Catalytic Partial Oxidation of Renewable Feedstocks

A dissertation submitted to the faculty of the Graduate School of the University of Minnesota by

Reetam Chakrabarti

In partial fulfillment of the requirements for the degree of Doctor of Philosophy.

Advisor: Lanny D. Schmidt

I owe a lot of gratitude to my advisor, Professor Lanny Schmidt for giving me the opportunity to join his group, and the freedom to pursue my ideas. His breadth of knowledge and unending curiosity about various things in the world have always amazed me. I thank him for always having an open door to discuss problems. I would like to thank Professor Aditya Bhan for his help, both academic and non-academic over the past four years, ever since I applied to the University of Minnesota. I would also like to thank Professor Alon McCormick and Professor Ulrike Tschirner for agreeing to be a part of my defense committee and reviewing my thesis. Thanks to the staff at the Characterization Facility at the University of Minnesota for answering questions regarding characterization techniques presented in this thesis. Thank you to all the teachers through school and college in India. I have enjoyed working with the graduate students and undergraduate students in the lab, all of whom have helped me in my experiments. In particular, I would like to thank Dr. Dave Rennard, Dr, Brian Michael, Dr. Josh Colby, Dr. Jake Kruger, and Sam Blass for their help during various stages of my research. I would like to thank Dr. Dave Rennard, Dr. Brian Michael and Dr. Josh Colby for helping me get started in my first summer with experiments starting with an empty hood. I thank Dr. Joshua Colby and Dr. Brian Michael for the great conversations in the lab, on our way down to the coffee room on the first floor, and in their house, the ‘Green Monster’. Both of them have been an inspiration with their engineering creativity and approach to doing experiments. Outside of his immense help in my research, I am grateful to Dr. Jake Kruger for a number of things which include the great conference trips, organizing a trip from the sky to the ground, and the trip to Sheboygan, WI for his wedding. I would like to thank Sam Blass for his comments on writing, and the research and non-research related conversations in the ping-pong room next to the lab, and for tennis and ice skating. I would like to thank Dr. Christine Balonek for her wonderful desserts and Jeremy W. Bedard for tickets for baseball games and

i ii rugby. Thanks to Richard ‘Walter’ Hermann for help with lot of the experiments in this thesis. Thanks to Michael Skinner for introducing me to number of things while initially in the US, including darts and Easter gifts. Thank you to Hui Sun, Alex Marvin, David Nare for their help in research and insightful comments during group meetings. Thanks to the undergraduate students, Nils Persson, for his great commedy shows and piano skills, Tyler Josephson, Eric Hansen, Ed Michor, and Sheila Hunter for their help. I would like to thank my friends, Shameek Bose, Sujit Jogwar, Srinivas Ran- garajan, Romain Le Picard, Aloysius Gunawan, Andrew Yongky, and my roommate Parthiv Daggolu for great moments in Minneapolis. In addition, thanks to my friends from undergrad, Umang Desai, Karan Kadakia, and Mansi Seth for great conversations that persisted through the last four years. I would like to thank my family in India, especially my parents and my sister Ishita for their support throughout my life. ABSTRACT

The current world energy and economic infrastructure is heavily reliant on fossil fuels such as coal, oil, and natural gas. The limited availability of fossil fuels along with environmental effects and economic uncertainties associated with their use has motivated the need to explore and develop other alternative sources of energy. Lig- nocellulosic biomass like fossil fuels is carbon-based and has the potential to partly supplant the energy supplied by fossil fuels. Lignocellulosic biomass is a complex mixture of polymers such as cellulose, hemicellulose, lignin along with small concentrations of inorganics and extractives. Recent research has shown that lignocellulosic biomass and biomass model compounds can be processed autothermally by catalytic partial oxidation in millisecond residence times over noble metal catalysts at high temperatures (600-1000 ◦C) to syngas (a mixture of carbon monoxide and hydrogen).1–3 The syngas stream can then be upgraded to fuels and chemicals. In Chapter 2, spatially resolved concentration and temperature profiles of methane and dimethyl ether, a model compound for biomass, are compared. Dimethyl ether can be produced renewably through syngas upgrading. Maximum temperature and concentration gradients were found within the oxidation zone. Most of the oxygen (∼ 95 %) was converted within the first 2.2 mm and syngas formation was observed despite the presence of oxygen. The catalytic partial oxidation process has been demonstrated using compounds which, unlike most biomass sources, contain negligible quantities of inorganics. Some of these inorganics have catalytic properties themselves and some may act as poisons for the Rh-based catalyst. The effects of common biomass-inorganics (silicon, calcium, magnesium, sodium, potassium, phosphorus, sulfur) on rhodium-based catalysts in autothermal reactors have been studied. To understand the effects of biomass inorganics on Rh catalysts, two sets of ex-

iii iv periments surveying common inorganics were performed - in the first set, inorganics were directly deposited on the rhodium catalyst and tested using steam methane reforming as a model reaction (Chapter 3); whereas in the second set, inorganics were introduced to a clean catalyst in an ethanol feed to simulate actual inorganic- containing biomass (Chapter 4). In both sets of experiments, performance testing, catalyst characterization and regeneration were carried out to probe the mechanism of inorganic interaction with the rhodium-based catalyst. Large decreases in reforming activity were observed on phosphorus- and sulfur-doped catalysts. Deactivation due to calcium and magnesium was primarily due to blocking of active sites. Potas- sium and silicon were volatile at the high temperatures within the reactor. Potassium introduced alkaline chemistry promoting acetaldehyde formation from ethanol while phosphorus introduced acid chemistry promoting formation of ethylene from ethanol. The effects of potassium and phosphorus on catalytic partial oxidation of methane and ethanol at different concentrations and temperatures have been studied in Chap- ter 5. The synergistic effects of potassium and phosphorus were studied by distribut- ing the inorganics together on the catalyst as monobasic potassium phosphate. The effects of both potassium and phosphorus were observed in the catalytic partial oxidation of methane on a potassium phosphate-doped catalyst at low temperatures. At high temperatures, only effects due to phosphorus were observed because of potassium volatilization. The results show that biomass-sources containing low concentrations of inorganics can be processed autothermally to a high selectivity syngas stream. The distribution and interactions of the inorganics within the catalyst can be used to design better pretreatment, processing, and regeneration strategies to minimize catalyst deactivation during biomass processing. Alcohols represent an important intermediate in different biomass upgrading routes. Chapter 6 discusses the behavior of butanol isomers, 1-butanol, isobutanol, 2-butanol, and tert-butanol over four different catalysts; Rh, Pt, RhCe, and PtCe at different fuel to oxygen (C/O) ratios. At low C/O ratios, equilibrium species such as CO,

CO2,H2 and H2O were obtained while non-equilibrium species such as carbonyls and oleﬁns were dominant at high C/O ratios. Low reforming activity was observed on Pt and PtCe catalysts. All isomers decompose primarily by dehydrogenation through a carbonyl intermediate except tert-butanol which decomposes by dehydration to isobutene; however, the reactivity of tert-butanol was unaﬀected. v

In Chapter 7, isobutanol autothermal reforming is integrated with a water gas shift stage downstream to produce hydrogen containing low concentrations of carbon monoxide for portable fuel cell applications. A RhCe-based catalyst was selected to carry out autothermal reforming of isobutanol while a PtCe catalyst was selected for the water gas shift stage. This staged reactor produced high yields of hydrogen (> 120 % selectivity) containing low concentrations of CO (< 2 mol %) in less than 100 ms making the effluent ideal for portable high temperature PEM fuel cell applications. The water gas shift stage also reduced the concentration of non-equilibrium products formed in the autothermal reforming stage by over 50 %. Thermodynamic analysis of the system showed that staged autothermal reforming of isobutanol integrated with a fuel cell can potentially lead to 2.5 times more efficient energy usage when compared to burning isobutanol in a conventional combustion engine. The results in this thesis give an insight into the mechanisms and processing challenges involved in converting renewable feedstocks to syngas by catalytic partial oxidation. Further experiments based on the conclusions of this thesis are discussed in Chapter 8. Spatial profile experiments to determine roles of mass transfer, steam reforming, and dry reforming during catalytic partial oxidation of oxygenates are proposed. Spatial profile studies for catalytic partial oxidation over inorganic-doped catalysts and feed are proposed to determine their concentrations and nature on the catalyst surface during reactor operation. CONTENTS

Acknowledgments i

Abstract iii

Table of Contents vi

List of Tables x

List of Figures xi

1 Introduction 1 1.1 Current World Energy Scenario ...... 1 1.1.1 Limited Fossil Fuel Reserves ...... 2 1.1.2 Developing countries ...... 2 1.1.3 Climate Impact ...... 3 1.1.4 Energy Security ...... 3 1.2 Biomass as an Energy Source ...... 3 1.2.1 Structure of Biomass ...... 4 1.2.2 Converting Biomass to Fuels and Chemicals ...... 5 1.3 Upgrading Syngas for Energy Use ...... 7 1.3.1 Fischer-Tropsch Process ...... 7 1.3.2 Methanol Route ...... 7 1.3.3 Hydrogen Generation for Fuel Cells ...... 8 1.4 Catalytic Partial Oxidation ...... 8 1.4.1 Evolution of Research ...... 9 1.4.2 Upgrading Products from Catalytic Partial Oxidation . . . . . 10 1.5 Inorganics and Biomass Processing ...... 10 1.5.1 Poisoning ...... 11 1.5.2 Fouling ...... 11 1.5.3 Sintering ...... 11 1.5.4 Attrition/Mechanical Failure ...... 12

vi CONTENTS vii

1.5.5 Solid State Transformation ...... 12 1.6 Summary ...... 12

2 Mechanism of Catalytic Partial Oxidation 14 2.1 Introduction ...... 14 2.2 Experimental ...... 17 2.2.1 Spatial Proﬁles ...... 17 2.2.2 Nature of Rhodium in Oxidation Zone ...... 20 2.3 Results ...... 21 2.3.1 Spatial proﬁles ...... 21 2.3.2 State of Rhodium in Oxidation Zone ...... 27 2.4 Conclusion ...... 30 2.5 Acknowledgements ...... 30

3 Eﬀects of Biomass Inorganics on Rhodium Catalysts: I. Steam Methane Reforming 31 3.1 Introduction ...... 32 3.2 Experimental ...... 34 3.2.1 Experimental Setup ...... 34 3.2.2 Product Analysis ...... 34 3.2.3 Catalyst Preparation ...... 35 3.2.4 Experimental Procedure ...... 35 3.2.5 Catalyst Characterization ...... 35 3.3 Results ...... 36 3.3.1 Performance Testing ...... 36 3.3.2 Catalyst Characterization ...... 37 3.3.3 Equilibrium Calculations ...... 42 3.4 Discussion ...... 42 3.4.1 Sulfur ...... 43 3.4.2 Phosphorus ...... 44 3.4.3 Silicon ...... 45 3.4.4 Sodium and Potassium ...... 45 3.4.5 Calcium and Magnesium ...... 46 3.5 Conclusions ...... 46 3.6 Acknowledgements ...... 47

4 Eﬀects of Biomass Inorganics on Rhodium Catalysts: II. Autother- mal Reforming of Ethanol 48 4.1 Introduction ...... 49 4.2 Experimental ...... 50 4.3 Results and Discussion ...... 52 4.3.1 Silicon ...... 52 4.3.2 Sulfur ...... 53 CONTENTS viii

4.3.3 Phosphorus ...... 54 4.3.4 Potassium ...... 55 4.3.5 Sodium ...... 57 4.3.6 Calcium ...... 59 4.3.7 Magnesium ...... 60 4.4 Comparison of Inorganics ...... 61 4.5 Conclusions ...... 63 4.6 Acknowledgements ...... 63

5 Effects of Potassium and Phosphorus on Rhodium Catalysts for Cat- alytic Partial Oxidation 64 5.1 Introduction ...... 65 5.2 Experimental ...... 67 5.3 Results ...... 69 5.3.1 Doping at Different Concentrations with Methane CPO . . . . 69 5.3.2 Doping at Different Concentrations with Ethanol CPO . . . . 74 5.3.3 Transient Studies with Methane CPO ...... 77 5.3.4 Transient Studies with Ethanol CPO ...... 79 5.4 Discussion ...... 80 5.4.1 Effect of Potassium ...... 81 5.4.2 Effect of Phosphorus ...... 83 5.5 Conclusions ...... 84 5.6 Acknowledgements ...... 84

6 Autothermal Partial Oxidation of Butanol Isomers 85 6.1 Introduction ...... 86 6.2 Experimental ...... 87 6.3 Results ...... 89 6.3.1 Conversion and Temperature ...... 89 6.3.2 Syngas and Combustion Products ...... 91 6.3.3 C4 Intermediates ...... 91 6.3.4 Other Intermediates ...... 95 6.4 Discussion ...... 95 6.4.1 Chemistry of the Isomers ...... 95 6.4.2 tert-Butanol ...... 99 6.4.3 Isobutanol ...... 100 6.4.4 Surface Chemistry ...... 100 6.4.5 Eﬀect of Catalyst ...... 101 6.5 Conclusion ...... 102 CONTENTS ix

7 Autothermal Reforming of Isobutanol 103 7.1 Introduction ...... 104 7.2 Experimental ...... 106 7.2.1 Catalyst Preparation ...... 108 7.2.2 Product Analysis ...... 108 7.3 Results ...... 108 7.3.1 CPO of Isobutanol ...... 109 7.3.2 Addition of WGS Stage ...... 111 7.4 Discussion ...... 113 7.4.1 CPO stage ...... 113 7.4.2 CPO + WGS ...... 114 7.5 Conclusions ...... 116 7.6 Acknowledgements ...... 116

8 Summary and Future Work 117 8.1 Spatial Proﬁles ...... 118 8.1.1 Examination of Mechanism of DME CPO ...... 118 8.2 Inorganics ...... 119 8.2.1 Extending Spatial Proﬁles to Inorganics ...... 119 8.2.2 Techniques to Minimize Catalyst Deactivation ...... 121

Bibliography 123 LIST OF TABLES

2.1 Conversions and product selectivities (both in %) at 2.2 mm (∼ 95 % oxygen conversion) and 10.2 mm (end of catalyst) for (a) methane and (b) dimethyl ether...... 26 2.2 Reactant conversions and product selectivities (both in %) during methane catalytic partial oxidation at C/O = 1 and total ﬂow rate 5 SLPM. . 28

3.1 Changes in product distributions upon doping with diﬀerent inorganics during steam methane reforming...... 40

4.1 Inorganic Content in Weight Percentage Dry Basis from Diﬀerent Biomass Sources ...... 50 4.2 Summary of Eﬀects Observed from Inorganics ...... 62

x LIST OF FIGURES

2.1 Experimental setup for measurement of concentration and temperature profiles during methane and dimethyl ether catalytic partial oxidation. 18 2.2 Experimental setup for methane catalytic partial oxidation to determine oxidation state of the rhodium catalyst...... 20 2.3 Methane, oxygen and temperature profile during methane catalytic partial oxidation...... 22 2.4 Product concentration profiles during methane catalytic partial oxidation...... 23 2.5 Dimethyl ether, oxygen and temperature profile during dimethyl ether catalytic partial oxidation...... 24 2.6 Product concentration profiles during dimethyl ether catalytic partial oxidation...... 25 2.7 Product flow rates at the end of 2.2 mm and 10.2 mm during (a) methane and (b) dimethyl ether catalytic partial oxidation...... 26 2.8 XRD patterns for alumina, fresh (calcined rhodium) and used rhodium catalysts...... 29

3.1 Methane conversion for catalysts doped with sulfur (A) and phosphorus (B) respectively. Hydrogen, carbon monoxide selectivities for catalysts doped with sulfur (C) and phosphorus (D) respectively. (A) also shows a schematic of the reactor setup...... 38 3.2 Methane conversion for catalysts doped with potassium (A) and sodium (B) respectively. Hydrogen, carbon monoxide selectivities for catalysts doped with potassium (C) and sodium (D) respectively...... 39

3.3 SEM images of (A) fresh 2.5 wt% Rh on α-Al2O3 catalyst (B) carbon ﬁlaments on catalyst doped with phosphorus (C) high resolution image of carbon ﬁlaments in catalyst doped with phosphorus showing rhodium particles at tip and (D) carbon structures on catalyst doped with potassium...... 41

xi LIST OF FIGURES xii

4.1 Doping and regeneration temperature profiles for Si (A). Doping and regeneration temperature profiles for S (B) ...... 53 4.2 Ethanol conversion, temperature profiles (A) and product selectivities (B) for doping with phosphorus...... 55 4.3 Ethanol conversion, temperature profiles (A) and product selectivities (B) for doping with potassium...... 56 4.4 Ethanol conversion, temperature profiles (A) and product selectivities (B) for doping with sodium...... 58 4.5 Ethanol conversion, temperature profiles (A) and product selectivities (B) for doping with calcium...... 59 4.6 Ethanol conversion and temperature profiles during doping and regeneration for magnesium...... 61

5.1 Carbon monoxide selectivities during methane catalytic partial oxidation at 1 % and 10 % loading of potassium...... 70 5.2 Methane and hydrogen selectivities during methane catalytic partial oxidation at 1, 10 and 100 % loading of phosphorus...... 72 5.3 Methane conversion, carbon monoxide and hydrogen selectivities during methane catalytic partial oxidation at 1 and 10 % loading of monobasic potassium phosphate...... 73 5.4 Carbon monoxide, hydrogen and ethylene selectivities during ethanol catalytic partial oxidation at 10 and 100 % loading of phosphorus. . . 75 5.5 Carbon monoxide, hydrogen and ethylene selectivities during ethanol catalytic partial oxidation at 10 and 100 % loading of monobasic potassium phosphate...... 76 5.6 Methane conversions and temperatures during methane catalytic partial oxidation for 6 h at C/O ratios of 0.75 and 1.5 at 10 % loading of potassium...... 78 5.7 (A) Methane conversions and temperatures during methane catalytic partial oxidation for 6 h at C/O ratios of 0.75 and 1.5 at 10 % loading of phosphorus. (B) Hydrogen selectivities over 6 h at C/O of 0.75 and 1.5...... 79 5.8 (A) Ethanol conversions and temperatures during catalytic partial oxidation with ethanol containing 0.05 mol % potassium. (B) Carbon monoxide and acetaldehyde selectivities during the doping period. . . 80 5.9 (A) Ethanol conversions and temperatures during catalytic partial oxidation with ethanol containing 0.05 mol % phosphorus. (B) Carbon dioxide, hydrogen and ethylene selectivities during the doping period. 81

6.1 Reactor conﬁguration for the autothermal CPO of the butanol isomers. 88 6.2 Conversion and catalyst backface temperature of the four butanol isomers...... 90 6.3 Selectivities to H2 and H2O from each butanol isomer...... 92 LIST OF FIGURES xiii

6.4 Selectivities to CO and CO2 from each butanol isomer...... 93 6.5 Selectivities to major carbonyl and C4 oleﬁns from each butanol isomer. 94 6.6 Selectivities to ethylene and propylene from each butanol isomer. . . 96 6.7 Proposed reaction mechanisms for each butanol isomer in the autothermal reactor...... 97

7.1 Reactor conﬁguration for autothermal catalytic partial oxidation of isobutanol...... 107 7.2 Product distributions and temperatures from catalytic partial oxidation of isobutanol...... 110 7.3 Product distributions and temperatures after integration of water gas shift with catalytic partial oxidation of isobutanol...... 112 7.4 CO conversion in the water gas shift stage...... 113 7.5 Thermodynamic analysis of the current system including photosynthesis, fermentation, catalytic partial oxidation-water gas shift, hydrogen use in fuel cell and combustion...... 115 CHAPTER ONE

INTRODUCTION

In this chapter, an overview is presented of the present global energy shortage and challenges associated with using biomass as an alternative to fossil fuels. Gasification, which is the principal topic of biomass upgrading in this thesis, involves producing synthesis gas (also called syngas), a mixture of carbon monoxide and hydrogen which can be upgraded through different routes to fuels and chemicals. Catalytic partial oxidation can convert different feedstocks to a high selectivity syngas stream autothermally over noble metal catalysts without any external heat input. Inorganics are an important constituent of lignocellulosic biomass and their mechanisms of altering catalyst activity are presented.

1.1 Current World Energy Scenario

Energy is essential to sustain the growth and development of human life. The quality of life and standard of living in any civilization is greatly inﬂuenced by its energy consumption.4 For example, there exists over an order of magnitude diﬀerence between the per capita energy consumption in North America and Africa.5 The current world energy consumption is ∼ 14 TWH and approximately 80 % of it is supplied from fossil fuels such as coal, oil, and natural gas.6,7 The world energy consumption is predicted to increase by approximately 35 % by the year 2035.8 The following sections describe the reasons for the present energy crisis and the why the current energy consumption model is unsustainable:

1 1.1 Current World Energy Scenario 2

1.1.1 Limited Fossil Fuel Reserves

Coal, oil, and natural gas have been the principal sources of energy for use in transportation, electricity generation, residential, and industrial sectors, supplying almost 80 % of the world’s energy.6,7 However, current estimates indicate that existing oil and coal reserves will be depleted to the point of being prohibitively expensive for exploration and recovery within approximately the next 50 and 120 years respectively.6,9 Peak production of both is expected to occur in approximately the next 20 years.10 Recently, unconventional sources of oil such as tar sands in Alberta, Canada have received attention. However, there exist concerns over the the high amount of energy required and environmental impact associated with extracting oil from such unconventional sources.11 Current estimates for natural gas consumption indicate that reserves will last for approximately 60 years.6 Natural gas use is predicted to increase with the discovery of large deposits of natural gas in unconventional sources. Over the past decade, large deposits of natural gas within shale formations in the US have been found resulting in gas production levels similar to the peak levels that existed during the 1970’s.12 From 2000 to 2010, the percentage of natural gas from unconventional sources has increased from 1 % to 20 %. Natural gas represents a more clean energy source than coal and can be upgraded to hydrogen or liquid fuels. However, the environmental eﬀects associated with use of chemicals for extracting natural gas from these unconventional sources such as shale formations by hydraulic fracturing are widely debated.13–18

1.1.2 Developing countries

The rapid increase in energy consumption by developing countries in Asia and Africa is another reason for the current energy crisis. Studies predict that over the next ∼ 25 years, the energy consumption of OECD nations (Organization of Economic Co-operation and Development countries, comprised mostly of North American and European Nations) will account for about 7 % of the global energy consumption increase while non-OECD countries (82 % of the current world population and 57 % of the current energy consumption19) account for the remaining 93 % of the world energy consumption increase.8 Of this increase, China and India alone account for more than 50 % of the predicted increase in global energy consumption. 1.2 Biomass as an Energy Source 3

1.1.3 Climate Impact

Perhaps the biggest driving force for reducing fossil fuel consumption and developing an energy infrastructure based on alternative energy sources is the issue of climate change associated with use of fossil fuels and subsequent global warming. Approxi- mately 30 Gigatonnes of CO2 are estimated to be released into the atmosphere every 20,21 year from fossil fuel combustion. As a consequence of rapid use of fossil fuels, CO2 levels in the atmosphere have increased from 280 ppm before the industrial revolution 20 to 390 ppm at present. Increasing CO2 levels may result in severe ecological and environmental impacts such as increased ﬂooding, unpredictable weather patterns, and decreased agricultural productivity.22,23 Hence, for a sustainable ecosystem it is necessary to reduce current CO2 emissions and explore CO2 neutral or non-carbon based sources of energy.

1.1.4 Energy Security

Existing fossil fuel reserves are concentrated in few geographical locations around the world. In addition, political issues in those regions can disrupt prices of fossil fuels in the global market leading to economic uncertainty and energy insecurity. Develop- ing local alternative sources of energy can reduce energy dependence on potentially unstable foreign nations.

1.2 Biomass as an Energy Source

Sources of energy that are being considered as potential alternatives to fossil fuel- based energy sources are biomass, photovoltaic, hydroelectric, nuclear, electrochemical (batteries). The issues of safety associated with use of nuclear energy as evidenced in March 2011 by the reactor meltdown in Japan has led many countries to develop plans to shut down existing nuclear power plants. Each of the above sources of energy has associated drawbacks, which combined with favorable economics of fossil fuels, has prevented widespread application of these alternative energy technologies. It is therefore likely that a combination of alternative energy sources rather than a single one may be the most feasible means of replacing fossil fuels. Lignocellulosic biomass has potential as an alternative source of energy, as it is abundant and represents the only sustainable source of renewable carbon.24–27 1.2 Biomass as an Energy Source 4

Furthermore, CO2 produced during the combustion of biofuels is consumed for the growth of biomass, preventing an increase in greenhouse gas emissions and making the overall process carbon neutral. A recent study conducted by the U.S. Department of Energy (DOE) and the U.S. Department of Agriculture showed that the U.S. has enough lignocellulosic biomass to sustainably produce biofuels replacing more than 30 % of its current petroleum consumption by 2030 if eﬃcient biomass processing technologies exist.28

1.2.1 Structure of Biomass

Biomass has a complex structure and the composition of various components in biomass changes depending upon the biomass source. A major diﬀerence between composition of biomass and fossil fuels is the presence of high amounts of oxygen in biomass as compared with fossil fuels (O/C ∼ 1 for carbohydrate rich biomass, O/C ∼ 0 for crude oil).29,30 As a result, the energy density of lignocellulosic biomass (∼ 8 MJ/kg) is much lower that that of gasoline (∼ 40 MJ/kg).31,32 In general, any biomass source is comprised of the following components:

Cellulose

Cellulose (C6H10O5)n (n ∼ 5000-10,000) is a polymer of glucose consisting of β-1,4 glycosidic linkages.29,33,34 Cellulose is present in the form of microﬁbres held together by hydrogen bonding and Van Der Waals forces. Cellulose comprises of approximately 40-50 wt % of dry weight of lignocellulosic biomass.30,33 The structural arrangement and intermolecular hydrogen bonding contribute towards making the cellulose structure highly resistant to both acid and enzymatic hydrolysis.29,35 Cellulose can exist in biomass in either crystalline or amorphous form, the crystalline form being more stable and resistant to hydrolysis.

Hemicellulose

Hemicellulose is a heteropolymer of diﬀerent C5 (xylose, arabinose) and C6 sugars (glucose, galactose, mannose).29,33,34 Hemicellulose comprises approximately 25-35 wt % of the dry weight of woody lignocellulosic biomass and consists of fewer repeating units than cellulose (n∼ 150).30,33 Compared to celluose, hemicellulose can be more 1.2 Biomass as an Energy Source 5

easily hydrolyzed by acids or enzymes. Unlike cellulose, hemicellulose exists in the amorphous form due to its branched structure.36

Lignin

Lignin is a complex 3-D network of phenolic derivatives and comprises of approximately 25-35 wt % of the dry weight of woody lignocellulosic biomass.33,34 Lignin binds cellulose ﬁbres together and helps in protecting plants from microbial attack.

Extractives

Extractives consists of components like fats, waxes, alkaloids, proteins, gums and resins.34,36 Extractives can generally be removed from woody biomass using polar solvents such as methanol and water.

Ash

Ash or inorganics in biomass consists of various elements such as Si, Ca, K, P, Mg, Na, etc. present in diﬀerent forms such as carbonates, chlorides, sulfates.37,38 Their overall and individual concentration changes depending upon the biomass source.

1.2.2 Converting Biomass to Fuels and Chemicals

The exact composition of individual components in lignocellulosic biomass can in- ﬂuence the type of processing technique being used or pretreatment technique being utilized. Currently there are three technologies being investigated for upgrading biomass to fuels and chemicals.

Gasiﬁcation

Gasification involves heating biomass in the presence of steam or oxygen at high temperatures (700-1000 ◦C) to produce syngas, a mixture of carbon monoxide and 36,39 hydrogen. Depending on the H2/CO ratio, different fuels and chemicals can be produced from syngas. Syngas is the building block for over 50 % of the petrochemicals synthesized worldwide.40 Gasification can take place with or without a catalyst and can be carried out in a fluidized bed reactor or updraft/downdraft gasifiers.36 1.2 Biomass as an Energy Source 6

Gasification generally results in the formation of tars and chars which can reduce performance efficiencies of downstream processes.36,39 Concentrations of tars are reduced by employing downstream techniques such as filtration or scrubbing.41

Pyrolysis

Pyrolysis involves heating lignocellulosic biomass in the absence of oxygen in an inert atmosphere at temperatures between 375 and 550 ◦C, followed by rapid cooling resulting in formation of a liquid product referred to as bio-oil.36,39,42 Depolymer- ization of the biomass structure takes place during heating and along with bio-oil, solid char and non-condensable gases are also produced. Fast pyrolysis processes require residence times less than 1 s and bio-oil yields of up to 80 % (on a dry basis) are obtained.42 Bio-oil is a mixture of over 400 compounds containing diﬀerent functional groups such as aldehydes, ketones, acids, alcohols, polyols and aromatic phenols.36,39,42 Bio-oil can be upgraded using catalysts by carrying out deoxygenation reactions to synthesize fuels and chemicals or to syngas by gasiﬁcation.36,43

Pretreatment and Hydrolysis

Pretreatment followed by hydrolysis involves depolymerizing the biomass structure to produce monomer sugars from cellulose and hemicellulose.36,39 Pretreatment results in removal of the lignin fraction and decreases crystallinity of cellulose. The sugar monomers obtained after hydrolysis can be upgraded by either fermentation (biolog- ical technique using enzymes) or using catalysts to produce fuels and chemicals. Each of these techniques has certain associated drawbacks. Gasiﬁcation generally results in formation of tars and chars, which can lead to additional expensive cleanup costs.36,39 During pyrolysis, bio-oil formed is acidic (pH 2-4), corrosive, toxic, and its composition changes with time.32,36,39 Pretreatment steps associated with hydrolysis are expensive and if fermentation is the next upgrading step, it involves long residence times (24 - 48 h).36 In case of ethanol production from corn starch, energy intensive distillation is required. 1.3 Upgrading Syngas for Energy Use 7

1.3 Upgrading Syngas for Energy Use

Syngas is an important intermediate for synthesis of chemicals (methanol, ammonia) and fuels. The following describe the various routes to upgrade syngas for producing energy.

1.3.1 Fischer-Tropsch Process

The Fischer-Tropsch process involves upgrading syngas to hydrocarbon based fuels. The process typically takes place over iron or cobalt based catalysts at temperatures of 220 - 250 ◦C for low temperature Fischer-Tropsch (LT-FT) and 330 - 350 ◦C for high temperature Fischer-Tropsch (HT-FT).44,45 Typical pressures are approximately 25 - 45 bar for LT-FT and 25 bar for HT-FT. Several companies have either Fischer- Tropsch plants in operation or are in the process of being setup.46 These technologies are mostly based on converting natural gas to syngas followed by upgrading through the Fischer-Tropsch process. The Fischer-Tropsch reaction may be used to synthesize higher hydrocarbons (Eq. 1.1) or mixed alcohols (Eq. 1.2) and may be represented as:45

(2n + 1)H2 + nCO −→ CnH2n+2 + nH2O (1.1)

2 nH2 + nCO −→ CnH2n+1OH + (n − 1)H2O (1.2)

1.3.2 Methanol Route

Another route for upgrading syngas is through a methanol intermediate. Methanol, one of the top manufactured chemicals in the world, is synthesized from syngas over copper-zinc oxide based catalysts at temperatures of 210-270 ◦C and pressures of approximately 50 - 100 bar.47,48 Methanol synthesis is equilibrium limited and the reactions may be represented as:

CO + 2 H2 −→ CH3OH ∆Hr = -91 kJ/mol (1.3)

CO2 + 3 H2 −→ CH3OH + H2O ∆Hr = -49 kJ/mol (1.4) 1.4 Catalytic Partial Oxidation 8

Nobel Laureate George Olah advocated the existence of a future methanol-based economy as it has the potential to be a valuable intermediate for production of fuels and chemicals.49 Methanol can be either converted to dimethyl ether (DME) or oleﬁns/gasoline by the MTO/MTG (methanol to oleﬁns/methanol to gasoline process).50–53 Both DME synthesis and MTO/MTG take place on acidic catalysts such as zeolites. DME is a clean burning fuel and can be used as a diesel substitute.54,55

1.3.3 Hydrogen Generation for Fuel Cells

Several studies have focused on the advantages of a hydrogen-based economy since hydrogen has a high energy density on a mass basis and can be used as an energy carrier.56–58 By use of hydrogen in fuel cells, chemical energy can be converted into electrical energy for use in residential, industrial and transportation sectors producing only heat and water. The CO tolerances are diﬀerent for various types of fuel cells.59 PEM (proton exchange membrane) fuel cells are an attractive option for transpora- tion, residence and commercial applications due to their high power densities. How- ever, the catalyst in a PEM fuel cell is sensitive to CO poisoning, thereby requiring CO concentrations in the ppm range. Higher temperatures result in improved CO tolerances in PEM fuel cells.60 Initial CO removal from syngas takes place through the water gas shift reaction (Eq. 1.5) shown below:

CO + H2O −→ CO2 + H2 ∆Hr = -41 kJ/mol (1.5) To reduce the CO concentration to ppm levels, CO methanation (Eq. 1.6) or preferential CO oxidation (Eq. 1.7) are employed.

CO + 3 H2 −→ CH4 + H2O ∆Hr = -206 kJ/mol (1.6)

1 CO + 2 O2 −→ CO2 ∆Hr = - 283 kJ/mol (1.7)

1.4 Catalytic Partial Oxidation

Schmidt and coworkers have used catalytic partial oxidation (CPO) has been used to convert diﬀerent feedstocks to syngas.1–3,61–72 The advantage of catalytic gasiﬁcation 1.4 Catalytic Partial Oxidation 9

over the non-catalytic route is that higher feed conversions and selectivities to desired syngas products are attainable. CPO takes place at high temperatures (600 - 1000 ◦C) over noble metal catalysts (Rh, Pt, Ru) with no tar or char formation, thereby reducing or eliminating cleanup costs. The residence times within the catalyst bed are on the order of milliseconds, thereby making it possible to have smaller reactor sizes as well as lower catalyst weights. Unlike fossil fuels, since biomass is a distributed resource, the cost of transporting biomass from its source to the processing location is an important economic consideration in implementing biomass as an alternative energy source. Due to the short residence times in CPO, it may be economically feasible to have small reactors at the biomass source and transport syngas produced to a central processing location.

1.4.1 Evolution of Research

Hickman and Schmidt in 1993 ﬁrst reported experiments on CPO.61 Methane CPO was carried out autothermally over rhodium- and platinum-based catalysts. The overall partial oxidation reaction with methane is shown below (Eq. 1.8)

1 CH4 + 2 O2 −→ CO + 2 H2 ∆Hr = -36 kJ/mol (1.8)

62–67 Research over the next ten years involved CPO of higher alkanes (C2−C16). Processing higher alkanes which exist in the liquid state at room temperature (eg. decane C10H22, hexadecane C16H34) required a preheating section before the catalyst to vaporize the feed prior to contacting the catalyst.64–67 In 2004, Deluga et al. showed that prevaporized ethanol could be converted to hydrogen with selectivities greater than 100 % over rhodium-based catalysts by steam addition.68 Further research over the next 3 years focused on CPO of oxygenated hydrocarbons with diﬀerent functional groups such as alcohols, acids, and esters.69–72 In 2006, Salge et al. demonstrated the CPO of nonvolatile liquid feeds to syngas with high selectivities by ‘Reactive Flash Volatilization’ (RFV).1 Dauenhauer (2007) and Colby (2008) used RFV to convert solid feedstocks such as aspen, polyethylene, and cellulose to high selectivity synthesis gas stream autothermally without any external heat input.2,3 Thus, CPO is a highly versatile and robust process capable of producing high yields of syngas from wide range of feedstocks, ranging from small molecules like methane to actual biomass, a complex network of polymers. 1.5 Inorganics and Biomass Processing 10

1.4.2 Upgrading Products from Catalytic Partial Oxidation

CPO is limited by thermodynamics over rhodium-based catalysts at high temperatures.3,70,72,73 CPO products over rhodium-based catalysts typically consist of thermodynamic equilibrium products - CO, CO2,H2 and H2O. Also, no tar or char formation is observed because thermodynamics does not predict char formation at the high temperatures involved with the feedstocks being processed.2,3 Since the contact times are in the order of milliseconds, these reactors run close to adiabatic conditions. As the product stream is at high temperatures, it can be upgraded to fuels (Section 1.3) and chemicals downstream using diﬀerent chemistries without any additional heat input.

