9a and 5.9b). Owing to the phase explosion phenomenon, larger nanoparticles of size above 100 nm are also observed on the surface. As the O2 pressure increases (4.5 x 10−2 and 8 x 10−2 mbar), small NPs of size between 3 10 nm appears on surface of the coating, while decrease in the density and size of bigger droplets was observed from SEM images (Figure 5.2b and 5.9c).

Figure 5.9: 10

−3

SEM images of Co3 O4 NPACs deposited by PLD using O2 pressures of (a) 3 x mbar, (b) 8 x 10−3 mbar and (c) 8 x 10−2 mbar.

O2 pressure causes the cooling down of highly energetic Cobalt ions or atoms by collision eects [150]. Collisional processes with O2 also leads to the condensation of liquid nanodroplets and also of vapor atoms into small Co NPs in oxidized form during ight and before reaching the substrate. These NPs later get deposited on the substrate to form nanoparticle assembled coatings. The morphology obtained above certainly contributes to the observed variation in the catalytic activity of the Co3O4 coating deposited at dierent O2 pressure. Coating deposited with lowest O2 pressure presents low surface area owing to its at and smooth surface. On the other hand, well organized NPs of smaller size and narrow distribution on the surface, obtained at high O2 pressure deposition, provide large number of active sites due to high eective surface area. However, it is not clear if the morphology or the structural variation is more eective in improving the catalytic activity of Co3O4 coating deposited at high O2 pressure. In order to put more light on this point, two sets of coatings were synthesized. In the rst set, the coatings were deposited at dierent laser uences (1, 3, 5 and 7 J/cm2) by keeping O2 pressure constant at 4.5 x 10−2 mbar and substrate temperature at 250 ºC. During the second set, the laser uence and O2 pressure 68


were maintained at 3 J/cm2 and 4.5 x 10−2 mbar while the substrate temperature was varied (25º, 200º, 250º and 300º C). In the former case, only the coating morphology is expected to be altered while the degree of crystallization might be varied with the latter condition.

5.5 Eect of variation in laser uence XRD pattern and Raman spectra of the coatings deposited at four dierent laser uences (1, 3, 5 and 7 J/cm2) showed similar characteristic of broad and low intensity peaks of only Co3O4 phase with nanocrystalline structure.

Figure 5.10:

SEM images of Co3 O4 NPACs deposited by PLD using laser uence of (a) 1 J/cm2 , (b) 5 J/cm2 and (c) 7 J/cm2 .

Figure 5.11:

Maximum H2 generation rate, obtained by hydrolysis of alkaline NaBH4 (0.025 M), as a function of dierent laser uences used for the deposition of Co3 O4 NPACs by PLD.

SEM images (Figure 5.10 and Figure 5.2b) shows that the coatings deposited at lowest uence (1 J/cm2) displays uniform and smooth surface with the presence 69


of big droplets. The surface of coatings deposited at other energies showed NPAC morphology. The eect of morphological variations realized by the dierent laser uences (1-7 J/cm2) was studied by measuring H2 generation rate (Figure 5.11) during hydrolysis of NaBH4. For all the laser uences, the coating showed almost similar HGR (∼4500 ml min−1 gcat−1) except for that deposited at the lowest laser uence of 1 J/cm2 (∼2720 ml min−1 gcat−1), which is lower by about 43% as compared to that of other coatings. The observed H2 generation rate decrease is mainly attributed to the surface morphology where coatings deposited at higher laser uences exhibit NPs assembled on the surface which provide high surface area and large number of active sites for favorable interaction and conversion of NaBH4.

5.6 Eect of variation in substrate temperature No major alteration was observed in the morphology of the coatings deposited at dierent substrate temperatures as conrmed by SEM (similar to Figure 5.2b). Nevertheless, the degree of crystallization was certainly dierent as indicated from the Raman spectra (Figure 5.12).

Figure 5.12: Raman spectra of Co O

NPACs deposited by PLD using substrate temperature of (a) RT, (b) 200 ºC, (c) 250 ºC and (b) 300 ºC. 3

4

All the modes of vibrations corresponding to Co3O4 phase are present when substrate temperature is applied. However, the Raman spectra of Co3O4 NPs prepared at room temperature (RT) shows only three peaks at 195, 485 and 693 cm−1 70


with low intensity and broad nature. The full width half maxima of Raman band A1g at 693 cm−1 decreases with increasing substrate temperature thus indicating an increase in the crystallinity of the Co3O4 NPs, while partial crystallization is obtained in the case of RT deposition. The structural information obtained by performing XRD (reported in a previous work [141]) were in good agreement with the present Raman spectroscopy results. During the catalytic hydrolysis of NaBH4, though all the coatings deposited at dierent substrate temperatures were able to produce expected H2 yield, the major dierence in the HGR was observed and presented in Figure 5.13. The HGR increases with the increase in substrate temperature reaching a maximum for 250 ºC and then decreases at highest temperature of 300ºC. The coating deposited at RT just shows maximum HGR of ∼1800 ml min−1 gcat −1 which is ∼60% lower than that obtained for the sample prepared at 250 ºC. Even though the NPs deposited at RT have similar size distribution than that deposited at 250 ºC, these NPs are mainly in amorphous phase. The increase of substrate temperature allows the NPs to crystallize. Highest HGR for the coating deposited at 250 ºC is mainly attributed to the mixed amorphousnanocrystalline phase obtained at this temperature as observed by Raman spectra (Figure 5.12) and XRD pattern [141].

Figure 5.13:

Maximum H2 generation rate, obtained by hydrolysis of alkaline NaBH4 (0.025 M), as a function of dierent substrate temperature used for the deposition of Co3 O4 NPACs by PLD.

The amorphous and nanocrystalline mixed phases are separated by grain boundaries which are linear defects in the nanoparticle; the grain boundaries have a width of 0.5 - 1 nm containing atoms with lower co-ordination number as compared to the atoms in ideal crystallites. Therefore, these grain boundary regions 71


are catalytically highly active sites with basic, acidic or redox functionality to generate higher HGR for the coating deposited at 250 ºC, higher than that of well-crystallized coating obtained at 300 ºC. By varying the laser uence and substrate temperature, it is now clear that both morphology as well as structure contribute in generating higher catalytic activity for the lm deposited at higher O2 pressures.

5.7 Recycling test for NPACs

Co3O4 NPAC, deposited at 250 °C with laser uence of 3J/cm2 and O2 pressure of 4.5 x 10−2 mbar, was recycled for 4 times for hydrolysis of NaBH4, in order to check the stability of the NPs on the catalyst surface. After each cycle, catalyst coating was washed gently with distilled water. Hydrogen generation yield, as a function of time, for a number of runs is reported in Figure 5.14. In 2nd cycle, the catalytic activity decreases by about 20% as compared to the rst cycle involving as prepared catalyst.

Figure 5.14: Reusability behavior of Co O

NPACs on hydrogen generation yield, as a function of reaction time, measured during hydrolysis of 0.025 M NaBH4 alkaline solution. 3

4

No major variation in the H2 production rate was observed in the subsequent cycles. This shows that during cycling, the NPs remained immobilized and strongly held on the catalyst surface. The initial decrement in the activity can be attributed to the loss of some NPs coated on the surface. 72


5.8 Conclusion Nanoparticles (NPs) of Cobalt oxide (Co3O4 and CoO) with average size of ∼5 nm, randomly arranged in the form of coating, with narrow size distribution (3 - 10 nm) and low degree of agglomeration, have been produced by Pulsed Laser Deposition using pure Co as target in oxygen atmosphere. Co3O4 crystalline powder were also chemically synthesized to make comparison of the H2 generation rate of the powder with that of the NPs deposited by PLD. With optimized choice of the PLD parameters, the maximum H2 generation rate obtained by Co3O4 coating (5010 ml min−1 gcat−1) was about 5 times higher than that produced by Co3O4 powder (1000 ml min−1 gcat−1). In addition, by making comparison between Co3O4 and CoO NPs, synthesized by PLD, it was found that the Co3O4 phase is more active for hydrolysis than CoO phase and the result was ascribed to the lower oxidation number of CoO phase. This was established by XPS analysis of the CoO and Co3O4 phases. Thus, Co3O4 was easily reduced to form active Co-B (a step in hydrolysis process of NaBH4) as compared to CoO that only exhibits the Co2+ state. The eect of morphology and crystallinity of Co3O4 NPs on H2 generation rate was studied with NPs obtained by tuning the laser uence and substrate temperature respectively. The results presented in this chapter showed that the NPs with mixed amorphous-nanocrystalline phase on the surface present the best active sites for favorable interaction and conversion of NaBH4. These results comprehensively establish that catalysts in the form of NPACs are highly suitable for industrial scale applications.

73

Chapter 6 Cobalt-Boride: An ecient and robust electrocatalyst for Hydrogen Evolution Reaction 6.1 Introduction Water electrolysis is an old-known process to produce pure hydrogen (and oxygen). However, today it contributes to only 4% of the total world hydrogen production. This is majorly due to the energy required to split water into hydrogen and oxygen. Also, the usage of noble metals (Pt, Pd, Ru) [151, 152] as electrocatalysts adds to the overall cost of H2 production. It is therefore essential to search for electrocatalysts made up of inexpensive and abundant materials that can imitate the eciencies of noble group elements and also withstand dierent pH conditions. In the past few years, there has been a plethora of earth-abundant electrocatalytic materials for hydrogen evolution with excellent eciency and stability [153, 154, 155, 156, 157, 158]. A lot of recent research is focused on using transition metals (Fe, Co, Ni, Mo, etc.) coupled with non-metals (P, N, C) and chalcogenides (S, Se) such as CoS [153], MoS [154], CoSe2 [155], Mo2C [156], MoP [157], Co2P [158] and so on. These metal/non-metal or metal/chalcogenide compounds have turned out to be highly ecient catalysts for hydrogen evolution reaction (HER) with many of them being active in multiple pH environments as well. However, there have been very few reports on transition metal borides as HER active material. About two decades ago, Lasia and Los [159] reported amorphous Ni2B electrocatalyst for HER in alkaline medium. Subsequently, there were more reports on electrodeposited Ni2B [160] and doped Ni2B [161] catalysts for alkaline water electrolysis. Since then, the interest in metal boride electrocatalysts almost vanished until Vrubel and Hu reported MoB [162] as HER active material 74

Chapter 6: Co-B electrocatalyst for hydrogen evolution reaction

in both acidic and basic conditions. Motivated by these reports and also on our past experience in replacing noble metal catalyst for H2 production by hydrolysis of chemical hydrides [163, 164], we reported, for the rst time, Cobalt-Boride (Co-B) as an electrocatalyst material for HER. In this chapter, we present the outstanding HER activity that was recorded with Co-B catalyst in wide range of pH values. Electronic transfer between Co and B possibly favors charge conduction during water reduction along with high resistance against deactivation and provides robustness to the catalyst. Co-B nanoparticles (NPs) were synthesized by chemical reduction of aqueous cobalt chloride by sodium borohydride, under continuous stirring. As bubble generation ceased, the black precipitate in the solution was ltered and extensively washed with double distilled water and ethanol to remove any traces of unreacted and unwanted ions. The black powder obtained after cleaning was dried in vacuum at room temperature. The powder thus obtained was pressed in a conventional hydraulic press at 7 tons of pressure to obtain disc-shaped pellets of diameter 18 mm and thickness 1 mm.

6.2 Characterization of Co-B catalyst Structural analysis: SEM (Figure 6.1a) and bright eld TEM image (Figure 6.1b) of Co-B catalyst shows particle-like morphology with spherical shape and size in the range of 30 - 50 nm. The amorphous nature of the powders with long range disorder and short range order was conrmed through XRD pattern (Figure 6.1c) and HRTEM image (Figure 6.1d), which was also veried from the SAED pattern (inset of Figure 6.1d). The broad peak centered at 45º in XRD pattern (Figure 6.1c) is assigned to the amorphous state of Co-B phase [165]. Elemental analysis: Elemental mapping with EDS (Figure 6.2) did not display any phase separation of Co and B hence suggesting that both the elements in the catalyst are well mixed at macroscopic level. Investigation of chemical states of each element in the catalyst was carried out using XPS (Figure 6.3a and b). It shows the presence of two peaks in Co 2p3/2 level with binding energies at 778.2 and 781.6 eV and a satellite peak at 785 eV indicating that Co metal exists in both elemental and oxidized state [Co(OH)2] (Figure 6.3a). Similar states were also detected in B1s level with peaks at binding energies of 188.2 and 192.1 eV respectively (Figure 75


6.3b). Most importantly, when compared to binding energy of pure B (187.1 eV), the elemental B in the catalyst is positively shifted by 1.1 eV suggesting an electron transfer from alloying B to vacant d-orbital of metallic Co making former electron decient while later enriched with electron as indicated by small negative shift (0.2 eV) in the Co elemental peak.

Figure 6.1: (a) SEM image, (b) Bright eld TEM image, (c) XRD pattern, (d) HRTEM image with inset showing SAED pattern of Co-B catalyst.

Figure 6.2:

Macroscopic elemental mapping of Co-B catalyst using Energy Dispersive Spectroscopy analysis.

These results suggest that the surface of Co-B catalyst is composed of electron enriched Co sites bonded with B, and of cobalt hydroxide [Co(OH)2] in ratio of 76


1:1.38 as inferred from the peak area. Similar oxide species are also noticed by Sun et al. (Co-S catalyst) [153] and Cobo et al. (Janus cobalt catalyst) [166]. The electronic transfer from B to Co allows B to act as the sacricial agent in order to partially protect Co from oxidation. The surface elemental atomic ratio of Co/B is 1.61, signifying that mixed composition of Co-B and oxidized cobalt, mainly Co(OH)2, is present on the surface.

Figure 6.3: X-ray Photoelectron Spectra of Co 2p and B 1s level of Co-B catalyst.

Computational analysis: To clarify the electron interaction between Co and B, rst principles study of charge transfer in Co-B alloy in their crystalline and amorphous forms was performed using Dmol3 module [167] of Materials Studio package based on the density functional theory. For the crystalline form, Co-B and Co2B were considered in orthorhombic (space group Pbnm) and tetragonal (space group I4/mcm) structures respectively, while for the amorphous phase, clusters containing 14 and 22 atoms for Co-B and Co2B were considered, respectively. From the Mulliken's charge transfer analysis, electrons were transferred from Co to B atom for both Co-B and Co2B in crystalline form which shows high electronegativity of B atom in Co-B alloys. On the other hand, in the case of Co-B amorphous clusters, reverse electron transfer was observed from B to Co atom making Co more electronegative in a disordered Co-B arrangement. In the present case, the synthesized Co-B alloy is in amorphous state, thus the above theoretical result conrms the XPS experimental nding of electron transfer from B to Co. These Co atom sites enriched with electron are highly active for catalytic reaction, as also suggested in the case of hydrogenation [168, 169] and hydrolysis reactions [163, 164]. Recently Carenco et al. [170] distinguished the electron transfer in transition metal (Co, Ni, and Fe) boride where electron transfer occurs from M to B in boron-rich systems (MBx, x ≥ 2) and from B to M for metal-rich borides (MBx, x ≤ 2). This nding of electronically enriched metal was also conrmed on the basis of magnetic 77


and Mossbauer measurements and metal binding energy shifts observed in X-ray photoelectron spectroscopy [171, 172]. In all the cases, the metal-boride was in amorphous state. However, most of the rst principle calculations [173, 174] show that the total electron transfer should occur from metal to boron in relation with relative electronegativities but these studies were carried out for crystalline Co-B.

6.3 Electrochemical measurements The catalyst powder was cold-pressed in the form of pellets and employed as working electrode in a typical three-electrode electrochemical cell. Water with neutral pH provides an ideal scene to produce H2 for fuel-cell in homes; therefore the synthesized catalysts were rst tested in potassium phosphate buer solution with pH 7. The linear polarization curves of Co-B catalyst along with that of Co and Pt metal are presented in Figure 6.4. The HER activity achieved by Co-B is signicantly higher than that obtained with Co metal and the onset overpotential observed (vs RHE), was as low as 70 mV (at 0.1 mA/cm2). Beyond this value, rapid rise in the cathodic current was induced at further negative potentials. The overpotential (η) of only 178 mV and 251 mV was required to attain moderate current densities of 2 mA/cm2 and 10 mA/cm2, respectively, where substantial H2 evolution occurs. These current densities were noted at 50 mV and 176 mV using Pt metal as working electrode. The overpotential required to achieve these current densities is considerably higher for most of the earth abundant element-based molecular and solid-state catalysts such as Janus cobalt catalyst [166], Cu2MoS4 [175], CoP4N2 [176] and NiWS [177] catalyst at neutral pH (Appendix D). The linear tting of Tafel plot (inset of Figure 6.4) gives a Tafel slope value of 75 mV/dec and exchange current density (J0) of 0.25 mA/cm2 for Co-B nanocatalyst. On the other hand, unfavorable values of these parameters were observed for Co metal (7.0 x 10−4 mA/cm2 and 133 mV/dec) (Table 6.1). The enrichment of d-band electron density, on Co sites of the Co-B catalyst, improves the electrondonating ability to enhance HER catalytic activity and therefore generate high current densities at lower overpotentials than that obtained by Co metal. Similar characteristic was reported for Ni-Mo-Nx alloy [178] where the presence of N modies the d-band electron density of Ni-Mo alloy to favor reaction kinetics. The obtained Tafel slope value is comparable to the values reported for Co-S [153], M-MoS3 (M = Fe, Co or Ni) [179], and Cu2MoS4 [175] HER catalyst at pH 7, but this value is nowhere near to the standard values for HER reaction steps that are: 120, 40, and 30 mV/dec for Volmer, Heyrovsky and Tafel mechanism, respectively. Thus, here it becomes dicult to establish the reaction mechanism; but the value of 75 mV/dec may suggest the Volmer-Heyrovsky route for HER [153, 175, 179]. 78


Figure 6.4:

Linear polarization curves with iR correction for Pt, Co-B, and Co in 0.5 M potassium phosphate buer with pH 7 obtained at a scan rate of 10 mV/s. Inset shows the corresponding Tafel plot for the Co-B catalyst.

The value of Tafel slope which is considerably lower than that of Volmer step (∼120 mV/dec) hints that the process of H+ ion adsorption is not the rate-limiting step. This adsorption process is conceivable because of the presence of electron enriched Co sites on the catalyst surface which can easily assist to reduce H2O and adsorb H according to the Volmer step (H+ + e− Hads). On the other hand, after the coverage of these Co active sites by Hads, unavailability of electron hinders desorption of H2 molecules by Heyrovsky step (Hads + H+ + e− H2). HER activity per site of a catalyst is investigated by evaluating turnover frequency (TOF) using BET surface area (16.42 m2/g) of Co-B catalyst (procedure reported in Appendix C). At η = 250 mV for exchange current density of 10 mA/cm2, TOF was calculated to be 0.088 s−1 (Table 6.2). This value is underestimated because the actual active Co sites on the surface (containing also boron) are not known. The TOF value is higher than that reported for Co-S (0.017 s−1) [153] and Janus cobalt (0.022 s−1) [166] at neutral pH. 79


Table 6.1: Electrochemical parameters obtained with Co-B, and Co metal for HER in aqueous solution of various pH values.

Table 6.2: Turnover Frequency of Co-B catalyst measured in aqueous solution of dierent pH values at certain overpotential.

6.4 Activity in dierent pH media The HER catalyst should be active and stable in any reaction medium even though the neutral medium is preferable for environmental purpose. Thus, the HER 80


activity of Co-B catalyst was investigated in aqueous solutions of dierent pH values, namely: 1, 4.4 and 9.2. These values were selected on the basis that sea water and rain water are the main sources of water which are relatively more basic (pH 8 - 9.5) and acidic (pH 4.2 - 6) in nature, respectively. For extreme pH value, strong acid was used while phosphate buer solution was tuned to establish the remaining pH values. As observed in linear polarization curves, CoB catalyst displayed extremely high HER activity for all pH values in contrast to Co metal (Figure 6.5a - c). The plot (Figure 6.5d) of overpotential (at 2 mA/cm2) and exchange current density values as a function of pH shows that pH 9.2 is the most favorable medium for Co-B catalyst (Table 6.1). Lowering the pH of the solution leads to a decrease in HER activity with lowest electrochemical parameters recorded in acidic medium (pH 1 and pH 4.4). The values of Jo = 5 x 10−2 mA/cm2 and η = 216 mV at 2 mA/cm2 in acidic solution are comparable to Ni and Mo based non-noble HER catalysts including Ni2P [180], Mo2C/CNT [181], MoN [178], MoB [162], and Mo2C [162].

Figure 6.5: Linear polarization curves with iR correction for Co-B catalyst compared with Co

metal in (a) pH 1 (0.1 M HClO4 ), (b) pH 4.4 (0.5 M KH2 PO4 ) and (c) pH 9.2 (0.4 M K2 HPO4 ) obtained with scan rate of 10 mV/s. (d) Plot of overpotential (at 2 mA/cm2 ) and exchange current density values as a function of pH values of the solution used to test the Co-B catalyst.

81


Water electrolysis in acidic medium proceeds by the formation of hydronium ions (H3O+) which later discharges on the catalyst surface for adsorption of H [182]. Higher Tafel slopes of 108 and 102 mV/dec in low pH (4.4 and 1 respectively) solutions indicate that the Volmer step for adsorption of H ion is the rate limiting step, which explains the poor adsorption of H3O+ ions in the acidic media, by the Co active sites. At higher pH (9.2), on the other hand, neutral water molecules are easily reduced on the Co active sites by hydrogen adsorption while OH− ions provide the necessary ion-conduction in the solution to deliver high current density at low overpotential. These results indicate that Co-B electrocatalyst is highly active in wide pH range.

Figure 6.6: Plot of charge build-up versus time for Co-B catalyst acquired at (a) pH 7 (b) pH 9.2 and (c) pH 4.4 at constant overpotentials. Inset of each plot shows the variation in current density over a long period of time.

82


Figure 6.7:

Recycling behavior of Co-B catalyst examined in aqueous solutions of (a) pH 7 (b) pH 9.2 and (c) pH 4.4 at a scan rate of 150 mV/s.

