CATALYST STUDIES ON THE CONVERSION OF BIOBASED INT ERMEDIATES TO BIOBASED PRODUCTS By Iman Nezam A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Chemical Engineering Doctor of Philosophy 2019 ABSTRACT CATALYST STUDIES ON THE CONVERSION OF BIOBASED INTERMEDIATES TO BIOBASED PRODUCTS By Iman Nezam The goal of this work is to enhance the production of fuels and chemicals from fermentation - derived materials via two routes. Route (a) focuses on Guerbet chemistry, the n - butanol production from ethanol; route (b) studies the production of acrylate esters from 2 - acetoxypropanoic acid (APA) esters. The catalytic condensation of ethanol to n - butanol and higher alcohols, known as the Gue rbet reaction, has attracted more attention in recent years due to the commercial availability of ethanol as a bio - renewable feed stock. Among various catalysts considered for this process, none have obtained stable and economically affordable yields; alumi na - supported metals have been less explored despite the ir promising primary results in the lower energy - demandi ng condensed - phase. Experiments on the continuous condensed - phase conversion of ethanol to n - butanol using Ni/ La 2 O 3 - Al 2 O 3 catalyst present a W HSV of >0.8 h - 1 and a temperature range of 210 - 250 o C as the ideal reaction conditions . Several nickel bimetall ic catalysts have been examined to optimize the reaction performance further ; characterization techniques have been employed to understand the be havior of these catalysts more effectively . Copper addition shifts the selectivity of the Guerbet products towa rd n - butanol rather than C 6 + alcohols, which is explained by the copper behavior reducing H 2 adsorption on the catalyst. Furthermore, the number of nickel atoms on the surface of the catalyst correlates directly with the performance of the Guerbet reaction, suggesting that the dehydrogenation of ethanol is the rate - limiting step of the reaction. Among diffe rent catalysts and reaction conditions st udied, the best results were obtained at the temperature of 250 o C and WHSV of 0.8 h - 1 using 1.0 wt% Ni/ 9.0 wt% La 2 O 3 - Al 2 O 3 with 41% ethanol conversion and 74% C 4 + alcohols selectivity. Fusel alcohol Guerbet studi es under the same conditions ha ve resulted in 88% higher alcohols selectivity at 12% conversion. Preliminary kinetic modeling analysis for the isoamyl alcohol - ethanol mixtures shows that the ethanol self - condensation reaction has the highest rate constant among the self - condensation and cross - condensation reactions in the system. Economic analysis for a first - generation facility producing 25 million gallons of n - butanol per year has been performed for several scenarios of catalytic performance and process co nfiguration to investigate the viability of the commercial use of this catalyst. Results indicate that the n - butanol required selling price at 25% return on investm ent (ROI) can vary between $1.30 - $1.60 per kg of n - butanol, which is reasonably competitive with the current n - butanol market price. The highly selective production of 2 - acetoxypropanoic acid (APA) from lactic acid and acetic acid through reactive distillation has motivated the study of the elimination rea ction of APA esters to acrylate esters. Among different APA esters studied, the best results are obtained for those with no hydrogen on the - carbon of the ester functionality . This hydrogen allows the elimination of the ester group as an alkene, leading t o the production of highly reactive materials that can decompose to other side - products and reduce the desired products selectivity. The use of CO 2 as the diluent gas reduces the amount of carbon deposited on the surface of the contact material and maintai ns the rate of the elimination reaction in extended operation. Highest yields of 35% for butyl acrylate and 70% for methyl acrylate and benz yl acrylate at 550 o C and LHSV of 1.9 h - 1 have been achieved in this study. Copyright by IMAN NEZAM 2019 v This dissertation is dedicated to my wonderful parents and beloved wife Neda vi ACKNOWLEDGMENTS First and foremost, my sincere thanks go to my research advisor, Dr. Dennis Miller for his invaluable mentorship, constant support, encouragement, and inspiration. It is difficult to overstate how I feel honored and thankful that I had the opportunity of b eing in his research group. From the first week of my life in the U.S. until now, I have been lucky enough to grow professionally and personally under his exceptional supervision. Throughout my Ph.D. study, Dr. Miller put his great effort to teach me his deep knowledge clearly and simply . He always pro vided me with a stimula ting and positive environment to learn and grow. His assistance and dedicated involvement in every single step throughout my Ph.D. study, helped me to accomplish my academics goals successfully . Besides the wonderful academic support , Dr. Miller and his wi fe, Dr. Daina Briedis were always there to support my wife and I living as international students far away from our families. I would like to thank Dr. Miller and Dr. Briedis for all their kind supports and vibes, which gave us hope and determination to ke ep going toward our goals. I would like to acknowledge Dr. Carl Lira for his valuable assistance throughout the research and his support of my doctorate studies, especially during the past year. I would like to thank my committee me mbers, Dr. James Jackso n and Dr. Christopher Saffron for their precious insight and contributions to this work. I would also like to thank Dr. Lars Peereboom for his training on instruments and guidance on experiments. I am also grateful of Jason Zak for c ontributing to my catal yst characterization and analytical studies. I would like to thank the National Corn Growers Association and the U.S. Department of Energy (Award no. DE - FG36 - 04G014216) for the financial support of this work. vii I would like to thank my wife and best friend, Neda Rafat, for being my happiness, my life, and my support at any time. I am grateful for her constant encouragement and assistance especially when the times got rough. I would also like to express my sincere gratitude for my lovely parents who guided me to where I am today. They have sacrificed a lot for me and constantly encouraged me to seek higher goals and have never doubted me. I would like to thank my wonderful sisters, Fahimeh and Forough, who I owe my childhood to and always making me feel loved a nd blessed, and my nephews and nieces, Alireza, Arnik, and Melika for their innocent love and bringing happiness in my life. Finally, during my time at Michigan State University, I met so many wonderful friends who always tried to su pport me through their great company and helpful advice. I would like to thank them all, especially my closest friends, Ali Tehrani, William Killian, and Aseel Bala. I am thankful for their kindness and enduring support. viii TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ......................... xi LIST OF FIGURES ................................ ................................ ................................ ..................... xiii KEY TO SYMBOLS AND A BBREVIATIONS ................................ ................................ ........ xvi 1 Literature Review and Background ................................ ................................ ........................ 1 1.1 Introduction ................................ ................................ ................................ ......................... 1 1.2 Butanol Production Methods ................................ ................................ .............................. 2 1.2.1 Petrochemical Processes ................................ ................................ ............................. 2 1.2.2 Fermentation Processes ................................ ................................ ............................... 2 1.2.3 Guerbet Process ................................ ................................ ................................ .......... 4 1.3 Mechanisms Suggested for the Guerbet Reaction ................................ .............................. 4 1.3.1 Aldol Condensation ................................ ................................ ................................ .... 4 1.3. 2 Direct Dehydration ................................ ................................ ................................ ...... 6 1.4 Catalysts ................................ ................................ ................................ .............................. 7 1.4.1 Homogeneous catalysts ................................ ................................ ............................... 7 1.4.2 MgO ................................ ................................ ................................ ............................ 9 1.4.3 Mg - Al Mixed Oxides ................................ ................................ ................................ 10 1.4.4 Hydroxyapatite (HAP) ................................ ................................ .............................. 12 1.4.5 Al 2 O 3 Support ed Catalysts ................................ ................................ ........................ 14 1.4.6 Other Catalysts ................................ ................................ ................................ .......... 17 1.5 Condensed Phase Reactions ................................ ................................ .............................. 18 1.6 C ontinuous Reactors ................................ ................................ ................................ ......... 18 1.7 Nickel Bimetallic Catalyst ................................ ................................ ................................ 19 1.8 Objectives ................................ ................................ ................................ ......................... 20 1.8.1 Process and Catalyst Studies on the Guerbet Reaction ................................ ............. 20 1.8.2 Techno - economic analysis of the industrial scale Guerbet reaction ......................... 22 1.8.3 Guerbet reaction studies of the Fusel Alcohols ................................ ........................ 22 1.8.4 Enhanced Acrylate Production from 2 - Acetoxypropanoic Acid Esters ................... 23 2 Ethanol Guer bet Reaction ................................ ................................ ................................ ..... 24 2.1 Introduction ................................ ................................ ................................ ....................... 24 2.2 Materials and Meth ods ................................ ................................ ................................ ...... 25 2.2.1 Material s and Catalyst Preparation ................................ ................................ ........... 25 2.2.2 Reactor System ................................ ................................ ................................ ......... 26 2.2.3 Analytical Methods ................................ ................................ ................................ ... 29 2.2.4 Catalyst Characterization ................................ ................................ .......................... 29 2.3 Experimental results ................................ ................................ ................................ .......... 31 2.3.1 Control Experiments ................................ ................................ ................................ . 31 2.3.2 Nickel Monometallic Catalytic Experiments ................................ ............................ 31 2.3.3 Nickel Bimetallic Catalytic Experiments ................................ ................................ .. 36 2.3.4 Mechanistic and Kinetic Studies ................................ ................................ ............... 45 2.4 Conclusions ................................ ................................ ................................ ....................... 48 ix APPENDICES ................................ ................................ ................................ .......................... 50 APPENDIX A: Catalyst Prepar ation Steps ................................ ................................ ........... 51 APPENDIX B: Feed Flow Rate Calculation ................................ ................................ ........ 54 APPENDIX C: Activation Energy Calculation ................................ ................................ .... 55 APPENDIX D: Nickel Monometallic Experimental Data ................................ .................... 57 APPENDIX E: Nickel Bimetallic Experimental Data ................................ .......................... 62 APPENDIX F: Temperature programmed Desorption (TPD) Profiles ................................ 69 3 Guerbet Economical Analysis ................................ ................................ ............................... 72 3.1 Introduction ................................ ................................ ................................ ....................... 72 3.2 Process Concept and Design Parameters ................................ ................................ .......... 73 3.2.1 Process Concept ................................ ................................ ................................ ........ 73 3.2.2 Definition of Design Parameters ................................ ................................ ............... 80 3.3 Techno - Economic Analysis Results ................................ ................................ ................. 86 3.3.1 Base Case ................................ ................................ ................................ .................. 86 3.3.2 Sensitivity An alysis of Key Cost Drivers ................................ ................................ . 89 3.3.3 Alternate Process Configurations ................................ ................................ ............. 91 3.4 Conclusions ................................ ................................ ................................ ....................... 95 APPENDICES ................................ ................................ ................................ .......................... 96 APPENDIX G. Formulas for Cost Estimation ................................ ................................ ..... 97 4 Fusel Alcohols Production Studies ................................ ................................ ..................... 100 4.1 Introduction ................................ ................................ ................................ ..................... 100 4.2 Materials and Methods ................................ ................................ ................................ .... 101 4.2.1 Materials and Catalyst Preparation ................................ ................................ ......... 101 4.2.2 Reactor System ................................ ................................ ................................ ....... 102 4.2.3 Analytical Methods ................................ ................................ ................................ . 103 4.3 Experimental Results ................................ ................................ ................................ ...... 103 4.3.1 Batch Experiments ................................ ................................ ................................ .. 103 4.3.2 Continuous Experiments ................................ ................................ ......................... 105 4.3.3 Kinetic Model Devel opment ................................ ................................ ................... 106 4.4 Conclusions ................................ ................................ ................................ ..................... 112 APPENDICES ................................ ................................ ................................ ........................ 113 APPENDIX H. Detailed Selec tivity Calculations ................................ .............................. 114 5 Acrylate Production from 2 Acetoxypropanoic Acid Esters ................................ ............. 115 5.1 Introduction ................................ ................................ ................................ ..................... 115 5.1.1 Glycerol Dehydration ................................ ................................ .............................. 116 5.1.2 Hydroxypropanoic acid Direct Dehydration ................................ ........................... 116 5.1.3 2 - Acetoxypropanoic Acid Indirect Dehydration ................................ .................... 117 5.2 Materials and Methods ................................ ................................ ................................ .... 119 5.2.1 Materials and Catalyst Preparation ................................ ................................ ......... 119 5.2.2 Reactor Configuration ................................ ................................ ............................. 120 5.2.3 Analytical Methods ................................ ................................ ................................ . 122 5.2.4 Contact Material Characterization ................................ ................................ .......... 123 5.3 Experimental Results ................................ ................................ ................................ ...... 123 x 5.3.1 Control Experiments ................................ ................................ ............................... 124 5.3.2 Feed Composition ................................ ................................ ................................ ... 125 5.3.3 Acrylate Yields from Different APA Esters ................................ ........................... 128 5.3.4 Reaction Temperature ................................ ................................ ............................. 130 5.3.5 Different Contact Materials ................................ ................................ .................... 131 5.3.6 Space V elocity ................................ ................................ ................................ ........ 133 5.3.7 Extended Reaction ................................ ................................ ................................ .. 133 5.4 Conclusions ................................ ................................ ................................ ..................... 134 APPENDICES ................................ ................................ ................................ ........................ 135 APPENDIX I. Rate constant and activation energy of APA conversion ............................ 136 APPENDIX J. Temperature programmed Desorption (TPD) Profiles ............................... 137 6 Summary and Recommendations for Future Work ................................ ............................ 139 6.1 Summary ................................ ................................ ................................ ......................... 139 6.2 Ethanol Guerbet Reaction ................................ ................................ ............................... 139 6.2.1 Catalyst Studies ................................ ................................ ................................ ....... 139 6.2.2 Separate Performance Improvement for Each Step of the Guerbet Reaction ......... 141 6.2.3 Kinetic Modeling of the Reaction Tree ................................ ................................ .. 141 6.3 Guerbet Economic Analysis ................................ ................................ ........................... 143 6.4 Fusel Alco hols Guerbet Reaction ................................ ................................ ................... 144 6. 5 Acrylate Production from 2 - Acetoxypropanoic Acid Esters ................................ .......... 144 REFERENCES ................................ ................................ ................................ ........................... 146 xi LI ST OF TABLES Table 2.1. Bimetallic catalyst compositions on a mass and molar basis. The balance is - Al 2 O 3 in all cases. The data on the first row is for the 8.0 wt% Ni/9.0 wt% La 2 O 3 / - Al 2 O 3 catalyst that second metals are impregnated on. M is the second metal. Ni - Cu bimetallic is made in two different metal ratios. ................................ ................................ ................................ .................... 27 Table 2.2. Compariso n of the performance of different Ni/La 2 O 3 - Al 2 O 3 catalysts and their surface characteristics. ................................ ................................ ................................ .................. 35 Table 2.3. Bimetallic catalysts expe rimental results and surface analysis comparison with the Ni (III) catalyst. Experiments are conducted at T= 230 o C and WHSV= 1.42 h - 1 with ethanol as the feed. a: Ethyl Acetate b: Diethyl Ether. All bimetallic catalysts are prepared with the separate impregnation method. ................................ ................................ ................................ ................... 38 Table 2.4. Metal particle diameter and dispersion for the Ni - Pd and Ni - Pt bimetallic catalysts. Numbers are calculated based on the H 2 uptake. ................................ ................................ .......... 42 Table 2.5. Surface elemental analysis conducted on bimetallic catalysts using SEM/EDS. ........ 43 Table 2.6. Different nickel, c opper, and cobalt based catalysts studied. All results are at T= 230 o C and WHSV= 1.42 h - 1 . ................................ ................................ ................................ ............... 44 Table 2.7. Acetaldehyde experiments over Ni (III) catalyst. All experiments are done at WHSV= 1.42 h - 1 . ................................ ................................ ................................ ................................ ......... 46 Table A.1. The spreadsheet used for obtaining the mass required for different chemicals in prepar ing catalysts . . 51 Table D.1 . Experimental details for the monometallic nickel experiments 7 Table D.2 monometalli c nickel experiments 8 Table D.3. monometallic nickel experiments 60 Table E .1. Experimental details for the bimetallic nickel experiments 2 Table E.2 . nickel experiments 4 Table E . 3 . bimetallic nickel experiments 6 T able 3.1. Parameters for techno - economic analysis ................................ ................................ .... 81 Table 3.2. Specifications of reactor and distillation columns for four base case scenarios. 1 Case designations: The number refers to p er - pass et hanol conversion in the fixed - xii refers to low selectivities of 60% to n - butanol, 16% to C 6 + al selectivities of 72% to n - butanol and 22% to C 6 + alcohols. ................................ ......................... 82 Table 3.3. Parameters for heat exchangers for Case 35L. 1 All process heat provided by steam (257 o C, 45 bar) from steam generator except for Heater 1. 2 All proces s cooling provided by direct air cooling in forced convection air - cooled heat exchangers. The area reported for air - cooled heat exchangers is bare tube area; fin:tube area ratio = 17:1. ................................ ........... 84 Table 3.4. Composition of Process Streams for 35L Case. ................................ .......................... 86 Table 3.5. Economic an alysis of base cases (Basis: 25 million gallons (75 million kg) n - buta nol/yr). 1 Fixed costs include capital depreciation, taxes, and insurance. 2 General Expenses include administration, distribution and selling costs, R&D, and financing costs ....................... 87 Table 3.6. Capital an d operating costs for different heat removal options for Case 40L. ............ 94 Table 4.1. Results of batch reactor experiments with 120 g of (ethanol - isoamyl alcohol) mixture at 230 o C,4.85 g of 8.0 wt% Ni/ 9.0 wt% La 2 O 3 - Al 2 O 3 is used as th e catalyst. ........................ 105 Table 4.2. Results of continuous cond ensed - phase experiments with 4/1 molar ratio of ethanol/isoamyl alcohol and 29.9 g of 8.0 wt% Ni/9.0 wt% La 2 O 3 / - Al 2 O 3 cat alyst. 1 Conversion is too low; thus, this set of data are unreliable. ................................ ................................ ........... 106 T able 4.3. Rate Constants developed for different batch experiments and their comparison with the one obtained for the cont inuous system at 230 o C. ................................ ............................... 111 Table 5.1. Summary of conditions, functional g roup balances, and product distributions for selected experiments R1 - R5. a Total carbon recovery doe s not include the carb on deposited on the contact material. ................................ ................................ ................................ .................... 125 Table 5.2. Properties and Experim ental Results for Contact Materials. Reaction conditions: T = 550°C; 80 wt% butyl APA/20 wt% acetic acid feed at 2.4 mL/h; CO 2 diluent gas at 20 ml/min; LHSV = 1.9 kg feed/kg contact material/h. ................................ ................................ ................ 132 xiii LIST OF FIGURES Figure 1.1. Petroleum - based methods for the production of butanol. (a) oxo synthesis, (b) Reppe synthe sis, and (c) crot onaldehyde hydrogenation [9] ................................ ................................ ..... 3 Figure 1.2. Reaction mechanism for the et hanol Guerbet reaction system [9] ............................... 5 Figure 1.3. Direct dehydration mechanism [17] ................................ ................................ ............. 6 Figure 2.1. Continuous flow reacto r for ethanol conversion. (a - stainless steel rods; b - SiC packing for feed preheating; c - catalyst bed; d - thermowell; e - oil jacket) ............................... 27 Figure 2.2. Typical C 6 + alcohols molar distribution. Negligible amounts of 1 - decanol are also present. ................................ ................................ ................................ ................................ .......... 31 Figure 2.3. Temperature dependence of the 8.0 wt% Ni/4.5 wt% La 2 O 3 / - Al 2 O 3 catalyst for the Guerbet reaction (WHSV= 1.4 h - 1 ) ................................ ................................ ............................... 33 Figure 2.4. Weight hourly sp eed velocity (WHSV) dependence of the 8.0 wt% Ni/4.5 wt% La 2 O 3 / - Al 2 O 3 catalyst for the Guerbet reaction (T=210 o C) ................................ ....................... 34 Figure 2.5. SEM picture from the co - impregnated 11.5 wt% (Ni/Cu 2/1)/8.7 wt% La 2 O 3 - Al 2 O 3 catalyst. ................................ ................................ ................................ ................................ ......... 42 Figure 2.6. Correlation between H 2 adsorpt ion on the surface of different 8 wt% nickel catalysts ................................ ................................ ................................ ... 47 Figure C.1 . Activation energy calculation for the 8.0 wt% Ni/4.5 wt% La 2 O 3 - Al 2 O 3 catalyst 6 Figure F . 1 9 Figure F . 2 70 Figure F . 3 . H 2 uptake measurement for bi 71 Figure 3.1. Process concept for the four base cases of ethanol conversion to n - butanol and higher alcohols. The values of reboiler, condenser and heat exchanger duties are for Case 35L. ........... 7 5 Figure 3.2. Residue curve map for ethanol/n - butanol/water system at 5.0 bar absolute (Aspen Plus V8.4 SR - Polar equation of state). Dashed lines represent material balances for distillation columns TC1 (S8 - S2 - S3) and TC3 (S6 - S3 - S7) for Cases 35L and 70H with feed and product streams exc luding light byproducts such that S3= n - butanol/water binary azeotrope; ethanol/water binary azeotrope. Distillation boundary is the curved solid line between azeotropes. ................................ ................................ ................................ ................................ .... 78 xiv Figure 3.3. Required n - butanol selling price vs. ROI f or four base cases in Table 3.5. ............... 88 Figure 3.4. Sensitivity analysis of required n - butanol selling price dependence on purchase equipment costs for Case 35L. Percentage in legend refers to increment in equipment purchase costs over Case 35L equipment purchase costs of $9.14 million. ................................ ................ 89 Figure 3.5. Sensitivity analysis o f required n - butanol selling price dependence on ethanol feed cost for Case 35L. Percentage in legend refers to increment in ethanol feed cost relative to Case 35L cost of $1.65/gallon ($0.53/kg). ................................ ................................ ............................ 90 Figure 3.6. Sensitivity analysis of n - butanol required selling price depen dence on utility costs for Case 35L. Percentage in legend refers to inc rement in total annual utility costs relative to Case 35L utility costs of $5.1 million. ................................ ................................ ................................ ... 90 Figure 4.1. Primary products observed from reaction of ethanol and isoamyl alcohol mixtures. ................................ ................................ ................................ ................................ ..................... 104 Figure 4.2. Comparison of simulated and experimental reactor outlet concentrations for the ethano l/isoamyl alcohol continuous Guerbet experiments at T= 210 o C using 8.0 wt% Ni/9.0 wt% La 2 O 3 / - Al 2 O 3 as the catalyst. (a)Ethanol, (b)Isoamyl alcohol. ................................ ......... 109 Figure 4.3. Comparison of simulated and experimental reactor outlet concentrations for the ethanol/isoamyl alcohol continuous Guerbet ex periments at T= 230 o C usi ng 8.0 wt% Ni/9.0 wt% La 2 O 3 / - Al 2 O 3 as the catalyst. (a)Ethanol, (b)Isoamyl alcohol, (c)2 - Isopropyl - 5 - methyl - hexanol. ................................ ................................ ................................ ................................ ....... 110 Figure 4.4. Co mparison of simulated and experimen tal reactor outlet concentrations for the ethanol/isoamyl alcohol continuous Guerbet experiments at T= 250 o C using 8.0 wt% Ni/9.0 wt% La 2 O 3 / - Al 2 O 3 as the catalyst. (a)Ethanol, (b)Isoamyl alcohol, (c)2 - Isopropyl - 5 - methyl - hexanol. ................................ ................................ ................................ ................................ ....... 110 Figure 5.1. Petroleum - based route to acrylic acid from propylene ................................ ............. 115 Figure 5.2. Methyl APA production using methyl lactate and acetic anhydride ........................ 118 Figure 5.3. Methyl APA production using lactic acid and methyl ace tate ................................ . 118 Figure 5.4. Schema tic of the reactor system ................................ ................................ ............... 121 Figure 5.5. Reaction pathways for APA ester elimination reactions. Path (a): elimination of acetic acid to form alkyl acrylate or acrylic acid; Path (b): elimination of alkene for ester R groups - carbon; Path (c): decarbonylation of APA to acetaldehyde, acetic acid, and CO. ................................ ................................ ................................ .................... 124 Figure 5.6. Effect of acetic acid feed concentration on steady state acrylate selectivity from butyl APA (R2, 45 mol% acetic acid; R4, 0% acetic acid). Reaction conditions: T = 550 °C; quartz (SiO 2 ) contact material; CO 2 diluent gas at material/h. ................................ ................................ ................................ ................................ ... 126 xv Figure 5.7. Total acrylate s electivity versus time on stream with N 2 or CO 2 as diluent gas (20 ml/min) for benzyl and n - butyl APA ester feed. Reaction conditions: T = 550 °C; 80 wt% APA ester/20 wt% acetic acid feed; LHSV = 1.9 kg feed/kg contact material/h. ............................... 127 Figure 5.8. Total acrylate selectivity vs time on stream for different APA ester feeds. Reaction conditions: T = 550 ° C; 80 wt% APA ester/20 wt% acetic acid feed; CO 2 diluent gas at 20 ml/min; LHSV = 1.9 kg feed/kg contact mat erial/h. ................................ ................................ .. 129 Figure 5.9. Steady state butyl APA conversion, butyl acrylate selectivity, and acrylic acid selectivity at steady state versus temperature (R5, 490 °C; R2, 550 °C). Reaction conditions: 80 wt% butyl APA/20 wt% acetic acid feed; CO 2 diluent gas at 20 ml/min; LHSV = 1.9 kg feed/kg contact material/h. ................................ ................................ ................................ ....................... 131 F igure 5.10. Extended time experiment. Reaction condit ions: T= 550 o C; 80 wt% butyl APA/20 wt% acetic acid feed; CO 2 diluent gas at 20 ml/min; LHSV = 1.9/kg feed/kg contact material/h. ................................ ................................ ................................ ................................ ..................... 133 Figure I.1 . Arrhenius plot of first ord er rate constant for APA decomposition 6 Figure J. 1 7 Figure J . 2 8 xvi KEY TO SYMBOLS AN D ABBREVIATIONS D p particle diameter k i rate constant for species i K equilibrium constant x reactant conversion superficial residence time r i rate of formation of species i C concentration in mol/vol E a activation energy C R reactor cost b base reactor cost D i shell I.D. p cm cost multiplier for O.D. f cm cost multiplier for TEMA - type front head r cm cost multiplier for TEMA - type rear head C T sum of base cost corrections for shell and tubes in the reactor A ht surface area for heat transfer E i escalation inde x (Chemical Engineering Plant Cost Index) C TC tray column cost C bs base shell cost C bt base tra y cost C pl base platforms and ladders cost f 1 cost multiplier for column material f 2 cost multiplier for column material f 3 cost multiplier for tray type f 4 cost multiplier for tray numbers N number of trays xvii W vessel weight L column height D colum n diameter T b head thickness T p shell thickness C SG steam generator cost C bg base steam generator cost C C air - cooled heat exchanger cost S reduced heat load C ac air - cooled heat exchanger cost function T i process inlet temperature T a ambient temper ature Q heat load C HE other heat exchangers cost C B other heat exchangers base cost A HE heat exchanger surface area F D cost multiplier for the exchanger type F P cost multiplier for the design - pressure F M cost multiplier for the material of construct ion C10 C 10 alcohols IA isoamyl alcohol APA 2 - acetoxypropanoic acid CS i calculated selectivity of species i based on GC results CS ij calculated selectivi ty of species I with respect to reactant j based on GC results S i total selectivity toward specie s i considering those molecules participated in other reactions S ij tot al selectivity toward species i with respect to reactant j considering those molecules participated in other reactions 1 1 Literature Review and Background 1.1 Introduction Petroleum resources have always been the subject of serious concerns for depletion. Ac cording to a study made in 2017, by maintaining current production levels, the re servoirs of coal, natural gas, and oil resources will be depleted in 114, 53, and 51 years, respec tively [ 1 ] . Furthermore, the role of fossil fuels in the current status of environmental sustainability has always been a controversial discussion. Some believe that regenerative and assimilative capacities shou ld be considered while using non - renewable resources [ 2 , 3 ] . Therefore, finding an alternative sourc e of energy and rchers for decades. This alternative needs to be renewable, envi ronmentally friendly, and efficiently produced to be considered as an appropriate option. The production of bioethanol (ethanol produced from bio - resources, C 2 H 5 OH) has been one of the signifi cant achievements in this ongoing challenge. Being produced from the fermentation of sugar, starch, or cellulosic - based glucose [ 4 ] , ethanol h as been the most common substitute for engine fuels [ 5 ] and one of the primary starting materials for the production of c ommodities which are already being produced using petroleum - base d resources. Ethanol can be used as a starting material for the production of butanol (C 4 H 9 OH ). Butanol has a wide range of applications in industry. Among them are its use as a chemical addi tive in the perfume industry, solvent in the paints and coatings industries, and extractant in the cosmetics and pharmaceuticals industry [ 6 ] . Moreo ver, butanol has some advantageous over ethanol as a fuel alternative. Chief among these are h igher energy density, more resistance to water contamination, and no phase separation while mixing with gasoline. Also, butanol is more akin to gasoline 2 consideri ng its chemical properties such as stoichiometric air - fuel ratio, heat of vaporization, resear ch octane number, and motor octane number [ 7 , 8 ] . 