1.5 Inorganics and Biomass Processing

The research on CPO to date has been performed using feedstocks that contain little to no inorganics with no observable catalyst deactivation. However, actual biomass, depending on the source can contain from less than 1 % to 25 % by weight inorganics.38 Even if deactivation is taking place, it may not be observed since CPO reactors at high temperatures operate close to thermodynamic equilibrium.2,3,70,73 Hence, due to thermodynamic equilibrium limitations, deactivation may be taking place without being observed on the time scale of experiments (tens of hours). CPO reactions over rhodium catalysts have typically been run for 10-100 hours with pure feedstocks without any observable catalyst deactivation. It is possible that the catalyst will deactivate over longer time scales with even small concentrations of impurities in the feed. For example ppm levels of impurities in glycerol feed resulted in significant loss of catalyst activity over ∼ 400 h.74 Inorganics in biomass present an important obstacle in any catalytic biomass upgrading process by changing the activity of the catalyst. Even after pyrolysis, bio-oil still contains up to 0.3 % inorganics which can influence subsequent catalytic upgrading processes.32,33 Since noble metal catalysts are expensive, efficient use of the catalyst is essential. Hence, regeneration of the catalyst is critical to maintain economic feasibility of the process. The reactor configuration needs to be altered based on the time-scales of deactivation and regeneration. For example, a catalyst which deactivates within a few seconds, as is the case with an FCC zeolite catalyst, an entrained-type reactor 1.5 Inorganics and Biomass Processing 11 with riser is used.75,76 For deactivation in the time scale of seconds or minutes, a swing type reactor may be used. For deactivation which takes place on the time scale of years, a fixed bed reactor is used. In general, catalyst deactivation can be attributed to any of the five following mechanisms:

1.5.1 Poisoning

Poisoning refers to deactivation of the catalyst due to strong chemisorption of impurity species on the catalyst surface.75,77,78 Poisoning can be either selective or non- selective depending on whether the impurity binds speciﬁcally to certain catalyst sites. Through electronic interactions with catalyst active sites, poisons prevent adsorption of other species or can restructure the catalyst surface thereby altering catalyst activity. Sulfur is a common industrial catalyst poison for many processes (steam reforming, water gas shift, ammonia synthesis); hence sulfur removal is necessary to minimize catalyst deactivation.

1.5.2 Fouling

Fouling refers to physical blocking of the active site by the impurity.75,77 This is usually the deactivation mechanism observed with coke formation. Due to physical blocking of active sites, reaction species are unable to access them leading to changes in catalyst activity. In addition, large amounts of fouling may also aﬀect heat transfer between catalyst particles which can alter reaction rates and product distributions.

1.5.3 Sintering

Sintering involves deactivation of the catalyst due to aggregation of metal clusters at high temperatures to form larger aggregates.75,77,78 Sintering is generally temperature dependent. By aggregating to form larger clusters, the surface area and hence surface energy is reduced thereby leading to a more energetically favored state. The metal support interaction and associated environment around the catalyst can also inﬂuence sintering. For example, in case of strong metal support interaction (SMSI), sintering is reduced.79 1.6 Summary 12

1.5.4 Attrition/Mechanical Failure

This mechanism involves structural disintegration of the catalyst due to mechanical or thermal stresses.75,77,78 Mechanical stresses may arise due to collision with other particles or at high ﬂow velocities leading to erosion. This is common in ﬂuidized bed reactors. Thermal stresses are encountered during rapid heating/cooling which may take place during startup/shutdown leading to structural disintegration.

1.5.5 Solid State Transformation

This mode of deactivation can refer to the active catalyst material reacting with the support to form a new phase leading to loss of catalytic activity or the active catalyst reacting with species in the gas or vapor phase to form volatile compounds leading to loss of catalyst.77,78 The nature of the environment around the catalyst and the temperature inﬂuence the transformation and change in catalytic activity.

1.6 Summary

This thesis presents an insight into the mechanisms, process design, and challenges associated with CPO of various renewable feedstocks. Upgrading lignocellulosic biomass to syngas by CPO represents an attractive technology to partly supplant the energy supplied by fossil fuels. Chapter 2 investigates the mechanism of CPO by spatial profiles and characterization techniques such as X-ray diffraction (XRD) and X-ray pho- toelectron spectroscopy (XPS). Biomass inorganics can affect the activity of rhodium catalysts by a single or a combination of the any of the above five modes of deactivation. Chapters 3 and 4 survey the effects of common biomass inorganics. Chapter 5 studies the effect of potassium and phosphorus in detail at different temperatures and concentrations. To understand the mechanism by which individual inorganics affect the rhodium catalyst, along with performance testing, different characterization techniques such as Scanning Electron Microscopy (SEM), XPS, XRD, and chemisorption are used. To amplify catalyst deactivation so that activity changes could be observed on time scales feasible in a laboratory, experiments were performed at lower temperatures ∼ 700 ◦C, such that the product distributions were away from thermodynamic equilibrium. At higher temperatures, the catalyst would be closer to thermodynamic equilibrium and the volatility of inorganics would increase thereby minimizing the ef- 1.6 Summary 13 fects of catalyst deactivation. Alcohols represent important intermediates in biomass upgrading processes such as pyrolysis and hydrolysis.36,39 Chapter 6 investigates the CPO of four butanol isomers on Rh, RhCe, Pt, and PtCe catalysts. For ultilizing hydrogen in syngas for fuel cells, carbon monoxide removal is necessary. In Chapter 7, integration of CPO and water gas shift for production of high-purity hydrogen for fuel cells is examined. CHAPTER TWO

MECHANISM OF CATALYTIC PARTIAL OXIDATION

Catalytic partial oxidation (CPO) reactors have complex concentration and temperature gradients which makes it difficult to understand the mechanism of autothermal operation. Previous research in the Schmidt group using the capillary sampling technique during methane CPO has shown the existence of two zones - oxidation and reforming.80–83 Approximately all of the oxygen in the feed is converted within the first 2 mm of the catalyst bed. In this chapter, temperature and concentration profiles during the CPO of methane and dimethyl ether CH3OCH3, an oxygenated hydrocarbon are presented. Results with the two feeds were compared and the nature of the profiles were similar with sharp gradients existing in the oxidation zone with slower chemistries in the reforming section. The oxidation state of the rhodium catalyst in the oxidation zone during methane CPO was also analyzed using XPS and XRD.

2.1 Introduction

Catalytic partial oxidation (CPO) has been shown to convert a wide range of feedstocks ranging from small molecules such as methane to actual biomass autothermally to a high selectivity thermodynamic equilibrium synthesis gas stream.1–3,61,63,84,85 The process takes place over noble metal based catalysts such as rhodium and platinum and typical temperatures in a CPO reactor are in the order of 600-1000 ◦C with residence times approximately 50 ms. CPO involves a combination of oxidation and reforming reactions. CPO was ﬁrst demonstrated by Hickman and Schmidt in 199361

14 2.1 Introduction 15

and comprises of the following reactions involving combustion (Eq. 2.1), partial oxida-

tion (Eq. 2.2), steam reforming (Eq. 2.3), CO2 reforming (also called dry reforming, Eq. 2.4) and water gas shift (Eq. 2.5).82

CH4 + 2 O2 −→ CO2 + 2 H2O ∆Hr = -803 kJ/mol (2.1)

1 CH4 + 2 O2 −→ CO + 2 H2 ∆Hr = -36 kJ/mol (2.2)

CH4 + 2 H2O −→ CO + 3 H2 ∆Hr = 206 kJ/mol (2.3)

CH4 + CO2 −→ 2 CO + 2 H2 ∆Hr = 247 kJ/mol (2.4)

CO + H2O −→ CO2 + H2 ∆Hr = -41 kJ/mol (2.5) Syngas stream can be upgraded to fuels or chemicals by the Fischer-Tropsch process, methanol, or hydrogen. Due to the short contact times (order of milliseconds), CPO is suitable for portable or small-scale applications like fuel cells unlike conventional steam reformers which have residence times in the order of seconds.80 The capillary sampling technique has been demonstrated with simple hydrocarbons such as methane80–83 and ethane.86 The noble metal catalyst during CPO can be divided into two sections - an oxidation section where oxygen is present and a reforming section downstream without any gas phase oxygen. Using the temperature and concentration profiles within the reactor, the mechanism of CPO on different catalysts can be determined. With oxygenated hydrocarbons, a problem with the technique is homogeneous chemistry within the reactor and sampling system which can introduce artifacts in the analysis. Kruger et al. showed that significant conversion of oxygenated hydrocarbons occur at temperatures in excess of 500 ◦C which 85,87 becomes more prominent in the presence of oxygen. Dimethyl ether, CH3OCH3 (DME) was shown to have negligible homogeneous chemistry even at temperatures in excess of 600 ◦C indicating its high stability in the gas phase. Therefore, all chemistry can be attributed to that taking place on the catalyst surface. DME concentration and temperature profiles will help in better understanding the mechanism of CPO of oxygenates and hence of lignocellulosic biomass. As in methane CPO, complete 2.1 Introduction 16

oxidation (Eq. 2.6), partial oxidation (Eq. 2.7), CO2 reforming (Eq. 2.8), steam reforming (Eq. 2.9) and water gas shift reactions (Eq. 2.5) take place during DME CPO.

DME is considered as a diesel substitute and has lower NOx and particulate emissions than diesel fuel.54,55 DME steam reforming as a source of hydrogen for fuel cells has been reported in the literature.88–90 CPO of DME to produce high selectivity hydrogen represents an attractive option for small-scale or portable applications.

7 CH3OCH3 + 2 O2 −→ 2 CO2 + 3 H2O ∆Hr = -1329 kJ/mol (2.6)

1 CH3OCH3 + 2 O2 −→ 2 CO + 3 H2 ∆Hr = -38 kJ/mol (2.7)

CH3OCH3 + CO2 −→ 3 CO + 3 H2 ∆Hr = 246 kJ/mol (2.8)

CH3OCH3 + H2O −→ 2 CO + 4 H2 ∆Hr = 204 kJ/mol (2.9) Initially, the interplay between the various reactions was unclear with both a direct and indirect mechanism of CPO being proposed.91–93 The direct mechanism suggests that syngas products, H2 and CO are formed directly from partial oxidation (Eq. 2.2) while the indirect mechanism suggests the existence of two distinct reaction zones - an oxidation zone producing combustion products (Eq. 2.1) and a reforming zone producing syngas (Eq. 2.3). Using the spatially resolved measurements by the capillary sampling technique, Schmidt and coworkers showed that all the oxygen is consumed within approximately 2 mm over Pt and Rh based catalysts and syngas production was observed even within the oxidation zone.80–83 Reaction rates, concentration and temperature gradients were highest in the oxidation zone where oxygen was present in the gas phase (∼ 2 mm). However, as reported by Horn et al. , the formation of syngas in the oxidation zone does not conﬁrm the prevalance of a direct mechanism of CPO formation due to the occurrence of multiple surface reactions on the catalyst surface.81 Detailed numerical simulation studies involving surface reactions taking place during methane CPO suggest that in short contact time reactors, at high temperatures, syngas formation in the oxidation zone takes place mostly by the direct mechanism along with syngas combustion, steam reforming and water gas 2.2 Experimental 17

shift reactions.94–97 After all the oxygen is consumed, syngas is produced by steam reforming. During CPO, the net environment is highly reducing due to high amounts of hydrogen produced, hence rhodium exists in a reduced state in the reforming section. However, within the oxidation zone, where oxygen is present, the state of the rhodium catalyst is subject of debate in literature. Grunwaldt et al. showed using in situ X-ray absorption spectroscopy studies at lower temperatures (∼ 400 ◦C) the presence of an indirect mechanism and in the oxidation zone, rhodium was present in the oxidized state.98,99 Numerical simulations of short contact time reactors at temperatures similar to CPO (600 - 1000 ◦C) suggest that rhodium is most likely to be in the reduced state even in the presence of oxygen.95–97 Since CPO reactors operate at high temperatures, it is possible that the state of the catalyst might change during shutting down the reactor. Typical shut down of a CPO reactor involves setting oxygen followed by fuel flow rates to zero and subsequent cooling of the reactor with an inert gas. It is therefore possible that the catalyst in the oxidation zone might be oxidized during reaction, however the catalyst might get reduced due to pyrolysis of the fuel at high temperatures after shutting off the oxygen flow. To avoid any change to the catalyst during shutdown, both oxygen and methane flows were shut off simultaneously using an on-off valve. Here, XPS and XRD are used to experimentally analyze the oxidation state of the rhodium catalyst in the oxidation zone.

2.2 Experimental

2.2.1 Spatial Proﬁles

Methane and DME CPO was carried out in a 19 mm ID quartz reactor. The catalyst consisted of a 80 ppi (pores per linear inch) α-alumina foam monolith (17 mm diameter, 10 mm long) coated with 5 wt% rhodium. Blank 80 ppi α-alumina foams were used as front and back heat shields. The three monoliths (front heat shield, rhodium catalyst, and back heat shield) were wrapped in aluminosilicate cloth to prevent bypassing of gases. A 700 µm hole was drilled along the axis of all three monoliths with a diamond drill bit. Flow rates of gases to the reactor were controlled using mass ﬂow controllers accurate to within ± 2%. 2.2 Experimental 18

Micrometer

To QMS/

Thermocouple Stainless steel union

Stainless steel guide Argon + Methane + Oxygen

Quartz reactor tube Fused silica capillary

Front heat shield 80 ppi

5 wt % Rh/α-Al O catalyst Back heat shield 2 3

Exhaust

Figure 2.1: Experimental setup for measurement of concentration and temperature proﬁles during methane and dimethyl ether catalytic partial oxidation. 2.2 Experimental 19

Rhodium nitrate was used as the rhodium precursor and catalysts were prepared by the incipient wetness technique as described previously,100 then dried in a vacuum oven and calcined in a furnace in air at 800 ◦C for 6h. Products were analyzed at the exit of the reactor for integral performance data and within front heat shield to test for homogeneous chemistry with a HP 5890 gas chromatograph with a 60 m Plot-Q column. For spatial profile studies, a fused silica capillary (650 µm OD) was connected to a 1/16 inch stainless steel union. The other end of the union was connected to a quadrupole mass spectrometer (QMS). The stainless steel union was mounted to a micrometer screw. The fused silica capillary moved along the axis of the reactor through the drilled holes within the 3 monoliths by turning the micrometer screw. Temperatures were measured by passing a thermocouple (Omega, K-type, 270 µm O.D.) through the union and the capillary, until the end of the thermocouple was slightly below the end of the capillary, similar to that shown by Horn et al.80 The experimental setup is shown in Fig. 2.1. After turning the micrometer screw, the system was allowed to equilibrate for approximately 5 min. The positions shown are accurate within 0.0125” (0.3 mm). The profiles along the axis of the catalyst were analyzed with the QMS which has much faster analysis times than a gas chromatograph. Errors in carbon balances were typically within 10 %. The water flow rate was quantified by closing hydrogen and oxygen balances on it and averaging the two values. Selectivity to a particular species was defined on a carbon (for all carbonaceous species) or hydrogen (for H2 and H2O) basis as number of atoms of carbon or hydrogen in a particular species to total number of carbon or hydrogen atoms in the products, not taking into account unconverted fuel. The carbon flow rate used in the experiments was approximately 25 mmol/min. The C/O ratio is defined as the ratio of carbon atoms in fuel to oxygen atoms from air. For spatial profile measurements of methane and DME, experiments were performed at a C/O ratio of 1.2 and at a total gas flow rate of 2.4 SLPM. This corresponds to a residence time of approximately 10 ms at reactor temperatures. The flow rate through the QMS sample line was approximately 5 ml/min which is negligible compared to the total flow rate through the system thereby causing negligible changes to flow behavior within the system. 2.2 Experimental 20

Argon

Methane + Oxygen On-off valve

Quartz reactor tube

0.25 g

80 ppi α-alumina 2.5 wt % Rh/α-Al2O3 catalyst monoliths

To GC and Incinerator

Figure 2.2: Experimental setup for methane catalytic partial oxidation to determine oxidation state of the rhodium catalyst.

2.2.2 Nature of Rhodium in Oxidation Zone

Experiments were carried out in a 19 mm ID quartz tube. Methane CPO was carried out at a total flow rate of 5 SLPM and a C/O ratio of 1. The methane flow rate was 1 SLPM. A 2.5 wt% rhodium catalyst was supported on 1.3 mm α-alumina spheres. 0.25 g of the catalyst was placed between two 80 ppi α-alumina foams. Another 80 ppi α-alumina foam was placed below the back heat shield and an Omega type-K thermocouple was placed between the two foam monoliths. The experimental setup is shown in Fig. 2.2. The monoliths were wrapped in aluminosilicate cloth to prevent bypassing of gases. Products were analyzed with a HP 5890 gas chromatograph as in the spatial profile experiments. CH4 was quantified on the flame ionization

detector while H2,O2, CO, and CO2 were quantified using the thermal conductivity detector. Errors in carbon balances were typically within 10 %. Water flow rate was quantified by closing hydrogen and oxygen balances on it and averaging the two values. Experiments were run for one hour before methane and oxygen flows were shut off simultaneously using an on-off valve. 2.3 Results 21

Catalysts were analyzed by XRD and XPS. XRD was performed in Bruker D5005 with a 2.2 kW Copper Source. The spheres were crushed into a powder and then placed in the XRD. A step size of 0.02 ◦ and a dwell time of 1 s was used. XPS was performed in SSX-100 with 1.38 A.U. aluminium K-α radiation. The spot size used was 800 µm and the Carbon 1s peak at 284.6 eV was used as a reference.

2.3 Results

2.3.1 Spatial proﬁles

Spatially resolved temperature and concentration profiles of methane and DME were measured during CPO using the capillary sampling technique. Up to the beginning of the catalyst (end of front face), no fuel or oxygen conversion was observed within the front heat shield on the gas chromatograph indicating absence of homogeneous chemistry with both methane and DME. Spatial profiles of methane showed that most of the methane conversion and almost all of the oxygen (∼ 95 %) consumption takes place within the first 2.2 mm of the catalyst as shown in Fig. 2.3. The temperature profile initially increased due to exothermic oxidation chemistry and then decreased due to endothermic reforming. The nature of the trends are consistent with previously observed results with spatial profiles of methane.80–82 Syngas formation was observed from the very beginning of the catalyst (Fig. 2.4a and Fig. 2.4b). Sharp gradients in all product distributions were observed in the oxidation zone. The water concentration profile (Fig. 2.4a) showed a peak at the end of the oxidation zone and then decreased due to steam reforming reaction (Eq. 2.3). The hydrogen profile increased throughout the length of the catalyst (Fig. 2.4a) while the CO concentration profile was almost constant after about 3.5 mm (Fig. 2.4b), just after all of the oxygen was consumed. The

CO2 concentration proﬁle was almost constant after ∼ 2mm indicating absence of

CO2 reforming (Eq. 2.4) over Rh catalysts consistent with previous results (Fig. 2.4b).80–82 Similar trends were observed during spatial proﬁles of DME. DME being more reactive than methane, higher overall conversion of DME was observed and ∼ 95 % of the oxygen was consumed within the ﬁrst 2.2 mm (Fig. 2.5). Similar to methane, syngas formation for DME was observed within the oxidation zone (Fig. 2.6a and 2.3 Results 22

2 6 9 0 0 5 w t % R h c a t a l y s t 2 4 8 0 0 2 2 T 2 0 7 0 0

) 1 8

n C H i

4 ) 1 6 m 6 0 0 C / o l (

o 1 4 F r o n t h e a t e r m s h i e l d 5 0 0 u t m 1 2 B a c k h e a t ( a

e s h i e l d e t 1 0 p a 4 0 0 r

8 m w e o T l 6 3 0 0 F 4 2 2 0 0 O 2 0 1 0 0 - 2 0 2 4 6 8 1 0 1 2 A x i a l C o - o r d i n a t e ( m m )

Figure 2.3: Methane (), oxygen (N)and temperature proﬁle (•) during methane CPO.

Fig. 2.6b). For both methane and DME, ≥ 99.9 % oxygen consumption was observed within 3.5 mm. CO production like during methane CPO was observed primarily in the oxidation zone with its profile almost constant after ∼ 3.5 mm as shown in Fig. 2.6b. Hydrogen concentration profile increased throughout the length of the catalyst (Fig. 2.6a) similar to methane CPO. Comparison of the H2 and CO concentration profiles during methane and DME CPO revealed that the profiles were steeper during DME CPO than with methane, which may be attributed to the higher overall reactivity of DME. Comparing the water evolution profiles, the initial profile was much steeper with a sharper peak for methane compared with DME. Examination of the CO2 concentration profile showed that the rate of CO2 increase was slower with

DME (Fig. 2.6b) than with methane (Fig. 2.4b). In case of methane, the CO2 proﬁle

was constant after approximately 2 mm. With DME, the rate of CO2 increase was much slower, gradually increasing up to 3.5 mm and was almost constant thereafter.

Thus, compared to methane, in case of DME, CO2 evolution takes place almost to the

point of ≥ 99.9 % oxygen conversion (3.5 mm). This may be due to CO2 reforming

of DME, which caused the CO2 evolution to be less steep as CO2 was consumed by 2.3 Results 23

1 6 5 w t % R h c a t a l y s t

1 4

1 2 H 2 ) n i 1 0 m / l o

m 8 F r o n t h e a t

m s h i e l d

( H O 2

e 6 t

a B a c k h e a t r s h i e l d

w 4 o l F 2

- 2 - 2 0 2 4 6 8 1 0 1 2 A x i a l C o - o r d i n a t e ( m m )

(a) 1 6 5 w t % R h c a t a l y s t 1 4

1 2 )

n 1 0 i m / l B a c k h e a t o 8 C O

m s h i e l d m

( F r o n t h e a t 6

e s h i e l d t a r 4 w C O o

l 2 F 2

- 2 - 2 0 2 4 6 8 1 0 1 2 A x i a l C o - O r d i n a t e ( m m )

(b) Figure 2.4: Product concentration profiles during methane CPO. (a) shows hydrogen() and water(N) concentration profiles. (b) shows carbon monoxide () and carbon dioxide(N) concentration profiles. 2.3 Results 24

1 4 9 0 0 5 w t % R h c a t a l y s t T 1 2 8 0 0

1 0 7 0 0 ) n i ) m F r o n t h e a t 6 0 0 C / o

l 8 (

o s h i e l d e r m B a c k h e a t

5 0 0 u t m

( 6 s h i e l d a

r e e t p a 4 0 0 r C H O C H

4 3 3 m w e o T l 3 0 0 F 2 O 2 0 0 0 2 1 0 0 - 2 0 2 4 6 8 1 0 1 2 A x i a l C o - o r d i n a t e ( m m )

Figure 2.5: DME (), oxygen (N) and temperature proﬁle (•) during DME CPO.

DME dry reforming. Due to competing pathways of CO2 and steam reforming of

DME, CO2 evolution was spread out compared to methane (2 mm vs 3.5 mm) and water concentration proﬁle showed a less steep rise and less sharp peak compared

to methane. After 3.5 mm, during DME CPO, due to reduced temperatures, CO2 reforming does not appear to play a role. Ma et al. reported 100 % conversion

of DME and 80 % conversion of CO2 during dry reforming of DME over Ni-based catalysts at temperatures similar to that involved in the current system at lower space velocities.101 Further experiments are necessary to gain a detailed understanding of the role of dry reforming over Rh-based catalysts during DME CPO. Temperature proﬁle trends were nearly identical in nature for DME and methane showing an initial steep rise due to highly exothermic oxidation chemistry and subsequent decrease due to endothermic reforming chemistry after all the oxygen was consumed (Fig. 2.3 and Fig. 2.5). Though most of the oxygen was consumed within ∼ 2 mm, the temperature proﬁles showed a peak at about 3.2 mm from the front face of the catalyst due to the gas and catalyst surface not being in thermal equilibrium.80,82 Temperatures during DME CPO were higher than methane CPO throughout the 2.3 Results 25

1 6 5 w t % R h c a t a l y s t H 2

1 4

1 2 ) n i 1 0

m H O

/ 2 l o

m 8 m

(

F r o n t h e a t B a c k h e a t

e 6 t s h i e l d s h i e l d a r

w 4 o l F 2

- 2 - 2 0 2 4 6 8 1 0 1 2 A x i a l C o - o r d i n a t e ( m m )

(a) 1 6 5 w t % R h c a t a l y s t 1 4 C O 1 2 )

n 1 0 i m / l

o 8 F r o n t h e a t B a c k h e a t

m s h i e l d s h i e l d m

( 6 e t a r 4 w o l F 2

C O 2 0

- 2 - 2 0 2 4 6 8 1 0 1 2 A x i a l C o - o r d i n a t e ( m m )

(b) Figure 2.6: Product concentration profiles during DME CPO. (a) shows hydrogen() and water(N) concentration profiles. (b) shows carbon monoxide () and carbon dioxide(N) concentration profiles. 2.3 Results 26

2 5 2 5 0 m m 0 m m 2 . 2 m m 2 . 2 m m 2 0 1 0 . 2 m m 2 0 1 0 . 2 m m ) ) n n i i m m / /

l 1 5 l 1 5 o o m m m m

( (

e e

t 1 0 t 1 0 a a r r

w w o o l l F F 5 5

0 0 H O C O C H H O D M E H 2 C O 2 2 4 H 2 C O 2 C O 2

(a) CH4 (b) DME Figure 2.7: Product ﬂow rates at the end of 2.2 mm and 10.2 mm during (a) methane and (b) dimethyl ether catalytic partial oxidation.

Species 2.2 mm 10.2 mm Species 2.2 mm 10.2 mm S 30.1 56.3 S 39.1 53.8 H2 H2 SCO 74.9 73.0 SCO 88.6 82.1 S 25.1 27.0 S 8.9 14.7 CO2 CO2 S 69.9 43.7 S 58.0 42.7 H2O H2O X 25.2 33.5 X 59.1 73.8 CH4 CH3OCH3 (a) (b) Table 2.1: Conversions and product selectivities (both in %) at 2.2 mm (∼ 95 % oxygen conversion) and 10.2 mm (end of catalyst) for (a) methane and (b) dimethyl ether.

catalyst by ∼ 50 ◦C due to the higher reactivity of DME. The ﬂow rates of syngas, combustion products and feed (DME or methane) during methane and DME CPO at various stages along the catalyst are shown in Fig. 2.7. 2.2 mm corresponds to the point where for both methane and DME CPO, ∼ 95 % oxygen conversion was observed. The product selectivities and conversions for methane and DME CPO are shown in Tables 2.1a and 2.1b: Based on the data in Table (2.1a) and (2.1b), the overall equations at 2.2 mm (Eq. 2.10) and 10.2 mm (Eq. 2.11) for methane CPO may be written as: 2.3 Results 27

CH4 + 0.5O2 −→ 0.75CO + 0.25CO2 + 0.6H2 + 1.4H2O ∆Hr = -445 kJ/mol (2.10)

CH4 + 0.5O2 −→ 0.73CO + 0.27CO2 + 1.12H2 + 0.88H2O ∆Hr = -325 kJ/mol (2.11)

Similarly, for DME CPO (neglecting CH4 and C2H6 as their selectivities are small, ≤ 1%), at 2.2 mm (Eq. 2.12) and 10.2 mm (Eq. 2.13):

CH OCH + 1.5O −→ 1.8CO + 0.2CO + 1.2H + 1.8H O 3 3 2 2 2 2 (2.12) ∆Hr = -529 kJ/mol

CH OCH + 1.5O −→ 1.67CO + 0.33CO + 1.65H + 1.35H O 3 3 2 2 2 2 (2.13) ∆Hr = -457 kJ/mol

The above equations highlight the diﬀerences in CO2 concentrations at 2.2 mm and 10.2 mm during methane and DME CPO. The less exothermic heat of reaction at 10.2 mm compared to 2.2 mm is due to the presence of endothermic reforming chemistry between 2.2 mm and 10.2 mm.

2.3.2 State of Rhodium in Oxidation Zone

The product distribution obtained during methane CPO at a total ﬂow rate of 5

SLPM with C/O = 1 and Ar/O2 ratio corresponding to air stoichiometry (79/21) is shown in Table 2.2. The temperature below the back heat shield was 635 ◦C and oxygen breakthrough was observed (∼ 62 % oxygen conversion). XRD and XPS measurements were used to verify the oxidation state of the rhodium catalyst in the oxidation zone. With the used rhodium catalyst, XPS mea-

surements showed the presence of rhodium in the reduced (metallic state) with 3d5/2

and 3d3/2 binding energies of 307.2 and 312.0 eV respectively. The 3d5/2 and 3d3/2 binding energies of the fresh (after calcination) 2.5 wt % rhodium catalyst were 308.3 2.3 Results 28 Species Exit value S 41.4 H2 SCO 60.4 S 39.6 CO2 S 58.6 H2O X 21.7 CH4 X 62.4 O2 Table 2.2: Reactant conversions and product selectivities (both in %) during methane catalytic partial oxidation at C/O = 1 and total ﬂow rate 5 SLPM.

and 313.1 eV respectively indicating presence of rhodium in the oxidized state. Values reported represent the average of XPS scans over two spheres. The XRD pattern of blank alumina spheres, fresh (after calcination) 2.5 wt % Rh spheres and used 2.5 wt % Rh spheres are shown in Fig. 2.8. With the used rhodium catalyst, XRD measurements showed the absence of rhodium oxide and the presence of metallic rhodium peaks at 2 theta values of 41, 48, and 70 degrees. Fresh 2.5 wt % Rh spheres showed the presence of rhodium oxide during XRD. These experiments show that despite the presence of high temperatures and oxygen within the oxidation zone, rhodium is present in the reduced state. This contra- dicts the results of Grunwaldt et al. who reported through in situ X-ray absorption studies that rhodium is present as rhodium oxide in the oxidation zone.98,99 However, those experiments were performed at much lower temperatures (∼ 400 ◦C). At the high surface temperatures within the oxidation zone in a typical CPO reactor (∼ 700-1000 ◦C),82,86 the kinetics of diﬀerent surface reactions are much faster. Previous research by Horn et al. has shown that the CPO process is mass transfer limited.81 This suggests that the kinetics of reactions involving oxygen containing species are very fast compared to reaction of oxygen with rhodium to give rhodium oxide. As a result, the surface concentration of oxygen containing species is very low. This has been shown to be true in simulation studies of methane CPO at comparable temperatures which have shown that OH∗ is the main oxidizer species in the oxidation zone and its coverage is in the order of 0.01 in the oxidation zone.95,96 This shows that within the oxidation zone, the reaction takes place by Regime II as described by Maestri et al.95 which involves primarily direct mechanism of syngas formation along with syngas combustion and water gas shift, and concentration of oxygen on the catalyst surface is almost zero. Spatially resolved concentration and temperature proﬁles were similar for methane 2.3 Results 29

R h p e a k R h a g e d

R h 2 O 3 p e a k ) . u . R h f r e s h a (

t i s n e t n I α b l a n k - A l 2 O 3

2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0 2 θ ( d e g r e e s )

Figure 2.8: XRD patterns for alumina, fresh (calcined rhodium) and used rhodium catalysts. Peaks are normalized to the largest α-alumina peak at 43 degrees. 2.4 Conclusion 30 and DME implying similar mechanism for CPO for diﬀerent molecules. In the absence of surface oxygen, as is likely the case in the current experiments, it has been shown that after adsorption of DME, dehydrogenation steps take place to form CHxOCHx 85,102 which ultimately form syngas, CO and H2 and desorb. It may therefore be expected that similar to methane, CPO of DME also proceeds primarily through a direct mechanism involving DME dehydrogenation (pyrolysis), syngas combustion, and water gas shift reactions within the oxidation zone.

2.4 Conclusion

Spatially resolved concentration and temperature proﬁles for methane and DME CPO were obtained using the capillary sampling technique. With both feedstocks, ∼ 95 % oxygen was consumed within approximately 2.2 mm of the beginning of the catalyst. Syngas formation was observed even within the oxidation zone. Largest temperature and concentration gradients were observed within the oxidation zone. Dry reforming appeared to play a role during DME CPO while no evidence of it was observed during methane CPO. Further experiments are necessary to gain a better understanding of the role of dry reforming during DME CPO. XPS and XRD measurements showed that even in the presence of oxygen, within the oxidation zone, rhodium was present in the reduced state as metallic rhodium indicating within the oxidation zone, syngas formation takes place mostly by the direct mechanism.

2.5 Acknowledgements

The author would like to acknowledge the Minnesota Corn Growers Association for ﬁnancial Support. The author thanks Dr. Jacob Kruger and Samuel Blass for valuable discussions regarding experiments performed in this chapter. CHAPTER THREE

EFFECTS OF BIOMASS INORGANICS ON RHODIUM CATALYSTS: I. STEAM METHANE REFORMING1

In Chapter 2, the complex concentration and temperature gradients existing during catalytic partial oxidation were presented. In this chapter, the effects of inorganics on rhodium catalysts were studied under controlled isothermal conditions, by depositing the inorganics directly on the catalyst and using steam methane reforming at 700 ◦C as a model system. Inorganic elements in biomass present a major challenge for large-scale application of catalytic processing. Rhodium-based catalysts have been shown to gasify lignocellulosic biomass to syngas with high selectivities by reactive flash volatilization. In this research, the effect of inorganics commonly found in biomass, Na, K, Ca, Mg, Si, P, and S, on rhodium catalysts was investigated using steam methane reforming (SMR) as a model system. SMR was carried out at 700 ◦C on a heated fixed bed of 2.5 wt % Rh/α-Al2O3 catalysts. The inorganics were uniformly deposited on the catalyst (1 inorganic atom/5 rhodium atoms), followed by performance testing and characterization through H2 chemisorption, SEM, XPS and XRD. Phosphorus, potassium and sulfur decreased the methane conversion the most (> 15 %) among the inorganics studied. No significant deactivation was observed upon doping with calcium, magnesium and silicon. Phosphorus and sulfur strongly reduced the dispersion

1Parts of this chapter appear in Reetam Chakrabarti, Joshua L. Colby, Lanny D. Schmidt, “Eﬀects of Biomass Inorganics on Rhodium Catalysts: I. Steam Methane Reforming,” Applied Catalysis B : Environmental 107 (2011) 88-94. c 2011 Elsevier B.V. Reproduced with permission from Elsevier.

31 3.1 Introduction 32

of the rhodium catalyst. Addition of phosphorus and potassium caused formation of carbon-based structures, and phosphorus also increased the rhodium binding energies by 0.6 eV in the XPS spectrum, indicating rhodium oxidation.