6.5 Stability and Reusability Furthermore, the stability of Co-B catalyst was investigated by measuring the current density under constant overpotential vs RHE as a function of time (Figure 6.6a). Under neutral pH, the current density varies by small value in 44 h while maintaining overpotential of 250 mV. Thus, over that time period, a linear buildup of charge is registered. Durability test conducted on pH values of 9.2 and 4.4 (Figure 6.6 b - c) demonstrated similar charge accumulation over the long period of time (44 h). To test the ability to withstand industrial workload, Co-B catalyst was cycled 1000 times in the potential range between 0 and -0.5 V vs RHE at pH 7 in phosphate buer (Figure 6.7a). After 1000 cycles, HER activity of Co-B catalyst remained unchanged for pH 7 while changed marginally for pH values of 9.2 and 4.4 (Figure 6.7 b - c). Under pH 1, after 10 cycles, the visible dissolution of Co-B 83


powder was monitored. Nevertheless, these results illustrate the robust nature of the Co-B catalyst under wide range (pH 4 - 9) of environmental condition.

6.6 Conclusion In summary, we introduced Co-B amorphous nanoparticles (30 - 50 nm size) catalyst, synthesized by facile method, as a possible substitute of noble metal based electrocatalyst for HER. In water with neutral pH, an overpotential of only 251 mV was required to attain a current density of 10 mA/cm2. The present overpotential, with related current density, is considerably lower than most of the previously reported results of the earth abundant element-based molecular and solid-state catalysts, with favorable values of Tafel slope of 75 mV/dec, and exchange current density of 0.25 mA/cm2 at neutral pH. Highly active Co surface sites, created by electronic transfer from B to Co, are responsible for the remarkable HER activity and robust nature in wide range of pH (4 - 9) values. Under neutral pH, the current density varies by small value during 44 h test while maintaining overpotential of 250 mV: this proves the stability of the Co-B catalyst. Finally, after 1000 cycles, the HER activity of Co-B at pH 7 remained unchanged.

84

Chapter 7 Co-Ni-B nanocatalyst for ecient hydrogen evolution reaction in wide pH range 7.1 Introduction In the last chapter, we discussed the viability of Co-B catalyst for electrochemical water splitting. However, there was still scope of improvement in the electrocatalytic performance of Co-B if it were to compete with noble metal catalysts. From the past literature [183], it was observed that inclusion of other transition elements (Ni, Mo, Cu, etc.) in Co-B leads to an increment in the catalytic activity for hydrolysis reactions. Following the same trend, we developed Co-Ni-B catalyst and tested it for electrochemical water splitting. As expected, an improvement in the electrocatalytic performance was observed when compared to Co-B catalyst. Co-Ni-B was also found to be stable almost across the entire pH range. In this chapter, we present Co-Ni-B nanocatalyst to be highly active for hydrogen evolution reaction (HER) in various pH media. It turns out to be an excellent alternative to noble metal electrocatalysts in the sense that it comprises all low-cost, earth-abundant, non-noble and non-toxic elements. Co-Ni-B catalyst powder was synthesized by reducing aqueous mixture of cobalt chloride and nickel chloride in presence of a strong reducing agent like sodium borohydride which also acts as the boron source. The molar ratio of metal to NaBH4 was taken as 1:3 to ensure complete reduction of Co and Ni ions. The proportion of Ni in Co-Ni-B was varied by adjusting the molar ratio χN i = Ni/(Ni+Co) in the starting aqueous mixture (from 10% to 50%). Co-Ni-B with molar ratio (χN i) of 30% (Co-30Ni-B) worked exceedingly well under benign neutral conditions with H2 onset potential of just 53 mV and a Tafel slope of 51 mV/decade. Co-30Ni-B outperformed Co-B 85

Chapter 7: Co-Ni-B electrocatalyst for hydrogen evolution reaction

[184] and many other cobalt-based electrocatalysts because of the inclusion of Ni in Co-B leading to a higher electron density at Co active sites thereby facilitating the HER process.

7.2 Characterization of Co-Ni-B catalyst Structural and elemental analysis: The reduction reaction of metal salts (Co and Ni) by sodium borohydride leads to the formation of well-dispersed spherical nanoparticles of narrow size distribution as observed in SEM image of Co-30Ni-B (Figure 7.1a).

Figure 7.1: (a) High resolution SEM image of Co-30Ni-B; (b) TEM image showing the average

size of Co-30Ni-B particles around 25 30 nm; (c) High resolution TEM image, SAED pattern (inset of c) and (d) XRD pattern showing the presence of amorphous state in Co-30Ni-B.

The average size of Co-30Ni-B NPs was determined to be around 27 ± 2 nm, acquired from the TEM image shown in Figure 7.1b and having a surface area of 16.8 ± 0.2 m2/g obtained by BET analysis which is similar to that of Co-B NPs (16.4 ± 0.2 m2/g). The vigorous nature of the synthesis reduction reaction 86


doesn't allow the formation of well-ordered structures and thus the resultant NPs are amorphous in nature. The absence of long-range ordering in Co-30Ni-B was conrmed from high resolution TEM micrograph (Figure 7.1c) which shows no signature of any crystalline phase; the result is also supported by the SAED pattern (inset of 7.1c) showing diused rings. XRD pattern (Figure 7.1d) of Co-30Ni-B shows a single broad peak centered at 2Θ = 45º conrming the amorphous nature of the catalyst powder formed. The molar ratio of Co and Ni in Co-Ni-B catalyst was conrmed by XRF (Table 7.1).

Table 7.1: Percentage of Co and Ni content in dierent Co-Ni-B catalyst determined by X-ray

uorescence measurement.

XPS studies: XPS spectra of Co-30Ni-B catalyst (Figure 7.2 (a-c)) display two peaks in both Co 2p3/2 and Ni 2p3/2 levels with binding energies of 774.45 and 780.45 eV for Co, and 852.2 and 855.5 eV for Ni indicating that the two metals are present in both elemental and oxidized states respectively. Apart from the two peaks stated above, an additional shake up satellite peak was also observed at BE of 784.45 eV for Co and 860.63 eV for Ni in 2p3/2 core level. These characteristic peaks are also present in corresponding 2p1/2 core levels of both metals. Similarly, two peaks were also observed for B 1s state with BE of 187.6 and 191.5 eV indicating the presence of elemental and oxidized boron respectively. The presence of surface oxidized states in all the samples can be attributed to the exposure of the catalyst samples to ambient atmosphere during the measurements. To make a comparative study, XPS analysis was also carried out for Co-50Ni-B (Figure 7.2 (d-f)) and Ni-B catalysts (Figure 7.2(g-h)). From the XPS results discussed in previous chapter, we know that in Co-B catalyst, the elemental boron peak (187.8 eV) is positively shifted by about 0.8 eV as compared to the BE of pure B (187.0 eV). 87


Figure 7.2:

XPS spectra of Co 2p, Ni 2p and B 1s states in Co-30Ni-B (a-c) and Co-50Ni-B (d-f); Ni 2p (g) and B 1s (h) states in Ni-B.

This shift results from an electron transfer from the alloying B to the vacant dorbital of metallic Co, thus making the B atom electron decient and the Co atom electron enriched. Since B is more electronegative than Co, this unusual electronic behavior in amorphous Co-B was also conrmed by computational studies in previous chapter. A similar phenomenon was also observed in the cases of Co-30Ni-B, Co-50Ni-B and Ni-B catalysts where B 1s peak was again found to be positively shifted (187.6 eV) by 0.6 eV as compared to that of pure B (187.0 eV) indicating an electron transfer from B to vacant d-orbitals of Co (or Ni in Ni-B), correlating well with the trend shown by amorphous metal borides [185, 186]. However, in both the Co-Ni-B catalysts, a negative shift in the BE peak of elemental Co (777.45 eV for Co- 30Ni-B and 777.5 eV for Co-50Ni-B) by 0.45 0.5 eV was observed which was not detected in Co-B catalyst. This shift in BE indicates the presence 88


of higher electron density on cobalt sites in the cases of both Co-Ni-B (χN i = 30 % and 50%) catalysts as compared to Co-B. One of the crucial requirements for a good HER catalyst is the presence of high electron density at the catalytically active sites which can facilitate reduction of water molecule more eciently [187]. From the above XPS results, it is certain that Ni plays a key role in increasing the electron density at Co sites to make them more active for HER. Also, on analyzing the surface composition from XPS data (Table 7.2), it is revealed that Co-30Ni-B shows boron enrichment on the surface which implies more electrons are available to be transferred from the surface of B atom to Co active sites as compared to Co-B catalyst.

Table 7.2:

BE peak positions and percentage surface composition for all catalysts obtained from XPS analysis.

X-Ray Absorption spectroscopy (XAS) studies: X-ray absorption spectroscopy (XAS) was also carried out to determine more details about the local geometric and electronic structure of the catalysts. The XANES spectra of Co-30Ni-B and Co-B measured at Co K edge are shown in Figure 7.3a along with that of Co metal foil and commercial CoO powder, where Co is present in elemental and +2 oxidation states respectively. The line shapes obtained here for the Co-B and Co-30Ni-B samples agree well with that reported by Luo et al [188]. Three points are indicated in the Figure 7.3a as A, B and C in the edge parts of XANES spectra. In the region B, cobalt oxide shows sharp white line, while Co-B and Co-30Ni-B displayed broad absorption edge like metallic cobalt. Though the XANES spectra of Co-B and Co-30Ni-B samples resemble more that of the Co metal, the features in the regions A and C are quite dierent which manifests that Co does not exist in pure elemental form in these 89


samples.

Figure 7.3:

(a) XANES spectra of Co-30Ni-B is shown along with Co metal foil and CoO reference spectra; (b) normalized EXAFS spectra of Co-B and Co-30Ni-B at Co K edge; Fourier transformed EXAFS spectra of (c) Co-B and (d) Co-30Ni-B measured at Co K edge (Scatter points) and theoretical t (Solid line).

A linear combination t of the XANES spectra of the samples with that of pure Co and CoO suggests the presence of Co in metallic as well as in Co2+ state in the samples. The feature denoted by A is shifted to slightly lower energy side for Co-30Ni-B compared to that for Co-B sample which is due to the excess electron density at Co site in Co-30Ni-B compared to Co-B. This correlates well with the XPS observations further supporting our conclusion on the charge transfer on Co. To investigate it further, we have carried out EXAFS analysis on these samples. Figure 7.3c and 7.3d respectively show the normalized FT-EXAFS spectra of CoB and Co-30Ni-B catalysts measured at Co K edge. The structure and lattice parameters of Co2B have been used to generate the theoretical FT-EXAFS spectra [189]. The bond distances and disorder (Debye-Waller) factors (σv2), which give the mean square uctuations in the distances, have been used as tting parameters. The best t results are summarized in Table 7.3. σv2 values are found to be very high due to the amorphous nature of the sample as discussed in XRD and HRTEM results. 90


Table 7.3: Bond length, coordination number and disorder factor obtain by EXAFS tting. In the case of Co-B, the rst peak around 1.9 Å (phase uncorrected spectra) in Figure 7.3c has a contribution from rst coordination shell of four B atoms and second coordination shell of six Co atoms at distances of 2.04 Å and 2.48 Å respectively and these bond lengths are consistent with other studies [188, 190]. However, the coordination numbers obtained in the present case are quite dierent from the reported results. The χ(R) versus R plot for the Co-30Ni-B sample is similar to that of Co-B sample where the 1st peak has contributions from rst neighboring shell of four B atoms and second neighboring shell of six Co atoms with a slight decrease in bond lengths to 2.02 Å and 2.43 Å, respectively. However, the contributions in the 2nd peak are quite dierent for Co-B and Co-30Ni-B samples. For Co-B, the second peak in the FT-EXAFS spectra can be tted by considering only Co-B and Co-Co shells while the doublet in Co-30Ni-B second peak is properly tted only after taking into account a Co-Ni shell along with CoCo and Co-B shells. Most importantly, B coordination increases to 10 whereas Co coordination decreases to 3 for the Co-30Ni-B sample as compared to that of Co-B catalyst which has 6 B coordination and 6 Co coordination. This clearly suggests that the inclusion of Ni creates an atomic arrangement where Co is surrounded by higher number of B atoms than that of Co atoms which will contribute to the increased electron density on Co sites by electron transfer from boron. This is in good agreement with XPS results showing boron enrichment. 91


Computational studies: The results obtained from XPS and XAS studies complement the fact that the addition of Ni in Co-B enhances the electron density on cobalt sites. To corroborate these results, we also performed theoretical calculations using density functional theory (DFT) and found the total charge on each element in Co-B and Co-25Ni-B catalyst. DFT formalism was implemented to perform geometry optimisation of the Co2B nanocluster which displayed random arrangement of Co and B atoms (Figure 7.4 a-b).

Figure 7.4:

Nanoclusters of (a-b) Co-25Ni-B and (c) Co-B used for the calculation of charge

transfer:

From the present calculations, the charge transfer in pure Co-B nanocluster was estimated from B to Co wherein the average charge on Co was calculated to be -0.0067 e− and average charge on B atoms was +0.0195 e−. To check the eect of Ni doping, 25% Ni doping was considered in Co-B cluster at various doping sites (Figure 7.4c). After Ni doping, the average charge obtained on Co atoms was -0.0098 e− which is higher than that obtained with pure Co-B nanocluster. On the other hand, positive average charge is noted on Ni (+0.0121 e−) and B (+0.01431 e−) atoms. Thus, the addition of Ni increases the negative charge on Co atom in Co-25Ni-B cluster which again conrms the results obtained by XPS and XAS analysis.

7.3 Electrochemical measurements All the synthesized nanocatalysts were tested for electrocatalytic activity at various pH values by loading them on a polished glassy carbon (GC) electrode. Figure 7.5 shows the linear polarization curves in neutral media (pH = 7) for Co-Ni-B catalyst with dierent χN i values. As the Ni concentration is increased, the HER improves up to χN i= 30% and then decreases. Thus Co-30NiB catalyst which showed the optimum HER activity was chosen for further studies. 92


Figure 7.5: Linear polarization curves in pH 7 (0.5 M KPi solution) of Co-Ni-B with dierent

Ni concentrations varying from 10% Ni to 50% Ni. Inset shows the plot of overpotential at 10 mA/cm2 versus Ni concentration in Co-Ni-B catalyst.

Figure 7.6 shows the linear polarization curves for Co-30Ni-B, Co-B, Ni-B, Pt and bare GC electrode in pH 7 (0.5 M KPi) potassium phosphate buer solution at a scan rate of 5 mV/s.

Figure 7.6: Linear polarization curves for Pt electrode, Co-30Ni-B, Co-B and Ni-B electrocatalysts in pH 7 (0.5 M KPi solution)

93


The obtained current densities for catalyst modied GC electrodes were normalized to the geometric surface area of bare GC electrode. To achieve the benchmark current density of 10 mA/cm2 (ideal value for solar fuel synthesis) [191], Ni-B and Co-B require 309 mV (vs RHE) and 203 mV (vs RHE) respectively whereas Co-30Ni-B could achieve the same at just 170 mV (vs RHE). Here, it must be noted that the activity shown by Ni-B powder is much lower than that reported for Ni-Bx lms [192] which could be due to dierent synthesis techniques adopted. The above overpotential values clearly indicate that the ternary alloy Co-30Ni-B yields better HER activity than binary alloys of Co-B and Ni-B which can be attributed to the higher electron density created at the Co sites of Co-30Ni-B as evident from our ndings. However, Pt still remains the best catalyst requiring just 50 mV (vs RHE) of overpotential to reach 10 mA/cm2 of current density. Co-30Ni-B shows H2 onset potential value of just 53 mV (vs RHE) (overpotential to achieve 0.5 mA/cm2) which is extremely competitive with other non-noble electrocatalysts employed in similar neutral conditions (Appendix D). Co-30NiB could achieve even higher current densities of 20 mA/cm2 and 40 mA/cm2 at benign overpotentials of 213 mV and 251 mV respectively. The overpotential values reported here were one of the best results ever obtained for any metal boride electrocatalysts in pH 7 (Appendix D). The Tafel plot is an important tool in evaluating the performance of an electrocatalyst. It leads us to obtain two crucial electrochemical parameters Tafel slope (b) and exchange current density (i0). Tafel slope refers to the overpotential required to raise the current density by one order of magnitude. Figure 7.7 shows Tafel plots for Co-30Ni-B, Co-B, Ni-B and Pt. For Co-30Ni-B, a Tafel slope of 51mV/dec was obtained which is much lower than that of Co-B (71 mV/dec) and also lower than that reported for the best electrocatalysts operating in neutral conditions such as Co-S lm (93 mV/dec) [193], Co-P/CC NWs array (93 mV/dec) [194], FeP/CC (70 mV/dec) [195], Ni0.33Co0.67S2 (67.8 mV/dec) [196], CoS2 NWs (87.1 mV/dec) [196] and Co-P/Ti (58 mV/dec) [197]. It is well known that the value of Tafel slope is indicative of the reaction steps in HER. The rst step is the proton discharge step (Volmer step, 120 mV/dec) followed by electrochemical desorption (Heyrovsky step, 40 mV/dec) or chemical desorption (Tafel step, 30 mV/dec). A value of 51 mV/dec for Co-30Ni-B suggests Volmer-Heyrovsky reaction mechanism [198, 199]. From Tafel analysis, we could also determine the exchange current density (i0). The exchange current density for Co-30Ni-B in pH 7 was 0.708 mA/cm2 which again establishes its supremacy over Co-B catalyst (i0 = 0.501 mA/cm2). However, Pt shows the highest i0 value of 1 mA/cm2. Here, it must be noted that the exchange current density of Co30Ni-B is only lower than that of Ni0.33Co0.67S2 (0.893 mA/cm2), NiBx lms (0.851 94


mA/cm2) and CoS2 NWs (0.976 mA/cm2) and is superior to all other non-noble electrocatalysts in pH 7 (Appendix D).

Figure 7.7:

Tafel plots for Pt electrode, Co-30Ni-B, Co-B and Ni-B electrocatalysts in pH 7 (0.5 M KPi solution).

Figure 7.8:

TOFs calculated for Co-30Ni-B catalyst as a function of overpotential (vs RHE)

at pH 7.

95


Table 7.4: Comparison of TOF values with other reports from literature in pH 7. The intrinsic activity of each catalytic site can be estimated on the basis of turnover frequency (TOF). Following the method reported by Popczun et al. [199], TOF was determined at dierent overpotentials in pH 7 using BET surface area (Figure 7.8) values. Co-30Ni-B demonstrated a TOF of 0.145 atom−1s−1 at an overpotential of 250 mV (vs RHE). This value is underestimated as we have considered all the surface atoms as active which is not true. The real HER activity per active site would be even higher if actual number of active Co sites can be estimated. Even after this underestimation, the TOF value reported here is higher than that of H2 Co-cat (0.022 atom−1s−1 at 385 mV) [200] and Co-S lm (0.017 atom−1s−1) [193], all for pH 7 (Table 7.4)[201].

7.4 Activity in dierent pH media It is always desirable to obtain an electrocatalyst that can work equally well under dierent pH conditions. Microbial electrolyzers operate in neutral pH conditions which make them more benign. To be used in an alkaline electrolyzer, the catalyst needs to work eciently in severe basic conditions whereas PEM electrolyzers demand extreme acidic conditions. All these electrolyzers function on potable water (clean water) which is less than 1 % of the earth's water resources. It would be a big leap if electrolyzers could use water directly from natural resources such as rain and sea (unclean water). To achieve this, we would need electrocatalysts that can function eciently in mild acidic conditions of pH 4.2 - 6 relevant to rain water and mild basic conditions of pH 8.1 9.5 relevant to sea-water. With these notions in mind, we tested Co-30Ni-B catalyst in dierent pH media, specically, 1, 4.4, 7, 9.2 and 14. The corresponding polarization curves are shown in Figure 7.9a. We observe that Co-30Ni-B works considerably well in dierent pH conditions and the performance improves as we go to higher pH values. Table 7.5 quanties the performance of Co-30Ni-B in various pH media. Figure 7.9b shows the variation in overpotential required to achieve a current density of 2 mA/cm2 in various pH 96


conditions. The overpotential decreases from acidic to basic media with values of 209 mV, 170 mV, 83 mV, 60 mV and 45 mV (all vs RHE) for pH 1, 4.4, 7, 9.2 and 14 respectively signifying that Co-30Ni-B is capable of HER under dierent pH conditions with most suitable in basic media (Table 7.5).

Figure 7.9:

(a) Linear polarization curves for Co-30Ni-B in dierent pH media (from pH 1 to pH 14); (b) Plot of overpotential (at 2 mA/cm2 ) and exchange current density values as a function of pH of the solution for Co-30Ni-B catalyst.

Under acidic conditions (pH 1), the overpotentials required to achieve 5 mA/cm2 and 10 mA/cm2 of current density were 333 mV and 479 mV (vs RHE) respectively (Figure 7.10a). These values are higher but are still comparable to that of transition metal based HER catalysts including Co-B [184], Mo2C [202] and MoN/C [203].

Table 7.5:

Table showing the current density values at various overpotentials for Co-30Ni-B catalyst in dierent pH solutions.

97


Figure 7.10: Polarization curves of Co-30Ni-B, Pt electrode and bare glassy carbon electrode

in (a) 0.1 M HClO4 solution (pH 1) solution and (b) 1 M NaOH solution (pH 14) obtained at a scan rate of 5 mV s−1

On the other hand, in extreme basic conditions (pH 14), Co-30Ni-B could achieve the current densities of 5 and 10 mA/cm2 at just 93 and 133 mV (vs RHE) respectively and could go up to 100 mA/cm2 at just 233 mV (vs RHE), as evident from Figure 7.10b. These values are higher than that of many nonnoble HER catalysts such as Mo2C [204], MoB [204], CoNx/C [198], Co-NCNTs [205], Co-NPs@N-C [206], MoS2-CPs [207] and Fe2P/NGr [208] working in alkaline conditions. Exchange current density of 0.830 mA/cm2 and 0.073 mA/cm2 and Tafel slope of 121 mV/dec and 123 mV/dec was obtained from the Tafel plot (Figure 7.11) for Co-30Ni-B catalyst in pH 14 and pH 1 respectively.

Figure 7.11: Tafel plots for Co-30Ni-B in pH 1 and pH 7 98


The higher HER activity in Co-B is mainly attributed to the electron transfer from boron to d-band of Co [184]. These Co active sites with higher electron density improve the electron donating ability of Co, thereby promoting HER. In Co-Ni-B, the presence of Ni causes boron enrichment on the surface which further activates these Co sites by providing excess electrons as conrmed by XPS, EXAFS, XANES and DFT calculations. This is the main reason of higher HER activity obtained for Co-Ni-B as compared to Co-B having similar morphology and surface area.

