The Chemistry of Selective Oxidation of Sulfur Compounds and its Relation to Fuel Desulfurization Citation Attar, Amir (1977) The Chemistry of Selective Oxidation of Sulfur Compounds and its Relation to Fuel Desulfurization. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/5923-b957. https://resolver.caltech.edu/CaltechTHESIS:10172025-215430915 Abstract See thesis PDF for abstract. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemical Engineering and Environmental Engineering Science) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemical Engineering Minor Option: Environmental Science and Engineering Thesis Availability: Public (worldwide access) Research Advisor(s): Corcoran, William Harrison Thesis Committee: Unknown, Unknown Defense Date: 9 July 1976 Funders: Funding Agency Grant Number National Science Foundation UNSPECIFIED Record Number: CaltechTHESIS:10172025-215430915 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:10172025-215430915 DOI: 10.7907/5923-b957 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 17724 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 27 Oct 2025 18:28 Last Modified: 27 Oct 2025 18:51 Thesis Files PDF - Final Version See Usage Policy. 52MB Repository Staff Only: item control page

THE CHEMISTRY OF SELECTIVE OXIDATION OF SULFUR COMPOUNDS AND ITS RELATION TO FUEL DESULFURIZATION Thesis by Amir Attar In Partial Fulfillment of the Requirements For the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1977 (Submitted July 9, 1976) -ii- Dedication This thesis is dedicated to Ilana, my wife, and to Eyal and Ron, my two sons. -iiiACKNOWLEDGMENTS I wish to thank my advisor, Dr. William H. Corcoran, for his support in all the phases of this work. The financial support of the National Science Foundation is gratefully acknowledged. The construction of the experimental system would have been impossible without the help of George Griffith, Ray Reed, Chico Nakawatase, Zig Jenner and Eric Siegel. Different aspects of the experimental work were done by Robert Forgey, Russell Bone, John Horn, Alex Brooks, Paul Goodson, Dale Shultz, Mike Porter and Mike Huster. Their ready help and understand- ing is greatly appreciated. Glenn Wood who worked with me many long nights in the computing center deserves special thanks. I wish to thank Russell Bone, who spent with me hours correcting English mistakes, and Ruth Stratton who typed the thesis. Their help made this work readable. It was a great pleasure knowing the Caltech community and working with so many nice people. -ivABSTRACT An alternative method to desulfurize fossil fuel is proposed. There are two steps in the method: l) Selective oxidation of the sulfur compound to the corresponding sulfoxide (I) and sulfone (II), s [OJ + so [OJ + I 2) Pyrolysis of the product to a hydrocarbon and sulfur dioxide, I':, so 2 + Hydrocarbon+ so 2 The critical step is the selective oxidation of the sulfur compound to I and II. This oxidation was carried out by peroxides that were formed_:!.!!_ situ when a hydrocarbon solution of the sulfide was oxidized with molecular oxygen in the presence of a catalyst like HBr, at 140-l60°C. Butyl sulfide was used as a model sulfur compound and toluene and Tetralin in bromobenzene were used to simulate the organic medium. l. The most important results are: In the absence of HBr the oxidation proceeds to the sul foxide only, but when HBr is present also sulfone is found. 2. HBr increases the rate of oxidation of the hydrocarbon and the reactivity of free radicals. The butyl radical is attacked about 14 times more rapidly in the presence of HBr. 3. Denote: SUM= [sulfoxide] + [sulfone] SUM ij;MPR = L(oxidation products of sulfides 1 sulfoxide, sulfone) -vLet the subscript o indicate the absence of catalyst and presence of catalyst. I II 1/JMPR > 1/JMPR o i/JMPR 0 l the Then -5 SUM< 10 mole/liter when l decreases when SUM increases. increases when SUM increases. III For a given [HBr],conditions can be found for which 1/JMPR -4 and SUM = SUM 1/JMPR l o SUM = 0 , e.g., when [HBr] = 6.4 • 10 mole/liter }3 . = SUM = 1.8 • 10 mole/liter, and 1/JMPR = 1/JMPR = 15.5 is o l o l achieved after 65 min. (When [HBr] = 0, 300 min are required to reach the same conditions.) 4. Denote D(R-H) the smallest bond energy from all hydrogens in the system. Then a) the oxidation of the sulfur compounds is less selec- tive the larger D(R-H) is, more products are found and intense oxidation of the butyl radical is observed; b) the oxidation potential of the medium is larger when D(R-H) is larger; c) more hydrocarbon is oxidized when D(R-H) is smaller. 5. When protic solvent is added, e.g., 1-octanol, the selectivity to sulfoxide increases and fewer products of free radical oxidation of the sulfide are found. was present. However, no sulfone was detected when 1-octanol The octanol stabilizes the sulfoxide by hydrogen bonding with an estimated energy of -1.55 kcal/mole. -vi Table of Contents Chapter l. INTRODUCTION l l. l 2 Desulfurization by Selective Oxidation 1.2 The Proposed Method to Achieve Selective Oxidation of Sulfur Compounds 5 l . 3 The Goal of the Work 9 1.4 The Scope of the Work 10 1. 5 Main Results 10 l. 5. l l. 5. 2 l. 5. 3 l. 5.4 l. 5. 5 Chapter 2. The Effect of the Catalyst The Results of the Use of HBr The Effect of the Hydrocarbon Medium The Effect of the Solvent Proticity Loss of Heating Value 10 12 13 14 15 INTRODUCTION TO THE AUTOCATALYTIC OXIDATION OF HYDROCARBONS WITH MOLECULAR OXYGEN 16 2. 1 Initiation Reactions 16 Propagation Reactions 20 2.3 Termination Reactions 22 2.2 2.4 The Effect of Oxidation Products on the Reaction 23 2.4. l Elementary Reactions of Hydroperoxides and Peroxy Radicals 2.4.2 Oxidation of Secondary Alcohols and Ketones Chapter 3. 