1.2 Butanol Production Methods 1.2.1 Petrochemical Processes Butanol is currently produced through petroleum - based methods. These methods are shown in Figure 1. 1 [ 9 ] and are: (a) oxo synthesis, (b) Reppe synth esis, and (c) crotonaldehyde hydrogenation [ 10 ] . Among them, oxo synthesis is the most widely used proce ss in the industry. In this process, first carbon monoxide and hydrogen are added to the carbon - carbon double bond of propylene in a hydroformylation process in the presence of various catalysts such as Co and Rh to produce butyraldehyde, and the n butyrald ehyde is hydrogenated to produce 1 - butanol. The Reppe reaction is essentially the same as oxo synthesis except that this process involves low temperature (100 o C) and low pressures (5 - 20 bar) conditions that facilitate the direct formation of alc ohols from the olefin. Crotonaldehyde hydrogenation was the preferred industrial process for 1 - butanol conversion until the mid - 1950s. This process involves three main steps including acetaldehyde condensation to acetadol in the presence of an alkaline cat alyst, deh ydration of acetadol by its acidification with acetic acid for the formation of crotonaldehyde, and finally hydrogenation of crotonaldehyde to 1 - butanol in th e presence of copper catalysts [ 10 ] . 1.2.2 Fermentation Processes Fermentation is one of the most common bio - based techniques for the production of butanol. There are two common fermentation methods used for the production of butanol. The first method, acetone - butano l - ethanol (ABE) fermentation, was discovered in 1911 by Strange Company & Graham Ltd. and a group of scientists [ 11 ] . In this method, hexose and pentose sugars are 3 fermented using Clostridi um acetobutylicum to produce acetone, butan ol, and ethanol in a standard 3:6:1 mass ratio, respectively [ 12 ] . There are a couple of issues involved with the ABE fermentation process. Among them are separation cost, low butanol titer, low yield, l ow productivity, and high cleaning costs resulting from potential bacteriophage invasion [ 13 ] . These challenges have been t he subject of several studies d uring the past decade. Developed genetic engineering and pre - treatment processes have led to promi sing accomplishments in addressing some of these challenges. Indeed, the use of modified Clostridium strains has led to improve d productivity of butanol compa red to the other two main products [ 14 ] . Nevertheless, the most significant difficulties lay ahead when it comes to the separation costs and overall ABE yield [ 11 ] . Figure 1. 1 . Petroleum - based m ethods for the production of butanol. (a) oxo sy nthesis, (b) Reppe synthesis, and (c) crotonaldehyde hydrogenation [9] 4 1.2.3 Gu erbet Process The catalytic pathway for the production of bio - butanol from bio - ethanol is called the Guerbet reaction . The availability of ethanol at a large scale as one of the most prominent sources of bio - based carbon is one of the biggest motivations for butanol production through this process [ 15 ] . Guerbet was the first scientist who produced th is process for the production of dimer alcohols from aliphatic alcohols with loss of one molecule of water [ 16 ] . While different mechanisms have been proposed for the production of 1 - butanol from ethanol, the one which is ac cepted by most scientists is a three - step reaction involving dehydrogenation, aldol condensation, and hydrogenation [ 17 ] . This mechanism is explained in more depth in t he next section. There have been several studies for conducting Guerbet reaction with high ethanol conversion and selectivity to higher alcohols. These experiments were done in batch or continuous reactors with the use of heterogeneous and homogeneous cata lysts at different reaction conditions [ 18 - 22 ] . Studying the proposed reaction mechanisms, reactor configurations, and catalyst structures used for obtaining higher product yields in previous similar studies wi ll enhance the familiarity with the reaction system a nd improve the chance of finding an alternative process for the current petrochemical process used for butanol production. 1.3 Mechanisms Suggested for the Guerb et Reaction 1.3.1 Aldol Condensation The primary me chanism accepted by most scientists for the Guerbet r eaction is the indirect mechanism known as aldol condensation. This mechanism, shown in detail s in Figure 1. 2 , in volves three main s teps: ethanol dehydrogenation to acetaldehyde, acetaldehyde aldol condensation to crotonaldehyde, and crotonaldehyde hydrogenation to butanol [ 9 ] . The produced butanol can 5 participate in further Guerbet react ions with other alcohols to produce 1 - hexanol, 2 - ethyl - 1 - butanol, 1 - octanol, 2 - ethyl - 1 - hexanol, etc. While dehydrogenation and hydrogenati on steps require a m etal catalyst to assist with hydrogen transfer, the aldol condensation step takes place in the pre sence of a support with both acid and base sites on it. Therefore, a multifunctional catalyst is required for conducting the reaction if t he suggested mechani sm is followed. Several observations support the proposed indirect mechanism. First, in most stu dies conducted on this reaction, the intermediate components for this mechanism are present as side - products [ 6 , 9 , 15 , 18 , 19 , 23 - 26 ] . Second, Ogo et al. studie d each intermediate component separately by Figure 1. 2 . Reaction mechanism for the ethanol Guerbet reaction system [9] 6 examining its reactivity at 300 o C and atmospheric pressure over hydroxyapatite (HAP) catalyst and observed high s electivity of the desired product in each experiment [ 24 ] . Third, Gines et al. studied the eff ect of adding labeled acetaldehyde to the feed and noticed improved butanol yield through the reaction of labeled carbon atoms [ 19 ] . Finally, the prer equisite for the aldol condensation step is the presence of a hydrogen atom on the - carbon. Weizmann tested and approved this requirement through feeding different types of al cohols to the Guerbet reaction chamber [ 27 ] . All these e vidence support the occurrence of this mechanism as the main one for butanol production from ethanol. 1.3.2 Direct Dehydration The dimerization of two ethanol molecules, first proposed by Yang et al. in 1993 [ 28 ] , is known is the direct dehydration mechanism for the Guerbet reaction. In this mec hanism, as shown in Figure 1 . 3 , a C - H bond in the - position in the ethanol molecule is first activated and then condense s with another ethanol molecu le to form butanol [ 17 ] . The most significant evidence for this mechanism is the lower activity of acetaldehyde compared to ethanol when fed to the reactor at reaction conditions [ 17 , 28 , 29 ] . However, the challenge with this mechanism is that all of the studies supporting it are conducted at high reaction tem pera tures (>350 o C) [ 26 ] . Under these conditions, ev en t hough selectivity to higher alcohols from acetaldehyde is lower than that from ethanol, the fact that higher alcohols are formed from acetaldehyde suggests that at high temperature the aldol condensation mechanism is still active. Figure 1 . 3 . Direct dehydration mech anism [17] 7 Scalbert et al. studied the Guerbet reaction using a non - metallic based catalyst (HAP) in the 350 - 410 o C temperature range and compared the reaction quotient (Q) to the thermodyna mic equilibrium constant (K) for the two proposed mechanisms [ 29 ] . While the Q/K ratio for the indirect mechanism was in the order of magnitude of two to three, this ratio was well below one for the direct condensation mechanism, suggesting that the direct condensation is the primary mechanism at the studied temperature s. Even then, the authors discuss that at lower temperatures (not thermodynamically controlled conditions) and in the presence of metal - promoted catalysts (with the ability of the formation of surface hydrogen atoms), the dominating mechanism is the acetal dehyde aldol condensation reaction. Based on the above discussions, one can claim that at high temperatures and over non - promoted oxides, the governing r eaction mechanism is the direct dehydration reaction with the indirect aldol condensation mechanism oc curring at slower rates. However, at lower temperatures and in the presence of metal - promoted supports, the accepted mechanism among most scientists is th e indirect reaction. 1.4 Catalysts 1.4.1 Homogeneous catalysts Homogeneous catalysts have shown significant pote ntial for obtaining high conversion and selectivity for the Guerbet reaction. Typically, these catalysts involve a metal complex, a phosphine ligand, and a basic inorganic solution. The metal is responsible for hydrogen transfer in the hydrogenation and de hydrogenation steps, and the inorganic base is responsible for the aldol condensation. Due to the presence of the phosphine ligands, these catalysts are s ubject to degradation to phosphine oxides when exposed to air. The primary advantage of these catalyst s is 8 their lower operating temperatures (<150 o C) compared to heterogeneous catalysts. Dowson et al. in 2013 were able to obtain 93% butanol selectivity a nd 22% ethanol conversion at 150 o C using ruthenium, bis(diphenylphosphanyl)methane ligand, and EtONa catalyst [ 30 ] . Same group was able to maintain the butanol selectivity at 31% ethanol conversion by upgrading the ligand to mixed - donor phosphine - amine (P - N) ligand [ 31 ] . Despite the high butanol selectivity obtained, one of the significant chall enges involved in the homogeneous catalysts is the deactivation of catalyst due to the decomposition of the ligand. None of the studies mentioned above ma intained their activity beyond 300 turnovers of the catalyst (moles of substrate reacted per mole of m etal). Ligand decomposition is attributed to, but not limited to, inadequate water tolerance of the ligand [ 26 ] . Tseng et al. used an amide - derived N,N,N - Ru II complex along with additional quantities of tri phenylphosphine (PPh 3 ) to improve the TON to 530 with ethanol conversion of 53%; however, the butanol selectivity decreased to 78% [ 32 ] . Xie et al. were able to show this correlation between the TON and reaction s electivity more clearly [ 33 ] . Although they achieved t he greatest TON of 18209 using acridine - based ruthenium pincer complexes at 150 o C, only 73% ethanol conversion and 60% butanol selectivity was obtained u sing this catalyst. So far, the highest selectivity for the ethanol Guerbet reaction has been reported by Jones et al. [ 34 ] . They modified the iridium catalyst with nickel hydroxide at 150 o C and were able to obtain > 99% butanol selectivity at 37% ethanol conversion and catalyst turnovers of 185. Despite the high butanol selectivity obtained, the stability of the catalyst is still the principal challenge in using this catalyst in industry. Other problems associated wit h this type of catalysts are separation costs and non - reusability of the catalyst. 9 1.4.2 MgO Metal oxides are generally among popular heterogeneous catalysts for the Guerbe t reaction because of the high alcohol selectivity they provide. They typically offer high basicity for the reaction environment, which is essential for the hydrogen exchange and cond ensation steps. Ueda et al. studied several metal oxides for the continuous cross - condensation of methanol with ethanol in the vapor phase, and observed the best reaction performance with MgO as the catalyst [ 35 ] . Other metal oxides studied in their experim ents were ZnO, CaO, and ZrO 2 . They reported an ethanol conversion of 30% and selectivity of 8 0% to higher alcohols at 360 o C and atmospheric pressure for the MgO catalyst. The selectivity obtained was even higher when using C 3 - C 5 alcohols as the second al cohol reacting with methanol [ 36 ] . MgO was also studied for the self - condensation reaction of alcohols. Ndou et al. investi gated vapor - phase ethanol self - condensation over several metal oxides such as MgO, CaO, BaO, and - Al 2 O 3 [ 17 ] . Optimal results were reported for MgO at 450 o C and 1 bar with 56% ethanol conversion and 19% selectivity to higher alcohols. The same group studied propanol self - condens ation over MgO catalyst at similar reaction conditions and observed 28% conversion and 50% se lectivity to 2 - methyl pentanol. Adding H 2 to the reaction environment improved the selectivity to 70%. Among multiple researchers studying the MgO catalyst for the Guerbet reaction, there is consensus that the reaction has to happen at high temperatures an d with vapor phase reactants. The minimum temperature used for a detectable activity of MgO as the catalyst for this reaction is 300 o C [ 15 ] . One of the methods t o improve the performance of MgO in the Guerbet reaction was the addition of metals to the ca talyst. However, the results were not very promising. Ueda et al. showed that the addition of transition metals to MgO typically maintained the rate of the dehydro genation reaction, but inhibited the hydrogenation step [ 36 ] . Ndou et al. had a similar observation using 10 alkaline earth an d transition metals [ 17 ] . In both cases, the selectivity to higher alcohols was reduced two - fold or more. Besides, adding alkali metals increased the MgO catalyst basicity, leading to increased selectivity to the dehydrogenation products and decreased selectivity to the condensati on products [ 37 ] . Olson et al. discovered that addition of nickel to activated carbon - supported MgO catalyst slightly improved the higher alcohols selectivity [ 38 ] . 1.4.3 Mg - Al Mixed Oxides Several researchers have addressed MgO as a catalyst that cannot solely provide all the functionalities required for the Guerbet mechanism [ 19 , 23 , 39 ] . Carvalho et al. claimed that adjacent acid and medium strength base sites are necessary for the dehydroge nation and condensation steps [ 40 ] . Di Cosimo et al. describe d the effect of Lewis acid sites on improving the rate of dehydrogenation reactions because of their enhanced hydrogen abstraction properties [ 23 ] . Tsuchida et al. compare d a dominantly basic MgO catalyst with another catalyst with both acid and base sites on it and observe d much higher selectivity to acetaldehyde and unsaturated alcohols in the reaction with MgO catalyst [ 39 ] . They justify this observation by stating that the acid sites have the capability of trapping the hydrogen molecules on the catalyst surface. Whereas on the MgO catalyst, H 2 is dissociated from the catalyst surface as a gas molecule. Because of these cat alytic requirements, several researchers have considered upgrading MgO catalyst using various metal oxides. Oxides of aluminum have been one of the most popular metal oxides that increase the acidity of the catalyst. Early in the 1930s, Fuchs et al. used a CuO/MgO/Al 2 O 3 mixtu re for the vapor phase Guerbet reaction [ 41 ] . They were able to obtain 15% butanol selectivity and 56 % conversion at 260 o C in a fixed bed reactor. Generally, the most common way of obtain ing a catalyst conta ining both magnesium and aluminum oxides is the calcination of hydrotalcite. Hydrotalcite is a basic layered double hydroxide (LDH) support with 11 Al positive charges neutralized between anionic hydroxide layers of Mg(OH) 2 . Calcination of hydrotalcites leads to the collapsing of this structure and results in a mixture of Mg/Al oxides. One of the advantages of using calcined hydrotalcite is to obtain an ideal number of acid and base sites on the catalyst, as the Mg/Al ratio can be adjusted as desired [ 15 , 23 , 42 - 47 ] . Di Cosimo et al. investigated different Mg /Al ratios for the Guerbet reaction and stated that 5>Mg/Al>1 is the ideal ratio [ 23 , 42 ] . Higher Mg conc entrations (more basicity) lowered the condensation reaction rate, and higher Al concentrations (more acidity) increased ethanol dehydration to ethylene and its coupling and dehydration to diethyl ether. Leon et al. confirmed this finding by stating that a cid sites are responsible for the ethanol dehydration reaction to ethylene, limiting the selectivity of ethanol to higher alcohols [ 43 , 44 ] . They also noted that at lower temperatures the basic sites are active for the condensation reaction, while acid sites are not active for dehydration reactions [ 45 ] . Another benefit of calcined hydrotalcite is the surface exchangeability of Mg and Al atoms with other metals to form other bi - metal or multi - metal mixtures. This characteris tic has provide d unique flexibility for this support. For instance, Leon et al. noticed that surface acidity can be reduced by replacing Al atoms with Fe, which leads to lower ethylene formation rates and increased desired products selectivity subsequently [ 44 ] . On the other hand, some researchers studied the eff ect of the addition of other metals by physically mixing them with calcined hydrotalcite. Carlini et al. mixed this catalyst with copper chromite for i - butanol production from methanol and n - propanol [ 46 ] . They were able to obtain 30% n - propanol conversion and almost complete iso - butanol selectivity at 200 o C and atmospheric pressure in the batch system . The same grou p tried to optimize this result using the impregnated Cu/Mg - Al mixed oxide, and were able to enhance the iso - butanol yield to 40% while maintaining complete selectivity [ 48 ] . They also 12 reported complete selectivity to iso - butanol and 80% n - propanol conversion with a vapor phase continuous reactor at 280 o C. Carlini et al. reported that their results were exclusive to the addition of copper, and no optimization effect was observed by precipitating other dehydrogenating metals such as Ni, Pd , Ag, and Pt. Several other studies have confirmed the positive impact of copper addition on vapor phase ethanol dehydrogenation to desired intermediate products [ 49 , 50 ] . A similar study made by Marcu et al. for the condensed phase Guerbet reaction showed the optimum loading of copper in calcined hydrotalcite to be between 5 wt% and 10 wt% for 9% ethanol conversion and 80% butanol selectivity [ 2 0 ] . Also, among various metals studied (Pd, Ag, Mn, Fe, Cu, Sm, and Yb), palladium had shown the best results, with 18% ethanol conversion 78% butanol selectivity at 300 o C and autogenous pressure [ 21 ] . A ccording to this study, Pd - Mg - Al had stronger basicity and lower acidity compared to Cu - Mg - A l catalyst. However, other metal combinations, such as Sm - Mg - Al, that had stronger basicity and lower acidity than Pd catalyst, did not perform as well as this cata lyst, confirming the imp ortance of base/acid ratio in the catalytic condensation of ethanol. Zhang et al. obtained 70% n - butanol selectivity and 28% ethanol conversion for Pd - Mg - Al catalyst at 290 o C in a vapor phase batch reactor [ 51 ] . 1.4.4 Hydroxyapatite (HAP) Hy droxyapatite (HAP) or Ca 10 (PO 4 ) 6 (OH) 2 is a catalyst with both acid and base functionalities studied in several papers for the Guerbet reaction. The molar ratio of C a/P can be controlled by adjusting the pH of the mixture solution while preparing the cataly st. This ratio corresponds to the base/acid sites on the catalyst surface. Tsuchida et al. found best butanol selectivity of 71% at 10% ethanol conversion for Ca/P molar ratio of 1.67, corresponding to a base/acid molar density ratio of 88 [ 39 ] . Hanspal et al. obtained 75% butanol selectivity and 24% acetaldehyde selectivity at 13 only 7% conversion at 330 o C and atmospheric pressure in a fixed bed reactor [ 52 ] . They conducted an experiment at similar conditions using basic MgO catalyst and obtained 40% and 49% selectivity for butanol and acetaldehyde, respectively, at a lower catalyst activity. They explained this result by performing acid - base analyses on both catalysts and observing higher density of both base and weak acid sites on hydroxyapatite. Finally, they compared different acetaldehyde concentrations at the reactor exit, and observed a linear depende nce of butanol f ormation on acetaldehyde concentration for MgO, and no specific relation between these two for the hydroxyapatite cata lyst. This suggests that the vapor - phase acetaldehyde participates in butanol formation solely for the MgO catalyst [ 53 ] . Similar to the calcined hydrotalcite, one of the advantages of hydroxyapatite is its flexibility of substituting calcium, phosphorous, and hydroxide atoms with other atoms such as strontium, vanadium, and fluorine, respectively. Ogo et al. examined four catalysts using this method (Ca - P HAP, Sr - P HAP, Ca - V HAP, and Sr - V HAP) and observed the best results with the Sr - P HAP catalyst [ 24 ] . T hey were able to obtain 81% butanol selectivity at 8% conversion at 300 o C and atmospheric pressure in a fixed - bed continuous reactor. This group optimized these results by tuning the Sr/P molar ratio and were able to obta in 86% butanol selectivity at Sr/P molar ratio of 1.70 corresponding to base/acid molar den sity of 4.5. A lthough the conversion of ethanol was improved to 11% using this catalyst, it was still too low for considering this catalyst as an ideal one for the Guerbet reaction. Furthermore, this conversion was obtained at a relatively low WHSV of 0.35 h - 1 , which indicates the high energy required to maintain the reaction conditions for a long time. Longer contact times also increase the chance of catalyst deactivation as a result of dimerization and polymerizati on of intermediate products such as acetaldehyde [ 26 ] . One of the methods of improving ethanol conversion while maintaining the WHSV in an acceptable range is 14 to increase the reaction temperature. However, this will lead to ethanol decomposition and side - reactions. For instance, Tsuchida et al. reported 76% ethanol conversion and 6% butanol selectivity at 450 o C [ 54 ] . Hansp al et al. recently studied the effect of cation and anion components on the performance of this catalyst by replacing Sr with Ca and OH - with F - . They conducted their experiments at 633 K and atmospheric pressure in a fixed bed reactor and obtained most op timum results with HAP at 5% ethanol conversion and 72% butanol selectivity, which emphasizes the beneficial role of the hydroxyl group of HAP. Among different catalysts tested in this s tudy, Sr - P HAP showed the highest rate of ethanol dehydrogenation to a cetaldehyde (91% acetaldehyde selectivity), with a quite low coupling to butanol observed. Low butanol formation rate is assigned to the weak binding affinity for acetaldehyde on this ca talyst. They also noticed that weaker base sites in HAP catalyst compa red to MgO are responsible for reversible water adsorption on HAP, which enables the catalyst to stay active for the aldol condensation reaction at a higher rate [ 55 ] . Silvester et al. found the optimized ethanol conversion of 40% and 83% selectivity to butanol at the base/acid molar density ratio of 0.2 [ 56 ] . Similar to most of the other studies using this catalyst, this selectivity was obtained in the gas phase and high contact t imes. 1.4.5 Al 2 O 3 Supported Catalysts Metal - supported alumina is one of the more recent catalysts suggested for the Guerbet reaction. This alumina support contains both acid and base sites, and the addition of metal sites improves hydrogen transfer at the reacti on conditions. Ndou et al. were one of the first research groups that studied this type of catalyst [ 17 ] . Using different alkali metals supported on alumina at 450 o C and atmospheric pressure, they were not able to obtain a selectivity of more than 4% for C 4 alcohols. Previous st udies on h ydrotalcite - based catalysts had shown the ability of nickel to facilitate the Guerbet reaction at lower temperatures [ 57 ] . Yang et al. tested several metallic (Fe, Co, and Ni) 15 catalyst s over - Al 2 O 3 support; they obtained 19% ethanol conversion and 64% butanol selectivity at 200 o C and atmospheric pressure in a packed bed reactor using 8 wt% Ni/ - Al 2 O 3 [ 58 ] . Other liquid side - products were acetaldehyde, butyraldehyde, and ethyl acetate. Fe/ - Al 2 O 3 suppressed the activity of the reaction compared to the Ni/ - Al 2 O 3 - Al 2 O 3 was as - Al 2 O 3 ; however, unlike nickel catalyst, ethyl acetate was the main product detected using cobalt . Finally , the same group stu died di - Al 2 O 3 catalyst and obtained the highest butanol yield at 8 wt% nickel loading. High pressure condensed phase reactions are popular conditions for the Guerbet reaction using alumina - supported catalysts. G haziask ar et al. studied the Guerbet reaction at different pressure (4 - 183 bar) and temperature (150 - 300 o C) ranges and were able to get C 4 + selectivity as high as - Al 2 O 3 catalyst at 250 o C and 176 bar [ 8 ] . They claimed that the catalyst is active for at least 18 hours without any need for regeneration. Riittonen et al. conducted Guerbet experiments in a conden sed phase batch reactor system at 250 o C and autogenous pressure using different metallic catalyst - Al 2 O 3 support [ 22 ] . Maximum butanol selectivity of 80% of liquid products at 25% ethanol conversion - Al 2 O 3 at long contact times (24 hours). The second best performance was with the Pt cata lyst. Later, the same group studied Ni, Co, and Cu catalysts over - Al 2 O 3 in a condensed phase fixed bed reactor at 240 o C, 70 bar, and LHSV of 4.3 h - 1 [ 59 ] . While Co catalyst showed the highest activity (28% ethanol conversion), Ni provided the highest selectivity toward butanol (69% of liquid products). C obalt results showed the highest selectivity - Al 2 O 3 were also studied, and the results indicated that increasing Cu loading would inhibit the selectivity toward butanol, and intro duce ethyl acetate as the main product. Due to the absence of qualitative and 16 quantitative analysis of gaseous products in these experiments, it is difficult to compare the selectivity results to similar studies. Aside from ethanol reactions, Al 2 O 3 - suppor ted metallic and bimetallic catalysts have also been used for the Guerbet reaction with other alcohols as the feed. Panchenko et al. compared the n - pentanol Guerbet reaction activity of different metallic catalysts (Pt, Pd, Ir, Ru, Rh) over different suppo rts (C, Al 2 O 3 , TiO 2 , CeO 2 , ZrO 2 ) doped with different solid bases (NaOH, MgO, Al(OH) 3 , Na 2 CO 3 , CaCO 3 ) in a condensed phase batch reactor at 180 o C and 10 bar [ 60 ] . Among the different catalysts studied, the highest conversion (12 %) and selectiv ity (90%) to 2 - propyl - 1 - heptanol was achieved using Pt/Al 2 O 3 in the presence of aqueous NaOH solution. Hernandez et al. studied the Guerbet reaction of n - octanol using co - impregnated 10 mol% Ni/Cu (3/1)/Al 2 O 3 at 225 o C and atmospheric pressu re in a batch r eactor [ 61 ] . They were able to obtain 95% n - octanol conversion and 75% C 16 alcohol selectivity after six hours of reaction. Although they suggested a new method for the preparation of this bimetallic catalyst, the classic wet impregnation of metal nitrates on the sur face of the support resulted in t he highest selectivity toward desired products. Due to the importance of acid - base surface interactions in the Guerbet reaction, obtaining the right - Al 2 O 3 support is a critical step in optimizing this catalyst. Ghaziaskar et al. showed that mixing basic Mn 2 O 3 - Al 2 O 3 - Al 2 O 3 resulted - Al 2 O 3 catalyst [ 8 ] . A similar observation was made by Ogo et al. during their experiments with HAP catalyst [ 18 ] . Jordison et al . modified the base sites - Al 2 O 3 catalyst by impregnating different amounts of La 2 O 3 . They observed optimum ethanol conversion of 55% and C 4 + selectivity of 71% for the 8 wt% Ni/9 wt% La 2 O 3 - Al 2 O 3 catalyst in a condensed phase batch react or at 230 o C, autogenous pressure, and 10 hours of reaction time [ 9 ] . They noticed that the addition of La 2 O 3 inhibited the rate of formation of ethyl 17 acetate produced by the reaction of acetaldehyde with ethanol . The biggest side - products in this system were gases such as CH 4 and CO 2 . The same authors studied the effect o f water removal on the Guerbet reaction pe rformance and were able to achieve 75% C 4 + selectivity and 50% ethanol conversion under similar reaction conditions and a water concentration of 5 wt% [ 25 ] . 1.4.6 Other Catalysts Similar to Al 2 O 3 , CeO 2 is a more recently used support for Guerbet reactions. This catalytic support is commonly incorporated with copper metal to enhance hydrogen transfer. In co mparison with - Al 2 O 3 supported catalysts, this catalyst gets activated at lower temperatures, with the drawback of being less selective to higher alcohols [ 26 ] . Earley et al., used the Cu - CeO 2 catalyst to achieve 67% ethanol conversion and 45% butanol selectivity at 260 o C and 100 bar pressure of CO 2 [ 62 ] . Jiang et al. impregnated copper on CeO 2 /AC (activated carbon) surface and obtaine d up to 46% ethanol conversion and 42% butanol selectivity at 250 o C and 20 bar [ 63 ] . Both experiments w ere conducted in continuous reactors. Activated carbon has also been used with other metals, metal oxides, and alkali metal salts for the Guerbet reaction. Onyestyak et al., obtained optimum results by modifying this support with nickel and KOH [ 64 , 65 ] . A maximum higher alcohols yield of 62% at 350 o C and 21 bar was reported for this catalyst. Mixed oxides have always been popular options for the Guerbet reaction. The use of Mg - Al mixed o xides has already been discussed in detail in previous sections. Gines et al. studied Mg/Ce oxides for the gas phase Guerbet reaction and observed the constructive effect of doping potassium for increasing the number of basic sites, and of impregnating cop per for enhancing the hydrogen transfer steps [ 19 ] . Unfortunately, they did not report any conversion/selectivity for their experiments. Another mixed oxide used for this reaction is Mg/Zr mixed oxide. Regardless of the 18 disappointing c atalytic performance of this material (8% ethanol conversion and 12% butanol selectivity), Kozlowski et al. showed the beneficial effect of adding sodium as the basic subs trate to this catalyst [ 37 ] . Another catalyst reported for the va por phase Guerbet reaction is alkali - zeolite catalysts. First introduced by Yang in 1993 [ 28 ] , other scientists h ave since developed it further. Yang et al. studied various Rb - impregnated alkali - zeo lite catalysts and found the optimum catalytic activity for Rb - LiX zeolite at 1 bar and 420 o C. Yoshioka et al. studied different catalysts for the ethanol Guerbet reactio n at different pressures and temperatures and were able to obtain up to 12% butanol y ield at 0.2 bar and 275 o C using Rb - Li ion - substituted zeolite catalyst [ 66 ] . Authors have not repo rted any conversion/selectivity results in this study. 1.5 Condensed Phase Reactions The critical point that differentiates alumina - supported catalysts from other catalysts used for the Guerbet experiments is their activity at condensed phase conditions. Being able to run the Guerbet experiments at higher pressures and lower temperatures not o nly forces the reaction thermodynamically toward the production of the desired products, but also saves the cost and energy required for bringing the feed to the reaction conditions. Several studies have been conducted in recent years to improve the perfor mance of the Guerbet reaction in the condensed phase using alumina - supported catalysts. These studies have been discussed in Section 1.4.5 . 1.6 Co ntinuous Reactors Continuous reactors are generally favored in handling heterogeneous catalytic reactions. They are easier to set - up and control, more environmentally friendly, and subsequent product separations and catalyst regeneration can be achieved in a shorter time an d lower cost compared to batch 19 reactors [ 15 , 67 ] . Moreover, continuous reactors are easy to scale - up, which is a big advantage for this type of process [ 8 ] . Wiles et al. discuss that besides quality, economical, and environmental advantages, sa fety perspectives are also superior in continuous reactors compared to batch processes, since heating systems and tempe rature control in these systems are more accurate [ 67 ] . All these mentioned benefits have made continuous processes more in dustrially relevant, especially for heterogeneous catalytic reactions. In Chapter 2 of this research, the continuous pr ocess for the catalytic Guerbet reaction has been optimized. In Chapter 3, the scale - up of this process has been studied and optimized in different aspects. 1.7 Nickel Bimetallic Catalyst The mechanism of the Guerbet reaction has been discussed in Section 1.3 . Hydrogen is a crucial component in the system, as it is involved in two steps of the reacti on. As discussed earlier, metallic sites are responsibl e for hydrogen adsorption (desorption) on (from) the surface of the catalyst. Liberation of the molecular hydrogen produced in the dehydrogenation step from the surface of the metal to the reaction env ironment can desirably shift the first step of the reac tion toward the formation of more acetaldehyde in one hand, and, on the other hand, create a challenge for the hydrogenation of the aldol condensation product (i.e. crotonaldehyde) toward the desired a lcohol. Conversely, a strong hydrogen bond with the met al surface can restrain the hydrogenation of crotonaldehyde (and other aldol condensation products) to higher alcohols [ 68 , 69 ] . Therefo re, finding a balanced number and strength of metal - hydrogen bonds are the most critical challenge s in obtaining an optimum selectivity for the Guerbet reaction. Furthermore, nickel is known for its strong capability of cracking the carbon - carbon bond pres ent in ethanol, which leads to the decomposition of ethanol to gaseous side - products. Therefore, nickel - based bimetallic catalysts 20 are potential alternatives to investigate the possibility of improving the performance of the Guerbet reaction. 1.8 Obj ectives 1.8.1 P rocess and Catalyst Studies on the Guerbet Reaction Experimental studies on the Guerbet reaction in this study divides into three main categories. These topics are discussed in detail in the following sections. 