Figure 7.12:

(a) Linear polarization curves of Co-30Ni-B before and after 1000 cycles in 0.5 M KPi (pH 7) and 1 M NaOH (pH 14) (at a scan rate of 100 mV/s); (b) Plot of charge build-up versus time and time-dependent current density curves (inset of b) at constant overpotentials of 110 mV in pH 7 and 120 mV in pH 14.

By electron transfer, boron also protects the Co sites from oxidation and deactivation, thus oering high stability in extreme conditions. Considering the HER rates of Ni-B and Co-B, it seems that Co is more active in borides than Ni. Thus, as the Ni concentration increases, the cobalt sites are further activated by electron density but after χN i= 30%, the Co sites are replaced by less active Ni sites causing reduction in HER activity.

7.5 Stability and Reusability To be used on an industrial scale, the electrocatalysts must be able to sustain long hours of operation and reuse. Hence, it becomes essential to test the catalysts for stability and long term durability. Based on the performance, we chose neutral and alkaline media for these tests. Figure 7.12a shows polarization curves for Co-30Ni-B catalyst cycled over 1000 times in pH 7 and pH 14. In both the pH conditions, we observe negligible loss in activity even after 1000 cycles of reuse 99


(less than 5% in pH 7 and no loss in pH 14). Co-30Ni-B was also subjected to continuous operation for over 45 hours at constant overpotentials of 110 mV (vs RHE) in pH 7 and 120 mV (vs RHE) in pH 14. From inset of Figure 7.12b, we see that in neutral pH, the current density falls marginally after 45 hours (maybe due to some atoms detachment from the surface) whereas it remains more or less the same in pH 14 leading to a linear charge build up over time (Figure 7.12b). These results illustrate the excellent stability and reusability of Co-30Ni-B catalyst in pH 7 and pH 14 making it ideal for industrial applications.

7.6 Eect of heat treatment To understand the eect of heat treatment on electrochemical performance, we annealed Co-30Ni-B catalyst at elevated temperatures of 200º, 300º and 400º C in vacuum. The annealed samples were also found to be amorphous as no crystallization was observed from XRD spectra which is consistent with that reported in literature [183].

Figure 7.13: Linear polarization curves of Co-30Ni-B annealed at various temperatures in 0.5

M KPi (pH 7).

The HER activity increased with the annealing temperature and was maximum for sample anealed at 400º C. Co-30Ni-B annealed at 400º C required just 122 mV and 166 mV (vs RHE) to achieve current densities of 10 mA/cm2 and 20 mA/cm2 respectively, in pH 7. The reason for this enhancement in HER activity is yet to be understood. At present, we speculate that partial crystallization of the catalyst upon annealing might create ample number of grain boundaries which are active catalytic centers, leading to the enhancement in activity. However, 100


detailed studies are required to probe the local structure and HER mechanism for annealed Co-30Ni-B catalyst.

7.7 Conclusion By a facile reduction method, we prepared Co-Ni-B powder nanocatalysts which are highly eective for HER in a wide pH range. The content of Ni was varied by tuning the molar ratio Ni/(Ni+Co) in Co-Ni-B from 10% to 50% to identify the most suitable composition (30 %) of Co-Ni-B for HER. High resolution TEM micrograph proved the absence of long-range ordering in Co-30Ni-B while XPS analysis of the Co-Ni-B catalysts showed a negative shift (0.45 eV) in the BE peak of elemental Co which indicates the presence of higher electron density on Co sites as compared to Co-B. By XPS, EXAFS, XANES and DFT calculations performed on random arrangement of atoms, it was proved that in Co-Ni-B, the presence of Ni causes B enrichment on the surface and also provides excess electrons to Co sites. The ternary alloy Co-30Ni-B yields better HER activity than binary alloys of Co-B and Ni-B at various pH values. The result is attributed to the higher electron density at the Co sites of Co-30Ni-B. Current densities of 20 mA/cm2 and 40 mA/cm2 at overpotentials of 213 mV and 251 mV, respectively, are observed which are one of the best results obtained for any metal boride electrocatalyst in pH 7. In addition, we observed that the overpotential decreases from acidic to basic media with values of 209 mV, 170 mV, 83 mV, 60 mV and 45 mV (all vs RHE) for pH 1, 4.4, 7, 9.2 and 14 respectively signifying that Co-30Ni-B is capable of HER under dierent pH conditions especially in basic media. Finally, negligible loss in activity was observed for Co-30Ni- B catalyst cycled over 1000 times in pH 7 and pH 14, making it an ideal material for industrial scale electrolysis.

101

Chapter 8 Co-Mo-B Nanoparticles as highly Ecient Bifunctional Electrocatalyst for Hydrogen and Oxygen Evolution 8.1 Introduction In the previous two chapters, we came across a number of transition-metal catalysts that show highly improved activity, for hydrogen evolution reaction (HER). In spite of these advances, none of these non-precious catalysts have been able to compete with noble group elements on site-specic activities [209]. Also, most of these catalysts face stability issues in alkaline media [209] whereas present industrial electrolyzers operate in alkaline conditions. Another major concern is that industrial alkaline electrolyzers operate at voltages of 1.8 2 V, a whopping 40% more energy than the ideal case (1.23 V). A major fraction of this energy is spent to overcome the hurdle of oxygen evolution reaction (OER). Even after a decade long research on OER materials, the quest for a catalyst that can slash down the overpotential considerably, is still a challenge. In the wake of present issues, it is highly desirable to obtain a single catalyst, made up of transition metals, to catalyse HER and OER equally well. There have been a number of reports on nonprecious metal catalysts that show bifunctional characteristic in alkaline media, such as Ni-P/NF [210], FeP nanotubes [211], Co-Se/Ti [212], Co-P MNA [213], Fe10Co40Ni40P [214] and so on. Recently, a renewed interest has surged in using transition metal borides (TMBs) for electrocatalytic applications [215, 216, 217]. Mono-metal borides such as Co-B [218], Ni-B [215], Mo-B [219] have shown signicant promise for HER in dierent pH conditions. Also, Co2B annealed at 500 102

Chapter 8: Co-Mo-B electrocatalyst for Hydrogen and Oxygen Evolution Reactions

°C (Co B-500) [217] was recently presented as a bifunctional catalyst active in 2

alkaline media. The superior performance of amorphous TMBs is due to the reverse electron transfer from boron to metal, providing higher electron density at these catalytic sites. From recent literature, one could see that ternary Co-based catalysts such as CoMoP [220], CoFeP [221], CoMoS [222], CoMoS2 [223], CoNiB [216], which contain an additional transition metal, show improved electrochemical response compared to their mono-metallic counterparts, owing to the synergy created between the two metals. In the light of these ideas, the present chapter discusses Co-Mo-B nanoparticles, made up of low-cost, non-toxic elements, as an ecient catalyst for HER in neutral and alkaline media. From the morphological, physico-chemical and electrochemical analysis, it was understood that addition of a small amount of Mo leads to signicant improvement in HER activity due to formation of well dispersed nanoparticles with increased specic and electrochemical surface area. Co-Mo-B was also found to be equally active and stable for OER in alkaline media. The presented results for Co-Mo-B catalyst superseded the reports made for Co2B500 and at present stands out to be the best TMB bifunctional electrocatalyst available.

8.2 Characterization of Co-Mo-B catalyst Co-Mo-B nanoparticles (NPs) were synthesized by reducing the aqueous mixture of cobalt and molybdenum salts with sodium borohydride. A series of samples were prepared with dierent molar concentrations of Mo with respect to Co [χM o = Mo/(Mo+Co)]. These samples are hereby denoted as Co-xMo-B, where x indicates the molar content of Mo (χM o = 1, 3, 5, 7, 9). Co-B NPs were also synthesized for comparative study.

Morphological and structural analysis: From SEM image of Co-B and Co-3Mo-B NPs (Figure 8.1), it is seen that Co-B NPs are highly agglomerated as compared to Co-3Mo-B. TEM image in Figure 8.2 gives better insight into the structure of Co-3Mo-B NPs showing uniformly sized spherical nanoparticles which are well separated and highly dispersed. From the corresponding histograms (inset of Figure 8.2), we see a narrow size distribution of Co-3Mo-B NPs with average size of about 18 nm ± 6 nm which is lower than that of Co-B (∼ 30 nm) (Figure 8.3). Thus, we observe that inclusion of a small amount of Mo in Co-B leads to a considerable change in particle size and dispersion creating smaller sized NPs with uniform distribution: conditions favourable for catalytic 103


processes.

Figure 8.1: SEM image of Co-B and Co-3Mo-B catalyst powder.

Figure 8.2: Bright-eld TEM image of Co-3Mo-B catalyst powder with size distribution.

104


Figure 8.3: Bright-eld TEM image of Co-B catalyst powder with size distribution and SAED pattern.

The amorphous structure of Co-B NPs is well known, as seen from SAED pattern in Figure 8.3 (also discussed in Chapter 6). From PXRD analysis of Co3Mo-B, we observe absence of any long-range ordering, as was the case in Co-B. However, on probing the local structure of these NPs at a higher magnication using HRTEM, presence of some short-range ordering in the form of nano-sized crystalline domains encapsulated in amorphous surrounding was observed. HRTEM image of a representative Co-3Mo-B nanoparticle is shown in Figure 8.4. One could observe that Co-3Mo-B NP is not completely amorphous but also consists of small domains showing lattice fringes with d-spacings of 1.97 Å and 1.77 Å that corresponds to (211) planes of Co2B phase as well as (303) planes of Co-hcp phase, respectively. These crystallographic domains are separated by grain boundaries, containing atoms with lower coordination number as compared to atoms in the ideal crystallites, making them highly active catalytic sites, enhancing the catalytic activity. The polycrystalline nature of Co-3Mo-B NPs was also conrmed by SAED pattern (Figure 8.4) which shows diraction spots in addition to the diused rings. Thus, it becomes clearly evident that Co-3Mo-B NPs lack any long-range ordering and possess only short-range ordering owing to the presence of nano-sized crystalline domains.

105


Figure 8.4:

HRTEM image of Co-3Mo-B nanoparticle and SAED pattern showinh polycrys-

talline nature.

Elemental and XPS analysis: Figure 8.5 shows scanning TEM (STEM) image and corresponding EDX elemental mapping images of Co and Mo in Co-3Mo-B, revealing a homogeneous distribution of Co and Mo in the sample. EDX spectra (Figure 8.6) also conrmed the composition of Co and Mo in Co-3Mo-B catalyst. The presence of B, however, could not be established owing to the limitations of these techniques. Nevertheless, XPS was used to establish the surface elemental composition, including that of boron (Table 8.1).

Figure 8.5: Scanning TEM (STEM) image and corresponding EDX elemental mapping images of Co and Mo in Co-3Mo-B.

Figure 8.7 shows XPS spectra of Co-3Mo-B catalyst. The two peaks of Co 2p3/2 levels at BE of 777.9 eV and 781.3 eV were observed corresponding to metallic and oxidized cobalt, respectively. Similarly, two peaks for B 1s level with BE of 187.7 eV and 191.9 eV were observed indicating the presence of metallic and oxidized 106


boron in the sample. Elemental boron peak (187.7 eV) is positively shifted by about 0.7 eV as compared to that of pure B (187.0 eV). This shift results due to the phenomena of reverse electron transfer from B to Co atoms, typical of amorphous transition metal borides [218, 215, 216]. The BE peaks of Co 2p3/2 and B 1s levels in Co-3Mo-B are consistent with that of Co-B catalyst and do not show any variation on inclusion of Mo.

Figure 8.6: EDX spectra showing composition of Co and Mo in Co-3Mo-B.

Table 8.1: Table showing surface elemental composition of Co-3Mo-B from XPS. The XPS spectra of Mo 3d levels (Figure 8.7), indicate that Mo is present only in oxidized state with BE peaks at 230.6, 232.3, 233.6, 235.1 and 235.9 eV. The BE peaks at 232.3 eV and 235.1 eV are assigned to Mo4+ state whereas peaks at 233.6 eV and 235.9 eV are assigned to Mo6+ state in 3d5/2 and 3d3/2 levels of Mo, respectively.

107


Figure 8.7: XPS spectra of Co 2p, B 1s and Mo 3d states in Co-3Mo-B.

BET surface area measurements: XPS data concludes that addition of Mo does not alter the electronic properties of Co-B. From the morphological analysis discussed above, it seems that inclusion of Mo, leads to formation of well-dispersed NPs with uniform size distribution, unlike Co-B. Based on these observations, we propose that in the present case, Mo (in oxidized form) acts like a barrier which prevents Co-B NPs from agglomeration.

Figure 8.8: BET adsorption-desorption isotherms for Co-B and Co-3Mo-B. To conrm this, BET measurements were carried out and the corresponding isotherms are shown in Figure 8.8. The BET surface area for Co-3Mo-B NPs was measured to be 38.8 ± 0.5 m2/g which is almost twice the value measured for Co-B (20.3 ± 0.3 m2/g). Most importantly, BET adsorption-desorption isotherms displayed a prominent hysteresis in Co-3Mo-B which is not visible in Co-B. The capillary action of the adsorbed layer is responsible for this hysteresis eect which is representative of Type H3 isotherms indicating mesoporous structure in Co3Mo-B. This conrms the enhancement in specic surface area of the catalyst 108


by addition of Mo, to produce more number of active sites for participation in electrocatalytic activity.

8.3 Electrochemical measurements 8.3.1 Hydrogen Evolution Reaction The electrochemical performance of the prepared catalysts was evaluated for HER by loading them on a polished glassy carbon (GC) electrode. The catalyst modied GC electrode was then tested in neutral and alkaline media, in a 3 electrode system equipped with a saturated calomel electrode as reference and a Pt electrode as counter.

Figure 8.9:

Linear polarization curves of Co-Mo-B with dierent Mo concentrations varying from 1% Ni to 9% Mo in pH 7 (0.5 M KPi solution). o Figure 8.9 shows polarization curves for Co-Mo-B catalyst with dierent (MMo+Co) ratios varying from 1% to 9%, measured in pH 7 phosphate buer. HER activity improves with increase in Mo content upto 3% and then decreases again. Thus, Co-3Mo-B catalyst, with best HER activity, was chosen for further measurements. Co-3Mo-B catalyst was rst tested in neutral media to establish its eectiveness for HER under benign pH conditions. Figure 8.10 shows the linear sweep curves recorded for Co-3Mo-B catalyst along with Co-B and Pt electrode for comparison. All the curves were iR corrected by measuring their Rohmic values using EIS and the current densities were normalized to the geometric surface area of the bare GC electrode. Co-3Mo-B shows negligible onset potential and could

109


achieve the benchmark current density of 10 mA/cm2 at mere 96 mV (vs RHE) which is just 46 mV higher than that obtained with Pt. When the cathodic potential is increased further, the current density rises sharply and higher current densities of 20 mA/cm2 and 40 mA/cm2 were reached at 141 mV and 213 mV, respectively, for Co-3Mo-B catalyst. The overpotential values reported here for Co-3Mo-B catalyst are signicantly lower than those reported for Co-B (197 mV for 10 mA/cm2) and all other metal boride catalysts, active in pH 7, with the exception of Ni-Bx lms which show even lower overpotential (∼54 mV) owing to its nanostructured lm assembly. Co-3Mo-B outclasses many other non-noble HER catalysts, such as CoNx/C [224], Mo2C@NC [225], Cu(0) catalyst [226], WP NAs/CC [227], Fe-P/CC [228] and H2-Cocat [229], employed in pH 7 (Appendix D). Tafel plot analysis (inset of 8.10) yields Tafel slope values of 56 mV/dec, 71 mV/dec and 38 mV/dec for Co-3Mo-B, Co-B and Pt electrode. The lower value of Tafel slope again indicates better HER kinetics. A Tafel slope value of 56 mV/dec for Co-3Mo-B suggests dominance of Volmer-Heyrovsky reaction mechanism. By extrapolation of the graph used for Tafel slope analysis, exchange current densities were determined for Co-3Mo-B (1.20 mA/cm2) which is much higher than that of Co-B and many other non-noble HER catalysts in pH 7 (Appendix D).

Figure 8.10:

Linear polarization curves and tafel plots (inset) for Pt electrode, Co-3Mo-B, Co-B and bare GC electrode in pH 7

As Co-B is known to work well in alkaline media [216], the HER performance of Co-3Mo-B catalyst was also tested in alkaline media (1 M NaOH). Figure 8.11 shows linear sweep curves for Co-3Mo-B, Co-B and Pt electrode in pH 14 recorded at a scan rate of 5mV/s. Co-3Mo-B produces exceedingly high current densities 110


at very meagre overpotentials. The current densities of 10, 50 and 100 mA/cm2 were achieved at 66 mV, 145 mV and 184 mV, respectively for Co-3Mo-B while that for Pt required 36, 81 and 92 mV respectively. The overpotentials for Co3Mo-B catalyst are amongst the lowest values reported for any non-noble HER electrocatalyst, lower than that of Ni-Bx lms [215], CoNx/C [224], NiO/Ni-CNT [230], Mo2C/NCNT [231], Ni0.33Co0.67S2 NWs [232], FeP nanorod array [233] and Co-P/CC nanowire array [234], all working in pH 14 (Appendix D). The Tafel slope and exchange current density for Co-3Mo-B in alkaline media was calculated to be 67 mV/dec and 1.95 mA/cm2(inset of 8.11).

Figure 8.11:

Linear polarization curves and tafel plots (inset) for Pt electrode, Co-3Mo-B, Co-B and bare GC electrode in pH 14.

To quantify the performance of a catalyst, it becomes crucial to probe the intrinsic catalytic activity obtained per catalytic site, which can be estimated by calculating the turn-over frequency (TOF). Following a previously reported method [235], TOF value for Co-3Mo-B was calculated to be 0.0595 atom−1s−1 at an overpotential of 200 mV, in pH 7. This value is not only higher than Co-B but also when compared to CoS lm [236] and H2 Co-cat [229], all reported at pH 7 (Table 8.2). Thus, the synergic eect created in Co-3Mo-B due to addition of Mo, not only increases the number of surface active sites, but also improves the HER activity per catalytic site. This improvement in activity of each site is attributed to the presence of nano-domains in each NPs, separated by grain boundaries having under co-ordinated active atoms. 111


Table 8.2: Comparison of TOF values with other reports from literature in pH 7.

Figure 8.12:

Potentiostatic curves for Co-3Mo-B catalyst at 100 mV in pH 7 and 80 mV in

pH 14.

Figure 8.13:

Linear polarization curves of Co-3Mo-B before and after 1000 cycles in 0.5 M KPi (pH 7) and 1 M NaOH (pH 14) (at a scan rate of 100 mV/s).

112


Figure 8.12 shows remarkable stability of Co-3Mo-B catalyst when tested under potentiostatic conditions in neutral and alkaline media for about 40 hours. Recycling tests (Figure 8.13) also show negligible loss in activity of Co-3Mo-B catalyst, even after 1000 cycles of operations in both neutral and alkaline solutions. These tests demonstrate excellent electrochemical stability of Co-3Mo-B in both the reaction media.

8.3.2 Electrochemical Surface Area (ESA) One of the key strategies to improve HER performance is by increasing the total number of surface active sites. Co-3Mo-B possesses high specic surface area, as conrmed from BET analysis, presenting a large number of active sites and hence the enhancement in its performance. To concrete these results, electrochemical surface area (ESA) was estimated by measuring the double layer capacitance formed at the electrode/electrolyte interface.

Figure 8.14: Cyclic voltammograms for (a) bare GC electrode, (b) Co-B, (c) Co-3Mo-B catalyst at dierent scan rates from 20 to 100 mV/s; (d) the measured capacitive currents plotted as a function of scan rate.

The electrochemical capacitance was determined by sweeping the potential in the window of 100 mV on either sides of open-circuit potential (OCP) in 0.1 M Na2SO4 solution, at increasing scan rates of 20, 40, 60, 80 and 100 mV/s [215]. 113


Usually, in the potential range of 100 to 150 mV from OCP, no Faradaic processes are involved and the current obtained is due to charging and discharging of the capacitive double layer (Cdl ) [237]. Cyclic voltammetry (CV) curves for bare GC electrode, Co-B and Co-3Mo-B are shown in Figure 8.14(a-c). The dierence in the capacitive currents (Δj = |jcathodic janodic|) were then measured at -0.2 V for bare GC, -0.4 V for Co-B and -0.15 V for Co-3Mo-B catalysts, corresponding to all scan rates. This dierence in capacitive current was plotted against the respective scan rates, as shown in Figure 8.14d. The slope of this graph gives the value of specic capacitance, which is an indicator of the electrochemical surface area of the catalyst. ESA for Co-3Mo-B, based on the Cdl value (0.761 mF/cm2), is almost 1.6 and 10 times higher than that of Co-B (0.481 mF/cm2) and bare GC electrode (0.049 mF/cm2). The increase in ESA of Co-3Mo-B is in accordance with the enhancement in specic surface area obtained by BET analysis, resulting in high electrochemical performance of Co-3Mo-B.

8.3.3 Oxygen Evolution Reaction In a recent work, Co2B-500 [217] was reported to portray bifunctional characteristics, being active for hydrogen and oxygen evolutions in alkaline media. Motivated by the report, we tested Co-3Mo-B for oxygen evolution reaction (OER) in 1 M NaOH solution.

Figure 8.15: Linear polarization curves for Pt electrode, Co-3Mo-B and bare GC electrode in pH 14.

114


Figure 8.16: Recycling and stability curves for Co-3Mo-B catalyst at 350 mV in pH 14. As speculated, Co-3Mo-B displayed superior performance for oxygen evolution as well, overhauling the report made for Co2B-500. Fig 8.15 shows anodic linear sweep curves for Co-3Mo-B catalyst in pH 14, along with that of Pt electrode for comparison. Co-3Mo-B requires an overpotential of 320 mV to attain 10 mA/cm2 which is lower than that required for Pt electrode (500 mV). This overpotential value is also lower than that reported for Co2B-500 (380 mV) [217], RuO2 (400 mV) [217], NG-CoSe/GC (370 mV) [238], Co/C/GC (390 mV) [239], Co3O4/NF (330 mV) [240], Ni/NiO(OH)/NC(390 mV) [241] and MnNixOy (430 mV) [242].