23 28 2.5 Oxidation of a Hydrocarbon in the Presence of a Homogeneous Catalyst 30 OXIDATION OF ORGANIC SULFUR COMPOUNDS BY PEROXIDES 34 3.1 Literature Survey of the Rate Data on the Oxida- 34 ·.tion of Sulfur Compounds by Peroxides 3.2 Chemistry and Mechanism of Oxidation of Sulfur Compounds by Peroxides 34 -viiChapter 4. THERMODYNAMIC ANALYSIS AND THE LOSS OF HEATING VALUE DUE TO OXIDATION 53 4.1 Calculation of the Equilibrium Concentration 53 4.2 The Trade-Off of Energy vs. Desulfurization 54 4.2.l A Method for the Evaluation of Enthalpy Changes Due to a Complex Reaction 4. 2.2 Application to a Reacting System 4.2.3 Case Study : The Enthalpy Loss in the Co-Oxidation of Tetralin and Dibutyl Sulfide Chapter 5. 55 56 57 EXPERIMENTAL 62 5. l On the Choice of the Experimental System 62 5. 1.l The Reactor 5.1 .2 The Choice of Sulfur Compound 5. 1.2. l The Choice of the Chemical System 5. 1. 2. 2 The Choice of Hydrocarbon Medium and Solvent 5. 1. 2.3 The Choice of Catalyst 62 63 63 64 5. 2 The Reactor and the Experimental Procedure 5.2.l Mechanical Design of the Experimental System 5.2.2 Experimental Procedure 5. 3 Methods of Chemical Analysis Methods of Qualitative Analysis 5.3.l.l IR Techniques 5.3.1.2 GC-MS Computer 5.3 . 2 Methods of Quantitative Analysis 5. 3.2.l Detennination of H20 2 and Hydroperoxides by Polarography 5.3.2.2 Acidometric Titration of Organic Acids 5. 3.2.3 Liquid Chromatography 5. 3.3 Analytical Procedure 5. 3. l 65 66 66 73 74 74 75 76 78 78 83 83 83 -viii - Chapter 6. 5.4 Estimation of the Possible Mass Transfer in the Reactor 84 EXPERIMENTAL RESULTS 86 6. 1 Experiment l 88 6.2 6. 1. l Experimental Conditions 6. l. 2 Qualitative Results 6. l. 3 Quantitative Results 89 Experiment 2 92 6. 2. l Experimental Conditions 6.2.2 Qualitative Results 6.2.3 Quantitative Results 92 92 93 6.3 Experiment 3 6.3. l Experimental Conditions 6.3.2 Qualitative Results 6.3.3 Quantitative Results 6.4 Experiment 4 6.5 6.6 6.7 88 91 95 95 95 96 97 6. 4. 1 Experimental Conditions 6.4.2 Qualitative Results 6.4.3 Quantitative Results 97 97 Experiment 5 l 01 6.5. l Experimental Conditions 6.5.2 Qualitative Results 6.5.3 Quantitative Results 101 101 Experiment 6 104 6. 6. l Experimental Conditions 6.6.2 Qualitative Results 6.6.3 Quantitative Results 104 105 105 Experiment 7 106a 6.7. 1 Experimental Conditions 6.7.2 Qualitative Results l 06a 98 l 02 107 -ix6.7 . 3 Quantitative Results 6.8 Experiment 8 108 6.8. l Experimental Conditions 6.8.2 Qualitative Analysis 6. 8.3 Quantitative Analysis 108 109 109 Experiment 9 110 6.9 . l Experimental Conditions 6.9.2 Qualitative Results 6.9.3 Quantitative Results 110 111 ANALYSIS AND DISCUSSION OF THE RESULTS 113 7.1 The Structure and Logic of the Discussion 113 7.2 Svstem Approach to the Simulation of the Chemi cal System 114 6.9 Chapter 7 l 07 111 7.3 The Interactions of the Sulfur Compounds in the 116 Subsystem NOCAT 7.4 The Coupling Reactions of Subsystems NOSC and NOCAT 7.5 Computer Simulation of the Reaction System 122 122 7.6 Simplifications Used in the Simulation and Their 123 Justification 7.6.l Rate Data for Individual Reaction Steps 125 7.6.2 Initial Values and the Concentration of 02127 and HBr in the Solution . 7.6.3 On the Comouter Program and the Manipula- 129 t ion of the Data 7.6.4 The Steady-State Approximation, Its Limitations and Application in the Simulation 131 7.7 The Results of the Computer Simulation 135 7.8 Evaluation of HBr as a Catalyst 153 7.8.l The Effect of HBr on the Overall Rate of 153 Oxidation 7.8.2 The Effect of HBr on the Rate and Selec- 158 tivity of the Oxidation of Aliphatic Sulfides in Tetralin -x7.8.3 Ionic Reactions and Their Relation to the HBr 7.8.4 The Potential of HBr as a Catalyst for the Selective Oxidation of Aliphatic Sulfides to Sulfones and Sulfoxides 7.8.5 Summary of the Evaluation of HBr as a Catalyst 160 7. 9 The Hydrocarbon Medi um and Its Effect on the Oxidation of Butyl Sulfide 171 The Kinetics of SO and so 2 Formation in Different Hydrocarbons 7.9.2 The Kinetics of COSO Formation in Different Hydrocarbons 7.9.3 Summary of the Discussion of the Effect of the Hydrocarbons 172 7.10 The Influence of the Protic Nature of the Solvent on the Rate of Reaction 180 7.10. 1 The Effect of the Protic Nature of the Solvent on the Selectivity of the Oxidation 7.10.2 On the Interaction between Sulfoxides and Protic Solvents 7. 10.3 Sul'l111ary of the Effect of the Solvent on the Oxidation of Butyl Sulfide 181 182 Loss of Heating Value Due to the Oxidation 185 7.9. l 7.11 7.12 The Oxidation of Dibenzothiophene Chapter 8 SUMMARY AND RECOMMENDATIONS 161 170 177 179 184 188 189 8. 1 Summary 189 8.2 Recommendations 190 References 192 Appendix A: Separation of Acidic Oxidation Products for MS 199 Appendix Bl: Purification Procedure for Tetralin 200 Appendix B2: Deactivation of the Reactor Wall 200 -xi List of Tables Table No. 1.1 Arrhenius Parameters for the Abstraction of Hydrogen by a Peroxy Radical and for .the Decomposition of Hydroperoxi des 8 l. 2 Initial Conditions for the Experiments 11 2.1 Activation Energies for the Dissociation of Initiators 18 2.2 R-H Bond Dissociation Energies 19 2.3 Rate Constants for the Hydrogen Abstraction Reaction 21 2.4 Rate Constants for the Recombination of Peroxy Radicals 23 2. 5 Kinetic Data for Homogeneous Processes 25 3.1 Rate Constants for the Oxidation of Sulfur Compounds by Peraci ds 35 3.2 Rate Constants for the Oxidation of Sulfur Compounds by Hydroperoxides and Peracids 40 4.1 The Change in the Functional Groups in the Co-Oxidation of Tetralin and Butyl Sulfide at 140°C 60 5.1 Major Components of the Experimental System 68 5.2 IR Absorption Bands for Sulfur Functional Groups 75 5.3 Methods of Qualitative Analysis 79 5.4 Methods of Analysis by Gas-Chromatography 81 6.1 Qualitative Results of Experiment l 89 6.2 Quantitative Results of Experiment l 91 6.3 Qualitative Results of Experiment 2 92 6.4 New IR Bands Formed in Experiment 2 93 6.5 Quantitative Results for Major Components, Experiment 2 94 6.6 Qualitative Results for Experiment 3 95 6.7 Quantitative Results for Experiment 3 96 -xiiTable No. 6.8 QuaJitative Results for Experiment 4 97 6.9 Quantitative Results for Major Components, Experiment 4 99 6.10 Quantitative Results for Components that Appeared after 100 Long Reaction Time, Experiment 4 6.11 Qualitative Results, Experiment 5 101 6.12 Quantitative Results, Experiment 5 103 6. 