1.8.1.1 Develop Ni/La 2 O 3 - Al 2 O 3 C atalyst i n a Conti nuous Condensed Phase Reactor Jordison et al. [ 9 , 25 ] have already studied Ni/La 2 O 3 - Al 2 O 3 catalyst in batc h reactors. However, no extensive work has been done so far for using this catalyst in a continuous reactor. Transitioning from batch to continuous reactors require a set of calculations to make sure that the starting point of experiments is in the right r ange of feed flow, reaction temperature, and catalyst quantity. Moreover, safety issues need to be considered and reviewed, since the operating process is different from the batch reaction. Besides designing a continuous condensed phase process for the Gue rbet reaction system, finding the ideal reactor configuration and operating conditions are the other objectives of this research. For this purpose, multiple reaction temperatures, feed flow rates, and catalyst compositions with different preparatio n condit ions will be tested, and the optimum reaction conditions will be investigated . Multiple surface analyses will be obtained to address the catalytic observations made throughout the research. The results obtained from this section will be the basis o f the st udies conducted on bimetallic catalysts, as discussed in the next section. 21 1.8.1.2 Improve the Reaction Yield by Finding the Ideal Bimetallic Catalyst Improving the rate of formation of higher alcohols compared to those of the side - reaction in the system i s the pr imary objective of studying bimetallic catalysts. On the effect of the addition of second metal to nickel on hydrogen chemisorption and desorption, palladium and platinum are metals that are recognized to enhance this amount. Ni - Pt and Ni - Pd bimeta llic cat alysts show lower hydrogen desorption temperatures than the monometallic nickel, meaning that they form weaker hydrogen bond strengths compared to nickel [ 70 - 73 ] . Addition of copper to nickel could suppress the amount of hydrogen chemisorbed (desorbed) to (from) the surface of the catalyst, since copper is known as a metal that does not promote dissociative hydrogen adsorption [ 74 ] . Similar to copper, molybdenum and iron suppress the amount of hydrogen uptake of the nickel catalyst, and cobalt does not significantly change this amount [ 73 , 75 ] . Copper and cobalt have also shown a weaker capability in cleaving the C - C bon d compared to nickel. These two metals increase the resistance toward carbon nucleation and growth on the surface of the nickel - containing catalysts that could potentially help in inhibiting the decomposition of ethanol to gaseous side - products [ 76 - 78 ] . Finally, according to previous studies [ 59 ] , one of the contributing factors in determining product selectivity is claimed to be the crystal structure of the metals. Therefore, adding a second metal could optimize the structure of the cataly st toward the production of more desired products. To cover as many varieties of catalyst optimization cases as possible, Cu, Co, Pd, Pt, Fe, and Mo bimetallic combination with Ni are examined in this study. 1.8.1.3 Understand the Mechan istic and Kinetic Behavior of the Reaction As discussed earlier, the mechanism of the Guerbet reaction has been a controversial topic among researchers in the field. Ethanol multi - step conversion to acetaldehyde, crotonaldehyde, and finally butanol is the more accepted mechanism, wh ile some researchers have not disregarded the direct 22 dehydration of ethanol. In this study, we design and construct experiments to better understand the governing mechanism for the condensed - phase Guerbet reaction using the nicke l - based La 2 O 3 - Al 2 O 3 catalysts. Besides the mechanistic studies, other experiments have been performed to understand and predict the behavior of the reaction more accurately. For this purpose, experiments with the intermediate and final products as the feed of the rea ction, a long with the results obtained from surface analysis, are employed to discuss the rate - limiting step and activation energy for the reaction. 1.8.2 Techno - economic analysis of the industrial scale Guerbet reaction A process concept is presented in which n - butano l and mixed C 6 + alcohols are produced as saleable products; ethanol is recycled to achieve nearly 100% overall conversion and minor byproducts are burned to provide process energy. A process design is conducted using Aspen Plus V8.4 process simulat ion soft ware, and economic analyses are carried out for several cases of ethanol conversion and alcohol selectivities. Several additional cases involving permutations of the base process configuration are also examined in attempts to improve process econom ics. At n - butanol selectivities achieved experimentally and for a facility producing 75 million kg n - butanol per year, the total capital costs, operating expenses, and the required n - butanol selling price for typical values of expected return on investment have be en calculated [ 79 ] . 1.8.3 Guerbet reaction studies of the Fusel Alcohols Guerbet reactions were also conducted with fusel alcohols obtained from collaborating research groups in the proje ct. Beca use the fusel alcohols obtained for these experiments contained approximately 5 wt% isoamyl alcohol (2 - methyl - 1 - butanol, IA) in ethanol, reactions were carried out with blends of IA alcohol and ethanol of different compositions in order to better u nderstan d 23 reaction rates and se lectivity to desired products. Reactions have been carried out in both batch autoclave reactor and in the continuous condensed - phase catalytic reactor. 1.8.4 Enhanced Acrylate Production from 2 - Acetoxypropanoic Acid Esters Acrylic acid and its esters are the starting materials for the production of polymers that are widely used in adhesives, paints, coatings, diapers, dispersants, etc. Acrylates are traditionally produced via a two - step propylene oxidation process, but recently thei r produc tion from renewable feedstocks is receiving significant attention. Glycerol, 3 - hydroxypropanoic acid, and lactic acid (2 - hydroxypropanoic acid) have all been investigated as feed sources; of particular interest here is the formation and subsequent pyrolysi s of 2 - acetoxypropanoic acid esters (APA esters). Although this route has been known for a long time, our ability to achieve near - quantitative yields of APA and its esters from lactic acid and acetic acid via reactive distillation provides an incen tive for further study of the pyrolysis step. Thus, reactor configurations and conditions for converting APA esters to acrylic acid esters in high yields have been examined and identified in this study [ 80 ] . 24 2 Ethanol Guerbet Reaction 2.1 Introduction H igher alcohols , such as n - butanol, can be produced via condensation of ethanol , also known as the Guerbet reactions. The avail ability of bioethanol is a major motivation for n - butanol production via this route [ 15 ] . Two possible reaction mechanisms for Guerbet reactions have been proposed : direct dehydration of the alcohols [ 17 , 28 , 81 ] , and the more generally accepted three - step mechanism involving dehydrogenation, aldol condensation, an d hydrogenat ion reactions. The latter is especially favored at lower reaction temperatures and over metal - containing catalysts [ 9 , 15 , 25 ] . Several heterogeneous and homo geneous catalytic systems have been proposed for Guerbet reactions. Heterogeneous catalysts are generally preferred due to lower separation costs and fewer environmental difficulties [ 15 , 26 ] . Chief among these catalysts are MgO [ 17 , 37 , 38 ] , multi - metal mixed oxides [ 21 , 82 , 83 ] , hydroxyap a tite [ 24 , 29 , 39 , 84 ] , alkali exchange d zeolites [ 28 , 66 ] , and alkaline activated carbon - supported catalysts [ 64 , 65 ] . While most studies have failed to exceed a C 4 + alcohol selectivity of 75% with ethanol co nversions above 40%, there is consensus that an ideal acid - base balance is the key to the desired catalyst activity and higher alcohols selectivity. Nickel metal supported on alumina has been recently introduced for the Guerbet reaction [ 8 , 22 , 59 ] . A significant property of this catalyst is its activity at lower temperatures (<250 o C ), whic h not only allows the reaction to be run in the condensed phase at elevated pressures but also slows ethanol decomposition to undesired gaseous products such as CO 2 and CH 4 . Jordison [ 9 , 25 ] further improved conversion and selectivity of - Al 2 O 3 by adding La 2 O 3 onto the support. This 25 modified catalyst gave a higher alcohol selectivity of 71% at an ethanol conversion of 55% in a stirred autoclave reactor at 230 o C and autogenous pressures. In this chapter, we desc ribe experiments conducted on the continuous condensed - phase Guerbet reaction. Several experiments have been perf ormed to find the ideal reactor configuration and catalytic behav ior of the Ni/La 2 O 3 - Al 2 O 3 catalyst. Later, several catalyst s of bimetallic combinations with nickel on the La 2 O 3 - Al 2 O 3 support are studied to further optimize the catalytic performance of the reaction . Surface analysis and materials characterization techniques have been employed to assist with the catalytic studies. Finally, some experiments have been performed to assist with and confirm the proposed mechanism for the reaction. 2.2 Materials and M ethods 2.2.1 Materials and C atalyst P reparation Anhydrous ethanol (Koptec, 200 proof) was used as the feed in all experiments. In the mono - metallic experiments, Ni(NO 3 ) 2 . 6H 2 O (99.999%, Aldrich) and La(NO 3 ) 3 . 6H 2 O (>99%, Fluka) were used as catalyst precursors, and spherical 1.6mm diameter - Al 2 O 3 (Strem Chemical) was used as the catalyst support. Three composition s of monometallic catalysts were prepared for this study: 8 .0 wt% Ni/ 9 .0 wt% La 2 O 3 / - Al 2 O 3 , 8 .0 wt% Ni/4.5 wt% La 2 O 3 / - Al 2 O 3, and 1 .0 wt% Ni/9 .0 wt% La 2 O 3 / - Al 2 O 3 . The first compos ition of catalyst was prepared three different times with different impregnation and calcination times ; that has affected the catalytic performance of the reaction. This will be discussed further in the following sections. Additionally, the first com positi on of the nickel catalyst (8 .0 wt% Ni/9 .0 wt% La 2 O 3 / - Al 2 O 3 ) was once made in big amounts (>500 g) to use for bimetallic catalyst preparation. This method of catalyst impregnation, named here as separate impregnation, assists with the integrity of the results while comparing 26 different bimetallic mixtures wi th each other. In addition to the separate impregnation method , some catalyst s were prepare d by impregnating nickel and the second metal at the same time on the surface of the catalyst, named here as the co - impregnation method . For the bimetallic experime nts, Ni(NO 3 ) 2 ·6H 2 O (99.999%, Aldrich), Cu(NO 3 ) 2 ·2.5H 2 O (98%, Sigma - Aldrich), Co(NO 3 ) 2 ·6H 2 O (98%, Sigma - Aldrich ), Pd(NO 3 ) 2 ·x H 2 O (40% Pd basis , Aldrich), Pt(NH 3 ) 4 (NO 3 ) 2 (99.995%, Sigma - Aldrich), Fe(NO 3 )· 9H 2 O (98+%, Sigma - Aldrich), NH 4 ·Mo 7 O 24 ·4H 2 O (81 - 83% MoO 3 basis, Sigma - Aldrich), and La(NO 3 ) 3 ·6H 2 O (>99%, Fluka) were used as catalyst precursors. Catalysts were prepared using incipient wetness impregnation according to prior work [ 9 ] . To the alumina support, La 2 O 3 was first impregnated by adding La(NO 3 ) 3 solution containing the desired quantity of lanthanum to the support in a quantity equal to the support pore volum e, followed by drying at 130 o C for 20 h and calcination at 600 o C for 18 h in 50 ml /min N 2 , to ensure the presence of La 2 O 3 on the support surface. The same process was used for the addition of nickel and the second metal (starting with their aqueous star ting material described above) to the La 2 O 3 / - Al 2 O 3 support, with an additional step of reducing metal oxides to the metal at 520 o C for 18 h using 50 ml /min H 2 at 1 atm. A complete description of the c alculations and steps taken for the preparation of catalysts are described in Appendix A . All of t he bimetallic catalysts in this study are prepared and reported on a weight basis . However, for simplicity, Table 2 . 1 shows the composition o f metals in each of the bimetallic catalysts prepared in both mass and mo lar basis. 2.2.2 Reactor System Reactions were performed in a 1.91 cm OD (1.57 cm ID) × 76 cm length jacketed 316 stainless steel up - flow packed bed reactor with a 3 mm OD internal thermowe ll to measure the temperature profile in the reactor during the reaction . The reactor was heated with silicon oil using a Julabo 27 (Model SE - 6) heating circulator, with the reaction temperature ranging from 17 0 o C to 250 o C. Silicon carbide (SiC, 20 - 50 mesh) was packed in the reactor inlet for preheating the feed, and stainless s teel rod fillers were used on the top and bottom of the reactor to reduce the dead space inside the reactor. Figure 2 . 1 shows a schematic of the reactor system. Table 2 . 1 . Bim etallic catalyst composition s on a mass and molar basis. The balance is - Al 2 O 3 in all cases. The data on the f irst row is for the 8 .0 wt% Ni/ 9 .0 wt% La 2 O 3 / - Al 2 O 3 catalyst that second metals are impregnated on . M is the second metal. Ni - Cu bimeta llic is made in two different metal ratios. M Mass % Mole % Ni/M Ni M Ni + M La 2 O 3 Ni/M Ni M Ni + M La 2 O 3 - - 8 - - 9 - 13.9 - - 2.8 Cu 2 7.7 3.8 11.5 8.7 2.2 13.1 6.0 19.1 2.7 Cu 20 8.0 0.4 8.4 9.0 22.7 13.6 0.6 14.2 2.7 Co 2 7.7 3.8 11.5 8.7 2.0 13 .0 6.5 19.5 2.6 Pd 4 7.8 1.9 9.7 8.8 7.2 13.7 1.9 15.6 2.8 Pt 4 7.8 1.9 9.7 8.8 13.8 13.8 1.0 14.8 2.8 Fe 2 7.7 3.8 11.5 8.7 1.9 13.0 6.8 19.8 2.6 Mo 2 7.7 3.8 11.5 8.7 3.4 13.4 4.0 17.4 2.7 Figure 2 . 1 . Continuous flow reactor for ethanol conversion. (a - stainless steel rods; b - SiC packing for feed preheating; c - catalyst bed; d - thermowell; e - oil jacket) 28 Typically, 35 g of SiC and 29.9 g of catalyst supported o n a quartz wool plug were placed in the reactor. The reactor was assembled and connected to the oil bath, and nitrogen gas was passed through the catalyst to purge air and ensure the absence of any leaks in the reactor system and connections. Then hydrogen gas was passed through the catalyst for 90 min after it reached the desired temperature to reduce surface nickel oxidized by exposure to air. A BioRad (Model 1350) liquid chromatography pump was used for dispensing liquid ethanol feed to the reactor. Liqu id flow rates varied from 0.3 ml/min to 1.3 ml /min, corresponding to a weight hourly spac e velocity (WHSV) of 0.5 - 2.1 kg ethanol/kg catalyst/h. A pressure relief valve was connected to the reactor outlet. The reactor effluent was cooled to ambient temperat ure in a double pipe heat exchanger using building water as the coolant. A Tescom (Model 26 - 1764 - 24) back pressure regulator was used to control reactor pressure at 100 bar and reduce the effluent pressure to near atmospheric. Following pressure reduction, condensable products were recovered in collection vessels submerged in an ice bath (0 o C), and gaseous products that passed through the collection vessels were collected periodically in gas bags. Steady state was assumed to be achieved after 6 - 8 superfici al residence times of feed materials through the reactor. Initial experiment s were condu c ted at the feed flow rate of 1.10 ml /min of ethanol at 230 o C using 29.9 grams of catalyst. This value was based on the observed conversion rate by Jordi son et al. [ 9 ] in the batch reactor experiments using the same catalyst to give a reasonable conversion in the flow system . Details of the calculations of obtaining this number are available in Appendix B . The flow rate and temperature of the reaction w ere further modified in the later experiments to achieve optimum values of ethanol conversion and C 4 + selectivity. 29 2.2.3 Analytical M ethods Liquid product samples were diluted 10 - fold in acetonitrile and analyzed using a Varian 450 gas chromatograph (GC) with a flame ionization detector. A 30 m SolGel - Wax column (0.53 mm ID, 1 mm film thickness) was used with th e following temperature p rogram: initial temperature 37 o C for 4 min; ramp at 10 o C /min to 90 o C , and hold at 90 o C for 3 min; ramp at 10 o C /min to 150 o C ; ramp at 30 o C / min to 230 o C and hold for 2 min. Butyl hexanoate 1% solution was used as an internal standard to improve the p recision of the analytical calculations. Gas samples were analyzed using a Varian 3300 GC with a thermal conductivity detector. A 4.57 m 1.25 mm SS 60/80 Carboxen 1000 column (2.1 mm ID) was used with the f ollowing temperature program: initial temperature 35 o C for 5 min, then ramp at 20 o C /min to 225 o C . The concentration of each species in the product mix was determined using response factors obtained from multi - point calibration curves. Response factors for unidentified components appearing in the chrom atogram were taken as average values for species in close proximity to unidentified components. The calculated concentrations were entered into an in - house Excel spreadsheet to calculate ethanol conversion, product selectivity (mol ethanol to product/mol e thanol converted), product yield (mol ethanol to product/mol ethanol fed ), and overall carbon recovery. Karl Fischer titration was used to determine the water content in the feed and reaction products. Thre e titrations of each sample were employed to confi rm the accuracy of the results. 2.2.4 Catalyst Characterization Total ( BET ) surface area, pore volume, and pore diameter measurements were done using a Micromeritics ASAP 2010 Plus Physisorption apparatus with nitrogen adsorption at - 195 o C. Prior to analysis, s amples were degassed at 150 o C for 24 h . 30 Acid and base site densities and H 2 uptake of the catalysts were measured using Micromeritics Autochem II chemisorption analyzer. Ammonia and carbon dioxide temperature programmed desorption technique (TPD) were use d for acid and base site density measurement, respectively. Catalysts were outgassed under helium flow by ramping the temperature at 10 o C/min to 600 o C and holding it at 600 o C for 60 minutes , followed by cooling the sample at 90 o C/min to 25 o C and holdi n g it for 10 minutes. A 50 ml/min flow of ammonia or carbon dioxide was passed across the samples for 30 minutes at 25 o C . The physisorbed gases were cleaned by the 50 ml/min flow of helium at 25 o C for 180 minutes. Chemisorbed gases were desorbed by rampi ng the temperature at 10 o C /min to 600 o C and holding it at 600 o C for 180 minutes under the 50 ml/min flow of helium ; the amount desorbed was measured by recordin g and integrating the derived intensity signal. For measuring the hydrogen uptake , samples we re outgassed by the flow of argon ramping at 10 o C/min to 600 o C and holding it at 600 o C for 60 minutes, followed by cooling it at 30 o C/min to 27 o C and holding it at 27 o C for 5 minutes. The hydrogen gas at the flow rate of 25 ml/min was passe d across t he catalysts at the same temperature for 120 minutes. Adsorbed hydrogen molecules were desorbed by 25 ml/min flow of argon ramping at 25 o C/min to 600 o C and holding it at 600 o C for 180 minutes. Desorption intensity signals were recorded and integrated fo r calculating the H 2 uptake of the catalysts. The surface elemental distribution of the catalysts was monitored and measured via scanning electron microscopy (SEM) coupled with energy - dispersive X - ray spectroscopy (EDS) using Carl Zeiss V ariable Pressure S EM EVO LS25 at high vacuum mode. Although s amples were sputter - coated with platinum to avoid their oxidation at room temperature before analysis , some oxidation was observed during the transition of the catalysts from preparation reactor to the sputter are a . 31 This oxidation affected the SEM images quality, but not the molar ratios obtained from EDS analysis. 2.3 Experimental results 2.3.1 Control Experiments Control experiments with La 2 O 3 / - Al 2 O 3 gave less than 2% ethanol conversion and less than 20% selectivity to desired condensation products, an indication that the metal - free support is inactive, and that Ni plays a key role in Guerbet reactions. Inert gas (N 2 ) and liquid ( t - butanol) fee d s were also fed to the reactor at typical reaction conditions (T= 230 o C, WHSV= 1.42 h - 1 ) to en sure that the mass balance close d to 100% and there wa s no leak in the system. 2.3.2 Nickel Monometallic Catalytic Experiments Throughout the discussions on this cha pter, all produced alcohols with more than six carbons are described as a single product denoted as C 6 + alcohols. In most of the experiments, C 6 + alcohols contain a similar product distribution. Figure 2 . 2 shows th e typical molar distribution of C 6 + alcohols in most of the experiments conducted. 62% 27% 7% 4% 1-hexanol 2-ethyl-1-butanol 1-octanol 2-ethyl-1-hexanol Figure 2 . 2 . Typical C 6 + alcohols molar distribution. Negligible amounts of 1 - decanol are also present. 32 The overall material balance ( = outlet flow/inlet flow × 100) had a closure of 95% - 101% in all of the experiments . Furthermore, carbon backbone recovery for most of the exp eri ments was in the range of 94% - 99 %. This recovery is reported for each single experiment via the total reaction selectivity in Appendix D.3 and Appendix E.3. Finally, to as sure that most of the water present in the system is the product of the Guerbet re action, the theoretical amount of water formed in samples were calculated based on the stoichiometric conversion of the moles of higher alcohols and aldehydes . This value was divided by the water content from Karl Fischer titrations for some representative samples at different ethanol conversion ranges. For most of the cases, this ratio was in the range of 75% - 85%. The difference could be due to the occurrence of water produci ng side - reaction , such as diethyl ether and methane production . Reactor configurat ion for the 8 .0 wt% Ni/4.5 wt% La 2 O 3 / - Al 2 O 3 catalyst was studied to find the range of parameters at which this catalyst and similar ones demonstrate optimum performance for the Guerbet reaction. For this purpose, ethanol conversion and higher alcohols sel ectivity were studied at different temperatures and feed flow rates. Figure 2 . 3 shows the performance of the Guerbet react ion at different temperatures. At the temperature of 210 o C , the maximum C 4 + alcohols selec tivity of 71% is observed . Furthermore , ethanol conversion and gases selectivity increase with increasing temperature . Based on the ethanol conversion at different temperatures and assuming second order kinetics for the reaction, an activation energy w as c alculated for this catalyst. The activation energy obtained for the Guerbet reaction and the 8 .0 wt% Ni/4.5 wt% La 2 O 3 / - Al 2 O 3 catalyst is around 121 kJ/mol. Details of the calculation s are shown in Appendix C . 33 Next, the effect of feed flow rate on reacti on performance at the temperature of 210 o C was studied. Weight hourly sp ace velocity of the ethanol feed was calculated by dividing the mass flow rate of ethanol by the mass of the catalyst present in the reaction zone . As shown in Figure 2 . 4 , a t WHSV of 0.8 h - 1 and higher, the selectivity of C 4 + components stay s around the maximum value of 73% . Moreover, t he selectivity toward gaseous products and ethanol conversion decrease as WHSV increase s . Thus, the optimum va lue for the WHSV is about 0.8 h - 1 for the 8 .0 wt% Ni/ 4.5 wt% La 2 O 3 / - Al 2 O 3 catalyst. B esides the 8.0 wt% Ni/ 4.5 wt% La 2 O 3 / - Al 2 O 3 catalyst, t hroughout the studies o f the nickel monometallic catalyst several other nickel - based catalysts were made that varie d by their nickel content, lanthanum content, preparation method, and metal dispersion. A detailed list of the experiments conducted on these catalysts , along with ethanol conversion and se lectivity toward different products , are shown in Appendix D . For s implicity, these catalysts are labeled with Figure 2 . 3 . Temperature dependence of the 8 .0 wt% Ni/4.5 wt% La 2 O 3 / - Al 2 O 3 catalyst for the Guerbet reaction ( WHSV= 1.4 h - 1 ) 0% 10% 20% 30% 40% 50% 60% 70% 160 170 180 190 200 210 220 230 240 Conversion/ Selectivity T ( o C) Butanol Conversion Gases C6+ Alcohol 34 specific Roman numerals ( i.e. Ni (I), Ni (II), etc.) . Furthermore , Table 2 . 2 offers a representati on of the performance of all of these nickel - based catalysts at a common experimental condition (T= 210 o C, WHSV= 1.42 h - 1 ) and other conditions with high product selectivity, along with the results of the surface analysis performed on them. Methane and ethy l acetate selectivity are also presented in the table as the key gaseo us and liquid side - products formed. Other gases formed in reaction include CO, CO 2 , CH 4 , C 2 H 6 , and C 3 H 8 ; other liquid byproducts were diethyl ether, acetaldehyde, butyraldehyde, and seve ral unidentified components. The best experimental results obtained fo r the nickel monometallic catalysts belong to the 1.0 wt% Ni/ 9.0 wt% La 2 O 3 / - Al 2 O 3 , with the C 4 + alcohols selectivity as high as 79% and the conversion of 22%, and the C 4 + alcohols selec tivity as high as 74% and the conversion of 41%. 0% 10% 20% 30% 40% 50% 60% 70% 0.2 0.5 0.8 1.1 1.4 1.7 2.0 Conversion/ Selectivity WHSV (hr - 1 ) Butanol Conversion Gases C6+ Alcohol Other Figure 2 . 4 . Weight hourly speed velocity (WHSV) dependence of the 8 .0 wt% Ni/4.5 wt% La 2 O 3 / - Al 2 O 3 catalyst for the Guerbet r eaction (T=210 o C) 35 Table 2 . 2 . Comparison of t he performance of different Ni/ La 2 O 3 / - Al 2 O 3 catalysts and their surface characteristics . T ( o C) WHSV (h - 1 ) Conv. (%) Selectivity (%) Acidic Sites ( mol/g) Basic Sites ( mol/g) H 2 Chemisorbed ( mol/g) C 4 OH C 6 + OH C 4 + OH CH 4 Ethyl Acetate Ni (I): 8 .0 wt% Ni/ 4.5 wt% La 2 O 3 / - Al 2 O 3 210 1.42 22 58 13 71 12 3 640 220 50 230 1.42 42 46 13 59 18 0 Ni (II): 8 .0 wt% Ni/ 9 .0 wt% La 2 O 3 / - Al 2 O 3 210 1.42 28 52 15 68 12 2 590 320 73 Ni (III): 8 .0 wt% Ni / 9 .0 wt% La 2 O 3 / - Al 2 O 3 210 1.42 8 60 13 73 6 1 550 300 11 230 1.42 16 61 18 79 6 2 250 1.42 30 54 18 72 8 2 Ni (IV): 8 .0 wt% Ni/ 9 .0 wt% La 2 O 3 / - Al 2 O 3 210 1.42 18 54 15 69 8 2 580 320 42 Ni (V): 1 .0 wt% Ni/ 9 .0 wt% La 2 O 3 / - Al 2 O 3 210 1.42 7 68 10 78 5 0 600 290 11 230 0.79 22 58 20 79 6 1 250 0.79 41 51 23 74 8 1 250 1.42 29 54 21 75 7 1 Comparing the quantity of basic sites in different catalysts shows the direct effect of La 2 O 3 loading on the basicity of the particles. Although the constructive effect of La 2 O 3 presence on the selectivity to desired products has been discussed before [ 9 ] , t he results of this study do not reveal any correlation between the quantit y of La 2 O 3 and the performance of the catalyst . Experimental results of catalysts with different nickel loadings at similar conditions show that t he 1 .0 wt% nickel catalyst has lowe r conversion than those with 8 .0 wt% nickel, but less gas and liquid byproduct formation and overall higher selectivity to higher alcohol products. Furthermore, as discussed earlier for the 8 .0 wt% Ni/ 4.5 wt% La 2 O 3 / - Al 2 O 3 catalyst (labeled as Ni (I)), for each 36 catalyst, as temperature increases or as space velocity decreases, ethanol convers ion increases, selectivity to C 4 + alcohols declines, and the quantity of gas and liquid byproducts formed increases. The cataly st labeled as Ni (III) shows a lower ethanol conversion rate than the two other catalysts with the same composition (Ni (II) and Ni (IV)). The main cause for this observation is that this catalyst was made in large scales (~2 kg) for the nickel bimetallic experiments . This could affect the n ickel particles during the calcination and reduction steps in two ways. First, nickel sinter ing or deactiva tion could occur during these large - scale exothermic processes in the areas with extremely high temperatures . Sec ond, nitrogen and hydrogen diffusion through central cataly tic particles could be interrupted by the surrounding particles , leading to inefficient and non - uniform calcination and reduction. These effects are further confirmed by measuring the hydrogen upta ke of each of the catalysts , as shown in Table 2 . 2 . The results indicate that the Nickel ( III) catalyst has the lowest hydrogen uptake to nickel content ratio of all the catalysts. Considering that surface nickel a toms alone are responsible for the hydrogen adsorption, this observation justifies the correlation between the catalyst activity for the Guerbet reaction and the number of nickel sites on the catalyst surface . This will be further discussed in S ection 2.3.4 . 2.3.3 Nickel Bimetallic Catalytic Experiments 2.3.3.1 Experimental Discussion Several metals ( Cu, Co, Pd, Pt, Fe, and Mo ) were impregnated on Ni catalyst as second metals in this study. As mentioned in previous sections, for consistency of the results, these metals were all impregnated to the 8 .0 wt% Ni/ 9 .0 wt% La 2 O 3 / - Al 2 O 3 catalyst labeled as Ni (III) that had been prepared in a large amount. A complete list of the experimental conditions of the bimetallic and nickel - fre e monometallic catalytic experiments with the selectivity toward different products are 37 reported in Appen dix E . Among different catalysts studied at different experimental conditions, best results are for the c o - i mpregnated 8.4 wt% (Ni/Cu 20 /1)/ 9.0 wt% La 2 O 3 / - Al 2 O 3 catalyst with 74% C 4 + selectivity and 35% ethanol conversion at T= 250 o C and WHSV= 1.42 h - 1 and 78% C 4 + selectivity and 15% ethanol conversion at T= 230 o C and WHSV= 2.06 h - 1 . Table 2 . 3 compar es the result s of these bimetallic catalysts with the Ni (III) catalyst at a common experimental condition (T= 230 o C, WHSV= 1.42 h - 1 ) along with presenting results of the surface studi es conducted on them. All of the catalysts that are reported in this table we re prep ared using the separate impregnation method. The Ni - Cu bimetallic catalyst show s an interesting behavior in having higher selectivity for butanol and lower selectivity toward longer chain alcohols (C 6 +) . This behavior could advantage processes that desire butanol as their main product sin ce it produces more of this component. An example of this type of process is discussed in detail in Chapter 3. The Ni - Co bimetallic cata lyst show s a substantial improvement in ethanol conversion , while the C 4 + activity is s till in the 60 %+ range. However, through analyzing other experimental conditions for this catalyst, not more than 66% C 4 + selectivity was achieved at conversions of more than 30%. These two bimetallic combinations are further studied based on their metal r atio, impregnation metho d, and monometallic behavior, which will be discussed later in this section. Other bimetallic combinations in Table 2 . 3 were not successful in either r educing the selectivity of one or more of the side - products or improving the selectivity/ conversion of the desired components. However, in all those experiments, butanol was still the main product. 2.3.3.2 Surface Characterization To get a better understanding of the trends observed for the bimetallic , BET surface area analysis, NH 3 , CO 2 , and H 2 temperature progra m med desorption (TPD) analysis were conducted on each catalyst. While the results of these studies are summarized in Table 2 . 3 , 38 Appendix F shows the detailed profiles of CO 2 , NH 3 , and H 2 TPD for the Ni (III) catalyst and all of the bimetallic catalysts in Table 2 . 3 . BET surface studies show that all of the catalysts are in the same range of available surf ace area (120 - 130 m 2 /g). Considering that the available BET surface area for both - Al 2 O 3 and 9.0 wt% La 2 O 3 / - Al 2 O 3 supports are in the range of 140 - 150 m 2 /g, BET results from this table indicate that the metal particles impregnated on the surface of the s upport are mostly small enough not to cover the pores of the catalyst . Thus, they do not reduce the available surfa ce area. Table 2 . 3 . Bimetallic catalysts experimental results and surface analysis comparison with the Ni (III) catalyst. Experiments are conducted at T= 230 o C and WHSV= 1.42 h - 1 with ethanol as the feed. a: Ethyl Acetate b: Diethyl Ethe r. All bimetallic catalysts are prepared with the separate impregnation method. Conv. (%) Selectivity (%) BET a rea (m 2 /g) Acidic Sites ( mol/g) Basic Sites ( mol/g) H 2 Chemisorbed ( mol/g) C 4 OH C 6 + OH C 4 + OH CH 4 CO 2 Eth. Ac. a DEE b Ni (III): 8 .0 wt% Ni/ 9 .0 wt% La 2 O 3 / - Al 2 O 3 16 61 18 79 6 1 2 2 130 550 300 11 11.5 wt% (Ni/Cu 2/ 1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 17 6 6 7 73 7 2 1 5 120 530 250 4 11.5 wt% (Ni/Co 2/ 1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 33 43 23 66 7 2 1 0 120 540 330 13 9.7 wt% (Ni/Pd 4/ 1)/ 8.8 wt% La 2 O 3 / - Al 2 O 3 20 53 14 67 5 1 1 7 120 600 390 21 9.7 wt% (Ni/Pt 4/ 1)/ 8.8 wt% La 2 O 3 / - Al 2 O 3 21 49 8 56 16 7 2 4 120 530 260 40 11.5 wt % (Ni/Fe 2/ 1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 22 44 5 49 15 3 8 8 120 510 250 9 11.5 wt% (Ni/Mo 2/ 1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 9 51 2 54 3 0 14 1 120 590 110 3 NH 3 and CO 2 TPD analysis demonstrate a correlation between selectivity toward the desi red products and the base/acid ratio on the surface of the catalyst. This correlation, discussed 39 extensively in Chapter 1 , has been previously acknowledged by many researchers [ 15 , 23 , 24 , 39 , 42 - 45 , 56 ] ; however, the reported base /acid ratio number seems to be varying based on the catalyst used and the operational conditions (phase, temperature, pressure, flow rate). In this study, the base /acid ratio f or the Ni (III) catalyst is 0.6 ; as we deviate from this number in nickel bimetallic combinations, the selectivity toward C 4 + alcohols declines. For instance, while this ratio for Ni - Cu bimetallic catalyst with the best performance among those studied is 0 .5 , for the Ni - Mo bimetallic ca talyst , with the lowest C 4 + alcohols selectivity , this number is 0.2 . Nevertheless, the base /acid ratio is not the only factor determining the performance of the reaction; several other parameters such as base /acid site stren gth, metal dispersion on the catalyst surface, and hydrogen - metal bond strength need to be considered as well. Trends obtained for the H 2 chemisorbed on the surface of the bimetallic catalysts in this study are in agreement with several papers in the lite rature [ 70 - 75 ] . Addition of copper to the nickel catalyst would lowe r the total amount hydrogen adsorbed on the surface of the catalyst since this catalyst do es not promote hydrogen adsorption and is covering the surface of some of the available nickel metal sites [ 74 ] . This could justify why Ni - Cu bimetallic catalyst tend s to produce more butanol rather than C 6 + alcohols. Having less available dissociated hydrogen on the surface of the metal not only will reduce the total selectivity of the Guerbet reactions, but also can preserve the produced butanol from further reacting into higher alcohols. A s imilar trend of hydrogen adsorpt ion is observed for the addition of iron and mo lybdenum to the nickel catalyst. C obalt addition , on the other hand, does not significantly promote or obstruct hydrogen adsorption on the surface of the catalyst. The diff erence observed in the performan ce of this catalyst compared to the nickel monometallic catalyst could be due to its different surfac e structure that promote s the production of side - products rather than higher alcohols [ 59 ] . 