Figure 8.17: Linear polarization and potentiostatic curve (inset) for Co-3Mo-B catalyst in pH

7.

115


A suitable comparison of the OER overpotential with other catalysts in literature is presented in Table 8.3. The Tafel slope analysis for Co-3Mo-B, yielded a value of 155 mV/dec (inset of Figure 8.15). Co-3Mo-B showed favourable recyclability for 1000 cycles (Figure 8.16) and stability (inset of Figure 8.16) for about 10 hours of OER operation at a constant overpotential of 350 mV. The test for OER was also performed in pH 7 solution, however the potential required to drive the current densities were pretty higher (2.03 V for 10 mA/cm2) as shown in Figure 8.17. The mechanism for oxygen evolution in alkaline media initiates with adsorption of OH− ions on the electrode surface [217, 243, 244]. In alkaline media, the Mo oxide in Co-3Mo-B plays the role of a Lewis acid site and readily attracts OH− ions (Lewis base) toward the electrode surface, thereby initiating OER process more eciently. The adsorbed OH then reacts with other OH− ions leading to formation of an intermediate oxide/hydroxide layer. The nal step involves interaction of OH− ions with this oxide/hydroxide layer to form O2 molecule which is later desorbed from the electrode surface.

Figure 8.18: XPS spectra of Co 2p state in Co-3Mo-B after oxidation tests in pH 14. It has been reported numerous times that formation of these oxide/hydroxide type species on the electrode surface play a vital role in improving the oxygen evolution rate [217, 245]. Figure 8.18 shows XPS spectra of Co 2p levels in Co3Mo-B electrode after testing it for OER in pH 14 for about 50 cycles. We see that the peak corresponding to metallic Co vanishes and two peaks were observed for Co 2p3/2 levels at BE of 780.1 eV and 781.3 eV, corresponding to formation of 116


Co(OH)O and Co(OH)2 species. As discussed above, these species are active for OER and appearance of these on the surface of Co-3Mo-B explains its eectiveness for OER.

8.4 Overall Water splitting using Co-Mo-B As a test of bifunctional nature, Co-3Mo-B was tested in alkaline media by sweeping the potential from cathodic region to anodic region and the corresponding graph is shown in Figure 8.19.

Figure 8.19:

Linear sweep voltammogram for Co-3Mo-B showing its ability to function both as a HER catalyst as well as an OER catalyst.

From the gure, it could be seen that Co-3Mo-B shows true bifunctional character, producing H2 and O2 eciently, in the respective potential regions. However, once we enter the anodic potential domain, formation of irreversible oxide/hydroxide layer on the surface takes place and the catalyst cannot be used again in cathodic region. Once the bifunctional nature of Co-3Mo-B was established, its practical use was demonstrated by using it as both cathode and anode in a 2-electrode electrolysis system. For the sake of comparison, a similar experiment was carried out where Pt served as both the electrodes, in an alkaline media. From Figure 8.20, one could see that this 2-electrode alkaline electrolysis assembly could produce 10 mA/cm2 at a cell voltage of 1.69 V for Co-3Mo-B || Co-3Mo-B, much lower than for Pt || Pt (1.9 V). This cell voltage for Co-3Mo-B || Co-3Mo-B is also pretty much lower than that reported for Co2B-500 || Co2B-500/NG (1.81 V) [217] and comparable to NiP/NF || NiP/NF (1.67 V) [210], FeP NTs || FeP NTs (1.69 V) [211], CoSe/Ti || CoSe/Ti (1.65 V) [212] and other totally non-noble electrocatalysts (Table 8.3). In this 2-electrode assembly, Co-3Mo-B could maintain 117


a constant current density at 1.69 V for about 25 hours of full water electrolysis (inset of Figure 8.20), making it highly attractive for industrial applications.

Figure 8.20:

Polarization curves for Co-30Mo-B||Co-30Mo-B and Pt||Pt for overall water splitting in 1 M NaOH (2-electrode assembly); inset shows potentiostatic water electrolysis at 1.69 V for Co-30Mo-B||Co-30Mo-B over 25 hours.

Catalyst

# ∗

Co-3Mo-B∗ Co2B-500 [217] NiMo NAs/Ti [246] NiP/NF [210] Ni0.33Co0.67S2 [232] FeP nanotubes [211] Co-Se/Ti [212] Co NPs/N-Carbon [247] CoP/MNA [213] CoP NS/C [248] CoP/NC [249] Fe10Co40Ni40P [214] Fe2Ni2N [250]

Overpotential HER OER

66 92 80 88 120 121 298

320 380 310 309 330 (onset) 288 292 370

54 111 191 68 180

290 277 354 250 -

Potential (V)#

1.69 1.81 1.64 1.67 1.65 (for 5 mA/cm2) 1.69 1.65 1.64

1.62 1.54 1.55 (onset) 1.57 1.65

Potential required to achieve 10 mA/cm2 in 2-electrode assembly.

This work

Table 8.3:

Comparison of the electrocatalytic activity of Co-3Mo-B to recently reported transition-metal based bifunctional catalysts in basic solutions.

118


8.5 Conclusion In summary, we presented Co-3Mo-B nanocatalyst, made up of low-cost, inexpensive, non-toxic elements, as an ecient catalyst for HER in neutral and alkaline media. The enhancement in activity of Co-3Mo-B was attributed to the synergic eect created by addition of Mo in Co-B, leading to an increment in the eective surface area and also improvement in the activity per catalytic site. Co-3Mo-B also turned out to be an ecient catalyst for oxygen evolution, owing to the formation of oxide/hydroxide species of Co on its surface, in alkaline media, as detected from XPS. The bifunctional characteristic of Co-3Mo-B was explored by employing it as both electrodes in a 2-electrode cell where it could generate 10 mA/cm2 at an operating voltage of 1.69 V, in alkaline media, with extremely high stability. It is worth mentioning that Co-3Mo-B stands out as the best bifunctional metal boride electrocatalyst reported so far. It also presents a unique scope of further improvement by nanostructuring or using suitable conducting supports. This work would certainly turn many more heads towards the untrodden and exciting realm of metal boride electrocatalysts for full water splitting.

119

Chapter 9 Conclusions and Future Outlook The ever-growing need for an ecient and clean energy source has given rise to a number of renewable fuels. However, hydrogen has emerged as one of the most promising energy carriers by virtue of its high gravimetric energy density and zero emission of green house gases. Hydrogen powered PEM fuel cells are attractive power sources for providing clean energy for transportation and personal electronic applications especially, where low system weight and portability are important. On our planet, hydrogen is not available in free form, but mainly exists in compounds such as hydrocarbons, chemical hydrides and water. At present, it is predominantly obtained from the steam reforming of hydrocarbons, relying on non-renewable energy and releasing enormous amounts of green house gases. The sustainable production of H2 fuel on an economical and industrial scale presents a daunting challenge. There are several processes for hydrogen production and tremendous amount of research is being pursued in the development of hydrogen generation and storage systems. Hydrolysis of chemical hydrides, in presence of a suitable catalyst is a simple, convenient and safe chemical process that generates high purity hydrogen gas on demand. Another proposed approach is via electrocatalytic water splitting, combined with photovoltaics, in which water molecules are dissociated into hydrogen and oxygen. The catalysts used in the above processes (like Pt and Pd) suer from high cost and less abundance, which hinders their mass scale implementation. This thesis was aimed at developing new kinds of catalysts that are economical, easy to synthesise and at the same time show high catalytic activity, comparable to that of noble metals for H2 production by hydrolysis of chemical hydrides and electrolysis of water. For hydrolysis of chemical hydrides (Sodium Borohydride and Ammonia Borane), the present study reported various new strategies to improve the catalytic performance of Co-B catalyst. These strategies include - supporting Co-B on mesoporous silica substrates (to reduce the agglomeration), using mesoporous Co-B (to increase the surface area) and using nanoparticle assembled coatings (NPACs) 120

Chapter 9: Conclusions and Future Outlook

(providing high surface area and good stability against aggregation). For water electrolysis, this thesis reports three transition metal borides (Co-B, Co-Ni-B and Co-Mo-B) as electrocatalytically active materials. Of these, Co-B and Co-Ni-B were found to be active for hydrogen evolution reaction (HER) alone while CoMo-B showed bifunctional nature, producing hydrogen and oxygen equally well.

9.1 Hydrolysis of chemical hydrides 9.1.1 Conclusions From the present study, it comes out that mesoporous (MSP) architecture improves the catalytic performance multifold, even when used as substrates or when mesoporous catalysts were used. Using MSP silica substrates, provide a unique opportunity to tune the size of the pores and thus that of the NPs enclosed within them. Also, they prevent the supported NPs from agglomeration, thereby improving the activity. On the other hand, when catalyst itself is made mesoporous, the eective surface area increases signicantly. Also, mesoporous assembly facilitates easy passage of reactants to the active sites, resulting in enhanced activity. Thus, as per this study, both these strategies employing mesoporous morphology, have found fair amount of success when the catalyst to be used is in the form of powder.

Table 9.1: Maximum H

2

generation rate and activation energy of all reported catalysts.

For applications where thin lm catalysts are required, Co3O4 nanoparticle assembled coatings (NPACs) were reported to serve as ecient catalyst for hydrolysis of sodium borohydride. The advantage of using catalyst in this form is 121


that it can function as ON/OFF switch for the reaction and can be easily recovered and reused. Dierent parameters were varied during deposition of these lms by PLD, to obtain the most eective catalyst composition in the lms. Table 9.1 summarizes the H2 production rates with corresponding activation energies for all the catalysts reported. It becomes clearly evident that all the three strategies lead to enhancement of H2 production rates signicantly, from hydrolysis reactions.

9.1.2 Future prospects The successful implementation of the above-mentioned strategies have paved way for a number of future prospects listed below.

MSP Co-B as well as Co-B supported on MSP silica, both serve as models

for other heterogeneous catalysis reactions, to improve upon the catalytic performance.

The success of Co-B based catalysts in sodium borohydride and ammonia

borane, demands their application to hydrolysis of other promising chemical hydrides, such as hydrazine borane.

Co-B NPs have endured their activity in various forms (supported, mesoporous and lms). New strategies to obtain dierent nanostructures of Co-B (such as urchin-like, nano-owers, etc.) can be undertaken with success.

Signicant eorts towards energy-ecient recycling of spent chemical hydrides will open up new doors for H2 production from chemical hydrides.

9.2 Electrocatalytic water splitting 9.2.1 Conclusions The study reported here presents three new transition-metal borides (TMBs) - CoB, Co-Ni-B and Co-Mo-B for electrochemical water splitting. Prior to this study, there were very few reports made on TMBs for such applications. In fact, this study reported for the rst time, the viability of TMBs in pH neutral solutions. For Co-Ni-B and Co-Mo-B catalyst, the molar ratio i.e. χM = CoM+etal was M etal varied in order to study the eect of dopant concentration on the electrochemical performance. It was identied that χN i = 30% and χM o = 3% showed the highest electroctalytic activity. Co-B and Co-Ni-B were active for HER in wide pH media, ranging from pH 4 to 14. In acidic media, catalyst degradation initiated after a few cycles of operation. Co-Mo-B also showed enhanced HER activity in neutral 122


and alkaline media. All the catalysts were characterized by various techniques such as XPS, BET, SEM, TEM and XRD. The excellent HER performance of these amorphous borides were due to the phenomenon of reverse electron transfer from boron to cobalt. Owing to this, electron density at Co active sites increases, thereby facilitating proton reduction. The occurence of reverse electron transfer in amorphous Co-B was established by experimental (XPS, EXAFS) as well as computational (DFT) evidences. Co-Mo-B catalyst was also found to be equally active for oxygen evolution reaction (OER) in alkaline media. The OER proceeded by formation of oxide/hydroxide species of Co on the surface of the catalyst, in an alkaline medium. The formation of these surface species ensured ecient oxidation of OH− ions to produce oxygen. Table 9.2 compares the electrochemical performance of all the three catalysts in neutral and alkaline media.

Table 9.2:

Comparison of HER performance in neutral and alkaline media for all the three reported catalysts.

It could be seen that HER performance increases in the order Co-Mo-B > CoNi-B > Co-B. In Co-Ni-B, the role of Ni is to create an atomic arrangement where Cobalt atoms are surrounded by higher number of boron atoms, which in turn promotes higher electron density at Co sites. Thus, the electrochemical activity per catalytic site improves signicantly by addition of Ni in Co-B. In Co-Mo-B, BET and electrochemical surface area mesurements indicated an increase in the specic surface area by 2 times and electrochemical surface area by 1.6 times, when compared to Co-B. Mo clusters act like a barrier between the Co-B NPs resulting in less agglomeration and better dispersion, as evident from electron microscopy. The enhancement in surface area, with high degree of dipersion, provides a large number of surface active sites for reaction to proceed, which explains the signicant enhancement in activity of Co-Mo-B catalyst.

123


9.2.2 Future prospects The following graph summarizes the evolution of transition-metal based HER catalysts over the last decade. The present study lends a signicant contribution towards re-igniting the lost interest in transition-metal borides for electrocatalytic water splitting. Eventually, transition-metal borides have shown enough potential to bridge the gap between non-noble and noble metal catalysts.

The success of this study has opened up new pathways towards achieving the goal of replacing noble-metal catalysts. It presents a number of exciting future prospects, some of which are listed below:

The inclusion of Ni and Mo in Co-B leads to signicant enhancement in elec-

trocatalytic activity. Similarly, the eect of other transition-metal dopants such as Cr, W, Cu etc. towards electrochemical performance of Co-B can be studied.

All the reported materials were used in the form of nanoparticles only.

A number of dierent strategies can be employed to further improve their activity such as, using suitable substrates, using mesoporous catalysts and making dierent nanostructures (urchin-like, nanowires, etc.). 124


The practical application of an electrocatalyst is in conjunction with a photovoltaic. An alkaline electrolyzer constituting Co-Mo-B catalyst as both the electrodes, powered by a photo-voltaic, can serve as a low-cost and clean H2 production system for household applications.

The reported catalyst materials can be coated on a photo-electrode to form an ecient assembly for photo-electrochemical water splitting.

125

List of Figures 1.1 1.2 1.3 1.4 1.5 1.6 2.1 2.2

World marketed energy consumption in Btu.

................. 2

Relation between human development index (HDI) and per capita electricity

............................... Variation in global temperature and CO concentration over the last century. . Schematic of an alkaline water electrolyzer. . . . . . . . . . . . . . . . . . Schematic of a PEM water electrolyzer. . . . . . . . . . . . . . . . . . . . Volcano curve for HER. . . . . . . . . . . . . . . . . . . . . . . . . . .

consumption.

2

3D view of the PLD apparatus.

3 4 11 12 14

. . . . . . . . . . . . . . . . . . . . . . . 23

Schematic diagram of the experimental setup for hydrogen gas measurement using gas-volumetric method. 1: Water bath; 2: Reaction chamber; 3: Syringe for insertion of hydride solution; 4: thermometer; 5: Erlenmayer ask; 6: Lab jack; 7: Electronic balance; 8: ball valve; 9: measuring cylinder; 10: RS 232 coupled acquisition software.

3.1 3.2 3.3

. . . . . . . . . . . . . . . . . . . . . . . . 25

Small-angle XRD pattern of NPS, MCM-41, FSM-16 and SBA-15 type silica materials.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Wide-angle XRD pattern of unsupported Co-B powder and that supported on NPS, MCM-41, FSM-16 and SBA-15 type silica materials. Nitrogen adsorption

. . . . . . . . . . 31

desorption isotherms of (a) MCM-41, (b) FSM-16 and

(c) SBA-15 type mesoporous silica supports with and without Co-B catalyst

gures shows the pore size distribution curves of the corresponding mesoporous silica supports. . . . . . . . . . . . . . . . . . . 33 3.4 SEM image of bare Co-B powder . . . . . . . . . . . . . . . . . . . . . . 35 3.5 Bright eld TEM micrograph of bare (a) NPS, (c) MCM-41, (e) FSM-16, (g) loading. Inset of the

and (i) SBA-15 type silica supports while (b), (d), (f), (h) and (j) shows mi-

3.6

crographs of corresponding supports with Co-B catalyst loading.

. . . . . . . 37

Hydrogen generation yield, as a function of reaction time, obtained by hydrolysis of AB (0.025 M) in the presence of unsupported Co-B catalyst powder, and that supported over NPS, MCM-41, FSM-16, and SBA-15 type silica supports.

126

39

LIST OF FIGURES 3.7

Hydrogen generation yield, as a function of reaction time, obtained by hydrolysis of AB (0.025 M) at 4 dierent solution temperatures in the presence of MCM-41, FSM-16, and SBA-15 type silica supported Co-B catalyst. Inset

3.8

shows the Arrhenius plot of the H2 generation rates for each support.

. . . . . 42

Hydrogen generation yield, as a function of reaction time, obtained by hydrolysis of AB (0.025 M) in the presence of untreated Co-B catalyst supported over SBA-15 type silica and that heat treated in Ar atmosphere for 2 h at 673, 773, and 873 K.

4.1 4.2 4.3 4.4

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

XRD pattern of Co-B/CTAB, Co-B/P123, and nonporous Co-B catalyst.

. . . 46

Nitrogen adsorption-desorption isotherms of Co-B/CTAB, Co-B/P123 and nonporous Co-B catalyst.

. . . . . . . . . . . . . . . . . . . . . . . . . . . 47

SEM image of (a) non-porous Co-B, (b) Co-B/CTAB, (c) Co-B/P123 and (d) bright eld TEM micrograph of Co-B/P123.

. . . . . . . . . . . . . . . . . 49

Hydrogen generation yield, as a function of reaction time, obtained by hydrolysis of NaBH4 (0.025 M) in presence of Co-B/CTAB, Co-B/P123 and non-porous

4.5

Co-B catalysts.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

Hydrogen generation yield, as a function of reaction time, obtained by hydrolysis of NaBH4 (0.025 M) at 4 dierent solution temperatures in presence of MSP Co-B/CTAB and Co-B/P123 catalysts. Inset shows the Arrhenius plot

4.6

of the H2 generation rates.

. . . . . . . . . . . . . . . . . . . . . . . . . 53

Hydrogen generation yield, as a function of reaction time, obtained by hydrolysis of NaBH4 (0.025 M) in presence of untreated MSP Co-B/CTAB and Co-B/P123 catalysts, and heat treated in Ar atmosphere for 2 h at 573,673,

4.7

and 773 K.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

SEM images of heat treated MSP Co-B/CTAB and Co-B/P123 catalysts in Ar atmosphere for 2 h at 773 K.

5.1

. . . . . . . . . . . . . . . . . . . . . . . . 55

XRD pattern of: (a) Co3 O4 NPAC prepared by PLD at 250° C with laser uence of 3J/cm2 and O2 pressure of 4.5 x 10−2 mbar, and (b) Co3 O4 powder

5.2

prepared by chemical method, and (c) standard JCPDS pattern of Co3 O4 .

. . 59

SEM images of (a) Co3 O4 powder prepared by chemical method and (b) Co3 O4 NPAC prepared by PLD at 250° C with laser uence of 3J/cm2 and O2 pressure

5.3

of 4.5 x 10−2 mbar.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

(a) Bright Field TEM image and (b) particle size distribution of Co3 O4 NPAC prepared by PLD at 250 °C with laser uence of 3J/cm2 and O2 pressure of 4.5

5.4

x 10−2 mbar. Inset of (a) shows HR-TEM of corresponding sample.

. . . . . . 61

XPS spectra with (a) Co 2p and (b) O 1s core levels of Co3 O4 powder prepared by chemical method and Co3 O4 NPAC prepared by PLD at 250° C with laser uence of 3J/cm2 and O2 pressure of 4.5 x 10−2 mbar.

127

. . . . . . . . . . . . 62

LIST OF FIGURES 5.5

Hydrogen generation yield as a function of reaction time obtained by hydrolysis of alkaline NaBH4 (0.025 M) using Co3 O4 powder prepared by chemical method and Co3 O4 NPAC prepared by PLD at 250 °C with laser uence of 3J/cm2 and

5.6

O2 pressure of 4.5 x 10−2 mbar.

. . . . . . . . . . . . . . . . . . . . . . . 64

Maximum H2 generation rate, obtained hydrolysis of alkaline NaBH4 (0.025 M), as a function of dierent O2 pressures used for the deposition of Co3 O4

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 5.7 XRD pattern of Co O NPACs deposited by PLD using dierent O pressures. 66 5.8 XPS spectra of Co 2p core level of Co O NPACs deposited by PLD using O pressures of (a) 3 x 10 mbar and (b) 8 x 10 mbar. . . . . . . . . . . 67 5.9 SEM images of Co O NPACs deposited by PLD using O pressures of (a) 3 x 10 mbar, (b) 8 x 10 mbar and (c) 8 x 10 mbar. . . . . . . . . . . . . 68 5.10 SEM images of Co O NPACs deposited by PLD using laser uence of (a) 1 J/cm , (b) 5 J/cm and (c) 7 J/cm . . . . . . . . . . . . . . . . . . . . . 69 5.11 Maximum H generation rate, obtained by hydrolysis of alkaline NaBH (0.025 NPACs by PLD.

3

4

2

3

3/2

4

−3

2

3

−3

−2

4

2

−3

3

2

−2

4

2

2

2

4

M), as a function of dierent laser uences used for the deposition of Co3 O4

5.12 5.13

NPACs by PLD.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

Raman spectra of Co3 O4 NPACs deposited by PLD using substrate temperature of (a) RT, (b) 200 ºC, (c) 250 ºC and (b) 300 ºC.

. . . . . . . . . . . . 70

Maximum H2 generation rate, obtained by hydrolysis of alkaline NaBH4 (0.025 M), as a function of dierent substrate temperature used for the deposition of

5.14

Co3 O4 NPACs by PLD.

. . . . . . . . . . . . . . . . . . . . . . . . . . 71

Reusability behavior of Co3 O4 NPACs on hydrogen generation yield, as a function of reaction time, measured during hydrolysis of 0.025 M NaBH4 alkaline solution.

6.1 6.2

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

(a) SEM image, (b) Bright eld TEM image, (c) XRD pattern, (d) HRTEM image with inset showing SAED pattern of Co-B catalyst.