3.1 Introduction

Fossil fuels such as coal, oil and natural gas have been a main energy source for more than a century. These fossil fuels are carbon-based and have a high energy density.103 Lignocellulosic biomass has potential as an alternative source of energy and represents a sustainable source of renewable carbon to produce fuels and chemicals.24–27 Gasification is a common technique for processing biomass to fuels and chemicals. It produces syngas, a mixture of hydrogen and carbon monoxide that can be upgraded to diesel fuel using the Fischer-Tropsch process or to methanol. Schmidt et al. recently established the reactive flash volatilization technique by which lignocellulosic biomass can be gasified autothermally to a high-selectivity syngas product stream free from tars and chars in millisecond contact times.1–3 This process takes place in an oxygen-deficient environment at high temperatures (600 - 1000 ◦C) over rhodium-based catalysts. One of the major challenges in catalytic gasification is the presence of inorganics in biomass which may alter the catalyst activity and process chemistry. The effect of these inorganics on rhodium catalysts in reactive flash volatilization has not yet been investigated in detail. Inorganic elements present in biomass are Na, K, Ca, Mg, Mn, P, Si, Cl, S, Fe, Al and small concentrations of heavy metals.37,104 Their quantities and concentrations vary depending on the type of biomass. For example, hardwoods and softwoods contain approximately 1 wt % inorganics whereas most annuals contain approximately 3 - 10 % and rice hulls contain up to 25 %.37 Understanding the nature of the interactions of inorganics with rhodium is necessary for efficient processing of biomass by catalytic gasification. During reactive flash volatilization of lignocellulosic biomass, inorganics can alter the activity of the catalyst by any combination of three different phenomena: (a) accumulation of fly-ash near the front face of the catalyst which affects mass transfer of reactants and inorganics as well as heat transfer in the system, (b) poisoning due to strong chemisorption on active sites and (c) fouling or physical blocking of the active sites. Mechanisms (a) and (c) may be expected to be more predominant with less volatile inorganics such 3.1 Introduction 33 as Ca and Mg. The objective is to understand the interactions of the inorganics with the active sites due to mechanisms (b) and (c). Direct introduction of inorganics in the feed makes it difficult to decouple mechanism (a) from mechanisms (b) and (c). Therefore, the inorganics were deposited uniformly throughout the catalyst prior to reaction. This system is not representative of how actual biomass-derived inorganics would ordinarily reach the catalyst. However, by eliminating the effect of the inorganics being filtered out near the front face, it is possible to attribute the deactivation to mechanisms (b) and (c). Steam methane reforming (SMR) was chosen as a model system to study the effect of inorganics on the rhodium-based catalyst. The primary reactions in SMR are:

CH4 + H2O −→ CO + 3 H2 (3.1)

CO + H2O −→ H2 + CO2 (3.2) SMR represents a controlled and well-defined system since it is kinetically lim- ited105,106 and the conversion of methane can be related to the activity of the rhodium catalyst. Also, methane exhibits very little homogeneous chemistry at the temperatures involved.107 Therefore, the change in conversion of methane can be directly related to the change in activity of the rhodium catalyst introduced upon addition of inorganics. By performing SMR over undoped and inorganic-doped rhodium catalysts, the effect of inorganics on the rhodium sites involved in reforming to produce syngas can be studied. It has been shown that the catalyst in reactive flash volatilization consists of three sections- volatilization, oxidation and reforming.3 Reforming, the downstream section of the catalyst in reactive flash volatilization, involves water gas shift and steam reforming of hydrocarbons to give a product stream with high selectivities to syngas. SMR simulates the reforming zone in reactive flash volatilization as it involves similar types of reactions. Also, by using methane as a probe, homogeneous chemistry in the system can be neglected.107 The effect of common biomass inorganics was examined by adding them through aqueous precursors to the rhodium-based catalyst at a concentration of 1 atom of inorganic for every 5 atoms of rhodium. The effect of these inorganics was studied by performance testing for both undoped and doped catalysts for SMR at 700 ◦C on 2.5 wt % Rh/α-Al2O3 catalysts. Catalysts were characterized by H2 pulsed chemisorp- 3.2 Experimental 34 tion, SEM, XRD and XPS before and after doping to study dispersion, microstruc- ture, crystallite formation and electronic interactions respectively upon introduction of inorganics. In Chapter 4, inorganics were introduced in the feed for reactive flash volatilization to study their effects on rhodium catalysts.108 In this study, we directly deposit them on the catalyst; which eliminates any deactivation that may occur by accumulation of inorganics near the front face of the catalyst when introducing them in the feed. Both techniques are necessary to gain a detailed understanding of the interaction of inorganics on rhodium catalysts for biomass processing applications.

3.2 Experimental

3.2.1 Experimental Setup

SMR was carried out in a quartz reactor (20 mm I.D.) over 3 g of 2.5 wt % Rh supported on 1.3 mm diameter α-Al2O3 spheres. A blank 80 ppi α-Al2O3 monolith (17 mm diameter, 10 mm long) was used to support the catalyst bed (Fig. 3.1A inset). Two blank 45 ppi α-Al2O3 monoliths were placed about 1cm upstream of the 80 ppi monolith which acted as static mixers. The monoliths were held against the reactor walls by wrapping them with aluminosilicate cloth. Mass flow controllers regulated the flow rate of gases (N2,H2, CH4) to the reactor, accurate to within ± 2 %. Water was fed using a liquid handling pump through a heated coil maintained at approximately 200 ◦C to generate steam. A type-K thermocouple was placed between the two 45 ppi monoliths and a benchtop temperature controller (Omega CSC 32) controlled the temperature between the monoliths to within ± 1 ◦C. The entire reactor tube was enclosed within a tube furnace. Reforming being an endothermic process, this arrangement helped to insulate the reactor from the heat effects of the reaction.

3.2.2 Product Analysis

Products were analyzed with an Agilent 6890 Gas Chromatograph equipped with a 30 m Plot-Q column and both thermal conductivity and ﬂame ionization detectors. Water was quantiﬁed by closing the hydrogen and oxygen balances and averaging the two results. During operation, the carbon, hydrogen and oxygen balances generally closed to within ± 5 %. 3.2 Experimental 35

Selectivities to CO and CO2 were calculated on a carbon basis and to H2 and H2O were on a hydrogen basis. Selectivity to a product is deﬁned as (atoms in product species)/(atoms in converted fuel). Steam fed to the reactor was not considered fuel. The sum of selectivities to products on carbon or hydrogen basis was 100 %, within the limits of experimental error.

3.2.3 Catalyst Preparation

Catalysts used in all the experiments were 2.5 wt % Rh on α-Al2O3 spheres (1.3 mm diameter, Saint Gobain Norpro). Catalysts were prepared by the incipient wetness technique [8]. Rhodium nitrate solution (Sigma-Aldrich) was used as the rhodium precursor. Catalysts were dried, then calcined in a furnace at 800 ◦C for 6 h.

3.2.4 Experimental Procedure

Each catalyst was initially aged at 850 ◦C for 3 h in a mixture of 20 % steam, 10 % methane and 70 % nitrogen at a total ﬂow rate of 2 SLPM. Subsequently, the temperature was reduced to 700 ◦C and the performance of the catalyst for SMR was measured. The residence time within the catalyst bed was approximately 20 ms at 700 ◦C. The inorganic species of interest was added to the catalyst (one inorganic atom for every ﬁve rhodium atoms) through its precursor by the incipient wetness technique and its performance was compared with the undoped catalyst at 700 ◦C.

3.2.5 Catalyst Characterization

SEM images were taken on JEOL 6700 equipped with a secondary electron detector. Most samples were coated with 100 A.U. of carbon to reduce charging effects. XRD was carried out on a Bruker D 5005 diffractometer equipped with a 2.2 kW sealed Cu source. Spheres were crushed to a fine powder increasing the peak intensities during analysis. XPS studies were carried out on Surface Science SSX 100 with a monochromatic Al K-α source (1486.6 eV). The spot size used in the measurements corresponded to 800 m. All peaks were calibrated with respect to the C 1s peak at 285 eV.

H2 pulsed chemisorption was carried out by injecting pulses of hydrogen from a 5 % hydrogen in argon mixture into a quartz tube containing 0.2 g of the catalyst 3.3 Results 36

sample. 15 pulses were injected into the catalyst at 1 minute intervals at 25 ◦C with a dwell time of 15 s. The outlet was connected to a thermal conductivity detector.

3.3 Results

The eﬀect of inorganics on rhodium was studied by measuring the reactor performance

for 5 h and through characterization by H2 chemisorption, SEM, XPS and XRD before and after doping with inorganics.

3.3.1 Performance Testing

All of the undoped catalysts showed similar baseline performance with about 67 % methane conversion. Hydrogen and carbon monoxide selectivities were about 68 % and 49 % respectively. These parameters were used as an indicator of reactor performance and plotted with time for each inorganic during the 5 h operation period. The changes in conversions and selectivities are expressed in terms of the absolute diﬀerence between the average doped (over 5 h) and undoped values. For example, if the conversion decreased from 60 % to 30 %, this is equivalent to (60-30) = 30 % and not 50 % decrease in conversion.

Sulfur

Dimethyl sulfoxide was used to add sulfur to the catalyst. Sulfur decreased the methane conversion the most among all the inorganics studied. However, partial regeneration of the catalytic activity was observed during the 5h test period. Methane conversion (Fig. 3.1A) and hydrogen selectivity increased (Fig. 3.1B) with time from 19 % and 32 % respectively at the beginning to 31 % and 42 % at the end of 5 h. The CO selectivity increased by about 10 % and was almost constant throughout the 5h duration (Fig. 3.1B).

Phosphorus

Ammonium phosphate was used as a precursor for introducing phosphorus to the catalyst. Phosphorus decreased the methane conversion by about 17 % (Fig. 3.1C).

Also, CO selectivity increased by 22 % whereas H2 selectivity decreased by 13 % (Fig. 3.3 Results 37

3.1D). After the 5 h performance testing period, a signiﬁcant fraction of the spheres appeared black due to coke formation.

Silicon

Tetraethylorthosilicate (TEOS) was added to the aged rhodium catalyst to introduce silicon. The changes in methane conversion and product selectivities (not shown) were within the limits of experimental error.

Sodium and Potassium

Acetate precursors were used to add sodium and potassium to the catalyst. Sodium and potassium decreased methane conversions by 9 % and 16 %, respectively (Fig.

3.2, A and C). Potassium decreased the H2 and CO selectivities by 5 % and 7 % respectively, whereas the decrease with sodium was within experimental error as shown in Fig. 3.2, B and D.

Calcium and Magnesium

Calcium acetate and magnesium acetate tetrahydrate were used as calcium and magnesium precursors. Conversions and syngas selectivities were unchanged within limits of experimental error for both calcium and magnesium (not shown). The results of performance testing are summarized in Table 3.1.

3.3.2 Catalyst Characterization

H2 chemisorption

The eﬀect of the inorganics on the dispersion of rhodium on the α-Al2O3 support was analyzed by H2 pulsed chemisorption. All the aged rhodium catalysts had a rhodium dispersion of approximately 10 %. Phosphorus and sulfur decreased the dispersion of the catalyst (10.7 % to 3.6 % for phosphorus, 9.5 % to 4.4 % for sulfur). For the other inorganics, the change in dispersion was within the limits of experimental error.

SEM

Images were taken for each of the catalysts studied. Fresh, aged and Na, Ca, Mg, Si, and S doped catalysts appeared similar in terms of morphology, each showing 3.3 Results 38

S u l f u r P h o s p h o r u s 8 0 8 0 A C 7 0 7 0

) P r o d u c t s t o G C a n d i n c i n e r a t o r %

( 6 0 6 0

n α− o R h / A l O c a t a l y s t i 2 3 s 5 0 5 0 r 8 0 p p i f o a m e v

T h e r m o c o u p l e n 4 5 p p i f o a m s R h d o p e d w i t h S

o 4 0 4 0 c

U n d o p e d R h e N i t r o g e n , M e t h a n e , S t e a m n

a 3 0 3 0 h t e R h d o p e d w i t h P M 2 0 2 0 U n d o p e d R h 1 0 1 0 0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 T i m e ( m i n ) T i m e ( m i n ) 8 0 8 0 B D C O H 7 0 2 7 0 H 6 0 6 0 2

) C O %

( 5 0 5 0

y C O t i

v i

t 4 0 4 0 c H 2 e l e C O s e l e c t i v i t y ( S d o p e d ) C O s e l e c t i v i t y ( P d o p e d ) S 3 0 3 0 C O s e l e c t i v i t y ( u n d o p e d ) C O s e l e c t i v i t y ( u n d o p e d ) H s e l e c t i v i t y ( S d o p e d ) 2 0 2 2 0 H 2 s e l e c t i v i t y ( P d o p e d ) H s e l e c t i v i t y ( u n d o p e d ) H 2 s e l e c t i v i t y ( u n d o p e d ) 2 1 0 1 0 0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 T i m e ( m i n ) T i m e ( m i n )

Figure 3.1: Methane conversion for catalysts doped with sulfur (A) and phosphorus (B) respectively. Hydrogen, carbon monoxide selectivities for catalysts doped with sulfur (C) and phosphorus (D) respectively. (A) also shows a schematic of the reactor setup. 3.3 Results 39

P o t a s s i u m S o d i u m 7 0 7 0 A C

6 5 6 5 ) % (

o R h d o p e d w i t h K i 6 0 6 0 s

r U n d o p e d R h e v

n o

c 5 5 5 5

e n a h

t R h d o p e d w i t h N a

e 5 0 5 0 U n d o p e d R h M

4 5 4 5 0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 T i m e ( m i n ) T i m e ( m i n ) 8 0 8 0 B D H 7 0 7 0 2 H 2 6 0 6 0 ) % (

y t

i 5 0 5 0

v C O i

t C O c e l 4 0 4 0 e C O s e l e c t i v i t y ( N a d o p e d )

S C O s e l e c t i v i t y ( K d o p e d ) C O s e l e c t i v i t y ( u n d o p e d ) C O s e l e c t i v i t y ( u n d o p e d ) H s e l e c t i v i t y ( N a d o p e d ) 3 0 H s e l e c t i v i t y ( K d o p e d ) 3 0 2 2 H s e l e c t i v i t y ( u n d o p e d ) H s e l e c t i v i t y ( u n d o p e d ) 2 2 2 0 2 0 0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 T i m e ( m i n ) T i m e ( m i n )

Figure 3.2: Methane conversion for catalysts doped with potassium (A) and sodium (B) respectively. Hydrogen, carbon monoxide selectivities for catalysts doped with potassium (C) and sodium (D) respectively. 3.3 Results 40

Inorganic -∆ X (%) ∆ S (%) ∆ S (%) CH4 CO H2 S(t=0) 48 8 -36 S(t=5h) 36 11 -26 P 17 22 -13 Si ∼ 0 ∼ 0 ∼ 0 Na 9 ∼ 0 ∼ 0 K 16 -7 -5 Mg ∼ 0 ∼ 0 ∼ 0 Ca ∼ 0 ∼ 0 ∼ 0 Table 3.1: Changes in methane conversion (∆ X ), carbon monoxide (∆ S ) and CH4 CO hydrogen selectivities (∆ S ) at 700 ◦C upon doping with different inorganics (1 H2 atom inorganic/ 5 atoms of rhodium) as compared to aged undoped catalyst (2.5 wt % Rh/α-Al2O3). Average values of methane conversion, carbon monoxide and hydrogen selectivities were 67 %, 49 % and 68 % whereas equilibrium values at 700 ◦C were 99 %, 73 % and 81 % respectively.109 rhodium particles around 50 nm in diameter (Fig. 3.3A). Carbon formation was observed on the surface of catalysts doped with phosphorus and potassium. In the case of phosphorus-doped catalyst, the carbon formed was filamentous (Fig. 3.3B) with rhodium particles at the tips (Fig. 3.3C), whereas the carbon formed with potassium-doped catalyst showed the presence of needle-like structures (Fig. 3.3D). After placing the phosphorus-doped catalyst in a furnace at 500 ◦C for 30 min, the filamentous carbon on the catalyst was almost completely eliminated. The carbon on the potassium-doped catalyst remained even on heating at 750 ◦C for 30 min.

XPS

The electronic interactions of the inorganics with rhodium were analyzed by XPS.

The aging process decreased the rhodium binding energies of 3d5/2 and 3d3/2 peaks +3 by about 1 eV showing conversion of rhodium oxide (Rh2O3, Rh ) to rhodium metal (Rh0). Sodium, calcium, magnesium and silicon lowered the binding energy of rhodium by about 0.3 eV, whereas potassium decreased it by about 0.6 eV compared to the aged rhodium catalyst. Sulfur addition caused no change whereas phosphorus increased the rhodium binding energies compared to the aged rhodium catalyst by approximately 0.6 eV. 3.3 Results 41

Figure 3.3: SEM images of (A) fresh 2.5 wt% Rh on α-Al2O3 catalyst (B) carbon ﬁlaments on catalyst doped with phosphorus (C) high resolution image of carbon ﬁlaments in catalyst doped with phosphorus showing rhodium particles at tip and (D) carbon structures on catalyst doped with potassium. 3.4 Discussion 42

XRD

To examine the possibility of formation of any crystallites, XRD studies were performed. During the aging process, XRD analysis showed that all the rhodium oxide was converted into rhodium metal. No additional peaks were observed upon doping with inorganics. With phosphorus, additional small peaks were observed at 33 and

47 degrees corresponding to Rh2P, rhodium phosphide.

3.3.3 Equilibrium Calculations

Equilibrium calculations were performed to study the eﬀect of the inorganics on the equilibrium of steam reforming (3.1) and water gas shift (3.2). The equilibrium constant for (3.1) and (3.2) at 700 ◦C was 12.1 and 1.6 respectively.109,110 The experimental value of the proportionality constant for reactions (2.3) and (2.5) and was calculated by the following equations :

3 PH PCO K = 2 for steam reforming exp,SR P P CH4 H2O

PH PCO K = 2 2 for water gas shift eq,W GS P P CO H2O For the undoped rhodium catalysts, equilibrium analysis indicated that methane steam reforming is not equilibrated, whereas the water gas shift reaction reached equilibrium (Kexp,W GS/Keq,W GS ∼ 1) showing SMR is kinetically limited by (1). Even upon addition of inorganics, water gas shift reaction was equilibrated for calcium, magnesium, sodium, potassium and silicon as the ratio (Kexp,W GS/Keq,W GS) was approximately 1. Doping with sulfur and phosphorus prevented the water gas shift reaction from proceeding to equilibrium as this ratio was about 0.3 and 0.25 respectively, showing the poisoning eﬀect of sulfur and phosphorus on water gas shift activity.

3.4 Discussion

The eﬀect of inorganics on rhodium is a combination of two parameters: the electronic interactions which may result in poisoning of catalytic activity, and the physical blocking of active sites due to carbon formation or volatility of the inorganic species. 3.4 Discussion 43

Sodium, potassium, calcium and magnesium are alkaline and electropositive, whereas phosphorus and sulfur are electronegative and acidic. In terms of volatility, the alkaline elements tend to form non-volatile species. Since acetates were used as precursors in the current experimental setup, they are probably present as carbonates/oxides at the high temperatures in the system.111–113 Non-metals such as phosphorus and sulfur most likely exist in the reduced form such as phosphine and hydrogen sulﬁde due to the reducing environment in SMR. The eﬀect of each inorganic is discussed below:

3.4.1 Sulfur

Poisoning of catalytic activity by sulfur has been attributed to a combination of geometric (structural changes) and electronic effects (affecting chemisorption).114 A single atom of sulfur can poison multiple catalyst sites around it, which decreases the catalyst dispersion from 9.5 % to 4.4 %. This explains the strong decrease in methane conversion upon doping with sulfur (>30 %). Dimethyl sulfoxide, the sulfur precursor used in the experiment on heating decomposes to sulfur dioxide. Sulfur dioxide dissociatively adsorbs on noble metals to give chemisorbed sulfur and oxygen atoms on the catalyst surface.115 Sulfur chemisorbs strongly on the catalyst surface thereby limiting the availability of rhodium active sites and decreasing methane conversion. It is interesting to note that the methane conversion increased from 19 % at the beginning to 32 % at the end of 5 h indicating partial regeneration of the catalyst. Similar regeneration of a sulfur poisoned rhodium catalyst has been observed upon switching from a sulfur containing feed to a sulfur free feed in methane catalytic partial oxidation.83,116 The regeneration of catalytic activity can be explained by the fact that sulfur can also desorb from the catalyst surface by the hydrogenation 117 reactions S+H → HS and HS+H → H2S. No shift in the rhodium binding energies in the XPS spectrum was observed indicating the absence of bulk sulfides/sulfates or other sulfur containing species. The sulfur concentration on the catalyst may also be below the detection limits of the instrument. The CO selectivity (Fig. 1B) was about 10 % greater than that with the undoped catalyst throughout the performance testing period which may be due to a combination of two phenomena. Firstly, in contrast with the undoped catalyst, the water gas shift reaction was not equilibrated upon doping with sulfur, indicating poisoning of the water gas shift reaction by sulfur. On noble metals such as rhodium, 3.4 Discussion 44

steam reforming and water gas shift occur through a bifunctional mechanism which involves reaction between hydroxyl groups formed by water activation on the support and carbonaceous species adsorbed on the metal.118,119 Sulfur inhibits hydroxyl group formation and migration for reaction which results in poisoning of water gas shift and steam reforming.83,120,121 This, in addition to the reduced rhodium dispersion results in poisoning of water gas shift and steam reforming. Secondly, electronegative elements have been shown to inhibit CO dissociation by increasing the work function of the metal through electrostatic interactions.122 This inhibition can result in increased CO selectivity by reducing the rate of CO dissociation.

3.4.2 Phosphorus

Doping with phosphorus decreased the methane conversion from 68 % to 51 % (Fig.

1C) and increased CO and H2O selectivities while decreasing H2 and CO2 selectivities (Fig. 1D). The CO selectivity increased by about 20 % compared to the undoped catalyst, likely due to mechanisms similar to that with S (poisoned water gas shift and increased work function122). SEM images showed the formation of carbon filaments on the catalyst after performance testing. Rhodium nanoparticles were observed at the tips of these filaments as seen in 3.3C. Such filamentous carbon, also called whisker carbon, has been observed previously on nickel catalysts.123,124 In the case of SMR, carbon formation on the surface takes place through two reactions: methane decomposition CH4 → C +

2 H2 and the Boudouard Reaction 2CO → CO2 + C. On nickel, it has been suggested that the surface carbon diffuses through the nickel particle from the gas side to the support/filament side leading to filament growth.125 The net growth rate is also in- fluenced by the rate of carbon gasification. Noble metal-based catalysts have been shown to have a high resistance to carbon formation in SMR.126 Formation of such whisker carbon has not been observed previously on rhodium catalysts as the surface carbon is gasified at a faster rate than its formation, thereby no filament growth takes place. Phosphorus poisons the surface-carbon gasification reactions, leading to accumulation of carbon on surface of rhodium. Thus, by a mechanism similar to that in nickel,125 filament growth takes place by diffusion of carbon atoms through the rhodium particle due to a concentration gradient of carbon between the rhodium-gas 3.4 Discussion 45

and rhodium-support boundaries. This results in structures as shown in 3.3C, with rhodium particles at the tip of carbon ﬁlaments. In addition to rhodium at the tips of the carbon ﬁlaments, there may still be some

rhodium on the α-Al2O3 support. The decrease in methane conversion may caused by deactivation of rhodium at either location by phosphorus or through diffusion limitations by filaments preventing access to the rhodium particles on the surface. Further work is necessary to determine the location of phosphorus and the nature of its interaction with rhodium at both locations to determine the relative contributions of the above mechanisms. Phosphorus increased the rhodium binding energy in the XPS spectrum by about 0.6 eV indicating oxidation of rhodium due to interaction with phosphorus. Complex formation with rhodium (in the form of rhodium phosphide) and carbon filament formation can contribute towards decreasing the rhodium dispersion, thereby decreasing methane conversion. Phosphorus has been shown to poison noble metal catalysts by hindering chemisorp- tion77,127 and causing structural rearrangement.77,127,128 The carbon filaments formed may also inhibit mobility of surface species. Along with reduced rhodium dispersion, the above mechanisms inhibit steam reforming (1) and water gas shift (2), preventing the latter from being equilibrated.

3.4.3 Silicon

Silicon caused no change in methane conversion and product selectivities. TEOS under high temperatures129 or in the presence of steam is converted to silica.130 At high temperatures, silica can be steamed out which reduces the concentration of 131 silicon on the catalyst. This is represented by the reaction SiO2 + H2O → Si(OH)4 and this reduced silicon concentration results in negligible change in dispersion and activity of the rhodium catalyst.

3.4.4 Sodium and Potassium

The effect of alkali metals on reforming systems has contradictory results in the literature. Alkali metals, especially in low concentrations, have been used as promoters for different reaction systems such as NOx oxidation and reforming reactions.132 Rostrup-Nielsen and Christiansen reported the poisoning effect of alkali metals on 3.5 Conclusions 46 nickel catalysts for SMR.133 They ascribed the poisoning effect to reconstruction of the metal surface by the alkali. In the current system, both sodium and potassium decreased the methane conversion. Potassium decreased syngas selectivities whereas negligible changes were observed with sodium. Unlike sulfur and phosphorus, the water gas shift reaction was still equilibrated despite catalyst poisoning. The potassium doped catalyst showed formation of needle-like structures which were stable upon burnoff in air at 750 ◦C. The higher stability of carbon species on a potassium-doped catalyst as compared to that on phosphorus-doped catalyst can be attributed to its graphitic nature which is more stable and unreactive.134 Potassium being more electropositive than sodium causes a larger decrease in rhodium work function122 and rhodium binding energies (0.6 eV for potassium, 0.3 eV for sodium). The decreased rhodium work function upon doping with potassium enhances CO dissociation and decreases CO selectivity. The decreased rhodium binding energy indicates electron transfer from potassium to rhodium which decreases the rhodium oxidation state. Potassium compared to sodium, results in stronger electronic modifications of the rhodium catalyst leading to greater poisoning of catalytic activity.

3.4.5 Calcium and Magnesium

The alkaline earth metals calcium and magnesium caused a slight decrease in rhodium binding energies (0.3 eV) in the XPS spectrum due to their electropositive nature. Calcium and magnesium being less electropositive and basic than sodium and potassium, resulted in weaker interactions with the rhodium catalyst and hence a negligible change in conversion.

3.5 Conclusions

The effects of inorganics commonly found in biomass on rhodium catalysts were studied using steam methane reforming (SMR) as a model system over a fixed bed of ◦ Rh/α-Al2O3 catalysts at 700 C. This was done by measuring the reactor performance and catalyst characterization by H2 chemisorption, SEM, XPS and XRD for both undoped and doped catalysts. Among all the inorganics studied, sulfur decreased the methane conversion the most. The conversion gradually increased with time indi- 3.6 Acknowledgements 47 cating partial regeneration of the catalyst. While phosphorus, potassium and sodium also decreased the methane conversion, calcium, magnesium and silicon caused negligible changes in methane conversion. The greater poisoning effect of alkali metals as compared to the less volatile alkaline earth metals is attributed to their stronger electropositive nature leading to stronger electronic interactions. The solubility of silica in steam resulted in no poisoning effect with silicon. Phosphorus resulted in formation of carbon filaments with rhodium particles at the tip which are formed by diffusion of surface carbon across the rhodium particle. Calcium, magnesium and silicon represent three of the common biomass inorganics and showed negligible change on the catalytic activity of rhodium. Other components such as potassium and phosphorus showed strong deactivation. Since the chemistry of the reforming zone during processing of lignocellulosic biomass by reactive flash volatilization is similar to that in SMR, it can expected that biomass sources containing higher concentrations of calcium, magnesium and silicon can demonstrate high reforming activity over time to produce syngas. Further research will focus on process and catalyst modification to improve tolerance towards inorganic components.

3.6 Acknowledgements

This work was supported by General Electric Global Research under United States Department of Agriculture Award No. DE-FG36-O8GO18085. Parts of this work were carried out in the University of Minnesota I.T. Characterization Facility, which receives partial support from NSF through the NNIN program. The authors would also like to acknowledge Dr. Brian C. Michael and Dr. Lingzhi Zhang (GE global research) for insightful discussions. CHAPTER FOUR

EFFECTS OF BIOMASS INORGANICS ON RHODIUM CATALYSTS: II. AUTOTHERMAL REFORMING OF ETHANOL1

In Chapter 3, the effects of biomass inorganics by predosing them directly on the catalyst with isothermal steam methane reforming as a model system were presented. In this chapter, to simulate how inorganics in actual biomass are distributed over rhodium catalysts during operation, inorganics were dissolved in an ethanol feed and their effect was studied by catalytic partial oxidation. Biomass contains inorganic elements such as Na, K, Ca, Mg, Si, S, and P which vary in concentration depending upon the source of biomass and can influence the chemistry in biomass-processing techniques. The effect of these inorganic constituents in biomass on Rh catalysts for catalytic gasification by ethanol autothermal reforming has been studied. The same concentration of each inorganic element (0.05 % atoms of inorganic/mol ethanol) was added through a precursor to ethanol, a model compound for biomass reforming in which the precursors were soluble. Reactor performance during doping and regeneration was used to analyze effects of inorganics on the Rh catalyst and to test the ability for regeneration. Of these elements, K, P, and Mg-doped ethanol operated autothermally over the 6 h test period whereas Si, S, Na, and Ca quenched the reaction. The ability of the

1Reproduced in part with permission from Reetam Chakrabarti, Sarah A. Tupy, Lanny D. Schmidt, “Eﬀects of Biomass Inorganics on Rhodium Catalysts: II. Autothermal Reforming of Ethanol,” Energy and Fuels 25 (2011) 4763-4769. c 2011 American Chemical Society.

48 4.1 Introduction 49 catalyst to be regenerated after doping was investigated by switching to inorganic- free fuel. Regeneration was possible for each of the inorganics studied. Silicon shut down autothermal operation the fastest, but complete regeneration of the catalyst was achieved by passing pure ethanol. Sulfur caused extensive coke formation and disintegration of the α-Al2O3 support. Phosphorus increased ethanol dehydration to ethylene, indicating acid chemistry which was observable even upon regeneration. Sodium and potassium increased acetaldehyde selectivities due to their basic nature.

Calcium and magnesium accumulated at the front face of the catalyst forming CaCO3 and MgO respectively. No liquid slag formation was observed with any of the inorganics.

4.1 Introduction

Biomass-derived fuels represent a renewable, carbon-based alternative to fossil fuels. Currently, there are three major processes for upgrading biomass: gasification to produce syngas, pyrolysis to give bio-oil, and hydrolysis to give sugars.36,135 Gasifica- tion is a high temperature process (> 500 ◦C) and can be catalytic or non-catalytic. The advantages of catalytic gasification are faster reaction rates and higher selectivities to desired syngas products. The syngas stream can then be converted to diesel (Fischer-Tropsch process), methanol, or mixed alcohols. Reactive flash volatilization has been shown to autothermally convert lignocellulosic biomass to a high selectivity syngas stream using Rh-based catalysts at temperatures of 600-1000 ◦C.1–3 The reaction takes place at millisecond residence times within the catalyst without tar or char formation. However, the effect of biomass inorganics on this process has yet to be explored. Inorganics are important components of biomass. Common inorganics found in biomass are Si, Ca, K, Na, Mg, S, P, Cl and heavy metals in small quantities.37,136 These inorganics are present in different forms such as chlorides, sulfates, oxides, silicates, carbonates and phosphates. There is a wide distribution in the composition of inorganics obtained from different biomass feedstocks as shown in Table 4.1.137–153 There is also a large variation of individual inorganic species. For example, rice hulls have a higher silica content than other biomass sources150 while algae contain alkali and alkaline earth metal salts in significant concentrations.154 Most research on the effect of inorganics on biomass processing has been carried out for pyrolysis and 4.2 Experimental 50

Source Inorganic content(wt%) corn cobs137,138 1.5-2.1 corn stover139,140 5.3-7.5 corn kernels141 1.2-2.4 hardwoods142 0.4-3.6 softwoods143,144 0.2-0.5 switchgrass145,146 4.6-7 algae147,148 10-27 rice hull149,150 16.5-20 coal151–153 1.2-27 Table 4.1: Inorganic Content in Weight Percentage Dry Basis from Different Biomass Sources combustion techniques. It has been observed that the inorganics themselves have catalytic properties, thereby altering process chemistry, and the amount and type of inorganic influence the rate of degradation and char formed in the process.37,155 In catalytic gasification, besides acting as individual catalysts, these inorganics can also interact with the catalyst active sites either through physical blocking (fouling) or chemical/electronic interaction or a combination of both.156 These interactions may result in inorganics altering the catalyst activity. Therefore, these inorganics present important technical challenges that must be understood and resolved to make autothermal catalytic gasification a feasible option. In Chapter 3, the catalysts were doped directly with inorganics for steam methane reforming and performance testing and catalyst characterization were used to understand interaction of inorganics with Rh.156 Here, we introduce the inorganics in ethanol feed (ethanol, a biomass surrogate) and survey their effects on Rh catalysts for reactive flash volatilization. The effect of the inorganics on the reactor performance during doping and regeneration was studied by monitoring the process parameters such as temperature, pressure and product distributions.

4.2 Experimental

The experimental setup is similar to that used by Rennard et al.73 Argon was used to create a ﬁne spray of ethanol by a nebulizer. Argon also acted as a diluent and internal standard in the Gas Chromatograph (GC). Mass ﬂow controllers accurate to 4.2 Experimental 51

within ± 2 % fed gases to the reactor at a controlled flow rate. The catalyst consisted of two cylindrical α-alumina 45 ppi (pores per linear inch) foams (Hi Tech ceramics) with a porous structure similar to that in.157 Each foam (17 mm diameter, 5 mm in length) was coated with 5 wt % Rh from Rh nitrate precursor (Sigma Aldrich) via incipient wetness impregnation,3 followed by drying and calcination in air at 600 ◦C for 6 h. The Rh-coated monoliths were placed on a 45 ppi α-alumina foam to prevent heat losses due to radiation. The reactor tube was made of quartz, and the combined assembly (two 45 ppi Rh foams and α-alumina foam) was pressed against the wall of the reactor by aluminosilicate cloth to prevent bypassing of gases. The reactor was surrounded by one inch thick insulation to minimize heat loss. The temperature at the back of the catalyst was measured with an Omega type-K thermocouple. The pressure drop across the catalyst bed was measured with a pressure transducer. The products were analyzed with a HP 5890 Series II GC with a 60 m Plot-Q capillary column with thermal conductivity and flame ionization detectors. Carbon, hydrogen and oxygen balances typically closed within ± 10 % and 95 % confidence intervals were generally within 5 %. The reactor was typically ignited with methane, and then the feed was switched to ethanol. The ethanol flow rate for all experiments was 2 ml/min fed through a HPLC pump. The oxygen and argon flow rates were 0.55 and 2.07 SLPM respectively, corresponding to air stoichiometry and a C/O ratio of 1.5 (carbon atoms in fuel/oxygen atoms from air). The concentration of each inorganic species in ethanol was kept constant at 0.05 % (atoms of inorganic atom)/mol ethanol, comparable to that found in hardwoods and softwoods.142–144 Each fresh catalyst was aged for one hour under operating conditions and then operated for another hour without inorganic to measure baseline performance with pure ethanol. This was followed by operation with inorganic-doped ethanol for up to 6 h. After doping, regeneration was also carried out with pure feed for 4 h. During doping, if the temperature dropped below 600 ◦C, the operation was shut down manually. All of the inorganic precursors chosen were completely soluble in ethanol. The product distribution was analyzed by calculating ethanol conversions (4.1) and product selectivities (4.2). Selectivities to carbonaceous products were calculated on a carbon basis and that to hydrogen and water on a hydrogen basis. FEtOH,out XEtOH (%) = 1 − ∗ 100 (4.1) FEtOH,in 4.3 Results and Discussion 52

 

 νi,C/H Fi  S(C/H)i(%) =  X  ∗ 100 (4.2)  νj,C/H Fj  j6=fuel F denotes the ﬂow rate of a particular species and ν denotes the number of atoms of carbon or hydrogen in that species. To identify the phase of the inorganic on the catalyst, XRD studies were performed on Bruker D 5005 diﬀractometer equipped with a 2.2 kW sealed Cu source.

4.3 Results and Discussion

At 0.05 % of inorganic in ethanol, approximately 30 mg of the inorganic atom is processed during doping for an hour (assuming atomic weight of inorganic = 30 g/mol, since the range of atomic weights in the current set is Na = 23 g/mol to Ca = 40 g/mol). Most inorganics operated autothermally for at least 100 min. Since the weight of Rh is approximately 100 mg, this shows that the catalyst can process signiﬁcant quantities of certain inorganics.

4.3.1 Silicon

Silicon was added to ethanol in the form of tetraethylorthosilicate (TEOS, Si(OC2H5)4). The reactor shut down within 5 minutes of introduction of Si after light oﬀ (Fig. 4.1A). During Si addition, the temperature decreased steadily, then dropped rapidly and autothermal operation was no longer sustainable. Upon regeneration by processing with undoped ethanol, the temperature dropped after methane was cutoﬀ; however, the rate of temperature drop was slightly slower than that during addition of Si-doped ethanol. After dropping to around 520 ◦C, it gradually increased and attained the same temperature as with pure ethanol. The product distribution in terms of selectivities and ethanol conversion during regeneration was similar to pure ethanol before doping (not shown). This behavior of Si can be explained by the following hypothesis. TEOS decom- 129,130 poses to SiO2 on heating or in the presence of water. Since the rate of formation of water in the reactor is approximately 500 times the rate of input of TEOS and also because of the high temperatures involved in the reactor, TEOS is most likely de- 4.3 Results and Discussion 53

A B 9 0 0 9 0 0 R e g e n e r a t i o n N o S i 0 . 0 5 % S i R e g e n e r a t i o n N o S 0 . 0 5 % S 8 0 0 m e t h a n e 8 0 0 7 0 0 c u t o f f 7 0 0 ) )

C 6 0 0 C 6 0 0 o o

e t h a n o l ( ( i n t r o d u c t i o n

e 5 0 0 e 5 0 0

r m e t h a n e r u u t 4 0 0 c u t o f f t 4 0 0 a l i g h t o f f a r r e e

p 3 0 0 p 3 0 0 R e a c t o r s h u t

m m d o w n w i t h

e e N o S 2 0 0 N o S 2 0 0 T T 0 . 0 5 % S p u r e e t h a n o l 0 . 0 5 % S i R e g e n e r a t i o n 1 0 0 R e g e n e r a t i o n 1 0 0 0 0 6 0 1 2 0 1 3 0 1 4 0 1 5 0 1 6 0 6 0 1 0 0 1 4 0 1 8 0 2 2 0 2 6 0 3 0 0 T i m e ( m i n ) T i m e ( m i n )

Figure 4.1: Doping and regeneration temperature profiles for Si (A). Doping and regeneration temperature profiles for S (B) composing to silica. Silica acts as a poison to Rh; hence autothermal operation shuts down quickly. Silicon has been shown to be a strong oxidation poison on noble metal catalysts.158,159 However, at the high temperatures in the reactor, silica can also be removed by volatilization in the form of silicic acids.131 During addition of Si-doped ethanol, silica is continually formed and removed from the Rh catalyst. Since the rate of formation of silica is greater, autothermal operation shuts down. Initially, during regeneration as most of the Rh sites remain poisoned by silica, the temperature keeps dropping. As silica is volatilized, the rate of temperature drop decreases until the in- flection point where enough sites are exposed to accelerate clean up and oxidation of ethanol. Due to the removal of silica, the temperature increases and attains the same value as that of pure ethanol before Si addition, indicating catalyst regeneration.