13 Quantitative Results for Components Fonned after Long Reaction Time, Experiment 5 104 6.14 Qualitative Results, Experiment 6 105 6.15 Quantitative Results, Experiment 6 106 6.16 Quantitative Results for Components Fonned after Long Reaction Period, Experiment 6 106a 6.17 Qualitative Results, Experiment 7 107 6.18 Quantitative Results, Experiment 7 107 6.19 Qualitative Results, Experiment 8 109 6.20 Quantitative Results, Experiment 8 109 6.21 Qualitative Results, Experiment 9 111 6.22 Quantitative Results, Experiment 9 111 7.1 Abbreviations of Species Names Used 118 7.2 Rate Constants for Elementary Steps in the Oxidation of Tetralin by Molecular Oxygen 126 7-:,,3 Reactions and Rate Cons tan ts for the Subsystem RH-0 2-t-tBr (NOSC) Reactions and Rate Constants for the Subsystem RH-0 2-s (NOCAT) Coupling Reactions of NOCAT and NOSC and Their Rate Constants 136 A List of Figures from the Simulation with Explanation of Their Significance 141 7.4 7.5 7.6 137 138 -xiiiTable No. 7.7 Reactions and Rate Constants for the Fit of Experiment 5 143 7.8 The Initial Conditions and Effect of [HBr] on the Rate of Oxidation of Tetralin 155 7.9 The Initial Conditions of Experiments 2, 4, and 6 171 -xiv- Li st of Figures Figure No. 1.1 Schematic Description of a Hydrodesulfurization Process 3 1.2 Schematic Description of a Process for Desulfurization by Selective Oxidation 4 5.1 Schematic Flow Diagram of the Experimental System 67 5.2 The Chemical Reactor 70 5.3 Sampling Device for Liquid Solutions 71 5.4 Flowmeter for Small Streams of Gas 72 7.1 The Oxidation of a Hydrocarbon in the Presence of HBr 119 7.2 The Interactions of an Aliphatic Sulfide with the System Hydrocarbon-0 2-HBr 120 7.3 The Free-Radicals vs. Time, Experiment 1 144 7.4 Major Oxidation Products vs. Time, Experiment 1 145 7.5 Water and Oxidizing Species vs. Time, Experiment l 146 7.6 Major Oxidation Products vs. Time, Experiment 3 147 7.7 Major Sulfur-Contreining Oxidation Products, Experiment 3 148 7.8 Water and Oxidizing Species, Experiment 3 149 7.9 Second-Generation Sulfur-Containing Oxidation Products vs. Time, Experiment 3 150 7.10 Hydrocarbon and Sulfi de vs. Ti_me, Experiment 5 151 7.11 The Absorbance of 310 nm vs. Time, Experiments 3, 4, and 5 , The Concentration of Butyl-Sulfone vs. Time, Experiments 4 and 5 The Concentration of Butyl-Sulfoxide vs. Time, Experiments 3, 4, and 5 The Relative Mechanistic Selectivity as a Function of Time, Experiments 3, 4, and 5 154 7.12 7.13 7. 14 156 162 164 -xvFigure No. Page 7. 15 The Relative Desulfurization Selectivity vs. Time, Experiments 3, 4, and 5 166 7. 16 The Sum of Concentrations of Sulfoxides and Sulfones vs. the Relative Mechanistic Selectivity The Effect of the Hydrocarbon Medium on the Formation of Sulfoxides 168 7. 18 The Concentration of Sulfones vs . Time, Experiments 2, 4, and 6 176 7. 19 The Concentration of a-keto-butyl-sulfoxide vs. Time, Experiments 2, 4, and 6 178 7.20 The Concentration of a-keto-butyl-sulfoxide and of Butyl Sul foxi de vs. Time, Experiments 4 and 7 183 7.21 The Loss of Heat of Combustion Due to the Oxidation, Experiment 5 186 7. 22 The Loss of Heating Value vs the Percent of Sulfide That Was Oxidized 187 7. 17 173 -xvi- LIST OF ABBREVIATIONS l. A list of the abbreviations of the species names is given in Table 7. l , page 118. 2. A list of the abbreviations of names of the methods of qualitative analysis is given in Table 5.3, page 79. 3. A list of the abbreviations of names of the methods of quantitative analysis is given in Table 5.4, page 81. Meaning Abbreviation BUS butyl sulfide CATSC The chemical system Hydrocarbon-O 2-catalyst-Sulfur compound. DBS di butyl sulfide DBT dibenzothiophene D( R-H) The smallest bond energy of all the carbon-hydrogen bonds of a molecule. EXP experiment FG functional group FGD functional group's distribution FR free radicals HA hydrogen abstraction, except for Chapter 3, where it denotes a protic molecule. MO molecular oxygen NOCAT the system Hydrocarbon-Oxygen-Sulfur compound NOSC the system Hydrocarbon-Oxygen-HBr OCTOL 1-octanol SC TRLN sulfur compound Tetralin ' -xvii- Nomenclature S_ymbol Meaning A frequency factor Ce HBr equilibrium concentration of HBr Ce 02 equilibrium concentration of o 2 c.l concentration of the ;th component c. concentration of the jth functional group co initial concentration D diffusivity J Do diffusivity of oxygen 2 E activation energy H enthalpy of formation H. enthalpy of formation of the ;th molecule HR enthalpy of reaction HS enthalpy of formation of the system per unit volume M molecular weight MS ratio of rate of mass transfer from the gas to the rate of fastest reaction. p concentrations of oxidizers 1 partial pressure of component j gas constant entropy of formation of activated complex concentration of sulfur compounds that were oxidized by a peroxide molecule [S]oxidized total concentration of oxidized sulfur compound -xviiiT absolute temperature V volume of reaction, except in Eq. 5.3 where it means the specific volume of the solvent total rate of initiation X association factor h. enthalpy of formation of the j th group k reaction rate constant J overall rate of oxidation of sulfide overall rate of oxidation of sulfoxide frequency factor of the ; th reaction mass transfer coefficient number of functional groups in the system number of times that the j th functional group appears in the ; th component m n .. Jl time t the time that is required to achieve 99 % of the equilibrium value of a concentration coefficient of proportion in Polayni equation, E = ooH ]J viscosity p density Notes l. 6 denotes difference 2. t H8 , tHH' and tH 0 are defined on p. 182 3. ljJ denotes selectivity. All .selectivities are defined on pp. 163 and 165. 