. . . . . . . . . . 76

Macroscopic elemental mapping of Co-B catalyst using Energy Dispersive Spec-

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 6.3 X-ray Photoelectron Spectra of Co 2p and B 1s level of Co-B catalyst. . . . . 77 6.4 Linear polarization curves with iR correction for Pt, Co-B, and Co in 0.5 M troscopy analysis.

potassium phosphate buer with pH 7 obtained at a scan rate of 10 mV/s.

6.5

Inset shows the corresponding Tafel plot for the Co-B catalyst.

. . . . . . . . 79

Linear polarization curves with iR correction for Co-B catalyst compared with Co metal in (a) pH 1 (0.1 M HClO4 ), (b) pH 4.4 (0.5 M KH2 PO4 ) and (c) pH 9.2 (0.4 M K2 HPO4 ) obtained with scan rate of 10 mV/s. (d) Plot of overpotential (at 2 mA/cm2 ) and exchange current density values as a function of pH values of the solution used to test the Co-B catalyst.

128

. . . . . . . . . . 81

LIST OF FIGURES 6.6

Plot of charge build-up versus time for Co-B catalyst acquired at (a) pH 7 (b) pH 9.2 and (c) pH 4.4 at constant overpotentials. Inset of each plot shows the

6.7

variation in current density over a long period of time.

Recycling behavior of Co-B catalyst examined in aqueous solutions of (a) pH 7 (b) pH 9.2 and (c) pH 4.4 at a scan rate of 150 mV/s.

7.1

. . . . . . . . . . . . 82 . . . . . . . . . . . 83

(a) High resolution SEM image of Co-30Ni-B; (b) TEM image showing the average size of Co-30Ni-B particles around 25 30 nm; (c) High resolution TEM image, SAED pattern (inset of c) and (d) XRD pattern showing the

7.2 7.3

presence of amorphous state in Co-30Ni-B.

. . . . . . . . . . . . . . . . . 86

XPS spectra of Co 2p, Ni 2p and B 1s states in Co-30Ni-B (a-c) and Co-50Ni-B (d-f); Ni 2p (g) and B 1s (h) states in Ni-B.

. . . . . . . . . . . . . . . . . 88

(a) XANES spectra of Co-30Ni-B is shown along with Co metal foil and CoO reference spectra; (b) normalized EXAFS spectra of Co-B and Co-30Ni-B at Co K edge; Fourier transformed EXAFS spectra of (c) Co-B and (d) Co-30Ni-B

7.4 7.5

measured at Co K edge (Scatter points) and theoretical t (Solid line).

. . . . 90

Nanoclusters of (a-b) Co-25Ni-B and (c) Co-B used for the calculation of charge transfer:

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

Linear polarization curves in pH 7 (0.5 M KPi solution) of Co-Ni-B with different Ni concentrations varying from 10% Ni to 50% Ni. Inset shows the plot

7.6 7.7 7.8 7.9

of overpotential at 10 mA/cm2 versus Ni concentration in Co-Ni-B catalyst.

. . 93

Linear polarization curves for Pt electrode, Co-30Ni-B, Co-B and Ni-B electrocatalysts in pH 7 (0.5 M KPi solution)

. . . . . . . . . . . . . . . . . . . 93

Tafel plots for Pt electrode, Co-30Ni-B, Co-B and Ni-B electrocatalysts in pH 7 (0.5 M KPi solution).

. . . . . . . . . . . . . . . . . . . . . . . . . . . 95

TOFs calculated for Co-30Ni-B catalyst as a function of overpotential (vs RHE) at pH 7.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

(a) Linear polarization curves for Co-30Ni-B in dierent pH media (from pH 1 to pH 14); (b) Plot of overpotential (at 2 mA/cm2 ) and exchange current

7.10

density values as a function of pH of the solution for Co-30Ni-B catalyst.

. . . 97

Polarization curves of Co-30Ni-B, Pt electrode and bare glassy carbon electrode in (a) 0.1 M HClO4 solution (pH 1) solution and (b) 1 M NaOH solution (pH

. . . . . . . . . . . . . . . . . . . 98 7.11 Tafel plots for Co-30Ni-B in pH 1 and pH 7 . . . . . . . . . . . . . . . . . 98 7.12 (a) Linear polarization curves of Co-30Ni-B before and after 1000 cycles in 0.5 14) obtained at a scan rate of 5 mV s−1

M KPi (pH 7) and 1 M NaOH (pH 14) (at a scan rate of 100 mV/s); (b) Plot of charge build-up versus time and time-dependent current density curves (inset of b) at constant overpotentials of 110 mV in pH 7 and 120 mV in pH 14.

129

. . . 99

LIST OF FIGURES 7.13

Linear polarization curves of Co-30Ni-B annealed at various temperatures in 0.5 M KPi (pH 7).

8.1 8.2 8.3 8.4 8.5

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

. . . . . . . . . . . . . 104 Bright-eld TEM image of Co-3Mo-B catalyst powder with size distribution. . 104

SEM image of Co-B and Co-3Mo-B catalyst powder.

Bright-eld TEM image of Co-B catalyst powder with size distribution and SAED pattern.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

HRTEM image of Co-3Mo-B nanoparticle and SAED pattern showinh polycrystalline nature.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

Scanning TEM (STEM) image and corresponding EDX elemental mapping

. . . . . . . . . . . . . . . . . . . . . 106 8.6 EDX spectra showing composition of Co and Mo in Co-3Mo-B. . . . . . . . . 107 8.7 XPS spectra of Co 2p, B 1s and Mo 3d states in Co-3Mo-B. . . . . . . . . . 108 8.8 BET adsorption-desorption isotherms for Co-B and Co-3Mo-B. . . . . . . . . 108 8.9 Linear polarization curves of Co-Mo-B with dierent Mo concentrations varying from 1% Ni to 9% Mo in pH 7 (0.5 M KPi solution). . . . . . . . . . . . . 109 8.10 Linear polarization curves and tafel plots (inset) for Pt electrode, Co-3Mo-B, Co-B and bare GC electrode in pH 7 . . . . . . . . . . . . . . . . . . . . 110 8.11 Linear polarization curves and tafel plots (inset) for Pt electrode, Co-3Mo-B, Co-B and bare GC electrode in pH 14. . . . . . . . . . . . . . . . . . . . 111 8.12 Potentiostatic curves for Co-3Mo-B catalyst at 100 mV in pH 7 and 80 mV in pH 14. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 8.13 Linear polarization curves of Co-3Mo-B before and after 1000 cycles in 0.5 M KPi (pH 7) and 1 M NaOH (pH 14) (at a scan rate of 100 mV/s). . . . . . . 112 8.14 Cyclic voltammograms for (a) bare GC electrode, (b) Co-B, (c) Co-3Mo-B images of Co and Mo in Co-3Mo-B.

catalyst at dierent scan rates from 20 to 100 mV/s; (d) the measured capacitive

8.15

currents plotted as a function of scan rate.

. . . . . . . . . . . . . . . . . . 113

Linear polarization curves for Pt electrode, Co-3Mo-B and bare GC electrode

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 8.16 Recycling and stability curves for Co-3Mo-B catalyst at 350 mV in pH 14. . . . 115 8.17 Linear polarization and potentiostatic curve (inset) for Co-3Mo-B catalyst in pH 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 8.18 XPS spectra of Co 2p state in Co-3Mo-B after oxidation tests in pH 14. . . . . 116 8.19 Linear sweep voltammogram for Co-3Mo-B showing its ability to function both as a HER catalyst as well as an OER catalyst. . . . . . . . . . . . . . . . . 117 8.20 Polarization curves for Co-30Mo-B||Co-30Mo-B and Pt||Pt for overall water in pH 14.

splitting in 1 M NaOH (2-electrode assembly); inset shows potentiostatic water electrolysis at 1.69 V for Co-30Mo-B||Co-30Mo-B over 25 hours.

A.1

SEM image of Co-P-B lm (4 layers) deposited on SS316L

130

. . . . . . . 118

. . . . . . . . . . 135

LIST OF FIGURES A.2

Steady-state polarization curves for Co-P-B lm and SS316L specimen in 1 M NaOH at a scan rate of 10 mV/s. The corresponding Tafel plots are shown in the inset.

B.1

Novel algorithm incorporated in the in-house instrumentation software in block diagram.

D.1

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138

Comparison of electrochemical parameters for HER between Co-B, Co-30Ni-B, Co-3Mo-B and various solid-state catalysts from literature, active in neutral and alkaline media.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

131

List of Tables 3.1

Physico-chemical properties of non-porous and three dierent mesoporous silica supports (MCM-41, FSM-16, and SBA-15) with and without Co-B catalyst

3.2

loading.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

Pore wall thickness and spacing between the two regular arrays of pore channel

desorption isotherms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 of mesoporous silica supports calculated from the SAXRD and N2 absorption

4.1

Physico-chemical properties of Co-B/CTAB, Co-B/P123 and non-porous Co-B catalyst.

5.1

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

A comparison of various Co-based catalysts used for hydrogen production by hydrolysis of aqueous sodium borohydride solution.

6.1 6.2

Electrochemical parameters obtained with Co-B, and Co metal for HER in aqueous solution of various pH values.

7.2

. . . . . . . . . . . . . . . . . . . . 80

Turnover Frequency of Co-B catalyst measured in aqueous solution of dierent pH values at certain overpotential.

7.1

. . . . . . . . . . . . . 64

. . . . . . . . . . . . . . . . . . . . . 80

Percentage of Co and Ni content in dierent Co-Ni-B catalyst determined by X-ray uorescence measurement.

. . . . . . . . . . . . . . . . . . . . . . 87

BE peak positions and percentage surface composition for all catalysts obtained

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 7.3 Bond length, coordination number and disorder factor obtain by EXAFS tting. 91 7.4 Comparison of TOF values with other reports from literature in pH 7. . . . . 96 7.5 Table showing the current density values at various overpotentials for Co-30NiB catalyst in dierent pH solutions. . . . . . . . . . . . . . . . . . . . . . 97 from XPS analysis.

8.1 8.2 8.3

. . . . . 107 Comparison of TOF values with other reports from literature in pH 7. . . . . 112 Table showing surface elemental composition of Co-3Mo-B from XPS.

Comparison of the electrocatalytic activity of Co-3Mo-B to recently reported transition-metal based bifunctional catalysts in basic solutions.

9.1

. . . . . . . . 118

Maximum H2 generation rate and activation energy of all reported catalysts.

132

. 121

LIST OF TABLES 9.2

Comparison of HER performance in neutral and alkaline media for all the three reported catalysts.

A.1

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Electrochemical Parameters.

. . . . . . . . . . . . . . . . . . . . . . . . 137

133

Appendix A Co-P-B lms for hydrogen evolution from alkaline water electrolysis In this work, Co-P-B lms were deposited on SS316L substrates using electroless deposition. These Co-P-B lms were used as cathodes in an electrochemical cell to produce H2 by alkaline water electrolysis. Steady state linear polarization curves were obtained for Co-P-B coated and uncoated SS316L substrates which were used to determine the overpotential, Tafel slope and exchange current density. For deposition, SS316L sheet (thickness = 1 mm) was chosen as the substrate material due to its high conductivity and corrosion resistance. The surface of SS316L specimen (2 cm x 2 cm) was roughened by emery paper (grade 50) and then ultrasonically cleaned with double distilled water (DDW) followed by acetone. The cleaned specimen was further treated with NaOH solution (1 M) at 333 K to remove any soaks, lubricants and ngerprints. The specimen was then extensively washed with DDW to remove any traces of NaOH on the surface. The cleaned specimen was nally activated by dipping in HCl solution (35% concentrated) for 2 min and then washed with DDW and dried in vacuum at room temperature. This activated specimen was immersed in 10 ml of aqueous solution of cobalt chloride (0.875 M, CoCl2.6H2O) and sodium hypophosphite (0.1 M, NaH2PO2) for 1 min under continuous stirring at room temperature. Ammonia solution (1 ml, 30%) and ammonium chloride (NH4Cl) were added as precipitators to this solution that was again stirred for about 2 minutes. This mixture was reduced by adding an equal volume of aqueous sodium borohydride solution (0.25 M, NaBH4) to it. The specimen was removed from the mixture after bubble generation ceased. The above process was repeated a number of times to ensure that the specimen was completely covered with the catalyst. Finally, the coated specimen was washed with DDW before drying in vacuum at room temperature. The molar ratios of (P+B)/Co and B/P were kept to 4 and 2.5 respectively. 134

Chapter A: Co-P-B lms for hydrogen evolution from alkaline water electrolysis

Structural characterization of Co-P-B coatings by XRD indicates amorphous nature of the catalyst lm. High magnication SEM image (Figure A.1) of the coating surface (4 layers) reveals the formation of two-dimensional nano-akes like structure with space separation of a few tens to hundreds of nanometers. During the deposition, it was observed that one layer of coating was not sucient to completely cover the surface of the specimen. Hence, 3 - 4 layers were coated to ensure complete coverage of the surface. After 4 layers of deposition, thickness of the coating was measured to be around 200 nm, using cross-sectional SEM. Similar to the case of Co-B, XPS spectra (reported in ref. [251]) of Co-P-B lms revealed a positive shift of 1.1 eV in B1s line indicating an electron transfer from B to the vacant d-orbitals of metallic Co and thus enriching Co with electrons. These electron-enriched Co sites serve as the binding site for H+ ions from the solution where reduction takes place. XPS analysis [251] also revealed the presence of metallic and oxidized phosphorous. The inclusion of P creates high number of Co metal active sites on the surface as suggested by the compositional analysis performed on Co-P-B and Co-B catalysts [252]. Thus it is concluded that the role of P is to create large number of Co sites on the surface while B provides electron density to these sites making them catalytically more active.

Figure A.1: SEM image of Co-P-B lm (4 layers) deposited on SS316L The electrochemical measurements were carried out in a 3 electrode at electrochemical cell (Princeton) where 1 cm2 circular area of the specimen was selectively exposed to the electrolyte solution (1 M NaOH). A linear polarization curve (iR corrected) for Co-P-B coated SS316L (4 layers) and bare SS316L specimenis shown in Figure A.2. All the electrochemical parameters which dene the performance of the electrocatalyst are summarized in Table A.1. Co-P-B lm shows onset of H2 production with overpotential as low as 94 mV (vs RHE) (Table A.1). The overpotentials required for current densities of 2 mA cm−2 and 10 mA cm−2 are 163 mV and 197 mv (vs RHE), respectively, for the Co-P-B coatings. These over135


potential values are lower than that reported for Mo2C and MoB catalysts [253] which exhibit overpotential of 250 mV for current densities of 6 mA cm−2 and 5 mA cm−2 respectively and that of Zr-Ni amorphous alloys [254] which show overpotential of 380 mV for a current density of 10 mA cm−2. From the Tafel plot (inset Figure A.2), the Tafel slope for Co-P-B lm is calculated to be 102 mV/decade (Table A.1) suggesting Volmer-Heyrovsky type of mechanism is involved in the reaction with H2 desorption as the rate-limiting step. Co-P-B lms not only exhibit lower overpotentials but their exchange current density (J0) is also signicantly high. Tafel plot analysis of Co-P-B lm gives the exchange current density of 0.125 mA cm−2: a value of about two orders of magnitude higher than that of bare SS316L specimen (3.1 x 10−3 mA cm−2) (Table A.1). The obtained current density is also higher than that of Mo2C catalyst (3.8 x 10−3 mA cm−2) [253], MoB catalyst (2 x 10−3 mA cm−2) [253], Ni-Mo catalyst (2.6 x 10−2 mA cm−2) [255] and Co-Mo catalyst (2.3 x 10−2 mA cm−2) [255], all used at pH 14.

Figure A.2:

Steady-state polarization curves for Co-P-B lm and SS316L specimen in 1 M NaOH at a scan rate of 10 mV/s. The corresponding Tafel plots are shown in the inset.

The higher values of J0 and lower values of overpotential for Co-P-B lms can be attributed to the presence of electron enriched Co metal sites created by transfer of electrons from B to the d-orbitals of Co metal (as conrmed by XPS), thereby providing large number of binding sites for H+ ions. Also, the presence of P plays a role in enrichment of the catalyst surface with Co sites [252] thereby providing more active sites for HER reaction to proceed. With this scenario, adsorption 136


occurs at a much faster rate followed by electrochemical desorption (Heyrovsky step) which acts as the rate-limiting step. Hence synergic eect created by the metalloids (B and P) is the main reason behind the established excellent HER at low overpotential for Co-P-B catalyst lm.

Table A.1: Electrochemical Parameters.

137

Appendix B Algorithm for data acqusition software to measure produced H2

Figure B.1:

Novel algorithm incorporated in the in-house instrumentation software in block

diagram.

The in-house developed instrumentation software can directly acquire the data from RS232 port with a baud rate of 4800. The software package acquires the continous variation in weight caused due to the produced H2. Knowing the volume of H2 produced, the H2 production yield and rate are also computed. Algorithm can export the graphically displayed data into ASCII text le, Excel le and bitmap le format. 138

Appendix C Calculations for Turn Over Frequency (TOF) Molar mass:

Co2B: 128.6774 g/mol; Co-30Ni-B: 128.5335 g/mol and Co-3Mo-B: 130.897 g/mol Density:

Co2B: 8.1 g/cm3; Co-30Ni-B: 8.04 g/cm3 and Co-3Mo-B: 8.116 g/cm3 Molar volume:

Co2B: 15.886 mL/mol; Co-30Ni-B: 15.986 mL/mol and Co-3Mo-B: 16.128 mL/mol BET surface area:

Co2B: 203.2 cm2/mg; Co-30Ni-B: 168.2 cm2/mg and Co-3Mo-B: 388.5 cm2/mg Current density at 200 mV overpotential in pH 7 for a catalyst loading of 2.1 mg/ cm2 :

Co2B: 10.81 mA/cm2; Co-30Ni-B: 16.39 mA/cm2 and Co-3Mo-B: 36.15 mA/cm2 Average surface atoms per 1

Co B: ( 2

3∗6.022∗1023 1 mol

Co-30Ni-B: ( Co-3Mo-B: (

∗

cm2

1 mol )2/3 15.886 cm3

of catalyst:

= 2.347 ∗ 1015

atoms cm2

;

3∗6.022∗1023 1 mol

∗

1 mol )2/3 15.986 cm3

= 2.337 ∗ 1015

atoms cm2

3∗6.022∗1023 1 mol

∗

1 mol )2/3 16.128 cm3

= 2.323 ∗ 1015

atoms cm2

139

;

Chapter C: Calculations for Turn Over Frequency (TOF) Surface atoms per testing area at 2.1 mg/ cm2 :

Co B: 2

2.1mg ∗ ( 1cm2 (glassy carbon)

203.2 cm2 (catalyst) mg

∗

2.347∗1015 atoms ) 1 cm2 (catalyst)

atoms = 10.015 ∗ 1017 test ; area


∗


atoms = 8.254 ∗ 1017 test ; area


∗


atoms = 1.895 ∗ 1018 test area

Co-30Ni-B: 2.1mg ( 1cm2 (glassy ∗ carbon)

Co-3Mo-B: 2.1mg ∗ ( 1cm2 (glassy carbon)

Turnover frequency (per surface atom) at overpotential of 200 mV:

Co B: 2

−3 A

1 mol ∗ 96485 ∗ 6.022∗10 C 1 mol

−3 A

1 mol ∗ 96485 ∗ 6.022∗10 C 1 mol

−3 A

1 mol ∗ 96485 ∗ 6.022∗10 C 1 mol

∗ 10.81∗10 ( 1 turnover 2e− 1 cm2

23 e−

1 test area −1 −1 ∗ 10.015∗10 s ; 17 atoms ) = 0.0337 atom

Co-30Ni-B: ( 1 turnover ∗ 16.39∗10 2e− 1 cm2

23 e−


23 e−


Co-3Mo-B: ∗ 36.15∗10 ( 1 turnover 2e− 1 cm2

140

Appendix D Comparison of HER performance from literature

141

Chapter D: Comparison of HER performance from literature

Figure D.1:

Comparison of electrochemical parameters for HER between Co-B, Co-30Ni-B, Co-3Mo-B and various solid-state catalysts from literature, active in neutral and alkaline media.