40 Palladium and platinum addition have expectedly improved the hydrogen uptake of the catalyst drastically. These two met als are known for their strong ability to both adsorb and desorb hydrogen [ 70 - 73 ] . This explain s the observed improv em ent in the activity of the se catalyst s . In addition , a declin e in the selectivity toward C 4 + alcohols is observed due to the strong metal - hydrogen bond formation that impedes the hydrogenation step and promotes the side - reactions . The strength of metal - hydrogen bonds can be assessed based on the H 2 TPD profiles that are presented in Figure F.3 in Appendix F . Based on the assumption that all nickel, platinum, and palladium particles are capable of adsorbing H 2 , and that these particles are on the surface of the support without covering each other, metal dispersion and particle diameter can be calculated from the H 2 uptake results obtained. Calculations below , starting with the bulk metal density of each metal, describe the process performed for calculating these two quantities. Assuming that bulk atoms are cubical a length , we can s ay that the atom volume i a 3 n exposed a 2 Equation 2 . 1 where a the crystalline metal (2.22 for Ni, 2.45 for Pd, and 2.47 for Pt) . Furthermore, c onsidering that each mole of molecular hydrogen adsorbs onto two moles of metal, the bulk catalyst dispersion can be calculated using the amount of H 2 uptake. Note that for th e bimetallic catalysts, the H 2 uptake by nickel is subtracted from the total H 2 uptake. 41 Equation 2 . 2 Multiplying Equation 2 . 1 and Equation 2 . 2 with the assumptions we have uniform spherical catalyst particles with diameter D give s : Equation 2 . 3 By p erforming the above calculations , the following equation s will be derived for each one of the three metals. In addition , Table 2 . 4 represents the dispersion and the derived particle diameter s for these metals. Results indicate that the second metals form significantly smaller particles and are more dispersed compared to the nickel particles. Equation 2 . 4 Equation 2 . 5 Equation 2 . 6 For understanding the distribution pattern of the components of the bime tallic catalysts, surface elemental analysis was conducted using Scanning Electron Microscopy (SEM) incorporated with Energy Dispersive Spectroscopy (EDS). Figure 2 . 5 shows a representative picture of the co - impregnated 11.5 wt% (Ni/Cu 2/1)/8.7 wt% La 2 O 3 / - Al 2 O 3 catalyst. Based on EDS results obtained at different regions of this catalyst, darker areas are rich in nickel and lighter areas are rich in copper. Nickel par ticles visible in this image are in the similar size range as the nickel particle diamete r obtained based on the calculations described above. 42 Table 2 . 4 . Metal particle diameter and dispersion for the Ni - Pd an d Ni - Pt bimetallic catalysts. Numbers are calculated based on the H 2 uptake. Ni Dispersio n (%) Ni Particle D (nm) 2 nd Metal Dispersion (%) 2 nd Metal Particle D (nm) 9.7 wt% (Ni/Pd 4/1)/ 8.8 wt% La 2 O 3 / - Al 2 O 3 1.6 82.5 10.9 13.5 9.7 wt% (Ni/Pt 4/1)/ 8.8 wt% La 2 O 3 / - Al 2 O 3 1.6 82.5 57.7 2.7 Table 2 . 5 shows the surface elemental distribution of the components of the bimetallic catalysts and the ir comparison wit h the bulk molar percentages impregnated . La/Al and Ni/Al molar ratios are in close proximity with the bulk molar ratios, which support s the result obtained from BET studies indicating that lanthanum and nickel particles are not covering the pores. However , in most cases (except the Ni - Co bimetall ic catalyst), the nickel to second metal ratio is smaller than the bulk ratio . This could be due to two main reasons. First, the second metal has formed smaller particle sizes compared to nickel particles, as sugge sted in the calculations performed for palladium and platinum particles, which increase s its available surface area compared to nickel. Second, the second metal is covering the surface of nickel particles, as proposed for the Ni - Cu Ni Cu Figure 2 . 5 . SEM picture from the co - im pregnated 11.5 wt% (Ni/Cu 2/1)/8.7 wt% La 2 O 3 / - Al 2 O 3 catalyst. 43 bimetallic catalyst , bas ed on its decrease in H 2 uptake compared to the nickel monometallic catalyst. Table 2 . 5 . Surface elemental analysis conducted on bimetallic catalysts using SEM/EDS. La/Al (molar) Ni/Al (molar) Ni/M (molar) su rface bulk surface bulk surface bulk Ni (III): 8 .0 wt% Ni/ 9 .0 wt% La 2 O 3 / - Al 2 O 3 0.03 0.03 0.07 0.08 - - 11.5 wt% (Ni/Cu 2/1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 0.03 0.03 0.09 0.08 2.12 2.2 11.5 wt% (Ni/Co 2/1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 0.04 0.03 0.11 0.08 3.06 2.0 9 .7 wt% (Ni/Pd 4/1)/ 8.8 wt% La 2 O 3 / - Al 2 O 3 0.02 0.03 0.09 0.08 1.03 7.3 9.7 wt% (Ni/Pt 4/1)/ 8.8 wt% La 2 O 3 / - Al 2 O 3 0.04 0.03 0.07 0.08 4.29 13.3 11.5 wt% (Ni/Fe 2/1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 0.04 0.03 0.08 0.08 0.86 1.9 11.5 wt% (Ni/Mo 2/1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 0.03 0.03 0.07 0.08 2.04 3.3 2.3.3.3 Ni - Cu and Ni - Co Detailed Experiments Since Ni - Cu and Ni - Co bimetallic catalysts show ed some improvement in butanol selectivity and ethanol activity, respectively, these catalysts were subjected to further investigatio ns to assess their potential for improving the overall performance of the Guerbet reaction. Ni - Cu bimetallic catalyst was tes ted at different metal ratios and impregnation methods. Besides , the monometallic copper catalyst was examined to better understand the trends observed for the Ni - Cu bimetallic catalyst. This study was performed on the cobalt monometallic catalyst as well. Table 2 . 6 represents the results of these studies. 44 Table 2 . 6 . Different nickel, copper, and cobalt based catalysts studied. All results are at T= 230 o C and WHSV= 1.42 h - 1 . Conversion (%) Selectivity (%) C 4 OH C 6 + OH C 4 + OH Ethyl Acetate Gases Ni (III): 8 .0 wt% Ni/ 9 .0 wt% La 2 O 3 / - Al 2 O 3 16 61 18 79 2 11 8.4 wt% (Ni/Cu 20/1)/ 9.0 wt% La 2 O 3 / - Al 2 O 3 (Co - Impregnation) 18 65 11 76 3 12 11.5 wt% (Ni/Cu 2/1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 (Co - Impregnation) 17 71 6 77 2 11 11.5 wt% (Ni/Cu 2/1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 (Sep - Impregnation) 17 66 7 73 1 13 8 .0 w t% Cu/ 9 .0 wt% La 2 O 3 / - Al 2 O 3 10 6 0 6 62 5 11.5 wt% (Ni/Co 2/1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 (Sep - Impregnation) 33 43 23 66 2 16 8 .0 wt% Co / 9 .0 wt% La 2 O 3 / - Al 2 O 3 24 7 0 7 53 22 For the Ni - Cu experiments, the impregnation method does not signi ficantly affect the experimental result s . On the other hand, as the copper content increases in the bimetallic catalyst, the effect of shifting the selectivity toward butanol instead of C 6 + alcohols become more evident. Furthermore, lower catalyst activity at high copper load ing is attributed to the formation of isolated CuO instead of the inverse spinel CuAl 2 O 4 structure on the surface [58] . Moreover , t he amount of gases formed in the monometallic copper catalyst is significantly lower compared to the othe r two monometal lic c atalysts. This is due to the ability of copper in preserving the carbon - carbon bond , therefore reducing the chance of ethanol decomposition . Although low selectivity for ethyl acetate is obtained in the Ni - Cu and Ni - Co bimetallic cataly sts, this chemi cal is the main product of the monometallic copper and cobalt catalysts . R eported 45 likewise by other research groups [ 59 , 85 - 88 ] , this observation is attributed to the metal specific causes. For the copper catalyst , coordinativel y unsaturated Cu + ions are the active centers for ethyl acetate formation. While octahedrally coordinated metal cations ar e known as idea l structures for the Guerbet reaction, the tetrahedral ly coordin ated Co 2+ sites are responsible for the ethyl acetate formation . In addition, the larger crystalline size of copper and cobalt particles compared to nickel on alumina surface su ggests that the se metals have different types of interactions with the support, leading to the formation of different products [ 59 , 85 , 86 ] . Finally, among the three monometallic catalysts studied, cobalt has the highest activity. This observation is justified by the higher d - band center compared to the o ther two cataly sts , corresponding to a lower energy required to activate th e metal - hydrogen interaction , and its subsequent higher activity [ 59 ] . 2.3.4 Mechanistic and Kinetic Studies Multiple experiments were conducted to investigate the active mechanism and the rate - limiting step of this mechanism for the cond ensed - phase Gue rbet reaction over metallic and bimetallic catalysts . These experiments are discussed in this section. Throughout the experiments on the Guerbet reacti on, there was evidence confirming the occurrence of the a ldol c ondensation mechanism. Firs t, the low etha nol conversion in control experiments with 9.0 wt% La 2 O 3 / - Al 2 O 3 catalyst supports the significant role of nickel as the promoting agent for activating the catalyst. As discussed in Chapter 1, m etallic sites are prerequisite for only the ald ol c ondensation mechanism and not the d irect d ehydration mechanism . Thus, the f act that non - metallic support shows ethanol conversion of less than 3% even at 250 o C supports aldol c ondensation as the main mechanism for this reaction. Second , experiments sh ow that the add ition of acetaldehyde to ethanol feed improve s the performance of the Guerbet reaction 46 accordingly, which confirms the role of acetaldehyde as an active component in the Guerbet mechanism. Finally , to investigate the mechanism of the reactio n further, acet aldeh yde was fed to the reactor as the sole reactant . The main product formed in a cetaldehyde experiments was th e conden sation product (crotonaldehyde), additional evidence supporting that the Guerbet reaction is proceeding through aldol con densation. Acet aldehyde experimental results , summarized in Table 2 . 7 , provides further information regarding the kinetics and thermodynamics of th e system . The temperature required to activate acetaldehyde is sign ificantly lower than that to activate ethanol ; experiments showed that at the temperature range of 210 o C - 250 o C, acetaldehyde completely reacts to dimers and polymers that can quickly bloc k the catalysis zone . Therefore, acetaldehyde experiments we re cond ucted in the te mperature range of 110 o C - 130 o C . This is an indicator that acetaldehyde is significantly more active compared to ethanol and its condensation could not be the rate - limiting step of the Guerbet mechanism. Table 2 . 7 . Acetaldehyde experiments over Ni (III) catalyst. All experiments are done at WHSV= 1.42 h - 1 . T ( o C) Conversion (%) Selectivity (%) Crotonaldehyde Ethanol Sum All 110 32.1 45.1 4.7 72.8 120 54.9 49.9 2.6 77.4 130 86.6 37.1 2.2 68 .1 The sum of all chemical selectivities in acetaldehyde experiments is in the range of 65% - 80%. This is due to the number and quantity of unidentified liquid compounds in this system. For measuring the unknown product selectivity, a hypothetical average stoi chiometry coefficient of three was considered , which might be too low since the chemicals in the system were highly active and capable of forming dimers and longer molecules with a higher stoichiometry ratio. Adjusting 47 this coefficient to 6.1 for the unkno wn product s will improve the sum of the selectivities to the range of 98% - 103%, which further supports this hypothesis. It is worthwhile to note that the mass closure (mass out/mass in) in all of the acetaldehyde experiments are in the range of 96% - 10 2%. A side from acetaldehyde experiments, H 2 uptake analysis provides additional evide nce for considering the first step of the reaction as rate limiting. It was discussed in Section 2.3.2 that there is a correla tion between th e amount of H 2 uptake in the 8% nickel catalysts and their activity for the Guerbet reaction. This correlation is shown in Figure 2 . 6 based on the data from Table 2 . 2 . The linear correlation suggest s that the number of available nickel sites for dehydrogenating ethanol molecules is the limiting factor in shifting the reaction toward higher C 4 + alcohols yield. This is a clear evidence that the dehydrogenation s tep is the rate - limiting step of the Guerbet reaction. Butanol was also fed to the reactor , and product analysis was done to help with understanding the behavior of the reaction. Feeding butanol to the reactor at 210 o C and 1.42 h - 1 flow rate resulted in 16 % butanol co nversion, 2% ethanol selectivity, and 31% C 6 + selectivity. Considering that feedin g ethanol under the similar conditions led to 22% ethanol conversion, 58% butanol selectivity and 0 5 10 15 20 0 20 40 60 80 C 4 + Alcohols Yield (%) H 2 Chemisorbed ( mol/g) Figure 2 . 6 . Correlation betwee n H 2 adsorption on the s urface of different 8 wt% nickel catalysts and the yield. 48 13% C 6 + selectivity, it can be concluded that the ethanol Gue rbet reaction a t the mentioned condition s is not at thermodynamic equilibrium. 2.4 Conclusions The continuous condensed - phase reaction of ethanol to n - butanol and C 6 + alcohols has been demonstrated in a laboratory fixed bed reactor. Experimental studies on dif ferent Ni/ La 2 O 3 / - Al 2 O 3 catalysts show that the WHSV of >0.8 h - 1 and the temperature range of 21 0 - 25 0 o C are the optimum conditions for the system . La 2 O 3 content of the catalyst is the main contributor to the number of basic sites on the surface of the cat alyst . Moreover , the number of Ni sites on the catalyst surface correlate directly with the performance of the Guerbet reaction, a strong evidence that the dehydrogenation step is the rate - limiting step for the Guerbet reaction. Among several bimetallic ni ckel catalysts st udied for improving the performance of the reaction, Ni - Cu bimetallic was able to shift the reaction selectivity toward n - butanol rather than C 6 + alcohols, and Ni - Co bimetallic catalyst improved the activity of ethanol with no evidence of improving the d es in separate experiments and favored the production of ethyl acetate rat her than higher alcohols. The base /acid molar ratio of 0.5 - 0.7 seems to be the necessary, but not suffici en t, condition to obtain optimum C 4 + selectivity. Presence of the right amount of metal sites that are capable of forming metal - hydrogen bonds with the ideal strength is another important benchmark . Mechanistic studies confirm the aldol cond ensation mechan is m as the governing one for condensed phase Guerbet reaction using the La 2 O 3 / - Al 2 O 3 supported nickel catalyst s and introduce the dehydrogenation step as the rate - limiting step of the reaction. Among different catalysts studied, 1.0 wt% Ni/ 9.0 wt% La 2 O 3 / - Al 2 O 3 with 41% ethanol conversion and 74% C 4 + 49 selectivity (T= 250 o C, WHSV= 0.79 h - 1 ) , and co - impregnated 8.4 wt% (Ni/Cu 20/1)/ 9.0 wt% La 2 O 3 / - Al 2 O 3 with 15% ethanol conve rsion and 78% C 4 + selectivity (T=230 o C , WHSV= 2.06 h - 1 ) are those w ith the best pe rformance. 50 APPENDI CES 51 APPENDIX A: Catalyst P reparation S teps Table A.1. The s preadsheet used for obtaining the mass required for different chemicals in preparing catalysts. A B C D E F 1 Component Mass (g) Mass (g) MW(g/mol) Conc.(wt%) 2 Desired wt% La 2 O 3 =F2*D 11 3.9 325.81 0.09 3 Ni =F3*D 11 3.47 58.69 0.08 4 Cu =F4*D 11 1.73 63.55 0.04 5 Raw m aterials used La(NO 3 ) 3 · 6H 2 O =D 2/E2*E5*2 10.37 433.01 - 6 Ni(NO 3 ) 2 · 6H 2 O =D 3/E3*E6 17.18 290.79 - 7 Cu(NO 3 ) 2 · 2.5H 2 O =D 4/E4*E7 6.33 232.59 - 8 Final amou nts - Al 2 O 3 =D 11*F8 35.97 101.96 0.8 9 La 2 O 3 / - Al 2 O 3 =D 8/F8*(F9) 39.87 123.86 0.89 10 Ni/La 2 O 3 / - Al 2 O 3 =D 11*(F8+F2+F3) 43.33 118.65 0.96 11 Catalyst 45 45 116.53 1 I. La 2 O 3 impregnation on - Al 2 O 3 : a. Have 35.97 g of - Al 2 O 3 . b. Have 10.37 g of La(NO 3 ) 3 · 6H 2 O. c. The p ore volume of - Al 2 O 3 is 0.5 ml/g. So, multiply the number from (a) by 0.5. The result is 17.99 ml of water and La(NO 3 ) 3 · 6H 2 O mixture. d. Add the number from part (b) to a small amount of water, and then add water again; until it reac hes the ml obta ined in (c). e. Add that solution to the amount of - Al 2 O 3 in part (a) drop by drop, and stir it gradually to get it absorbed by the support. f. Let the mixture stay overnight. g. Dry the mixture in an oven at 130 o C for 18h. 52 h. Calcine it at 600 o C for 20 h in 35 ml/min of N 2 flow. The final mass of this step will be 39.87 g. II. Ni impregnation on La 2 O 3 / - Al 2 O 3 : a. Take the whole amount obtained from (1.h). b. Have 17.18 g of Ni(NO 3 ) 2 · 6H 2 O. c. The p ore volume of La 2 O 3 / - Al 2 O 3 is still close to 0.5 ml/g. So, mul tiply the numbe r from (a) to that. The result is 19.94 ml of water and the bimetal mixture. d. Add the number from part (b) to a small amount of water, and then add water again; until it reaches the volume obtained in part ( c). e. Add that solution to the amoun t of La 2 O 3 / - Al 2 O 3 in part (a) drop by drop, and stir it gradually to get it absorbed by the support. f. Let the mixture stay overnight. g. Dry the mixture in an oven at 130 o C for 18h. h. Calcine it at 600 o C for 20 h in 35 ml/ min of N 2 flow. i. Reduce it at 520 o C in a tubul ar f low reactor for 20 h under 35 ml/min H 2 . The final mass of this step will be 43.33 g. III. Cu impregnation on Ni/ La 2 O 3 / - Al 2 O 3 : a. Take the whole amount obtained from (2.h). b. Have 17.18 g of Cu(NO 3 ) 2 .2.5H 2 O. c. The p ore volume of Ni - La 2 O 3 - Al 2 O 3 is still close to 0.5 ml/g. So, multiply the number from (a) to that. The result is 21.66 ml of water and the f bimetal mixture. d. Add the number from part (b) to a small amount of water, and then add water again; until it reaches the volume obtained in part (c). 53 e. Add that sol ution to the amount of Ni/ La 2 O 3 / - Al 2 O 3 in part (a) drop by drop, and stir it gradually to get it absorbed by the supp ort. f. Let the mixture stay overnight. g. Dry the mixture in an oven at 130 o C for 18h. h. Calcine it at 600 o C for 20 h in 35 ml/min of N 2 flow. i. Reduce it at 520 o C in a tubul ar flow reactor for 20 h under 35 ml/min H 2 . The final mass of this step will be 45 g. 54 APPENDIX B: Feed F low R ate C alculation Based on the work reported by Jordison et al. [ 9 ] on t he batch Guerbet experiments with 8 .0 wt% Ni/9 .0 wt% La 2 O 3 / - Al 2 O 3 at 230 °C and autogenous pressure, by using 0.093 grams of catalyst pe r grams of ethanol fed , after ten hours of the catalyst screening experiments maximum conversion of 55% was achieved . S o, This number correlates the batc h and continuous experiments. For the continuous experiment s at 230 °C , for achieving 35 % conversion of et hanol with 29.9 g of catalyst we have: 55 APPENDIX C: Activation E nergy C alculation Assuming a second order reaction for the consumption of ethanol, we have: Equation C . 1 Equation C . 2 By integrating from both sides of the equation : Equation C.3 Equation C.4 Equation C.5 Where C 0 is the feed concentration in mol/ m 3 , x is the feed conversion, k 1 is the reaction rate constant in m 6 of so lution / kg of catalyst/ mol ethanol /min assuming a single - ste p mechanism, cat is the bulk density of the catalyst in kg/m 3 , and is the superficial residence time in min. Using the above equation at different reaction temperatures, we will be able to get m ultiple k 1 s . Then we need to involve the activation energy equation for each data point. Equation C.6 Equation C.7 Ne xt, we can draw a diagram for R ln (k 1 ) from Equation vs. 1/T for different reaction temperatu res and find the slope of the diagram to derive an activation energy for the reaction. For instance, for 56 the 8 .0 wt% Ni/4.5 wt% La 2 O 3 / - Al 2 O 3 catalyst t he diagram presented in the next page will be obtained: Usi ng the slope obtained in Figure , the activation energy of the Guerbet reaction for this c atalyst will be 121.4 kJ/mol. Rlnk 1 = - 121364/T + 189.67 -95 -85 -75 -65 -55 -45 0.00195 0.00205 0.00215 0.00225 0.00235 Rln(k 1 ) (J/mol/K) 1/T (1/K) Figure C.1 . Activation energy calculation for the 8 .0 wt% Ni/4.5 wt% La 2 O 3 / - Al 2 O 3 catalyst. 57 APPENDIX D: Nickel Monometallic Experimental Data D.1. Experimental Details Table D.1 . Experimen tal details for the monometallic nickel experiments. Exp # Feed T ( o C) WHSV (h - 1 ) Co nversion (%) Selectivity (%) BuOH C 6 + OH Gasses Control Experiment: 9 .0 wt% La 2 O 3 / - Al 2 O 3 IN116 - 2 Ethanol 210 1.42 0.8 19.8 0.3 0.0 IN116 - 1 E thanol 230 1.42 1.3 17.4 0.0 0.0 IN116 - 3 Ethanol 250 1.42 2.4 13.0 0.3 17.3 Ni (I): 8 .0 wt% Ni/ 4.5 wt% La 2 O 3 / - Al 2 O 3 IN28 - 4 Ethanol 170 1.42 4.2 27.2 0.4 20.1 IN28 - 1 Ethanol 190 1.42 10.0 55. 1 6.8 10.8 IN48 - 2 Ethanol 210 0.47 40.6 47.4 12.7 27.0 IN29 - 3 Ethanol 210 0.79 29.0 59.2 14.0 22.4 IN29 - 2 Ethanol 210 1.11 25.8 57.3 13.2 17.9 IN28 - 2 Ethanol 210 1.42 21.9 57.9 12.9 16.0 IN54 - 1 Ethanol + 0.7wt% Acetald ehyde 210 1.42 23.8 56.6 17.1 15.0 IN56 - 1 Ethanol + 0.4wt% Ethyl Acetate 210 1.42 21.9 54.1 13.6 16.9 IN58 - 1 Butanol 210 1.46 16.2 - 31.1 19.4 IN65 - 1 Ethanol + 4wt% Water 210 1.46 17.3 54.4 8.4 25.3 IN48 - 3 Ethanol 210 1.74 21.7 56.2 12.8 15.5 IN28 - 3 Ethanol 230 1.42 42.3 45.5 13.2 28.4 Ni (II): 8 .0 wt% Ni/ 9 .0 wt% La 2 O 3 / - Al 2 O 3 IN66 - 1 Ethanol 210 1.42 27.8 52.1 15.4 18.8 Ni (III): 8 .0 wt% Ni/ 9 .0 wt% La 2 O 3 / - Al 2 O 3 IN83 - 1 Ethanol 210 0.79 11.2 58.8 16.1 10.0 IN83 - 2 Ethanol 210 1.42 8.2 60.1 13.1 8.6 IN83 - 3 Ethanol 210 2.06 7.2 59.4 10 .6 8.0 IN84 - 1 Ethanol 230 0.79 23.7 57.1 19.8 15.2 IN84 - 2 Ethanol 230 1.42 16.3 61.4 17.7 10.8 IN84 - 3 Ethanol 230 2.06 12.8 63.9 16.2 8.4 IN85 - 1 Ethanol 250 0.79 40.2 50.6 19.6 18.3 58 Table D.1 (c ) IN85 - 2 Ethanol 250 1.42 29.6 54.2 17.9 15.7 IN85 - 3 Ethanol 250 2.06 25.6 54.9 19.2 12.2 IN124 - 2 t - butanol+ 5wt% Acetaldehyde 110 1.42 32.1 0.0 0.0 0.0 IN124 - 3 t - butanol+ 5wt% Acetaldehyde 120 1.42 54.9 0.0 0.0 0.0 IN124 - 1 t - butanol+ 5wt% Acetaldehyde 130 1.42 86.6 0.0 0.0 0.0 Ni (IV): 8.0 wt% Ni/ 9.0 wt% La 2 O 3 / - Al 2 O 3 IN99 - 1 Ethanol 210 1.42 18.1 54.4 14.8 12.8 IN99 - 2 Ethanol 230 1.42 36.1 45.5 16.4 23.7 IN99 - 3 Ethanol 250 1.42 60.7 29.7 12.9 46.3 Ni (V): 1.0 wt% Ni/ 9.0 wt% La 2 O 3 / - Al 2 O 3 IN111 - 1 Ethanol 210 1.42 7.0 68.0 10.4 5.5 IN109 - 1 Ethanol 230 0.79 22.4 58.2 20.5 9.4 IN109 - 2 Ethanol 230 1.42 14.9 61.6 17.1 8.1 IN109 - 3 Ethanol 230 2.06 12.0 62.9 14.4 7.7 IN110 - 1 Ethanol 250 0.79 41.1 50.9 23.4 14.7 IN110 - 2 Ethanol 250 1.42 28.9 54.4 20.9 12.4 IN110 - 3 Ethanol 250 2.06 23.2 58.1 10.4 11.2 D.2. Liquid Products Selectivity Table D.2 monometallic nickel experiments. 1 Ethanol selectivity for these experiments are: IN58 - 1= 2.2%, IN124 - 2= 4.7%, IN124 - 3= 2.6%, IN124 - 1= 2.2%. Exp # Selectivity (%) Ethyl Acet ate Diethyl Ether Acetal Acetaldehyde Butyraldehyde Crotonaldehyde Other Liquids IN116 - 2 0.0 7.8 1.2 0.0 0.0 0.0 62.0 IN116 - 1 0.0 19.6 2.2 0.0 0.0 0.0 47.2 IN116 - 3 0.0 15.0 2.3 0.0 0.0 0.0 60.1 IN28 - 4 59.2 3.8 0.0 8.5 0.0 1.0 3.1 IN28 - 1 13.6 1.6 0.0 8 .2 1.4 0.9 1.3 IN48 - 2 1.5 1.3 0.7 2.3 0.4 0.3 1.2 IN29 - 3 2.2 1.8 0.9 2.9 0.6 0.4 1.3 IN29 - 2 2.0 1.7 1.2 3.0 0.7 0.3 1.2 IN28 - 2 2.8 1.4 0.0 6.2 1.8 0.2 0.6 59 Table D.2 (c ) IN54 - 1 2.1 1.3 2.0 - 0.6 1.1 0.2 1.2 IN56 - 1 2.9 2.1 2.7 2.9 1.3 0.3 1.3 IN5 8 - 1 1 1.0 0.2 0.0 0.6 11.0 0.0 19.7 IN65 - 1 1.5 1.2 1.8 3.4 0.9 0.0 0.6 IN48 - 3 1.8 1.5 1.8 3.6 1.0 0.2 1.1 IN28 - 3 0.5 0.5 0.0 5.3 2.5 0.3 0.3 IN66 - 1 2.2 1.3 1.4 2.5 1.0 0.3 1.3 IN83 - 1 1.4 1.9 3.3 2.1 0.7 0.3 1.7 IN83 - 2 0.7 2.3 5.5 2.4 0.8 0.4 2.1 IN83 - 3 1.0 2.6 6.6 2.5 0.8 0.4 2.4 IN84 - 1 1.9 1.7 0.6 1.5 0.6 0.1 1.2 IN84 - 2 1.9 1.9 1.6 1.9 0.8 0.2 1.5 IN84 - 3 1.7 2.2 2.5 2.3 0.9 0.2 1.5 IN85 - 1 3.0 2.1 0.1 1.4 0.6 0.0 1.5 IN85 - 2 2.3 2.2 0.4 1.6 0.7 0.1 1.5 IN85 - 3 2.4 1.8 0.6 2.0 1.0 0.1 1.4 IN124 - 2 1 0.0 0.0 0.0 - 0.0 45.1 22.9 IN124 - 3 1 0.0 0.0 0.0 - 0.0 49.9 24.9 IN124 - 1 1 0.0 0.0 0.0 - 0.0 37.1 28.8 IN99 - 1 1.7 2.2 2.5 2.2 1.2 0.0 2.3 IN99 - 2 2.0 1.7 0.7 1.6 0.9 0.0 3.8 IN99 - 3 2.2 1.2 0.0 1.1 0.8 0.2 5.5 IN111 - 1 0.0 1.0 6.0 2.2 0.3 0.4 5.8 IN109 - 1 0.9 0.8 1.1 1.6 0.7 0.1 4.6 IN109 - 2 1.0 0.9 1.9 2.0 0.7 0.1 4.4 IN109 - 3 1.0 1.0 2.7 2.3 0.7 0.0 4.6 IN110 - 1 1.1 0.8 0.5 1.2 0.9 0.1 4.7 IN110 - 2 1.2 0.8 0.8 1.6 1.0 0.1 5.2 IN110 - 3 1.3 0.9 1.2 1.9 1.1 0.1 5.1 60 D.3. Gaseous Products Selectivity Tab le D . 3 . monometallic nickel experiments. 1 Sum of both liquid and Exp # Selectivity (%) Carbon Monoxide Carbon Dioxide Methane Ethane Propane Sum All 1 IN28 - 4 0.1 1.3 17.9 0.5 0.2 123.2 IN2 8 - 1 0.1 0.6 9.2 0.5 0.4 99.7 IN48 - 2 0.2 5.0 15.9 2.1 3.7 94.7 IN29 - 3 0.6 3.6 14.5 0.9 2.4 105.1 IN29 - 2 0.4 2.6 12.1 0.8 1.7 98.1 IN28 - 2 0.3 2.0 11.6 0.7 1.3 99.5 IN54 - 1 0.1 2.0 9.0 1.2 2.6 96.1 IN56 - 1 0.1 1.8 10.8 1.6 2.4 98.0 IN58 - 1 0.1 1.2 2.4 0.7 15.0 85.2 IN65 - 1 0.3 3.7 16.5 1.5 3.0 97.2 IN48 - 3 0.1 1.7 10.8 1.1 1.7 95.3 IN28 - 3 0.4 5.3 17.8 1.2 3.6 96.2 IN66 - 1 0.2 2.3 11.5 1.4 3.3 96.2 IN83 - 1 0.0 0.7 6.0 1.9 1.3 96.2 IN83 - 2 0.1 0.5 6.4 1.5 0.2 96.1 IN83 - 3 0.0 0.4 6.2 1.3 0.0 94.3 IN84 - 1 0. 1 1.9 6.2 2.3 2.0 97.1 IN84 - 2 0.1 1.1 6.2 2.1 1.7 100.0 IN84 - 3 0.1 0.6 6.5 1.6 0.0 100.0 IN85 - 1 0.2 1.9 10.1 3.0 3.1 97.2 IN85 - 2 0.7 2.0 7.8 2.1 3.1 96.5 IN85 - 3 0.3 0.9 8.2 1.8 1.0 95.5 IN124 - 2 0.0 0.0 0.0 0.0 0.0 72.8 IN124 - 3 0.0 0.0 0.0 0.0 0.0 77 .4 IN124 - 1 0.0 0.0 0.0 0.0 0.0 68.1 IN99 - 1 0.4 1.1 7.9 1.4 2.1 94.2 IN99 - 2 0.6 2.9 11.0 3.4 5.9 96.2 IN99 - 3 1.5 6.6 19.8 3.8 14.6 99.9 IN111 - 1 0.1 0.2 4.6 0.2 0.2 99.5 61 Table D.3 (c ) IN109 - 1 0.2 1.3 6.0 0.7 1.3 98.0 IN109 - 2 0.2 0.8 5.8 0.5 0.8 9 7.7 IN109 - 3 0.2 0.7 5.7 0.4 0.7 97.3 IN110 - 1 0.2 2.6 8.0 0.8 3.1 98.1 IN110 - 2 0.2 1.9 7.3 0.7 2.3 98.4 IN110 - 3 0.1 5.2 4.2 0.7 1.0 91.3 62 APPENDIX E: Nickel Bi metallic Experimental Data E.1. Experimental Details Table E . 1 . Experimental details for th e bimetallic nickel experiments . All bimetallic catalysts are prepared using the separate impregnation method unless otherwise noted. Exp # T ( o C) WHSV (h - 1 ) Conv. (%) Selectivity (%) BuOH C 6 + OH C 4 + OH Gasses 11.5 wt% (Ni / Cu 2 / 1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 IN86 - 1 230 0.79 22.2 63.6 7.7 71.3 14.7 IN86 - 2 230 1.42 16.8 66.1 7.0 73.0 12.6 IN86 - 3 230 2.06 14.3 66.7 6.8 73.5 11.6 IN87 - 1 250 0.79 39.1 54.2 10.5 64.7 22.8 IN87 - 2 250 1.42 29.5 61.9 10.3 72.2 15.7 IN87 - 3 250 2.06 25.4 62.2 10.2 72.4 14.2 11. 5 wt% (Ni/Cu 2/ 1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 (Co - Impregnated ) IN74 - 1 210 1.42 7.3 67.4 3.1 70.5 10.1 IN76 - 3 230 1.42 16.9 70.8 6.4 77.2 11.3 IN76 - 2 250 1.42 30.0 61.5 8.6 70.0 15.0 IN76 - 3 250 3.32 22.3 65.7 8.0 73.7 10.2 8.4 wt% (Ni/Cu 20/ 1)/ 9.0 wt% La 2 O 3 / - Al 2 O 3 (Co - Impregnated ) IN79 - 3 210 0.79 10.9 67.7 9.1 76.8 8.7 IN78 - 2 210 1.42 8.8 67.8 7.6 75.5 8.2 IN80 - 1 210 2.06 7.6 65.3 7.0 72.3 8.1 IN79 - 2 230 0.79 25.6 60.3 12.1 72.5 16.6 IN78 - 1 230 1.42 17.9 65.2 11.3 76.5 12.1 IN80 - 2 230 2.06 15.1 66.1 11.5 77.7 9.2 IN79 - 1 250 0.79 43. 9 52.3 15.2 67.5 20.8 IN80 - 4 250 1.42 34.6 58.8 15.2 74.0 17.1 IN80 - 3 250 2.06 29.5 58.2 14.3 72.6 15.5 8 .0 wt% Cu / 9 .0 wt% La 2 O 3 / - Al 2 O 3 IN97 - 1 210 1.42 6.6 5.2 0.0 5.2 4.1 IN95 - 1 230 0.79 13.0 5.2 0.0 5.2 5.1 IN95 - 2 230 1.42 10.2 5.6 0.0 5.6 5.3 IN95 - 3 230 2.06 9.3 5.7 0.0 5.7 5.8 63 Table E.1 (c ) IN96 - 1 250 0.79 26.1 6.7 0.0 6.7 9.9 IN96 - 2 250 1.42 20.3 7.3 0.0 7.3 5.8 IN 96 - 3 250 2.06 19.1 6.7 0.0 6.7 15.7 11.5 wt% (Ni / Co 2/ 1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 IN105 - 1 210 1.42 11.8 56.8 15.3 72.0 9.3 IN103 - 1 230 0.79 45.9 40.7 24.6 65.3 18.7 IN103 - 2 230 1.42 33.4 43.3 22.9 66.2 15.6 IN103 - 3 230 2.06 26.9 46.0 21.9 67.9 13.9 IN10 4 - 1 250 0.79 59.5 22.4 11.9 34.3 45.9 IN104 - 2 250 1.42 51.4 34.0 17.8 51.8 29.0 IN104 - 3 250 2.06 43.3 41.6 20.1 61.7 16.1 8 .0 wt% Co / 9 .0 wt% La 2 O 3 / - Al 2 O 3 IN115 - 2 210 1.42 12.5 7.5 0.0 7.5 20.7 IN114 - 1 230 0.79 38.0 0.7 0.0 0.7 24.9 IN115 - 3 230 1.42 24.0 6.5 0.1 6.6 21.8 IN115 - 1 250 1.42 49.2 4.3 0.1 4.4 23.6 9.7 wt% (Ni/ Pd 4 / 1)/ 8.8 wt% La 2 O 3 / - Al 2 O 3 IN107 - 1 210 1.42 9.9 54.3 8.9 63.1 5.9 IN106 - 1 230 0.79 28.3 51.1 15.7 66.8 14.0 IN106 - 2 230 1.42 20.3 53.4 13.8 67.1 12.0 IN106 - 3 230 2.06 16.8 53 .1 12.1 65.2 11.7 IN108 - 1 250 0.79 46.8 41.1 14.9 56.0 24.1 IN108 - 2 250 1.42 34.4 47.2 14.2 61.4 13.4 IN108 - 3 250 2.06 27.9 49.4 13.3 62.7 12.1 9.7 wt% (Ni/ Pt 4 / 1)/ 8.8 wt% La 2 O 3 / - Al 2 O 3 IN91 - 1 210 1.42 9.2 55.8 4.9 60.7 19.8 IN89 - 1 230 0.79 28.2 43.9 7.3 51.3 36.4 IN89 - 2 230 1.42 20.9 48.9 7.6 56.5 31.3 IN89 - 3 230 2.06 16.9 51.2 7.3 58.5 28.6 IN90 - 1 250 0.79 59.2 19.6 4.4 24.0 66.1 IN90 - 2 250 1.42 40.8 32.9 6.8 39.7 47.9 IN90 - 3 250 2.06 35.4 35.5 7.9 43.5 46.4 64 Table E.1 (c ) 11.5 wt% (Ni/Fe 2/ 1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 IN94 - 1 210 1.42 8.9 45.9 1.8 47.7 15.0 IN92 - 1 230 0.79 29.7 42.6 5.6 48.2 27.4 IN92 - 2 230 1.42 21.9 44.2 5.0 49.2 23.8 IN92 - 3 230 2.06 17.5 44.3 4.3 48.6 21.9 IN93 - 1 250 0.79 43.9 37.2 8.7 45.9 31.5 IN93 - 2 250 1.42 31.8 44.5 9.6 54.1 23.0 IN 93 - 3 250 2.06 25.8 47.0 9.3 56.3 20.1 11.5 wt% (Ni/Mo 2/ 1)/ 8.7 wt% La 2 O 3 / - Al 2 O 3 IN102 - 1 210 1.42 5.2 51.3 2.1 53.4 5.9 IN100 - 1 230 0.79 9.7 56.3 2.4 58.7 11.2 IN100 - 2 230 1.42 9.3 51.3 2.4 53.7 10.5 IN100 - 3 230 2.06 8.1 51.9 2.6 54.5 11.1 IN101 - 1 25 0 0.79 16.9 55.3 5.0 60.2 1.8 IN101 - 2 250 1.42 15.3 52.7 4.8 57.5 9.9 IN101 - 3 250 2.06 13.3 55.5 4.9 60.5 7.1 E.2. Liquid Products Selectivity Table E.2 . bimetallic nickel experiments. Exp # Selectivity (%) Ethyl Ace tate Diethyl Ether Acetal Acetaldehyde Butyraldehyde Crotonaldehyde Other Liquids IN86 - 1 1.4 6.4 1.5 1.3 0.0 0.0 1.0 IN86 - 2 1.4 5.4 2.0 1.7 0.0 0.0 1.1 IN86 - 3 1.4 5.1 2.5 2.0 0.0 0.0 1.1 IN87 - 1 2.1 4.5 0.0 1.2 0.4 0.0 1.3 IN87 - 2 2.4 4.6 0.0 1.4 0.0 0. 0 1.1 IN87 - 3 2.7 4.5 0.0 1.7 0.2 0.0 1.0 IN74 - 1 1.1 7.5 5.3 1.2 0.0 0.0 1.9 IN76 - 3 2.3 5.0 1.4 1.2 0.0 0.0 1.1 IN76 - 2 2.3 4.5 0.7 1.0 0.3 0.0 1.7 IN76 - 3 2.6 3.9 0.9 1.6 0.4 0.0 1.5 65 Table E.2 (c ) IN79 - 3 1.1 2.8 3.2 1.3 0.2 0.3 1.7 IN78 - 2 1.7 2. 7 4.2 1.6 0.2 0.4 1.7 IN80 - 1 1.6 2.6 5.3 1.7 0.2 0.4 1.9 IN79 - 2 2.4 2.3 0.8 0.9 0.4 0.1 1.3 IN78 - 1 2.7 2.2 1.4 1.4 0.4 0.2 1.4 IN80 - 2 1.8 2.3 1.9 1.6 0.5 0.3 1.4 IN79 - 1 3.4 1.9 0.1 0.9 0.5 0.0 1.4 IN80 - 4 2.8 2.0 0.4 1.2 0.6 0.1 1.2 IN80 - 3 2.9 2.0 0. 5 1.4 0.7 0.1 1.2 IN97 - 1 66.6 0.6 0.0 0.0 0.0 0.0 15.4 IN95 - 1 61.9 1.8 0.0 0.0 0.0 0.0 17.7 IN95 - 2 61.8 1.3 0.0 0.0 0.0 0.0 19.8 IN95 - 3 61.9 1.1 0.0 0.0 0.0 0.0 19.2 IN96 - 1 51.9 2.1 0.0 0.6 0.0 0.0 21.1 IN96 - 2 56.5 1.6 0.0 0.7 0.0 0.0 21.3 IN96 - 3 51 . 9 1.1 0.0 0.5 0.0 0.0 19.2 IN105 - 1 0.0 0.3 4.4 0.0 0.0 0.0 6.2 IN103 - 1 1.4 0.4 0.4 0.9 1.1 0.0 7.3 IN103 - 2 1.4 0.4 0.8 0.8 1.6 0.0 6.6 IN103 - 3 1.3 0.5 1.1 0.7 1.8 0.0 6.3 IN104 - 1 1.3 0.4 0.3 2.4 0.9 0.0 7.3 IN104 - 2 1.5 0.4 0.3 1.4 1.3 0.0 7.3 IN104 - 3 1.7 0.5 0.4 1.0 1.7 0.0 6.9 IN115 - 2 52.2 1.8 0.0 0.6 0.0 0.0 12.7 IN114 - 1 63.4 1.7 0.0 0.3 0.0 0.0 2.7 IN115 - 3 53.3 2.4 0.0 0.4 0.0 0.0 10.9 IN115 - 1 55.4 1.2 0.0 0.5 0.0 0.0 8.4 IN107 - 1 0.9 12.2 3.7 1.9 0.6 0.1 6.3 IN106 - 1 1.3 5.8 0.7 1.7 0.7 0.1 4 .9 IN106 - 2 1.1 6.6 1.3 2.0 0.8 0.1 4.4 IN106 - 3 1.1 8.2 1.8 2.3 0.9 0.1 4.3 IN108 - 1 1.8 4.9 0.0 1.0 0.7 0.1 7.3 IN108 - 2 1.7 6.5 0.3 1.4 0.7 0.1 7.3 66 Table E.2 (c ) IN108 - 3 1.4 7.4 0.7 1.6 0.8 0.1 7.2 IN91 - 1 0.0 4.2 4.6 1.8 0.0 0.0 0.9 IN89 - 1 1.4 4 .7 0.7 1.1 0.0 0.0 0.9 IN89 - 2 1.5 4.5 1.0 1.3 0.0 0.0 0.9 IN89 - 3 1.4 4.5 1.4 1.6 0.0 0.0 1.0 IN90 - 1 0.8 2.3 0.2 0.4 0.0 0.0 1.0 IN90 - 2 1.4 3.0 0.4 0.8 0.0 0.0 0.9 IN90 - 3 1.4 2.5 0.5 1.1 0.0 0.0 0.9 IN94 - 1 7.6 7.9 9.3 2.8 0.0 0.0 2.3 IN92 - 1 7.3 7.1 0 .7 2.0 0.6 0.6 2.9 IN92 - 2 7.6 8.0 1.0 2.7 0.9 0.6 2.5 IN92 - 3 6.4 9.3 3.1 3.3 1.0 0.0 2.2 IN93 - 1 6.5 4.8 0.0 1.6 0.8 0.7 4.7 IN93 - 2 5.1 5.2 0.5 2.1 1.1 0.7 3.8 IN93 - 3 5.3 5.7 0.5 2.5 1.2 0.7 3.1 IN102 - 1 7.3 0.7 9.6 1.8 0.0 0.0 9.4 IN100 - 1 11.5 2.2 1 . 3 1.4 0.0 0.0 8.1 IN100 - 2 14.1 1.3 4.3 1.5 0.0 0.0 6.2 IN100 - 3 11.2 0.9 6.1 1.8 0.0 0.0 6.4 IN101 - 1 13.1 2.2 0.0 1.2 0.0 0.0 16.7 IN101 - 2 7.5 1.6 0.6 1.5 0.0 0.0 15.9 IN101 - 3 10.4 1.5 0.2 1.9 0.0 0.0 13.8 E.3. Gaseous Products Selectivity Table E . 3 . bimetallic nickel experiments. 1 Sum of both liquid and selectivity. Exp # Selectivity (%) Carbon Monoxide Carbon Dioxide Methane Ethane Propane Sum Al l 1 IN86 - 1 0.1 2.5 8.2 2.4 1.4 97.7 IN86 - 2 0.2 2 .0 7.1 2.4 1.0 97.2 IN86 - 3 0.2 1.8 7.0 1.6 0.9 97.1 IN87 - 1 0.2 5.1 11.3 2.6 3.5 96.9 67 Table E. 3 (c ) IN87 - 2 0.2 3.2 7.6 2.8 2.0 97.5 IN87 - 3 0.2 2.9 7.2 2.0 1.8 96.7 IN74 - 1 0.2 1.3 7.8 0.6 0.1 97.3 IN76 - 3 0.1 2.9 7.2 0.8 0.6 99.9 IN76 - 2 0.2 4.2 7. 8 0.9 1.8 95.5 IN76 - 3 0.2 2.6 5.1 1.2 1.1 94.7 IN79 - 3 0.2 0.9 6.3 1.1 0.2 95.9 IN78 - 2 0.2 0.7 6.5 0.5 0.2 96.1 IN80 - 1 0.4 0.8 5.9 0.6 0.2 93.9 IN79 - 2 0.2 2.4 10.3 1.4 2.1 97.1 IN78 - 1 0.2 1.2 8.3 0.9 1.3 98.1 IN80 - 2 0.2 1.1 6.9 0.8 0.1 96.5 IN79 - 1 0 .4 4.4 12.0 1.5 1.1 95.1 IN80 - 4 0.4 2.9 9.9 1.5 2.8 99.7 IN80 - 3 0.3 2.5 9.3 1.3 1.9 96.6 IN97 - 1 0.9 0.5 1.6 1.0 0.0 91.9 IN95 - 1 1.2 0.3 1.8 1.7 0.0 91.6 IN95 - 2 1.0 0.2 2.6 1.5 0.0 93.8 IN95 - 3 0.9 0.6 2.8 1.5 0.0 93.7 IN96 - 1 1.5 2.2 3.2 2.9 0.0 92. 2 IN96 - 2 1.4 1.3 1.2 1.8 0.0 93.2 IN96 - 3 1.2 2.1 8.5 2.7 1.3 95.3 IN105 - 1 0.2 0.6 6.1 1.1 1.3 92.2 IN103 - 1 0.4 2.6 8.0 1.1 6.5 95.5 IN103 - 2 0.4 1.8 7.4 1.1 5.0 93.4 IN103 - 3 0.4 1.2 7.0 1.3 4.0 93.5 IN104 - 1 0.9 7.7 18.6 2.9 15.9 92.7 IN104 - 2 0.6 4.2 1 2.1 2.6 9.6 93.1 IN104 - 3 0.2 2.6 6.3 0.9 6.1 90.1 IN115 - 2 1.6 0.0 17.0 0.0 2.0 95.5 IN114 - 1 2.2 0.0 20.8 0.0 1.9 93.7 IN115 - 3 1.8 0.0 17.8 0.0 2.2 95.4 IN115 - 1 2.1 0.2 18.3 0.0 3.0 93.5 68 Table E. 3 (c ) IN107 - 1 0.1 0.3 3.5 1.5 0.5 94.7 IN106 - 1 0 .3 1.7 6.1 3.0 2.9 96 .0 IN106 - 2 0.2 1.2 5.4 3.1 2.2 95.6 IN106 - 3 0.2 1.0 5.6 3.2 1.7 95.5 IN108 - 1 0.4 3.4 8.6 5.5 6.2 96.0 IN108 - 2 0.2 1.6 4.7 3.9 2.9 92.9 IN108 - 3 0.2 1.3 4.6 3.5 2.5 94.0 IN91 - 1 0.2 3.7 12.3 2.4 1.3 91.9 IN89 - 1 0.5 9.0 18.7 3.8 4.3 9 6.4 IN89 - 2 0.4 7.3 16.0 4.0 3.7 97.1 IN89 - 3 0.4 6.2 14.6 4.2 3.2 97.1 IN90 - 1 0.3 18.4 29.6 7.0 10.7 94.9 IN90 - 2 0.5 14.6 24.1 5.3 3.5 94.0 IN90 - 3 0.5 12.4 20.2 6.4 6.8 96.3 IN94 - 1 0.1 0.4 10.5 2.9 1.0 92.7 IN92 - 1 0.2 3.7 15.0 4.4 4.1 97.0 IN92 - 2 0. 2 2.7 14.7 2.9 3.4 96.3 IN92 - 3 0.2 2.0 14.3 2.7 2.7 95.9 IN93 - 1 0.1 4.8 14.9 5.4 6.3 96.5 IN93 - 2 0.2 3.0 12.6 3.0 4.2 95.8 IN93 - 3 0.2 2.3 11.2 3.0 3.5 95.5 IN102 - 1 0.0 0.1 1.4 4.3 0.1 88.1 IN100 - 1 0.6 0.2 1.4 9.0 0.0 94.3 IN100 - 2 0.6 0.4 2.8 6.4 0 .2 91.7 IN100 - 3 0.6 0.6 4.4 4.7 0.7 92.0 IN101 - 1 0.2 0.1 0.3 1.2 0.0 95.3 IN101 - 2 0.6 1.0 2.8 4.6 1.0 94.5 IN101 - 3 0.5 0.3 1.0 5.3 0.0 95.3 69 APPENDIX F: Temperature programmed Desorption (TPD) Profiles F.1. NH 3 TPD (Acidic Sites) -0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0 100 200 300 400 500 600 700 Concentration (cm 3 /min) Temperature ( o C) NH 3 TPD Ni/Mo (2/1) Ni/Pd (4/1) Ni/Co (2/1) Ni/Fe (2/1) Ni/Pt (4/1) Ni/Cu (2/1) Ni (III) (b) Figure F . 1 . Acidic sites measurement for bimetallic catalysts . (a) Temperature profil e, (b) NH 3 Temperature Programmed Desorption (TPD) diagram. 0 100 200 300 400 500 600 700 0 20 40 60 80 100 120 140 160 180 200 Temperature ( o C) Time (min) Temperature vs Time (a) 70 F.2. CO 2 TPD (Basi c Sites) 0 100 200 300 400 500 600 700 0 20 40 60 80 100 120 140 160 180 200 Temperature ( o C) Time (min) Temperature vs. Time (a) -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 100 200 300 400 500 600 700 Concentration (cm 3 /min) Temperature ( o C) CO 2 TPD Ni/Co (2/1) Ni/Mo (2/1) Ni/Pd (4/1) Ni/Pt (4/1) Ni/Fe (2/1) Ni/Cu (2/1) Ni (III) (b) Figure F . 