4. ~ under a variable denotes a vector. -1- Chapter l INTRODUCTION Desulfurization of fossil fuels has become a major issue in the world's economics and politics. It is not surprising, therefore, that many investigators are currently examining the technology and science of desulfurization. One of the major contributors to the problem known as the "energy crisis" is the lack of desulfurization technology. Because of air pol- lution regulations and the lack in the United States of low-sulfur fuel, the huge U.S. energy market looked for other sources of fuel abroad. However, high sulfur fuel in the form of coal does exist in the United States and it would be possible to use it, if it could be desulfurized first. The function of science and technology is to provide a process which will convert high-sulfur fuel into a clean energy source. The task of desulfurizing fuel is different in many ways from the classic problems the chemical industry has faced, for example, the scale of the problem and of its solution. No desulfurization process should be considered before the impact of its nationwide use on the steel and transportation industries is studied. The energy-efficiency of a process, i.e., how much of the input energy will be lost in order to achieve desulfurization, is important, as is the disposal of the refuse. The above-mentioned technological questions are still without solutions. One .way to approach desulfurization is presented here, as well as exploratory experiments on one special system. -21. 1 Desulfurization by Selective Oxidation* Current desul furi zati on processes can be cl ass ified into two categories: l) processes for the desulfurization of gaseous combustion products; and 2) processes for the desulfurization of dense fuel prior to combustion. Many processes have been proposed for desulfurization of gaseous combustion products; however, in the mind of the author, gas-phase desulfurization cannot be more than a temporary solution. The concen- tration of so 2 in the combustion products is small, and the resistance to so 2 mass transfer is in the gas phase. The particular solution or solid used to scrub it is of secondary importance only. The penalty for gas-phase desulfurization is paid in the fonn of large capital investments in huge processing units and in large operation costs. The only method practiced commercially to desulfurize liquid fuel prior to combustion is hydrodesulfurization. Hydrodesulfurization of fuels is widely practiced and much literature is available on the process (H-10, S-5, S-7, M-5). Work is currently done on hydrodesulfurization by many research groups. However, the basic facts about hydrodesul furi zation have remained unchanged in the last forty years: Different aspects of desulfurization were reviewed by the author in previous reports and therefore will not be detailed here. The industrial processes for hydrodesulfurization of oil were reviewed in 1972(A-1). A comparison of desulfurization by hydrogenation and by selective oxidation was made in 1973 (A-2, A-6). The area of organic sulfur compounds in coal was reviewed in 1976 (A-3). -3- l. Hydrodesulfurization is a process which needs hydrogen. The , only large-scale commercial process to obtain H2 is from fossil fuels, 2. and its cost is continuously on the rise. Hydrodesulfurization needs high pressures and temperatures, and therefore ties up large amounts of capital in huge processing equipment. 3. Many peripheral processes are required which add to the cost of hydrodesulfurization and reduce its energy efficiency. 4. The selectivity of hydrodesulfurization is poor when thiophenes are desulfurized. High pressures and temperatures are required and considerable amount of H2 is used to saturate aromatics. In Figure l.,l ' the scheme of a general hydrodesulfurization process is demonstrated. ,-- , 0 _ _ _ _ _ _ j L. I I Reactor High Sul_!ur CO 2 2 H2+co 2 -- -· .,eparator - - 0 ,,2 ,__ Claus ProrPSS . 7 H, ' -II I - l • I I ~ J• I Fuel , ' I -- Reactor I - Separator - I L--. I I _J - 7 r· I -· -- Figure 1.1. --·--- - • --1 r low sulfur sulfur fuel Schematic Description of a Hydrodesulfurization Process -4- Many high-pressure and high-temperature processing stages are used between the feed and the final product; however, this staging increases the capital irivestment and the cost of operation. The next question that may be asked is whether the same end products can be obtained in a more efficient way. Hydrogen was re- quired as an intermediate because it can selectively attack the sulfur-to-carbon bond, C-S,which is very similar to the carbon-tocarbon bond, C-C. The C-S bond can, however, be cleaved thermally if the sulfur is oxidized first to the corresponding sulfoxide (I) and sulfone (II) so [OJ (1.1) ➔ I so 2 ~ · hydrocarbon+ so 2 ( l. 2) A second scheme for desulfurization based on reactions 1.1 and 1.2 is given in Figure 2: 02 ---------- -I High Sulfur I Selectiv - - - - - 1 Oxida- Fuel ; Low i----~--◄ Pyrolysi+----~ Sulfur '---"-.:..x.&....__, ,, I Fuel I L. Figure 1.2. ___I Schematic Description of a Process for Desulfurization by Selective Oxidation. -5- Several differences are noted between Figures 1 and 2: 1. The sulfur outlet of hydrodesulfurizatjon is elementary sulfur, while a gaseous so 2 is produced when selective oxidation is used. 2. Fewer steps are required in scheme 2 (although the method to accomplish selective oxidation was not specified.) 