[256, 257, 258, 259, 260, 261, 262, 263, 264, 265]

142

Bibliography [1] A. Midilli, M. Ay, I. Dincer, and M. Rosen, Renewable and Sustainable energy Reviews, vol. 9, pp. 255271, 2005. [2] U.S. Energy Information Administration, 2010. [3] Oil and Gas Journal, vol. 107, pp. 2021. [4] BP Statistical Review of World Energy, 2010. [5] I. E. Agency, 2010. [6] A. Zuttel, Naturwissenschaften, pp. 157172, 2004. [7] Wikipedia, http://en.wikipedia.org/wiki/Engine-eciency. [8] M. Ni, D. Y. C. Leung, M. K. H. Leung, and K. Sumathy, Fuel Processing Technology, vol. 87, pp. 461472, 2006. [9] J. Gan, X. Lu, and Y. Tong, Nanoscale, vol. 6, pp. 71427164, 2014. [10] K. S. Jeong and B. Oh, J. Power Sources, vol. 105, pp. 5865, 2002. [11] A. Bauen and D. Hart, J. Power Sources, vol. 86, pp. 482494, 2000. [12] U. Eberle, M. Felderho, and F. Schuth, Angew. Chem. Int. Ed., vol. 48, pp. 66086630, 2009. [13] X. L. Ding, X. Yuan, C. Jai, and Z. F. Ma, Int. J. Hydrogen Energy, vol. 35, pp. 11 07711 084, 2010. [14] Z. T. Xia and S. H. Chan, J. Power Sources, vol. 152, pp. 4649, 2005. [15] N. Patel, R. Fernandes, G. Guella, and A. Miotello, App. Catal. B: Environ., vol. 95, pp. 137143, 2012. [16] S. B. Kalidindi, A. A. Vernekar, and B. R. Jagirdar, Phy. Chem. Chem. Phy., vol. 11, pp. 770775, 2009. 143

BIBLIOGRAPHY [17] S. Orimoa, Y. Nakamoria, G. Kitaharaa, K. Miwab, N. Ohbab, S. Towatab, and A. Zuttel, J. Alloys Compounds, vol. 427, pp. 404406, 2005. [18] T. Umegak, J. Yan, X. Zhang, H. Shioyama, N. Kuriyama, and Q. Xu, Int. J. Hydrogen Energy, vol. 34, pp. 23032311, 2008. [19] F. H. Stephens, V. Pons, and R. Baker, Dalton Transactions, vol. 25, pp. 26132626, 2007. [20] S. C. Amendola, S. L. Sharp-goldman, M. S. Janjua, N. C. Spencer, M. T. Kelly, P. J. Petillo, and M. Binder, Int. J. Hydrogen Energy, vol. 25, pp. 969975, 2000. [21] Z. P. Lin, N. Morigazaki, B. H. Liu, and S. Suda, J. Alloys Compounds, vol. 349, pp. 232236, 2003. [22] H. Schlesinger, H. Brown, A. Finholt, J. Gilbreath, H. Hockstra, and E. Hyde, J. .Am. Chem. Soc., vol. 75, pp. 215219, 1953. [23] R. W. Kreevoy and M. M. Jacobson, Ventron Alembic, vol. 15, pp. 23, 1979. [24] J. C. Walter, A. Zurawaski, D. Montgomery, M. Thornburg, and S. Revankar, J. Power Sources, vol. 179, pp. 335339, 2008. [25] C. M. Kaufman and B. Sen, J. Chem. Soc. Dalton Trans., vol. 2, pp. 307 313, 1985. [26] Y. Kojima, K. Suzuki, and Y. Kawai, J. Power Sources, vol. 155, pp. 325 328, 2006. [27] Y. Bai, C. Wu, F. Wu, and B. Yi, Materials Letters, vol. 60, pp. 22362239, 2006. [28] G. Guella, C. Zanchetta, B. Patton, and A. Miotello, J. Phy. Chem. B, vol. 110, pp. 17 02417 033, 2006. [29] P. Krishnan, S. G. Advani, and A. K. Prasad, App. Cat. B, vol. 86, pp. 137144, 2009. [30] C. H. Hsueh, C. Y. Chen, J. R. Ku, S. F. Tsai, Y. Y. Hsu, F. Tsu, and M. S. Jeng, J. Power Sources, vol. 177, pp. 485492, 2008. [31] S. Ozkar and M. Zahmakiran, J. Alloys Compounds, vol. 404, pp. 728731, 2005. 144

BIBLIOGRAPHY [32] D. Hua, Y. Hanxi, and A. Xinping, Int. J. Hydrogen Energy, vol. 28, pp. 10951100, 2003. [33] H. Ma, W. Ji, J. Zhao, J. Liang, and J. Chen, J. Alloys Compounds, vol. 474, pp. 584589, 2009. [34] M. Diwan, Int. J. Hydrogen Energy, vol. 33, pp. 11351141, 2008. [35] Q. Xu and M. Chandra, J. Alloys Compounds, vol. 446, pp. 729732, 2007. [36] O. Metin, S. Sahin, and S. Ozkar, Int. J. Hydrogen Energy, vol. 34, pp. 63046313, 2009. [37] F. Durap, M. Zahmakiran, and S. Ozkar, Int. J. Hydrogen Energy, vol. 34, pp. 72237230, 2009. [38] S. Ozkar and M. Zahmakiran, App. Cat. B: Environ., vol. 89, pp. 104110, 2009. [39] Q. Xu and M. Chandra, J. Power Sources, vol. 163, pp. 364370, 2006. [40] T. J. Clark, G. R. Whittell, and I. Manners, Inorganic Chemistry, vol. 46, pp. 75227527, 2007. [41] O. Metin and S. Ozkar, Energy and Fuel, vol. 23, pp. 35173527, 2009. [42] C. F. Yao, L. Zhuang, Y. L. Cao, X. P. Ai, and H. X. Yang, Int. J. Hydrogen Energy, vol. 33, pp. 24622467, 2008. [43] T. Umegaki, J. Yan, X. Zhang, H. Shioyama, N. Kuriyama, and Q. Xu, J. Power Sources, vol. 191, pp. 209216, 2009. [44] W. Kreuter and H. Hofman, Int. J. Hydrogen Energy, vol. 23, p. 661, 1998. [45] O. Ulleberg, Int. J. Hydrogen Energy, vol. 28, p. 21, 2003. [46] E. Navarro-Flores and S. Omenovic, J. Mol. Cat. A: Chem., vol. 242, p. 182, 2005. [47] W. Kreuter and H. Hofman, Hydrogen Energy Progress XI, vol. 1, p. 537, 1996. [48] W. T. Grubb, J. Phy. Chem., vol. 63, pp. 5567, 1959. [49] P. Millet, R. Ngameni, S. A. Grigoriev, and V. N. Fateev, Int. J. Hydrogen Energy, vol. 36, pp. 41564163, 2011. 145

BIBLIOGRAPHY [50] S. Gupta, N. Patel, A. Miotello, and D. C. Kothari, J. Power Sources, vol. 279, pp. 620625, 2015. [51] S. Trasatti, J. Electroanal. Chem., vol. 39, p. 163, 1972. [52] M. H. Miles and M. A. Thomson, J. Electrochem. Soc., vol. 123, p. 1459, 1976. [53] B. E. Conway and J. O. M. Bockris, J. Chem. Phy., vol. 26, p. 532, 1957. [54] L. I. Krishtalik, Adv. Electrochem. Eng., 1970. [55] S. Trasatti, J. Chem. Soc. Farad. Trans., vol. 68, p. 229, 1972. [56] D. V. Esposito and J. G. Chen, Energy Environ. Sci., vol. 4, pp. 39003912, 2011. [57] J. Greeley, T. F. Jaramillo, J. Bonde, I. B. Chorkendor, and J. K. Norskov, Nat. Mater., vol. 5, pp. 909913, 2006. [58] D. M. F. Santos and C. A. C. Sequeira, Quim. Nova, vol. 36, pp. 11761193, 2013. [59] H. E. G. Rommal and P. J. Moran, J. Electrochem. Soc., vol. 135, pp. 343 346, 1988. [60] J. L. Dempsey, B. S. Brunschwig, J. R. Winkler, and H. B. Gray, Acc. Chem. Res., vol. 42, p. 1995, 2009. [61] V. Artero and M. Fontecave, Chem. Soc. Rev., vol. 42, p. 2338, 2013. [62] A. Wang and T. Zhang, Nat. Energy, vol. 1, pp. 12, 2016. [63] Y. Sun, C. Liu, D. C. Grauer, J. Yano, J. R. Long, P. Yang, and C. J. Chang, J. Am. Chem. Soc., vol. 135, pp. 17 69917 702, 2013. [64] J. Tian, Q. Liu, A. M. Asiri, and X. Sun, J. Am. Chem. Soc., vol. 136, pp. 74877590, 2014. [65] P. Zhang, M. Wang, Y. Yang, T. Yao, H. Han, and L. Sun, Nano Energy, vol. 19, pp. 98107, 2016. [66] X. Zou, X. Huang, A. Goswami, R. Silva, B. R. Sathe, E. Mikmekova, and T. Asefa, Angew. Chem. Int. Ed., vol. 53, p. 4372, 2014. [67] J. Wang, D. Gao, G. Wang, S. Miao, H. Wu, J. Li, and X. Bao, J. Mater. Chem. A, vol. 2, p. 20067, 2014. 146

BIBLIOGRAPHY [68] D. Kong, J. J. Cha, H. Wang, H. R. Lee, and Y. Cui, Energy Environ. Sci., vol. 6, p. 3553, 2013. [69] S. Peng, L. Li, X. Han, W. Sun, M. Srinivasan, S. G. Mhaisalkar, F. Cheng, Q. Yan, J. Chen, and S. Ramakrishna, Angew. Chem. Int. Ed., vol. 53, p. 12594, 2014. [70] Z. Peng, D. Jia, A. M. Al-Enizi, A. A. Elzatahry, and G. Zheng, Adv. Energy Mater, vol. 5, p. 1402031, 2015. [71] E. J. Popczun, J. R. McKone, C. G. Read, A. J. Biacchi, A. M. Wiltrout, N. S. Lewis, and R. E. Schaak, J. Am. Chem. Soc., vol. 135, p. 9267, 2013. [72] Z. Huang, C. Lv, Z. Chen, Z. Chen, F. Tian, and C. Zhang, Nano Energy, vol. 12, pp. 666674, 2015. [73] J. Tian, Q. Liu, A. M. Asiri, and X. Sun, J. Am. Chem. Soc., vol. 136, pp. 74877590, 2014. [74] P. Los and A. Lasia, J. Electroanal. Chem., vol. 333, pp. 115125, 1992. [75] E. Ndzebet and O. Savadogo, Int. J. Hydrogen Energy, vol. 19-8, pp. 687 691, 1994. [76] J. J. Borodzilqski and A. Lasia, J. Appl. Electrochem., vol. 24, pp. 1267 1275, 1994. [77] H. Vrubel and X. Hu, Angew. Chem. Int. Ed., vol. 51, pp. 12 70312 706, 2012. [78] J. O. Bockris and T. Otagawa, J. Phys. Chem., vol. 87, p. 29601, 1983. [79] U. Demirci and P. Miele, Energy and Environmental Science, vol. 2, pp. 627637, 2009. [80] F. Stephens, V. Pons, and R. Baker, Dalton Transactions, vol. 25, pp. 2613 2626, 2007. [81] A. Baiker, Faraday Discussions of the Chemical Society, vol. 87, pp. 239251, 1989. [82] Z. Wu and S. Ge, Catalysis Communications, vol. 13, pp. 4043, 2011. [83] R. Fernandes, N. Patel, A. Miotello, R. Jaiswal, and D. C. Kothari, Int. J. Hydrogen Energy, vol. 37, pp. 23972406, 2012. 147

BIBLIOGRAPHY [84] R. Fernandes, N. Patel, A. Miotello, and D. C. Kothari, Int. J. Hydrogen Energy, vol. 36, pp. 13 37913 391, 2011. [85] N. Patel, R. Fernandes, and A. Miotello, Journal of Catalysis, vol. 271, pp. 315324, 2010. [86] D. Tong, W. Chu, Y. Luo, H. Chen, and X. Ji, Journal of Molecular Catalysis A: Chemical, vol. 269, pp. 149157, 2007. [87] N. Patel, R. Fernandes, N. Bazzanella, and A. Miotello, Catal. Today, vol. 170, pp. 2026, 2011. [88] N. Patel, R. Fernandes, A. Santini, and A. Miotello, International Journal of Hydrogen Energy, vol. 37, pp. 20072013, 2012. [89] I. Yuranov, P. Moeckli, E. Suvorova, P. Buat, L. Kiwi-Minsker, and A. Renken, Journal of Molecular Catalysis A: Chemical, vol. 192, pp. 239 251, 2003. [90] J. Beck, J. Vartuli, W. Roth, M. Leonowicz, C. Kresge, K. Schmitt, C. Chu, D. Olsen, E. Sheppard, S. McCullen, J. Higgins, and J. Schlenker, Journal of the American Chemical Society, vol. 114, pp. 10 83410 843, 1992. [91] Q. Huo, D. Margolese, U. Ciesla, D. Demuth, P. Feng, T. Gier, P. Sieger, A. Firouzi, B. Chemelka, F. Schuth, and G. Stucky, Chemistry of Materials, vol. 6, pp. 11761191, 1994. [92] V. Kocherbitov and V. Alfredsson, Langmuir, vol. 27, pp. 38893897, 2011. [93] T. B. K. Moller, Chemistry of Materials, vol. 109, pp. 29502963, 1998. [94] G. Attard, S. Leclerc, S. Maniguet, A. Russell, I. Nandhakumar, and P. Bartlett, Electrochim. Acta, vol. 13, pp. 14441446, 2001. [95] H. Shi, L. Zhang, and W. Cai, Materials Research Bulletin, vol. 35, p. 1689, 2000. [96] Y. Han, J. Kim, and G. Stucky, Chemistry of Materials, vol. 12, pp. 2068 2069, 2000. [97] X. Chen, S. Wang, J. Zhuang, M. Qiao, K. Fan, and H. He, Journal of Catalysis, vol. 227, pp. 419427, 2004. [98] H. Li, J. Liu, H. Yang, and H. Li, Chinese Journal of Chemistry, vol. 27, pp. 23162322, 2009. 148

BIBLIOGRAPHY [99] D. Zhao, J. Feng, Q. Huo, N. Melosh, G. Fredrickson, B. Chmelka, and G. Stucky, Science, vol. 279, pp. 548552, 1998. [100] K. Sing, D. Everett, R. Haul, L. Moscou, R. Pierotti, J. Rouquerol, and T. Siemieniewska, Pure and Applied Chemistry, vol. 57, p. 603, 1985. [101] D. Zhao, Q. Huo, J. Feng, B. Chmelka, and G. Stucky, Journal of the American Chemical Society, vol. 120, pp. 60246036, 1998. [102] Y. Luo, Z. Hou, R. Li, and X. Zheng, Microporous and Mesoporous Materials, vol. 109, pp. 585590, 2008. [103] D. Tong, X. Zeng, W. Chu, D. Wang, and P. Wui, Journal of Materials Science, vol. 45, pp. 28622867, 2010. [104] A. Figen and B. Coskuner, International Journal of Hydrogen Energy, vol. 33, pp. 24622467, 2008. [105] H. Li, W. Wang, H. Li, and J. Deng, Journal of Catalysis, vol. 194, pp. 211221, 2000. [106] N. Mohajeri, A. Raissi, and O. Adebiyi, Journal of Power Sources, vol. 167, pp. 482485, 2007. [107] N. Patel, R. Fernandes, G. Guella, A. Kale, A. Miotello, B. Patton, and C. Zanchetta, Journal of Physical Chemistry C, vol. 112, pp. 69686976, 2008. [108] M. Zahmakiran and S. Ozkar, Applied Catalysis B: Environmental, vol. 891, pp. 104110, 2009. [109] C. Yao, L. Zhuang, Y. Cao, X. Ai, and H. Yang, International Journal of Hydrogen Energy, vol. 33, pp. 24622467, 2008. [110] F. Durap, M. Zahmakiran, and S. Ozkar, International Journal of Hydrogen Energy, vol. 34, pp. 72237230, 2009. [111] Q. Xu and M. Chandra, Journal of Power Sources, vol. 163, pp. 364370, 2006. [112] H. Li, H. Yang, and H. Li, Journal of Catalysis, vol. 251, pp. 233238, 2007. [113] H. Li, X. Chen, M. Wang, and Y. Xu, Applied Catalysis A-General, vol. 225, pp. 117130, 2002. 149

BIBLIOGRAPHY [114] D. Tong, W. Chu, Y. Luo, H. Chen, and X. Ji, J.l Mol. Catal. A: Chem., vol. 269, pp. 149157, 2007. [115] H. Li, H. Yang, and H. Li, J. Catalysis, vol. 251, pp. 233238, 2007. [116] D. Tong, X. Han, W. Chu, H. Chen, and X. Ji, Mat. Res. Bulletin, vol. 43, pp. 13271326, 2008. [117] S. C. Amendola, S. L. Sharp-Goldman, M. S. Janjua, N. C. Spencer, M. T. Kelly, P. J. Petillo, and M. Binder, Int. J. Hydrogen Energy, vol. 25, pp. 969975, 2000. [118] Z. P. Li, N.Morigazaki, B. H. Liu, and S. Suda, J. Alloys Compd., vol. 349, pp. 232236, 2003. [119] A. Baiker, Faraday Discussions of the Chemical Society, vol. 87, pp. 239251, 1989. [120] R. IUPAC, Pure Applied Chem., vol. 66, p. 1739, 1994. [121] IUPAC, Pure Applied Chem., vol. 57, p. 603, 1985. [122] L. Qi, J. Ma, H. Cheng, and Z. Zhao, Chem. Mater., vol. 10, pp. 16231626, 1998. [123] Q. Huo, D. Margolese, U. Ciesla, D. Demuth, P. Feng, T. Gier, P. Sieger, A. Firouzi, B. Chemelka, F. Schuth, and G. Stucky, Chem. Mater., vol. 6, pp. 11761191, 1994. [124] Y. Kojima, K. Suzuki, K. Fukumoto, M. Sasaki, T. Yamamoto, Y. Kawai, and H. Hayashi, Int. J. Hydrogen Energy, vol. 27, pp. 10291034, 2002. [125] B. H. Liu, Z. P. Li, and S. Suda, J. Alloys Compd, vol. 415, pp. 288293, 2006. [126] S. U. Jeong, R. K. Kim, E. A. Cho, H. J. Kim, S. W. Nam, I. H. Oh, S. A. Hong, and S. H. Kim, J. Power Sources, vol. 144, pp. 129134, 2005. [127] P. Krishnan, T. H. Yang, W. Y. Lee, and C. S. Kim, J. Power Sources, vol. 143, pp. 1723, 2005. [128] J. S. Zhang, W. N. Delgass, T. S. Fisher, and J. P. Gore, J. Power Sources, vol. 164, pp. 772781, 2007. [129] N. Patel, R. Fernandes, G. Guella, and A. Miotello, Appl. Catal. B: Env., vol. 137-143, p. 95, 2010. 150

BIBLIOGRAPHY [130] J. Zhao, H. Ma, and J. Chen, Int. J. Hydrogen Energy, vol. 32, pp. 4711 4716, 2007. [131] J. Lee, K. Y. Kong, C. R. Jung, E. Cho, S. P. Yoon, J. Han, T. G. Lee, and S. W. Nam, Catal. Today, vol. 120, pp. 305310, 2007. [132] P. Krishnan, S. G. Advani, and A. K. Prasad, Appl. Catal. B: Env., vol. 86, pp. 137144, 2009. [133] A. Garron, D. Swierczynski, S. Bennici, and A. Auroux, Int. J. Hydrogen Energy, vol. 34, pp. 11851199, 2009. [134] P. Krishnan, S. Advani, and A. Prasad, Int. J. Hydrogen Energy, vol. 33, pp. 70957102, 2008. [135] T. Pfeil, T. Pourpoint, and L. Groven, Int. J. Hydrogen Energy, vol. 39, pp. 21492159, 2014. [136] J. Mahmood, S. Jung, S. J. Kim, J. Park, J. W. Yoo, and J. B. Baek, Chem. Mater., vol. 27, pp. 48604864, 2015. [137] V. Simagina, O. Komova, A. Ozerova, O. V. Netskina, G. Odegova, D. Kellerman, O. Bulavchenko, and A. Ishchenko, Appl. Catal. A, vol. 394, pp. 8692, 2011. [138] N. Patel, G. Guella, A. Kale, A. Miotello, B. Patton, C. Zanchetta, L. Mirenghi, and P. Rotolo, Appl. Catal. A, vol. 323, pp. 1824, 2007. [139] R. Edla, N. Patel, Z. E. Koura, R. Fernandes, N. Bazzanella, and A. Miotello, Appl. Surf. Sci., vol. 302, pp. 105108, 2014. [140] N. Patel, A. Miotello, and V. Bello, Appl. Catal. B: Env., vol. 103, pp. 3138, 2011. [141] T. Warang, N. Patel, A. Santini, N. Bazzanella, A. Kale, and A. Miotello, Appl. Catal. A, vol. 423-424, pp. 2127, 2012. [142] T. Warang, N. Patel, R. Fernandes, N. Bazzanella, and A. Miotello, Appl. Catal. B: Env., vol. 132-133, pp. 204211, 2013. [143] N. Patel, A. Santini, V. Bello, G. Mattei, and A. Miotello, Surf. Coat. Technol., vol. 235, pp. 784791, 2013. [144] N. Patel, R. Fernandes, G. Guella, A. Kale, A. Miotello, B. Patton, and C. Zanchetta, J. Phys. Chem. C, vol. 112, pp. 69686976, 2008. 151

BIBLIOGRAPHY [145] O. Akdim, R. Chamoun, U. B. Demirci, Y. Zaatar, A. Khoury, and P. Miele, Int. J. Hydrogen Energy, vol. 36, pp. 14 52714 533, 2011. [146] N. Malvadkar, S. Park, M. U. M. Donald, H. Wang, and M. C. Demirel, J. Power Sources, vol. 182, pp. 323328, 2008. [147] D. Xu, H. Zhang, and W. Yeb, Catal. Commun., vol. 8, pp. 17671771, 2007. [148] O. Komova, V. Simagina, O. Netskina, D. Kellerman, A. Ishchenko, and N. Rudina, Catal. Today, vol. 138, pp. 260265, 2008. [149] S. Laureti, E. Agostinelli, G. Scavia, G. Varvaro, V. Albertini, A. Generosi, B. Paci, A. Mezzi, and S. Kaciulis, Appl. Surf. Sci., vol. 254, pp. 51115115, 2008. [150] A. Miotello and R. Kelly, Appl. Phys. Lett., vol. 67, pp. 35353537, 1995. [151] M. Wua, P. K. Shena, Z.Weib, S. Songa, and M. Nie, J. Power Sources, vol. 166, pp. 310316, 2007. [152] M. C. Tavares, S. A. S. Machado, and L. H. Mazo, Electrochim. Acta, vol. 46, pp. 43594369, 2001. [153] Y. Sun, C. Liu, D. C. Grauer, J. Yano, J. R. Long, P. Yang, and C. J. Chang, J. Am. Chem. Soc., vol. 135, pp. 17 69917 702, 2013. [154] T. W. Lin, C. J. Liu, and J. Y. Lin, Appl. Catal. B: Environ., vol. 134, pp. 7582, 2013. [155] K.Wang, D. Xi, C. Zhou, Z. Shi, H. Xia, G. Liu, and G. Qiao, J. Mater. Chem. A, vol. 3, pp. 94159420, 2015. [156] P. Xiao, Y. Yan, X. Ge, Z. Liu, J. Y. Wang, and X. Wang, Appl. Catal. B: Environ., vol. 154, pp. 232237, 2014. [157] W. Cui, Q. Liu, Z. Xing, A. M. Asiri, K. A. Alamry, and X. Sun, Appl. Catal. B: Environ., vol. 164, pp. 144150, 2015. [158] Z. Huang, Z. Chen, Z. Chen, C. Lv, M. G. Humphrey, and C. Zhang, Nano Energy, vol. 9, pp. 373382, 2014. [159] P. Los and A. Lasia, J. Electroanal. Chem., vol. 333, pp. 115125, 1992. [160] E. Ndzebet and O. Savadogo, Int. J. Hydrogen Energy, vol. 19-8, pp. 687 691, 1994. 152