2 . Basic sites measurement for bimetallic catalysts. (a) Temperature profile, (b) CO 2 Temperature Programmed Desorption (TPD) diagram. 71 F.3. H 2 TPD -0.05 0 0.05 0.1 0.15 0.2 0 100 200 300 400 500 600 700 Concentration (cm 3 /min) Temperature ( O C) H 2 TPD Ni/Mo (2/1) Ni/Pt (4/1) Ni/Pd (4/1) Ni/Fe (2/1) Ni/Cu (2/1) Ni/Co (2/1) Ni (III) (b) 0 100 200 300 400 500 600 700 0 50 100 150 200 250 Temperature ( o C) Time (min) Temperature vs Time (a) Figure F . 3 . H 2 uptake meas urement for bimetallic c atalysts. (a) Temperature profile, (b) H 2 Temperature Programmed Desorption (TPD) diagram. 72 3 Guerbet Economical Analysis 3.1 Introduction In addition to the studies of catalyst performance for the Guerbet reaction, a series of papers have been published [ 89 - 92 ] that compare economics and environmental impacts of n - butanol formation from ethanol via the Guerbet reaction with n - butanol formation via the traditional petroleum - based [ 92 ] and with n - butanol production via ABE fermentation. Feed stocks examined for ABE fermentation include Brazilian sugar cane [ 91 , 93 ] , corn grain [ 93 - 95 ] , corn stover [ 95 ] , wheat straw [ 94 ] , Canadian pulp pre - hydrolysate [ 96 ] , and others [ 97 ] . In the above studies, n - butanol produ ction from ethanol is integrated into the biorefinery, and the overall economics and environmental impacts of th e integrated refinery are reported. Two scenarios for Guerbet reactions are examined: vapor phase reaction at high temperature and low pressure as proposed by Tsuchida et al. [ 39 , 54 ] , and condensed phase reaction at lower temperature and elevated pressure [ 22 ] . These studies make several assumptions regarding reaction rates, catalyst stability, and byproduct purific at ion that have significant effects on the overall economic and environmental im pacts from the different processes. Because n - butanol production is integrated into the overall ethanol biorefinery, it is difficult to gain a realistic assessment of the envir on mental impact or economic potential related specifically to n - butanol producti on. The economics of producing 2 - ethylhexanol from ethanol via Guerbet reactions has also been examined [ 98 ] , but yields used in the analysis do not reflect available experimental values so the validity of the analysis is uncertain. In this chapter, we describe the economic analysis of a continuous con de nsed phase stand - alone process for the production of n - butanol and higher alco hols using ethanol as feed stock. The 73 process includes an improved distillation concept to purify n - butanol, and produces a mixed C 6 + alcohols stream as a byproduct instead of se parating out individual minor products formed. Several alternate process confi gurations are evaluated as well in order to better ascertain the required selling price of n - butanol from ethanol. The economic analysis for each process configuration investig at ion is based upon a rigorous Aspen Plus V8.4 simulation using the S - R Polar eq uation of state from our prior work ; all parameters were fitted by Jordison et al [ 25 ] . 3.2 Process Concept and Design Parameters 3.2.1 Process Concept The initial results in the continuous fixed bed reactor from Chapter 2 form the basis for a process to convert ethanol to n - butanol a nd higher alcohols. The process concept consists of the reactor for ethanol conversion and a separation train to produce pure n - butanol and a mixture of C 6 + alc ohols as saleable products. In contrast to some other reports of Guerbet reaction processes [ 89 - 92 , 99 ] , in this work minor gas and liquid byproducts are burned as fuel for steam generation to provide energ y for the process; capital and operating expenditures to recover them as pure products are not justified because of their low c oncentration in the product mix. The ethanol - to - higher alcohols Guerbet reaction process is intended to be located adjacent to an e thanol plant, and thus there are two strategies for configuring the combined facility. The first approach, used in several ec onomic analyses of Guerbet reactions with sugar cane - based ethanol production [ 89 - 92 ] , is to pass the entire product stream from ethanol fermentation through the Guerbet process, wherein partial conversion to higher alcohols takes place and the unreacted ethanol becomes the primary plant outpu t. The second approach, adopted in this work, is to divert 74 only a portion of the product et hanol to the Guerbet process, and then recycle unreacted ethanol within the Guerbet process to give an overall conversion approaching 100%. The key reaction paramete r for most renewables - based chemical processes is selectivity to desired products. In this ethanol condensation process, both selectivity and ethanol conversion are important, because reactions are sequential and thus selectivity depends on conversion. Fur th er, because the overall process conversion of ethanol is essentially 100%, the extent of ethanol recycle and thus distillation column size depends on per - pass reactor conversion. To illustrate the dependence of process economics on these parameters, a se t of four base cases with different combinations of ethanol conversion and alcohol selectiv ities is examined. These are designated by the two - n - butanol, 16% C 6 + r high (72% n - butanol, 22% C 6 + alcohols) selectivity. The case designated 35L (35% ethanol conversion, low selectivity) represents results obtained in the experimental fixed - bed reactor studies reported in Chapter 1. The other cases (40L, 35H, 70H) reflect h igher ethanol conversion or selectivity, in order to ascertain future potential for n - but anol production via ethanol condensation. The flow diagram used as the basis for the Aspen Plus V8.4 simulation of the four base cases for the proposed ethanol conde ns ation process is given in Figure 3 . 1 . Fresh ethanol (Stream F) is combined with recycled ethanol (S6), pressurized to 100 bar, and preheated to the reaction temperature of 230 o C in heat exchangers H1 and H2. The reactor (FBR) i s a fixed - bed shell and tube reactor in which the heat of reaction generated on the tube side, where catalyst is loaded, is removed by generating steam on the shell side at approximately 230 o C. This steam is used to partially preheat the reactor feed in H 1. 75 The effluent stream (S2) from the fixed - bed reactor is partially flashed from 100 bar to a lower pressure (Valve V1) and fed to Column TC1, where ethanol, water, and light byproducts are taken as the distillate (S3) and n - butanol, mixed C 6 + alcohols, a nd in some cases water exit as the bottoms strea m (S8) of TC1. Further details regarding the ethanol/ n - butanol separation carried out in Column TC1 are given below, along with an explanation of the two process alternatives highlighted in Figure 3 . 1 for recovering pure n - butanol and mixed C 6 + alcohols. The light byproducts formed in reaction (chosen for simulation as ethyl acetate, diethyl ether, CH 4 , CO 2 , and H 2 , see Appendix D for the complete byproduct slate) are recovered as a m ix ed distillate stream (S4) in Column TC2 and used as fuel for steam generation. The bottoms stream from Figure 3 . 1 . Process concept for the four base cases of ethanol conversion to n - butanol and h igher alcohols. The valu es of reboiler, condenser and heat exchanger duties are for Case 35L. 76 Column TC2, consisting of ethanol and water, is separated in Column TC3 into an azeotropic ethanol/water mixture (S6) for re cycling and pure water (S7) a s a bottoms product. Several pressures were examined in preliminary simulations of the separation train; an absolute pressure of 5.0 bar was chosen for columns TC1 and TC2, and an absolute pressure of 1.0 bar was chosen for Col umn TC3 and the n - butanol/C 6 + alcohol separation (Column TC4 or Columns TC4 - TC6). These pressures are high enough to allow column condensers to operate with air cooling instead of refrigeration, yet are low enough to avoid excessively high temperatures in column reboilers. Other com bi nations of pressures could conceivably further improve process efficiency and economics, but no other pressure combinations were evaluated in this work. The two process alternatives highlighted in the lower right side of Figure 3 . 1 represent two process scenarios depending on whether or not Stream S8 from Column TC1 contains water. If Stream S8 is dry, then n - butanol and C 6 + alcohols are se parated in Column TC4 into pure (99.87 ± 0.04 mol%) n - butanol (S9) and C 6 + alcohols (chosen f or simulation as equimolar quantities of 1 - hexanol and 2 - ethyl - 1 - butanol, together >99.75 mol% in S10) as shown in the upper block in Figure 3 . 1 . If Stream S8 contains water, then C 6 + alcohols are recovered as the bottoms stream ( >99.75 mol%, S12) of Column TC4, and the distillate (S11) of Column TC4 containing n - butanol and wat er, which form a heterogeneous azeotrope, is separated using the classic approach of two distillation columns (TC5 and TC6) with a decanter as shown in th e lower block in Figure 3 . 1 . Capital and operating costs associated with this latter approach are significant, as the n - butanol - water separation columns (TC5 and TC6) require high reflux ratios and recycling of intermediate stream s to obtain dry (99.87 ± 0.04 mol%) n - b utanol. The composition of the reactor effluent (Stream S2) entering Column TC1 determines whether or not Stream S8 from Column TC1 contains water, and thus which of the two process scenarios 77 above is used. The ethano l/ water/ n - butanol residue curve map sho wn in Figure 3 . 2 (generated using the S - R Polar equation of state in Aspen Plus V8.4) illustrates the two cases. In this residue curve map, a distillation boundary ar ises because ethanol and n - butanol both form minimum boiling azeotropes with water. If C 6 + alcohol products (recovered in Column TC4) and light byproducts (recovered in Column TC2) are momentarily neglected, so that Columns TC2 and TC4 can be neglected and Stream S5 can be conside re d equivalent to Stream S3 in Figure 3 . 1 , then the residue curve map in Figure 3 . 2 applies directly to Columns TC1 and TC3. If sufficient ethanol is present in the reactor effluent (S2), generally corres po nding to a maximum per - pass ethanol conversion in the FBR of 40% (Cas es 35L, 35H, 40L), then all water in Stream S2 (both produced in reaction and present in the reactor feed) can be removed with ethanol into the distillate (S3) of Column TC1. Negligible n - butanol is carried to the distillate (S3), leaving dry n - butanol as the bottoms product of TC1. This separation is represented as the straight line S8 - S2 - S3 for Case 35L in Figure 3 . 2 ; the distillate (S3) composition lies close t o the distillation boundary and contains less than 0.001 mole fraction n - butanol, while the bottoms (S8) composition approaches pure n - butanol. Separation of the distillate mixture (S3 equivalent to S5 without light byproducts present) in the subsequent Co lumn T C3, represented by the line S7 - S3 - S6 for Case 35L in Figure 3 . 2 , produces the ethanol - water azeotrope as distillate (S6) and water containing trace n - butanol as the bo ttoms product (S7). This separation in Column TC3 cross es the distillation boundary on the residue curve map. While crossing the distillation boundary is not usually possible, such crossings are possible in cases where the boundary has significant curvatur e [ 100 ] . To ensure that crossing the distillation boun da ry in the Aspen Plus V8.4 simulation of this system was not an artifact of the S - R Polar properties package, the ethanol/ n - butanol/water residue curve map was also generated and the separations in Co lumns TC1 and TC3 were also simulated using 78 the Non - Ran do m Two Liquid (NRTL) properties package. The same results were obtained using the NRTL properties package as with the S - R Polar equation of state; the use of very different thermodynamic properties pa ckages to obtain the same result provides validation of t he crossing of the distillation boundary and the separations in Columns TC1 and TC3. Figure 3 . 2 . Residue curve map for ethanol/n - butanol/water system at 5.0 bar absolute (Aspen Plus V8.4 SR - Polar eq uation of state). Dashed lines represent material balances for distillation columns TC1 (S8 - S2 - S3) and TC3 (S6 - S3 - S7) for Cases 35L and 70H with feed and product streams excluding light n - butanol/water binary azeotrope; ethanol/water binary azeotrope. Distillation boundary is the curved solid line between azeotropes. 79 As per - pass ethanol conversion in the FBR increase ab ove 40% (Case 70H), the reactor effluent (S2) contains less ethanol and more water and n - butanol, thus shifting po in t S2 in Figure 3 . 2 away from the ethanol vertex of the residue curve map (to 41 mol% H2O, 32 mol% ethanol, and 27 mol% n - butanol for Case 70H). Attempting to recover pure n - butanol from Column TC1 as this shift in S2 occurs woul d result in ~8 mol% n - butanol in the distillate stream (S3) from Column TC1, a s Stream S3 is limited in composition by the distillation boundary. This n - butanol in the distillate Stream S3 from Column TC1 would be lost from the process, or would have to be r ecovered via additional separation. Thus, for high ethanol conversion in the FBR, a different separation in Column TC1 must be carried out. This separation is shown as the line S3 - S2 - S8 for Case 70H in Figure 3 . 2 , where an ethan ol /water mixtu re is taken as the distillate stream (S3) and a binary mixture of n - butanol and water is taken as the bottoms product (S8). The binary n - butanol/water mixture is then separated into pure n - butanol and pure water with the two column and decant er system show n in the lower box in Figure 3 . 1 . In the actual process simulation of Column TC1 with the complete reactor effluent (Stream S2), the light byproducts formed in reacti on rapidly move upward in the vapor phase of Colum n TC1 and have little influence on the ethanol/ n - butanol separation. The C 6 + alcohols in Stream S2, which have low miscibility with water, move rapidly downward in the liquid phase of Column TC1 and actuall y aid in facilitating the separation of water from n - butanol in Column TC1 as described above. It is noted here that there are other scenarios possible for recovery of n - butanol from mixed alcohol streams. Patrascu propose a double wall distillation column integrated with other columns to recovery of dry n - butanol from ABE fermentation broth [ 101 ] . Pervaporation has also been examined as another route to recovery of n - butanol from fermentation [ 102 ] . Michaels et al. 80 describe d the use of benzene to break the n - butanol - water azeotrope i n an etha nol condensation reaction process [ 99 ] . Energy for distillation reboilers and heating process streams is provided by high pressu re steam at 257 o C and 45 bar produced in a steam generator fueled by natural gas and by the light byproducts from the Guerbet reaction. Because the plant is proposed to be located adjacent to an ethanol facility and not necessarily close to an external co oling wat er source, for the base case analysis all energy removed from the process (reactor, coolers, and column condensers), if not used elsewhere for heating, is ultimately rejected to air via air - cooled heat exchangers. The use of external water cooling is exami ned as an alternate process configuration. 3.2.2 Definition of Design Parameters Capital and operating costs for ethanol conversion to higher alcohols have been evaluated from Aspen Plus V8.4 simulations with the SR - Polar equation of state as the proper ties pack age [ 9 ] . The four base case scenarios and several variations of the base process concept shown in Figure 3 . 1 have been examined. The outcome of the analysis determines the required selling price of n - butanol for these different cases as a function of return on investm ent (ROI). As in any economic analysis, assumptions must be made, and va lues of parameters defined, in order to properly carry out the calculations. General design parameters used in the techno - economic analysis are summarized in Table 3 . 1 ; additional parameters defining unit operations for specific c ases are given in Table 3 . 2 and Table 3 . 3 . 81 Table 3 . 1 . Parameters for techno - economic anal ysis Parameter Value Location Midwest U.S. Plant Capacity (10 6 kg n - bu tanol/yr) 75 (25 million gallons) Plant Lifetime 10 years Ethanol feed cost $0.53/kg ($1.65/gallon) C 6 + alcohol selling price ($/kg) 0.75 × n - butanol selling price UTILITIES Natural gas cost $3.21/10 6 kJ ($3.39/10 6 Btu) Electricity cost $18.6/10 6 kJ ($0.067/kWh) Cooling water cost $14.80/10 3 m 3 GENERAL DESIGN PARAMETERS REACTOR Tube size 2.5 cm 20 BWG, 316 SS Tube spacing 3.2 cm triangular centers Catalyst particle diameter 2.0 mm Catalyst cost (8% Ni/8% La 2 O 3 / - Al 2 O 3 ) $100/kg DISTILLATION COLUMNS Tray type Sieve Tray efficiency 60% Tray spacing 0.61 m Approach to flooding 80% AIR - COOLED HEAT EXCHANGERS AND CONDENSERS Configuration Forced - air finned tube Heat transfer coefficient (bare tube area basis) 7 70 W/m 2 /K Fin area / bare tube area ratio 17 REBOILERS Configuration Shell and tube, single pass Energy source Steam (45 bar, 257 o C) Heat transfer coefficient 850 W/m 2 /K WATER - or GLYCOL - COOLED HEAT EXCHANGERS Configuration Shell and tube, single p ass Heat transfer coefficient 680 W/m 2 /K 82 Table 3 . 2 . Specifications of reactor and distillation columns for four base case scenarios. 1 Case designations: The number refers to per - pass ethanol conversion in t he fixed - low selectivities of 60% to n - butanol, 16% to C 6 refers to high selectivities of 72% to n - butanol and 22% to C 6 + alcohols. CASE 1 35L 40L 35H 70H Reactor (100 bar, 230 o C) 5144 5144 5144 8117 Tube length (m) 16.5 17.8 16.5 18.8 Catalyst mass (10 4 kg) 3.60 3.93 3.60 6.56 WHSV (kg EtOH/kg catalyst/h) 1.45 1.15 1.45 0.33 Cost (10 3 $) 2,350 2,580 2,350 4,500 Column 1 (TC1) (5 bar absolute pressure) Number of actual sta ges 44 45 49 67 Feed stage 33 28 33 21 Diameter (m) 3.0 2.8 2.7 1.5 Reflux ratio (L o /D) 1.2 1.2 1.5 3.0 Boilup ratio (V N /B) 12.7 11.2 10.8 3.2 Cost (10 3 $) 670 670 630 380 Column 2 (TC2) (5 bar absolute pressure) Number of actual stages 57 47 5 7 49 Feed stage 35 33 23 40 Diameter (m) 1.7 1.6 1.2 0.6 Reflux ratio 4.5 4.5 6.8 3.5 Boilup ratio 0.9 1.0 0.5 0.6 Cost (10 3 $) 370 330 270 140 Column 3 (TC3) (1 bar absolute pressure) Number of actual stages 40 44 50 55 Feed stage 25 28 35 1 6 Diameter (m) 4.2 3.5 3.5 1.9 Reflux ratio 3.0 2.5 2.5 3.5 Boilup ratio 16.8 11.7 12.5 22.7 Cost (10 3 $) 1,110 850 920 420 Column 4 (TC4) (1 bar absolute pressure) Number of actual stages 25 25 25 25 Feed stage 13 11 13 13 83 Table 3.2 (c ) Diameter (m) 1.9 1.9 1.9 2.7 Reflux ratio 2.5 2.5 2.5 2.0 Boilup ratio 16.6 16.6 14.4 29.1 Cost (10 3 $) 290 390 280 460 Column 5 (TC5) (1 bar absolute pressure) Number of actual stages - - - 12 Feed stage - - - 11 Diameter (m) - - - 2.4 Reflux ratio - - - 2.5 Boilup ratio - - - 6.2 Cost (10 3 $) - - - 280 Column 6 (TC6) (1 bar absolute pressure) Number of actual stages - - - 20 Feed stage - - - 6 Diameter (m) - - - 2.0 Reflux ratio - - - 1.5 Boilup ratio - - - 7.4 Cost (10 3 $) - - - 270 Steam Generator (45 bar, 257 o C) Capacity (10 3 kg/h) 97.4 80.3 67.9 61.0 Cost (10 3 $) 1,520 1,240 1,050 970 Location and scale for an initial n - butanol - from - ethanol facility fit with existing U.S. ethanol production facilities. The price of ethanol ($1.65/gallon, or $0.53/kg) is taken as the average wholesale price over a one year period of 2016 - 2017 [ 103 ] ; across - the - fence prices would be expected to be lower. Utility p rices are taken as conservative values for the Midwestern U.S. [ 104 , 105 ] ; a sensitivity analysis of process economics vs. ethanol price and utility prices is presented later in thi s chapter. Reactor and distillation column specifications are taken from standard references [ 106 , 107 ] and are typic al for initial process equipment design. Catalyst cost is conservatively estimated, as nickel and lanthanum oxide are both inexpensive materials. Air - 84 cooled heat exchanger properties are taken from a handbook on air - cooled exchanger design [ 108 ] ; heat transfer coefficients for shell and tube heat exchangers were taken as conservative (low) values within the range of typical coefficients for each type of heat exchanger [ 109 ] . Table 3 . 3 . Parameters for heat exchangers for Case 35L. 1 All process heat provided by steam (257 o C, 45 bar) from steam generator except for Heater 1. 2 All process cooling provided by direct air cooling in forced convection air - cooled heat exchangers. The area repor ted for air - cooled heat exchangers is bare tube area; fin:tube area ratio = 17:1. Unit Target T ( o C) Approach o C) Heat Duty 1,2 (MW) Area (m 2 ) Cost ( 10 3 $) Function H1 177 53 5.8 68 60 Feed preheat using steam from reactor H2 230 27 5.5 133 80 Feed pr eheat using generated steam H3 95 55 - 10.2 1330 160 Cooler between TC1 and TC2 C1 122 82 - 13.7 218 180 Partial condenser on column TC1 R1 176 81 18.1 263 310 Partial reboiler on column TC1 C2 57 17 - 8.1 619 520 Partial condenser on column TC2 R2 126 1 30 8.7 78 120 Partial reboiler on column TC2 C3 78 38 - 34.8 1180 980 Total condenser on column TC3 R3 99 158 33.1 246 290 Partial reboiler on column TC3 C4 118 78 - 3.6 59 50 Total condenser on column TC4 R4 152 105 4.4 50 80 Partial reboiler on column TC4 In the Aspen Plus simulations, the fixed bed reactor is simulated using the RStoic module with the required catalyst volume calculated from simple second - order kinetics of ethanol conversion with a fixed value of the rate constant (k = 5.7 × 10 - 5 m 6 / kmol E tOH/kg cat/h). Fixing the catalyst activity for all four cases allows comparison of the effect of selectivity on economics (cases 35L vs. 35H) and the comparison of designated per - pass conversion on economics (35L vs. 40L; 35H vs. 70H). In addition, in our prior work [ 25 ] we reported that small quantities of water present in the ethanol feed hav e litt le effect on conversion rate for that reason, we neglect that up to 2.7 wt% water, which comes from ethanol recycle (Stream S6) as the azeotropic composition with water, may be present in the ethanol feed stream (S1) to the reactor. 85 Distillation co lumns are simulated using the rigorous column module RadFrac. Number of stages and reflux ratios are adjusted to achieve desired product purities from each column. Cases 35L, 35H, and 40L use the single - column block in the lower part of Figure 3 . 1 to separate n - butanol from C 6 + alcohols; Case 70H uses the three column plus decanter block in Figure 3 . 1 to produce the C 6 + alcohol product and dry n - butanol. The specifications of each piece of equipment from t he Aspen Plus V8.4 simulation are entered into an in - house Excel spreadsheet that includes standard formulas for calculating the purcha se cost of individual pieces of equipment [ 106 - 112 ] ; these cost formulas are given in Appendix G. The reactor and distillation column specifications for each case are given in Table 3 . 2 . Specifications for the heat exchanger s in Case 35L, including column reb oilers and condensers, are given in Table 3 . 3 . To determine total capital costs of the process, individual equipment purchase costs are su mmed and multiplied by standard multipliers for installation, facilities, engineering, working capital, etc. to arrive at the total capital costs for the process, which is then normalized with the CPI index to 2016 $US [ 113 ] . Operating costs are similarly determi ned by entering raw material and utility requirements for each unit operation and using standar d multipliers as required for labor, site maintenance, marketing, taxes, etc. [ 114 ] . The multipliers used in calculation of various contributions to the overa ll process economic analysis are given in Table 3 . 5 , where total capi tal costs and operating costs for each of the four cases are presented. Depreciation of plant capi tal costs (10 years, straight line) is combined with operating costs and feed stock costs to determine a total cost of production of n - butanol. This value is subtracted from product sale income to give annual net revenue, which is readily converted to an e stimated annual return on investment. 86 3.3 Techno - Economic Analysis Results 3.3.1 Base Case The compositions and flow rates of each of the streams in Figure 3 . 1 for Case 35L are given in Table 3 . 4 . The energy demands for each column reboiler and condenser, and for each heat exchanger for Case 35L are shown in Figure 3 . 1 . A summary of total plant ca pital costs and operating costs for the four base cases is given in Table 3 . 5 . A graph of required n - butanol selling price versus desired ROI for the cases is presented in Figure 3 . 3 . Table 3 . 4 . Composition of Process Streams for 35L Case. Stream F S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 Temperature ( o C) 25.0 230.0 210.0 121.9 57.2 126.5 78.5 98.9 122.1 117.9 151.8 Pressure (bar) 1.0 100.0 100.0 5.0 5.0 5.0 1.0 1.0 5.0 1.0 1.0 Mole Flows (kmol/h) Ethanol 402.8 1124.3 730.7 730.6 9.0 721.6 721.5 0.1 0.1 0.1 0.0 n - Butanol 0.0 0.0 118.0 0.1 0.0 0.1 0.0 0.1 117.9 117.8 0.1 1 - Hexanol 0.0 0.0 10.5 0.0 0.0 0.0 0.0 0.0 10.5 0.0 10.5 2 - Ethyl - 1 - Butanol 0.0 0.0 10.5 0.0 0 .0 0.0 0.0 0.0 10.5 0.1 10.4 Water 0.0 70.4 242.5 242.5 1.0 241.5 70.4 171.1 0.0 0.0 0.0 Carbon Dioxide 0.0 0.0 29.5 29.5 29.5 0.0 0.0 0.0 0.0 0.0 0.0 Methane 0.0 0.0 99.6 99.6 99.6 0.0 0.0 0.0 0.0 0.0 0.0 Diethyl Ether 0.0 0.0 6.5 6.5 6.5 0.0 0.0 0.0 0.0 0.0 0.0 Ethyl Acetate 0.0 1.1 9.5 9.5 8.4 1.1 1.1 0.0 0.0 0.0 0.0 Hydrogen 0.0 0.0 6.0 6.0 6.0 0.0 0.0 0.0 0.0 0.0 0.0 Total 402.8 1195.8 1263.3 1124.3 160.0 964.3 792.9 171.3 139.0 118.0 21.0 Figure 3 . 3 shows that produc ing n - butanol via catalytic ethanol Guerbet condensation has significant potential for chemical applications of n - butanol, especially if higher selectivity to the desired alc ohol products can be achieved. At present, n - butanol market prices range from $1.3 0/kg to $2.00/kg [ 115 , 116 ] depending on geographic location, so the required selling price reflected in 87 Figure 3 . 3 falls in the range of those values under most conditions. It should be noted that n - butanol potent ial as a fuel component or oxygenate in gasoline in the U.S. is not economical in the present scenario, as n - bu tanol does not compete with current ethanol ($0.50 - 0.60/kg) or gasoline ($0.80 - 1.00/kg) prices. Table 3 . 5 . Economic analysis of base cases (Basis: 25 million gallons (75 million kg) n - butanol/yr). 1 Fixed costs include capital depreciation, taxes, and insurance. 2 General Expenses include administration, distribution and selling costs, R&D, and financing costs Case (Ethanol conversion, Selectivity) 35L 40L 35H 70H Ethanol conversion (%) 35 40 35 70 n - Butanol selectivity (%) 60 60 72 72 C 6 + alcohol selectivity (%) 16 16 22 22 Parameter Reference CAPITAL COSTS (10 6 $) Equipment Purchase Cost (PC) Cal culated 9.14 8.33 7.64 9.61 Instrumentation, controls 1.21 PC 11.02 10.05 9.21 11.59 Facility 0.75 PC 6.77 6.18 5.66 7.12 Contractor, engineering 0.82 PC 7.47 6.81 6.24 7.86 Contingency 5% CAPEX 1.88 1.72 1.57 1.98 Working Capital 5% CAPEX 1.81 1.65 1 .52 1.91 TOTAL CAPITAL COSTS (CAPEX) 38.09 34.75 31.84 40.07 OPERATING COSTS (10 6 $/yr) Raw Materials Calculated 88.05 89.26 73.83 75.75 Utilities Calculated 5.10 3.81 6.13 2.58 Labor and maintenance 6% OPEX 8.18 7.89 6.93 7.46 Laboratory / Analytic al 1% OPEX 1.30 1.26 1.10 1.16 Royalties / Licensing Fees 2% OPEX 2.68 2.64 2.28 2.30 Fixed Costs 1 16% CAPEX 6.21 5.66 5.19 6.53 Overhead 6% OPEX 8.04 7.92 6.84 6.90 General Expenses 2 11% OPEX 14.35 13.96 12.17 12.75 TOTAL OPERATING COSTS (OPEX) (10 6 $/yr) 133.90 132.42 114.47 115.43 88 The economic analysis shows that increasing the per - pass ethanol conversion from 35% to 40% at low selectivity values reduces r equired selling price of n - butanol, because the ethanol recycle stream and thus the qu antity of ethanol fed to the reactor are smaller at higher conversion. Increasing ethanol conversion also reduces utility costs, as the contribution of the light byproduc ts formed to overall energy requirements for the process increases from 50% for Case 3 5L to 60% for case 40L, reducing utility costs by $1.3 million annually. Required n - butanol selling price increases when per - pass conversion is increased from 35% to 70% at high selectivity, because 1) a much larger reactor is required to achieve 70% conv ersion, and 2) the smaller quantity of ethanol in the reactor effluent (S2) at 70% conversion requires that the n - butanol/water azeotrope be broken with the traditional t wo column/decanter approach shown in the lower block of Figure 3 . 1 . Thus, for this process, the preferred design is to run the reactor at a per - pass conversion of 35 - 40% to lower capital costs and avoid having to separate the n - 1.20 1.30 1.40 1.50 1.60 1.70 1.80 0% 10% 20% 30% 40% 50% Required BuOH Selling Price ($/kg) Return on Investment (ROI) 35L 40L 70H 35H Figure 3 . 3 . Required n - butanol selling price vs. ROI for four base cases in Table 3.5. 89 butanol/water azeotrope using multiple columns. Of course, if a more act ive catalyst can be de veloped, then the higher per pass conversion through the reactor may become preferred. 3.3.2 Sensitivity Analysis of Key Cost Drivers An analysis of n - butanol required selling price dependence on ethanol feed cost, equipment purchase cost, and utility costs has been carried out for the Case 35L. For each of these parameters, the effect of increasing the base case value by 10%, 20%, and 30% on n - butanol required selling price has been evaluated. All other cost calculations in each case are sc aled according to guid elines given in Table 3 . 5 . Results for the complete range of ROI are given in Figure 3 . 4 - Figure 3 . 6 . Figure 3 . 4 . Sensitivity analysis of required n - butanol selling price dependence on purchase equipment costs for Case 35L. Per centage in legend refers to increment in equipment purchase costs over Case 35L equipment purchase costs of $9.14 million. 1.45 1.50 1.55 1.60 1.65 1.70 1.75 1.80 0% 10% 20% 30% 40% 50% Required BuOH Selling Price ($/kg) Return on Investment (ROI) 35L 35L+10% 35L+20% 35L+30% 90 1.40 1.50 1.60 1.70 1.80 1.90 2.00 2.10 0% 10% 20% 30% 40% 50% Required BuOH Selling Price ($/kg) Return on Investment (ROI) 35L 35L+10% 35L+20% 35L+30% Figure 3 . 5 . Sensitivity analysis of req uired n - butanol selling price dependence on ethanol feed cost for Case 35L. Percentage in legend refers to increment in ethanol feed cost relative to Case 35L cost of $1.65/gallon ($0.53/kg). 1.45 1.50 1.55 1.60 1.65 1.70 1.75 0% 10% 20% 30% 40% 50% Required BuOH Selling Price ($/kg) Return on Investment (ROI) 35L 35L+10% 35L+20% 35L+30% Figure 3 . 6 . Sen sitivity analysis of n - butanol required selling price dependence on utility costs for Case 35L. Percentage in legend refers to increment in total annual utility costs relative to Case 35L utility costs of $5.1 million. 91 For 25% ROI, increasing equipment purchase cost by 30% (from $9.14 million to $11.88 million) increases the required n - butanol selling price by $0.09/kg ( Figure 3 . 4 ). For the same ROI, increasing ethanol feed cost by 30% (from $1.65/gallon to $2.15/gallon) increases the require d n - butanol selling price by $0.36/kg ( Figure 3 . 5 ). Fin ally, at the same ROI, increasing total utility (natural gas + electricity) costs by 30% (from $5.10 million to $6.33 million) increases the required n - butanol selling price b y $0.02/kg ( Figure 3 . 6 ). Clearly, the process economics for n - butanol production from ethanol are most heavily affected by ethanol feed cost. 3.3.3 Alternate Process Configurations In addition to the four base cases examined, several al ternate design sce narios have been simulated in Aspen Plus V8.4, and the techno - economic analysis has been carried out for each scenario. The results of these analyses are described in the following paragraphs. 3.3.3.1 Two - fold Process Scale - up Case 35L was simula ted and analyzed f or n - butanol production of 150 million kg (50 million gallons) per year, twice the size of the base case capacity. Capital cost increased from $38.1 million to $71.1 million; operating costs related to raw materials and utilities essentia lly double for the larger scale. The overall reduction in required n - butanol selling price in scaling up to 150 million kg/yr from 75 million kg n - butanol annually is $0.03/kg n - butanol at 25% ROI. 3.3.3.2 Heat Integration The process configuration in Figure 3 . 1 with the specified process pressures offers the opportunity for heat integrating the condensers of columns TC1 (121 - 122 o C) and TC4 (118 o C) with the reboiler of TC3 (99 o C). The heat - integrated process does not require a dedicated heat exchanger or utilities for the condensers of TC1 and TC4, but instead, direct heat exchange can take place between the distillate vapor streams of TC1 and TC 4 and the reboiler liquid of TC3. Heat 92 integration reduces the overall utility costs for the process from $5.1 million to $3.2 million annually. A smaller steam generator is also required for the heat integrated case. Unfortunately, modestly larger heat ex changers are required for the heat integrated process, because thermal driving forces are sma ller. Because of this, total capital costs for the heat integrated process are $38.5 million vs. $38.1 million for base case 35L. But overall, the required selling price of n - butanol for an ROI of 25% is approximately $0.03/kg lower for the heat integrated case than for the base case 35L. 3.3.3.3 Drying of Ethanol/Water Recycle Stream In the base - case process, unreacted ethanol is recycled as its azeotropic composition with water (Stream S6 in Figure 3 . 1 ); it is assumed, based on our prior studies [ 25 ] , that the recycled water has a negligible effect on ethanol conversion and selectiv ity to higher alcohols. Nevertheless, the cost of removing water from the ethanol recycle stream (S6 ) via a conventional two - unit pressure swing molecular sieve unit was estimated for Case 35L. The molecular sieve unit is assumed to adsorb at 1.7 bar absol ute and regenerate at 0.14 bar absolute, with 10% of the dry product ethanol stream directed to the regeneration of the saturated bed with an 8 h cycle time. The purchase cost of the molecular sieve unit is $2.48 million [ 117 , 118 ] , which increases the total process CAPEX from $38.1 million to $47.7 million. Utility costs are essentially unchanged with addi tion of the molecular sieve unit, as the absence of water in S6 reduces the reboiler loads in separation to offset the increased reboiler duty in TC3 from t he regenerating ethanol stream. The increased capital and operating costs associated with the molecu lar sieve unit together increase the required n - butanol selling price at 25% ROI by $0.07/kg n - butanol. It is worth noting that removing water from the recy cle stream (S6) makes the process equivalent to a process with once - through ethanol flow of a rate e qual to Stream S1 and the same conversion 93 and selectivity. The increase in cost noted above results from drying unreacted ethanol via molecular sieve before sending it to market. Similarly, the base cases are also equivalent to a process with once - through ethanol feed at a rate equal to S1 where only part of the feed ethanol and the exiting wet ethanol are dried by molecular sieves. 