3. CO 2 and H2O are produced in scheme 1. No such products are recognized in scheme 2, provided that the oxidation is truly selective. Hydroxyl and carbonyl groups, as well as olefinic groups will be formed in the desulfurized fuel as a result of poor selectivity. The comparison of the two schemes shows that if a sufficiently selective oxidation process can be devised, desulfurization of fossil fuels might be done in a much cheaper and more efficient way than by hydrogenation. The goal of this present work w~s to probe how selective oxidation of sulfur compounds can be achieved. 1.2 The Proposed Method to Achieve Selective Oxidation of Sulfur Compounds Selective oxidation of sulfur compcrunds to the corresponding sulfoxi des and sul fones can be achieved by hydroperoxi des and by other peroxy species. (A detailed literature review is given in .Chapter 3.) Hydroperoxides and other peroxy compounds are intermediates in the oxidation •of hydrocarbons- , (A detailed literature review is given in Chapter 2.) If these two facts are coupled, the following scheme for selective oxidation of sulfur compounds emerges: l) oxidize the hydrocarbon medium -6- to the extent that a sufficiently large concentration of peroxides is formed; 2) the peroxides that are formed in the medium will selectively oxidize the sulfur compounds to the sulfoxides and sulfones. The next question is how to effect selective oxidation of the ' hydrocarbon to maximize the concentration of the intermediate peroxides. The concentration of the peroxides in an organic solvent is controlled for each hydrocarbon by the relative rates of formation and of annihilation of the peroxide. The general scheme of oxidation in excess of oxygen is: In it i at i on : Propagation: Termination: ( l. 3) kl R· + o + R0 , 2 2 fast ( l. 4) k2 R0 • + RH + R0 H + R· 2 2 rate limiting ( l. 5) 2Ro • 2 k6 -►- ROH + R• c~---0 + o 2 ( l. 6) The hydroperoxide R0 2H decomposes and provides additional means of branching: (l. 7) The radicals RO· and ·OH abstract hydrogens and enhance the conversion: RO• + RH kg + kg ·OH+ RH + ROH + R• ( l . 8) H 0 + R• 2 ( l. 9) R· is an unspecified organic radical. -7- The concentration of the hydroperoxide in the system is determined essentially by the rate of the decomposition reaction (l. 7) and the rate of reaction (1.5) which controls the rate of conversion of the hydrocarbon. the temperature When excess hydrocarbon and oxygen are present, determines the relative rates of reactions (1.5) and (1.7), and thus the concentration of the hydroperoxide .is controlled by the temperature. If the rate of reactions (1 . 5) and (1.7) is expressed in the Arrhenius form, then: (1. 10) -E/RT (1.11) k3 = k30 e where ki is the rate constant, kio and Ei are the frequency factor and the activation energy, respectively, R is the gas constant, and T is the absolute temperature. The sensitivity of the rate constant to the temperature is given by: cl l n k clT = E RT2 (1. 12) Equation (1.12) shows that the sensitivity of the rate constant to the temperature is determined at each temperature, by the value of the activation energy. Values of the Arrhenius parameters of reactions (l.5) and (1.7) are given in Table 1.1. -8Table l. l Arrhenius Parameters for the Abstraction of Hydrogen by a Peroxy Radical and for the Decomposition of Hydroperoxides Methyl Primary Second. Tert. Tetralyl Ref. D(R-H) 102.5 96.9 93 . 2 90.0 '\., 83. 5h K-4 log k20 8.8 8. 7 8.2 8.2 5. 75e K-4 E2 10. 3 7.5 5. l 4.7 8.3e K-4 13.4a 15. 2b 13. 7C K-5 E3 13 .7a 40.0b 37.8c K-5 log k30 10.0f 12.0 9 11. l d 26.9f 31.7 9 29.0d Hydrogen log k30 E3 gas liquid aValues for ethyl radicals bValues for i-propyl radicals cValues fort-butyl radicals dRef. (T- l ) ' eRef. (H-6) fvalues for octyl radicals in white oil, Ref. (T-1) 9values for isodecyl radicals in white oil, Ref. (T-2) hEstimated in this work, see Section 7. 10. The data of Table 1.1 show that the rate of decomposition of the hydroperoxides has much larger activation energy than the rate of hydrogen abstraction, and is therefore much more sensitive to the temperature. Therefore, if the oxidation is carried out at lower temperatures, the half life of the hydroperoxides will be significantly increased . . However, the rate determining step, reaction (1-5), will be even slower. -9- The natural deduction from the previous discussion is the following: If the oxidation is carried out at a sufficiently low temperature and the rate of the hydrogen abstraction stage is catalyzed, it will be possible to obtain relatively large concentrations of hydroperoxides, and thus to oxidize selectively the sulfur compounds that are present. HBr was reported in the literature to be a good catalyst for the hydrogen abstraction step, even at low temperatures (150-230°C) (R-1). Full discussion of the role of HBr is given in Section 2.5. The mech- anism of the enhancement of the hydrogen abstraction by HBr is k10 R0 • + HBr ➔ R0 H + Br, 2 2 kll Br,+ RH ➔ HBr + R• (1. 13) (1. 14) ( l . 4) The result of the sequence of reactions (1.13) and (1 .14) is the same as that of (1 .5); however, the rate of propagation is much larger. The previous discussion leads to the following conclusion: in order to oxidize the sulfur compound selectively, oxidize in a hydrocarbon at low temperature and in the presence of a catalyst for the hydrogen abstraction step . In this work this concept was probed as a potential method for the selective oxidation of sulfur compounds. 1.3 The Goals of the Work The main goals were: 1) verify experimentally if selective oxidation of sulfur compounds can be accomplished by low temperature oxidation in the presence of a catalyst to the hydrogen abstraction step . . -102) Derive kinetic data on the oxidation for selected systems and use it to study i) the mechanism of the reaction; ii) the effect of the catalyst; iii) the variation of the kinetics with the hydrocarbon medium; iv) the effect of the solvents on the oxidation. 1.4 Scope of the Work Nine experiments are described. The details of these experiments and their results are given in Chapter 6. given in Chapter 7. Discussion of results is In Table 1.2 the most important parameters of the initial conditions of the experiments are summarized. 