4.3.2 Sulfur

Sulfur was added to ethanol in the form of dimethyl sulfoxide, [(CH3)2SO]. Autother- mal operation lasted around 115 min. Extensive coke formation was visible on the catalyst as well as on the reactor walls which turned black. The physical structure of the α-alumina support was destroyed, especially the front face which was converted to powder. The back face temperature and pressure drop across the catalyst fluctu- ated, which may be due to channeling from breakdown of the support. Temperature variations are shown in Fig. 4.1B. Errors in carbon, hydrogen and oxygen balances 4.3 Results and Discussion 54 also increased during the deactivation period, which may be attributed to the large amount of coke formed. During regeneration, the reactor lit off with methane and was able to run at a steady temperature and pressure drop at a C/O of 1.5 with methane as a feed, but shut down whenever ethanol was introduced. Sulfur has been shown to severely poison reforming,114 resulting in coke formation. Large amount of coke formation has been shown to destroy the physical structure of catalysts77 and reduces the heat transfer coefficient.160,161 Coke formed may be diffus- ing into the struts of the monolithic foams and causing it to crumble to powder. This along with poisoning of Rh sites prevents sustainable autothermal operation, decreasing the temperature and ultimately causing the reactor to shut down. Regeneration was possible with methane but not with ethanol. The reduced heat transfer to the front face prevents ethanol volatilization, which results in quenching the reactor and shutting down autothermal operation. Methane being a gas, partial oxidation can be carried over the remaining catalyst sites thereby sustaining autothermal operation.

4.3.3 Phosphorus

Phosphorus was added to ethanol in the form of phosphoric acid (H3PO4). Au- tothermal operation was possible for the full 6 h during P addition and the reactor maintained steady operation as shown in the temperature-time proﬁle in Fig. 4.2A. The temperature during steady-state operation with P was higher than that with pure ethanol by about 50 ◦C. Ethanol conversion increased from 59 % to 72 % and was almost constant during the doping period. During the doping period, the CO2 selectivity decreased from 17 % to approximately 7 % (Fig. 4.2B). Ethylene selectivity increased from approximately 4 % to 15 %, which most likely corresponds to increased dehydration chemistry resulting from the introduction of acid in the feed. Hydrogen selectivity decreased from 29 % to 24 %. During regeneration, the temperature, ethanol conversion and product selectivities were similar to the values during doping indicating the catalytic eﬀect of P being present even upon regeneration (Fig. 4.2A, 4.2B). Phosphorus has been shown to be a poison for reforming reactions.162,163 During P addition, the temperature was greater than that with pure ethanol by about 50 ◦C suggesting endothermic reforming reactions are poisoned by P in the system. The higher temperature also increases ethanol conversion due to increased reaction rates. 4.3 Results and Discussion 55

A B 1 0 0 9 0 0 4 0 N o 0 . 0 5 % P R e g e n e r a t i o n N o 0 . 0 5 % P R e g e n e r a t i o n P 9 0 P T 8 0 0 H 8 0 7 0 0 3 0 2

X )

6 0 0 ) C

7 0 o ( %

( )

5 0 0 e r y % t ( i 6 0 u 2 0 C O

t 2 v H i

4 0 0 a t O r t c E 5 0 e e l p X 3 0 0

X e E t O H T e m p e r a t u r e m S 4 0 N o P N o P 2 0 0 e 1 0 0 . 0 5 % P 0 . 0 5 % P T 3 0 R e g e n e r a t i o n R e g e n e r a t i o n 1 0 0 C H 2 0 0 0 2 4 6 0 1 6 0 2 6 0 3 6 0 4 6 0 5 6 0 6 6 0 7 6 0 6 0 1 6 0 2 6 0 3 6 0 4 6 0 5 6 0 6 6 0 7 6 0 T i m e ( m i n ) T i m e ( m i n )

Figure 4.2: Ethanol conversion, temperature proﬁles (A) and product selectivities (B) for doping with P. Horizontal lines without symbols indicate average values of temperature, conversion and product selectivities with pure ethanol. During doping/regeneration, hydrogen selectivites are represented as squares (), carbon dioxide as diamonds () and ethylene as circles (•).

The steady temperature during doping indicates a constant interaction of P with Rh. The increase in ethylene selectivity is most likely due to the acidic nature of the feed, as the acidic protons of phosphoric acid can catalyze the dehydration of ethanol to give ethylene. The decrease in CO2 and H2 selectivities suggests poisoning of the water gas shift reaction by P. Also, P increases the work function of the Rh catalysts, thereby inhibiting CO dissociation.122 Poisoned reforming decreases the rate of production of CO; but this change is likely oﬀset by decreased water gas shift and the increased Rh work function, which reduce the rate of CO consumption. Water selectivity is unchanged because additional water is produced by the ethanol dehydration reaction and decreased water gas shift reduces rate of water consumption. The product distributions and temperatures during regeneration were similar to that during doping, indicating some P was still present on the catalyst and interacting with Rh.

4.3.4 Potassium

Potassium was added to ethanol in the form of potassium acetate (CH3COOK). Au- tothermal operation was possible for 6 h with K-doped ethanol before the operation was shut down manually (Fig. 4.3A). During doping, the back face temperature of the 4.3 Results and Discussion 56

A B 1 0 0 9 0 0 5 0 N o 0 . 0 5 % K R e g e n e r a t i o n N o 0 . 0 5 % K R e g e n e r a t i o n K 8 0 0 K 9 0 T 4 5 C H C H O 3 8 0 7 0 0 4 0 )

6 0 0 ) C O C

7 0 o ( X %

( )

3 5 5 0 0 e r y % t ( i 6 0 u

t v H i H 4 0 0 a

t 2

O 3 0 r t c E 5 0 e e l p X 3 0 0 e

X T e m p e r a t u r e m 2 5 E t O H S 4 0 2 0 0 e N o K N o K T 3 0 0 . 0 5 % K 0 . 0 5 % K 2 0 R e g e n e r a t i o n R e g e n e r a t i o n 1 0 0 2 0 0 1 5 6 0 1 6 0 2 6 0 3 6 0 4 6 0 5 6 0 6 6 0 7 6 0 6 0 1 6 0 2 6 0 3 6 0 4 6 0 5 6 0 6 6 0 7 6 0 T i m e ( m i n ) T i m e ( m i n )

Figure 4.3: Ethanol conversion, temperature proﬁles (A) and product selectivities (B) for doping with K. Horizontal lines without symbols indicate average values of temperature, conversion and product selectivities with pure ethanol. During doping/regeneration, hydrogen selectivites are represented as squares (), carbon monoxide as diamonds () and acetaldehyde as circles (•).

catalyst increased by about 70 ◦C over that with pure ethanol. Ethanol conversion also increased from 63 % to 71 % at the end of 6 hours. During doping, an increase in acetaldehyde selectivity by 20 % upon addition of K was observed (42 % with K vs. 22 % without K) (Fig. 4.3B). Also, the hydrogen and carbon dioxide (not shown)

selectivities increased (H2: 29 % vs. 22 %, CO2: 25 % vs. 20 %), while the carbon

monoxide and water (not shown) selectivities decreased (CO: 21 % vs. 37 %, H2O: 31 % vs. 41 %) compared to pure ethanol. Upon regeneration with pure ethanol, the temperature ultimately steadied at around 755 ◦C, about 20 ◦C higher than the initial steady state temperature with pure ethanol. The catalyst demonstrated higher selectivities to CO2 (31 % during regeneration vs. 20 % before doping) compared to pure ethanol before doping (Fig. 4.3B). The CO selectivity was lower than that with pure ethanol before doping (28 % vs. 37 %). Ethanol conversion, acetaldehyde and hydrogen selectivities, at the end of regeneration were similar to their original values. Alkali metals can decrease the work function of Rh which promotes CO dissociation.122 Also, alkali metals have been shown to poison reforming activity.133 Due to these reasons, the CO selectivity decreases upon doping with potassium. The increased temperatures due to decreased endothermic reforming lead to higher ethanol 4.3 Results and Discussion 57

conversions. The H2 selectivity does not decrease like CO because of increased dehydrogenation chemistry; alkali’s have been shown to promote ethanol dehydrogenation, increasing acetaldehyde selectivity.164 Previous research has shown that alkali metals promote water gas shift on noble metal catalysts which can contribute to increas- 165 ing H2 and CO2 selectivities. During K addition, there exists a steady state K concentration on the surface as indicated by the constant temperature and product distribution. During regeneration, the ethanol conversion, most product selectivities and reactor temperature were similar to that before doping indicating volatilization of K from the catalyst surface.

4.3.5 Sodium

Sodium was added to ethanol in the form of sodium acetate (CH3COONa). The reactor did not attain steady state as indicated by the temperature-time plot shown in Fig. 4.4A. The temperature dropped below 600 ◦C in approximately 110 minutes. Ethanol conversion initially increased with the higher catalyst temperature and then decreased with time. Towards the end of the doping period, the reforming activity

decreased as indicated by lower selectivities to CO and H2 (CO: 10 % vs. 42 %, H2: 13 % vs. 23 %) compared to pure ethanol (Fig. 4.4B). Carbon dioxide (42 % vs. 19 %), acetaldehyde (not shown) (36 % vs. 23 %) and water (not shown) (55 % vs. 45 %) showed increased selectivities. Also, a white deposit was observed on the front face of the catalyst, which was shown by XRD to be a mixture of sodium oxide and sodium peroxide. Upon regeneration, the catalyst temperature gradually increased and ultimately attained a steady value about 60 ◦C hotter than that with pure ethanol as shown in Fig. 4.4A. Ethanol conversions also followed a similar trend, reaching a ﬁnal value of

68 % compared to 60 % before doping. CO2 (24 % vs. 19 %) and acetaldehyde (not shown) selectivities (27 % vs. 23 %) were higher than that with pure ethanol while

CO (29 % vs. 42 %) and H2O (not shown) (39 % vs. 45 %) selectivities were lower (Fig. 4.4B). The white deposit was still present on the front face of the catalyst, which was shown by XRD analysis to be sodium carbonate with no sodium peroxide. This diﬀerence in reactor performance among the alkali metals is a combination of both the diﬀerence in electronic interactions with Rh and their volatilities. Sodium forms more non-volatile compounds which accumulate near the front face of the cat- 4.3 Results and Discussion 58

A 1 0 0 9 0 0 5 0 B N o N a 0 . 0 5 % N a R e g e n e r a t i o n N o N a 0 . 0 5 % N a R e g e n e r a t i o n T 9 0 8 0 0 4 5 C O

8 0 7 0 0 4 0 )

6 0 0 ) 3 5 C

7 0 o ( %

( )

X 5 0 0 e 3 0 r y % t ( i 6 0 u

v H H i 2 4 0 0 a 2 5 t O r t c E 5 0 e e l p X 3 0 0 2 0 e m

X S C O

T e m p e r a t u r e e 4 0 E t O H 2 0 0 1 5 2 N o N a N o N a T 3 0 0 . 0 5 % N a 0 . 0 5 % N a 1 0 0 1 0 R e g e n e r a t i o n R e g e n e r a t i o n 2 0 0 5 6 0 1 2 0 1 8 0 2 4 0 3 0 0 3 6 0 4 2 0 6 0 1 2 0 1 8 0 2 4 0 3 0 0 3 6 0 4 2 0 T i m e ( m i n ) T i m e ( m i n )

Figure 4.4: Ethanol conversion, temperature proﬁles (A) and product selectivities (B) for doping with Na. Horizontal lines without symbols indicate average values of temperature, conversion and product selectivities with pure ethanol. During doping/regeneration, hydrogen selectivites are represented as squares (), carbon monoxide as diamonds () and carbon dioxide as circles (•).

alyst forming a white deposit; no such deposit was observed with K. Initially, during doping of Na, the eﬀect on process chemistry was similar to K as acetaldehyde selectivity increased and CO selectivity decreased, likely due to base catalysis164 and reduced work function respectively.122 Sodium acts as a poison for

reforming as indicated by the decreasing CO and H2 selectivities. Upon regeneration, CO selectivity was greater than that during doping but lower than the original value with pure ethanol. Hydrogen selectivity gradually increased towards its original value, though the increased acetaldehyde selectivity indicates that the amount of hydrogen formed by reforming chemistry is lower. Additional hydrogen could also be formed 165 by water gas shift which is promoted by Na, as evidenced by increased CO2 and

decreased CO and H2O selectivities. The higher temperatures are most likely due to decreased endothermic reforming caused by Na deposited on Rh sites. After doping, sodium oxide and sodium peroxide were present on the catalyst, whereas sodium carbonate was present after regeneration. This indicates that the oxides combine with carbon dioxide to give sodium carbonate, possibly due to the higher temperatures during regeneration. 4.3 Results and Discussion 59

A B 1 0 0 9 0 0 4 5 N o 0 . 0 5 % C a N o 0 . 0 5 % C a R e g e n e r a t i o n C a 8 0 0 C a 9 0 T T 4 0 8 0 7 0 0 X 3 5 )

6 0 0 ) X C 7 0 R e g e n e r a t i o n o C O ( %

( )

3 0 5 0 0 e r y % t ( i 6 0 u

v H i

4 0 0 a t

O 2 5 r t c E 5 0 e e l p X 3 0 0 e C O X T e m p e r a t u r e m 2 0 2 S

4 0 E t O H e N o C a 2 0 0 C H C H O N o C a T 3 0 . 0 5 % C a 0 . 0 5 % C a 1 5 3 0 R e g e n e r a t i o n R e g e n e r a t i o n ( m e t h a n e ) 1 0 0 R e g e n e r a t i o n ( e t h a n o l ) 2 0 0 1 0 6 0 1 5 0 2 4 0 3 3 0 4 2 0 5 1 0 6 0 0 6 0 1 5 0 2 4 0 3 3 0 4 2 0 5 1 0 6 0 0 T i m e ( m i n ) T i m e ( m i n )

Figure 4.5: Ethanol conversion, temperature proﬁles (A) and product selectivities (B) for doping with Ca. Horizontal lines without symbols indicate average values of temperature, conversion and product selectivities with pure ethanol. During doping/regeneration, carbon monoxide selectivites are represented as squares (), carbon dioxide as diamonds () and acetaldehyde as circles (•).

4.3.6 Calcium

Calcium was added to ethanol in the form of calcium nitrate tetrahydrate, [Ca(NO3)2.

4 H2O]. Unlike Na, K, and Mg, calcium acetate or formate could not be used because of low solubility in ethanol. During doping with Ca, the reactor shut down after about 160 minutes on stream as shown in Fig. 4.5A. Ethanol conversion gradually increased with temperature from 54 % to 79 %. Selectivities to CO and H2 were unchanged within the limits of experimental error. Carbon dioxide selectivity decreased from 26 % to 17 % (Fig. 4.5B). A white deposit was observed on the front face of the catalyst which was shown by XRD analysis to be calcium carbonate. During regeneration, the reactor lit off with methane, but shut down initially when pure ethanol was introduced. However, after regenerating with methane for 1 h and then switching to pure ethanol, autothermal operation was achieved. The final temperature (732 ◦C) was similar to the temperatures observed with pure ethanol (745 ◦C) before doping (Fig. 4.5A). During regeneration, higher ethanol conversions were obtained (82 % vs. 54 %). Selectivities to acetaldehyde (15 % vs. 25 %), carbon dioxide (14 % vs. 26 %) and water (37 % vs. 46 %, not shown) decreased. Selectivities to carbon monoxide (37 % vs. 30 %) and methane (19 % vs. 11 %, not shown) were higher (Fig. 4.5B). Calcium carbonate persisted as a white deposit on the front face 4.3 Results and Discussion 60 of the catalyst even after regeneration. The gradual increase in temperature during the doping period may be explained by the accumulation of Ca near the front face, pushing the oxidation zone downstream leading to increased back face temperatures. Calcium addition has been shown to promote ethanol reforming,166,167 which explains the increasing ethanol conversion with time during doping and higher ethanol conversion during regeneration. During regeneration, initially the accumulated calcium carbonate near the front face affects heat transfer to ethanol at the front face of the catalyst. Therefore, autothermal reforming is possible only with methane. After operation with methane for an hour and then switching to ethanol, autothermal operation was possible, probably due to some cleanup of the surface during methane CPO. During regeneration, the CO and

CH4 selectivities were higher and the acetaldehyde selectivity was lower, similar to that obtained by Biswas and Kunzru.166 A strong correlation (Pearson Correlation

Coeﬃcient = 0.86 over 11 data points) was observed between the CO, CH4 and acetaldehyde molar ﬂow rates indicating increased decarbonylation of acetaldehyde.

4.3.7 Magnesium

Magnesium was added to ethanol in the form of magnesium acetate tetrahydrate,

[(CH3COO)2Mg.4 H2O]. The reactor maintained autothermal operation for 6 h before it was shutdown manually (Fig. 4.6). The temperature during doping was initially similar to that of pure ethanol. After about 200 min on stream during doping, a decrease in temperature was observed. Initially, the ethanol conversion was higher than with pure ethanol (65 % vs. 53 %) and it decreased with temperature to about 60 % towards the end of the doping period. XRD analysis showed that magnesium oxide was present as a white deposit on the front face of the catalyst, which possibly blocks the Rh sites and decreases catalytic activity resulting in a drop in temperature and conversion with time. During regeneration, the ethanol conversion decreased gradually to its original value. Final steady state temperature during regeneration was slightly lower than that with pure ethanol by about 15-20 ◦C (Fig. 4.6). The changes in product selectivites during doping and regeneration were generally within 6 % (absolute). It is still unclear why Mg despite being less volatile than Na can maintain autothermal operation for the full 6h. We surmise that Mg has promoting eﬀects similar to Ca as indicated 4.4 Comparison of Inorganics 61

Figure 4.6: Ethanol conversion and temperature proﬁles during doping and regeneration for Mg. by the initial higher conversions, which promote ethanol reforming and help sustain autothermal operation.

4.4 Comparison of Inorganics

The results discussed in previous sections are summarized in Table 4.2. The deactivation of Rh due to the inorganics is a combination of electronic interactions and physical blocking of active sites. The physical blocking of Rh active sites by the inorganics depends on the volatility of the inorganics. Calcium, magnesium and silicon represent non-volatile inorganics and their oxides are often used as catalyst supports. Potassium is more volatile than sodium whereas phosphorus and sulfur represent the most volatile inorganics. Silicon showed the strongest poisoning effect on Rh, however complete regeneration was possible due to volatilization in the form of silicic acids. Sulfur poisoning was due to the strong chemisorption on Rh sites and destruction of the support. Phos- phorus due to its acidic nature increased ethylene selectivity and ethanol conversion which persisted even upon regeneration. Alkali metals, Na and K, both increased acetaldehyde selectivity due to their basic nature and reduced CO selectivity due to reduced Rh work function. Potassium being more volatile than sodium maintained 4.4 Comparison of Inorganics 62 Inorganic Run time (min) Effects (doping) Effects (regeneration) Si 5 fastest deactivation complete regeneration observed S 115 disintegration of alu- regeneration only with mina support methane P 360 increased tempera- similar to that during ture, ethylene selec- doping tivity, and ethanol conversion K 360 increased acetalde- similar to that with hyde selectivity pure ethanol during doping Na 110 decreased reforming higher temperatures during doping and ethanol conversion Ca 160 higher ethanol conver- initial regeneration sion only with methane and then with ethanol Mg 360 no significant activity no significant activity change change Table 4.2: Summary of Effects Observed from Inorganics

steady state autothermal operation during doping. Calcium showed a strong promoting effect on ethanol conversion unlike magnesium which showed no significant changes in conversions/selectivities. With calcium and sodium, the accumulation on the front face affects heat transfer from the oxidation zone to vaporize ethanol which can contribute to shutting down autothermal operation. In the present system, after regeneration, Mg, Ca, and Na, were present in the form

of MgO, CaCO3, Na2CO3 respectively as shown by XRD analysis. These compounds are low vapor pressure compounds, but some of them must pass through the catalyst as indicated by the change in color of the exhaust ﬂame. This might be occurring through the formation of more volatile compounds by combination with species in

the reactor such as H2, CO, CO2 or hydrocarbons. Further investigation of this phenomenon is necessary to determine the nature of these species and how they leave the catalyst. In contrast, PH3, Phosphorus oxides, H2S and sulfur oxides do not poison the catalyst by blocking as they are gases. Their chemisorbed or complexed form instead acts as an electronic poison to the catalyst. No slag formation was observed with any of the inorganics studied. 4.5 Conclusions 63

4.5 Conclusions

The effect of biomass inorganics Na, K, Ca, Mg, S, P, Si on Rh catalysts was studied by ethanol autothermal reforming. Potassium, phosphorus and magnesium maintained autothermal operation over the 6 h test period whereas sodium, calcium, silicon and sulfur extinguished reaction. Silicon shut down the autothermal operation the fastest. Auothermal regeneration was possible for each of them by passing clean fuel, either methane or ethanol. Potassium and phosphorus showed the presence of base and acid chemistry by increasing acetaldehyde and ethylene selectivities respectively. Silicon forms volatile compounds at the high temperatures involved and hence complete regeneration of catalytic activity was observed. Extensive coke formation was observed upon doping with S which led to disintegration of the α-Al2O3 support. A significant fraction of Na, Ca and Mg accumulated near the front face of the catalyst due to their non-volatile nature. However, a certain fraction of these non-volatile inorganics pass through the catalyst indicating that at some lower concentration of inorganics, steady-state operation may be possible. Future efforts will focus on investigating thermal and chemical regeneration techniques for various inorganics.

4.6 Acknowledgements

We acknowledge funding from the Minnesota Corn Growers Association. Parts of this work were carried out in the University of Minnesota I.T. Characterization Facility, which receives partial support from NSF through the NNIN program. We thank Dr. Brian Michael and Dr. Joshua Colby for their helpful consultation with experiments. CHAPTER FIVE

EFFECTS OF POTASSIUM AND PHOSPHORUS ON RHODIUM CATALYSTS FOR CATALYTIC PARTIAL OXIDATION

In chapters 3 and 4, the effects of common biomass inorganics on rhodium catalysts were compared by introducing them directly on the catalyst(with steam reforming) and in the feed (with catalytic partial oxidation) respectively. The surveys were performed at a particular concentration of inorganics and feed composition. Potassium and phosphorus represent inorganics commonly found in biomass and can significantly affect catalytic gasification of biomass. The alkaline and acidic nature of potassium and phosphorus can introduce different reaction chemistries as shown in Chapter 4. In this chapter, the effects of potassium and phosphorus on Rh catalysts for catalytic gasification have been studied at different fuel to oxygen ratios and temperatures. Catalytic partial oxidation of methane and ethanol was carried out over potassium, phosphorus and monobasic potassium phosphate-doped catalysts at loadings of 0.1, 1, 10, and 100 % of Rh atoms over a range of fuel to oxygen ratios. In addition, time-on-stream studies were also carried out with methane by predosing inorganics on the catalyst (10 % loading) and ethanol by introducing inorganics in the feed (0.05 mol %) to understand their effects on stability of autothermal operation. With methane, potassium-doped catalysts showed a strong decrease in CO selectivity at lower temperatures and the effect decreased with increasing temperatures. This regeneration of product distributions at higher temperatures was attributed to the volatilization of potassium from the catalyst surface. Phosphorus-doped catalysts

64 5.1 Introduction 65

greatly reduced hydrogen selectivities with methane. With ethanol, synthesis gas selectivities did not change significantly with potassium-doped catalysts whereas they decreased with phosphorus-doped catalysts. Monobasic potassium phosphate showed synergistic effects of potassium and phosphorus. Unlike potassium, phosphorus did not appear to show any thermal regeneration and exhibited a stronger poisoning effect on rhodium for catalytic partial oxidation than potassium.

5.1 Introduction

Gasification of biomass to produce syngas represents an attractive route to upgrade biomass to carbon-based fuels and chemicals. Cellulose, hemicellulose, and lignin represent the prinicipal components of lignocellulosic biomass. In addition, different biomass sources contain different concentrations of inorganics, collectively referred to as ash, such as Si, Ca, Na, K, Ca, Mg, P, S, Cl, Fe, Cu, and Mn.136,168,169 These inorganics in biomass can change catalyst activity, reaction rates, and product distributions. Schmidt and coworkers have shown that lignocellulosic biomass containing small concentrations of inorganics and other biomass model compounds can be gasified autothermally over Rh-based catalysts to syngas with high yields in millisecond contact times at temperatures between 600-1000 ◦C.1–3 No tars or chars are formed in this process. For catalytic gasification of lignocellulosic biomass to be feasible, the effect of inorganics on catalyst activity should be minimal and the catalyst should be easily regeneratable. At high temperatures in catalytic partial oxidation (CPO), certain inorganics may be volatile or may combine with other species in the reactor

such as CO, CO2,H2,H2O or other inorganics to form volatile compounds and leave the catalyst.170,171 In such a case, biomass feedstocks containing low concentrations of certain inorganics may be processed to syngas by catalytic gasification without significant catalyst deactivation. Potassium and phosphorus represent two of the major inorganic constituents of biomass. Potassium content in ash can vary from ∼ 4 % to 50 % and can compose up to 3 % of the total dry weight of biomass.172,173 Potassium content is high in biomass feedstocks like straws and grasses. Phosphorus concentration in biomass ash, ∼ 0.1 % to 8 % is lower than potassium concentration. Potassium and phosphorus can also enter biomass sources like plants through fertilizers. Due to the significant concentrations of phosphorus and potassium in biomass, understanding their effects 5.1 Introduction 66 on the catalyst is essential for CPO of lignocellulosic biomass to be feasible. In addition, potassium and phosphorus also have widely different electronic properties; potassium is electropositive while phosphorus is electronegative. Most other biomass inorganics have electronegativities in between the two; thus studying the effects of potassium and phosphorus encompasses almost the entire range of interactions of common biomass inorganics with rhodium. In this chapter, CPO of methane and ethanol is used to understand the effects of potassium and phosphorus on rhodium catalysts on reaction chemistries. Methane represents the simplest carbon species and has negligible homogeneous chemistry at 600-1000 ◦C which are typical temperatures for CPO.107 CPO of methane also has an easily quantifiable product spectrum, consisting of CO, H2, CO2, and H2O. Hence, any change in chemistry or product distribution can be attributed to the change in activity introduced by the inorganics on rhodium catalysts. Ethanol represents a simple oxygenated molecule, and contains all of the bonds present in biomass: C- C, C-H, C-O, and O-H. In addition to producing syngas and combustion products, CPO of ethanol can also form other products such as methane, and non-equilibrium products such as ethylene, ethane, and acetaldehyde. Inorganics in biomass can also introduce acid or base chemistry which are not typically observed in noble metal catalysis. Additional reactions can promote or inhibit production of non-equilibrium products. This can be observed with ethanol, but not methane. Chapters 3 and 4 present surveys of the effects of common biomass inorganics on CPO such as Na, K, Ca, Mg, S, P, and Si on rhodium catalysts.108,156 Here, the effects of potassium and phosphorus on rhodium catalysts are studied in detail by depositing them at different loadings (inorganic concentration = 0.1, 1, 10 or 100 % of Rh atoms) using CPO of methane and ethanol. Also, their effect on reactor performance and product distributions was measured at different fuel to oxygen ratios, and consequently different temperatures. In actual biomass, inorganics also exist in the combined state with one or more inorganics. To understand the synergistic effect of potassium and phosphorus, monobasic potassium phosphate was added at similar loadings. Time-on-stream studies were also performed with methane and ethanol to understand the transients involved during operation with inorganics long-term stability of reactor operation. 5.2 Experimental 67

5.2 Experimental

Experiments were carried out in a quartz tube (19 mm I.D.). Gas ﬂow rates to the reactor were controlled by mass ﬂow controllers accurate to within ± 2 %. The catalyst consisted of a 45 ppi (pores per linear inch) α-alumina cylindrical monolith (17 mm I.D., 10 mm long) which was coated with 5 wt % Rh. Catalysts were prepared by the incipient wetness technique using rhodium nitrate as the rhodium precursor.100 Catalysts were then dried in a vacuum oven and calcined at 800 ◦C for 6 h. An

uncoated 45 ppi α-Al2O3 monolith was used as a back heat shield. The entire assembly was wrapped in aluminosilicate cloth to prevent bypassing of gases and inserted into a quartz reactor tube. The reactor was wrapped with one inch thick ceramic fiber insulation to minimize heat losses. For methane CPO, another uncoated 45 ppi monolith was used as a front heat shield. For ethanol CPO, the experimental setup was similar to that reported by Rennard et al.73 Ethanol was fed in through an HPLC pump. The reactor was lit off with hydrogen and then switched to either methane or ethanol. The carbon molar flow rate with methane and ethanol was approximately 0.04 mol/min which is equivalent to a flow rate of 1 SLPM (standard liters per minute) for methane and 1.2 ml/min for ethanol. A nebulizer was used to create a fine mist of ethanol which impinged upon the front face of the catalyst. A pyrex annulus was placed just above the tip of the nebulizer to prevent convective recirculation in the ethanol experiments. Argon, in place of nitrogen, was fed with oxygen at air stoichiometry (3.76). Argon acted as a diluent and as an internal standard for gas chromatograph (GC) analysis. The C/O ratio is defined as the ratio of carbon atoms in fuel to the oxygen atoms in air; oxygen in ethanol was not counted in the denominator. Products were analyzed and quantified using a HP 5890 series II gas chromatograph (GC) with a 60 m PLOT-Q column. Response factors of each compound were determined by calibrating with known concentrations of species with respect to argon as an internal standard. Hydrocarbon species were quantified on the flame ionization

detector while H2, CO, Ar and CO2 were quantified with the thermal conductivity detector. Hydrogen and oxygen balances were closed on water and average of the two results was used to determine water flow rate. Carbon, hydrogen and oxygen balances typically closed within ± 10 % and 95 % confidence intervals were typically less than 7 %. The effluent was sampled after approximately an inch of the bottom 5.2 Experimental 68 edge of the catalyst. Three experiments were performed to study the interactions of inorganics with the rhodium-based catalyst: doping the catalyst at different levels for CPO of methane and ethanol, a transient study with methane by doping the inorganics on the catalyst and a transient study with ethanol by introducing inorganics in the feed. The changes with respect to baseline values are expressed in terms of absolute values. Product selectivities were defined on a carbon or hydrogen (for H2 and H2O) basis as number of atoms of carbon or hydrogen in a particular species to total number of carbon or hydrogen atoms in the products, unconverted fuel was not counted in the product stream. For the first set of experiments, each catalyst was aged with either methane (C/O = 0.75) or ethanol (C/O = 0.8) for one hour. For experiments with methane feed, baseline performance with an undoped rhodium catalyst was measured at C/O ratios from 0.75 to 2, in steps of 0.25. For ethanol at C/O = 0.75, upstream flames were observed. Hence, baseline performance was measured at C/O = 0.8 and C/O = 1 to C/O = 1.75 (steps of 0.25). At C/O = 2 for ethanol, oxygen breakthrough was observed as was small amount of coke deposition on the catalyst. Hence, this data point was not examined. After aging the catalyst at the lowest C/O ratio for each fuel for an hour, baseline performance was measured from the lowest C/O (highest temperatures) to highest C/O (lowest temperatures). Inorganics were then added to the catalyst through precursors (potassium acetate for potassium, phosphoric acid for phosphorus and potassium phosphate monobasic to test for their combined effect) by the incipient wetness technique and then dried. Loadings are expressed in terms of ratio of number of atoms of potassium/phosphorus (or molecules of monobasic potassium phosphate) to number of atoms of rhodium. Performance testing with the inorganic-doped catalyst was then carried out from the highest C/O ratio to the lowest C/O ratio, to minimize the effect of inorganics volatilizing at higher C/O ratios. After a particular doping experiment, if the range of differences of a particular quantity (conversion/selectivity) over a C/O range compared to baseline was greater than 10 % (absolute), then the experiment was repeated (from the highest C/O ratio to the lowest C/O ratio) to test for thermal regeneration by volatilization at higher temperatures. For example, if the difference between inorganic-doped and undoped CO selectivities was 20 % at C/O = 2 and 4 % at C/O = 0.75, then the experiment was again repeated from C/O = 2 5.3 Results 69 to C/O = 0.75 to investigate for thermal regeneration. As a result of volatilization, catalytic activity can be partially or completely restored at lower temperatures. Each data point shown represents the average of three injections on the GC. The above experiments were performed for ∼ 1 h at each C/O ratio and show the effect of doping potassium and phosphorus at different concentrations during methane and ethanol CPO. At every C/O ratio, some fraction of inorganics volatilize which increases with temperature. To rule out transient effects, the three GC injections representing each data point were measured at temperatures differing by ≤ 5 ◦C. To gain a greater understanding of the transients involved during volatilization of potassium and phosphorus, a time-on-stream study was performed with methane CPO. 10 % potassium and 10 % phosphorus were added to the rhodium catalyst (5 wt %) after baseline performance testing at C/O = 0.75 and C/O = 1.5. After doping, the catalyst was lit off with hydrogen and switched to methane at C/O = 0.75 and performance was measured for ∼ 60 min. The C/O was then switched to 1.5 for ∼ 20 min and performance was measured after which it was again switched back to 0.75 and the cycle was repeated for total 6 h. The experiments described previously involve predosing the catalyst with inorganics. To simulate more closely how inorganics in actual biomass reach the catalyst, third set of experiments were performed by dissolving the inorganic in ethanol and performing CPO of inorganic-doped ethanol. Two sliced 5 wt % Rh foams (semicir- cular cross-section from top) were used as the catalyst and placed in the reactor. The concentration of inorganics in the feed was 0.05 % mol inorganic/mol ethanol. The same ethanol flow rate of 1.2 ml/min was used at a C/O of 1, which corresponded to a baseline temperature of ∼ 800 ◦C , typical of CPO.

5.3 Results

5.3.1 Doping at Diﬀerent Concentrations with Methane CPO

Negligible changes were observed in product distributions and temperatures with methane CPO at 0.1 % loading of potassium, phosphorus and monobasic potassium phosphate (≤ 5% change). 5.3 Results 70

1 0 0 1 0 0 A R h B R h R h + 1 % K R h + 1 0 % K R h + 1 0 % K r e g 8 0 8 0 ) ) % % ( (

6 0 6 0 y y t t i i v v i i t t c c e e l 4 0 l 4 0 e e s s

O O C C 2 0 2 0

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 C / O C / O

Figure 5.1: CO selectivities at 1 % (A) and 10 % (B) loading of potassium. White bars represent performance of the undoped rhodium catalyst. Bars with slanted lines indicate performance of the potassium-doped catalyst from high C/O ratios to low C/O ratios. Grey bars represent data when the reactor was operated again with the doped catalyst from high C/O ratios to low C/O ratios to investigate for thermal regeneration.