3.3.3.4 Method of Heat Removal The four cases reported in Table 3 . 5 use direct air cooling in forced air convection heat exchangers to remove process heat from distillation column condensers and the intercooler between TC1 and TC2. These exchangers are sized accor d the electrical power requirements for the fans are included in utility costs for the processes [ 108 ] . The process design for Case 40L has also been carried out with two alternative heat exchange systems: 1) the use of cooling water from an external source or cooling ponds on - site in direct exchan ge with process streams; and 2) a clo sed - loop glycol cooling system in which process energy is removed via glycol cooling and the glycol is cooled in a secondary air - cooled heat exchanger. For cooling with an external water source, the incoming water is as sumed to be at 25 o C and a maximum co oling water o C. With these design parameters, cooling water demand for the process is 5.3 × 10 3 m 3 /h. For the closed - loop coolant, two heat exchangers are required for each location, leading to an increase in c apital and operating cost. This increase in capital cost results from the large heat exchanger areas required, as the overall temperature drivin g force for heat removal to air must be divided between two heat exchangers. This renders the cost of a facility with closed loop cooling substantially more expensive, as seen in Table 3 . 6 . The higher costs render the closed loop cooling unattractive; the only likely advantage of such a system would be improved control of condenser heat dut y and temperature. In contrast, external water cooling, if such a source is avai lable, is slightly less expensive than air cooling, and has advantages of reduced noise and possible 94 mechanical failures related to air - cooled heat exchange. Returning external water to a river or lake has possible environmental consequences regarding ther mal pollution, so cooling ponds are preferable alternatives if makeup water and sufficient space are available. Table 3 . 6 . Capita l and operating costs for different heat removal options for Case 40L. Parameter Air cooling Glycol closed loop / Air External water TOTAL Capital Costs (CAPEX) ($10 6 ) 34.7 66.5 32.5 Utilities ($10 6 /yr) 3.81 4.17 4.30 TOTAL Production Costs ($10 6 /yr) 13 2.4 144.8 132.3 Required n - butanol selling price at 25% ROI ($/kg) 1.56 1.78 1.55 3.3.3.5 Alternate Location of Ethanol Feed As stated earlier, for Case 70H (70% ethanol conversion) there is insufficient ethanol present in the reactor effluent to remove all wat er produced in reaction to the distillate of Column TC1 without taking substantial n - butanol into the distillate as well . To further examine the possibility of improving the economics of Case 70H, an alternate scenario was examined in which fresh ethanol t o the process is fed into Column TC1 instead of into the FBR. Simulation of this configuration shows that dry n - butanol can be produced as a bottoms of Column TC1 with little n - butanol in the distillate (similar to Column TC1 represented as the line S8 - S2 - S3 for Case 35L in Figure 3 . 2 ), thus eliminating the need for the two column/decanter system to separate the n - butanol/water azeotrope. However, the economic analysis shows that feeding fresh ethanol to Column TC1 leads to substan tial increases in capital and utility costs of Columns TC 1, TC2, and TC3 that more than offset the capital and operating cost savings from eliminating the butanol/water separation columns. At 25% ROI, the required n - butanol selling price with ethanol fed t o Column TC1 is $1.45/kg, as opposed to $1.35/kg for the base case 70H where fresh ethanol is fed to the 95 FBR. Thus, the alternate ethanol feed location is impractical, despite the fact that it eliminates separating the n - butanol/water azeotrope using two c olumns and the decanter. 3.4 Conclusions Based on experimenta l results and analysis of phase equilibria for the continuous condensed - phase reaction of ethanol to n - butanol and C 6 + alcohols, a process concept has been developed that converts all ethanol and pro duces pure n - butanol and a mixed higher (C 6 +) alcohols st ream as products, while minor byproducts formed are burned to provide process energy. The process has been simulated in Aspen Plus V8.4 for several combinations of per - pass ethanol conversion and sel ectivity to product alcohols, and an economic analysis of each combination has been carried out. For n - butanol selectivities achieved in laboratory studies, the required selling price is $1.55 - $1.60/kg, close to current n - butanol selling prices for chemica l applications. If selectivity to higher alcohols can be improved, then the production of n - butanol and C 6 + alcohols from ethanol for use as industrial chemicals becomes attractive. 96 APPENDI CES 97 APPENDIX G. Formulas for Cost Estimation The following secti on contains the formulas used in the estimation of equipment purchase costs for the proposed process. Where identical, the definition of various symbols is not repeated after each equation. G.1. Reactor (Shell and tube configuration) [ 107 ] where D i (in)= shell I.D. p cm = cost multiplier for O.D., pitch and layout angle= 0.98 f cm = cost multiplier for TEMA - type front head= 0.95 r cm = cost multiplier for TEMA - type rear head= 0.9 C T = sum of base cost corrections for shell type, expansion joint, tube length, number of tube passes, shell side and tube side design pressures, materials of construction, and tube gage= 2.3 A ht (ft 2 ) = su rface area for heat transfer E i = escalation index (Chemical Engineering Plant Cost Index) [ 113 ] G.2. Tray column [ 106 ] 98 where f 1 = cost multiplier for column material= 2.1 f 2 = cost multiplier for column material= 1.7 f 3 = cost mu ltiplier for tray type= 0.95 f 4 = cost multiplier for tray numbers= 0.1 - 1.5 N= number of trays W (lb)= vessel weight L (ft)= column height D (ft)= column diameter T b (in)= head thickness T p (in)= shell thickness G.3. Steam generator [ 114 ] where C bg ($)= base steam generator cost 99 G.4. Air cooled heat exchanger [ 108 ] where C ac ( o C $/kW)= cost function= 5 - 11 (process inlet and outlet temperature dependent) T i ( o C)= process inlet temperature T a ( o C)= ambient temperature= 40 o C (*Except for one heat exchanger in the ethylene glycol case where Ta = 30 o C) Q (kW)= heat load G.5. Other heat exchangers [ 119 ] where A HE (ft 2 )= exchanger surface area F D = cost multiplier for the exchanger type = 0.7 for fixed head and 1.4 f or kettle reboiler F P = cost multiplier for the design - pressure (surface area dependent) F M = cost multiplier for the material of construction (surface area dependent) 100 4 Fusel Alcohols Production Studies 4.1 Introduction The predominant pathway for the conversion of hexose sugars (sucrose, glucose, maltose, etc.) in yeast of the Saccharomyces family is the Embden Meyerhof Pathway [ 120 ] . In this pathway, t he [ 121 , 122 ] . Fusel alcohols, the equivalent German uor consist of higher carbon number alcohols such as 3 - methyl - 1 - butanol (also known as isoamyl alcohol, which is the major component), n - propanol, isobutanol, and optically active amyl alcohol [ 123 ] . Produced primarily from fermentation - derived amino acids through a pathway proposed by Ehrlich [ 124 ] , low concentrations of these compounds have essential applications as aroma and flavoring agents in the food and beverage industry [ 122 , 125 - 128 ] . Furthermore, their mixture has the potential for uses in industry such as solvents or cleaners, and can also reac t with various organic acids to make mixed esters that may have desirable properties for the same general applications or as fuels [ 129 ] . F usel alcohols contain at least one hydrogen on the - position of the carbon adjacent to their OH group. Thus, they can contribute to Guerbet reactions if exposed to desirable reaction con ditions. The Guerbet reaction can be a direct aldol condensation betw een two fusel alcohol with the same molecular structure, or it can be a cross - condensation reaction between ethanol and a fusel alcohol, or two different fusel alcohols. As mentioned earl ier, ethanol and fusel alcohols are the products of the same fermenta tion process. Large - scale production of bioethanol through the fermentation process leaves large quantities of fusel alcohols available for further processing. Therefore, trying to produc e a value - added product from ethanol - fusel alcohols mixture enhances the profitability 101 of the fermentation process. If successful , the cost of producing ethanol may be reduced significantly through the sale of higher - value fusel alcohol products. There hav e been few studies on the Guerbet reaction of fusel alcohols. Matsu - u ra et al. studied the conversion of fusel alcohols over a homogeneous Ir - based catalyst at 120 o C and atmospheric pressure and obtained yields of as high as 98% for the self - coupling of C 5 and C 6 alcohols, and 86% for the self - coupling of C 12 alcohol [ 130 ] . They also studied isoamyl alcohol as the feed and were able to get 50 % yield of C 10 alcohol at the same reaction conditions. Later, Busch et al. confirmed the feasibility of the synthesis of branched C 10 alcohols through the Guerbet reaction of i soamyl alcohols at 180 o C and elevated pressure ranges (1.4 - 4.6 bar) using a Pd/C based homogeneous catalyst [ 131 ] . Unfortuna tely, the authors did not provide any analytical data for this reacti on. No studies have been conducted on the Guerbet reaction of an ethanol and isoamyl alcohol mixture thus far. Therefore, the prospects of the batch production of value - added products fro m different combinations of this mixture have been addressed in this study. Later, the feed mixture with the highest isoamyl products selectivity is employed in a continuous reactor to confirm the batch studies and find an ideal reactor configuration for the system. Finally, kinetic modeling has been developed to c ompare th e rate of formation of the main products in the system. 4.2 Materials and Methods 4.2.1 Materials and Catalyst Preparation The materials used and the catalyst preparation method in this chapter are the same as those described in Section 2.2.1. Isoamyl alcoh ol (>98%, Sigma - Aldrich) was mixed with anhydrous ethanol ( Koptec , 200 proof) in the desired ratios as the feed of the experiments. Moreover, the 102 catalyst known as Ni (IV) in Chapter 2, with the composition of 8.0 wt% Ni/9.0 wt% La 2 O 3 / - Al 2 O 3 was used for both batc h and continuous experiments. 4.2.2 Reactor System 4.2.2.1 Batch Experiments Batch reactions were performed in a 300 ml Parr reactor (Model 4842, Parr Instruments, Chicago, Illinois) with reaction times between 22 and 51 hours. Typically, 120 g of the feed mixt ure along with the desired amount of catalyst were placed into the reactor. The reactor was purged with nitrogen and sealed with 1 atm of nitrogen overpressure. The reactions were carried out at autogenous pressure. The Parr reactor contained an Omega 1/8 ss steel Type J thermocouple which was not calibrated for this study; nevertheless, previous works had shown its accuracy in the range of ±1 ºC [ 9 , 25 ] . Pressure measurements for all experiments were done using an electronic pressure transducer (200 atm) that was calibrated against a 100 atm mechanical gauge wit h increments of 0.7 atm. The mechanical stirrer was set at 1000 rpm during the reaction. An initial liquid sample was taken after purging nitrogen to the reactor and before heating the reactor to the reaction temperature. Usually, a second sample of the li quid phase was taken after 1 - 2 h via a dip tube int end to isolate the liquid sample from the reactor. The sample tube was vented after isolating, and the liquid sample was analyzed b y gas chromatography. The reactor pressure was moni tored during the reaction and after cooling the reactor at the end of reaction to help with determining product compositions and quantities of gas formed. The quantity of gaseous products in the reaction was determined at the end of each experiment by weig hing the entire cooled reactor with 103 chemicals both before and after depressuri zation. The gas exhausted during depressurization was collected in a gas bag and analyzed by gas chromatography. 4.2.2.2 Continuous Experiments Reactor setup and experimental steps in th e continuous experiments are the same as ethanol Guerbet experiments discussed in Section 2.2.2. Experiments were done using 29.9 g of catalyst, and for the integrity of the results, the feed had the composition of 80 mol% ethanol and 20 mol% isoamyl alcoh ol. Experimental temperature was changed from 170 o C to 250 o C, and the liquid fe ed flow rate was varied from 0.3 ml/min to 1.3 ml/min, corresponding to the WHSV of 0.5 h - 1 to 2.1 h - 1 . Products were collected in two different traps; an ice/water trap for c ondensing liquid components at room temperature, and a gas bag for gaseous pro ducts. These samples were analyzed using different gas chromatography instruments discussed in the next section. 4.2.3 Analytical Methods The analytical methods and instruments used in this chapter are the same as those described in Section 2.2.3. 4.3 Experimental Results 4.3.1 Batch Experiments Guerbet reactions with two alcohols lead to a significantly wider variety of product species than for a single alcohol. The key products of the mixed iso amyl alcohol (IA) and ethanol experiments are shown on the right side of Figure 4 . 1 ; the alcohols responsible for forming the products are shown on the left side of this figure. Results obtained from batch reaction studies are su m marized in Table 4 . 1 . There are two sets of results presented: Reactions B2 - B6 were run for approximately 24 h, and reaction B7 was run for 104 51 h. Both sets were studi ed at 230 o C on the scale of 120 g of feed. Each reaction had a different composition of alcohols, ranging from 100% ethanol to 100% isoamyl alcohol. The results in Table 4 . 1 show that ethanol selectivity to n - butanol and C 6 alcohols declin e as IA concentration increases until a majority of ethanol is reacting with isoamyl alcohol (B4) instead of with itself. Figure 4 . 1 . Primary products observed from reaction of ethanol and isoamyl alcohol mixtures. 105 Table 4 . 1 . Results of batch reactor experiments with 120 g of (ethanol - isoamyl alcohol) mixture at 230 o C,4.85 g of 8.0 wt% Ni/9.0 wt% La 2 O 3 - Al 2 O 3 is used as the catalyst. Exp. Initial Molar Ratio EtOH/ IAOH Reaction time (h) Conv. (%) Selectivity (%) EtOH IAOH EtOH Prod. EtOH + IA Cross Prod. IAOH Prod. w.r.t. EtOH w.r.t. IA OH C 4 C 6 C 8 C 7 C 9 C 7 C 9 C 10 B2 1/0 22 21.1 - 67.5 22.0 5.0 0.0 0.0 - - - B3 3.8/1 23 30.1 13.7 50.8 15.5 3.3 6.0 0.0 50.1 6.8 0.0 B5 1/1 22 35.2 11.4 39.0 9.6 3.0 16.8 0.3 52.1 5.6 0.9 B4 1/4 24 43.3 9.7 25.5 3.2 8.6 39.5 2.9 43.8 2.9 3.2 B6 0/1 24 - 11.1 - - - - - 0.0 0.0 4.8 B7 3.8/1 51 40.5 14.6 47.9 15.9 3.6 5.9 0.1 62.9 8.7 0.6 From the two reactions with pure alcohols (B2 and B6), the rate of ethanol conversion is approximately twice that of isoamyl alcohol. A detailed kinetic model for the combined ethanol/isoamyl alcohol reaction sys te m is developed in Section 4.3.3 . 4.3.2 Continuous Experiments Reactions were also carried out in the continuous condensed - phase reactor over the standard 8 wt% Ni/9 wt% La 2 O 3 /Al 2 O 3 catalyst with a reactor feed mixture of 80 mol% ethanol and 20 mol% isoamyl alcohol. Results of these experiments are given in Table 4 . 2 . The results show that isoamyl alcohol reacts with ethanol and with itself to form a variety of straight - chain and branched - chain alcohols. Recoverie s for both batch a nd continuous reactions are reasonable (70 - 90%) for these preliminary experiments, especially given that gases contribute another 5 - 10% of total alcohols converted. Nonetheless, unlike alcohol products, the source of gaseous products (eth anol or IA) is ind istinguishable. 106 Table 4 . 2 . Results of continuous condensed - phase experiments with 4/1 molar ratio of ethanol/isoamyl alcohol and 29.9 g of 8.0 wt% Ni/9.0 wt% La 2 O 3 / - Al 2 O 3 catalyst. 1 Conversion is too low; t hus, this set of data are unreliable. Conditions Conversion (%) Selectivity (%) EtOH Products EtOH + IAOH Cross Products IAOH Products w.r.t. EtOH w.r.t. IAOH T ( o C) WHSV (h - 1 ) EtOH IAOH C 4 C 6 C 8 C 7 C 9 C 7 C 9 C 10 210 0.8 10.1 3.5 67.0 11.8 1.5 5.2 0.9 62.2 5.6 0.0 1.4 7.8 2.1 66.6 8.9 0.8 4.5 0.7 69.0 5.1 0.0 2.1 5.8 0.8 1 62.4 6.4 0.3 3.8 0.5 144.8 9.2 0.0 230 0.8 2 3.4 9.4 58.1 14.4 2.6 6.3 1.4 62.2 7.0 0.3 1.4 17.1 5.3 59.8 12.3 1.9 5.9 1.1 76.4 7. 4 0.2 2.1 12.7 6.8 67.6 12.5 1.7 5.7 1.1 42.3 4.1 0.0 250 0.8 4 2.4 11.9 47.9 13.7 2.8 5.5 1.4 78.1 9.8 0.5 1.4 38.1 13.7 45.7 14.3 3.2 5.6 1.5 61.5 8.4 0.4 2.1 31.2 11.0 47.4 13.0 2.6 5.2 1.3 59.4 7.6 0.3 4.3.3 Kinetic Model Development A kinetic model has been developed for the ethanol/isoamyl reaction mixture. The following key reactions are considered for this modeling: C 2 H 5 OH + C 2 H 5 OH C 4 H 9 OH + H 2 O Reaction 4 . 1 C 2 H 5 OH + C 5 H 11 OH C 7 H 15 OH + H 2 O Reaction 4 . 2 C 5 H 11 OH + C 5 H 11 OH C 10 H 21 OH + H 2 O Reaction 4 . 3 For simplicity, C 2 H 5 OH, C 5 H 11 OH, and C 10 H 21 OH are show n as E, I, and C10 in the equations, respectively. Initially , the continuous system was modeled for the kinetic studies since it provided experimental results at three different superficial residence times ( ). The following rate equations were developed f or the continuous system: 107 Equation 4 . 1 Equation 4 . 2 Equation 4 . 3 where r E , r I , and r C10 are in the unit of mol/kg of catalyst/h, and k 1 , k 2 , and k 3 are in the unit of m 6 of solution/kg of catalyst/mol/h. Euler method of integration with the step size of one minute was used for numerically integrating the above equations: Equation 4 . 4 Equation 4 . 5 Equation 4 . 6 Mo deling was performed for three temperatures (210 o C, 230 o C, and 250 o C) and the contact times of as high as 70 minutes, based on the experimental data available for the continuous system. The experimental data were calculated based on the selectivities re ported in Table 4 . 2 . For each one of the components modeled, the following equations were us ed for calculating the experimental concentrations at different contact times: Equation 4 . 7 Equation 4 . 8 Equation 4 . 9 108 where , , and are the initial concentrations of each of the components in mol/L of solution. Moreover, S B , S C 7E , S C7I , and S C10 ar e the selectivities toward butanol, C 7 alcohol products (with respect to ethanol), C 7 alcohol products (with respect to isoamyl alcohol), and C 10 alcohol product (i.e. 2 - iso propyl - 5 - methyl - hexanol). In the selectivity calculations, the share of secondary p roducts (such as butanol products) has also been considered and added to the selectivity toward primary products; these calculations are shown in detail in Appendix H. Figure 4 . 2 - Figure 4 . 4 show the comparison of the modeling results with experimental data at each temperature for the (4/1) molar ratio of ethanol/isoamyl alcohol mixture. The modeling has not been performed for 2 - iso propyl - 5 - methyl - hexanol (C10) at 210 o C since this chemical was not detected at th at temperature. The results indicate a good fit between modeling and experimental data at different temperatures for each of the three species modeled. Based on the modeling results, rate constants were developed for the three reactions at each temperature . For the three temperatures, the ratio of the rate constants for Reaction 4 . 1 and Reaction 4 . 2 ( k 1 / k 2 ) was between 1.4 and 1.6; this ratio for Reaction 4 . 3 and Reaction 4 . 2 ( k 3 / k 2 ) was between 0 .007 and 0.010. This shows that ethanol - ethanol Guerbet reaction is the fastest reaction among the three, and isoamyl - isoamyl condensation is the slowest, which is explained by the inductive effect; the electron - donating alky l groups surrounding the OH group of the isoamyl al cohol destabilize the negatively charged oxygen and prevent the dehydrogenation reaction from happening. Based on the rate constants obtained, an activation energy was calculated for each one of the react ions. The activation energies for ethanol - ethanol ( Reaction 4 . 1 ), ethanol - isoamyl alcohol ( Reaction 4 . 2 ), and isoamyl alcohol - isoamyl alcoho l ( Reaction 4 . 3 ) reactions were found to be 80 kJ/mol, 89 kJ/mol, and 110 kJ/mol, respectively. The E a obtained for Reaction 4 . 1 is comparable to the global activation energy of 121 kJ/mol calculated based on ethanol conversion in Chapter 2 109 using Ni(I) (8 wt% Ni/4.5 wt% La 2 O 3 ) ca talyst, considering that these two catalysts have different lanthanum compositions. Figure 4 . 2 . Comparison of simulated and experimental reactor outlet concentrations for the ethanol/isoamyl alcohol c ontinuous Guerbet experiments at T= 210 o C using 8.0 wt% Ni/9.0 wt% La 2 O 3 / - Al 2 O 3 as the catalyst. (a)Ethanol, (b)Isoamyl alcohol. 10.4 10.8 11.2 11.6 12.0 0 25 50 75 C E (mol/L) (min) Mod. Exp. 2.84 2.86 2.88 2.90 2.92 2.94 0 25 50 75 C I (mol/L) (min) Mod. Exp. 9.0 9.5 10.0 10.5 11.0 11.5 12.0 0 25 50 75 C E (mol/L) (min) Mod. Exp. (a) 2.70 2.75 2.80 2.85 2.90 2.95 0 25 50 75 C I (mol/L) (min) Mod. Exp. (a) (b) (b) 110 Figure 4 . 3 . Comparison of simulated and experimental reactor outle t concentrations for the ethanol/isoamyl alcohol continuous Guerbet experiments at T= 230 o C using 8.0 wt% Ni/9.0 wt% La 2 O 3 / - Al 2 O 3 as the cat alyst. (a)Ethanol, (b)Isoamyl alcohol, (c)2 - Isopropyl - 5 - methyl - hexanol. Figure 4 . 4 . Comparison of simulated and experimental reactor outlet concentrations for the ethanol/isoamyl alcohol continuous Guerbet experiments at T= 250 o C using 8.0 wt% Ni/9.0 wt% La 2 O 3 / - Al 2 O 3 as the catalyst. (a)Ethanol, (b)Isoamyl alcohol, (c)2 - Isopropyl - 5 - methyl - hexanol. 0 0.0001 0.0002 0.0003 0.0004 0.0005 0 25 50 75 C C10 (mol/L) (min) Mod. Exp. (c) 8.0 9.0 10.0 11.0 12.0 0 25 50 75 C E (mol/L) (min) Mod. Exp. 2.55 2.65 2.75 2.85 2.95 0 25 50 75 C I (mol/L) (min) Mod. Exp. 0.0000 0.0003 0.0006 0.0009 0.0012 0 25 50 75 C C10 (mol/L) (min) Mod. Exp. (a) (b) (c) 111 Similar calculations were made for the batch system results, except that the following equations were used in the modeling process: Reaction 4 . 4 Reaction 4 . 5 Reaction 4 . 6 For the batch system, there is only one data point for each one of the feed compositions. Therefore, the rate constants determined could be less accurate because there are fewer experimental data to confirm the modeling results. Table 4 . 3 shows the rate constants obtained for each one of the batch experiments and their comparison with the continuous system results obtained at the same reaction temperature. It can be observed that the va lues of k 1 , k 2 , and k 3 are in the same order of magnitude for each batch experiment and the continuous experiment. Furthermore, the ratios of rate constants are in the same range as those mentioned earlier (1.3 - 1.9 for k 1 /k 2 , and 0.005 - 0.013 for k 3 /k 2 ). Th is indicates that even if reaction conditions (such as temperature integrity, catalytic activity, etc.) affect the rate constants, each reaction proceeds at the same proportional rate compared to the others. Table 4 . 3 . Rate Constan ts developed for different batch experiments and their comparison with the one obtained for the continuous system at 230 o C. Experiment k 1 k 2 k 3 k 1 /k 2 k 3 /k 2 (m 6 solution/kg catalyst/mol/h) B2 9.3E - 06 - - - - B3 1.3E - 05 9.7E - 06 4. 7E - 08 1.3E+00 4.8E - 03 B5 2.5E - 05 1.9E - 05 1.3E - 07 1.4E+00 6.8E - 03 112 B4 6.8E - 05 3.6E - 05 2.6E - 07 1.9E+00 7.2E - 03 B6 - - 3.8E - 07 - - B7 8.9E - 06 6.8E - 06 9.0E - 08 1.3E+00 1.3E - 02 Continuous System 9.2E - 06 6.6E - 06 4.7E - 08 1.4E+00 7.1E - 03 4.4 Conclusions Guerbet r eactions of ethanol - isoamyl alcohol mixtures were conducted using the 8.0 wt% Ni/9.0 wt% La 2 O 3 / - Al 2 O 3 catalyst in both batch and continuous systems. While ethanol selectivity toward C 4 + alcohols stayed as high as 72% at 42% conversion, isoamyl alcohols se lectivity of 88% toward higher alcohols (mainly cross - condensation products with ethanol) at 12% conversion was achieved . Kinetic modeling for three primary Guerbet reactions (C 2 - C 2 , C 2 - C 5 , and C 5 - C 5 ) for both continuous and batch systems provided consiste nt results with respect to the experimental data and different feed compositions. A fixed ratio of the rate constants at different conditions were also obtained . Finally, activation ene rgies determined from the rate constants at different temperatures indi cated that the ethanol - ethanol and isoamyl alcohol - isoamyl alcohol reactions possess the smallest and largest barriers for activating the molecules, respectively. 113 APPENDI CES 114 APPENDIX H. Detailed Selectivity Calculations The formulas used for the calculat ion of the total selectivity of different chemicals in Equation 4 . 7 and Equation 4 . 9 are as follows: Equation 4 . 10 Equation 4 . 11 Equation 4 . 12 Equation 4 . 13 where: CS B = Calculated selectivity for butanol, CS C6 = Calculated selectivity for C 6 alcohols, CS C2C6 = Calculated selectivity of C 8 alcohols that are the result of the reaction of an ethanol and a C 6 alcohol, CS C4C4 = Calculated selectivity of C 8 alcohols that are the result of the reaction of two C 4 alcohols. CS C7E = Calculated selectivity of C 7 alcohols with respect to ethanol, CS C9E = Calculated selectivity of C 9 alcohols with respect to ethanol, CS C7I = Calculated selectivity of C 7 alcohols with respect to i soamyl alcohol, CS C9I = Calculated selectivity of C 9 alcohols with re spect to isoamyl alcohol, CS C10 = Calculated selectivity of C 10 alcohols. 115 All the calculated selectivities are based on the analytical results obtained from GC. 5 Acrylate Production from 2 Acetoxypropanoic Acid Esters 5.1 Introduction Acrylic acid ( 2 - propenoic acid , C 3 H 4 O 2 ) is the simplest alkenoic acid . It can react with itself or other monomers to form polymers that have extensive application as adhesives, polishes, binders, coatings, paints, detergents, fibers, polyelectrolytes, flocculants, diapers, and disp ersants. Based on a 2015 report, the acrylic acid market size was 5.8 million metric tons with demand growing at 6.3% per year and prices ranging from $1600 - $220 0 per ton depending on its grade [ 132 , 133 ] . Acrylic acid is traditionally produced by a petroleum - based two - step gas - phase process ( Figure 5 . 1 ) which involves the catalytic oxidation of propylene to acrolein, foll owed by the reaction of acrolein with oxygen in the presence of a catalyst [ 134 ] . Acrolein yields of 83 - 90% and ac rylic acid yield of 5 - 10% are obtained in the first step, and a maximum 97.5% yield of acrylic acid is reported for the second step. Recently, renewable biobased routes to acrylic acid have rece ived attention, both to address resource and climate issues and to ensure a stable, inexpensive supply in li ght of volatile petroleum markets. A few of those routes have garnered the most attention to replace the existing propylene - based process; these include dehydration of glycerol to acrolein, d irect dehydratio n of lactic acid Figure 5 . 1 . Petroleum - based route to acrylic acid from propylene 116 (2 - hydroxypropanoic acid) or 3 - h ydroxypropanoic acid, and pyrolysis of acetoxy isopropionic acids esters or salts. 5.1.1 Glycerol Dehydration In 1933, Schwenk introduced a method for hydrolyzing glycerol to acrolein in the vapor phase with 80% acrolein yield [ 135 ] . Later, s everal catalysts were proposed for increasing the acrolein yield for this reaction in the gas phase, liquid phase, or in subcritical or supercritical water [ 136 - 140 ] . Among those, Fe 3 PO 4 had the best results with 92% acrolein yield [ 139 ] . Despite the high yields obtain ed, catalyst de activation and by - product formation are challenges for acrolein production via this route. According to a recent review article [ 132 ] , there have been attempts to directly produce acrylic acid from glycerol in the presence of different catalysts, with the highest yield of 75% obtained so far. However, in both routes, catalyst deactivation and sustaining selectivity are the primary challen ges to commercialization . 5.1.2 H ydroxypropanoic acid Direct Dehydration Acrylic acid is also produced by the direct dehydration of lactic acid (2 - hydroxypropanoic acid) or 3 - hydroxypropanoic acid. H olmen first introduced direct lactic acid dehydration by examin ing several lactate substrates (free lactic acid, ammonium lactate , alkyl lactates) and achieving 68% acrylic acid yield with free lactic acid at 400 o C with Na 2 SO 4 and CaSO 4 as the catalyst [ 141 ] . Subsequent studies were primarily focused on phosphate [ 142 - 149 ] and sulfate [ 1 50 - 152 ] catalysts, but none of those gave acrylic acid yields higher than those obtained by Holmen. Naito and Abe from the same research group used molecular sieve 13X ion - exchanged with cesium and ruthenium and untreated molecular sieve 13X for the dire ct dehydration of methyl lactate [ 147 , 153 ] . An unprecedented methyl acrylate yield of 92 - 93 % , which was claimed to be stable over 117 minimum operation time of 40 h, w as reported in these studies. Lat er, additional attempts were made to modify the structure of molecular sieves with sodium, potassium, alkali phosphates, and lanthanum; but none achieved yields of above 66% [ 148 , 149 , 154 - 160 ] . Several of these studies show ed that diluting the reactant with some material such as met hanol [ 148 , 153 ] or water [ 156 , 158 - 161 ] increase s the reactant conversion and acrylate ester / acrylic acid selectivity . A recent review [ 162 ] gives an excellent summary of lactic acid and lactate ester dehydration pathways. Besides lactic acid, 3 - hydroxypropionic acid is another substrate which can be directly dehydrated to acrylic acid. Studies on this reaction have shown an acrylic acid yield of 88% using solid acid catalysts [ 163 - 165 ] . In 2015, Cargill Corporation started an investment on the commercial development of acrylic ac id production through acquiring OPX Biotechnologies, which has proprietary technology for producing acry lic acid from 3 - hydroxypropionic acid [ 166 ] . 5.1.3 2 - Acetoxypropanoi c A cid Indirect Dehydration L act ic acid, lact ate esters, or lactate salts can react with acetic anhydride or acetic acid to produce 2 - acetoxypropanoi c acid (APA), its esters, or its salts in high yields. T h e produced APA or its derivatives can lose an acetic acid molecule to produce the corresponding acrylate via a high - temperat ure pyrolysis reaction . I n 1935 , Burns et al. studied the reaction of methyl and butyl lactate esters with acetic anhydride ( Figure 5 . 2 ) to produc e alkyl APA esters, w hich were then pyrolyzed at 450 o C with quartz chips as the contact material to give methyl acrylate and butyl acrylate yield s of 76% and 15 - 25%, respectively [ 167 ] . The same group optimized the operational conditions for this catalyst and obtained 89% acrylate yields at 550 o C and WHSV of 0.6 h - 1 using pyrex and quartz as the contact material [ 168 - 170 ] . Godlewsk i et al. [ 171 ] used grounded fused quartz as the packing material for the methyl lactate elimination reaction at 560 o C and were able to obtain 90% 118 yield for methyl acrylate . Unfortunately, in all of these st udies, the yield invariably declines with time on stream , posing a barrier for the practical application of this p athway. Several studies show that low - cost acetic acid or acetate esters can be used instead of acetic anhydride in the formati on of the AP A species. Rehberg et al. [ 172 ] was the first one that obtained 28% methyl APA yield through the reaction of lactic acid and m ethyl acetate ( Figure 5 . 3 ). Studies in recent years were focusing on the production of APA from lactic acid and acetic acid. Among th em, APA yields of more than 90% were reported using homogeneous acid catalysts and solid acid cat alysts such as ion exchange resin, zeolites, Amberlyst 70, Nafion, and sulfonated graphene [ 173 , 174 ] . Besides lactic acid, dilactide is another substrate that can react with acetic acid to produce APA [ 175 ] . Recently, a new reactive distillation method has been developed to produce APA from lactic acid and acetic acid [ 176 ] . During this process, water is removed from the column as it is produced in order to drive the APA production reaction toward completion. Maximum APA yield of 95% is reported using this method. Besides the high yield obtained, other advantages of this method are Figure 5 . 2 . Methyl APA production using methyl lactate and acetic anhydride Figure 5 . 3 . Methyl APA production using lactic acid and methyl acetate 119 the absence of acidic homogeneous catalysts that facilitate the separation process, and the use of a continuous reactor that could use the recovered acetic acid that is liberated in the subsequent acrylate for mation step. Therefore, dev eloping a stable process for the production of acrylates from APA in high yields is an essential challenge in considering this pathway as a commercially viable one compared to the current methods used for acrylate production. In this work, we present condi tions that provide sustained high yields of acrylate from lactic acid - derived APA esters. 5.2 Materials and Methods 5.2.1 Materials and Catalyst Preparation Methyl (S) - - lactate (98%), ethyl (S) - - lactate (98%), butyl (S) - - lactate (97%), benzyl (S) - - lact ate (90%), isobutyl (R) - (+) - lactate (97%), and acetic acid (99.7%) were purchased from Sigma - Aldrich and use d as received. Methyl - 2 - acetoxypropa n o a t e (MAPA) (99%), ethyl - 2 - acetoxypropa n o at e (EAPA) (96%), butyl - 2 - acetoxy prop a n o ate (BAPA) (98%), benzyl - 2 - acetoxypropan oate (BeAPA ) (96%), and isobutyl - 2 - acetoxy prop ano ate (IBAPA) (99%) were produced by reaction of the corresponding lactate ester with exce ss acetic anhydride under acidic conditions at 25 °C. The acetoxy ester was recove red by mixing the reaction solution with diethyl ether and an aqueous solution of sodium bicarbonate in a separatory funnel. The ether phase, which contain the APA ester, was separated and then washed again with an aqueous solution of sodium chloride. Fina lly, the APA ester was isolated and purified by centrifuging and distilling off the ether under vacuum. Several materials were evaluated as contact materials for the process . Nonporous granular quartz (SiO 2 , Sigma - Aldrich) as the primary material was grou nd and sieved to 30 - 50 mesh size. 120 Additional contact materials included the porous silica Spherosil (XOA 400) and silicon carbide (SiC). Modified 13X molecular sieve was prep ared as a pyrolysis catalyst by soaking the 13X molecular sieves overnight in 38 w t% cesium acetate (Sigma - Aldrich) in water solution, drying for 24 hours at 100 o C, and calcining at 400 o C for 5 hours. The sieves were then soaked overnight in 0.5 wt% ruth enium chloride (Sigma - Aldrich) in ethanol s olution, dried for 24 hours at 100 o C, and finally calcined at reaction temperature for 5 hours before use in reaction. Additional catalyst materials evaluated in this work include porous SiO 2 treated with CsOH or KOH, CeZrO x - Al 2 O 3 , MSU - F structured pore zeolite, Zeolite - - H, and Zeolite - Y - H. All materials were used as obtained from vendors or other laboratories. 