1.5 Main Results Detailed discussion of the results is given in Chapter 7, so only a brief summary of the most important conclusions is given here. 1.5. 1 The Effect of the Catalyst Tables 7.3 and 7.4 show the rate data of the elementary steps of the catalyzed oxidation of Tetralin and of the uncatalyzed oxidation of butyl sulfide in Tetralin. The data indicate that v , the ratio of the propagation step in the presence of a catalyst and in its absence is: k11 [RH][Br•] v = k2[RH][R02·] ~ 10 The increased rate of oxidation as measured by the rate of formation of ketone is first order with the concentration of the HBr. 5.514 5.002 II II 9 5.002 Tetral in 7 8 2.177 Toluene 6 5.614 II 5 5.514 Tetra I in 3 5.395 - - 2 II 5. 039 Tetral in 1 4 RH Cone. mol e/1 it Hydrocarbon ( RH) Experiment No. 0. 1292 0. 1292 Dibenzoth i ophene II 140 o. 1292 II 160 160 136.5 0.1424 II 140 140 140 0. 1424 0.2784 0. 1424 140 140 Temperature oc II II II 0.1423,, - Butyl Sulfide SC Cone. mole/1 it 6.4 6.4 ,. 2 . 5.5 5.6 5.6 1. 2 l. 2 0.9 1.0 6.0 5.7 0.64 5.6 5.6 mol e/1 it 103 o2 Cone. l. l l. 34 HBr Cone. mol e/1 it 103 Initial Conditions of the Experiments Sul fur Compound (SC) Table 1.2 7.078 2. 130 2.081 2 .130 2. 130 1-0ctano 1 1 . 889 Bromobenzene 1-0ctanol 1.882 II II II II 9. 281 __, __, 2 .130 Bromebenzene II SOL Cone. mol e/1 it Sol vent (SOL) -12- 1.5.2 The Results of the Use of HBr The main results of the use of HBr are: l. Increased rate of oxidation. The rate constant of ketone formation is about 5.3 times larger when the ox i dation is catalyzed by HBr . 2. Formation in the solut i on of compounds with larger oxidation potential than the compounds produced when HBr is absent. The oxida- tion potential that is required to oxidize sulfoxides to sulfones is larger than that required to oxidize the sulfides to the sulfoxides. Sulfones are found when HBr is present, but are not found when HBr is absent . 3. The activity of free radicals is increased about 100 times by 1. 3 • l0- 3 mole/liter HBr. The increased activity of the radicals shows as an increase in the side chain reactions of the sulfide, e.g . , the formation of about 100 times more of a-keto butyl sulfoxide (COSO) than the amount that is formed when HBr is absent. The ratio of the concentrations of COSO remained about constant for the first two hours. The addition of only 0.64 • 10- 3 mole/liter HBr doubled the activity of the free radicals, and hence increased the concentration of species like COSO by a factor of 2-3 only. This shows that the dominant mech- anism of the oxidation of the sulfide is a complex function of the concentration of HBr. 4. Change in the selectivity of the reaction. [SJ per = [sulfoxide] + [sulfone] and Define: (1.15) -13- l/JMPR = where (1.16) [S]oxidized - [S]per [S]per is the molar concentration of sulfur compounds that were produced by a reaction with at least one peroxide molecule, [SJ oxidized is the total number of moles of sulfide that were oxidized, l/JMPR is the mechanistic selectivity which defines the ratio of moles of sul11 11 fur which were oxidized by a homolytic reaction with the peroxides and that reacted by a free radical reaction. A plot of [S]per vs l/JMPR for the uncatalyzed oxidation defines a region_ in which, by the use of HBr, better selectivity and larger conversion to sulfoxides and sulfones can be obtained. Whether such con- ditions will be realized depends on the concentration of HBr and on the residence time in the reactor. When [HBr] = 6.4 • 10- 4 mole/liter, better l/JMPR and larger [S]per were obtained after 65 min. (Comparable l/JMPR and [S]per conditions were obtained in the uncatalyzed reaction after about 310 min.) When (HBr] = 1.34 • 10- 3 mole/liter the equivalence point was not reached in five and a half hours. 5. Sulfoxides oxidize HBr to Br2 and water. This process takes place in ionic mechanisms and is considerably accelerated when water separates as a suspension in the solution. A list of estimated rate constants for the reactions of Br containing species is given in Table 7.3. 1.5.3 The Effect of the Hydrocarbon Medium The hydrocarbon medium serves both as a solvent and as the -14source of radical R· . The oxidation potential of the solution and the reactivity of the radicals depends on the structure of R· . The tolyl radical is very unstable and therefore less selective than the tetralyl radical . Many more products of free radical side chain re- actions are found when toluene is used ins t ead of Tetralin . Moreover, less stable R· forms less stable R0 2• radicals and oxidizers with a larger oxidation potential. There f ore , in toluene, the sulfoxide and the sulfone are oxidized to more advanced oxidation products than when the hydrocarbon medium is Tetralin. by the energy of the bond R-H. The reactivity of R, is measured The rate of hydrogen abstraction of all the common aromatic hydrocarbons is between that of toluene and Tetralin . * Therefore, one may infer that when a mixed medium is used the rates and selectivities which can be expected are between those found for Tetralin and those for toluene . The initial rate of fonnation of COSO in toluene is 7. 5 times that in Tetralin. However, the initial rates of formation of sulfoxides is approximately the same, so when small conversions are considered one may expect the same yield of sulfoxides in any aryl hydrocarbon, but different concentrations of products of free radicals attack on the side chain. 1.5.4 The Effect of the Solvent Proticity Protic solvents stabilize the sulfoxides by formation of the complex[-S-o-+lo-R]and solvate the free radical R0 2• . Therefore, large concentrations of sulfoxides and fewer products of reactions of free *The difference in energy of the R-H bond between tetralin and toluene was estimated in this work to be about 1.5 kcal/mole. -15radicals are found when the hydrocarbon is modified by a protic solvent, e.g., octyl alcohol. The energy of solvation of butyl sulfoxide by octyl alcohol was estimated to be -1 . 