Eﬀect of Potassium

Autothermal operation was possible with methane at potassium loadings from 0.1 % to 10 %. At a potassium loading of 100 % of Rh atoms, lightoff with the potassium- doped catalyst was possible with hydrogen, however, the reactor shut down whenever methane was introduced. Potassium addition had a negligible effect on methane conversion and hydrogen selectivities (≤ 5% change). At loadings of 1 % and 10 % at C/O ratio of 2, strong decrease in CO selectivites of ∼ 7 % and 22 % respectively was observed, however the selectivities approached their baseline values (undoped catalyst) with decreasing C/O ratios and was similar to the original value at C/O = 0.75 as shown in Figure 5.1 (A and B). The backface temperatures increased slightly at higher C/O ratios at 1 and 10 % loadings (∆T ≤ 50 ◦C) and approached their baseline values at lower C/O ratios, the temperature difference being ≤ 25 ◦C. This regeneration in temperatures and CO selectivities indicates the reduced effect of potassium on CPO chemistry at low C/O ratios due to its volatility at high temperatures in a CPO reactor. 5.3 Results 71

Eﬀect of Phosphorus

Phosphorus maintained autothermal operation with methane at all loadings from 0.1 % to 100 %. The effect on CO selectivity at different levels of phosphorus loading was negligible. A significant decrease in methane conversion and hydrogen selectivities was observed at loadings of 1 to 100 % (Figure 5.2). The decrease in methane conversion compared to baseline values was negligible at high C/O ratios and was more significant at lower C/O ratios. Hydrogen selectivities were closer to baseline values at low C/O ratios compared to high C/O ratios (Figure 5.2 B,D,F). Unlike potassium, no significant changes in product distributions were observed upon regeneration. Phosphorus also increased the backface catalyst temperatures by ∼ 75 ◦C at 1 % loading, 95 ◦C at 10 % loading and up to 200 ◦C at C/O = 0.8 at 100 % loading. No significant changes were observed upon regeneration. The decrease in hydrogen selectivities may be attributed to a combination of reduced water gas shift and steam reforming due to phosphorus poisoning the rhodium catalyst.

Eﬀect of Monobasic Potassium phosphate

Potassium phosphate (monobasic) was used to study the effects of potassium and phosphorus in the presence of each other. With monobasic potassium phosphate, autothermal operation was possible at loadings from 0.1 % to 10 %. The reactor shut down with methane at a loading of 100 %. Negligible effect on methane conversion was observed at loadings of 0.1 and 1% (≤ 5% change, Figure 5.3 A). At a 10 % loading, the methane conversion decreased from baseline by ∼ 7 % at C/O = 0.75 and was similar to baseline values at other C/O ratios (Figure 5.3 B). A strong decrease in CO selectivities was observed at high C/O ratios at 1 and 10 % loadings; however they approached baseline values with decreasing C/O ratios (Figure 5.3 C,D). Upon regeneration, the CO selectivities at high C/O ratios were similar to their undoped values (Figure 5.3 D). Hydrogen selectivities decreased at 10 % loading, however no significant changes were observed upon regeneration (Figure 5.3 E,F). The temperatures were ∼ 30, 75 and 100 ◦C hotter at loadings of 0.1, 1 and 10 % loading respectively. Upon regeneration at higher C/O ratios at 10 % loading, temperatures at higher C/O ratios decreased by ∼ 35 ◦C. Thus, monobasic potassium phosphate showed the cumulative effect of potassium and phosphorus with the potassium component reducing CO selectivities and volatilizing at high temperatures, whereas phosphorus 5.3 Results 72

1 0 0 1 0 0 R h R h A B R h + 1 % P R h + 1 % P

8 0 8 0 ) % (

) n % o i

6 0 ( 6 0

s r y t i e v v i t n c o e c l

4 0 4 0 e e s n

2 a h H t e 2 0 2 0 M

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 C / O C / O

1 0 0 1 0 0 C R h D R h R h + 1 0 % P R h + 1 0 % P R h + 1 0 % P r e g R h + 1 0 % P r e g 8 0 8 0 ) % (

) n % o i

6 0 ( 6 0

s r y t i e v v i

t n c o e c l

4 0 4 0 e e s n

2 a h H t e 2 0 2 0 M

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 C / O C / O

1 0 0 1 0 0 E R h F R h R h + 1 0 0 % P R h + 1 0 0 % P R h + 1 0 0 % P r e g R h + 1 0 0 % P r e g 8 0 8 0 ) % (

) n % o i

6 0 ( 6 0

s r y t i e v v i t n c o e c l

4 0 4 0 e e s n

2 a h H t e 2 0 2 0 M

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 C / O C / O

Figure 5.2: CH4 conversion and H2 selectivities during methane CPO at 1, 10 and 100 % loading of phosphorus. White bars represent performance of the undoped rhodium catalyst. Bars with slanted lines indicate performance of the phosphorus- doped catalyst from high C/O ratios to low C/O ratios. Grey bars represent data when the reactor was operated again with the doped catalyst from high C/O ratios to low C/O ratios to investigate for thermal regeneration. 5.3 Results 73

1 0 0 1 0 0 R h B A R h R h + 1 % K P R h + 1 0 K P R h + 1 0 % K P r e g 8 0 8 0 ) ) % % ( (

n n o o

i 6 0 i 6 0 s s r r e e v v n n o o c

c 4 0

4 0 e e n n a a h t h t e e 2 0 M 2 0 M

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 C / O C / O

1 0 0 1 0 0 C P u r e R h D R h R h + 1 % K P R h + 1 0 K P R h + 1 0 K P r e g 8 0 8 0 ) ) % %

( 6 0 ( 6 0

y y t t i i v v i i t t c c e e l l

e 4 0 e 4 0 S S

O O C C

2 0 2 0

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 C / O C / O

1 0 0 1 0 0 E P u r e R h R h + 1 % K P F R h R h + 1 0 % K P 8 0 8 0 R h + 1 0 % K P r e g ) ) % 6 0 % (

( 6 0

y y t t i i v v i i t t c c e l e

4 0 l 4 0 e e s

s 2

2 H H 2 0 2 0

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 2 . 0 0 2 . 2 5 C / O C / O

Figure 5.3: CH4 conversion, CO and H2 selectivities during methane CPO at 1 and 10 % loading of KH2PO4. White bars represent performance of the undoped rhodium catalyst. Bars with slanted lines indicate performance of the monobasic potassium phosphate-doped catalyst from high C/O ratios to low C/O ratios. Grey bars represent data when the reactor was operated again with the doped catalyst from high C/O ratios to low C/O ratios to investigate for thermal regeneration. 5.3 Results 74 showed poisoning activity by decreasing hydrogen selectivities which persisted even upon regeneration.

5.3.2 Doping at Diﬀerent Concentrations with Ethanol CPO

Negligible changes in product distributions and temperatures were observed at 0.1 % loading of potassium, phosphorus and monobasic potassium phosphate (≤ 5% change).

Eﬀect of Potassium

In case of ethanol CPO with potassium-doped catalysts, autothermal operation was possible at all loadings from 0.1 % to 100 %. Negligible changes in ethanol conversion and temperatures were observed at all loadings (≤ 5% ), unlike methane. CO selectivities were essentially unchanged from 0.1 to 10 % loadings. At 100 % loading, the CO selectivity decreased by ∼ 7 % at all C/O ratios. Hydrogen selectivities increased by ∼ 7 % at higher C/O ratios at loadings of 10 and 100 %. The ability of the catalyst at 100 % potassium loading to maintain autothermal operation with ethanol is likely due to the higher reactivity of ethanol compared to methane.

Eﬀect of Phosphorus

Autothermal operation was possible with all loadings of phosphorus from 0.1 % to 100 %. Ethanol conversion showed negligible changes at all loadings. The largest decrease in CO selectivities was observed at C/O = 1 at 10 and 100 % loadings (Figure 5.4 A,B). The decrease in hydrogen selectivities compared with baseline values increased with temperatures (Figure 5.4 C,D). At 100 % loading, it decreased by ∼ 10 % at C/O = 1.75 and ∼ 30 % at C/O = 0.8. Ethylene selectivites increased by ∼ 10 % at 100 % loading at high C/O ratios which decreased slightly at upon regeneration (Figure 5.4 E,F). Changes in temperatures were generally less than 50 ◦C. No signiﬁcant changes were observed upon regeneration at all loadings. Thus, phosphorus with ethanol CPO showed similar eﬀects to methane CPO as hydrogen selectivities strongly decreased.

Eﬀect of Monobasic Potassium phosphate

Similar to potassium and phosphorus, autothermal operation was possible with potassium phosphate (monobasic) at all loadings from 0.1 % to 100 %. Ethanol conversions 5.3 Results 75

1 0 0 1 0 0 A R h B R h R h + 1 0 % P R h + 1 0 0 P R h + 1 0 % P r e g R h + 1 0 0 % P r e g 8 0 8 0 ) ) % % ( (

6 0 6 0 y y t t i i v v i i t t c c e e l 4 0 l 4 0 e e s s

O O C C 2 0 2 0

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 C / O C / O

1 0 0 1 0 0 R h C D R h R h + 1 0 % P R h + 1 0 0 % P R h + 1 0 % P r e g R h + 1 0 0 % P r e g 8 0 8 0 ) ) % %

6 0 ( 6 0 (

y y t t i i v v i i t t c c e l e

4 0 l 4 0 e e s

s 2

2 H H 2 0 2 0

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 C / O C / O

2 0 2 0 R h R h E F R h + 1 0 % P R h + 1 0 0 % P R h + 1 0 % P r e g R h + 1 0 0 % P r e g 1 6 1 6 ) ) % % ( (

y y t t

i 1 2 i 1 2 v v i i t t c c e e l l e e s s 8 8 e e n n e e l l y y h h t t

E 4 E 4

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 C / O C / O

Figure 5.4: CO, H2, and C2H4 selectivities during ethanol CPO at 10 and 100 % loading of phosphorus. White bars represent performance of the undoped rhodium catalyst. Bars with slanted lines indicate performance of the phosphorus-doped catalyst from high C/O ratios to low C/O ratios. Grey bars represent data when the reactor was operated again with the doped catalyst from high C/O ratios to low C/O ratios to investigate for thermal regeneration. 5.3 Results 76

1 0 0 1 0 0 A R h B R h R h + 1 0 % K P R h + 1 0 0 % K P R h + 1 0 % K P r e g R h + 1 0 0 % K P r e g 8 0 8 0 ) ) % % ( (

6 0 6 0 y y t t i i v v i i t t c c e e l 4 0 l 4 0 e e s s

O O C C 2 0 2 0

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 C / O C / O

1 0 0 1 0 0 R h R h C D R h + 1 0 % K P R h + 1 0 0 % K P R h + 1 0 % K P r e g R h + 1 0 0 % K P r e g 8 0 8 0 ) ) % %

( 6 0 ( 6 0

y y t t i i v v i i t t c c e e

l 4 0 l 4 0 e e s s

2 2 H H 2 0 2 0

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 C / O C / O

2 0 2 0 R h R h E F R h + 1 0 % K P R h + 1 0 0 % K P R h + 1 0 % K P r e g R h + 1 0 0 % K P r e g 1 6 1 6 ) ) % % ( (

y y t t

i 1 2 i 1 2 v v i i t t c c e e l l

e e s s 8 8 e e n n e e l l y y h h t t

E 4 E 4

0 0 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 0 . 7 5 1 . 0 0 1 . 2 5 1 . 5 0 1 . 7 5 C / O C / O

Figure 5.5: CO, H2, and C2H4 selectivities during ethanol CPO at 10 and 100 % loading of KH2PO4. White bars represent performance of the undoped rhodium catalyst. Bars with slanted lines indicate performance of the monobasic potassium phosphate- doped catalyst from high C/O ratios to low C/O ratios. Grey bars represent data when the reactor was operated again with the doped catalyst from high C/O ratios to low C/O ratios to investigate for thermal regeneration. 5.3 Results 77 increased by about 10 % from C/O = 1.75 to 1.25 and decreased to their original values upon regeneration. Signiﬁcant regeneration was observed in CO selectivities at higher C/O ratios (Figure 5.4 A,B). The decrease in hydrogen selectivities compared to baseline values was more at lower C/O ratios (Figure 5.4 C,D). At 100 % loading, ethylene selectivities increased by ∼ 11% at high C/O ratios (1.75 - 1.25) and 4 % at C/O = 0.8 (Figure 5.4 E,F). The ethylene selectivites at higher C/O ratios decreased by about 6 % upon regeneration. At loadings other than 100 %, temperatures were within 50 ◦C of their baseline values. At 100 % loading, they increased by ∼ 120 ◦C, but then decreased at higher C/O ratios to within 50 ◦C of their original values. Similar to methane, cumulative eﬀects of potassium and phosphorus were observed with potassium phoshphate. The higher ethanol conversions compared to baseline at high C/O ratios result from higher temperatures thereby increasing reaction rates.

5.3.3 Transient Studies with Methane CPO

To study the transients involved during volatilization of potassium and phosphorus, methane CPO of doped catalysts was performed by switching between C/O ratios of 0.75 (∼ 60 min) and 1.5 (∼ 20 min). The results with potassium and phosphorus are shown below.

Eﬀect of Potassium

The temperatures and methane conversion during the 6 h performance testing period with 10 % potassium loading are shown in Figure 5.6. After lightoﬀ on the doped catalyst with hydrogen, the temperature increased to ∼ 1050 ◦C with methane at C/O = 0.75 compared to baseline value of 900 ◦C. At the end of 1 h, the temperature was ∼ 940 ◦C. Methane conversions also gradually increased as shown during the initial period. After switching to C/O = 1.5, the temperatures and product distributions were close to undoped values. Upon switching back to C/O = 0.75, the temperature continued to drop though at a slower rate than before, with methane conversions similar to baseline values. Temperatures steadied out after ∼ 100 min on stream, the product distributions and temperatures at both C/O = 0.75 and 1.5 thereafter being similar to their original values indicating the absence of catalytic activity of potassium due to potassium volatilization. Further experiments are necessary to determine the concentration of potassium remaining on the catalyst surface. 5.3 Results 78

1 0 0 1 2 0 0 P o t a s s i u m X C / O = 0 . 7 5 9 0 1 1 0 0 )

% 8 0 (

1 0 0 0 )

n T C

o C / O = 0 . 7 5 o

i 7 0 (

r 9 0 0 e r e v 6 0 u t n a o

8 0 0 r c e 5 0 p e

n r e a c t o r s h u t d o w n T C / O = 1 . 5 7 0 0 m a 4 0 e h T t e 6 0 0 M 3 0 X C / O = 1 . 5 2 0 5 0 0 0 8 0 1 6 0 2 4 0 3 2 0 4 0 0 T i m e ( m i n )

Figure 5.6: CH4 conversions () and temperatures (solid lines) during methane catalytic partial oxidation for 6 h at C/O ratios of 0.75 and 1.5 at 10 % loading of potassium. Dashed and dotted horizontal lines indicate undoped values of conversions and temperatures respectively.

Eﬀect of Phosphorus

The temperatures and methane conversions during the 6 h performance testing period with 10 % potassium are shown in Figure (5.7 A). After lightoﬀ, the temperature with the 10 % phosphorus doped catalyst was about 1030 ◦C after which it dropped to a steady value of 985 ◦C in ∼ 50 min. Methane conversion increased with time reaching 79 % after 50 min, compared to the baseline value of 88 %. Hydrogen selectivity was 71 % compared to a original value of 82 % (Figure 5.7 B). At C/O = 1.5, the temperature steadied at ∼ 835 ◦C and the methane conversion was similar to undoped values. Hydrogen selectivity was much lower at 48 % compared to a baseline of 68 %. Upon returning to C/O = 0.75, the temperature was steady at ∼ 980 ◦C, 75 ◦C higher than the baseline. Methane conversion and hydrogen selectivites were about 10 % below the baseline values at C/O = 0.75. At C/O = 1.5, temperatures were also higher by about 75 ◦C while hydrogen selectivities were about 20 % lower than the baseline values. No further change was observed with time. Negligible changes in CO selectivities at both C/O ratios were observed during the 6 h study. The higher temperatures and lower methane conversions and hydrogen selectivities indicate inhibition of endothermic reforming activity by phosphorus which is largely unaﬀected by maintaining higher temperatures. 5.3 Results 79

1 0 0 1 1 0 0 1 0 0 A B X C / O = 0 . 7 5 P h o s p h o r u s 9 0 9 0 S 1 0 0 0 C / O = 0 . 7 5 )

% 8 0 8 0 (

) ) n 9 0 0 C o o % i 7 0 T 7 0 ( C / O = 0 . 7 5 (

s r y e t r e i v v 6 0 8 0 0 u 6 0 i

t S n C / O = 1 . 5 a c o r e c l e 5 0 5 0 T e

C / O = 1 . 5 p e 7 0 0 s

n r e a c t o r s h u t d o w n m 2 a e 4 0 H 4 0 h T t e X 6 0 0 M 3 0 C / O = 1 . 5 3 0

2 0 5 0 0 2 0 0 8 0 1 6 0 2 4 0 3 2 0 4 0 0 0 8 0 1 6 0 2 4 0 3 2 0 4 0 0 T i m e ( m i n ) T i m e ( m i n )

Figure 5.7: (A)CH4 conversions () and temperatures (solid lines) during methane catalytic partial oxidation for 6 h at C/O ratios of 0.75 and 1.5 at 10 % loading of phosphorus. Dashed and dotted horizontal lines indicate undoped values of conver-

sions and temperatures respectively. (B) H2 selectivities over 6 h at C/O of 0.75 and 1.5, undoped values are shown by dashed horizontal lines. Red arrows indicate decrease in conversions/selectivities, blue arrows show increased temperatures.

5.3.4 Transient Studies with Ethanol CPO

The eﬀect of adding 0.05 mol % potassium and phosphorus in ethanol feed is described below:

Eﬀect of Potassium

Potassium maintained steady-state operation for the full 8 h doping period as shown in Figure 5.8 A. Potassium increased the ethanol conversion from 89 % to 100 % (Figure 5.8 A). Temperatures with potassium during doping were slightly higher than the baseline values by about 25 ◦C. Potassium decreased the carbon monoxide selectivities by ∼ 10 % (65 % before doping vs 55 % after doping) while acetaldehyde selectivities increased from about 8 % to 16 % (Figure 5.8 B). Hydrogen selectivities were similar to baseline values and water selectivities also decreased from 40 % to 33 %. 5.4 Discussion 80

1 0 0 9 0 0 7 0 0 . 0 5 % K N o K B 9 0 8 0 0 6 0 A C O 7 0 0 8 0

) 5 0 6 0 0 C o ( 7 0 )

) e % r (

% 5 0 0

( u 4 0 t y H N o K

t 0 . 0 5 % K O i 6 0 a t r

E v i e t X 4 0 0 p c e l 5 0 m 3 0 e 3 0 0 e T S 4 0 E t h a n o l c o n v e r s i o n 2 0 C H C H O 2 0 0 3 T e m p e r a t u r e 3 0 1 0 0 1 0 2 0 0 6 0 1 5 0 2 4 0 3 3 0 4 2 0 5 1 0 6 0 0 6 0 1 5 0 2 4 0 3 3 0 4 2 0 5 1 0 6 0 0 T i m e ( m i n ) T i m e ( m i n )

Figure 5.8: (A) Ethanol conversions () and temperatures (solid lines) during catalytic partial oxidation with ethanol containing 0.05 mol % potassium. (B) CO () and CH3CHO (•) selectivities during the doping period. Arrows indicate the trends (increase/decrease) introduced upon doping.

Eﬀect of Phosphorus

Phosphorus similar to potassium maintained steady state autothermal operation for the full 8 h as shown in Figure 5.9 A. Doping with phosphorus resulted in rise in temperatures of about 100 ◦C. Ethanol conversion also increased from 88 % to about

98 % (Figure 5.9 A). CO2,H2 and C2H4 selectivities are shown in Figure 5.9 B. Phosphorus resulted in a strong decrease in hydrogen selectivities from 53 % to about 19 %. Carbon dioxide selectivities decreased from about 22 % to 4 % while ethylene selectivities increased by from 2 % to about 17 %. Water selectivities increased from 35 % to 57 % and acetaldehyde selectivities decreased from about 10 % to 4 % (not shown).

5.4 Discussion

The effects of potassium and phosphorus on rhodium catalysts have been studied by three sets of experiments: doping at different concentrations for methane and ethanol CPO, studying the effect of higher temperatures in a time-on-stream study for methane CPO, and a transient study by introducing inorganics in the feed for ethanol CPO. These experiments combined give insight into the mechanism of interactions of 5.4 Discussion 81

1 0 0 1 0 0 0 6 0 N o P 9 0 0 0 . 0 5 % P B 9 0 5 0 A 8 0 0 8 0 7 0 0 N o P 0 . 0 5 % P ) 4 0 C ) 7 0 o 6 0 0 (

)

% C O

e 2 (

r % ( y u

C H t t 2 4 6 0 5 0 0 i 3 0 H a

v O r i t H t

E 2 e c p X

4 0 0 e 5 0 l m e

e 2 0 S

3 0 0 T 4 0 E t h a n o l c o n v e r s i o n 2 0 0 1 0 3 0 T e m p e r a t u r e 1 0 0

2 0 0 0 6 0 1 5 0 2 4 0 3 3 0 4 2 0 5 1 0 6 0 0 6 0 1 5 0 2 4 0 3 3 0 4 2 0 5 1 0 6 0 0 T i m e ( m i n ) T i m e ( m i n )

Figure 5.9: (A) Ethanol conversions () and temperatures (solid lines) during catalytic partial oxidation with ethanol containing 0.05 mol % phosphorus. (B) CO2 (), H2 (N) and C2H4 (•) selectivities during the doping period. Arrows indicate the trends (increase/decrease) introduced upon doping.

potassium and phosphorus with rhodium catalysts for CPO. The first set of experiments involving doping at different concentrations for methane and ethanol CPO compares the tolerance of rhodium catalysts to different concen-

trations of inorganic impurities. Rhodium catalysts supported on α-Al2O3 foams have been shown to have dispersions ≤ 1 %.174 Since significant changes were observed at loadings ≥ 10% (effectively 10 atoms inorganic for every atom rhodium considering rhodium dispersion ∼ 1%), the results demonstrate strong resistance to poisoning of rhodium-based catalysts. The second set of experiments involving time- on-stream studies of methane CPO help in understanding the volatility of potassium and phosphorus during CPO. The third set of experiments introduces potassium and phosphorus in the feed, similar to actual biomass and at higher temperatures than that in Chapter 4.108 Comparing the two helps understand how temperature affects the activity of inorganics.

5.4.1 Eﬀect of Potassium

Eﬀect on Lightoﬀ

Potassium appeared to be a strong poison for methane CPO after lightoﬀ with hydrogen. Lightoﬀ occured at higher temperatures with increasing levels of potassium 5.4 Discussion 82

addition. At 100 % loading of potassium, lightoff was not possible with methane after initial lightoff with hydrogen. It has been reported in the literature that potassium reduces the sticking probability of methane on metal surfaces.175–177 The inability of catalysts at 100 % loading of monobasic potassium phosphate catalysts to lightoff may also be attributed to the same reason. Also, methane being highly stable and unreactive in the gas phase,107 the higher reactivity of ethanol may explain operation with ethanol at 100 % loading of both potassium and monobasic potassium phosphate.

Eﬀect on Chemistry

An observation consistent in most of the potassium experiments is that potassium caused a strong decrease in CO selectivities. This is probably a combination of two reasons: alkali metals have been shown to act as promoters to the water gas shift reaction,165,178 thereby increasing consumption of carbon monoxide by reaction with steam. Secondly, due to their electropositive nature, alkali metals reduce the rhodium work function, thereby promoting CO dissociation.179 Similar eﬀects of alkali metals reducing CO selectivities in reforming systems have been reported.180 Introduction potassium in the feed at a lower C/O ratio (∼ 70 ◦C higher temperatures) than that in Chapter 4 has similar eﬀects such as increased ethanol conversions and acetaldehyde selectivities, and decreased CO selectivities.108 The increased acetaldehyde selectivity may be attributed to the alkaline nature of potassium.108,164

Volatility of Potassium

With potassium, catalyst regeneration was possible at higher temperatures in all experiments. In methane experiments, at 1 and 10 % loading, potassium results in reduced CO selectivities. However, at higher temperatures (low C/O ratios), partial catalyst regeneration was observed which led to an increase in CO selectivities of almost 11 % at C/O of 2 with methane. Similar trends were observed at corresponding loadings of monobasic potassium phosphate, in that the eﬀect of decrease in carbon monoxide selectivities is reduced at higher temperatures. However, since the monobasic potassium phosphate-doped catalysts run hotter, complete regeneration of the catalyst was observed even at higher C/O ratios due to potassium volatilization. The time-on-stream studies with methane CPO shows the volatility of potassium at 5.4 Discussion 83

typical temperatures in CPO. Operation for less than an hour at C/O = 0.75 resulted in complete regeneration of catalyst activity at C/O = 1.5 as well. In the ethanol transient experiment, steady-state operation with potassium was obtained indicating the rate of potassium coming in through the feed is balanced by the rate at which it volatilizes. The results of the first set of experiments involving doping potassium at different concentrations show a greater effect on the catalytic activity of methane than ethanol, indicating the presence of oxygen may help in volatilization of potassium species.

5.4.2 Eﬀect of Phosphorus

Eﬀect on Lightoﬀ

With regard to lightoff, phosphorus did not appear to inhibit lightoff as potassium. Lightoff was possible for catalysts containing 0.1 to 100 % phosphorus with both methane and ethanol.

Eﬀect on Chemistry

Phosphorus, unlike potassium is electronegative and causes an increase in the rhodium work function thereby reducing CO dissociation.179 Hence, negligible changes in CO selectivities are observed with phosphorus with methane CPO at all loadings and during transient experiments. A common trend in all phosphorus experiments was the decrease in hydrogen selectivities. This can be attributed to a combination of poisoning of the water gas shift and reforming reactions by phosphorus on rhodium catalysts.156 Also, phosphorus increased ethylene selectivities in both the doped and transient experiments with ethanol due to acidic nature of phosphoric acid. Water selectivities in the transient experiments increased due to a combination of reduced water gas shift, reduced steam reforming and increased ethanol dehydration to ethylene.

Volatility of Phosphorus

In the methane CPO experiments, the catalytic activity of phosphorus was observable at all temperatures even upon regeneration. Even in the methane time-on-stream experiment, catalyst temperatures were stable at around 980 ◦C which resulted in no 5.5 Conclusions 84 observable regeneration. In the ethanol experiments, no regeneration was observed for CO selectivities but regeneration was observed for ethylene selectivities. It may be possible that some phosphorus catalyzing ethanol dehydration is volatilized at higher temperatures leading to lower ethylene selectivities upon regeneration.

5.5 Conclusions

The effects of potassium and phosphorus at different levels of doping on rhodium catalysts were studied on a 5 wt % Rh/α-alumina catalyst. In addition, time-on-stream experiments were also carried out with methane and ethanol CPO to evaluate the volatility of the inorganics. Potassium at low temperatures decreased CO selectivities and its effect on the rhodium catalyst decreased at higher temperatures due to its volatile nature. Phosphorus at higher loadings showed a much more severe poisoning effect than potassium which persisted even at high temperatures and strongly reduced hydrogen selectivities. Potassium phosphate (monobasic) showed cumulative poisoning effects wherein the potassium component could be volatilized at higher temperatures with only the catalytic effect of phosphorus being observed.

5.6 Acknowledgements

The author would like to acknowledge funding from the Minnesota Corn Growers As- sociation and Samuel Blass for discussions related to the experiments in this chapter. CHAPTER SIX

AUTOTHERMAL PARTIAL OXIDATION OF BUTANOL ISOMERS1

Alcohols represent an important intermediate in various biomass upgrading routes. They can be upgraded to syngas or to other hydrocarbons for use as chemicals. The four isomers of butanol oﬀer an interesting platform from which to study the reaction pathways of alcohols in an autothermal partial oxidation system, as they comprise one tertiary, one secondary, and two primary alcohols with the same number of carbon atoms. We demonstrate high yields of syngas or unsaturated molecules at contact times on the order of 10 ms, and investigate the reaction pathways of each isomer over Rh, RhCe, Pt, and PtCe catalysts for a range of carbon-to-oxygen (C/O) ratios. For each isomer, conversion to equilibrium syngas products is essentially complete at C/O = 0.8. As C/O ratio increases, the major product from the primary and secondary butanols switches to the corresponding carbonyl, producing butyraldehyde, isobutyraldehyde, and butanone from 1-butanol, isobutanol and 2-butanol, respectively. Selectivity to the carbonyls approached 30 - 50% as C/O approached 2.0. Dehydration to the corresponding butenes is relatively minor in comparison, representing less than 20% selectivity at C/O = 2.0. tert-butanol reacted diﬀerently, selecting mainly for the dehydration product isobutene. Acetone was the main carbonyl product from tert-butanol, but selectivity to acetone was always ≤ 10%. Global mechanisms in an autothermal reactor, based on pyrolysis, combustion and surface

1Parts of this chapter appear in Jacob S. Kruger, Reetam Chakrabarti, Richard J. Hermann, Lanny D. Schmidt, “Autothermal Partial Oxidation of Butanol Isomers,” Applied Catalysis A : General 411-412 (2012) 87-94. c 2011 Elsevier B.V. Reproduced with permission from Elsevier.

85 6.1 Introduction 86

science literature, are proposed for each alcohol. Surface chemistry likely accounts for much of the syngas formation and heat generation, while the carbonyls and alkenes may be formed primarily through homogeneous routes.

6.1 Introduction

Significant research effort in recent years has focused on the development of renewable liquid fuels in order to decrease the economic and environmental impacts of fossil fuel consumption. Four-carbon alcohols have received particular attention because they can be produced in significant quantities from a variety of renewable feedstocks181–187 and have several advantages over short-chain alcohols. In particular, the butanols have many properties similar to currently consumed fossil fuels, including high energy density, low hygroscopicity, and low corrosivity.188,189 Autothermal partial oxidation over noble metal catalysts has shown promise as a technique to convert a variety of feedstocks (both renewable and fossil fuel-based)

into syngas (a mixture of H2 and CO) and longer-chain molecules, both of which are important intermediates in the synthesis of a wide range of molecules.69–72,85,190,191 Additionally, these reactions are carried out at high temperatures with heat generated in-situ and occur on millisecond time scales, allowing for the use of small and simple reactors that are appropriate for utilizing diﬀuse renewable resources. Recently, St. Clair and Lee192 studied the autothermal partial oxidation of isobutanol over Rh and γ-Al2O3 catalysts and found that Rh gave high yields of syngas while γ-Al2O3 selected primarily for the dehydration product isobutene. We have recently shown that the addition of Ce to Rh catalysts improves yields of hydrogen from several fuels,69,71,191,193 and that autothermal short contact time (SCT) reactors can also give high yields to nonequilibrium products from light oxygenates.73,85 Addi- tionally, we proposed that the behavior of a model compound of diﬀerent functional group classes (e.g. C2 alcohols, aldehydes, and acids) in an autothermal SCT reactor could be extrapolated to other molecules in the same class.85 In this work we apply autothermal partial oxidation of the four butanol isomers over Rh, RhCe, Pt, and PtCe catalysts to show that, depending upon operation parameters, high selectivity to syngas or nonequilibrium products may be obtained. We note that although the formation of the nonequilibrium products may be primarily homogeneous, the exothermic reactions that occur mainly on the catalyst surface are 6.2 Experimental 87 required to provide heat for gas-phase dehydrogenation and dehydration reactions. Additionally, the behavior of the primary butanols during autothermal processing appears to be similar to ethanol,85 supporting the hypothesis that members of a functional group family behave similarly in autothermal SCT reactors.

6.2 Experimental

Experiments were carried out with a 19 mm ID quartz tube as shown in Figure 6.1. Liquid fuel was fed through a stainless steel nebulizer at the top of the reactor at −1 1 mL min and was atomized with 0.7 SLPM N2. Remaining N2 and O2 were fed around the nebulizer, and total N2 and O2 were maintained at air stoichiometry. Gas flow rates were varied to obtain different carbon-to-oxygen (C/O) ratios, defined as the molar ratio of carbon in the feed molecule to atomic oxygen in the O2 feed. The catalyst bed was positioned 30 cm below the nebulizer and reactor walls between the nebulizer and catalyst were maintained at 175 ◦C to vaporize the fuel. An 80 pores-per-linear-inch (ppi) α-Al2O3 foam monolith (Südchemie) was positioned 3.5 cm upstream of the catalyst bed to mix reactants. The catalyst bed was comprised of 3.25 g of 1.3 mm catalytic spheres. The spheres (St. Gobain-Norpro) consisted of 1 wt% M or 1 wt% M-1 wt% Ce supported on α-Al2O3, where M = Rh or Pt. A 65 ppi α-Al2O3 foam monolith served as a back heat shield and to hold thermocouples in place. Type K thermocouples (Omega) were located at the upstream and downstream edges of the sphere bed, and the entire bed was wrapped with aluminosilicate cloth insulation to prevent heat loss and gas bypass during the experiment. Spheres were prepared by incipient wetness impregnation of aqueous catalyst precursor salts (Rh(NO3)3,

H2PtCl6, and Ce(NO3)3) followed by drying under vacuum. Rh and RhCe catalysts were produced in multiple cycles of deposition-dry-calcination at 600 ◦C, where the calcination duration was ten minutes (to decompose the nitrate salt), followed by a ﬁnal calcination at 600 ◦C for 6 hours. Pt and PtCe catalysts were prepared without ◦ the brief calcination steps and were reduced under ﬂowing H2 and N2 at 600 C for 6 hours. The multistep calcination procedure was adopted for the Rh-based catalysts to avoid loss of catalyst metal through delaminating bubbles that formed when a concentrated precursor solution was used. For the Pt-based catalysts, the reduction procedure was employed to mitigate metal loss through volatile intermediates in the decompositon of the precursor salt. 6.2 Experimental 88

Fuel, O 2, N 2

Preheat thermocouple

Heating tape

80 ppi static mixer

Catalyst bed Back heat shield Sample Insulation port

Exhaust

Thermocouples Figure 6.1: Reactor conﬁguration for the autothermal CPO of the butanol isomers. 6.3 Results 89

Analysis of products was performed on an HP6890 gas chromatograph equipped with thermal conductivity (TCD) and ﬂame ionization detectors (FID). The column

was 30m with a PLOT-Q stationary phase. N2 from the feed was used as an internal standard. H2,O2, CO and CO2 were quantiﬁed with the TCD, while all others were

quantified with the FID, except H2O, which was calculated by difference from an oxygen atom balance. Carbon and hydrogen atom balances typically closed within 10%. Each data point shown represents the average of three runs, and 95% confidence

intervals were generally 5-10% absolute. Products up to C6 were analyzed, but no

products > C4 were detected in selectivity > 1%.

6.3 Results

6.3.1 Conversion and Temperature

Each of the butanols sustained autothermal operation for 0.8 ≤ C/O ≤ 2.0, and higher C/O ratios were not attempted. At C/O = 0.8, conversion was ≥ 99% and selectivities closely reﬂect those predicted by thermodynamic equilibrium, which is primarily syngas. Conversion and back face temperature data are shown in Figure 6.2. At the temperatures observed here, residence times ranged from 6-20 ms, calculated at the temperatures reported in Figure 6.2. Conversion was similar for 1-butanol across all four catalysts and was essentially complete at C/O ≤ 1.0, as shown in Figure 6.2. The Pt catalysts yielded a higher temperature at the downstream edge of the catalyst bed than their Rh counterparts by 50-100 ◦C. 2-Butanol was similar to 1-butanol in that conversion was generally similar across catalysts, and temperatures were higher over Pt-based than Rh-based catalysts. Conversion of tert-butanol was generally higher than the other butanols (Figure 6.2), possibly due to its greater tendency toward homogeneous pyrolysis. Like 1- butanol and 2-butanol, the Pt catalysts gave a higher back face temperature that

corresponded to a lower selectivity to CO and H2 (Figure 6.3). Conversion of isobutanol fell oﬀ faster than the other isomers as C/O ratio increased (Figure 6.2). For isobutanol, the Pt catalyst operated roughly 100 ◦C higher than PtCe over the entire C/O range; for the other three isomers, PtCe gave temperatures comparable to or higher than the Pt catalyst. 6.3 Results 90

100% 1200 100% 1200 1-Butanol i-Butanol

1100 1100 Rh 80% 80% Rh RhCe RhCe (Temperature Pt 1000 (Temperature Pt 1000 PtCe 60% PtCe 60% 900 900 800 40% 40% o o Conversion Conversion 800 C) C) 700 20% 20% 700 600

0% 600 0% 500 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 C/O Ratio C/O Ratio

(a) 1-butanol (b) isobutanol

100% 1200 100% 1200 2-Butanol

1100 1100 80% Rh 80% Rh

RhCe (Temperature (Temperature 1000 1000 Pt RhCe PtCe Pt 60% 60% 900 PtCe 900 t-Butanol

800 800 40% 40% o o Conversion Conversion C) C) 700 700 20% 20% 600 600

0% 500 0% 500 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 C/O Ratio C/O Ratio

(c) 2-butanol (d) t-butanol Figure 6.2: Conversion and catalyst backface temperature of the four butanol isomers for each catalyst as a function of C/O ratio. 6.3 Results 91

6.3.2 Syngas and Combustion Products

At C/O ≤ 1.2, CO and H2 were favored from all four isomers, as shown in Figure 6.3 and Figure 6.4. Rh and RhCe generally gave higher conversion to syngas than the Pt catalysts, and PtCe gave signiﬁcantly lower syngas yields than the other three catalysts for C/O ≥ 1.4. The lower selectivity to syngas from the Pt-based catalysts is consistent with the lower steam reforming activity of Pt,106,194 although the selectivity to H2O was not signiﬁcantly higher.