5.2.2 Reactor Configuration Reactions were performed in a 0.5 in. ID × 40 cm long quartz tube reactor pac ked with con tact mate temperature. A schematic of the reactor system is given in Figure 5 . 4 . Quartz (SiO 2 ) was chosen for the reactor material because it has been shown [ 167 , 169 ] to minimize undesired reactant a nd product d ecomposition relative to other reactor materials in APA conversion to acrylates. The reactor was sealed on each end with 316 SS O - ring fittings to facilitate reactant input and product collection and p laced in a tube furnace. Liquid reactants w ere fed to t he top of the reactor through Initially, diluent gas was fed through the annulus surrounding the liquid feed tu be; however, it was found that better vapor ization of t he liquid feed took place and higher acrylate yields were obtained therefore used in all experiments repo rted in this work. 121 Products were collect ed in two traps in series: one trap immersed in ice and a second immersed in dry ice/acetone. Gas that passed through both traps was sampled periodically during reaction using gas bags. Experiments were initiated by loading 1.0 g of c ontact ma terial suppo rted on quartz wool into the reactor. The reactor was assembled in the tube furnace, and diluent gas was passed through the liquid feed was me tered int o the reacto All experiments were conducted at atmospheric pressure. FURNACE Liquid f eed from pump Carrier gas flow from mass flow controller Contact material bed Thermowell Downward feed flow Vent for non condensibles Collector in dry ice/acetone bath Collector in ice bath Figure 5 . 4 . Schematic of the reactor system 122 Product samples were collected from the ice trap (0 °C) after the first 60 min of reaction and then at 90 min intervals. Product collected in the dry ice trap was analyzed only at the conclusion of the experiment. Liquid and gas samples were analyzed by gas chromatography; details of the analytical methods and yield calculations are given in the Supporting Information. 5.2.3 Anal ytical Me thods Liquid samples from all experiments except those with ethyl APA as the feed material were diluted 10 - fold in acetonitrile and analyzed using a Varian 450 gas chromatograph with flame ionization detector. A 30 m Sol Gel Wax column (0.25 mm ID film thickness) was used with the following temperature p rogram: initial temperature 37 o C for 4 min; ramp at 10 o C /min to 90 o C , and hold at 90 o C for 3 min; ramp at 10 o C /min to 150 o C ; ramp at 30 o C /min to 230 o C, and hold for 2 min. Liquid sa mples from ethyl APA experiments were diluted 20 - fold in acetonitrile and analyzed using a Perkin Elmer Autosys GC with a thermal conductivity detector. A 2 m long Chromosor b packed column (2 mm ID) was u sed with the following temp erature program: initial temp erature 130 o C for 1 min; ramp at 10 o C /min to 250 o C, and hold at 250 o C for 5 min. Gas samples w ere analyzed using the same ins trument (Perkin Elmer Autosys GC) and column, but with the following temperature program: initial temperature of 130 o C for 1 min; ramp at 30 o C /min to 250 o C , and hold at 250 o C for 5 min. Gas samp les were also analyzed using a Varian 3300 GC - 1000 column (2.1 mm ID) w as used with the f ollowing temp erature prog ram: initial temperature 35 o C for 5 min, then ramp at 20 o C/min to 225 o C . Species concentratio n from liquid and gas samples were determined using response factors obtained f rom the slope of multi - point ca libration curves made wi th reactant and product s pecies solu tions of known concentrations. Calculated conc entrations were entered into a Microsoft Excel 123 spre adsheet where reactant fraction al conversion, product yields and selectivities, overa ll carbon recovery, and recover y of mo lecular f ragments were ca lculated. Yield is defined as the number of moles of the desi red products per mole of limiting reactant fed, and s electivity is defin ed as the number of moles of de sired produc ts formed per mole of limiting reactant converted. Unless other wise noted, yield a nd selectivity re fer to total acrylate produced (acryla te ester + free acrylic acid). 5.2.4 Contact Material Characterization cs ASAP 2010 instrument was used to measure the BET surface area of contact mate rials. Materials were outgassed at 260 °C for 24 hours before adsorption measurements. Surface acidity and basicity were measured by temperature - programmed desorption (TPD) of NH 3 and CO 2 , respectively, in a Micromeritics Autochem 2910 instrument. For surf ace acidity, approximately 0.6 g of contact material was thermally pretreated at 800 °C in 50 cm 3 /min of He (99.999%) for 60 min, cooled to 25 °C in He, saturated at 25 °C with NH 3 by flowing a mixture of 14.8 vol% NH 3 in helium at 50 cm 3 /min for 1 hour, a nd then purged with He (50 cm 3 /min) for 2 hours to remove all physisorbed NH 3 . Desorption was carried out by heating in He from ambient temperature to 600 °C at 10 °C/min and h olding for 30 min. The resulting NH 3 peak area was quantified using an NH 3 gas s tandard of known composition. An identical procedure was used for measuring surface basic site density, except that pure CO 2 (99.8%) was used instead of the NH 3 /He blend. 5.3 Exper imental Results Figure 5 . 5 describes t he reaction pathways observed in experiments. Path (a) is acetic acid elimination to form acrylates or, in anoth er sense, elimination of APA as an alkene (acrylate) from acetic acid. Path (b) is elimination of the alkyl ester group as an alkene that can oc cur when there 124 - carbon of the ester functionality. Both eliminations are facilitated via the formation of a cyclic six - - carbon [ 177 ] . Complete product distributions and functional group balances (acetate (C 2 ), lactate (C 3 ), and ester text and figures, are given in Table 5 . 1 . 5.3.1 Control Experiments Two control experiments were conducted by feeding buty l acrylate and acetic acid at 550 °C with a liquid hourly space velocity (LHSV) of 1.9 kg solution/k g contact material/h to determine the stability of these chemicals under typical reaction conditions. The recovery of acetic acid from the reactor after 4 h ours of steady state operation was 94%, which is close to complete recovery within the uncertainty o f the experiment. Under the same conditions, butyl acrylate was 92% converted to acrylic acid with accompanying quantitative formation of 1 - butene. Figure 5 . 5 . Reaction pathways for APA ester elimination reactions. Path (a): elimination of acetic acid to form alkyl acrylate or acrylic acid; Path (b): elim ination of alkene for ester R groups containing hydrogen on - carbon; Path (c): decarbonylation of APA to acetaldehyde, acetic acid, and CO. 125 Table 5 . 1 . Summary of conditions, functional group balances, and product distributions for selected experiments R1 - R5. a Total carbon recovery does not include the carbon deposited on the contact material. Experiment R1 R2 R3 R4 R5 Reaction Conditions and APA Ester Conversion Feed Composition (wt%) APA Ester Acetic Acid 80 20 80 20 80 20 100 0 80 20 Ester Group Butyl Butyl Methyl Butyl Butyl Diluent Gas N 2 CO 2 CO 2 CO 2 CO 2 Temperature ( o C) 550 550 550 550 490 APA Ester Conversion (%) 100 100 88 97 46 Product Selectivities (mol product/mol APA ester converted) Alkyl Acrylate (%) 7.9 3.5 77.8 10.1 21.5 Acrylic Acid (%) 18.6 31.4 0 22.2 4.8 Acetic Acid (%) 41.6 60.7 71.0 75.5 26.1 Alkene (%) 79.3 94.8 0 97. 6 51.6 Acetaldehyde (%) 12.1 14.9 0 13.6 13.9 Carbon monoxide (%) 8.8 59.2 1.7 56.5 46.7 Alkyl alcohol (%) 0.8 0.5 0 0.6 1.6 Unknown (%) 1.3 2.0 0 1.4 1.7 Functional Group and Carbon Recoveries (mol group/mol APA ester fed) Methyl (C1) (%) - - 86 - - Acetate (C2) (%) 94 94 93 95 89 Lactate (C3) (%) 52 76 91 90 87 Butyl (C4) (%) 80 99 - 109 93 Total Carbon Recovery a (%) 76 90 91 98 90 5.3.2 Feed Composition 5.3.2.1 Diluent Liquid Adding acetic acid to butyl APA feed ( Figure 5 . 6 ) modestly improves acrylate selectivity to a provide acidity to aid elimination or may interact wi th the contact material surface to limit side 126 reactions, but ace tic acid most likely aids in dispersing and volatilizing the APA esters. Acetic acid concentrations above 50 mol% lead to reduced acrylate yields because the large excess of acetic acid may en hance unwanted decomposition reactions or thermodynamically limi t the extent of acetic acid elimination from APA. Subsequent experiments were carried out with 55 mol% butyl APA + 45 mol% acetic acid (80 wt% APA ester + 20 wt% acetic acid) as the liquid fee d. 5.3.2.2 Diluent Gas Carbon dioxide and nitrogen were examined as di luent gases with butyl APA and benzyl APA as feed materials over quartz contact material. Conversion of b oth APA esters is essentially complete under reaction conditions for both gases, but CO 2 increases acrylate selectivity relative to N 2 ( Figure 5 . 7 ). With butyl APA, the combined acrylate selectivity (butyl acrylate + acrylic acid) with CO 2 is approximately twice that with N 2 as steady state is approached (300 min on - stream). Acrylate selec tivity with N 2 declines with time on stream , whereas it remains essentially constant 0.25 0.27 0.29 0.31 0.33 0.35 0.37 0 0.1 0.2 0.3 0.4 0.5 0.6 Total Acrylate Selectivity Mole Fraction Acetic Acid in Feed Figure 5 . 6 . Effect of acetic acid feed concentration on steady sta te acrylate selectivity from butyl APA (R2, 45 mol% acetic acid; R4, 0% acetic acid). Reaction conditions: T = 550 °C; quartz (SiO 2 ) con tact material; CO 2 127 with CO 2 as diluent. The effect is much less dramatic with benzyl APA as the feed, but it is significant during the ramp - up to steady state. Carbon monoxide (CO) is present in the effluent gas stream with both N 2 and CO 2 as diluent gases. If CO was formed only via decarbonylation of the lactate backbone, then on a molar basis, CO formation should be at most equal to t hat of acetaldehyde (because decarboxylation of lactate can also occur). This is indeed the case for butyl APA conversion with N 2 (R1 in Table 5 . 1 ). I n contrast, CO formation with CO 2 as diluent is 8 - fold greater than that of acetaldehyde. Given that carbon is deposited on the contact material during the reac tion by partial decomposition of APA and other species, the large quantity of carbon monoxide produced can be explained by the Bo udouard reaction of CO 2 with this deposited carbon (CO 2 + C 2 CO). 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 50 100 150 200 250 300 350 400 Total Acrylate Selectivity Time On Stream (min) Benzyl APA, CO Bezyl APA, N Butyl APA, CO Butyl APA, N Butyl APA, N , Repeated Figure 5 . 7 . Total acrylate selectivity versus time on stream with N 2 or CO 2 as diluent gas (20 ml/min) for benzyl and n - butyl APA ester feed. Reaction conditions: T = 550 °C; 80 wt% APA ester/20 wt% acetic acid feed; LHSV = 1.9 kg fee d/ kg contact material/h. 128 The deposited carbon is likely amorphous and porous in nat reactants and products in the system, causing them to remain in the high - temperature envi ronment and further decompose. The presence of CO 2 as diluent gas continually removes these carbon deposits from the contact ma terial, thus ensuring a short residence time for reactants and products that minimizes decomposition and results in higher acryla te selectivity. To further support the presence of the Boudouard reaction, the contact material was removed from the reactor f ollowing experiments with N 2 and CO 2 as diluent gases, weighed, and heated in air to 500 °C. The weight loss upon heating was 3 t o 4 times higher for the experiment with N 2 as the diluent gas than with CO 2 as the diluent, indicating greater carbon depositi on with N 2 than with CO 2 . This result, along with increased pressure drop through the reactor and reduction in acrylate selectivi ty after 300 min on stream with the N 2 diluent, supports the hypothesis that CO 2 maintains the integrity of the contact materia l surface by removing deposited carbon during the reaction. Finally, the composition of the effluent gas suggests that the react ion of CO 2 with deposited carbon is rapid and even approaches equilibrium. With butyl APA feed and CO 2 as the diluent gas, the reactor effluent contains 10 mol% CO and 50 mol% CO 2 . These quantities of CO and CO 2 are close to the composition at P = 1 atm di ctated by the equilibrium constant for the Boudouard reaction at 550 °C (K p 0.02 atm = ) [ 178 ] . 5.3.3 Acrylate Yields from Different APA Esters Methyl APA, benzyl APA, isobutyl APA, butyl APA, and ethyl APA were fed over the quartz fixed bed at 550 °C with LHSV = 1.9 kg feed/kg contact material/h. Conversion of APA ester rang ed from 90% to nearly 100% for the different esters; total ac rylate selectivity is given in Figure 5 . 8 for each ester. 129 Figure 5 . 8 - carbon of the ester functionality (ethyl, isobutyl, bu tyl) exhibit signi ficantly lower selectivity to acrylates than those without hydrogen on - carbon (methyl, benzyl). Furthermore, selectivity to acryla - carbon hydrogen atoms increases, and the acrylate product distribution shifts from exclu sively acrylate ester for methyl and benzyl APA feeds to a majority of free acrylic acid for ester groups - carbon hydrogens. As shown in Figure 5 . 5 (Path ( b)), the ability of ester groups - carbon hydrogens to eliminate as alkenes from alkyl acrylate, leads to free APA or free acrylic acid formation. These components are significantly more r eactive toward undesired decomposition or polymerization than their ester counterparts. Free APA is not observed in the reactor effluent under any conditions in this study, even though nearly stoichiometric quantities of alkene relative to feed are formed in some experiments. Thus, free APA formed must quickly Figure 5 . 8 . Total acrylate selectivity vs time on stream for different APA ester feeds. Reaction conditions: T = 550 ° C; 80 wt% APA ester/20 wt% acetic acid feed; CO 2 diluent gas at 20 ml/min; LHSV = 1.9 kg feed/kg contact material/h. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 50 100 150 200 250 300 350 Yield Time On Stream (min) Methyl Benzyl Isobutyl n-Butyl Ethyl 130 d ecompose under the reaction conditions, partially to acrylic acid via the desired acetic acid elimination but also to undesired products via decarbonylation ((Path (c), Figure 5 . 5 ) and other decomposition reactions to form acetaldehyde, gases, free acetic acid, and carbon on the contact material surface. Acrylic acid likewise must undergo polymerization or decomposition, leading to reduced yields. The decomposition of f ree AP A and free acrylic acid in the reactor is further supported by low lactate (C3) group recovery for butyl APA reactions (R1 and R2 in Table 5 . 1 ) relative to methyl APA reactions (R3 in Table 5 . 1 ), partiall y because acetaldehyde (bp 20 °C) formed via decarbonylation escapes from collection traps and is lost in the gas effluent stream. 5.3.4 Reaction Temperature Figure 5 . 9 illustrates the dependence of butyl APA conversion, acrylic acid se lectivity, and butyl acrylate selectivity on reaction temperature from 420 °C to 690 °C over quartz contact material. Conversion of butyl APA increases from 10% at 420 °C to nearly 100% at 550 °C, indicating that Paths (a) and (b) in Figure 5 . 5 go to completion. Selectivity to acrylic acid increases and selectivity to butyl acrylate decreases above 500 °C, as butene elimination from butyl acrylate (Path (b) in Figure 5 . 5 ) becomes rapid. Total acrylate yiel d remains nearly constant at its maximum value between 550 °C and 620 °C, suggesting that acrylic acid is relatively stable over this temperature range and only conversio n from butyl acrylate to acrylic acid occurs. Acrylic acid decomposition predominates above 620 °C such that no acrylic acid exits the reactor at the highest temperature examined (690 °C). An estimate of the activation energy assuming butyl APA conversion as a simple first - order reaction gives a value of 132 kJ/mol (R 2 = 0.99). The Arrheniu s plot is provided as Figure I.1 in the Appendix I. 131 5.3.5 Differen t Contact Materials Experiments were performed at 550 °C with 1.0 g of Spherosil (XOA 400) porous silica and silicon carbide (SiC) as contact materials in the quartz tube reactor. Table 5 . 2 compares acrylate selectivity from butyl APA with total surface area, surface acid site density, and surface basic site density of each of the contact materials. At 550 °C, nearly complete bu tyl APA conversion was observed for all thr ee materials, and the highest acrylate selectivity was obtained for quartz and SiC. The porous silica contact material does not perform well, most likely because any feed material entering the pores is trapped and decomposes to carbon and gases. In contras t, SiC and quartz are nonporous and do not trap reactants or products; instead, they provide a short contact time heat transfer surface area for volatilization and reaction of APA esters. Silicon carbide and quart z have such low surface areas that their su rface acidity and basicity were not measurable. Figure 5 . 9 . Steady state butyl APA conversion, butyl acrylate selectivity, and acrylic acid selectivity at steady state versus temperature (R5, 490 °C; R2, 550 °C). Reaction conditions: 80 wt% butyl APA/20 wt% acetic acid feed; CO 2 diluent gas at 20 ml/min; LHSV = 1.9 kg feed/kg contact material/h. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 400 450 500 550 600 650 700 Molar selectivity/ Conversion Butyl APA Conversion Butyl Acrylate Selectivity Acrylic Acid Selectivity Total Acrylate Selectivity 132 Table 5 . 2 . Properties and Experimental Results for Contact Materials. Reaction conditions: T = 550°C; 80 wt% butyl APA /20 wt% acetic acid feed at 2.4 mL/h; CO 2 d iluent gas at 20 ml /min; LHSV = 1.9 kg feed/kg contact material/h. Catalyst Surface Area (m 2 /g) Surface Acidity (µmol/g) Surface Basicity (µmol/g) Butyl APA Conversion % Acrylate Selectivity % SiO 2 (35 - 50mesh) 0.08 <10 <10 97.4 36.5 SiC (50 mesh) 0.03 <1 0 <10 99.5 29.4 - Al 2 O 3 161 495 75 96.7 24.8 13X Molecular Sieve 175 152 176 99.0 9.0 Spherosil (XOA400) 375 137 45 100 5.1 One additional set of experiments was carried out with the modified 13X m olecular sieve , reported by Mitsubishi Gas Chemical Co mpany research gr oup [ 147 , 153 ] . Although high acrylate yields - hydroxy isobutyrate, - methoxy isobutyr ate, and methyl lactate at temperatures of 300 - 350 o C was reported in those studies, no significant acrylate formation was observed from butyl APA using this catalyst over the range of reaction conditions. To further evaluate this catalyst , the reaction wa s re peated with methyl lactate at the exact reaction conditions as those studies [ 153 ] for methyl acrylate production. While a small quantity of methyl acrylate was observe d (<5%), the yiel d ob tained was significantly lower compared to the reported values. As results confirm, lower surface acidity and basicity lead to higher selectivity toward acrylates. Wang et al. discussed that the ideal surface acidity/basicity ratio for this reaction is con tact material dependent [ 179 ] ; nevertheless, it seems that increasing acidic and basic sites improves the selectivity of side reactio ns such as acid - c atal yzed decarbonylation and decarboxylation and base - catalyzed condensation reactions rather than the APA elimination reaction. Thus, contact materials with low concentrations of acid and base sites are the best choices for this reaction. 133 5.3.6 Space Velocity B utyl APA conversion and acrylate selectivity were measured at 550 °C with different quantities of quartz (0.3 g, 1 g, and 3g) in the reactor that correspond to LHSV of 6.3, 1.9, and 0.65 kg feed/kg contact material/h. Nearly complete conve rsion of butyl AP A wa s observed over this range of LHSV along with similar acrylate selectivity, an indication that even higher space velocities should be suitable for APA ester conversion to acrylates at 550 °C. 5.3.7 Extended Reaction To demonstrate the improv ed stability of a cryl ate formation with CO 2 as diluent gas and inclusion of acetic acid as a co - feed, an extended time experiment was carried out for 30 h. The results ( Figure 5 . 10 ) show that butyl APA conversion (98%) and overall acrylate selecti were stable over the 30 h period of operation. The run was terminated at 30 h because of depletion of the feed material. Following the reaction, the quartz contact material was removed from the reactor and was found to have a quantity of carb on deposited that was essentially the same as that gment balances were excellent for this extended experiment. 0 10 20 30 40 50 60 70 80 90 100 0 5 10 15 20 25 30 Selectivity/Yield (%) Time on Stream (h) Conversion Butyl Acrylate Yield Acrylic Acid Yield Total Yield Figure 5 . 10 . Extended time experiment. Reaction conditions: T= 550 o C; 80 wt% butyl APA/ 20 wt% acetic acid feed; CO 2 diluent gas at 20 ml/min; LHSV = 1. 9/ kg feed/kg contact material/h. 134 5.4 Conclusions Feed composition and reaction conditio ns for enhanced a crylate production from APA esters have been identified. Acrylate selectivity of 35% was achieved from butyl APA at 55 0 °C and contact material , 20 wt% acetic a cid in the liquid feed, and CO 2 as the diluent gas. Experiments with methyl or benzyl APA ester feeds resulted in acry late selectivity of 70+% under the same conditions and are thus clearly preferred feed materials. Higher alkyl APA esters have hydrogen at - carbon of their ester functionality that allows elimination of the ester group as an alkene, liberating f ree acrylic acid and APA that decompose or polymerize to reduce acrylate yield. Using CO 2 as diluent gas increases acryla te formation from butyl APA and reduces the quantity of deposited carbon on the contact material, thus maintaining reaction rate over e xtended operation. 135 APPENDI CES 136 APPENDIX I. Rate constant and activation energy of APA conversion The rate constant acti vation energy for APA conversion is estimated by assuming that APA reacts via a simple first - order reaction as shown below. A first - order rate constant was determined at each temperature shown in Figure 5 . 9 ; the rate constant is p lotted in Figure I .1 to obtain the activation energy and pre - exponential factor. The value of the acti vation energy obtained from the Arrhenius plot is 131.5 kJ/mol; the pre - exponential is 3.9 × 10 4 (s - 1 ). k = 39566e - 15822/T R² = 0.9986 0.000001 0.00001 0.0001 0.001 0.01 0.1 1 0.0012 0.00125 0.0013 0.00135 0.0014 0.00145 0.0015 k (s - 1 ) 1/T (1/K) Figure I.1 . Arrhenius plot of first order rate constant for APA decomposition. 137 APPENDIX J. Temperature programmed Desorption (TPD) Profiles I.1. NH 3 TPD (Acidic Sites) (a) 0 100 200 300 400 500 600 700 0 10 20 30 40 50 60 70 80 90 100 Temperature ( o C) Time (min) Temperature vs. Time -2 0 2 4 6 8 10 0 100 200 300 400 500 600 Concentration (cm 3 /min/g) Temperature ( o C) NH 3 TPD Blank Spherosil Gamma Alumina Molecular Sieve (b) Figure J. 1 . Acidic sites measurement for differe nt materials studied. (a) Temperature profile, (b) NH 3 Temperature Programmed Desorption (TPD) diagram. (a) 138 J.2. CO 2 TPD (Basic Sites) 0 100 200 300 400 500 600 700 0 10 20 30 40 50 60 70 80 90 100 Temperature ( o C) Time (min) Temperature vs. Time -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 100 200 300 400 500 600 Concentration (cm 3 /min/g) Temperature ( o C) CO 2 TPD Blank Spherosil Gamma Alumina Molecular Sieve (a) (b) Figure J . 2 . Basic sites measurement for different materials studied. (a) Temperature profile, (b) CO 2 Temperature Programmed Desorption (TPD) diagram. 139 6 Summary and Recommendati ons for Future Wo rk 6.1 Summary The main objective of this work is to make highe r alcohols, specifically butanol, from efficiently produced fermentation - based ethanol through a process that is known as the Guerbet chemistry. The work done in this study achieve d 74% C 4 + alcohol s selectivity at 41% ethanol conversion for the Ni/ La 2 O 3 / - Al 2 O 3 catalyst. Attempts have been made to improve this result by optimizing reaction conditions such as temperature and feed velocity, characterizing the catalyst via changing nickel density and particle size and studying bi - metallic catalysts, and prel iminary screenings on the cross - condensation of fermentation - derived ethanol - fusel alcohols mixtures. Results o btained have been evaluated via the economic analysis of an industrial - scale butanol production plant and proposed a butanol selling price of $1. 55 - $1.6 0/kg for 25% ROI. To gain consideration for commercial development , the butanol price needs to be in the range of $1.30 - $1.40/kg, which requires the butanol selectivity to be >90% at 30 - 35% ethanol conversion. Here are some recommendations for futur e work that could lead to this objective. 6.2 Ethanol Guerbet Reaction 6.2.1 Catalyst Studies Guerbet reaction requires a multi - functional metallic acid - base supported catalyst. Metal sites are essential due to their hydrogen bond formation, while acid - base supports provid e ideal sites for the aldol condensation step. The presence of these sites in molecular distance from each other can be the key to improve the Guerbet products selectivity rather than the side - reactions. 140 Two strategies could be employed to enhance the num ber of metal and acid - base sites with molecular proximity to each other . The first one is to reduce the metal particle size and the number of segregated particles so that the dehydrogenated molecules can quickly condense on the neighbor acid - base si tes. Ca talytic preparation methods that have been developed recently, such as strong electrostatic adsorption (SEA) method [ 180 ] , or different impregnation, calcination, and reduction techniques [ 181 , 182 ] need to be employed to this end. Most of these new methods and techniques are still under development and can lead to the formation of coatings on the surface of the metal particles or the leaching and sintering of me tal. The refore, the main challenge facing this strategy is to activate metal particles for the hydrogen exchange. Nickel has been used as the primary metal in this study. Several studies have shown the effect of calcination and reduction temperatures on ni ckel par ticle sizes and its reducibility [ 183 - 185 ] . Therefore, optimizing the calcination and reduction temperature of the nickel particles not only enhances the deh y drogenation step but also affects the overall Guerbet reaction performance through the formation of smaller particles. The second strategy for enhancing the number of metal and acid - base sites close to each other is using atomic layer deposition (ALD) tec h nique to create an overcoat layer of support over the metal sites on the surface of the catalyst. This technique has been developed in several studies for the Al 2 O 3 supported metals [ 186 - 189 ] . A reduction in available surface area [ 186 ] has been reported in these studies as a result of covering the surface of the support, that can be addressed by calcining the overcoat at 700 o C [ 182 ] and creating multiple fractures on the surface of the deposited layer. Besides increasing the available surface area, this calcination helps with the re - exposure of metal sites to the surface of the catalyst. Therefore, the chances of the exposure of dehydrogenated molecules to the Al 2 O 3 sites for the fo llowing condensation step can improve significantly. 141 6.2.2 Separate Performance Improvement for Each Step of the Guerbet Reaction The Guerbet mechanism consists of three main steps: ethanol dehydrogenation to acetaldehyde, aldol condensation of acetaldehyde to c rotonaldehyde, and crotonaldehyde hydrogenation to n - butanol. So far, most of the studies have focused on improving the selectivity of butanol using a multi - functional catalyst that facilitates all three steps simultaneously. Nevertheless, this multi - funct ionality could lead to the production of undesired chemicals during each step of the reaction. Therefore, separately studying the proce ss design, catalyst characterization, chemistry, and kinetic modeling of each step of the Guerbet reaction could be an ap propriate strategy for further optimization of the Guerbet reaction. Literature review on the first step of the Guerbet reaction shows several reports with >97% acetaldehyde selectivity and ethanol conversion of >37% [ 190 - 192 ] . However, little work has been performed on distinctly improving the last two steps of the Guerbet mechanism. Our initial screenings on the aldol condensation reaction show that while aceta ldehyde is significantly more active compared to ethanol, the selectivity to crotonaldehyde is within the same range as the n - butanol selectivity from ethanol in the successive Guerbet reaction using a multi - func tional catalyst. This suggests that the aldo l condensation step could be the key reaction in the Guerbet mechanism for improving the butanol selectivity from ethanol. Nevertheless, aldol condensation reaction conditions have not been optimized in this stud y. Thus, designing reaction setups for exclu sively studying the last two steps of the Guerbet reaction can provide an economically and environmentally sustainable process to produce n - butanol at high selectivities. 6.2.3 Kinetic Modeling of the Reaction Tree Whi - essent ial methods for investigating the performance of catalytic systems, kinetic modeling of the system can provide valuable insight into the behavior of 142 different reacting components . The more comprehensive the model ing is in terms of reaction steps and side - r eactions considered , the more accurate and predictable results are obtained . Initial work has been done to model the Guerbet reaction with the batch reaction data using the multi - functional Ni/La 2 O 3 / - Al 2 O 3 catal yst [ 9 ] . Reactions considered for the current modeling are the four main steps of the ethanol Gue rbet reaction (including two hydrogenation steps), secondary Guerbet reactions involving ethanol and butanol, and ethyl acetate production from acetaldehyde and ethan ol . These reactions are considered as equilibrium reactions, and thermodynamic equilibrium cons tants (K) for each one of them are optimized based on the fit between experimental data and modeling results. Methane production reactions can also be added to t his modeling for further accuracy. Continuous experiments at 230 o C and a wide range of c ontac t times have been developed to verify the modeling results that are obtained based on the batch experimental data. For instance, experimental batch results show that ethyl acetate production reaction moves toward thermodynamic equilibrium at reaction condi tions. However, this observation is not verified by the initial continuous reactor experimental results. Thus, continuous data in a wide range of contact times c an be employed to fit the modeling results to a more accurate data. Another advantage that cont inuous data provide compared to the batch results is the precise analytical quantification of gaseous products at different contact times. Experimental results obtained from each step of the Guerbet reaction that was discussed in Section 6.2.3 will also have a significant role in finding kinetic rate constants. These experiments should be conducted at multiple temperatures to provide an activation energy and a pre - expone ntial factor for each reaction. Although the rate consta nts derived from each single reaction might be affected by the presence of other components, the data obtained from these studies provide an ideal starting 143 point for finding the rate constants or a re liable verification of the previously developed rate con stants. 6.3 Guerbet Economic Analysis One of the challenges in the economic analysis of the Guerbet process was the separation of the azeotropic ethanol/water mixture. As discussed earlier, studies have s hown the destructive effect of the presence of water on ethanol conversion [ 9 ] . Nevertheless, this effect was assumed to be negligible in the process simulation of th e ethanol Guerbet reaction, and the azeotropic ethanol/water mixture was recycled back to the system. Studies were also done to assess the economic feasibility of sep arating this mixture via a convent ional two - unit pressure swing molecular sieve unit. Resu lts indicated that utilizing this unit would increase the final selling price of the n - butanol. Among other alternative options for separating these two chemicals suc h as pervaporation membranes [ 102 ] a nd heterogeneous azeotropic distillation [ 193 ] , extractive distillation process seems to be the one with t he most promising results for commercial development [ 194 ] . This process involves the addition of a solvent th at interacts differently with the components that are forming an azeotropic mixture. This solvent typically alters the relative volatility of one of the components ag ainst the other one, which mak es the separation of the two components feasible [ 194 ] . This process has been si mulated and optimized in sever al studies for the ethanol/water mixture [ 195 , 196 ] and can be considered as a pote ntial approach that might improve the costs and quality of the butanol produ ction simultan eously. The four cases studied in the Guerbet reaction simulation involve two selectivity scenarios. The tanol and C 6 + alcohols, respectively; the high selectivity scenario correspo nds to 72% sel ectivity toward butanol and 22% selectivity toward C 6 + alcohols. However, experimental results in Chapter 2 have shown that the contribution 144 of the selectivity toward these two components is different from the mentioned cases and involves hig her ratios of C 6 + alcohols. Running simulation process and economic calculations at the selectivity distributions closer to the experimental results will significantly improve the precision of the economic analysis. 6.4 Fusel Alcohols Guerbet Reaction Fusel al cohols Guerbet reaction can be optimized in certain ways for both batch and continuous processes. Among those are finding batch products distribution at different reaction times, d oing extended batch reactions to address the stability of the catalyst, and studying a wid er range of feed compositions and reaction temperatures for the continuous system. Furthermore, the accuracy of the kinetic modeling results can improve if side - react ions and secondary and tertiary Guerbet reactions are also considered in the modeling stud ies. 6.5 Acrylate Production from 2 - Acetoxypropanoic Acid Esters Several approaches can be suggested for further improving the efficiency of the acrylate production proce ss. First, the carrier gas flow rate can be optimized to lower the contact time of the che micals with the catalyst and suppress the rate of the decomposition of reactive components. Seco nd, while materials with low (<1 m 2 /g) and high (>100 m 2 /g) surface ar eas have already been tested for the APA elimination reaction, the performance of moderate surface area contact materials (1 - 10 m 2 /g) has yet to be determined. Silica - based materials with low acidity and basicity are appropriate options for the initial exp eriments. The acidity and ba sicity of the contact material can be adjusted after optimizin g the surface area . Furthermore , 2 - acetoxy propanoic acid has not been detected in any of the analytical studies. Therefore, it is assumed that this chemical is being decomposed to smalle r molecules or eliminated 145 to acrylic acid upon its formation. Running control experiments with pure 2 - acetoxy propanoic acid, and studyin g the behavior of the 2 - acetoxy propanoic acid and APA ester mixtures as the feed at different temper atures would signifi cantly help in better understanding the mechanism and kinetics of the reaction. Finally, although the best results are obtained when methyl APA and benzyl APA were used as the feed, reaction conditions for these chemicals were never opt imized. Studying mul tiple temperatures, liquid feed flow rates, and gas feed flow rates ca n help in finding higher acrylate es ter yields from these chemicals. 