55 kcal/mole, which is smaller than the energy of salvation of dimethyl sulfoxide by phenol (-6.5 kcal/mole) but larger than the "heat of mixing" of water and dimethyl sulfoxide (at 0. 1 M), -0 . 36 kcal/mole. 1.5.5 Loss of Heating Value Duet~ the Oxidation Experimental measurements have indicated that the loss of heating value due to the oxidation was less than the experimental error in the calorimetric measurement that was used (1-1.5%). Estimates of the loss of heating value, based on the computer simulation of the reaction, have shown that a loss of about 1-2 cal/gram (or 1-2 kcal/liter) occurred. The results are encouraging, since a loss of such a small magnitude for 25% conversion of the sulfide was not anticipated. -16Chapter 2 INTRODUCTION TO THE AUTOCATALYTIC OXIDATION OF HYDROCARBONS WITH MOLECULAR OXYGEN A considerable amount of literature is available on the liquid phase oxidation of hydrocarbons, therefore the discussion will be limited to the homogeneous, liquid-phase oxidation of hydrocarbons by molecular oxygen in the absence of salts of transition metals. Furthermore, only the reactions that take place at low temperatures (below 200°C) and with oxygen pressure of one atmosphere or less will be described. The material was collected from several books and review articles (E-1, S-3, I-1, I-2, G-3), Many other papers were used which will not be referenced specifically. The reactions of peroxides, hydroperoxides and the properties of peroxy compounds have recently been described in a three-volume manuscript edited by Swern (S-4). Oxidation of hydrocarbons by molecular oxygen proceeds via a free radical mechanism (FRM). In a FRM there are three types of reactions: (l) reactions in which more FR are produced than with the initial reactants; such reacti ans are termed II i niti ati on or branchi ng 11 II 11 reacti ens; (2) reactions in which the number of free radicals is the same in the reactants and in the products; or transfer 11 11 reactions; such reactions are termed propagation 11 (3) reactions in which the number of FR de- decreases; such reactions are termed terminati on 11 2.1 11 II reactions. Initiation Reactions Two kinds of initiation reactions can be distinguished: (1) decom- position of unstable compounds ( initiators 11 11 ), and (2) hydrogen abstraction -17- by molecular oxygen. Typical unstable initiators are azo and peroxy compounds~ Typical initiation reactions are: ( CH3)2 y--N = NJ~ CH3l2 + 2( CH3l2 ~~ • ( 2. 1 ) Two radicals are obtained from each azoisobutyronitrile molecule. Initiation via peroxides can be by a first or by a second order reaction, or by both. First order decompositions of hydrogen peroxide yields two·OH radicals: (2.2) Second order decomposition oft-butyl hydroperoxide (I) yields an alkoxy and a peroxy radical, and a water molecule is eliminated: (2.3) The first order decomposition yields an alkoxy and a hydroxy radical: (2.4) Several generalizations can be made on the decomposition reactions of peroxide: 1. Peroxide decomposition reactions have a large activation energy (28- 45 kcal/mole),and therefore the rate of initiation changes very rapidly with temperature. Values of typical activation energies for initiators are given in Table 2. 1. -18- Table 2. 1 Activation Energies for the Dissociation of Initiators Structure Activation Energy kcal/mole HO-OH 48 t-Butyl hydroperox i de t-Bu0-0-+1 42 t-Butyl peroxide t-Bu0-043u-t 0 37 t-Butyl pe rben zoate t-BuO-OC ph 0 0 34 Benzoyl peroxide ph CO-OC ph 0 ?i II CH 3 CO-OCCH 3 30 Name Hydrogen peroxide II II Acetyl peroxide 2. II 30-32 The structure of the organic ligand strongly influences the rate of decomposition of the peroxide and its activation energy. Electron repelling groups and resonance with n bonds stabilize the products of the decomposition of the peroxides. The peroxides of this group are relatively more stab 1e. Peroxides are intermediates in oxidation reactions with molecular oxygen (MO). MO can abstract a hydrogen atom and form a hydroperoxy radical (I) and a FR (II) (reaction (2.5)). The hydroperoxy radical reacts with a second molecule of hydrocarbon producing hydrogen peroxide and another FR (reaction 2.6). RH+ o 2 ➔ H0 2• + R· I (2.5) II (2.6) -19At temperatures below 90°C the H2o2 is stable; however, it may decompose at higher temperatures and cause further branching. The rate of initiation with molecular oxygen is strongly dependent on the bond energy of the hydrogens. Larger rates of initiation are observed when smaller bond energies exist. Bond energies for classes of hydrogens are given in Table 2.2: Table 2.2 R-H Bond Dissociation Energies Bond Energy kcal/mole Structure CH 3-t1 103 Ph CH 2--H 85 n-{;3H7-+I 99 RC;::::::O 'H 86 i-{;3H7-+I 94 CH 3S-+1 88 t-C 4H9-+I 90 CH 3pH-+1 85 CH 2=CH--H 105 phO-+I 88 C5H5--H 103 phNH--H 80 CH 3=CH-CH 2--H 85 ROO-H 90 Structure Bond Energy kcal/mole Larger rates of production of peroxide are observed when the FR that is formed is more stable. TI bonds or n electrons in a. position to the abstracted hydrogen enhance the rate of hydrogen abstraction with MO. The general form of the initiation reactions with peroxide R-0--0--R' a re: R--0-0-R' ~ RO·+ R'O· (2. 7) -2ORO• + R H + 11 ROH+ R 11 ( 2. 8) • Initiation with molecular oxygen can be written as: (2.9) 2.2 Propagation Reactions Propagation is a chain reaction, with at least two steps. The FR that is formed in the initiation step reacts very rapidly with molecular oxygen producing a peroxy radical: fast (2.10) The peroxy radical abstracts a hydrogen and forms a new FR and a hydro peroxide molecule: (2.11) The chain is completed by the last step, because R· can react with MO (2 . 