H2O and CO2 were present at 20-30% selectivity regardless of catalyst or isomer;

H2O was present in slightly higher selectivity over the PtCe catalyst, although the magnitude of diﬀerence is near the limit of experimental uncertainty.

6.3.3 C4 Intermediates 1-butanol, 2-butanol and isobutanol gave high selectivity to carbonyl products as

C/O ratio increased, with relatively minor selectivity to C4 oleﬁns as shown in Figure 6.5. The trend for tert-butanol was reversed, with isobutene as the major product. From the primary butanols, selectivity to the aldehyde reached 20-30% as C/O approached 2.0, though the Rh catalyst showed lower selectivity to isobutyraldehyde. Selectivity to butanone from 2-butanol reached 35-45% for the same C/O range. For 1-butanol over all catalysts and for 2-butanol over the Pt-based catalysts, selectivity to the carbonyl at C/O ≤ 1.4 was comparable to butene selectivity (shown as a sum of 1-butene and both 2-butene isomers), but at C/O = 2.0, the carbonyl was favored by a factor of 2-3. The ratio of carbonyl to butene was even higher over the Rh catalysts for 2-butanol, reaching a factor of 5-10 in preference of the carbonyl. For isobutanol, selectivity at C/O = 2.0 favored the carbonyl over isobutene by a factor of 3-5. Because tert-butanol lacks an α-H atom, formation of a carbonyl cannot occur through simple hydrogen abstraction reactions. A dehydration route to isobutene is thus predominant, with selectivities to isobutene reaching 60-70% over the Pt catalysts and 30-50% over the Rh catalysts. Acetone was the main carbonyl product formed from tert-butanol, but represented ≤ 10% selectivity at all C/O ratios investigated. 6.3 Results 92

100% 100% 1-Butanol i-Butanol

Rh 80% RhCe 80% Rh Pt RhCe PtCe Pt 60% 60% H PtCe H2 2

40% 40% HSelectivity HSelectivity

20% 20%

H2O H2O

0% 0% 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 C/O Ratio C/O Ratio

(a) 1-butanol (b) isobutanol

100% 100% 2-Butanol t-Butanol

Rh Rh 80% RhCe 80% Pt RhCe PtCe Pt PtCe 60% H2 60% H2

40% 40% HSelectivity HSelectivity

20% 20%

H2O H2O 0% 0% 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 C/O Ratio C/O Ratio

Figure 6.3: Selectivities to H2 and H2O from each isomer as a function of C/O Ratio over all four catalysts. 6.3 Results 93

100% 100% Rh 1-Butanol Rh i-Butanol RhCe RhCe Pt Pt 80% PtCe 80% PtCe PtCe PtCe

CO Other 60% 60% Other Pt CO

40% 40% Rh CSelectivity CSelectivity

CO 2 CO 2 20% 20%

0% 0% 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 C/O Ratio C/O Ratio

(a) 1-butanol (b) isobutanol

100% 100% Rh 2-Butanol Rh t-Butanol RhCe RhCe Pt Pt 80% 80% PtCe PtCe PtCe PtCe

60% CO Other 60% CO Other Rh

RhCe 40% 40% CSelectivity CSelectivity

CO 2 CO 2 20% 20%

0% 0% 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 C/O Ratio C/O Ratio

Figure 6.4: Selectivities to CO and CO2 from each isomer as a function of C/O Ratio over all four catalysts. ‘Other’ represents the sum of all carbonaceous products other than CO and CO2. For clarity, only the minimum and maximum lines are shown. 6.3 Results 94

70% 70% 1-Butanol i-Butanol

60% 60% Rh Rh RhCe RhCe 50% Pt 50% Pt PtCe PtCe 40% 40%

30% 30% i-C 4H8O CSelectivity CSelectivity i-C 4H8O 20% 20%

10% Butenes 10%

Butenes 0% 0% 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 C/O Ratio C/O Ratio

(a) 1-butanol (b) isobutanol

70% 2-Butanol 70% t-Butanol 60% Rh Rh 60% RhCe RhCe 50% Pt Pt 50% PtCe PtCe Butanone 40% 40% Butenes 30% 30% CSelectivity CSelectivity 20% Butenes 20%

10% 10% Acetone

0% 0% 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 C/O Ratio C/O Ratio

Figure 6.5: Selectivities to major carbonyl and C4 oleﬁns from each isomer as a function of C/O Ratio over all four catalysts. 6.4 Discussion 95

6.3.4 Other Intermediates

The other species observed in significant amounts were generally only ethylene and propylene. Selectivities to these two olefins were similar over 1-butanol at 5-15% for C/O ≥ 1.4, regardless of catalyst. Isobutanol, however, showed higher selectivity to propylene at 10-20% for C/O ≥ 1.4 and only minor selectivity to ethylene, at ≤ 5%. The difference in selectivity to propylene from isobutanol across catalysts cannot be explained by temperature alone. Pt and Rh, which displayed the highest and lowest back-face temperatures, respectively (Pt was over 100 ◦C higher than Rh at all C/O ratios), gave comparable selectivities to propylene, while RhCe and PtCe give noticeably higher selectivity to propylene. This observation suggests that either production or consumption of propylene (or both) occurs to some extent on the catalyst surface. 2-butanol and tert-butanol showed lower selectivities to these two olefins, although 2-butanol produced significant selectivity to each over PtCe. It is difficult to discern the role of the PtCe catalyst in the formation of ethylene and propylene from 2-butanol because the range over which their selectivities are significant corresponds to the range where PtCe temperature was nearly 100 ◦C higher than the other catalysts. Their formation may therefore be homogeneous, possibly from the thermal decomposition of 1- and 2-butenes, which is discussed below.

6.4 Discussion

6.4.1 Chemistry of the Isomers

Reaction schemes within a CPO reactor are complex networks of homogeneous and heterogeneous reactions that are diﬃcult to untangle from analysis of integral data. However, the reaction schemes shown below are not unprecedented in the literature, and by observation of reaction intermediates and comparison between catalysts, a qualitative formulation of the dominant reaction pathways is possible.

1-Butanol

Alcohols generally decompose by either dehydrogenation to produce a carbonyl or dehydration to produce an alkene.195,196 For 1-butanol, dehydration and dehydro- ◦ genation routes are competetive in the gas phase in the absence of O2 up to 500 C, 6.4 Discussion 96

50% 50% 1-Butanol i-Butanol

40% Rh 40% Rh RhCe RhCe Pt Pt PtCe 30% 30% PtCe

C H 20% C3H6 20% 3 6 CSelectivity CSelectivity

C H 10% 2 4 10%

C2H4

0% 0% 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 C/O Ratio C/O Ratio

(a) 1-butanol (b) isobutanol

50% 50% 2-Butanol t-Butanol

Rh 40% 40% Rh RhCe RhCe Pt Pt PtCe PtCe 30% 30%

20% 20% CSelectivity CSelectivity C3H6 C H 10% 2 4 10%

C2H4 C3H6 0% 0% 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 C/O Ratio C/O Ratio

(c) 2-butanol (d) t-butanol Figure 6.6: Selectivities to ethylene and propylene as a function of C/O Ratio over all four catalysts. 6.4 Discussion 97

(a) 1-butanol (b) isobutanol

(c) 2-butanol (d) t-butanol Figure 6.7: Proposed reaction mechanisms for each isomer in the autothermal reactor. For brevity, only products with ≥ 5% selectivity are included. Arrow thickness is proportional to selectivity; intermediate species may also adsorb to surface and react. 6.4 Discussion 98

while dehydrogenation predominates at higher temperatures.195,197 This scheme is consistent with the product spectrum observed here, as temperatures were well above 500 ◦C and selectivity to the primary products at high C/O ratios (where secondary reactions are less favored) demonstrates 2 to 3-fold diﬀerence in selectivity in favor of the aldehyde over the alkenes. As C/O decreases (and temperature increases), these primary products can further decompose. The butyraldehyde intermediate can react by decarbonylation to yield CO, H2 and C3H6, or by demethylation to yield propanal

and a CHx radical; the butenes may decompose to produce propylene and a CHx radical.195,197,198 Ethylene may be formed by reactions of several intermediate species via demethylation and dehydration reactions.197,199 Other than the lower selectivity to syngas species for PtCe, there are few features in the selectivity data to aid in distinguishing between the diﬀerent catalyst surfaces for 1-butanol. Further discussion of potential surface reactions of 1-butanol will therefore not be attempted here; potential surface mechanisms of alcohols in general are discussed below.

2-Butanol

The major non-equilibrium product from 2-butanol for C/O ≥ 1.2 was butanone, which reached 35-45% selectivity as C/O incereased to 2.0. At C/O ≤ 1.4 over the Pt catalysts, dehydration reactions appeared to be competitive with dehydrogenation, as selectivities to the butenes and butanone were comparable. When the butenes were produced in ≥ 2% selectivity, 1-butene was favored over either isomer of 2-butene, with selectivities to 1-butene, cis-2-butene, and trans-2-butene present generally in the ratio 2:1:1, respectively. Negligible selectivity to acetaldehyde (< 1%) and only minor selectivity to propanal were observed in these experiments over any of the catalysts, despite their prominence in the 2-butanol pyrolysis literature.195,200 We note

here that O2 conversion was less than 100% under some reaction conditions (PtCe

for C/O ≥ 1.2); in other similar experiments, the presence of O2 signiﬁcantly decreased selectivity to acetaldehyde,85,87 and in 2-butanol combustion, acetaldehyde and propanal were much less prominent.201 These observations likely mean that oxygen persists until near the downstream edge of the catalyst bed, at least for the PtCe catalyst. This hypothesis is also consistent with the very small amounts of hydrocarbon products observed both here and in 2-butanol combustion.201 PtCe showed a higher selectivity to the ethylene and propylene than the other 6.4 Discussion 99

three catalysts, despite the observed O2 breakthrough. As discussed above, the range of C/O ratios over which these two alkenes show signiﬁcant selectivity corresponds to signiﬁcantly higher temperatures over PtCe; we thus surmise that they may be produced by homogeneous decomposition of either butanone (via unstable ketene in- 202 198,203 termediates ) or the butenes. Indeed, CH4 (not shown) was produced in nearly equal quantities as C3H6, which is expected from the pyrolysis of 1- and 2-butene; 196 202 the additional C2H4 may have been produced by oxidative or nonoxidative decomposition of butanone. CH4 can also be produced by pyrolysis of butanone by reactions that also produce acetone and propanal,195 but the negligible selectivity to acetone and propanal suggests that these pathways are minor.

6.4.2 tert-Butanol

Other than the four equilibrium species, tert-butanol yielded only one main product, isobutene, which approached 70% selectivity over PtCe at C/O = 2.0. Other light oleﬁns were never present in ≥ 1% selectivity, and acetone, the primary carbonyl- containing product, was always ≤ 10% selectivity. Over PtCe, selectivity to propanal (not shown) was comparable to acetone. Isobutene may result mainly from homogenous dehydration, either molecularly or through sequential H and OH radical abstractions.199 It is also possible that isobutene could be formed from surface reactions as well. We note that selectivity to isobutene from tert-butanol across catalysts (lowest on RhCe, highest on Pt and PtCe) more closely matches the conversion than the temperature trends. Thus, at least some of the isobutene may come from surface reactions; although selectivity to isobutene may also be related to the ability of each catalyst to reform this intermediate. It is also of interest that the activity of the catalysts toward C–O bond scission may be a factor in determining the conversion and selectivity trends of each catalyst. The Pt-containing catalysts may be more active for C–O bond scission,204 resulting in a higher conversion of tert-butanol. However, some of the initial dehydration of tert-butanol may also be homogeneous; the lower temperatures observed over the Rh- based catalysts (likely due to greater activity for endothermic reforming) may also result in lower initial conversion of the incoming fuel. The other nonequilibrium product was acetone, which may have been produced in 199 the gas phase by subsequent H and CH3 abstractions. We also note that the neg- 6.4 Discussion 100

ligible selectivty to CH4,C2H4 and C3H6 suggests that homogeneous decomposition of isobutene is not signiﬁcant.198,205,206

6.4.3 Isobutanol

The main nonequilibrium products for isobutanol were isobutyraldehyde, propylene, and isobutene. Gas phase mechanisms for the production of these species are relatively straightforward, and closely reﬂect the patterns of 1-butanol and 2-butanol. Isobutene may have been produced by either molecular dehydration or by abstraction of H and OH radicals.199,201 Sequential H abstractions may have produced isobutyraldehyde, which in turn may have decomposed to propylene and syngas species.195 Alternatively, propylene may have been produced either by abstraction of a primary H radical from isobutanol, followed by loss of a CH2OH radical, or by demethylation of the parent alcohol followed by abstraction of an OH radical.199,201 PtCe and RhCe showed relatively high selectivities to the aldehyde and propylene, Rh showed relatively low selectivities to both, and Pt showed high selectivity for the aldehyde and low selectivity for propylene. As discussed above, these features cannot be explained solely by temperature eﬀects, suggesting that some isobutyraldehyde and propylene are either produced or consumed on the catalyst surface.

6.4.4 Surface Chemistry

Although the product spectra generally match a homogeneous mechanism well, surface reactions are certainly contributing to the observed products. Previous experiments have found that alcohols typically adsorb onto Pt and Rh surfaces by an oxygen lone pair, followed by O–H scission.207–214 Surface reactions after this point may di- verge on Pt and Rh,207 and to our knowledge have not been well studied on PtCe or RhCe. A feature shared by alcohol decomposition pathways on these surfaces is the dehydrogenation and decarbonylation of intermediate species, and desorption of H2 and CO from the surface. Under certain conditions, intermediates may desorb,207,215 although we are unable to discern the relative contribution to the observed product spectrum with the current analytical setup. It is also possible that intermediate carbonyl and oleﬁn species formed in the gas phase are reacting on the catalyst surface. Under diﬀerent conditions, Pt and Rh surface mechanisms for carbonyls have been found to be similar,207–211 and initial 6.4 Discussion 101

reactions in combustion of the primary and secondary butanols lead primarily to carbonyls that may react on the surface.196 The actual situation is likely a convolution of the alcohol reacting on the surface and in the gas phase, vapor-phase intermediates reacting further in the gas phase and on the surface, and surface intermediates reacting further on the surface and desorbing to the gas phase, where they may or may not react further. To avoid unnecessary speculation, we have kept our proposed mechanisms as general as possible, proposing only what appear to be the dominant overall pathways based on observed products and intermediates with ≥ 5% selectivity. Figure 6.7 shows general reaction schemes for the CPO of each isomer.

6.4.5 Eﬀect of Catalyst

Trends for each alcohol were similar across the four catalysts studied, namely, high

selectivity to CO and H2 at low C/O ratios, steady selectivity to H2O and CO2 across C/O ratios, and increasing selectivity to intermediate species (e.g. aldehydes and oleﬁns) at high C/O ratios. However, the absolute selectivities were diﬀerent

across catalysts, with PtCe consistently exhibiting lower selectivity to CO and H2 than the other catalysts. PtCe was conversely more selective for intermediate species than the other catalysts, particularly for 2-butanol and iso-butanol. The observed temperature and syngas selectivity trends across catalysts appears to be related to endothermic reforming reactions of the feed alcohols and intermediate

carbonyls and alkenes to CO and H2. The trend of selectivity for CO and H2 across catalysts, in order of decreasing selectivity, was RhCe ≈ Rh > Pt > PtCe, while the trend of catalyst backface temperature was generally PtCe > Pt > Rh ≈ RhCe. The addition of Ce likely has different effects for the Pt and Rh catalysts. For Pt, addition of Ce may lead to less active Pt centers by promoting bulk oxidation of the Pt metal, while the oxidation state of Rh (and hence the catalytic activity) is more sensitive to the surrounding environment than the presence of Ce.216 In particular, the addition of Ce to Pt has been shown to inhibit oxidation of hydrocarbons,217 consistent with the relatively higher selectivity to olefins in the current experiments,

and the observation that O2 consumption was less than 100% over PtCe over some of the C/O range investigated for each alcohol. In the present experiments, it appears that adding Ce to Pt also inhibits reforming reactions, as discussed above, at least for small oxygenates and alkenes. 6.5 Conclusion 102

6.5 Conclusion

Primary and secondary alcohols decompose in the autothermal system mainly via a carbonyl intermediate in surface and homogeneous reactions. This result provides support to our previous proposal that molecules within in a functional group class behave similarly in a SCT reactor.85 Tertiary alcohols, with no α-H atom available for dehydrogenation, decompose by dehydration, although the lack of a dehydrogenation pathway does not necessarily lead to lower overall reactivity in an autothermal reactor. Because of the similarity in trends across catalysts, it is difficult to discern relative contributions of homogeneous and heterogeneous reactions. The actual situation is likely a convolution of multiple reaction schemes, although some effect of catalyst is observed in the selectivities to CO and H2. To that end, our results are consistent with previous work106,194 that found Pt and PtCe to be less active reforming catalysts than Rh and RhCe for these molecules, although as C/O approaches 0.8, all catalysts reform the alcohol isomers to an equilibrium syngas stream. Alternatively, as C/O approaches 2.0, high selectivities to carbonyls and light olefins, which are important chemical intermediates, are achieved.

Acknowledgement

Funding for this research was graciously provided by the Department of Defense (DOD) through Fuel Cell Energy, Inc. in Danbury, CT. CHAPTER SEVEN

AUTOTHERMAL REFORMING OF ISOBUTANOL1

In Chapter 6, the catalytic partial oxidation of butanol isomers to a high selectivity thermodynamically equilibrated synthesis gas stream at low C/O ratios was demonstrated. As mentioned in Section 1.3.3, one of the energy uses of syngas is to produce hydrogen for fuel cells. In this chapter, catalytic partial oxidation (CPO) of isobutanol to produce syn-

thesis gas with varying H2/CO ratios has been performed autothermally in a staged millisecond-contact-time reactor. A 1 wt% Rh-1 wt% Ce/α-alumina catalyst was used to carry out CPO of isobutanol over a range of fuel to oxygen ratios (C/O) in ∼ 20 ms residence times. Steam was added (S/C = 0 to S/C = 3) to promote the water gas shift (WGS) reaction and increase hydrogen production. Without steam

addition, maximum CO and H2 selectivities obtained were greater than 70 %. Steam addition increased the maximum hydrogen selectivity to 103 %, at C/O = 1 and S/C = 2. Conversion of isobutanol and oxygen was always > 99 % at all C/O and S/C ratios after CPO where the autothermal temperatures were typically between 600 and 1000 ◦C. A 1 wt% Pt-1 wt% Ce catalyst was added downstream of the CPO stage

to further reduce the CO and increase the H2 concentrations by the WGS reaction. Addition of the WGS stage resulted in a product stream containing 1.7 to 3 mol % CO at S/C of 2 and 3 which is comparable to the exit stream CO concentration of

1Portions of this chapter appear in Reetam Chakrabarti, Jacob S. Kruger, Richard J. Hermann, Lanny D. Schmidt, “Autothermal reforming of isobutanol,” RSC Advances 2 (2012) 2527-2533. c 2012 The Royal Society of Chemistry. Reproduced by permission of The Royal Society of Chemistry and available online at http://pubs.rsc.org/en/content/articlelanding/2012/ra/c2ra01348g.

103 7.1 Introduction 104

an industrial high temperature shift catalyst. With steam addition, selectivities to hydrogen greater than 100 % were obtained at all C/O ratios, with a maximum of

122 % at C/O = 0.9 and S/C = 3. At S/C = 3, the H2/CO ratio increased from ∼ 3 after the CPO stage to ∼ 12 after the WGS stage. Also, the WGS catalyst reduced the selectivities of non-equilibrium products like isobutylene and isobutyraldehyde by almost half compared to that after the CPO stage, which can reduce syngas cleanup costs after a reformer. Since the total residence time in the reactor is ∼ 100 ms and the product distribution can be tuned by addition of steam, this represents a compact and simple system for producing renewable syngas or hydrogen.

7.1 Introduction

Biomass represents a renewable alternative to fossil fuels to partly fulfil the increasing demand for carbon-based fuels and chemicals. However, the distributed nature of biomass leading to increased transportation costs and its low energy density compared to fossil fuels represent challenges for large-scale processing to produce biofuels.218 Upgrading biomass to a higher density liquid intermediate reduces transportation costs and is economical for centralized processing. Isobutanol is produced by fermentation of sugars such as glucose and cellulose by different strains of bacteria or yeast.181,219,220 Efforts are currently underway to commercialize the manufacture of isobutanol by fermentation of different biomass sources.220 Unlike other biofuels such as ethanol, isobutanol has a higher energy-density, lower volatility, and it does not ab- sorb moisture. Transporting a liquid fuel such as isobutanol is more economical than solid biomass, so that operation can be carried out in a central processing location and utilize economies of scale. Gasification of isobutanol by partial oxidation (Eq. 7.1) produces synthesis gas, a mixture of hydrogen and carbon monoxide.

3 i − C4H9OH + 2 O2 −→ 4 CO + 5 H2 ∆Hr = -160kJ/mol (7.1) Syngas is the starting material for ∼ 50 % of petrochemicals synthesized world- 40 wide. Depending upon the H2/CO ratio, diﬀerent chemicals can be synthesized.

For example, a H2/CO = 2 is required for synthesis of Fischer Tropsch diesel and 3 methanol, whereas H2/CO = 1.2 is preferred for mixed alcohols. Hydrogen is re- 7.1 Introduction 105

garded as an important energy carrier since it is clean and has a high speciﬁc energy. High purity hydrogen with low amounts of CO is necessary for fuel cells, which can be used to power vehicles or to generate electricity. To make high purity hydrogen free from CO for fuel cells, the output stream from a reformer containing products

such as H2, CO, CO2,H2O is processed by the WGS reaction (Eq. 7.2) represented below:

CO + H2O −→ H2 + CO2 ∆Hr = -41kJ/mol (7.2) WGS typically takes place in industry in two stages - a high temperature shift ◦ stage on Fe2O3 - Cr2O3 at 350 - 500 C followed by a low temperature shift stage on Cu-ZnO catalysts at 180 - 250 ◦C.119,221 The CO concentration is reduced to about 3- 5 % after the high temperature shift stage and about 0.5 % after the low temperature shift stage. Further CO removal steps include preferential oxidation (PROX) or CO methanation which reduce the CO concentration to ppm levels; ≤ 50 ppm is desired for PEM fuel cells.222,223 Noble metals such as Rh and Pt have been shown to convert diﬀerent feedstocks to a high selectivity synthesis gas stream autothermally in milliseconds by catalytic partial oxidation (CPO).1–3,61 This process takes place without the formation of un-

desired tars or chars. The H2/CO ratio can be manipulated by adjusting the fuel to oxygen ratio and also by addition of steam to drive the WGS reaction. Due to the short contact times involved, heat losses are minimal; making these reactors run close to adiabatic conditions. Hence, the product stream coming from these reactors contains substantial thermal energy which can be used for downstream reactions. In this experiment, we integrate the exothermic CPO with a WGS stage without any external heat input to the catalyst. Pt-Ce based catalysts have been shown to equilibrate the WGS reaction at short contact times (order of milliseconds) at temperatures above 550 ◦C.224 In contrast, commercial WGS catalysts have residence times in order of seconds. By integrating the noble metal stage and Pt-Ce based WGS stage, we have developed an integrated system which produces CO levels comparable to that after a commercial high temperature shift stage in a total residence time of ∼ 100 ms. This results in a compact system suitable for small-scale distributed applications or onboard a fuel cell powered vehicle. The reactions in this integrated system are oxidation of isobutanol (Eq. 7.3) which provides the necessary heat energy to drive the entire system, endothermic 7.2 Experimental 106

steam reforming (Eq. 7.4) and the slightly exothermic WGS reaction (Eq. 7.2), represented below:

i − C4H9OH + 6 O2 −→ 4 CO2 + 5 H2O ∆Hr = -2500kJ/mol (7.3)

i − C4H9OH + 3 H2O −→ 4 CO + 8 H2 ∆Hr = 566kJ/mol (7.4) In addition to the above products, hydrocarbon-based products such as methane, ethylene, ethane, propylene, propane, acetone, propanaldehyde, isobutylene, cis- and trans-2-butene and isobutyraldehyde are also formed. Thermodynamic equilibrium at reactor temperatures exclusively predicts syngas and combustion products with small amounts of methane (≤ 1% at C/O ≤ 1) and negligible selectivities to the other non- equilibrium products (≤ 0.1%). These undesired products become signiﬁcant with decreasing temperatures in a CPO reactor and they tend to further react to form tars and chars. Since these products do not likely desorb on rhodium surface, their formation may be attributed mostly to homogeneous chemistry within the reactor.85 The above three equations can be combined to give the overall reaction (Eq. 7.5) for the integrated system consisting of CPO and WGS.

3 i − C4H9OH + 4 H2O + 2 O2 −→ 9 H2 + 4 CO2 ∆Hr = -324kJ/mol (7.5)

In this work, we convert isobutanol to a high selectivity synthesis gas stream using Rh-Ce based catalysts. The effect of different fuel to oxygen ratios was studied to maximize syngas yields. Additionally, the effect of steam addition to drive the WGS reaction was investigated. To produce a high purity hydrogen stream free from CO for fuel cell applications, a Pt-Ce based WGS catalyst was added downstream of the Rh-Ce catalyst. This resulted in a stream containing ∼ 2% CO.

7.2 Experimental

A schematic of the reactor conﬁguration is shown in Figure 7.1. Experiments were carried out in a 19 mm I.D. quartz tube. Isobutanol was fed in through a HPLC pump (Chromtech) and water was fed through a syringe pump. Both isobutanol and 7.2 Experimental 107

Argon, Oxygen

Vaporized Isobutanol Steam

Front Heat Shield Insulation 1% Rh-1% Ce Catalyst

Back Heat Shield Sample Port 1

1% Pt-1% Ce Catalyst

Back Heat Shield Sample Port 2 Exhaust

Figure 7.1: Reactor configuration for autothermal catalytic partial oxidation of isobutanol. Black circles indicate positions of thermocouples. water were vaporized by a resistively heated rod maintained at ∼ 200 ◦C. Gases were fed to the reactor at a controlled flow rate by mass flow controllers accurate to within

± 2 %. The catalyst consisted of 5 g of 1 %Rh-1 %Ce supported on 1.3 mm α-Al2O3 spheres. Uncoated 80 ppi α-Al2O3 monoliths (17 mm diameter, 10 mm long) were used as front and back heat shields to minimize heat losses due to radiation. The reactor tube was wrapped in 2 cm thick insulation to minimize heat losses. 30 mm downstream of the back heat shield, 15 g of 1 %Pt-1 %Ce supported on 1.3 mm α-

Al2O3 catalyst was placed for WGS. Products were analyzed after each stage with a HP 5890 Series II GC. The temperature at the back of the CPO stage and at the top and back of the WGS stage were measured with type-K thermocouples. Catalysts were run for up to 20 h without any observable deactivation. 7.3 Results 108

7.2.1 Catalyst Preparation

The CPO catalyst consisted of 1 wt% Rh-1 wt% Ce supported on 1.3 mm α-Al2O3 spheres while the WGS catalyst consisted of 1 wt% Pt-1 wt% Ce supported on 1.3 mm α-Al2O3 spheres. Both the catalysts were prepared by the incipient wetness technique using rhodium nitrate, cerium nitrate and chloroplatinic acid as precursors for Rh, Ce and Pt respectively, as described previously.100 Catalysts were dried in a vacuum oven. The Rh-Ce catalyst was calcined in a furnace at 800 ◦C for 6h and the ◦ Pt-Ce catalyst was reduced at 500 C in a 5 % H2 atmosphere for 6h.

7.2.2 Product Analysis

Products were analyzed after each stage with a HP 5890 Series II GC with thermal conductivity and flame ionization detectors, and a 60 m Plot-Q column. Argon was used as an internal standard. Water flow rate was quantified by closing the hydrogen and oxygen balances and averaging the two. Carbon, hydrogen and oxygen balances typically closed with ± 10 %. Each data point represents the average of three runs. 95 % confidence intervals were typically less than 5 %. Selectivities to carbon-based products were defined on a carbon-basis and selectivities to H2 and H2O on a hydrogen basis. Selectivity (C or H) to a species was defined as the number of atoms of C or H in a species divided by the total number of atoms of carbon or hydrogen in the fuel, respectively. Steam, which was co-fed was not considered fuel, and made it possible to obtain H2 selectivities greater than 100 %. Equilibrium calculations were performed with HSC 7.1 software by minimizing Gibbs free energy.

7.3 Results

For all the experiments presented, the backface temperature of the catalyst was less than 1000 ◦C to limit deactivation due to sintering. In all experiments, the flow rate of isobutanol was 0.65 ml/min. The results are reported after the CPO stage and after the WGS stage. For each of the two configurations, the fuel to oxygen ratio and the steam to carbon ratio was varied. The C/O ratio is defined as the ratio of number of carbon atoms in the fuel to number of oxygen atoms in the O2 feed. Argon was used as a substitute nitrogen in air; the ratio of argon to oxygen corresponded to air stoichiometry (79:21). The steam to carbon ratio is defined as the ratio of molar flow 7.3 Results 109

rates of steam to that of carbon in the fuel.

7.3.1 CPO of Isobutanol

For all the conditions studied, the conversion of isobutanol and oxygen at all C/O ratios and S/C ratios was greater than 99 %. The results of CPO of isobutanol are shown in Figure 7.2. At low C/O ratios, nearly equilibrium product selectivities were obtained. The hydrogen selectivities for diﬀerent C/O and S/C ratios are shown in Panel A. At all S/C ratios, the hydrogen selectivity passes through a maximum.

At lower C/O ratios, combustion chemistry dominates, which reduces H2 selectivity, whereas at higher C/O ratios, reforming activity is supressed due to lower temperatures (Panel B), which also reduces H2 selectivity. Steam addition from S/C = 0 to S/C = 2 caused an increase in hydrogen selectivities due to increased WGS activity. Hydrogen selectivities greater than 100 % were observed at C/O = 1 and 1.1 for S/C = 2. Further addition of steam (S/C) = 3, decreased hydrogen selectivities, most likely due to quenching of the reactor from steam addition, which lowers temperatures and inhibits syngas-producing reforming reactions. The temperatures in the reactor decreased with both increasing C/O and S/C ratios (Panel B). Backface temperatures within the catalyst bed ranged from ∼ 950 ◦C to 600 ◦C. At S/C ≤ 1, at C/O = 0.7, the backface temperature was greater than 1000 ◦C , hence this regime was not investigated. Since experiments were carried out at a constant fuel ﬂow rate, greatest decline in temperatures was observed from C/O of 0.7 to 0.8 (∼ 100 ◦C ) compared to ∼ 20 ◦C from 1.1 to 1.2. Similar to the trends in temperatures, the CO selectivities decreased with increasing C/O and increasing S/C ratios (Panel C). At a S/C = 0, CO selectivities ranged from 77 % (C/O = 0.8) to 59 % (C/O = 1.2), whereas at S/C = 3, they varied from

37 % (C/O = 0.7) to 30 % (C/O = 1.2). At the same C/O ratio, CO2 selectivities increased with increasing steam addition due to increased WGS and reduced reforming as a result of lower temperatures (Panel D). At S/C = 0, CO2 selectivities ranged from 16 % (C/O = 0.8) to 10 % (C/O = 1.2), whereas at S/C = 3, they ranged from 57 % (C/O = 0.7) to 43 % (C/O = 1.2). Non-equilibrium products from isobutanol CPO included ethylene, ethane, propylene, propane, isobutylene, cis- and trans-2-butene, acetone, propanal and isobutyraldehyde. At a given S/C ratio, their selectivities increased with increasing C/O ra- 7.3 Results 110

1 5 0 1 0 0 0 1 4 0 A B 1 3 0 9 0 0 1 2 0 ) ) C % E q S / C = 3 o ( (

1 1 0 8 0 0 y e t r i u v i 1 0 0 t

t a c r

e S / C = 0 e l 9 0 7 0 0 p e s m S / C = 2 2 8 0 e H S / C = 1 T

S / C = 1 E q S / C = 0 S / C = 2 7 0 S / C = 3 6 0 0 S / C = 3 6 0 S / C = 0 5 0 5 0 0 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3

C / O C / O 9 0 9 0 8 0 E q S / C = 0 C 8 0 D 7 0 7 0 E q S / C = 3 ) )

% S / C = 0 % ( 6 0 6 0 (

S / C = 3 y y t i t i

v S / C = 2 i

5 0 v 5 0 i t t c c

e l e 4 0 S / C = 1 l 4 0 e e

s S / C = 1 s

S / C = 2 2 O 3 0 3 0

S / C = 3 O C

C 2 0 2 0 E q S / C = 3 S / C = 0 1 0 1 0 E q S / C = 0 0 0 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3 C / O C / O

1 0 1 6 S / C = 0 E F 8

) 1 2 I s o b u t y l e n e ) % (

S / C = 3 S / C = 1 % y ( t i

6 y v t i i

t S / C = 2 v c i t

e S / C = 2 8 l c

S / C = 3 e e l s

S / C = 1 e

6 4 s

H 4

3 S / C = 0 C C

4 2

S / C = 3 I s o b u t y r a l d e h y d e 1 S / C = 0 0 0 2 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3

C / O C / O

Figure 7.2: Product distributions and temperatures from catalytic partial oxidation of isobutanol. represents S/C = 0, • S/C = 1, N S/C = 2 and S/C = 3. Dashed lines represent equilibrium values at S/C = 0 and 3, calculated at the catalyst backface temperaure. 7.3 Results 111

tios due to reduced reforming activity at lower temperatures. Propylene, isobutyraldehyde and isobutylene were generally the major non-equilibrium products. Propylene selectivities were mostly invariant to steam addition and were generally 5 % or lower (Panel E). Isobutylene selectivities decreased with steam addition whereas isobutyraldehyde selectivities increased (Panel F).

7.3.2 Addition of WGS Stage

A 1 % Pt-1 % Ce/α-alumina catalyst was placed 30 mm downstream of the CPO catalyst to produce a high purity hydrogen stream. The product distributions are shown in Figure 7.3. The hydrogen selectivites obtained with this staged reactor system are shown in Panel A. The hydrogen selectivities were higher than that with just the CPO stage and increased with addition of steam due to more favorable equilibrium of the WGS reaction. As with the CPO stage, for a given S/C ratio, the hydrogen selectivites displayed a maximum at a particular C/O ratio. Without any steam addition (S/C = 0), the highest H2 selectivity obtained was 84 % at a C/O of

0.9. With steam addition, H2 selectivities were generally greater than 100 %. The highest hydrogen selectivity of 122 % was obtained at a C/O of 0.9 and S/C = 3 as compared to the theoretical maximum of 180 % as predicted by Eq. 7.5. The backface temperatures in the WGS stage are shown in Panel B. At a given C/O ratio, from S/C = 0 to S/C = 2, the backface temperature in the WGS stage increased due to the exothermic nature of the WGS reaction and increasing conversion of CO with increasing S/C. Temperatures at S/C = 2 were similar to that at S/C = 3. For a given S/C ratio, temperatures decreased with increasing C/O ratios. The typical diﬀerence between the backface temperatures in the CPO stage and outlet of the WGS stage was ∼ 300 ◦C.