146 REFERENCES 147 REFERENCES 1. Ritchie, H., How Long Before We Run Out of Fossil F uels? , in Ou r World in Data 2017. 2. Daly, H.E., Toward some operational principles of sust ainable development. Ecological economics, 1990. 2 (1): p. 1 - 6. 3. Kulshreshtha, S., Agricultural practices as barriers to sustainability. Public Policy in Food and Ag riculture, 2009: p. 220. 4. Sánchez, Ó.J. and C.A. Cardona, Trends in biotechnological pro duction of fuel ethanol from different feedstocks. Bioresource Technology, 2008. 99 (13): p. 5270 - 5295. 5. Nagy, Z.K., Model based control of a yeast fermentation bior eactor using optimally designed artificial neural networks. Chemical Engineering Journal, 2007. 127 (1): p. 95 - 109. 6. Ndaba, B., I. Chiyanzu, and S. Marx, n - Butanol derived from biochemical and chemical routes: A review. Biotec hnology Reports, 2015. 8 : p. 1 - 9. 7. Wass, D.F.D., George Richard Michael C onversion of alcohols , 2013, BP Biofuels UK Limited (Middlesex, GB): United States. 8. Ghaziaskar, H.S. and C. Xu, One - step continuous process for the production of 1 - butanol and 1 - hexanol by catalytic conversi on of bio - ethanol at its sub - /supercritical state. RSC Advances, 2013. 3 (13): p. 4271 - 4280 . 9. Jordison, T.L., C.T. Lira, and D.J. Miller, Condensed - Phase Ethanol Conversion to Higher Alcohols. Industrial & Engineering Chemistr y Research, 2015. 54 (44): p. 10991 - 11000. 10. Uyttebroek, M., W. Van Hecke, and K. Vanbroekhoven, Sustainability metric s of 1 - butanol. Catalysis Today, 2015. 239 : p. 7 - 10. 11. Algayyim, S.J.M., et al., Production and application of ABE as a biofuel. Renewa ble and Sustainable Energy Re views, 2018. 82 : p. 1195 - 1214. 12. Patakova, P., et al., Novel and neglected issues of ace tone butanol ethanol (ABE) fermentation by clostridia: Clostridium metabolic diversity, tools for process mapping and continuous fermenta tion systems. Biotechnology A dvances, 2013. 31 (1): p. 58 - 67. 13. Pfromm, P.H., et al., Bio - butanol vs. bio - ethanol: A t echnical and economic assessment for corn and switchgrass fermented by yeast or Clostridium acetobutylicum. Biomass and Bioenergy, 2010. 34 (4): p. 515 - 524. 148 14. Khamai seh, E.I., et al., Enhanced butanol production by Clostridium acetobutylicum NCIMB 13357 g rown on date fruit as carbon source in P2 medium. The Scientific World Journal, 2014. 2014 . 15. Gabriëls, D., et al., Review of catalytic systems and thermodynamics f or the Guerbet condensation reaction and challenges for biomass valorization. Catalysis Sc ience & Technology, 2015. 5 (8): p. 3876 - 3902. 16. Guerbet, M., méthylisobutylcarbinol et du diméthyl - 2.4 - heptanol - 6. Comptes rendus, 1909. 149 : p. 129 - 132. 17. Ndou, A.S., N. Plint, and N.J. Coville, Dimerisation of ethanol to butanol over solid - base catalysts. Applied Catalysis A: General, 2003. 251 (2): p. 337 - 345. 18. Ogo, S., et al., 1 - Butanol synthesis from ethanol over strontium phosphate hydroxyapatite catalysts with various Sr/ P ratios. Journal of Catalysis, 2012. 296 : p. 24 - 30. 19. Gines, M.J. and E. Igl esia, Bifunctional condensation reactions of alcohols on basic oxides modified by coppe r and potassium. Journal of Catalysis, 1998. 176 (1): p. 155 - 172. 20. Marcu, I. - C., et al., Catalytic valorization of bioethanol over Cu - Mg - Al mixed oxide catalysts. Cata lysis Today, 2009. 147 (3): p. 231 - 238. 21. Marcu, I. - C., et al., Catalytic Conversion o f Ethanol into Butanol over M Mg Al Mixed Oxide Catalysts (M = Pd, Ag, Mn, Fe, Cu, Sm, Yb) Obtained from LDH Precursors. Catalysis Letters, 2013. 143 (1): p. 23 - 30. 22. R iittonen, T., et al., One - Pot Liquid - Phase Catalytic Conversion of Ethanol to 1 - Butanol over Aluminium Oxide The Effect of the Active Metal on the Selectivity. Catalysts, 2012. 2 (1): p. 68. 23. Di Cosimo, J.I., et al., Structural Requirements and Reaction Pathways in Condensation Reactions of Alcohols on MgyAlO x Catalysts. Journal of Catalys is, 2000. 190 (2): p. 261 - 275. 24. Ogo, S., A. Onda, and K. Yanagisawa, Selective synthesis of 1 - butanol from ethanol over strontium phosphate hydroxyapatite catalysts. A pplied Catalysis A: General, 2011. 402 (1): p. 188 - 195. 25. Jordison, T.L., L. Peereboom , and D.J. Miller, Impact of Water on Condensed Phase Ethanol Guerbet Reactions. Industria l & Engineering Chemistry Research, 2016. 55 (23): p. 6579 - 6585. 26. Wu, X., et al., Catalytic Upgrading of Ethanol to n - Butanol: Progress in Catalyst Development. Che mSusChem, 2018. 11 (1): p. 71 - 85. 149 27. Weizmann, C., E. BERGMANN, and M. Sulzbache r, Further observations on the Guerbet reaction. The Journal of Organic Chemistry, 1950. 15 (1): p. 54 - 57. 28. Yang, C. and Z.Y. Meng, Bimolecular Condensation of Ethanol to 1 - B utanol Catalyzed by Alkali Cation Zeolites. Journal of Catalysis, 1993. 142 (1): p. 37 - 44. 29. Scalbert, J., et al., Ethanol condensation to butanol at high temperatures over a basic heterogeneous catalyst: How relevant is acetaldehyde self - aldolization? Jo urnal of Catalysis, 2014. 311 : p. 28 - 32. 30. Dowson, G.R.M., et al., Catalytic C onversion of Ethanol into an Advanced Biofuel: Unprecedented Selectivity for n - Butanol. Angewandte Chemie International Edition, 2013. 52 (34): p. 9005 - 9008. 31. Wingad, R.L., e t al., Catalytic Conversion of Ethanol to n - Butanol Using Ruthenium P N Ligand C omplexes. ACS Catalysis, 2015. 5 (10): p. 5822 - 5826. 32. Tseng, K. - N.T., et al., Upgrading ethanol to 1 - butanol with a homogeneous air - stable ruthenium catalyst. Chemical Commun ications, 2016. 52 (14): p. 2901 - 2904. 33. Xie, Y., et al., Highly Efficient Proc ess for Pr oduction of Biofuel from Ethanol Catalyzed by Ruthenium Pincer Complexes. Journal of the American Chemical Society, 2016. 138 (29): p. 9077 - 9080. 34. Jones, W.D.C., Su mit M ethods for producing butanol , 2017, University of Rochester (Rochester, NY, US): Unit ed States. 35. Ueda, W., et al., A low - pressure guerbet reaction over magnesium oxide catalyst. Journal of the Chemical Society, Chemical Communications, 1990(22): p. 1558 - 1559. 36. Ueda, W., et al., Condensation of alcohol over solid - base catalyst to form higher alcohols. Catalysis Letters, 1992. 12 (1 - 3): p. 97 - 104. 37. Kozlowski, J.T. and R.J. Davis, Heterogeneous Catalysts for the Guerbet Coupling of Alcohols. ACS C atalysis, 2013. 3 (7): p. 1588 - 1600. 38. Olson, E.S., R.K. Sharm a, and T.R. Aulich, Higher - alcohols biorefinery. Applied Biochemistry and Biotechnology, 2004. 115 (1): p. 913 - 932. 39. Tsuchida, T., et al., Reaction of ethanol over hydroxyapatite affected by Ca/P ratio of catalyst. Journal of Catalysis, 2008. 259 (2): p. 183 - 189. 40. Carvalho, D.L. , et al., Mg and Al mixed oxides and the synthesis of n - butanol from ethanol. Applied Catalysis A: General, 2012. 415 - 416 : p. 96 - 100. 41. Otto, F., Wilhelm, Querfurth , Process for the production of higher alcohols, particularly b utyl alcohol, from ethyl al cohol , 1935, Degussa: United States. 150 42. Di Cosimo, J., et al., Structure and surface and catalytic properties of Mg - Al basic oxides. Journal of Catalysis, 1998. 178 ( 2): p. 499 - 510. 43. León, M., E. Díaz, and S. Ordóñez, Ethanol catalytic condensation over Mg Al mixed oxides derived from hydrotalcites. Catalysis Today, 2011. 164 (1): p. 436 - 442. 44. León, M., et al., Consequences of the iron aluminium exchange on the pe rformance of hydrotalcite - derived mixed oxides for ethanol condensation. Applied Catalysis B: Environmental, 2011. 102 (3): p. 590 - 599. 45. Ordóñez, S., et al., Hydrotalcite - derived mixed oxides as catalysts for different C C bond formation reactions from b ioorganic materials. Catalysis Today, 2011. 167 (1): p. 71 - 76. 46. Carlini, C., et al., Gue rbet condensation of methanol with n - propanol to isobutyl alcohol over heterogeneous copper chromite/Mg Al mixed oxides catalysts. Journal of Molecular Catalysis A: C hemical, 2004. 220 (2): p. 215 - 220. 47. Benito, P., et al., Tunable copper - hydrotalcite der ived mixed oxides for sustainable ethanol condensation to n - butanol in liquid phase. Journal of Cleaner Production, 2019. 209 : p. 1614 - 1623. 48. Carlini, C., et al., Guerbet condensation of methanol with n - propanol to isobutyl alcohol over heterogeneous bi functional catalysts based on Mg Al mixed oxides partially substituted by different metal components. Journal of Molecular Catalysis A: Chemical, 2005. 232 (1): p. 13 - 20. 49. Bravo - Suárez, J.J., B. Subramaniam, and R.V. Chaudhari, Vapor - phase methanol and e thanol coupling reactions on CuMgAl mixed metal oxides. Applied Catalysis A: General, 2013. 455 : p. 234 - 246. 50. Hosoglu, F., et al., High resolution NMR unraveling C u substi tution of Mg in hydrotalcites ethanol reactivity. Applied Catalysis A: General, 20 15. 504 : p. 533 - 541. 51. Zhang, C.B., Mason ; Weiner, Heiko, Coated Hydrotalcite Catalysts and Processes for Producing Butanol , 2014, C elanese I nternational C orporati on (Irvi ng, TX, US): United States. 52. Hanspal, S., et al., Multiproduct Steady - State Iso topic Transient Kinetic Analysis of the Ethanol Coupling Reaction over Hydroxyapatite and Magnesia. ACS Catalysis, 2015. 5 (3): p. 1737 - 1746. 53. Birky, T.W., J.T. Koz lowski, and R.J. Davis, Isotopic transient analysis of the ethanol coupling reaction over magnesia. Journal of Catalysis, 2013. 298 : p. 130 - 137. 54. Tsuchida, T., et al., Direct Synthesis of n - Butanol from Ethanol over Nonstoichiometric Hydroxyapatite. Ind ustrial & Engineering Chemistry Research, 2006. 45 (25): p. 8634 - 8642. 151 55. Hanspal, S., et al., Influence of surface acid and base sites on the Guerbet coupling of ethanol to butanol over metal phosphate catalysts. Journal of Catalysis, 2017. 352 : p. 182 - 19 0. 56. S ilvester, L., et al., Reactivity of ethanol over hydroxyapatite - based Ca - enriched catalysts with various carbonate contents. Catalysis Science & Technology, 2015. 5 (5): p. 2994 - 3006. 57. Kourtakis, K.O., Ronnie ; D'amore, Michael B. , P rocess for p roducing guerbet alcohols using water tolerant basic catalysts , 2010, e.i. du pont de nemo urs and company (Wilmington, DE, US): United States. 58. Yang, K. - W., X. - Z. Jiang, and W. - C. Zhang, One - step synthesis of n - butanol from ethanol condensation over alu mina - supported metal catalysts. Chinese Chemical Letters, 2004. 15 (12): p. 1497 - 1500. 59. Riittonen, T., et al., Continuous liquid - phase valorization of bio - ethanol towards bio - butanol over metal modified alumina. Renewable Energy, 2015. 74 : p. 369 - 378. 60 . Panchenko, V.N., et al., Solid Base Assisted n - Pentanol Coupling over VIII Group Metals: Elucidation of the Guerbet Reaction Mechanism by DRIFTS. Industrial & Engineering Chemistry Research, 2017. 56 (45): p. 13310 - 13321. 61. Hernández, W.Y., et al., O ne - pot preparation of Ni - - Al 2 O 3 as selective and stable cataly st for the Guerbet reaction of 1 - octanol. Catalysis Communications, 2017. 98 : p. 94 - 97. 62. Earley, J.H., et al., Continuous catalytic upgrading of ethanol to n - bu tan ol and >C4 products over Cu/CeO 2 catalysts in supercritical CO 2 . Green Chemistry, 2015. 17 (5): p. 3018 - 3025. 63. Jiang, D., et al., Continuous catalytic upgrading of ethanol to n - butanol over Cu CeO 2 /AC catalysts. Chemical Communications, 2016. 52 (95): p. 13749 - 13752. 64. Onyestyák, G., et al., Guerbet self - coupling for ethanol valorization ove r activated carbon supported catalysts. Reaction Kinetics, Mechanisms and Catalysis, 2017. 121 (1): p. 31 - 41. 65. Onyestyák, G., Carbon Supported Alkaline Catalysts fo r Guerbet Coupling of Bioethanol. Periodica Polytechnica Chemical Engineering, 2018. 62 (1) : p. 91 - 96. 66. Yoshioka, T.T., Takashi ; Kubo, Jun ; Sakuma, Shuji Method for producing alcohol by guerbet reaction , 2015, Kabushiki Kaisha Sangi (Tokyo, JP): Uni ted States. 67. Wiles, C. and P. Watts, Continuous flow reactors: a perspective. Green Chemis try, 2012. 14 (1): p. 38 - 54. 152 68. Olcese, R. and M. Bettahar. Thermodynamics conditions for Guerbet ethanol reaction . in MATEC Web of Conferences . 2013. EDP Sciences. 6 9. Pang, J., et al., Upgrading ethanol to n - butanol over highly dispersed Ni MgAlO catalys ts. Journal of Catalysis, 2016. 344 : p. 184 - 193. 70. Seshu Babu, N., N. Lingaiah, and P.S. Sai Prasad, Characterization and reactivity of Al 2 O 3 supported P d - Ni bimeta llic catalysts for hydrodechlorination of chlorobenzene. Applied Catalysis B: Environmenta l, 2012. 111 - 112 : p. 309 - 316. 71. Sanchez - Sanchez, M.C., et al., Mechanistic Aspects of the Ethanol Steam Reforming Reaction for Hydrogen Production on Pt, Ni, and Pt - Al 2 O 3 . The Journal of Physical Chemistry A, 2010. 114 (11): p. 3873 - 3882. 72. Huang, W., Selective hydrogenation of acetylene on zeolite - supported bimetallic catalysts , 2007, University of Delaware. 73 . Weatherbee, G.D. and G.A. Jarvi, E ffects of carbon deposits on the specific activity of nickel and nickel bimetallic catalysts au - B artholomew , C.H. Chemical Engineering Communications, 1980. 5 (1 - 4): p. 125 - 134. 74. Muhammad, S., et al., Borohydride reduction of Al 2 O 3 supported Ni Cu bimetallic catalysts for the hydrogenation of styrene: study of surface properties. Rea ction Kinetics, Mechanisms and Catalysis, 2016. 118 (2): p. 537 - 556. 75. Dong, X., et al., The synergy effect of Ni - M (M= Mo, Fe, Co, Mn or Cr) bicomponent catalysts o n partial methanation coupling with water gas shift under low H 2 /CO conditions. Catalysts, 2017. 7 (2): p. 51. 76. Davda, R., et al., A review of catalytic issues and process conditions for re newable hydrogen and alkanes by aqueous - phase reforming of oxygen ated hydrocarbons over supported metal catalysts. Applied Catalysis B: Environmental, 2005 . 56 (1 - 2): p. 171 - 186. 77. Chen, L. - C. and S.D. Lin, Effects of the pretreatment of CuNi/SiO 2 on etha nol steam reforming: Influence of bimetal morphology. Applied Cat alysis B: Environmental, 2014. 148 - 149 : p. 509 - 519. 78. Influence of active metal loading and oxygen mobility on coke - free dry reforming of Ni Co bimetallic catalysts. A pplied Catalysis B: Environmental, 2012. 125 : p. 259 - 270. 79. Nez am, I., L. Peereboom, and D.J. Miller, Continuous condensed - phase ethanol conversion to hi gher alcohols: Experimental results and techno - economic analysis. Journal of Cleaner Production, 2019 . 209 : p. 1365 - 1375. 153 80. Nezam, I., L. Peereboom, and D.J. Miller , Enhanced Acrylate Production from 2 - Acetoxypropanoic Acid Esters. Organic Process Resear ch & Development, 2017. 21 (5): p. 715 - 719. 81. Chieregato, A., et al., On the Chemistry of Ethanol on Basic Oxides: Revising Mechanisms and Intermediates in the Lebed ev and Guerbet reactions. ChemSusChem, 2015. 8 (2): p. 377 - 388. 82. Hernández, W.Y., et al. , - Derived Mixed Oxides as Highly Selective and Stable Catalysts for the Synthesis - Branched Bioalcohols by the Guerbet Reaction. ChemSusChem, 2 016. 9 (22): p. 3196 - 3205. 83. Silvester, L., et al., Guerbet Reaction over Strontium - Substit uted Hydroxyapatite Catalysts Prepared at Various (Ca+Sr)/P Ratios. ChemCatChem, 2017. 9 (12): p. 2250 - 2261. 84. Ho, C.R., S. Shylesh, and A.T. Bell, Mechanism and K inetics of Ethanol Coupling to Butanol over Hydroxyapatite. ACS Catalysis, 2016. 6 (2): p. 93 9 - 948. 85. Wang, L., et al., Direct transformation of ethanol to ethyl acetate on Cu/ZrO 2 catalys t. Reaction Kinetics, Mechanisms and Catalysis, 2010. 101 (2): p. 36 5 - 375. 86. Voß, M., D. Borgmann, and G. Wedler, Characterization of Alumina, Silica, and Tit ania Supported Cobalt Catalysts. Journal of Catalysis, 2002. 212 (1): p. 10 - 21. 87. Sato, A.G., et al., Effect of the ZrO 2 phase on the structure and behavior of sup ported Cu catalysts for ethanol conversion. Journal of Catalysis, 2013. 307 : p. 1 - 17. 88. Si rijaruphan, A., et al., Cobalt Aluminate Formation in Alumina - Supported Cobalt Catalysts: Effects of Cobalt Reduction State and Water Vapor. Catalysis Letters, 2003 . 91 (1): p. 89 - 94. 89. Dias, M.O.S., et al., Butanol production in a sugarcane biorefinery u sing ethanol as feedstock. Part I: Integration to a first generation sugarcane distillery. Chemic al Engineering Research and Design, 2014. 92 (8): p. 1441 - 1451. 90. Pereira, L.G., et al., Butanol production in a sugarcane biorefinery using ethanol as feedst ock. Part II: Integration to a second generation sugarcane distillery. Chemical Engineering Resea rch and Design, 2014. 92 (8): p. 1452 - 1462. 91. Pereira, L.G., et al ., Life cycle assessment of butanol production in sugarcane biorefineries in Brazil. Journal of Cleaner Production, 2015. 96 : p. 557 - 568. 92. Pereira, L.G., et al., Economic and environment al assessment of n - butanol production in an integrated first and se cond generation sugarcane biorefinery: Fermentative versus catalytic routes. Applied Energy, 2015. 160 : p. 120 - 131. 154 93. Väisänen, S., et al., Carbon footprint of biobutanol by ABE fermentat ion from corn and sugarcane. Renewable Energy, 2016. 89 : p. 401 - 410 . 94. Brito, M. and F. Martins, Life cycle assessment of butanol production. Fuel, 2017. 208 : p. 476 - 482. 95. Tao, L., et al., Comparative techno - economic analysis and reviews of n - butanol production from corn grain and corn stover. Biofuels, Bioproducts & amp; Biorefining, 2014. 8 (3): p. 342 - 361. 96. Levasseur, A., et al., Assessing butanol from integrated forest biorefinery: A combined techno - economic and life cycle approach. Appl ied Energy, 2017. 198 : p. 440 - 452. 97. Quiroz - Ramírez, J.J., et al., Optimal Planning of Feedstock for Butanol Production Considering Economic and Environmental Aspects. ACS Sustainable Chemistry & Engineering, 2017. 5 (5): p. 4018 - 4030. 98. Patel, A.D., et al., Analysis of sustainability metrics and application to the catalytic pro duction of higher alcohols from ethanol. Catalysis Today, 2015. 239 : p. 56 - 79. 99. Michaels, W., et al., Design of a separation section in an ethanol - to - butanol pro cess. Biomass and Bioenergy, 2018. 109 : p. 231 - 238. 100. Fidkowski, Z.T., M.F. Doherty, and M.F. Malone, Feasibility of separations for distillation of nonideal ternary mixtures. AIChE Journal, 1993. 39 (8): p. 1303 - 1321. 101. A.A. Kiss, Eco - efficient butanol separation in the ABE fermentation process. Separation and Purification Technology, 2017. 177 : p. 49 - 61. 102. Kießlich, S., et al., Pervaporative buta nol removal from PBE fermentation broths for the bioconversion of glycer ol by Clostridium pasteurianum. Journal of Membrane Science, 2017. 535 : p. 79 - 88. 103. Anony mous, Ethanol Historical Prices , 2017, Business Insider. 104. Anonymous, Electric Power Mont hly , 2017, Energy Information Administration. 105. Anonymous, U.S. Natur al Gas Prices , 2017, Energy Information Administration. 106. Couper, J.R., et al., 21 - Cost s of Individual Equipment , in Chemical Process Equipment (Third Edition) , J.R. Couper, et al ., Editors. 2012, Butterworth - Heinemann: Boston. p. 731 - 741. 107. Purohi t, G., Estimating costs of shell - and - tube heat exchangers. Chemical engineering, 1983. 90 (17 ): p. 56 - 67. 108. Anonymous, Process Engineering Guide: Air Cooled Heat Exchanger Design . Vo l. 1. 2010: GBH Enterprises, Ltd. 80. 155 109. Kern, D.Q., Process heat transfer 1950: Tata McGraw - Hill Education. 110. Megyesy, E.F. and P. Buthod, Pressure vessel hand book 2008, Oklahoma: PV Pub. 111. Primo, J., Shell and tube heat exchangers basic calculation s. nd): n. pag. PDH Online. PDH, 2012. 112. McKetta Jr, J.J., Encycl opedia of Chemical Processing and Design . Vol. 50. 1987: CRC Press. 113. Anonymous, Economic Ind icators. Chemical Engineering, 2017. 124 (10): p. 1. 114. Peters, M.S., et al., Plant design and economics for chemical engineers . Vol. 4. 1968: McGraw - Hill New York. 115. Anonymous, n - Butanol: Chemical Market Insight and Foresight , in Chem - Net Facts 2013, T ecnon OrbiChem. p. 1. 116. Mariano, A.P., et al., Butanol production in a first - generation B razilian sugarcane biorefinery: Technical aspects and economics of g reenfield projects. Bioresource Technology, 2013. 135 : p. 316 - 323. 117. Bastidas, P., et al., Al cohol Distillation Plant Simulation: Thermal and Hydraulic Studies. Procedia Engineering, 20 12. 42 : p. 80 - 89. 118. Dale, R.T. and W.E. Tyner, Economic and techn ical analysis of ethanol dry milling: Model description. Staff Paper (06 - 04), Dept. of Ag. Econ. , Purdue University, 2006. 119. Fraas, A.P., Heat exchanger design 1989: John Wiley & Sons. 1 20. Al Abdallah, Q., B.T. Nixon, and J.R. Fortwendel, The enzymatic conversion of major algal and cyanobacterial carbohydrates to bioethanol. Frontiers in energy re search, 2016. 4 : p. 36. 121. Tesfaw, A. and F. Assefa, Current trends in bioethanol producti on by Saccharomyces cerevisiae: substrate, inhibitor reduction, grow th variables, coculture, and immobilization. International Scholarly Research Notices, 2014. 201 4 . 122. Hazelwood, L.A., et al., The Ehrlich Pathway for Fusel Alcohol Production: a Century of Research on <em>Saccharomyces cerevisiae</em> Metabo lism. Applied and Environmental Microbiology, 2008. 74 (8): p. 2259. 123. Boumba, V.A., K.S. Ziav rou, and T. Vougiouklakis, Biochemical pathways generating post - mortem volatile compounds co - detected during forensic ethanol analyses. Forensic Science International, 2008. 174 (2): p. 133 - 151. 156 124. Ehrlich, F., Über die Bedingungen der Fuselölbildung und über ihren Zusammenhang mit dem Eiweissaufbau der Hefe. Berichte der deutschen chemischen Ge sellschaft, 1907. 40 (1): p. 1027 - 1047. 125. Jackson, R.S., 7 - Fermentation , in Wine Science (Third Edition) , R.S. Jackson, Editor 2008, Academic Press: San Diego. p. 332 - 417. 126. Jackson, R.S., Chapter 6 - Qualitative Wine Assessment , in Wine Tasting (Th ird Edition) , R.S. Jackson, Editor 2017 , Academic Press. p. 253 - 291. 127. Vilanova, M., I.S. Pretorius, and P.A. Henschke, Chapter 58 - Influence of Diammonium Phos phate Addition to Fermentation on Wine Biologicals , in Processing and Impact on Active Compo nents in Food , V. Preedy, Editor 2015, Academic Press: San Diego. p. 483 - 491. 128. Marullo, P. and D. Dubourdieu, Yeast selection for wine flavour modulation , in Ma naging Wine Quality: Oenology and Wine Quality 2010, Elsevier. p. 293 - 345. 129. Branduardi, P. and D. Porro, n - butanol: challenges and solutions for shifting natural metabolic pathways into a viable microbial production. FEMS microbiology letters, 2016. 36 3 (8). 130. Matsu - ura, T., et al., - Alkylate d Dimer Alcohols Catalyzed by Iridium C omplexes. The Journal of Organic Chemistry, 2006. 71 (21): p. 8306 - 8308. 131. Busch, S.F. - s., Ingo; Mack, Sandra ; Mahnke, Eik e Ulf ; Wick, Anja Biocide composi tions comprising alkoxylation products of isoamyl alcohol derivatives , 2016, C ognis IP M anagement GMBH (Dusseldorf, DE): United States. 132. Beerthuis, R., G. Rothenberg, and N.R. Shiju, Catalytic routes towards acrylic ac id, - caprolactam starting from biorenewables. Green Chemistry, 2015. 17 (3): p. 1341 - 1361. 133. Anonymous, Acrylic Acid Market Analysis, By Product, By End - Use, Bio Acrylic Acid Downstream Potential And Segment Forecasts To 2022 , 2015, Gran d View Research. p. 120. 134. Mata r, S., M.J. Mirbach, and H.A. Tayim, Catalytic Oxidation Reactions , in Catalysis in Petrochemical Processes , S. Matar, M.J. Mirbach, and H.A. Tayim, Editors. 1989, Springer Netherlands: Dordrecht. p. 84 - 108. 135. Erwin, S. , Max, Gehrke, Franz, Aichner, Pro duction of acrolein , 1933, S chering K ahlbaum AG: United States. 136. Bühler, W., et al., Ionic reactions and pyrolysis of glycerol as competing reaction pathways in near - and supercritical water. The Journal of Supercritic al Fluids, 2002. 22 (1): p. 37 - 53. 157 137. Corma, A., et al., Biomass to chemicals: Catalytic conversion of glycerol/water mixtures into acrolein, reaction network. Journal of Catalysis, 2008. 257 (1): p. 163 - 171. 138. Ott, L., M. Bicker, and H. Vogel, Catalyti c dehydration of glycerol in sub - and supercritical water: a new chemical process for acrolein production. Green Chemi stry, 2006. 8 (2): p. 214 - 220. 139. Deleplanque, J., et al., Production of acrolein and acrylic acid through dehydration and oxydehydration of glycerol with mixed oxide cata lysts. Catalysis Today, 2010. 157 (1): p. 351 - 358. 140. Lourenço, J.P., M.I. Macedo, and A. Fernandes, Sulfonic - functionalized SBA - 15 as an active catalyst for the gas - phase dehydration of Glycerol. Catalysis Communications , 2012. 19 : p. 105 - 109. 141. Holme n, R.E., Production of acrylates by catalytic dehydration of lactic acid and alkyl l actates , 1958, M innesota M ining & MFG: United States. 142. Odell, B., G. Earlam, and D.J. Cole - Hamilton, Hydrothermal reactions of lactic acid catalysed by group: VIII Meta l complexes. Journal of Organometallic Chemistry, 1985. 290 (2): p. 241 - 248. 143. Saw icki, R.A.S., NY), Catalyst for dehydration of lactic acid to acrylic acid , 1988, Texaco Inc. (White Plains, NY): United States. 144. Papa rizos, C.W., OH), Dolhyj, Serge R. (Parma, OH), Shaw, Wilfrid G. (Lyndhurst, OH), Catalytic conversion of lactic acid and ammonium lactate to acrylic acid , 1988, The Standard Oil Company (Cleveland, OH): United States. 145. Mok, W.S.L., M.J. Antal, and M. Jones, Formation of acrylic acid f rom lactic acid in supercritical water. The Journal of Organic Chemistry, 1989. 54 (19): p. 4596 - 4602. 146. Lira, C.T. and P.J. McCrackin, Conversion of lactic acid to acrylic acid in near - critical water. Industrial & Engin eering Chemistry Research, 1993. 3 2 (11): p. 2608 - 2613. 147. Abe, T.N., JP), Hieda, Shinichi (Niigata, JP), Process f or preparing unsaturated carboxylic acid or ester thereof , 1993, Mitsubishi Gas Chemical Company, Inc. (Tokyo, JP): United States. 148. Wang , H., et al., Rare earth metal mod ified NaY: Structure and catalytic performance for lactic acid dehydration to acry lic acid. Catalysis Communications, 2008. 9 (9): p. 1799 - 1803. 149. Sun, P., et al., Potassium modified NaY: A selective and durable catalyst for dehydration of lactic acid to acrylic acid. Catalysis Communications, 2009. 10 (9): p. 1345 - 1349. 150. Walkup, P .C.R., WA), Rohrmann, Charles A. (Kennewick, WA), Hallen, Richard T. (Richland, WA), Eakin, David E. (Kennewick, WA), Production of esters o f lactic acid, 158 esters of acrylic a cid, lactic acid, and acrylic acid , 1993, Battelle, Memorial Institute (Richland, WA): United States. 151. Zhang, J., et al., Evaluation of Catalysts and Optimization of Reaction Conditions for the Dehydration of Methyl La ctate to Acrylates*. Chinese Journ al of Chemical Engineering, 2008. 16 (2): p. 263 - 269. 152. Zhang, J., J. Lin, and P. Cen, Catalytic dehydration of lactic acid to acrylic acid over sulfate catalysts. Canadian Journal of Chemical Engineering, 2008. 86 (6): p . 1047. 153. Naito, S.K., Takao; I keda, Ritoko, Process for preparing unsaturated carboxylic acid ester , 1991, Mitsubishi Gas Chemical Company, Inc. (Tokyo, JP): United States. 154. Zhang, J., et al., Efficient Acrylic Acid Production through Bio Lactic Ac id Dehydration over NaY Zeolite Mo dified by Alkali Phosphates. ACS Catalysis, 2011. 1 (1): p. 32 - 41. 155. Sun, P., et al., NaY Zeolites Catalyze Dehydration of Lactic Acid to Acrylic Acid: Studies on the Effects of Anions in Potassium Salts. Industrial & En gineering Chemistry Research, 2010 . 49 (19): p. 9082 - 9087. 156. Näfe, G., et al., True Cataly tic Behavior of Lactic Acid Dehydration on Zeolite Na - Y in the Gas Phase Measured by Means of a Novel Apparatus Design. Catalysis Letters, 2014. 144 (7): p. 1144 - 115 0. 157. Tang, C., et al., Catalyti c dehydration of lactic acid to acrylic acid over dibarium pyrophosphate. Catalysis Communications, 2014. 43 : p. 231 - 234. 158. Li, C., et al., Efficient catalytic dehydration of methyl lactate to acrylic acid using sulphat e and phosphate modified MCM - 41 ca talysts. Applied Catalysis A: General, 2014. 487 : p. 219 - 2 25. 159. Ghantani, V.C., et al., Catalytic dehydration of lactic acid to acrylic acid using calcium hydroxyapatite catalysts. Green Chemistry, 2013. 15 (5): p. 1211 - 1217. 160. Zhang, J., et al., Sodi um nitrate modified SBA - 15 and fumed silica for efficient production of acrylic acid and 2,3 - pentanedione from lactic acid. Journal of Industrial and Engineering Chemistry, 2014. 20 (4): p. 1353 - 1358. 161. Ghantani, V.C., M .K. Dongare, and S.B. Umbarkar, No nstoichiometric calcium pyrophosphate: a highly efficient and selective catalyst for dehydration of lactic acid to acrylic acid. RSC Advances, 2014. 4 (63): p. 33319 - 33326. 162. Mäki - Arvela, P., et al., Production of Lactic Acid/Lactates from Biomass and Th eir Catalytic Transformations to Commodities. Ch emical Reviews, 2014. 114 (3): p. 1909 - 1971. 159 163. Kuppinger, F ranz - f elix; H engstermann , Axel ; Stochniol, Guido ; Bub, Günther; Mosler, Jürgen; Sabbagh, Andreas P rocess for pr eparing acrylic acid purified by c rystallization from hydroxypropionic acid and ap paratus therefore , 2011: United States. 164. Decoster, D., S. Hoyt, and S. Roach, Dehydration of 3 - hydroxypropionic acid to acrylic acid. Patent WO2013192451, USA, 2013. 165. Craciun, L.B., Gerald P. ; Dewing , John;Schriver, George W.; Peer, William J. ; S iebenhaar, Bernd ; Siegrist, Urs Preparation of acrylic acid derivatives from alpha - or beta - hydroxy carboxylic acids , 2005: United States. 166. Lane, J. Cargill acquires OPX Biotechnologies . 2015. 167. Burns, R., D.T. Jones, and P.D. Ritchie, 88. Studies i n pyrolysis. Part I. The pyrolysis of - acetoxypropionic acid, and related substances. Journal of the Chemical Society (Resumed), 1935: p. 400 - 406. 168. Smith, L.T., et al., Pyrolysis of lactic acid derivatives. Industrial & Engineering Chem istry, 1942. 34 (4): p. 473 - 479. 169. Fisher, C., W. Ratchford, and L.T. Smith, Methyl acrylate production by pyrolysis of methyl acetoxypropionate. Industrial & Engineering C hemistry, 1944. 36 (3): p. 229 - 234. 170. Ratchford, W. and C. Fisher, Methyl acrylate by pyrolysis of methyl acetoxypropionate. Industrial & Engineering Chemistry, 1945. 37 (4): p. 382 - 387. 171. Godlewski, J.E.V., Juan Esteban ; Collias, Dimitris Ioannis Con version of Methyl - 2 - Acetoxy Propio nate t o Methyl Acrylate and Acrylic Acid , 2013, G odlewski J ane E llen ,V elasquez J uan E steban, C ollias D imitris I onnis : United States. 172. Rehberg, C., W. Faucette, and C. Fisher, Preparation of methyl acetoxypropionate: re action of lactic acid with methyl acetat e. Industrial & Engineering Chemistry, 1944. 36 (5): p. 469 - 472. 173. Lilga, M.A.W., Todd A. ; Holladay, Johnathan E., Methods of forming alpha, beta - unsaturated acids and esters , 2006, Battelle Memorial Institute (Ri chland, WA, US): United States. 17 4. Bee rthuis, R., et al., Catalytic acetoxylation of lactic acid to 2 - acetoxypropionic acid, en route to acrylic acid. RSC Advances, 2015. 5 (6): p. 4103 - 4108. 175. Fruchey, O.S.H., WV, US), Malisezewski, Thomas A. (Charles ton, WV, US), Sawyer, John E. (Cha rlesto n, WV, US), Acrylic acid from lactide and process , 2015, F ruchey Olan S.,M alisezewski T homas A.,S awyer J ohn E.: United States. 176. Miller, D.L., Carl T. ; Peereboom, Lars ; Kolah, Aspi K. , M ethod for producing acyl oxy carboxylic acids and derivativ es the reof , 2013, board of trustees of michigan state university (East Lansing, MI, US): United States. 160 177. Chuchani, G., et al., Kinetics and mechanism of the gas - phase elimination of primary, secondary and tertiary 2 - ac etoxycarboxylic acids. Journal of Physical Organic Chemistry, 2000. 13 (11): p. 757 - 764. 178. Elliott, J.R. and C.T. Lira, Introductory chemical engineering thermodynamics . Vol. 184. 1999: Prentice Hall PTR Upper Saddle River, NJ. 179. Wang, B., et al., The effect of K 2 HPO 4 and Al 2 (SO 4 ) 3 mo dified MCM - 41 on the dehydration of methyl lactate to acrylic acid. RSC Advances, 2014. 4 (86): p. 45679 - 45686. 180. Hakim, S.H., et al., Synthesis of supported bimetallic nanoparticles with controlled size and composition distributions for active site eluc idation. Journal of Catalysis, 2015. 328 : p. 75 - 90. 181. Chen, D.H. and S. H. Wu, Synthesis of Nickel Nanoparticles in Water - in - Oil Microemulsions. Chemistry of Materials, 2000. 12 (5): p. 1354 - 1360. 182. Cargnello, M., et a l., Efficient Rem oval of Organic L igands from Supported Nanocrystals by Fast Thermal Annealing Enables Catalytic Studies on Well - Defined Active Phases. Journal of the American Chemical Society, 2015. 137 (21): p. 6906 - 6911. 183. Al - Fatesh, A.S.A. and A.H. F akeeha, Effects o f calcination and activation temperature on dry reforming catalysts. Journal of Saudi Chemical Society, 2012. 16 (1): p. 55 - 61. 184. Manukyan, K.V., et al., Nickel Oxide Reduction by Hydrogen: Kinetics and Structural Transformations. The Jo urnal of Physical Chemistry C, 201 5. 119 (28): p. 16131 - 16138. 185. Smoláková, L., et al., Effect of Calcination Temperature on the Structure and Catalytic Performance of the Ni/Al 2 O 3 and Ni Ce/Al 2 O 3 Catalysts in Oxidative Dehydrogenation of Ethane. Industr ial & Engineering Chemistry Resear ch, 2015. 54 (51): p. 12730 - 12740. 186. O'Neill, B.J., et al., Stabilization of Copper Catalysts for Liquid - Phase Reactions by Atomic Layer Deposition. Angewandte Ch emie International Edition, 2013. 52 (51): p. 13808 - 13812. 187. O'Neill, B.J., et al., Contro l of Thickness and Chemical Properties of Atomic Layer - Al 2 O 3 Catalysts. ChemSusChem, 2014. 7 (12): p. 3247 - 3251. 188. O'Nei ll, B.J., et al., Operando X - ray Absorption Spectroscopy S tudies of Sintering for Supported Copper Catalysts during Liquid - phase Reaction. ChemCatChem, 2014. 6 (9): p. 2493 - 2496. 189. Catalyst Design with Atomic Layer Deposition. ACS Catalysis, 2015. 5 (3): p. 1804 - 1825. 161 190. Morales, M.V., e t al., Bioethanol dehydrogenation over copper supported on functionalized graphene materials and a high surface area graphite. Carbon, 2016. 102 : p. 426 - 436. 191. Zhang, P., et al., A Highly Porous Carbon Support Rich in Graphitic - N Stabilizes Copper Nanoc atalysts for Efficient Ethanol Deh ydrogenation. ChemCatChem, 2017. 9 (3): p. 505 - 510. 192. Lu, W. - D., et al., Copper Supported on Hybrid C@SiO = Hollow Submicron Spheres as Active Ethanol Dehydrogenat ion Catalyst. ChemNanoMat, 2018. 4 (5): p. 505 - 509. 193. Go mis, V., et al., Dehydration of Et hanol Using Azeotropic Distillation with Isooctane. Industrial & Engineering Chemistry Research, 2007. 46 (13): p. 4572 - 4576. 194. Gil, I., L. García, and G. Rodrígu ez, Simulation of ethanol extractive distillation with mix ed glycols as separating agent. Br azilian Journal of Chemical Engineering, 2014. 31 (1): p. 259 - 270. 195. Gil, I., et al., Separation of ethanol and water by extractive distillation with salt and solvent as entrainer: process simulation. Brazilian Journal o f Chemical Engineering, 2008. 25 (1 ): p. 207 - 215. 196. Kiss, A.A., R.M. Ignat, and C.S. Bildea, Optimal Extractive Distillation Process for Bioethanol Dehydration , in Computer Aided Chemical Engin eering Varbanov, and P.Y. Liew, Editors. 2 014, Elsevier. p. 1333 - 1338.