BIBLIOGRAPHY [161] J. J. Borodzilqski and A. Lasia, J. Appl. Electrochem., vol. 24, pp. 1267 1275, 1994. [162] H. Vrubel and X. Hu, Angew. Chem. Int. Ed., vol. 51, pp. 12 70312 706, 2012. [163] N. Patel, A. Miotello, and V. Bello, Appl. Catal. B: Environ., vol. 103, pp. 3138, 2011. [164] N. Patel, R. Fernandes, G. Guella, A. Kale, A. Miotello, B. Patton, and C. Zanchetta, J. Phys. Chem. C, vol. 112, pp. 69686976, 2008. [165] Y. Wang, X. Ai, Y. Cao, and H. Yang, Electrochem. Comm., vol. 6, pp. 780784, 2004. [166] S. Cobo, J. Heidkamp, P. Jacques, J. Fize, V. Fourmond, L. Guetaz, B. Jousselme, V. Ivanova, H. Dau, S. Palacin, M. Fontecave, and V. Artero, Nat. Mater., vol. 11, pp. 802807, 2012. [167] B. Delley, J. Chem. Phys., vol. 92, pp. 508513, 1990. [168] H. Li, X. Chen, M. Wang, and Y. Xu, Appl. Catal. A: Gen, vol. 225, pp. 117130, 2002. [169] H. Li, Y. Wua, H. Luo, M. Wang, and Y. Xu, J. Catal., vol. 214, pp. 1525, 2003. [170] S. Carenco, D. Portehault, C. Boissiere, N. Mezailles, and C. Sanchez, Chwm. Rev., vol. 113, pp. 79818065, 2013. [171] J. Cooper, T. Gibb, N. Greenwood, and R. Paris, Trans. Faraday Soc., vol. 60, p. 2097, 1964. [172] Y. Okamoto, Y. Nitta, T. Imanaka, and S. Teranishi, J. Chem. Soc., Faraday Trans., vol. 75, p. 2027, 1979. [173] B. Xiao, J. Xing, S. Ding, and W. Su, Phys. B, vol. 403, p. 1723, 2008. [174] C. Zhou, J. Xing, B. Xiao, J. Feng, X. Xie, and Y. Chen, Comput. Mater. Sci., vol. 44, p. 1056, 2009. [175] P. Tran, M. Nguyen, S. Pramana, A. Bhattacharjee, S. Chiam, J. Fize, M. Field, V. Artero, L. Wong, J. Loo, and J. Barber, Energy Environ. Sci., vol. 5, pp. 89128916, 2012. 153

BIBLIOGRAPHY [176] L. Chen, M. Wangi, K. Han, P. Zhang, F. Gloaguen, and L. Sun, Energy Environ. Sci., vol. 7, pp. 329334, 2014. [177] P. Tran, S. Chiam, P. P. Boix, Y. Ren, S. Pramana, J. Fize, V. Artero, and J. Barber, Energy Environ. Sci., vol. 6, pp. 24522459, 2013. [178] Z. Pu, Q. Liu, A. M. Asiri, A. Y. Obaid, and X. Sun, J. Power Sources, vol. 263, pp. 181185, 2014. [179] D. Merki, H. Vrubel, L. Rovelli, S. Fierro, and X. Hu, Chem. Sci., vol. 3, pp. 25152525, 2012. [180] E. Popczun, J. McKone, C. Read, A. Biacchi, A. Wiltrout, N. Lewis, and R. Schaak, J. Am. Chem. Soc., vol. 135, pp. 92679270, 2013. [181] W. Chen, C. Wang, K. Sasaki, N. Marinkovic, W. Xu, J. Muckerman, Y. Zhu, and R. Adzic, Energy Environ. Sci., vol. 6, pp. 943951, 2013. [182] Y. Li, H. Wang, L. Xie, Y. Liang, G. Hong, and H. Dai, J. Am. Chem. Soc., vol. 133, pp. 72967299, 2011. [183] R. Fernandes, N. Patel, A. Kale, A. Miotello, and M. Filippi, J. Mol. Cat. A: Chem., vol. 298, pp. 16, 2009. [184] S. Gupta, N. Patel, A. Miotello, and D. C. Kothari, J. Power Sources, vol. 279, pp. 620625, 2015. [185] N. N. Greenwoord, V. Parish, and P. Thornton, Q. Rev. Chem. Soc., vol. 20, pp. 441464, 1966. [186] Y. Okamoto, Y. Nitta, T. Imanaka, and S. Teranishi, J. Chem. Soc. Faraday. Trans. 1, vol. 75, pp. 20272039, 1979. [187] C. G. Morales-Guio, L. A. Stern, and X. Hu, Chem. Soc. Rev., vol. 43, pp. 65556569, 2014. [188] Y. C. Luo, Y. H. Liu, Y. Hung, X. Y. Liu, and C. Y. Mou, Int. J. Hydrogen Energy, vol. 38, pp. 72807290, 2013. [189] H. Kadomatsu, F. ishii, and H. Fujiwara, J. Phys. Soc. Japan, vol. 47, pp. 10781085, 1979. [190] B. Shen, S. Wei, K. Fang, and J. F. Deng, Appl. Phys. A, vol. 65, pp. 295299, 1997. [191] M. Zeng and Y. Li, J. Mater. Chem. A, vol. 3, pp. 14 94214 962, 2015. 154

BIBLIOGRAPHY [192] P. Zhang, M. Wang, Y. Yang, T. Yao, H. Han, and L. Sun, Nano Energy, vol. 19, pp. 98107, 2016. [193] Y. Sun, C. Liu, D. C. Grauer, J. Yano, J. R. Long, P. Yang, and C. J. Chang, J. Am. Chem. Soc., vol. 135, pp. 17 69917 702, 2013. [194] J. Tian, Q. Liu, A. M. Asiri, and X. Sun, J. Am. Chem. Soc., vol. 136, pp. 74877590, 2014. [195] J. Tian, Q. Liu, Y. Liang, Z. Xing, A. M. Asiri, and X. Sun, ACS Appl. Mater. Interfaces, vol. 6, pp. 20 57920 584, 2014. [196] Z. Peng, D. Jia, A. M. Al-Enizi, A. A. Elzatahry, and G. Zheng, Adv. Energy Mater., vol. 5, p. 1402031. [197] Z. Pu, Q. Liu, P. Jiang, A. M. Asiri, A. Y. Obaid, and X. Sun, Chem. Mater., vol. 26, pp. 43264329, 2014. [198] H. W. Liang, S. Bruller, R. Dong, J. Zhang, X. Feng, and K. Mullen, Nat. Comm., vol. 6, 2015. [199] E. Popczun, C. Read, C. W. Roske, N. Lewis, and R. Schaak, Angew. Chem. Int. Ed., vol. 53, pp. 54275430, 2014. [200] S. Cobo, J. Heidkamp, P. Jacques, J. Fize, V. Fourmond, L. Guetaz, B. Jousselme, V. Ivanova, H. Dau, S. Palacin, M. Fontecave, and V. Artero, Nat. Mater., vol. 11, pp. 802807, 2012. [201] D. Merki, H. Vrubel, L. Rovelli, S. Fierro, and X. Hu, Chem. Sci., vol. 3, pp. 25152525, 2012. [202] W. Chen, C. Wang, K. Sasaki, N. Marinkovic, W. Xu, J. Muckerman, Y. Zhu, and R. Adzic, Energy Environ. Sci., vol. 6, pp. 943951, 2013. [203] W. Chen, K. Sasaki, C. Ma, A. I. Frenkel, N. Marinkovic, J. Muckerman, Y. Zhu, and R. Adzic, Angew. Chem. Int. Ed., vol. 51, pp. 61316135, 2012. [204] H. Vrubel and X. Hu, Angew. Chem. Int. Ed., vol. 51, pp. 12 70312 706, 2012. [205] Z. Xing, Q. Liu, W. Xing, A. M. Asiri, and X. Sun, ChemSusChem, vol. 8, pp. 18501855, 2015. [206] H. Fei, Y. Yang, Z. Peng, G. Ruan, Q. Zhong, L. Li, E. L. G. Samual, and J. M. Tour, ACS Appl. Mater. Interfaces, vol. 7, pp. 80838087, 2015. 155

BIBLIOGRAPHY [207] T. N. Ye, L. B. Lv, M. Xu, B. Zhang, K. X. Wang, J. Su, X. H. Li, and J. S. Chen, Nano Energy, vol. 15, pp. 335342, 2015. [208] Z. Huang, C. Lv, Z. Chen, Z. Chen, F. Tian, and C. Zhang, Nano Energy, vol. 12, pp. 666674, 2015. [209] M. Zeng and Y. Li, J. Mater. Chem. A, vol. 14942-14962, p. 15032, 2015. [210] C. Tang, A. M. Asiri, Y. Luo, and X. Sun, ChemNanoMat, vol. 1, pp. 558 561, 2015. [211] Y. Yan, B. Y. Xia, X. Ge, Z. Liu, A. Fisher, and X. Wang, Chem. Eur. J., vol. 21, pp. 17, 2015. [212] T. Liu, Q. Liu, A. M. Asiri, Y. Luo, and X. Sun, Chem. Commun., vol. 51, pp. 16 68316 686, 2015. [213] Y. P. Zhu, Y. P. Liu, T. Z. Ren, and Z. Y. Yuan, Adv. Funct. Mater., vol. 25, pp. 73377347, 2015. [214] Z. Zhang, J. Hao, W. Yang, and J. Tang, RSC Adv., vol. 6, pp. 96479655, 2016. [215] P. Zhang, M. Wang, Y. Yang, T. Yao, H. Han, and L. Sun, Nano Energy, vol. 19, pp. 98107, 2016. [216] S. Gupta, N. Patel, R. Fernandes, R. Kadrekar, A. Dashora, A. K. Yadav, D. Bhattacharyya, S. N. Jha, A. Miotello, and D. C. Kothari, Appl. Cat. B: Environ., 2016. [217] J. Masa, P. Weide, D. Peeters, I. Sinev, W. Xia, Z. Sun, C. Somsen, M. Muhler, and W. Schuhmann, Adv. Energy Mater., p. 1502313, 2016. [218] S. Gupta, N. Patel, A. Miotello, and D. C. Kothari, J. Power Sources, vol. 279, pp. 620625, 2015. [219] H. Vrubel and X. Hu, Angew. Chem. Int. Ed., vol. 51, pp. 12 70312 706, 2012. [220] D. Wang, D. Zhang, C. Tang, P. Zhou, Z. Wu, and B. Fang, Catal. Sci. Technol., vol. 6, pp. 19521956. [221] J. Hao, W. Yang, Z. Zhang, and J. Tang, Nanoscale, vol. 7, pp. 11 055 11 062, 2015. [222] J. S.-J. et al, Nat. Mater., vol. 15, pp. 197203, 2016. 156

BIBLIOGRAPHY [223] H. Zhang, Y. Li, T. Xu, J. Wang, Z. Huo, P. Wana, and X. Sun, J. Mater. Chem. A, vol. 3, pp. 15 02015 023, 2015. [224] H. W. Liang, S. Bruller, R. Dong, J. Zhang, X. Feng, and K. Mullen, Nat. Comm., vol. 6, 2015. [225] Y. Liu, G. Yu, G. D. Li, Y. Sun, T. Asefa, W. Chen, and X. Zou, Angew. Chem. Int. Ed., vol. 54, pp. 16, 2015. [226] X. Liu, S. Cui, Z. Sun, and P. Du, Chem. Comm., vol. 51, pp. 12 95412 957. [227] Z. Pu, Q. Liu, A. M. Asiri, and X. Sun, ACS Appl. Mater. Interfaces, vol. 6, pp. 21 87421 879, 2014. [228] J. Tian, Q. Liu, Y. Liang, Z. Xing, A. M. Asiri, and X. Sun, ACS Appl. Mater. Interfaces, vol. 6, pp. 20 57920 584, 2014. [229] S. Cobo, J. Heidkamp, P. Jacques, J. Fize, V. Fourmond, L. Guetaz, B. Jousselme, V. Ivanova, H. Dau, S. Palacin, M. Fontecave, and V. Artero, Nat. Mater., vol. 11, pp. 802807, 2012. [230] M. G. et.al., Nat. Comm., vol. 5, pp. 16, 2014. [231] K. Zhang, Y. Zhao, D. Fu, and Y. Chen, J. Mater. Chem. A, vol. 3, pp. 57835788, 2015. [232] Z. Peng, D. Jia, A. M. Al-Enizi, A. A. Elzatahry, and G. Zheng, Adv. Energy Mater., vol. 5, p. 1402031. [233] Y. Liang, Q. Liu, A. M. Asiri, X. Sun, and Y. Luo, ACS Catal., vol. 4, pp. 40654069, 2014. [234] J. Tian, Q. Liu, A. M. Asiri, and X. Sun, J. Am. Chem. Soc., vol. 136, pp. 75877590, 2014. [235] E. Popczun, C. Read, C. W. Roske, N. Lewis, and R. Schaak, Angew. Chem. Int. Ed., vol. 53, pp. 54275430, 2014. [236] Y. Sun, C. Liu, D. C. Grauer, J. Yano, J. R. Long, P. Yang, and C. J. Chang, J. Am. Chem. Soc., vol. 135, pp. 17 69917 702, 2013. [237] C. C. L. McCrory, S. Jung, J. C. Peters, and T. F. Jaramillo, J. Am. Chem. Soc., vol. 135, pp. 16 97716 987, 2013. [238] M. Gao, X. Cao, Q. Gao, Y. Xu, Y. Zheng, J. Jiang, and S. Yu, ACS Nano, vol. 8, pp. 39703978, 2014. 157

BIBLIOGRAPHY [239] L. Wu, Q. Li, C. H. Wu, H. Zhu, A. Mendoza-Garcia, B. Shen, J. Guo, and S. Sun, J. Am. Chem. Soc., vol. 137, pp. 70717074, 2015. [240] A. J. Esswein, M. J. McMurdo, P. N. Ross, A. T. Bell, and T. D. Tilley, J. Phys. Chem. C, vol. 113, pp. 15 06815 072, 2009. [241] J. Ren, M. Antonietti, and T. P. Fellinger, Adv. Energy Mater., vol. 5, p. 1401660, 2015. [242] M. Ledendecker, G. Clavel, M. Antonietti, and M. Shalom, Adv. Func. Mat., vol. 25, pp. 393399. [243] S. Trasatti, Electrochim. Acta, vol. 29, p. 15032, 1984. [244] J. Rossmeisl, Z. W. Qu, H. Zhu, G. J. Kroes, and J. K. Norskov, J. Electroanal. Chem., vol. 607, p. 83, 2007. [245] Y. Liang, Q. Liu, Y. Luo, X. Sun, Y. He, and A. M. Asiri, Electrochim. Acta, vol. 190, pp. 360364, 2016. [246] J. Tian, N. Cheng, Q. Liu, X. Sun, Y. He, and A. M. Asiri, J. Mater. Chem. A, vol. 3, pp. 20 05620 059, 2015. [247] X. Li, Z. Niu, J. Jiang, and L. Ai, J. Mater. Chem. A, vol. 4, pp. 32043209, 2016. [248] J. Chang, L. Liang, C. Li, M. Wang, J. Ge, C. Liu, and W. Xing, Green Chem., 2016. [249] B. You, N. Jiangn, M. Sheng, S. Gul, J. Yano, and Y. Sun, Chem. Mater., vol. 27, pp. 76367642, 2015. [250] M. Jiang, Y. Li, Z. Lu, X. Sun, and X. Duan, Inorg. Chem. Front., 2016. [251] N. Patel, A. Kale, and A. Miotello, Appl. Catal. B: Environ., vol. 111-112, pp. 178184, 2012. [252] N. Patel, R. Fernandes, and A. Miotello, J. Power Sources, vol. 188, pp. 411420, 2009. [253] H. Vrubel and X. Hu, Angew. Chem. Int. Ed., vol. 51, pp. 12 70312 706, 2012. [254] M. M. Hukovic and A. Jukic, Electrochim. Acta, vol. 45, pp. 41594170, 2000. 158

BIBLIOGRAPHY [255] C. Fan, D. L. Piron, A. Sleb, , and P. Paradis, J. Electrochem. Soc., vol. 141, pp. 382387, 1994. [256] Z. Pu, C. Tang, and Y. Luo, Int. J. Hydrogen Energy, vol. 40, pp. 50925098, 2015. [257] J. F. Callejas, J. M. McEnaney, C. G. Read, J. C. Crompton, A. J. Biacchi, E. J. Popczun, T. R. Gordon, N. S. Lewis, and R. E. Schaak, ACS Nano, vol. 8, pp. 11 10111 107, 2014. [258] H. Du, S. Gu, R. Liu, and C. M. Li, J. Power Sources, vol. 278, pp. 540545, 2015. [259] J. Shi, Z. Pu, Q. Liu, A. M. Asiri, J. Hu, and X. Sun, Electrochim. Acta, vol. 154, pp. 345351, 2015. [260] Y. Hou, B. Zhang, Z. Wen, S. Cui, X. Guo, Z. He, and J. Chen, J. Mater. Chem. A, vol. 2, pp. 13 79513 800, 2014. [261] C. Tang, Z. Pu, Q. Liu, A. M. Asiri, and X. Sun, Electrochim. Acta, vol. 153, pp. 508514, 2015. [262] L. Feng, G. Li, Y. Liu, Y. Wu, H. Chen, Y. Wang, Y. C. Zou, D. Wang, and X. Zou, ACS Appl. Mater. Interfaces, vol. 7, pp. 980988, 2015. [263] J. R. McKone, B. F. Sadtler, C. A. Werlang, N. S. Lewis, and H. B. Gray, ACS Catal., vol. 3, pp. 166169, 2013. [264] A. B. Laursen, K. R. Patraju, M. J. Whitaker, M. Retuerto, T. Sarkar, N. Yao, K. V. Ramanujachary, M.Greenblatt, and G. C. Dismukes, Energy Environ. Sci., vol. 8, pp. 10271034, 2015. [265] H. Fei, Y. Yang, Z. Peng, G. Ruan, Q. Zhong, L. Li, E. L. G. Samuel, and J. M. Tour, ACS Appl. Mater. Interfaces, vol. 7, pp. 80838087, 2015.

159

SYNOPSIS OF THE THESIS TO BE SUBMITTED TO THE

UNIVERSITY OF MUMBAI FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY IN PHYSICS IN THE FACULTY OF SCIENCE Title of the thesis

Synthesis of nanocatalysts and their applications to hydrogen production by electrolysis of water and hydrolysis of chemical hydrides

Name of the candidate

Gupta Suraj Gulab Sonkali

Name of the research guide

Dr. D.C. Kothari Professor, Department of Physics, University of Mumbai, Vidyanagari, Santacruz (East), Mumbai 400 098.

Name of the research co-guide

Dr. Nainesh Patel Assistant Professor, UGC-FRP Faculty, Department of Physics, University of Mumbai

Place of the research

Department of Physics, University of Mumbai, Vidyanagari, Santacruz (East) Mumbai-400098.

Number and date of registration

105 – 20/06/2013

Date of submission of synopsis

07/01/2016

Signature of the candidate

Signature of the guide

Signature of the co-guide

Signature of head of department

Synthesis of nanocatalysts and their applications to hydrogen production by electrolysis of water and hydrolysis of chemical hydrides Introduction Energy has been one of the primary needs of human beings since ages. Earlier, we relied on naturally available resources (fossil fuels, oil etc.) as sources of energy. In the past, these natural resources seemed plentiful but the scenario has changed signicantly over a period of time. Ever since the advent of industrialization, our energy demands have increased multi-fold and they keep on increasing exponentially. Another worrisome aspect of using fossil fuels is the large amount of greenhouse gases emitted by their use. These greenhouse emissions have already caused serious damage to our environment and keep on adding signicantly to global warming. This may eventually lead to catastrophic disaster that cannot be reversed. In the wake of above mentioned issues, all the countries have now started taking measures to avert this forthcoming energy crisis. Looking for alternatives to non-renewable natural resources is one of the key strategies towards this goal.

When we talk about

alternative resources, we expect them to be as ecient, if not more, as the presently used fossil fuels and other non-renewable energy resources. Thus, the present issue demands that the alternate resources - must be free of greenhouse emissions, must be inexpensive for large scale use and must rely on resources that are plentiful on earth. A lot of alternatives t this list - wind energy, solar energy, nuclear energy etc. In addition to these, use of hydrogen as an energy carrier has attracted major attention from researchers all round the globe. H2 as a fuel satises all the requirements of a primary energy carrier

1

and certainly has the potential to replace conventional non-renewable energy resources. In the past few decades, a lot of research has been focused on the sole purpose of realizing this idea of using H2 fuel on practical scale. The long foreseen idea of H2 based economy has now become a present day reality owing to the advances in renewable H2 production techniques. One of the major hurdles in using H2 as a fuel is its unavailability in free form.

Hydrogen as a gas doesn't exist on earth and has to be produced from

dierent sources. Traditional methods of hydrogen production involve steam reforming of natural gas where nal product contains various greenhouse gases (CO2 and CO) [1]. These poisonous emissions combined with consumption of fossil fuels are major concerns in the production of hydrogen. Renewable hydrogen can be produced in several ways: water electrolysis, biomass conversion, solar conversion and from chemical hydrides. Of these listed methods, biomass-to-hydrogen conversion is complex because of the technical details of the conversion processes [2], whereas the eciency of solar based processes are extremely low for practical uses [3]. Water and chemical hydrides, on the other hand, are preferred sources to produce H2 at room temperature without any emission of harmful gases with high conversion eciency. Hydrogen can be eciently produced either by water-splitting in an electrolyzer or by dissociation of chemical hydrides (NaBH4 & NH3 BH3 ). The rate of H2 production, from either technique, is a key factor in deciding their feasibility for particular applications. The rate of these reactions and thus that of H2 production can be increased multi-fold by employing suitable catalysts.

As catalysis is a surface phenomena, it is desirable

to use catalysts with high surface area that can provide more surface sites for reaction to take place.

In such a case, nanocatalysts are highly preferred owing to their nano-

scale architecture which lead to exceptional catalytic performance. Nevertheless, the key challenge lies in the synthesis of fairly dispersed and size-controlled nanoparticles (NPs). However, no general strategy exists for synthesizing NPs of various materials with narrow size distribution, tailored properties, and desired morphologies.