Similar to the CPO stage, the largest changes in H2, CO and CO2 selectivities

were observed from S/C = 0 to S/C = 1. The CO and CO2 selectivities are shown in Panels C and D respectively. At S/C = 0, at C/O ratios greater than 0.9, the CO selectivities were similar to that with just the CPO stage due to negligible shift activity in the WGS stage. As S/C ratios increased, the amount of CO converted in the shift stage increased, reducing CO selectivities. The lowest CO selectivity of ∼

12 % was obtained at a C/O of 0.8 and S/C ratio of 3. CO2 selectivities decreased with increasing C/O ratio and increased with increasing S/C ratio. The WGS stage 7.3 Results 112

1 3 0 6 5 0 A B 1 2 0 5 8 0 1 1 0 S / C = 3 )

S / C = 1 ) %

C 5 1 0 ( o

1 0 0 S / C = 2 (

y t e i r

v i u

t S / C = 3 t

c 9 0 a

r 4 4 0 e l e e

p S / C = 2 s S / C = 0

m 2 8 0 e H

T 3 7 0 S / C = 1 7 0 S / C = 0 6 0 3 0 0 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3 C / O C / O

9 0 9 0 S / C = 3 8 0 C 8 0 D S / C = 2 7 0 7 0 S / C = 0 )

% S / C = 1 ) ( 6 0 6 0 % y ( t i

v y i t t i 5 0 5 0 c v i t e l

c e

e 4 0 4 0 l s

e S / C = 1 2 s

S / C = 0 3 0 O 3 0 O C

S / C = 2 2 0 2 0 S / C = 3 1 0 1 0 0 0 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3 C / O C / O

1 0 6 S / C = 0 E S / C = 0 F 5 8 S / C = 3 I s o b u t y l e n e ) % ) 4 S / C = 3 (

% y ( t 6 S / C = 2 i

v y

i S / C = 1 t t i c v 3

S / C = 1 i e t l c

e S / C = 3 e s 4 l

e 6 S / C = 2 s 2 H 3 4 C C

S / C = 0 2 1 2 1 I s o b u t y r a l d e h y d e 0 0 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3

C / O C / O

Figure 7.3: Product distributions and temperatures from WGS after catalytic partial oxidation of isobutanol. represents S/C = 0, • S/C = 1, N S/C = 2 and S/C = 3. 7.4 Discussion 113

7 0

6 0 S / C = 3

) S / C = 2 5 0 % ( S / C = 1 n

o 4 0 i s r

e 3 0 v n

o S / C = 0 c

2 0 O

C 1 0

0 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2 1 . 3 C / O r a t i o

Figure 7.4: CO conversion in the water gas shift stage. represents S/C = 0, • S/C = 1, N S/C = 2 and S/C = 3. also resulted in reduced concentrations of isobutyraldehyde and isobutylene (Panel F) while propylene selectivities were mostly unchanged (Panel E). The CO conversion in the WGS stage increased with increasing amount of steam and was higher at lower C/O ratios as shown in Figure 7.4. The highest CO conversion was 66 % at C/O = 0.8 and S/C = 3. The WGS reaction was close to equilibrium ◦ (Kobs/Keq ≥ 0.75) at temperatures greater than 530 C. Increasing the amount of steam results in a more favorable equilibrium conversion, whereas decreasing C/O ratios led to higher temperatures which led to more favorable kinetics driving the reaction to equilibrium.

7.4 Discussion

7.4.1 CPO stage

The Rh-Ce based catalyst produced high yields of synthesis gas with > 99 % conversion of isobutanol and oxygen. At low C/O ratios (C/O ≤ 0.9), nearly equilibrium syngas products were obtained while non-equilibrium products became signiﬁcant at 7.4 Discussion 114 high C/O ratios. Similar results in terms of high syngas selectivities and conversions have been observed with Rh-Ce based catalysts with other oxygenated fuels.69,70 The formation of non-equilibrium products can be attributed mostly to homogeneous gas- phase chemistry; since they have low tendency to desorb from rhodium surface and their adsorption on the rhodium surface generally leads to syngas formation.85 In the C/O range of interest, the range of temperatures at a given S/C ratio was about 250-300 ◦C. CPO has been shown to occur in two stages: oxidation and reforming; > 99 % of the oxygen is consumed within ∼ 2mm.81 With increasing C/O ratios, conversion of the fuel starts to spread out with increasing C/O ratios as compared to mostly within the oxidation zone at low C/O ratios.70 The absence of oxygen in the reforming zone at high C/O ratios and lower temperatures results in slower surface kinetics and increased formation of homoegeneous non-equilibrium products from isobutanol. At lower C/O ratios closer to combustion stoichiometry, higher temperatures and faster kinetics on the catalyst surface increase yields to equilibrium products and reduce selectivities to non-equilibrium products. Steam addition reduced the reactor temperature and caused a signiﬁcant change in

H2, CO, CO2 and H2O selectivities due to increased WGS. The maximum in hydrogen selectivity shifts to a higher C/O ratio from S/C = 0 to S/C = 2 which may be due to the more favorable equilibrium of the WGS reaction at lower temperatures caused by steam addition.

7.4.2 CPO + WGS

After the WGS stage, the WGS reaction was equilibrated at temperatures greater than 530 ◦C , consistent with results previously reported by Wheeler et al.224 At S/C ratios of 2 and 3, for most C/O ratios, the CO conversion in the WGS stage was greater than 40 % thereby reducing the amount of CO in the exit stream to concentrations comparable to or lower than after an industrial high temperature shift catalyst. Also, selectivities to isobutylene and isobutyraldehyde reduced by almost half, indicating cleanup of hydrocarbons in the WGS stage over the Pt-Ce catalyst, likely due to the steam reforming of hydrocarbons. The reduction in concentrations of isobutylene and isobutyraldehyde as compared to propylene may be attributed to the higher steam reforming activity of larger hydrocarbons.225 In the WGS stage with steam addition, an increase in temperatures were observed with increasing S/C 7.4 Discussion 115

Figure 7.5: Thermodynamic analysis of the current system including photosynthesis, fermentation, CPO-WGS, hydrogen use in fuel cell and combustion. ratios in contrast with the CPO stage where increasing amounts of steam reduced temperatures. This increase in temperatures in the WGS stage may be attributed to the exothermic nature of the WGS reaction and the increasing conversion of CO with increasing S/C ratios.

Energetic Analysis of Isobutanol Reforming

Thermodynamic analysis of this proposed staged system is shown in Fig 7.5. The figure shows the integrated pathway wherein renewable hydrogen can be synthesized from CO2 and H2O. In photosynthesis, plants synthesize glucose in the presence of sunlight from CO2 and H2O. Glucose can then be fermented to isobutanol with CO2 and H2O as coproducts in a process which is slightly exothermic. By the integrated CPO and WGS process as described in this chapter, isobutanol can be converted to high purity hydrogen free from CO. This hydrogen can then be used to produce electical energy by combination with oxygen in a fuel cell, with water as a coproduct. Oxygen released during photosynthesis is partly used for CPO and the remainder in a fuel cell. Energy losses from fermentation and integrated CPO-WGS are less than 15 %, thereby making available most of the energy from glucose for use in a fuel cell. Isobutanol can also be used as a gasoline blend or substitute in which case the oxygen released from photosynthesis can be used to combust isobutanol to H2O and CO2. Typical energy efficiencies for fuel cells are about 60 % while that for combustion is 7.5 Conclusions 116 around 20 %.68 Thus, available energy from the current process is 60 % of 2176 ≈ 1300 kJ/mol while that from combustion is about 20 % of 2500 ≈ 500 kJ/mol. Thus, the current integrated system represents a process more than 2.5 times as efficient as combustion of isobutanol. This shows that converting isobutanol to hydrogen with the current process reprsents a more efficient use of energy than using it as a gasoline blend or substitute. With improving technologies for fuel cells, steam reforming may yield a slightly more energy efficient process than integrated CPO-WGS, however, the current system represents a more scalable and portable technology than steam reforming. Steam reforming is highly endothermic and requires external heating which is more economical for large-scale applications. CPO being exothermic, requires no external heat input and involves short startup and shutdown times making it ideal for small-scale or onboard applications.226

7.5 Conclusions

CPO of isobutanol was carried out in an autothermal millisecond contact time reactor with > 99% conversion of isobutanol over a 1 % Rh-1 % Ce catalyst. The H2 and CO yields could be adjusted by varying the fuel to oxygen ratio and also by adding steam. Steam addition promoted the WGS reaction and signiﬁcantly increased yields of H2. In addition, a WGS stage consisting of 1 % Pt-1 % Ce was added downstream over the CPO stage, which resulted in H2/CO ratio of ∼ 12 and a CO concentration of about 1.7-3%, with negligible yields of non-equilibrium products (≤ 1%). Thus, this integrated autothermal system producing high yields of hydrogen in residence times less than 100 ms represents an attractive option as an onboard reformer or other portable systems.

7.6 Acknowledgements

We acknowledge funding for this research provided by the Department of Defense (DOD) through Fuel Cell Energy, Inc. in Danbury, CT. We would like to thank Stephen Jolly (Fuel Cell Energy) for valuable discussions and Dr. Brian Michael for help with setup of experiments. CHAPTER EIGHT

SUMMARY AND FUTURE WORK

The chapters presented in this thesis provide an insight into the catalytic partial oxidation (CPO) of different renewable feedstocks. Chapter 2 shows the general trend of concentration and temperature gradients that are observed in a CPO reactor during methane and dimethyl ether (DME) CPO. The largest gradients in both concentration and temperatures were observed within the oxidation zone. Approximately 95 % of the oxygen was consumed within the first 2.2 mm of the catalyst and syngas formation was observed even within the oxidation zone. For upgrading lignocellulosic biomass to synthesis gas by catalytic gasification, biomass inorganics present a processing challenge. Chapters 3 and 4 provide a survey of the effects of common biomass inorganics on rhodium catalysts by two sets of experiments. In Chapter 3, individual inorganics were predosed on the catalyst and studied using steam methane reforming while in Chapter 4, inorganics were added to the ethanol feed and then onto the catalyst for CPO. The results from these chapters help compare severities of different inorganics and provide a platform for further investigation of inorganics with interesting chemistry and product distributions. Since potassium and phosphorus showed alkaline and acid chemistries with ethanol resulting in increased acetaldehyde and ethylene yields while maintaining autothermal operation during doping, their effects were studied in greater detail in Chapter 5. The results indicate that lignocellulosic biomass sources containing low concentrations of inorganics can be processed autothermally to a high selectivity thermodynamically equilibrated synthesis gas stream. In both pyrolysis and hydrolysis routes of biomass upgrading, alcohols represent

117 8.1 Spatial Proﬁles 118 an important class of functional groups found in intermediates in both the above routes.36,39 Butanol is being considered as a next generation biofuel due to its superior properties to ethanol.181,220 Chapter 6 compares the behavior of butanol isomers over diﬀerent noble metal catalysts during CPO. Chapter 7 demonstrates integration of CPO and WGS to produce high purity hydrogen stream from isobutanol for fuel cell applications.

8.1 Spatial Proﬁles

During CPO over noble metal catalysts, spatially resolved concentration and temperature measurements give an insight into the mechanism and nature of reactions taking place during autothermal operation. Homogeneous chemistry has prevented extension of this technique to heavier hydrocarbons and oxygenates. In Chapter 2, DME (a molecule with negligible homogeneous chemistry85,87) showed similar behavior as methane, with the presence of oxidation and reforming sections. For the most part, trends in temperatures, feed and product concentration proﬁles were similar for both species. However, this study of DME CPO was performed over just one catalyst at a particular ﬂow rate. Further experiments are necessary to gain a more detailed understanding of the behavior of oxygenates in a CPO reactor.

8.1.1 Examination of Mechanism of DME CPO

Role of Mass transfer Limitations

Horn et al. by spatially resolved measurements using the capillary sampling technique compared the length of the oxidation zone at diﬀerent C/O ratios for methane CPO and concluded that methane CPO is mass transfer limited over Rh catalysts and kinetically limited over Pt based catalysts.81 Similar experiments with DME CPO will help determine the role of mass transfer and reaction kinetics over diﬀerent catalysts.

Surface temperature Measurements

In Chapter 2, the thermocouple used measures the gas phase temperature and not surface temperatures. To measure surface temperatures during CPO, an optical pyrometer with a 45◦ tip has been used.82,227 By measuring the surface temperatures, 8.2 Inorganics 119 additional information regarding approach to equilibrium of steam reforming (Eq. 8.1) and water gas shift reactions (Eq. 8.2) can be measured along the length of different catalysts. The experimental values obtained through equations (8.3) and (8.4) can be compared with that calculated from thermodynamics to measure approach to equilibrium.

CH3OCH3 + H2O −→ 2 CO + 4 H2 (8.1)

CO + H2O −→ H2 + CO2 (8.2)

4 2 PH PCO K = 2 for steam reforming (8.3) exp,SR P P CH3OCH3 H2O

PH PCO K = 2 2 for water gas shift (8.4) eq,W GS P P CO H2O

Role of CO2 reforming

In Chapter 2, comparison of CO2 and CH4 concentration proﬁles revealed that during

DME CPO, evolution of CO2 concentration proﬁles is less steep than methane. Simi- larly, H2O proﬁles during methane CPO are more sharp than that during DME CPO.

During methane CPO over a Rh catalyst, no CO2 reforming was observed consistent 80–82 with previous results. This may be due to a more dominant role of CO2 reforming during DME CPO as compared to methane CPO. Measuring reaction kinetics of DME dry reforming and steam reforming over rhodium catalysts can help compare the relative rates of the two reactions and determine the extent of CO2 reforming of DME in short contact time reactors.

8.2 Inorganics

8.2.1 Extending Spatial Proﬁles to Inorganics

Spatially resolved concentration and temperature proﬁles can help in obtaining a better understanding of mechanism of interaction of inorganics with rhodium catalysts. This technique cannot be used with ethanol due to homogeneous gas phase 8.2 Inorganics 120

chemistry,85,87 but as shown in Chapter 2, can be used with methane or DME. An experimental strategy similar to that used in Chapter 5 can be employed - measuring baseline performance with undoped catalysts, doping with inorganics and measuring performance of inorganic doped catalysts. Though the results in Chapter 3 show the effect of different inorganics on reforming on rhodium catalysts, due to complex concentration and temperature gradients, the effect on reforming during autothermal CPO operation might be different. By spatial profile studies, the relative contributions of inorganics to the oxidation and reforming sections can be determined during autothermal operation.

Concentration proﬁles of inorganics

As shown in Chapter 2, large concentration and temperature gradients exist during operation of a CPO reactor. These gradients can influence the volatility and the nature of the inorganic species within the catalyst bed. For example, in the oxidation zone, it is likely that inorganics are present in the oxidized state while they may exist in the reduced state in the reforming section where oxygen is no longer present. The temperature gradients during CPO can also influence the concentration profile of inorganics during CPO. In the oxidation zone, where temperatures are highest, inorganics will have higher volatility compared to that in the reforming section. However, since the oxidation zone is present upstream of the reforming section, it may have a higher concentration of inorganics. Use of analytical techniques such as XPS and EDS will help in measuring concentrations and oxidation states of the inorganics along the length of the catalyst after operation with inorganics.

Nature of Inorganic Species during reaction

The characterization techniques used in this thesis and those described in the above section help determine state of the inorganics and their nature after reaction. Due to high temperatures within a CPO reactor, implementing in situ characterization techniques to detect surface species and nature of catalyst during operation has been diﬃcult. With developments in the ﬁeld of in situ characterization, there are ex- amples in the literature of in situ characterization techniques during methane CPO. For example, in situ IR (DRIFTS, FTIR),228,229 Raman,230 X-ray absorption98,99,231 based techniques have been reported in the literature to determine surface species 8.2 Inorganics 121 and state of the catalyst during methane CPO. By using such in situ characterization techniques, the nature of the inorganic surface species along the length of the catalyst and the molecular state in which the inorganics leave the catalyst surface during reaction can be determined, which may help in better catalyst design to minimize deactivation.

8.2.2 Techniques to Minimize Catalyst Deactivation

Chemical Regeneration

Chapter 5 shows the thermal regeneration of potassium-doped catalyst at higher temperatures, even in conjunction with phosphorus as monobasic potassium phosphate. Thermal regeneration may not always be feasible due to deactivation of the catalyst at high temperatures by sintering. Chapter 4 shows the regeneration of a Si-poisoned rhodium catalyst likely due to combination of silica with steam. Also in Chapter 4, the presence of less volatile inorganics such as Ca and Na in the effluent stream is most likely due to combination with species in the reactor to form more volatile compounds. To investigate the effect of different cofeeds such as H2O and CO2 on inorganic volatilization, experiments can be performed with inorganic-doped catalysts in presence of an inert gas such as nitrogen or argon in a heated isothermal system. The effect of cofeeding can be studied by weighing the catalyst before and after the cofeeding process or measuring concentrations by analytical techniques such as XPS or XRD. Steam addition does not represent a significant processing challenge for CPO as it has been shown that CPO reactors can operate autothermally with up to 50 wt % water in the feed.3

Altering Reactor Design

As shown in Chapter 4, with time, non-volatile inorganics such as Ca, Mg, and Na accumulate near the front face of the catalyst. Similar results have been reported by Dauenhauer et al. while processing aspen containing ∼ 0.5 % ash.2 As the ash layer accumulates, it gradually slows heat transfer from the oxidation zone to the top surface, preventing biomass pyrolysis and volatilization and gradually shutting down autothermal operation. Hence, to minimize accumulation of inorganics near the catalyst front face, periodic removal of accumulated ash using a moving bed type reactor can be implemented. Also, addition of a blank conductive ﬁlter which prevents 8.2 Inorganics 122 direct contact of ash with the catalyst at the same time maintaining heat transfer will minimize the rate at which inorganics reach the catalyst and prolong catalyst life. After removal of the inorganics accumulated near the front face of the catalyst during the regeneration stage in a moving bed conﬁguration, based on the results in Chapters 3, 4 and 5, partial regeneration of the catalyst may be carried out to restore catalytic activity by CPO with clean fuel.

Pretreatment of Feed

For gasiﬁcation, leaching is a commonly used pretreatment technique to reduce concentration of inorganics in the biomass feed.232 It involves washing the biomass with water to reduce the concentration of inorganics and can result in signiﬁcant reduction in the inorganic content of biomass, especially elements such as alkali metals sodium and potassium, chlorine, sulfur and phosphorus.233,234 BIBLIOGRAPHY

[1] J.R. Salge, B.J. Dreyer, P.J. Dauenhauer, and L.D. Schmidt. Renewable Hydrogen from Nonvolatile Fuels by Reactive Flash Volatilization. Science, 314(5800):801–804, 2006. [2] P.J. Dauenhauer, B.J. Dreyer, N.J. Degenstein, and L.D. Schmidt. Millisec- ond Reforming of Solid Biomass for Sustainable Fuels. Angewandte Chemie International Edition, 46(31):5864–5867, 2007. [3] J.L. Colby, P.J. Dauenhauer, and L.D. Schmidt. Millisecond autothermal steam reforming of cellulose for synthetic biofuels by reactive ﬂash volatilization. Green Chem., 10:773–783, 2008. [4] N. Armaroli and V. Balzani. The Future of Energy Supply: Challenges and Opportunities. Angewandte Chemie International Edition, 46(1-2):52–66, 2007. [5] R.E.H. Sims, R.N. Schock, A. Adegbululgbe, J. Fenhann, I. Konstantinaviciute, W. Moomaw, H.B. Nimir, B. Schlamadinger, J. Torres-Martnez, C. Turner, Y. Uchiyama, S.J.V. Vuori, N. Wamukonya, X. Zhang. Energy supply. In Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, pages 251–322. [B. Metz, O.R. Davidson, P.R. Bosch, R. Dave, L.A. Meyer (eds)], Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA., 2007. [6] British Petroleum. BP Statistical Review of World Energy 2011. British Petroleum, 2011. [7] International Energy Agency. Key World Energy Statistics 2011. World Wide Web, http://www.iea.org/textbase/nppdf/free/2011/key_world_ energy_stats.pdf, 2011. [8] International Energy Agency. World Energy Outlook 2010: Executive Summary. World Wide Web, http://www.iea.org/Textbase/npsum/ weo2010sum.pdf, 2011.

123 BIBLIOGRAPHY 124

[9] J.R. Bartels, M.B. Pate, and N.K. Olson. An economic survey of hydrogen production from conventional and alternative energy sources. International Journal of Hydrogen Energy, 35(16):8371 – 8384, 2010.

[10] R.A. Kerr. Do We Have the Energy for the Next Transition? Science, 329 (5993):780–781, 2010.

[11] D. Schindler. Tar sands need solid science. Nature, 468(7323):499–501, 2010.

[12] R.A. Kerr. Natural Gas From Shale Bursts Onto the Scene. Science, 328(5986): 1624–1626, 2010.

[13] C. Mooney. The truth about Fracking. Scientiﬁc American, 305(5):80–85, 2011.

[14] R.W. Howarth, A. Ingraﬀea, and T. Engelder. Natural gas: Should fracking stop? Nature, 477(7364):271–275, 2011.

[15] S.G. Osborn, A. Vengosh, N.R. Warner, and R.B. Jackson. Methane contamination of drinking water accompanying gas-well drilling and hydraulic fracturing. Proceedings of the National Academy of Sciences, 108(20):8172–8176, 2011.

[16] S.G. Osborn, A. Vengosh, N.R. Warner, and R.B. Jackson. Reply to Saba and Orzechowski and Schon: Methane contamination of drinking water accompanying gas-well drilling and hydraulic fracturing. Proceedings of the National Academy of Sciences, 108(37):E665–E666, 2011.

[17] T. Saba and M. Orzechowski. Lack of data to support a relationship between methane contamination of drinking water wells and hydraulic fracturing. Pro- ceedings of the National Academy of Sciences, 108(37):E663, 2011.

[18] S.C. Schon. Hydraulic fracturing not responsible for methane migration. Pro- ceedings of the National Academy of Sciences, 108(37):E664, 2011.

[19] OECD. Oecd factbook 2011-2012. doi: http://dx.doi.org/10.1787/ factbook-2011-en. URL /content/book/factbook-2011-en. [20] A. Goeppert, M. Czaun, G.K.S. Prakash, and G.A. Olah. Air as the renewable carbon source of the future: an overview of co2 capture from the atmosphere. Energy and Environmental Science, 5:7833–7853, 2012.

[21] P. Friedlingstein, R.A. Houghton, G. Marland, J. Hackler, T.A. Boden, T.J. Conway, J.G. Canadell, M.R. Raupach, P. Ciais, and C.L. Quere. Update on

CO2 emissions. Nature Geoscience, 3(12):811–812, 2010. [22] S. Solomon, G.-K. Plattner, R. Knutti, and P. Friedlingstein. Irreversible climate change due to carbon dioxide emissions. Proceedings of the National Academy of Sciences, 106(6):1704–1709, 2009. BIBLIOGRAPHY 125

[23] M.Z. Jacobson. Review of solutions to global warming, air pollution, and energy security. Energy and Environmental Science, 2:148–173, 2009. [24] A.V. Bridgwater. Renewable fuels and chemicals by thermal processing of biomass. Chemical Engineering Journal, 91(2 - 3):87 – 102, 2003. [25] A.J. Ragauskas, C.K. Williams, B.H. Davison, G. Britovsek, J. Cairney, C.A. Eckert, W.J. Frederick, J.P. Hallett, D.J. Leak, C.L. Liotta, J.R. Mielenz, R. Murphy, R. Templer, and T. Tschaplinski. The Path Forward for Biofuels and Biomaterials. Science, 311(5760):484–489, 2006. [26] J.N. Chheda, G.W. Huber, and J.A. Dumesic. Liquid-Phase Catalytic Processing of Biomass-Derived Oxygenated Hydrocarbons to Fuels and Chemicals. Angewandte Chemie International Edition, 46(38):7164–7183, 2007. [27] T.E. Bull. Biomass in the Energy Picture. Science, 285(5431):1209, 1999. [28] R.D. Perlack, L.L. Wright, A.F. Turhollow, R.L. Graham, B.J. Stokes, and D.C. Erbach. Biomass as feedstock for a bioenergy and bioproducts industry: the tecnical feasibility of a billion-ton annual supply. Technical report, US Department of Energy and Department of Agriculture, DOE/GO-102005-2135, ORNL/TM-2005/66, 2005. [29] R. Rinaldi and F. Schuth. Design of solid catalysts for the conversion of biomass. Energy and Environmental Science, 2:610–626, 2009. [30] L. Petrus and M.A. Noordermeer. Biomass to biofuels, a chemical perspective. Green Chemistry, 8:861–867, 2006. [31] A. Demirbas. Biomass resource facilities and biomass conversion processing for fuels and chemicals. Energy Conversion and Management, 42(11):1357 – 1378, 2001. [32] S. Czernik and A.V. Bridgwater. Overview of Applications of Biomass Fast Pyrolysis Oil. Energy and Fuels, 18(2):590–598, 2004. [33] D. Mohan, C.U. Pittman, and P.H. Steele. Pyrolysis of Wood/Biomass for Bio-oil: A Critical Review. Energy and Fuels, 20(3):848–889, 2006. [34] P. Kumar, D.M. Barrett, M.J. Delwiche, and P. Stroeve. Methods for Pretreatment of Lignocellulosic Biomass for Eﬃcient Hydrolysis and Biofuel Production. Industrial and Engineering Chemistry Research, 48(8):3713–3729, 2009. [35] M.E. Himmel, S.-Y. Ding, D.K. Johnson, W.S. Adney, M.R. Nimlos, J.W. Brady, and T.D. Foust. Biomass Recalcitrance: Engineering Plants and Enzymes for Biofuels Production. Science, 315(5813):804–807, 2007. BIBLIOGRAPHY 126

[36] G.W. Huber, S. Iborra, and A. Corma. Synthesis of Transportation Fuels from Biomass: Chemistry, Catalysts, and Engineering. Chemical Reviews, 106(9): 4044–4098, 2006. [37] K. Raveendran, A. Ganesh, and K.C. Khilar. Influence of mineral matter on biomass pyrolysis characteristics. Fuel, 74(12):1812 – 1822, 1995. [38] C.D. Blasi. Modeling chemical and physical processes of wood and biomass pyrolysis. Progress in Energy and Combustion Science, 34(1):47 – 90, 2008. [39] J.C. Serrano-Ruiz and J.A. Dumesic. Catalytic routes for the conversion of biomass into liquid hydrocarbon transportation fuels. Energy and Environmen- tal Science, 4:83–99, 2011. [40] E.R. Riegel and J.A. Kent. Riegel’s Handbook of Industrial Chemistry. Kluwer Academic/Plenum Publishers: New York, NY, tenth edition, 2003. [41] A.V. Bridgwater. The technical and economic feasibility of biomass gasification for power generation. Fuel, 74(5):631 – 653, 1995. [42] A.V. Bridgwater and G.V.C. Peacocke. Fast pyrolysis processes for biomass. Renewable and Sustainable Energy Reviews, 4(1):1 – 73, 2000. [43] D. Rennard, R. French, S. Czernik, T. Josephson, and L. Schmidt. Production of synthesis gas by partial oxidation and steam reforming of biomass pyrolysis oils. International Journal of Hydrogen Energy, 35(9):4048 – 4059, 2010. [44] D.B. Dadyburjor, Z. Liu, and B.H. Davis. Coal Liquefaction. Kirk-Othmer Encyclopedia of Chemical Technology, John Wiley and Sons, Inc., 2000. ISBN 9780471238966. doi: 10.1002/0471238961.1209172102011204.a01.pub3. URL http://dx.doi.org/10.1002/0471238961.1209172102011204.a01.pub3. [45] T. Kaneko, F. Derbyshire, E. Makino, D. Gray, and M. Tamura. Coal Lique- faction. Ullmann’s Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH and Co. KGaA, 2000. ISBN 9783527306732. doi: 10.1002/14356007. a07\ 197. URL http://dx.doi.org/10.1002/14356007.a07\_197. [46] R. Luque, A.R. de la Osa, J.M. Campelo, A.A. Romero, J.L. Valverde, and P. Sanchez. Design and development of catalysts for Biomass-To-Liquid-Fischer- Tropsch (BTL-FT) processes for biofuels production. Energy and Environmen- tal Science, 5:5186–5202, 2012. [47] A. English, J. Rovner, J. Brown, S. Davies, and Updated by Staff. Methanol. Kirk-Othmer Encyclopedia of Chemical Technology, John Wiley and Sons, Inc., 2000. ISBN 9780471238966. doi: 10.1002/0471238961.1305200805140712.a01. pub2. URL http://dx.doi.org/10.1002/0471238961.1305200805140712. a01.pub2. BIBLIOGRAPHY 127

[48] E. Fiedler, G. Grossmann, D.B. Kersebohm, G. Weiss, and C. Witte. Methanol. Ullmann’s Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH and Co. KGaA, 2000. ISBN 9783527306732. doi: 10.1002/14356007.a16\ 465.pub2. URL http://dx.doi.org/10.1002/14356007.a16\_465.pub2. [49] G.A. Olah. Beyond Oil and Gas: The Methanol Economy. Angewandte Chemie International Edition, 44(18):2636–2639, 2005. [50] M. Xu, J.H. Lunsford, D.W. Goodman, and A. Bhattacharyya. Synthesis of dimethyl ether (DME) from methanol over solid-acid catalysts. Applied Catal- ysis A: General, 149(2):289 – 301, 1997. [51] F.J. Keil. Methanol-to-hydrocarbons: process technology. Microporous and Mesoporous Materials, 29(1-2):49 – 66, 1999. [52] J. Fei, Z. Hou, B. Zhu, H. Lou, and X. Zheng. Synthesis of dimethyl ether (DME) on modiﬁed HY zeolite and modiﬁed HY zeolite-supported Cu-Mn-Zn catalysts. Applied Catalysis A: General, 304:49 – 54, 2006. [53] T. Mokrani and M. Scurrell. Gas Conversion to Liquid Fuels and Chemicals: The Methanol Route-Catalysis and Processes Development. Catalysis Reviews, 51(1):1–145, 2009. [54] T.A. Semelsberger, R.L. Borup, and H.L. Greene. Dimethyl ether (DME) as an alternative fuel. Journal of Power Sources, 156(2):497 – 511, 2006. [55] M.Y. Kim, S.H. Bang, and C.S. Lee. Experimental Investigation of Spray and Combustion Characteristics of Dimethyl Ether in a Common-Rail Diesel Engine. Energy and Fuels, 21(2):793–800, 2007. [56] G. Marban and T. Valdes-Solis. Towards the hydrogen economy? International Journal of Hydrogen Energy, 32(12):1625 – 1637, 2007. [57] L. Barreto, A. Makihira, and K. Riahi. The hydrogen economy in the 21st century: a sustainable development scenario. International Journal of Hydrogen Energy, 28(3):267 – 284, 2003. [58] J. Graetz. New approaches to hydrogen storage. Chemical Society Reviews, 38: 73–82, 2009. [59] C. Song. Fuel processing for low-temperature and high-temperature fuel cells: Challenges, and opportunities for sustainable development in the 21st century. Catalysis Today, 77(1 - 2):17 – 49, 2002. [60] J. Zhang, Z. Xie, J. Zhang, Y. Tang, C. Song, T. Navessin, Z. Shi, D. Song, H. Wang, D.P. Wilkinson, Z.-S. Liu, and S. Holdcroft. High temperature PEM fuel cells. Journal of Power Sources, 160(2):872 – 891, 2006. BIBLIOGRAPHY 128

[61] D.A. Hickman and L.D. Schmidt. Production of Syngas by Direct Catalytic Oxidation of Methane. Science, 259(5093):343–346, 1993.

[62] M. Huﬀ, P.M. Torniainen, and L.D. Schmidt. Partial oxidation of alkanes over noble metal coated monoliths. Catalysis Today, 21(1):113 – 128, 1994.

[63] R.P. O’Connor, E.J. Klein, and L.D. Schmidt. High yields of synthesis gas by millisecond partial oxidation of higher hydrocarbons. Catalysis Letters, 70: 99–107, 2000.

[64] J.J. Krummenacher, K.N. West, and L.D. Schmidt. Catalytic partial oxidation of higher hydrocarbons at millisecond contact times: decane, hexadecane, and diesel fuel. Journal of Catalysis, 215(2):332 – 343, 2003.

[65] G.J. Panuccio, B.J. Dreyer, and L.D. Schmidt. A comparison of the catalytic

partial oxidation of C1 to C16 normal paraﬃns. AIChE Journal, 53(1):187–195, 2007.

[66] R. Subramanian, G.J. Panuccio, J.J. Krummenacher, I.C. Lee, and L.D. Schmidt. Catalytic partial oxidation of higher hydrocarbons: reactivities and selectivities of mixtures. Chemical Engineering Science, 59(22 - 23):5501 – 5507, 2004.

[67] R. Subramanian and L.D. Schmidt. Renewable Oleﬁns from Biodiesel by Autothermal Reforming. Angewandte Chemie International Edition, 44(2):302– 305, 2005.

[68] G.A. Deluga, J.R. Salge, L.D. Schmidt, and X.E. Verykios. Renewable Hydrogen from Ethanol by Autothermal Reforming. Science, 303(5660):993– 997, 2004.

[69] J.R. Salge, G.A. Deluga, and L.D. Schmidt. Catalytic partial oxidation of ethanol over noble metal catalysts. Journal of Catalysis, 235(1):69 – 78, 2005.

[70] P.J. Dauenhauer, J.R. Salge, and L.D. Schmidt. Renewable hydrogen by autothermal steam reforming of volatile carbohydrates. Journal of Catalysis, 244 (2):238 – 247, 2006.

[71] E.C. Wanat, B.Suman, and L.D. Schmidt. Partial oxidation of alcohols to produce hydrogen and chemicals in millisecond-contact time reactors. Journal of Catalysis, 235(1):18 – 27, 2005.

[72] D.C. Rennard, P.J. Dauenhauer, S.A. Tupy, and L.D. Schmidt. Autothermal Catalytic Partial Oxidation of Bio-Oil Functional Groups: Esters and Acids. Energy and Fuels, 22(2):1318–1327, 2008. BIBLIOGRAPHY 129

[73] D.C. Rennard, J.S. Kruger, and L.D. Schmidt. Autothermal Catalytic Partial Oxidation of Glycerol to Syngas and to Non-Equilibrium Products. Chem- SusChem, 2(1):89–98, 2009.

[74] D.C. Rennard, J.S. Kruger, B.C. Michael, and L.D. Schmidt. Long-Time Behavior of the Catalytic Partial Oxidation of Glycerol in an Autothermal Reactor. Industrial and Engineering Chemistry Research, 49(18):8424–8432, 2010.

[75] J.A. Moulijn, A.E. van Diepen, and F. Kapteijn. Catalyst deactivation: is it predictable?: What to do? Applied Catalysis A: General, 212(1 - 2):3 – 16, 2001.

[76] G. Busca. Acid Catalysts in Industrial Hydrocarbon Chemistry. Chemical Reviews, 107(11):5366–5410, 2007.

[77] C.H. Bartholomew. Mechanisms of catalyst deactivation. Applied Catalysis A: General, 212(1 - 2):17 – 60, 2001.

[78] P. Forzatti and L. Lietti. Catalyst deactivation. Catalysis Today, 52(2 - 3):165 – 181, 1999.

[79] R.W. Mccabe, R.K. Usmen, K. Ober, and H.S. Gandhi. The Eﬀect of Alumina Phase-Structure on the Dispersion of Rhodium/Alumina Catalysts. Journal of Catalysis, 151(2):385 – 393, 1995.

[80] R. Horn, K.A. Williams, N.J. Degenstein, and L.D. Schmidt. Syngas by catalytic partial oxidation of methane on rhodium: Mechanistic conclusions from spatially resolved measurements and numerical simulations. Journal of Cataly- sis, 242(1):92 – 102, 2006.

[81] R. Horn, K.A. Williams, N.J. Degenstein, A. Bitsch-Larsen, D.D. Nogare, S.A. Tupy, and L.D. Schmidt. Methane catalytic partial oxidation on autothermal Rh and Pt foam catalysts: Oxidation and reforming zones, transport eﬀects, and approach to thermodynamic equilibrium. Journal of Catalysis, 249(2):380 – 393, 2007.

[82] A. Donazzi, B.C. Michael, and L.D. Schmidt. Chemical and geometric eﬀects of

Ce and washcoat addition on catalytic partial oxidation of CH4 on Rh probed by spatially resolved measurements. Journal of Catalysis, 260(2):270 – 275, 2008.

[83] A. Bitsch-Larsen, R. Horn, and L.D. Schmidt. Catalytic partial oxidation of methane on rhodium and platinum: Spatial proﬁles at elevated pressure. Ap- plied Catalysis A: General, 348(2):165 – 172, 2008. BIBLIOGRAPHY 130

[84] C.M. Balonek, J.L. Colby, and L.D. Schmidt. Millisecond catalytic reforming of monoaromatics over noble metals. AIChE Journal, 56(4):979–988, 2010.

[85] J.S. Kruger, D.C. Rennard, T.R. Josephson, and L.D. Schmidt. Eﬀect of Functional Groups on Autothermal Partial Oxidation of Bio-oil. part 1: Role of Catalyst Surface and Molecular Oxygen. Energy and Fuels, 25(7):3157–3171, 2011.

[86] B.C. Michael, D.N. Nare, and L.D. Schmidt. Catalytic partial oxidation of ethane to ethylene and syngas over Rh and Pt coated monoliths: Spatial proﬁles of temperature and composition. Chemical Engineering Science, 65(12):3893 – 3902, 2010.

[87] J.S. Kruger, D.C. Rennard, T.R. Josephson, and L.D. Schmidt. Eﬀect of Functional Groups on Autothermal Partial Oxidation of Bio-oil. part 2: Role of homogeneous and support-mediated reactions. Energy and Fuels, 25(7):3172– 3185, 2011.

[88] K. Takeishi and H. Suzuki. Steam reforming of dimethyl ether. Applied Catalysis A: General, 260(1):111 – 117, 2004.

[89] V.V. Galvita, G.L. Semin, V.D. Belyaev, T.M. Yurieva, and V.A. Sobyanin. Production of hydrogen from dimethyl ether. Applied Catalysis A: General, 216(1-2):85 – 90, 2001.

[90] T. Kawabata, H. Matsuoka, T. Shishido, D. Li, Y. Tian, T. Sano, and K. Takehira. Steam reforming of dimethyl ether over ZSM-5 coupled with

Cu/ZnO/Al2O3 catalyst prepared by homogeneous precipitation. Applied Catal- ysis A: General, 308(0):82 – 90, 2006.

[91] A.P.E. York, T. Xiao, and M.L.H. Green. Brief Overview of the Partial Oxidation of Methane to Synthesis Gas. Topics in Catalysis, 22:345–358, 2003. ISSN 1022-5528.

[92] A.P.E. York, T. Xiao, M.L.H. Green, and J.B. Claridge. Methane Oxyforming for Synthesis Gas Production. Catalysis Reviews, 49(4):511–560, 2007.

[93] B.C. Enger, R. Lodeng, and A. Holmen. A review of catalytic partial oxidation of methane to synthesis gas with emphasis on reaction mechanisms over transition metal catalysts. Applied Catalysis A: General, 346(1 - 2):1 – 27, 2008.

[94] G. Veser and J. Frauhammer. Modelling steady state and ignition during catalytic methane oxidation in a monolith reactor. Chemical Engineering Science, 55(12):2271 – 2286, 2000. BIBLIOGRAPHY 131

[95] M. Maestri, D. Vlachos, A. Beretta, P. Forzatti, G. Groppi, and E. Tronconi.

Dominant Reaction Pathways in the Catalytic Partial Oxidation of CH4 on Rh. Topics in Catalysis, 52:1983–1988, 2009.

[96] A. Donazzi, M. Maestri, B.C. Michael, A. Beretta, P. Forzatti, G. Groppi, E. Tronconi, L.D. Schmidt, and D.G. Vlachos. Microkinetic modeling of spa-

tially resolved autothermal CH4 catalytic partial oxidation experiments over Rh-coated foams. Journal of Catalysis, 275(2):270 – 279, 2010.

[97] D.D. Nogare, N.J. Degenstein, R. Horn, P. Canu, and L.D. Schmidt. Modeling spatially resolved data of methane catalytic partial oxidation on Rh foam catalyst at diﬀerent inlet compositions and ﬂowrates. Journal of Catalysis, 277(2): 134 – 148, 2011.

[98] J. Grunwaldt and A. Baiker. Axial variation of the oxidation state of Pt-

Rh/Al2O3 during partial methane oxidation in a ﬁxed-bed reactor: An in situ X-ray absorption spectroscopy study. Catalysis Letters, 99:5–12, 2005. ISSN 1011-372X.

[99] J.-D. Grunwaldt, S. Hannemann, C.G. Schroer, and A. Baiker. 2d-Mapping of the Catalyst Structure Inside a Catalytic Microreactor at Work:? Partial

Oxidation of Methane over Rh/Al2O3. The Journal of Physical Chemistry B, 110(17):8674–8680, 2006.

[100] A.S. Bodke, S.S. Bharadwaj, and L.D. Schmidt. The Eﬀect of Ceramic Supports on Partial Oxidation of Hydrocarbons over Noble Metal Coated Monoliths. Journal of Catalysis, 179(1):138 – 149, 1998.

[101] Z. Ma, Q.-Z. Jiang, X. Wang, W.-G. Zhang, and Z.-F. Ma. CO2 reforming of dimethyl ether over Ni/γ-Al2O3 catalyst. Catalysis Communications, 17:49 – 53, 2012.

[102] L. Bugyi and F. Solymosi. Adsorption and dissociation of dimethyl ether on clean and oxygen-dosed Rh(111). Surface Science, 385(2-3):365 – 375, 1997.

[103] S.E. Koonin. Getting Serious About Biofuels. Science, 311(5760):435, 2006.

[104] A.A.T. Masia, B.J.P. Buhre, R.P. Gupta, and T.F. Wall. Characterising ash of biomass and waste. Fuel Processing Technology, 88(11-12):1071 – 1081, 2007.

[105] A.P. Murray and T.S. Snyder. Steam-methane reformer kinetic computer model with heat transfer and geometry options. Industrial and Engineering Chemistry Process Design and Development, 24(2):286–294, 1985. BIBLIOGRAPHY 132

[106] G. Jones, J.G. Jakobsen, S.S. Shim, J. Kleis, M.P. Andersson, J. Rossmeisl, F. Abild-Pedersen, T. Bligaard, S. Helveg, B. Hinnemann, J.R. Rostrup-Nielsen, I. Chorkendorﬀ, J. Sehested, and J.K. Norskov. First principles calculations and experimental insight into methane steam reforming over transition metal catalysts. Journal of Catalysis, 259(1):147 – 160, 2008.

[107] M. Bizzi, L. Basini, G. Saracco, and V. Specchia. Short contact time catalytic partial oxidation of methane: analysis of transport phenomena eﬀects. Chemical Engineering Journal, 90(1-2):97 – 106, 2002.

[108] R. Chakrabarti, S.A. Tupy, and L.D. Schmidt. Eﬀects of Biomass Inorganics on Rhodium Catalysts: II. Autothermal Reforming of Ethanol. Energy and Fuels, 25(10):4763–4769, 2011.

[109] Stanjan Chemical Equilibrium Calculation, 2012. URL http://navier.engr. colostate.edu/tools/equil.html.

[110] Y. Choi and H.G. Stenger. Water gas shift reaction kinetics and reactor modeling for fuel cell grade hydrogen. Journal of Power Sources, 124(2):432 – 439, 2003.

[111] M.D. Judd, B.A. Plunkett, and M.I. Pope. The thermal decomposition of calcium, sodium, silver and copper(II) acetates. Journal of Thermal Analysis and Calorimetry, 6:555–563, 1974.

[112] M. Afzal, P.K. Butt, and H. Ahmad. Kinetics of thermal decomposition of metal acetates. Journal of Thermal Analysis and Calorimetry, 37:1015–1023, 1991.

[113] M. Gabor and L. Poppl. Thermoanalytical investigations in diﬀerent controlled atmospheres. Journal of Thermal Analysis and Calorimetry, 11:231–240, 1977.

[114] C.H. Bartholomew, P.K. Agrawal, and J.R. Katzer. Sulfur Poisoning of Metals. volume 31 of Advances in Catalysis, pages 135 – 242. Academic Press, 1982. doi: 10.1016/S0360-0564(08)60454-X. URL http://www.sciencedirect.com/ science/article/pii/S036005640860454X.

[115] T. Wilke, X. Gao, C.G. Takoudis, and M.J. Weaver. Surface-enhanced Raman spectroscopy at transition metal-gas interfaces: adsorption and reactions of sulfur dioxide on platinum-, rhodium-, and ruthenium-coated gold. Journal of Catalysis, 130(1):62 – 75, 1991.

[116] S. Cimino, R. Torbati, L. Lisi, and G. Russo. Sulphur inhibition on the catalytic partial oxidation of methane over Rh-based monolith catalysts. Applied Catalysis A: General, 360(1):43 – 49, 2009. BIBLIOGRAPHY 133

[117] B. McAllister and P. Hu. A density functional theory study of sulfur poisoning. Journal of Chemical Physics, 122(8), 2005.

[118] J. Barbier Jr. and D. Duprez. Steam eﬀects in three-way catalysis. Applied Catalysis B: Environmental, 4(2-3):105 – 140, 1994.

[119] C. Ratnasamy and J. Wagner. Water Gas Shift Catalysis. Catalysis Reviews, 51:325–440(116), July 2009.

[120] S. Peucheret, M. Feaviour, and S. Golunski. Exhaust-gas reforming using pre- cious metal catalysts. Applied Catalysis B: Environmental, 65(3-4):201 – 206, 2006.

[121] R. Torbati, S. Cimino, L. Lisi, and G. Russo. The Eﬀect of Support on Sulphur Tolerance of Rh Based Catalysts for Methane Partial Oxidation. Catalysis Letters, 127:260–269, 2009.

[122] N.D. Lang, S. Holloway, and J.K. Norskov. Electrostatic adsorbate-adsorbate interactions: The poisoning and promotion of the molecular adsorption reaction. Surface Science, 150(1):24 – 38, 1985.

[123] D. Duprez, K. Fadili, and J. Barbier. Regeneration of Nickel Catalysts Deactivated by Filamentous Carbon. Industrial and Engineering Chemistry Research, 36(8):3180–3187, 1997.

[124] B.D. Gould, X. Chen, and J.W. Schwank. n-Dodecane reforming over nickel- based monolith catalysts: Deactivation and carbon deposition. Applied Catal- ysis A: General, 334(1-2):277 – 290, 2008.

[125] J.-W. Snoeck, G.F. Froment, and M. Fowles. Filamentous Carbon Formation and Gasiﬁcation: Thermodynamics, Driving Force, Nucleation, and Steady- State Growth. Journal of Catalysis, 169(1):240 – 249, 1997.

[126] J.R. Rostrup-Nielsen and I. Alstrup. Innovation and science in the process industry: Steam reforming and hydrogenolysis. Catalysis Today, 53(3):311 – 316, 1999.

[127] S.Y. Christou, H. Birgersson, J.L.G. Fierro, and A.M. Efstathiou. Reactivation of an Aged Commercial Three-Way Catalyst by Oxalic and Citric Acid Washing. Environmental Science and Technology, 40(6):2030–2036, 2006.

[128] W.√ Liu,√ K.C. Wong, and K.A.R. Mitchell. Novel Surface Structure: Rh(111)- ( 7× 7)R19.1◦-P. Journal of the American Chemical Society, 117(49):12344– 12345, 1995. BIBLIOGRAPHY 134

[129] S. Romet, M.F. Couturier, and T.K. Whidden. Modeling of Silicon Dioxide Chemical Vapor Deposition from Tetraethoxysilane and Ozone. Journal of The Electrochemical Society, 148(2):G82–G90, 2001.

[130] M.-H. Kim, H.-X. Li, and M.E. Davis. Synthesis of zeolites by water-organic vapor-phase transport. Microporous Materials, 1(3):191 – 200, 1993.

[131] S.A. Greenberg. Thermodynamic Functions for the Solution of Silica in Water. The Journal of Physical Chemistry, 61(2):196–197, 1957.

[132] W.D. Mross. Alkali Doping in Heterogeneous Catalysis. Catalysis Reviews- Science and Engineering, 25(4):591–637, 1983.

[133] J.R. Rostrup-Nielsen and L.J. Christiansen. Internal steam reforming in fuel cells and alkali poisoning. Applied Catalysis A: General, 126(2):381 – 390, 1995.

[134] T.V. Choudhary and D.W. Goodman. Stepwise methane steam reforming: a route to CO-free hydrogen. Catalysis Letters, 59:93–94, 1999.

[135] D.M. Alonso, J.Q. Bond, and J.A. Dumesic. Catalytic conversion of biomass to biofuels. Green Chemistry, 12:1493–1513, 2010.

[136] I. Obernberger, F. Biedermann, W. Widmann, and R. Riedl. Concentrations of inorganic elements in biomass fuels and recovery in the diﬀerent ash fractions. Biomass and Bioenergy, 12(3):211 – 224, 1997.

[137] Q. Cao, K.-C. Xie, Y.-K. Lv, and W.-R. Bao. Process eﬀects on activated carbon with large speciﬁc surface area from corn cob. Bioresource Technology, 97(1):110 – 115, 2006.

[138] J. Yanik, C. Kornmayer, M. Saglam, and M. Yuksel. Fast pyrolysis of agricultural wastes: Characterization of pyrolysis products. Fuel Processing Technol- ogy, 88(10):942 – 947, 2007.

[139] S. Xiong, J. Burvall, H. Orberg, G. Kalen, M. Thyrel, M. Ohman, and D. Bostrom. Slagging Characteristics during Combustion of Corn Stovers with and without Kaolin and Calcite. Energy and Fuels, 22(5):3465–3470, 2008.

[140] S. Mani, L.G. Tabil, and S. Sokhansanj. Grinding performance and physical properties of wheat and barley straws, corn stover and switchgrass. Biomass and Bioenergy, 27(4):339 – 352, 2004.

[141] S. Flint-Garcia, A. Bodnar, and M. Scott. Wide variability in kernel composition, seed characteristics, and zein proﬁles among diverse maize inbreds, lan- draces, and teosinte. TAG Theoretical and Applied Genetics, 119:1129–1142, 2009. ISSN 0040-5752. BIBLIOGRAPHY 135

[142] H. Mansilla, R. Garcia, J. Tapia, H. Duran, and S. Urzua. Chemical characterization of Chilean hardwoods. Wood Science and Technology, 25:145–149, 1991. ISSN 0043-7719. [143] A.B. Anderson. Chemistry of Western Pines. Industrial and Engineering Chem- istry, 36(7):662–663, 1944. [144] M. Larsson, M. Galbe, and G. Zacchi. Recirculation of process water in the production of ethanol from softwood. Bioresource Technology, 60(2):143 – 151, 1997. [145] D.C. Dayton, R.J. French, and T.A. Milne. Direct Observation of Alkali Vapor Release during Biomass Combustion and Gasification. 1. Application of Molecular Beam/Mass Spectrometry to Switchgrass Combustion. Energy and Fuels, 9(5):855–865, 1995. [146] R. Lemus, E.C. Brummer, K.J. Moore, N.E. Molstad, C.L. Burras, and M.F. Barker. Biomass yield and quality of 20 switchgrass populations in southern Iowa, USA. Biomass and Bioenergy, 23(6):433 – 442, 2002. [147] S. Polat and Y. Ozogul. Biochemical composition of some red and brown macro algae from the Northeastern Mediterranean Sea. International Journal of Food Sciences and Nutrition, 59(7-8):566–572, 2008. [148] H.Y. Lee and L.E. Erickson. Characterization of algae using regularities. Biotechnology and Bioengineering, 26(7):758–763, 1984. ISSN 1097-0290. [149] M.-H. Rei, S.J. Yang, and C.H. Hong. Catalytic gasification of rice hull and other biomass. The general effect of catalyst. Agricultural Wastes, 18(4):269 – 281, 1986. [150] N.A. Sanchez-Flores, G. Pacheco-Malagan, P. Perez-Romo, H. Armendariz, M. de L. Guzman-Castillo, J.M. Saniger, and J.J. Fripiat. Mesoporous silica from rice hull ash. Journal of Chemical Technology and Biotechnology, 82 (7):614–619, 2007. [151] K.R. Smith, J.M. Veranth, J.S. Lighty, and A.E. Aust. Mobilization of Iron from Coal Fly Ash Was Dependent upon the Particle Size and the Source of Coal. Chemical Research in Toxicology, 11(12):1494–1500, 1998. [152] M. Ikeda, H. Makino, H. Morinaga, K. Higashiyama, and Y. Kozai. Emission characteristics of NOx and unburned carbon in fly ash during combustion of blends of bituminous/sub-bituminous coals. Fuel, 82(15-17):1851 – 1857, 2003. [153] J.-R. Qiu, F. Li, Y. Zheng, C.-G. Zheng, and H.-C. Zhou. The influences of mineral behaviour on blended coal ash fusion characteristics. Fuel, 78(8):963 – 969, 1999. BIBLIOGRAPHY 136

[154] X. Hou and X. Yan. Study on the concentration and seasonal variation of inorganic elements in 35 species of marine algae. Science of The Total Environment, 222(3):141 – 156, 1998. [155] R. Fahmi, A.V. Bridgwater, L.I. Darvell, J.M. Jones, N. Yates, S. Thain, and I.S. Donnison. The effect of alkali metals on combustion and pyrolysis of Lolium and Festuca grasses, switchgrass and willow. Fuel, 86(10-11):1560 – 1569, 2007. [156] R. Chakrabarti, J.L. Colby, and L.D. Schmidt. Effects of biomass inorganics on rhodium catalysts: I. Steam methane reforming. Applied Catalysis B: Environmental, 107(1-2):88 – 94, 2011. [157] D.D. Nogare, N.J. Degenstein, R. Horn, P. Canu, and L.D. Schmidt. Modeling spatially resolved profiles of methane partial oxidation on a Rh foam catalyst with detailed chemistry. Journal of Catalysis, 258(1):131 – 142, 2008. [158] K. Arnby, M. Rahmani, M. Sanati, N. Cruise, A.A. Carlsson, and M. Skoglundh.

Characterization of Pt/γ-Al2O3 catalysts deactivated by hexamethyldisiloxane. Applied Catalysis B: Environmental, 54(1):1 – 7, 2004. [159] L. Colin, A. Cassuto, J.J. Ehrhardt, M.F. Ruiz-Lopez, and D. Jamois. Adsorp- tion and decomposition of hexamethyldisiloxane on platinum: an XPS, UPS and TDS study. Applied Surface Science, 99(3):245 – 254, 1996. [160] P. Das, S. Prasad, and D. Kunzru. Organophosphorus compounds as coke in- hibitors during naphtha pyrolysis. Eﬀect of benzyl diethyl phosphite and triph- enylphosphine sulﬁde. Industrial and Engineering Chemistry Research, 31(9): 2251–2255, 1992. [161] E. Turpeinen, R. Raudaskoski, E. Pongracz, and R.L. Keiski. Thermodynamic analysis of conversion of alternative hydrocarbon-based feedstocks to hydrogen. International Journal of Hydrogen Energy, 33(22):6635 – 6643, 2008. [162] T.S. Christensen. Adiabatic prereforming of hydrocarbons - an important step in syngas production. Applied Catalysis A: General, 138(2):285 – 309, 1996. [163] K. Haga, Y. Shiratori, Y. Nojiri, K. Ito, and K. Sasaki. Phosphorus Poisoning of Ni-Cermet Anodes in Solid Oxide Fuel Cells. Journal of The Electrochemical Society, 157(11):B1693–B1700, 2010. [164] R.J. Chimentao, J.E. Herrera, J.H. Kwak, F. Medina, Y. Wang, and C.H.F. Peden. Oxidation of ethanol to acetaldehyde over Na-promoted vanadium oxide catalysts. Applied Catalysis A: General, 332(2):263 – 272, 2007.

[165] P. Panagiotopoulou and D.I. Kondarides. Eﬀects of alkali promotion of TiO2 on the chemisorptive properties and water gas shift activity of supported noble metal catalysts. Journal of Catalysis, 267(1):57 – 66, 2009. BIBLIOGRAPHY 137

[166] P. Biswas and D. Kunzru. Steam reforming of ethanol on Ni-CeO2-ZrO2 catalysts: Effect of doping with copper, cobalt and calcium. Catalysis Letters, 118 (1-2):36–49, 2007. [167] H. Song and U.S. Ozkan. Changing the Oxygen Mobility in Co/Ceria Catalysts by Ca Incorporation: Implications for Ethanol Steam Reforming. The Journal of Physical Chemistry A, 114(11):3796–3801, 2010. [168] D. Nutalapati, R. Gupta, B. Moghtaderi, and T.F. Wall. Assessing slagging and fouling during biomass combustion: A thermodynamic approach allowing for alkali/ash reactions. Fuel Processing Technology, 88(11-12):1044 – 1052, 2007. [169] C.D. Blasi. Modeling chemical and physical processes of wood and biomass pyrolysis. Progress in Energy and Combustion Science, 34(1):47 – 90, 2008. [170] D.C. Dayton, B.M. Jenkins, S.Q. Turn, R.R. Bakker, R.B. Williams, D. Belle- Oudry, and L.M. Hill. Release of Inorganic Constituents from Leached Biomass during Thermal Conversion. Energy and Fuels, 13(4):860–870, 1999. [171] L.L. Baxter. Ash deposition during biomass and coal combustion: A mechanistic approach. Biomass and Bioenergy, 4(2):85 – 102, 1993. [172] T.R. Miles, T.R. Miles Jr., L.L. Baxter, R.W. Bryers, B.M. Jenkins, and L.L. Oden. Boiler deposits from firing biomass fuels. Biomass and Bioenergy, 10 (2-3):125 – 138, 1996. [173] B.M. Jenkins, L.L. Baxter, T.R. Miles Jr., and T.R. Miles. Combustion properties of biomass. Fuel Processing Technology, 54(1-3):17 – 46, 1998. [174] J.L. Colby. Catalytic Autothermal Reforming of Biomass to Synthesis Gas. PhD thesis, University of Minnesota, 2010. [175] J. Juan-Juan, M.C. Romn-Martinez, and M.J. Illan-Gomez. Effect of potassium

content in the activity of K-promoted ni/Al2O3 catalysts for the dry reforming of methane. Applied Catalysis A: General, 301(1):9 – 15, 2006.

[176] T. Osaki and T. Mori. Role of Potassium in Carbon-Free CO2 Reforming of Methane on K-Promoted Ni/Al2O3 Catalysts. Journal of Catalysis, 204(1):89 – 97, 2001. [177] H.S. Bengaard, I. Alstrup, I. Chorkendorﬀ, S. Ullmann, J.R. Rostrup-Nielsen, and J.K. Norskov. Chemisorption of methane on Ni(100) and Ni(111) surfaces with preadsorbed potassium. Journal of Catalysis, 187(1):238–244, 1999. [178] X. Zhu, M. Shen, L.L. Lobban, and R.G. Mallinson. Structural eﬀects of Na

promotion for high water gas shift activity on Pt-Na/TiO2. Journal of Catalysis, 278(1):123 – 132, 2011. BIBLIOGRAPHY 138

[179] J.K. Norskov, S. Holloway, and N.D. Lang. Microscopic model for the poisoning and promotion of adsorption rates by electronegative and electropositive atoms. Surface Science, 137(1):65 – 78, 1984.

[180] F. Frusteri, S. Freni, V. Chiodo, L. Spadaro, G. Bonura, and S. Cavallaro. Potassium improved stability of Ni/MgO in the steam reforming of ethanol for the production of hydrogen for MCFC. Journal of Power Sources, 132(1-2):139 – 144, 2004.

[181] S. Atsumi, T. Hanai, and J.C. Liao. Non-fermentative pathways for synthesis of branched-chain higher alcohols as biofuels. Nature, 451(7174):86–U13, 2008.

[182] S. Atsumi, W. Higashide, and J.C. Liao. Direct photosynthetic recycling of carbon dioxide to isobutyraldehyde. Nature Biotechnology, 27(12):1177–1180, 2009.

[183] S. Atsumi, T.-Y. Wu, E.-M. Eckl, S. Hawkins, T. Buelter, and J. Liao. Engi- neering the isobutanol biosynthetic pathway in Escherichia coli by comparison of three aldehyde reductase/alcohol dehydrogenase genes. Applied Microbiology and Biotechnology, 85:651–657, 2010.

[184] N. Qureshi, B.C. Saha, B. Dien, R.E. Hector, and M.A. Cotta. Production of butanol (a biofuel) from agricultural residues: Part I-Use of barley straw hydrolysate. Biomass and Bioenergy, 34(4):559 – 565, 2010.

[185] N. Qureshi, B.C. Saha, R.E. Hector, B. Dien, S. Hughes, S. Liu, L. Iten, M.J. Bowman, G. Sarath, and M.A. Cotta. Production of butanol (a biofuel) from agricultural residues: Part II- use of corn stover and switchgrass hydrolysates. Biomass and Bioenergy, 34(4):566 – 571, 2010.

[186] K. Smith, K.-M. Cho, and J. Liao. Engineering Corynebacterium glutamicum for isobutanol production. Applied Microbiology and Biotechnology, 87:1045– 1055, 2010.

[187] W. Higashide, Y. Li, Y. Yang, and J.C. Liao. Metabolic Engineering of Clostrid- ium cellulolyticum for Production of Isobutanol from Cellulose. Applied and Environmental Microbiology, 77(8):2727–2733, 2011.

[188] V. Garcia, J. Pakkila, H. Ojamo, E. Muurinen, and R.L. Keiski. Challenges in biobutanol production: How to improve the eﬃciency? Renewable and Sustainable Energy Reviews, 15(2):964 – 980, 2011.

[189] M. Kumar and K. Gayen. Developments in biobutanol production: New in- sights. Applied Energy, 88(6):1999 – 2012, 2011. BIBLIOGRAPHY 139

[190] L.D. Schmidt, R. Subramanian, J.R. Salge, and G.A. Deluga. Hydrogen and Chemicals from Renewable Fuels by Autothermal Reforming. Indian Chem. Eng. Sec. B, 47:100–105, 2005.

[191] N.J. Degenstein, R. Subramanian, and L.D. Schmidt. Partial oxidation of n- hexadecane at short contact times: Catalyst and washcoat loading and catalyst morphology. Applied Catalysis A: General, 305(2):146 – 159, 2006.

[192] I.C. Lee, J.G. St. Clair, and A.S. Gamson. Catalytic partial oxidation of isobutanol for the production of hydrogen. International Journal of Hydrogen Energy, 37(2):1399 – 1408, 2012.

[193] A. Donazzi, A. Beretta, G. Groppi, and P. Forzatti. Catalytic partial oxidation

of methane over a 4% Rh/α-Al2O3 catalyst: Part I: Kinetic study in annular reactor. Journal of Catalysis, 255(2):241 – 258, 2008.

[194] J. Barbier Jr. and D. Duprez. Reactivity of steam in exhaust gas catalysis I. Steam and oxygen/steam conversions of carbon monoxide and of propane over PtRh catalysts. Applied Catalysis B: Environmental, 3(1):61 – 83, 1993.

[195] G. Ingram and S.M.H. Rizvi. The pyrolytic identiﬁcation of organic molecules: IV. The pyrolytic behavior of the isomeric butanols. Microchemical Journal, 20 (3):324 – 352, 1975.

[196] J.A. Barnard. Chapter 4 gas phase combustion of organic compounds other than hydrocarbons and aldehydes. In C.H. Bamford and C.F.H. Tipper, editors, Gas-Phase Combustion, volume 17 of Comprehensive Chemical Kinetics, pages 441 – 500. Elsevier, 1977. doi: 10.1016/S0069-8040(08)70082-7. URL http: //www.sciencedirect.com/science/article/pii/S0069804008700827.

[197] M.R. Harper, K.M. Van Geem, S.P. Pyl, S.S. Merchant, G.B. Marin, and W.H. Green. Erratum to comprehensive reaction mechanism for n-butanol pyrolysis and combustion [Combust. Flame 158 (2011) 16-41]. Combustion and Flame, 158(10):2075 –, 2011.

[198] A.A. Shoaibi and A.M. Dean. Kinetic Analysis of C4 Alkane and Alkene Pyrolysis: Implications for SOFC Operation. Journal of Fuel Cell Science and Technology, 7(4):041015, 2010.

[199] R. Grana, A. Frassoldati, T. Faravelli, U. Niemann, E. Ranzi, R. Seiser, R. Cat- tolica, and K. Seshadri. An experimental and kinetic modeling study of combustion of isomers of butanol. Combustion and Flame, 157(11):2137 – 2154, 2010. BIBLIOGRAPHY 140

[200] A.M. El-Nahas, A.H. Mangood, H. Takeuchi, and T. Taketsugu. Thermal Decomposition of 2-Butanol as a Potential Nonfossil Fuel: A Computational Study. The Journal of Physical Chemistry A, 115(13):2837–2846, 2011.

[201] C. Togbe, A. Mze-Ahmed, and P. Dagaut. Kinetics of Oxidation of 2-Butanol and Isobutanol in a Jet-Stirred Reactor: Experimental Study and Modeling Investigation. Energy and Fuels, 24(9):5244–5256, 2010.

[202] C.D. Hurd, P. Perletz, and S.S. Drake. Attempts to Isolate Methylketene by Pyrolytic Methods. The Journal of Organic Chemistry, 10(1):62–66, 1945.

[203] P. Jeﬀers and S.H. Bauer. The homogeneous pyrolysis of 2-butenes. Interna- tional Journal of Chemical Kinetics, 6(5):763–771, 1974.

[204] M.P. Zum Mallen and L.D. Schmidt. Oxidation of Methanol over Polycrystalline Rh and Pt: Rates, OH Desorption, and Model. Journal of Catalysis, 161(1): 230 – 246, 1996.

[205] S. Douhou, D. Perrin, and R. Martin. Etude cinétique et modélisationde la réactionthermique de l’isobuténe vers 800 K. i: Isobuténepur. J. Chim. Phys., 91(10):1597 – 1627, 1994.

[206] S. Santhanam, J.H. Kiefer, R.S. Tranter, and N.K. Srinivasan. A shock tube, laser-schlieren study of the pyrolysis of isobutene: Relaxation, incubation, and dissociation rates. International Journal of Chemical Kinetics, 35(8):381–390, 2003.

[207] M. Mavrikakis and M.A. Barteau. Oxygenate reaction pathways on transition metal surfaces. Journal of Molecular Catalysis A: Chemical, 131(1-3):135 – 147, 1998.

[208] N.F. Brown and M.A. Barteau. Epoxides as probes of oxametallacycle chemistry on Rh(111). Surface Science, 298(1):6 – 17, 1993.

[209] C.J. Houtman and M.A. Barteau. Divergent pathways of acetaldehyde and ethanol decarbonylation on the Rh(111) surface. Journal of Catalysis, 130(2): 528 – 546, 1991.

[210] N.F. Brown and M.A. Barteau. Reactions of 1-propanol and propionaldehyde on rhodium(111). Langmuir, 8(3):862–869, 1992.

[211] N.F. Brown and M.A. Barteau. Comparison of the Reactions of Branched Alcohols and Aldehydes on Rh(111). The Journal of Physical Chemistry, 100 (6):2269–2278, 1996. BIBLIOGRAPHY 141

[212] C.J. Kliewer, C. Aliaga, M. Bieri, W. Huang, C.-K. Tsung, J.B. Wood, K. Komvopoulos, and G.A. Somorjai. Furan Hydrogenation over Pt(111) and Pt(100) Single-Crystal Surfaces and Pt Nanoparticles from 1 to 7 nm: A Kinetic and Sum Frequency Generation Vibrational Spectroscopy Study. Journal of the American Chemical Society, 132(37):13088–13095, 2010.

[213] I. Lee and F. Zaera. Enantioselectivity of Adsorption Sites Created by Chiral 2- Butanol Adsorbed on Pt(111) Single-Crystal Surfaces. The Journal of Physical Chemistry B, 109(26):12920–12926, 2005.

[214] X. Hu, R.A. Rosenberg, and M. Trenary. Thermal and Electron-Induced Decomposition of 2-Butanol on Pt(111). The Journal of Physical Chemistry A, 115(23):5785–5793, 2011.

[215] B.E. Bent, C.M. Mate, J.E. Crowell, B.E. Koel, and G.A. Somorjai. Bonding and thermal decomposition of propylene, propadiene, and methylacetylene on the rhodium(111) single-crystal surface. The Journal of Physical Chemistry, 91 (6):1493–1502, 1987.

[216] S.H. Oh, P.J. Mitchell, and R.M. Siewert. Methane oxidation over alumina- supported noble metal catalysts with and without cerium additives. Journal of Catalysis, 132(2):287 – 301, 1991.

[217] Y.-F.Y. Yao. Oxidation of Alkanes over Noble Metal Catalysts. Industrial and Engineering Chemistry Product Research and Development, 19(3):293–298, 1980.

[218] Y. Yu, J. Bartle, C.-Z. Li, and H. Wu. Mallee Biomass as a Key Bioenergy Source in Western Australia: Importance of Biomass Supply Chain. Energy and Fuels, 23(6):3290–3299, 2009.

[219] C. Weber, A. Farwick, F. Benisch, D. Brat, H. Dietz, T. Subtil, and E. Boles. Trends and challenges in the microbial production of lignocellulosic bioalcohol fuels. Applied Microbiology and Biotechnology, 87(4):1303–1315, 2010.

[220] N. Savage. The ideal biofuel. Nature, 474(7352):S9–S11, 2011.

[221] D.L. Trimm. Minimisation of carbon monoxide in a hydrogen stream for fuel cell application. Applied Catalysis A: General, 296(1):1 – 11, 2005.

[222] A.S. Chellappa, C.M. Fischer, and W.J. Thomson. Ammonia decomposition

kinetics over Ni-Pt/Al2O3 for PEM fuel cell applications. Applied Catalysis A: General, 227(1-2):231 – 240, 2002.

[223] J.M. Zalc and D.G. Loﬄer. Fuel processing for PEM fuel cells: transport and kinetic issues of system design. Journal of Power Sources, 111(1):58 – 64, 2002. BIBLIOGRAPHY 142

[224] C. Wheeler, A. Jhalani, E.J. Klein, S. Tummala, and L.D. Schmidt. The water- gas-shift reaction at short contact times. Journal of Catalysis, 223(1):191 – 199, 2004.

[225] X. Wang and R.J. Gorte. A study of steam reforming of hydrocarbon fuels on Pd/ceria. Applied Catalysis A: General, 224(1-2):209 – 218, 2002.

[226] B.C. Michael, A.Donazzi, and L.D. Schmidt. Eﬀects of H2O and CO2 addition in catalytic partial oxidation of methane on Rh. Journal of Catalysis, 265(1): 117 – 129, 2009.

[227] A. Donazzi, D. Livio, A. Beretta, G. Groppi, and P. Forzatti. Surface tem-

perature proﬁles in CH4 CPO over honeycomb supported Rh catalyst probed with in situ optical pyrometer. Applied Catalysis A: General, 402(1 - 2):41 – 49, 2011.

[228] K. Walter, O.V. Buyevskaya, D. Wolf, and M. Baerns. Rhodium-catalyzed partial oxidation of methane to CO and H2. In situ DRIFTS studies on surface intermediates. Catalysis Letters, 29:261–270, 1994.

[229] W.Z. Weng, M.S. Chen, Q.G. Yan, T.H. Wu, Z.S. Chao, Y.Y. Liao, and H.L. Wan. Mechanistic study of partial oxidation of methane to synthesis gas over supported rhodium and ruthenium catalysts using in situ time-resolved FTIR spectroscopy. Catalysis Today, 63(2 - 4):317 – 326, 2000.

[230] J.-M. Li, F.-Y. Huang, W.-Z. Weng, X.-Q. Pei, C.-R. Luo, H.-Q. Lin, C.-J.

Huang, and H.-L. Wan. Eﬀect of Rh loading on the performance of Rh/Al2O3 for methane partial oxidation to synthesis gas. Catalysis Today, 131(1-4):179 – 187, 2008.

[231] J.-D. Grunwaldt, L. Basini, and B.S. Clausen. In Situ EXAFS Study of

Rh/Al2O3 Catalysts for Catalytic Partial Oxidation of Methane. Journal of Catalysis, 200(2):321 – 329, 2001.

[232] M. Ni, D.Y.C. Leung, M.K.H. Leung, and K. Sumathy. An overview of hydrogen production from biomass. Fuel Processing Technology, 87(5):461 – 472, 2006.

[233] S.Q. Turn, C.M. Kinoshita, and D.M. Ishimura. Removal of inorganic constituents of biomass feedstocks by mechanical dewatering and leaching. Biomass and Bioenergy, 12(4):241 – 252, 1997.

[234] S. Arvelakis and E.G. Koukios. Physicochemical upgrading of agroresidues as feedstocks for energy production via thermochemical conversion methods. Biomass and Bioenergy, 22(5):331 – 348, 2002.