The development of a

technique, which is able to prepare ecient and size-controlled nanocatalysts, is thus a goal of great importance to avoid present trial and error approaches. On the other hand,

2

because of their high surface energies, NPs tend to agglomerate and/or change shape during the catalytic reactions or treatment at moderate temperatures, a process which could cause the loss of their initial activity and selectivity.

NPs partially embedded

in appropriate matrix, may be a solution to the agglomeration problem.

Pulsed laser

deposition (PLD) has been widely investigated [4] to synthesize NPs assembled in the form of thin lms on various substrates. Thus the NPs will remain separated and stable during the course of the catalytic reaction. In addition to PLD, nanocatalyst thin lms can also be formed by chemical deposition techniques such as electroless deposition and electroplating which provide convenient and cost-eective alternatives. Chemical hydrides (such as NaBH4 , NH3 BH3 , LiH, KBH4 , etc.), appear promising materials for storing hydrogen owing to their high volumetric and gravimetric hydrogen storage capacity as well as for their stability. Pure hydrogen can be generated from these hydrides at room temperatures and therefore they are expected to be suitable for on-board as well as o-board applications. To eectively control the hydrolysis of chemical hydrides, an ecient catalyst is needed. Noble metal catalysts have been widely investigated and are found to be suitable for such purposes [5]. However, the use of noble metals needs to be minimized owing to their high cost and scarcity. This has led to the search for nonnoble catalysts, which are more earth-abundant and have eciency comparable to their noble counterparts. Transition metal (Co, Ni) catalysts have been investigated in the past [6, 7] and are found to be ecient in accelerating the hydrogen production rate. Cobalt boride (Co-B) has been a promising candidate for hydrogen generation [8, 9] considering its high activity and low cost. However, during Co-B synthesis, particle agglomeration takes place which reduces the eective surface area and limits the catalytic activity. In the past, researchers have used innovative synthesis methods to avoid agglomeration such as doping with transition metals [6, 10], using organic templates [11] or supporting the catalyst on high surface area materials such as rough carbon [12]. In the present thesis, we have reported various new strategies to improve the catalytic activity of Co-B: supporting on mesoporous silica substrates (to reduce the agglomeration), using mesoporous Co-B (to increase the surface area) and using nanoparticle assembled coatings (NPACs) (providing

3

high surface area and good stability against aggregation). Water electrolysis is an old-known process to produce pure hydrogen (and oxygen). However, today it contributes to only 4% of the total world hydrogen production. This is majorly due to the energy required to split water into hydrogen and oxygen.

Also,

the usage of noble metals (Pt, Pd, Ru) [13, 14] as electrocatalysts adds to the overall cost of H2 production.

It is therefore essential to search for electrocatalysts made up

of inexpensive and abundant materials that can imitate the eciencies of noble group elements and also withstand dierent pH conditions.

In the past few years, there has

been a plethora of earth-abundant electrocatalytic materials for hydrogen evolution with excellent eciency and stability [15, 16, 17, 18, 19, 20]. A lot of recent research is focused on using transition metals (Fe, Co, Ni, Mo, etc.)

coupled with non-metals (P, N, C)

and chalcogenides (S, Se) such as MoS [15], CoS [16], CoSe2 [17], Mo2 C [18], MoP [19], Co2 P [20] and so on.

These metal/non-metal or metal/chalcogenide compounds have

turned out to be highly ecient catalysts for hydrogen evolution reaction (HER) with many of them being active in multiple pH environments as well.

However, there have

been very few reports on transition metal borides as HER active material. About two decades ago, Lasia and Los [21] reported amorphous Ni2 B electrocatalyst for HER in alkaline medium. Subsequently, there were more reports on electrodeposited Ni2 B [22] and doped Ni2 B [23] catalysts for alkaline water electrolysis. Since then, the interest in metal boride electrocatalysts almost vanished until Vrubel and Hu reported MoB [24] as HER active material in both acidic and basic conditions.

In the present thesis, we

report our work carried out on binary and ternary transition metal borides for HER in various pH media. We have reported Cobalt boride (Co-B) as an excellent electrocatalyst for HER active in wide pH range (4-9). We have also reported the vast improvement in activity and stability of Co-B electrocatalyst obtained after introducing other transition metals, specically Ni and Mo in Co-B. The role played by each element in enhancing the HER rate was studied in details by using numerous characterization tools such as XPS, XAS, BET, TEM, SEM, XRD and theoretical calculations. This thesis presents a detailed study on the various strategies stated above to enhance

4

H2 production by both the techniques adopted i.e. hydrolysis of chemical hydrides and water electrolysis. The present thesis will include a brief introduction and in-depth literature survey followed by detailed chapters on the various strategies employed and then a summary of the work and future plans based on the results presented. A short summary of the dierent strategies and the subsequent chapters is presented below.

1

Hydrolysis of chemical hydrides

1.1 Co-B catalyst supported over mesoporous silica for hydrogen production by catalytic hydrolysis of Ammonia Borane Three kinds of mesoporous silica (MCM-41, FSM-16 and SBA-15) of dierent pore size and texture were synthesized by templating method. Co-B nanoparticle catalysts were supported over these mesoporous silica by impregnationreduction method in order to study the eect of support pore structure on the catalytic properties in H2 production by hydrolysis of Ammonia Borane (AB). TEM and N2 adsorptiondesorption isotherm results clearly revealed that size, dispersion degree and location of Co-B particle is aected by the pore texturing of the support. It also showed that the catalyst particle acquires directly the size of the support pores only for SBA-15 whereas there is no correlation of the particle size and pore size for MCM-41 and FSM-16. Co-B supported over SBA-15 silica was found to be the most active catalyst as inferred from the observed hydrogen generation rates in the hydrolysis reaction compared to that produced by MCM-41 and FSM-16 supported catalyst.

Higher activity for SBA-15 support is mainly attributed

to the geometrical connement of Co-B particles within the pores which creates smaller Co-B particles (6 nm) with uniform size distribution and higher degree of dispersion as compared to MCM-41 and FSM-16 supports where Co-B particles lie on the external surface with broad size distribution. Open and interconnected pores of SBA-15 can also provide easy passage for reactant and product during the course of reaction. The Co-B particles supported in the interconnected pores of SBA-15 reduces the activation energy barrier related to the hydrolysis process of AB more eciently than that established with

5

MCM-41 and FSM-16 supported catalyst. Most importantly, the thicker pore walls of SBA-15 assist in avoiding the agglomeration of Co-B particles and even provide high stability at elevated temperatures (873 K) at which unsupported Co-B catalyst gets completely destroyed.

Thus, mesoporous silica substrates provide an eective way to

increase the surface area and stability of the desired nanocatalysts thereby enhancing the reaction rates.

1.2 Mesoporous Co-B nanocatalyst for ecient hydrogen production by hydrolysis of Sodium Borohydride Taking a cue from the synthesis technique used for mesoporous silica, we synthesized two types of mesoporous Co-B nanocatalysts by the reduction of cobalt chloride with sodium borohydride (SBH) in the presence of cationic and non-ionic surfactant templates, namely n-cetyl-trimethyl-ammonium bromide (CTAB) and pluronic (P123) respectively. Nitrogen adsorption-desorption isotherms revealed the presence of slit-like pores on the catalyst surface which provide high eective surface area. These surface enhanced catalysts were tested for hydrogen production by hydrolysis of sodium borohydride. The mesoporous Co-B catalysts showed much higher activity for hydrogen production (4 times) in comparison to the non-porous Co-B, which can be attributed to the higher surface area of the mesoporous structures. Co-B/P123 catalyst showed the highest hydrogen generation rate owing to the presence of wide uniform pores which facilitated easier interaction of the reactants to release hydrogen. A lack of stability in the pore structures was observed at elevated temperatures for both the mesoporous Co-B catalyst. Using mesoporous silica templates limit the amount of catalyst that can be loaded, whereas making the catalyst itself mesoporous provides a way to increase the eective surface area of the catalyst without using any foreign substrates.

6

1.3 Co O nanoparticles assembled coatings for enhanced H production from hydrolysis of Sodium Borohydride 3

4

2

When catalysts are used in the form of powder, separating the catalyst for their reuse is a major hurdle to overcome. On the other hand, catalyst in form of coatings or thin lm will serve as an environmentally friendly green catalyst for easy recovery, reuse and can function as ON/OFF switch for the reaction. Mainly, nanoparticle assembled coatings (NPACs) are best suited catalyst providing high surface area and good stability against aggregation owing to the immobilized and adherent nature of NPs on the suitable substrate. With this motivation, nanocatalyst in form of nanoparticles assembled coatings (NPACs) of Co3 O4 were synthesized by pulsed laser deposition (PLD) with optimized parameters. Phase explosion phenomena occurring at high laser uences produce nanoparticles (NPs) with average size of ~5 nm, which are randomly arranged in form of coating on the substrate surface with narrow size distribution (3-10 nm) and low degree of agglomeration. In comparison with chemically synthesized Co3 O4 crystalline powder, the NPACs deposited by PLD showed signicantly higher catalytic activity for H2 generation by hydrolysis of NaBH4 . Maximum H2 generation rate obtained by Co3 O4 coating is about 5 times higher than that produced by Co3 O4 powder which is mainly attributed to high surface area and large number of active sites provided by the Co3 O4 NPs in the coating owing to their size and shape.

By varying the O2 pressure during PLD, two

dierent cobalt oxide phases, namely Co3 O4 and CoO, were achieved in NPACs and it was found that Co3 O4 phase is more active for hydrolysis than CoO phase with lower oxidation number. The eect of morphology and crystallinity of Co3 O4 NPACs on H2 generation rate was studied by tuning the laser uence and substrate temperature, respectively. The result showed that the NPs with mixed amorphous-nanocrystalline phase on the surface act as the active sites for favorable interaction and conversion of NaBH4 .

7

2

Water Electrolysis

2.1 Cobalt-Boride as an ecient and robust electrocatalyst for hydrogen evolution reaction We reported Cobalt-Boride (Co-B), for the rst time, as a non-noble, ecient and robust electrocatalyst for Hydrogen Evolution Reaction (HER) active in aqueous solution of wide pH values. In water with neutral pH, an overpotential of only 251 mV was required to

2 attain a current density of 10 mA/cm . The reported overpotential, with related current density, was considerably lower than almost all the previously reported results of the earth abundant element-based molecular and solid-state catalysts, with favorable values

2 of Tafel slope (75 mV/dec) and exchange current density (0.25 mA/cm ) in neutral pH. From XPS and theoretical calculations, it was seen that a partial electron transfer takes place from boron atom to the metallic cobalt. These highly active Co surface sites are responsible for the remarkable HER activity and robust nature in wide range of pH (4-9) values. Under neutral pH, the current density varies by small value during 44 h test while maintaining overpotential of 250 mV: this proves the stability of Co-B catalyst. Finally, after 1000 cycles, the HER activity of Co-B at pH 7 remained unchanged proving it to be an ideal candidate for long runs in industrial conditions.

2.2 Co-Ni-B nanocatalyst for ecient hydrogen evolution reaction in wide pH range To improve upon the results obtained with Co-B electrocatalyst, Ni metal was incorporated into Co-B. Thus, amorphous Co-Ni-B nanocatalyst was successfully synthesized with high electrocatalytic activity towards hydrogen evolution reaction (HER) in wide pH range.

The content of Ni was varied by adjusting the molar ratio Ni/(Ni+Co) in

Co-Ni-B from 10% to 50% to identify the most suitable composition for HER. Co-30Ni-B (with 30% Ni) showed the highest catalytic activity and could reach a current density of 10 mA/cm

2

at just 170 mV in pH 7 and 133 mV in pH 14. It also exhibited a Tafel slope

8

value of 51 mV/dec in pH 7 suggesting Volmer-Heyrovsky reaction mechanism for HER. The role of each element in improving the activity was justied with results from XPS, XAS and DFT calculations.

It was observed that the presence of Ni promotes higher

electron density at Co active sites of Co-Ni-B which in turn facilitates ecient reduction reaction to enhance HER rate. Co-30Ni-B could sustain 1000 cycles and prolonged hours of operation for about 40 hours without losing activity making it an excellent low-cost electrocatalyst material.

2.3 Co-Mo-B nanocatalyst for enhanced hydrogen evolution in neutral and alkaline media A key strategy in improving the HER rate is to increase the number of surface active sites. This can be achieved by increasing the eective surface area of the nanocatalyst used.

To employ this strategy, we synthesized Co-Mo-B nanoparticles to catalyze the

hydrogen evolution reaction. It was observed from N2 adsorption-desorption isotherms that by addition of a small amount of Mo in Co-B, the surface area increases considerably presenting a large number of active sites on the surface. TEM micrographs showed that the average particle size of Co-Mo-B nanoparticles was around 18.4 nm which was lower compared to that of Co-B (30 nm) which explains the reason for increase in the surface area of Co-Mo-B. Co-3Mo-B with Mo/(Mo+CO) molar ratio of 3% showed exceptional electrochemical activity in both neutral and alkaline media where the overpotentials re-

2 quired to achieve the benchmark current density of 10 mA/cm were just 95 mV and 66 mV respectively. Co-3Mo-B also portrayed excellent stability in both media where it could maintain a constant current density in potentiostatic conditions for about 40 hours and could sustain 1000 cycles of operation without losing its activity, making it highly suitable for large scale applications.

9

Conclusion The present thesis consists of a compilation of the research work carried out on various strategies employed using nanocatalysts to improve the hydrogen production rates by two important processes - hydrolysis of chemical hydrides and electrolysis of water. The work done in this thesis holds signicance in the context of solving the global energy crisis by strengthening the case of hydrogen fuel.

We have not only explored various novel

strategies in employing nanocatalysts but have also ensured that we nd an inexpensive, earth-abundant and non-toxic alternative to noble metal catalysts which we have successfully achieved in the form of Co-B and its various alloys. We expect that the compilation of work in this thesis will pave way for many new research ideas and prove benecial for the research community at large.

References [1] Z. E.Huang, A. Su, and Y. C. Liu, Int. J. Energy Research, vol. 37, pp. 11871195, 2013. [2] M. Ni, D. Y. C. Leung, M. K. H. Leung, and K. Sumathy, Fuel Processing Technology, vol. 87, pp. 461472, 2006. [3] J. Gan, X. Lu, and Y. Tong, Nanoscale, vol. 6, pp. 71427164, 2014. [4] N. Patel, A. Miotello, and V. Bello, Appl. Catal. B: Environ., vol. 103, pp. 3138, 2011. [5] U. B. Demirci and F. Garin, Catal. Commun., vol. 9, pp. 11671172, 2008. [6] R. Fernandes, N. Patel, and A. Miotello, Int. J. Hydrogen Energy, vol. 34, pp. 2893 2900, 2009. [7] F. Rosalbino, G.Borzone, E. Angelini, and R. Raggio, Electrochim. Acta, vol. 48, pp. 39393944, 2003. [8] N. Patel, R. Fernandes, G. Guella, A. Kale, A. Miotello, B. Patton, and C. Zanchetta, J. Phys. Chem. C, vol. 112, pp. 69686976, 2008.

[9] N. Patel, G. Guella, A. Kale, A. Miotello, B. Patton, C. Zanchetta, L. Mirenghi, and P. Rotolo, Appl. Catal. A: General, vol. 323, pp. 1824, 2007. [10] R. Fernandes, N. Patel, A. Miotello, R. Jaiswal, and D. C. Kothari, Int. J. Hydrogen Energy, vol. 37, pp. 23972406, 2012.

10

[11] D. Tong, W. Chu, Y. Y. Luo, H. Chen, and X. Y. Ji, J. Mol. Catal. A: Chemical, vol. 269, pp. 149157, 2007. [12] N. Patel, R. Fernandes, N. Bazzanella, and A. Miotello, Catal. Today, vol. 170, pp. 2026, 2011. [13] M. Wua, P. K. Shena, Z.Weib, S. Songa, and M. Nie, J. Power Sources, vol. 166, pp. 310316, 2007. [14] M. C. Tavares, S. A. S. Machado, and L. H. Mazo, Electrochim. Acta, vol. 46, pp. 43594369, 2001. [15] T. W. Lin, C. J. Liu, and J. Y. Lin, Appl. Catal. B: Environ., vol. 134, pp. 7582, 2013. [16] Y. Sun, C. Liu, D. C. Grauer, J. Yano, J. R. Long, P. Yang, and C. J. Chang, J. Am. Chem. Soc., vol. 135, pp. 17 69917 702, 2013.

[17] K.Wang, D. Xi, C. Zhou, Z. Shi, H. Xia, G. Liu, and G. Qiao, J. Mater. Chem. A, vol. 3, pp. 94159420, 2015. [18] P. Xiao, Y. Yan, X. Ge, Z. Liu, J. Y. Wang, and X. Wang, Appl. Catal. B: Environ., vol. 154, pp. 232237, 2014. [19] W. Cui, Q. Liu, Z. Xing, A. M. Asiri, K. A. Alamry, and X. Sun, Appl. Catal. B: Environ., vol. 164, pp. 144150, 2015.

[20] Z. Huang, Z. Chen, Z. Chen, C. Lv, M. G. Humphrey, and C. Zhang, Nano Energy, vol. 9, pp. 373382, 2014. [21] P. Los and A. Lasia, J. Electroanal. Chem., vol. 333, pp. 115125, 1992. [22] E. Ndzebet and O. Savadogo, Int. J. Hydrogen Energy, vol. 19-8, pp. 687691, 1994. [23] J. J. Borodzilqski and A. Lasia, J. Appl. Electrochem., vol. 24, pp. 12671275, 1994. [24] H. Vrubel and X. Hu, Angew. Chem. Int. Ed., vol. 51, pp. 12 70312 706, 2012.

Suraj G. Gupta (Research student)

Dr. Nainesh Patel (Research Co-Guide)

11

Prof. D. C. Kothari (Research Guide)

List of Research Publications: 1. S. Gupta, N. Patel, R. Fernandes, D.C. Kothari, A. Miotello Mesoporous Co-B nanocatalyst for efficient hydrogen production by hydrolysis of sodium borohydride. International Journal of Hydrogen energy 38 (2013) 14685-14692. (I. F. 3.31) 2. N. Patel, R. Fernandes, S. Gupta, R. Edla, D.C. Kothari, A. Miotello Co-B catalyst supported over mesoporous silica for hydrogen production by catalytic hydrolysis of Ammonia Borane: A study on influence of pore structure. Applied Catalysis B: Environmental 140–141 (2013) 125– 132. (I. F. 7.44) 3. S. Gupta, N. Patel, D.C. Kothari, A. Miotello Cobalt-Boride: an Efficient and Robust Electrocatalyst for Hydrogen Evolution Reaction Journal of Power Sources 279 (2015) 620-625. (I. F. 6.22) 4.

S. Gupta, N. Patel, R. Fernandes, R. Kadrekar, A. Dashora, A.K. Yadav, D. Bhattacharyya, S.N. Jha, A. Miotello, D.C. Kothari Co-Ni-B nanocatalyst for efficient hydrogen evolution reaction in wide pH range Applied Catalysis B: Environmental 192 (2016) 126-133. (I. F. 7.44)

5.

R. Edla, S. Gupta, N. Patel, N. Bazzanella, R. Fernandes, D.C. Kothari, A. Miotello Enhanced H2 Production from hydrolysis of Sodium Borohydride using Co3O4 Nanoparticles Assembled Coatings Prepared by Pulsed Laser Deposition Applied Catalysis A: General. 515 (2016) 1-9. (I. F. 3.94)

6.

S. Gupta, N. Patel, R. Fernandes, S. Hanchate, A. Miotello, D.C. Kothari Co-Mo-B Nanoparticles as highly Efficient Bifunctional Electrocatalyst for Hydrogen and Oxygen Evolution Communicated to Angewandte Chemie Int. Ed. (I. F. 11.26)

7. S. Gupta, N. Patel, R. Fernandes, D.C. Kothari, A. Miotello Co-B Nanoparticles Supported Over FSM Type Mesoporous Silica: An Efficient Nanocatalyst for Hydrogen Production by Hydrolysis of Ammonia Borane. AIP Conf. Proc. 1512, 284 (2013). 8. U.V. Patil, N.S. Ramgir, A.K. Debnath, D.K. Aswal, S.K. Gupta, S. Gupta, N.Patel, D.C. Kothari Room Temperature NH3 Sensing Properties of Co-B-PANI Nanocomposite Films 2nd International Symposium on Physics and Technology of Sensors (ISPTS), 2015, IEEE XPLORE pp. 19-21, 2015.

9. H. Jadhav, A.K. Singh, N. Patel, R. Fernandes, S. Gupta, D.C. Kothari, A. Miotello, S. Sinha Pulsed Laser Deposition of Nanostructured Co-B-O thin films as efficient catalyst for H2 production Communicated to Applied Surface Science. (I. F. 2.71)

Conferences, Workshop Participation and Invited talks: 1. Contributed talk at National workshop on “Advanced Functional materials: Synthesis to Applications” held during 21st-22nd March, 2016 organized by Department of Physics, University of Mumbai, India. 2. Invited talk at workshop on “Nanotechnology and non-conventional Energy Production” held on 15th March, 2016 at Atharva College of Engineering, Mumbai. 3. Oral presentation in 102nd Indian Science Congress held during 3rd – 7th January, 2015 at the University of Mumbai, India. 4. Oral and poster presentations in Euro-mediterranean Hydrogen Technologies Conference - 2014 held during 9th – 12th December, 2014 at Taormina, Italy. 5. Oral presentation in International Conference on Energy Materials (ICEM – 2014) held during 28th – 30th July, 2014 at Sathyabama University, Chennai, India. 6. Participated in SERB school on “Materials Modeling and Simulations across Length Scales” held during 9th - 20th December, 2013 jointly organized by BARC, Mumbai and Department of Physics, University of Mumbai, India. 7. Oral and poster presentations in National Workshop on “Physics for Energy and Environment” held during 14th-15th June, 2013 organized by Department of Physics, University of Mumbai, India. 8. Poster presentation in 57th “DAE Solid State Physics Symposium” held during 3rd -7th December, 2012 jointly organized by BARC, Mumbai and IIT Bombay, India. 9. Participated in workshop on “Fluorescence Steady State and Lifetime Analysis” held on 7th Dec, 2010 jointly organized by NCNN University of Mumbai, ISS USA and Sinsil International, India.