10), thus the chain propagates; the result of the chain is the production of a hydroperoxide molecule from a hydrocarbon and an oxygen molecule. Since (2. 11) is considerably slower than (4 . 10), the rate of oxidation is determined by the rate of hydrogen abstraction by the peroxy radicals. Several generalizations may be made: l. Larger rates of hydrogen abstractions (HA) are obtained when the heat of reaction of the HA is larger. Thus, larger rates of oxidation are obtained for hydrocarbons with hydrogens with smaller carbon to hydrogen bond dissociation energy. The relative rates of the HA reac- tions for primary, secondary, and tertiary hydrogens is approximately -21- 1:30:300. Allylic, benzylic, thiotic and aminic radicals react more rapidly. 2. Hydrogens or carbons which are connected to oxygen, e.g . , in alcohols, ethers or aldehydes, can be easily abstracted because the free radical that is produced can resonate with then electrons of the oxygen. Therefore, partially oxidized hydrocarbon continues to oxidize much more rapidly than the parent hydrocarbon. 3. The bond energy of the hydrogens of hydroperoxides is smaller than that of alcohol, so hydrogen abstraction by R0 2 • is more selective than HA by RO· The free radicals can isomerize either before or after reacting with oxygen. A full discussion of such reactions is beyond the scope of this thesis, and the reader is referred to Vol. I of reference (S-5). A short discussion of the secondary oxidation reactions is given in Section 2.4.2. Typical rate constants for the HA reaction are given in Table 2.3. Table 2.3 Rate Constants for the Hydrogen Abstraction Reactions Temp. oc Rate Const. lit/mole/sec cumyl Tetralin tetralyl Tetralin cyclohexenyl cyclohexene 1-phenethyl ethyl benzene a-cumyl -j sopropyl benzene octenyl octene-1 benzoyl benzaldehyde hyd roxy-decyl decano l polyperoxyntyryl phenol 2,6-Di-t-butylphenol po lyperoxyntyryl 30 30 40 30 30 25 5 5 65 65 1. 65 po lyperoxys tyryl 65 Hyd roca rbo n RH di phenyl amine Peroxy Radi ca 1 RO· 2 6.4 3.7 1. 3 0.18 0.03 1900 720 5· l o3 5· ,a3 20· 10 3 -222.3 Termination Reactions Termination can occur by the recombination of two FR. Usually the most important termination reaction is that between the two most abundant FRs. If the partial pressure of o2 is greater than 100 torr, R0 2 • will be the most abundant radical because reaction(2.10) is very fast. Thus, the dominant termination reaction will be the recombination of two R0 2 • radicals. Such a reaction can produce a peroxide molecule: (2. 12) or other non~radical products, like an alcohol and a ketone molecule: 2RO 2• + ROH+ R'C=O + o2 (2.13) The particular products will depend on the structure of the organic radical. Typically, termination reactions will have zero or negligible activation energy, and a large frequency factor. has a weak temperature dependence. So, the rate of termination However, large changes in the temper- ature affect the rate of HA reactions by the peroxy radicals and the rates of their isomerization and decomposition. Selected average values of the rate of termination reactions by recombination of peroxy radicals are given in Table 2.4. -23Table 2.4 Rate Constants for the Recombination of Peroxy Radicals Alkylperoxy radicals 2k t ( i /mo 1e/sec) at 30 °C 8 primary, RCHif 2-4 • 10 secondary, R2CH0 2• 20-40 106 benzylic allylic tertiary 1-10 • 106 6-8 • 106 0. 1-60 • 10 4 R/0 2• H0 2• 1. 2 • 10 cyclic 9 2.4 The Effect of Oxidation Products on the Reaction Primary oxidation products influence both the rate and the selectivity of the oxidation. The oxidation potential of oxidation products is larger than that of molecular oxygen, so the hydrogen abstraction becomes less selective to the most labile hydrogens. In the presence of oxidation products, all the hydrogens become prone to abstraction. Furthermore, the overall rate of oxidation increases, since the decomposition of the primary products is in most cases a branching reaction. 2.4.1 Elementary Reactions of Hydroperoxides and Peroxy Radicals Peroxy radicals can react i n two important ways: 1) abstract a hydrogen from another molecule and form a hydroperoxide; this reaction is important at temperatures below 200°C and when hydrogens with a small bond energy are present. 2) decompose with a rupture of a C-C bond. An aldehyde or ketone (I) molecule is produced plus an alkoxy radical (II): -240-0 0-0 J_ C I When I O· 0 I I -+ -£· C I I I I I II II -+ --c + --c- (2.14) is an aldehyde, it branches according to: R'C'(_: + o2 -+ R'C0· + H0 2• (2.15) II The alkoxy radical., II, abstracts a hydrogen from a hydrocarbon mole- cule much more readily than does the peroxy radical. Decomposition of the radical to an olefin and a hydroperoxy radical becomes important at higher temperatures (450-500°C). In the presence of olefins, the peroxy radicals can add to the double bond. (2. 16) The addition reaction is reversible. absorb o2 or add The peroxide radical III can either to another olefin molecule; polyperoxide is formed. III can also decompose to an epoxide molecule, IV, and a hydroxy radical v, 0-0 · 0--0H I -C- CH- 0 I . -+ -C--C- -+ I\ --C - C- + • OH IV (2.17) V Polyperoxide formation occurs in solution only. Epoxide formation is important in the gas phase at high temperatures (above 300°C). Typical values for the frequency factors and the activation -25energies of some of the reactions are given in Table 2.5: Table 2.5 Kinetic Data for Homogeneous Processes Reaction log A* E kcal/mole RO 2• -+ aldehyde+ R1 0· 13 20 RO· 2 RO 2• -+ c/Q'c + · OH 11 15 -+ C= C 9 33 R0 • -+ R· + 0 2 2 RO 2• + RH -+ R0 2H + R· 14. 5 29 8 12 RO 2• + C=C -+ 8.5 5-7 R· + 02 -+ RO 2• 8.5 3 R· + 02 -+ C=C + HO· 2 9.2 4 15 43 • +HO· 2 R0 2H -+ RO · + ·OH * A is in sec - l or lit/sec/mole Once sufficient primary hydroperoxide and aldehyde molecules are formed in the system, the hydroperoxide can add to the carbonyl double bond : OH RCH200H + RCHO I -+ RCH2 OOTR VI (2.18) H The adduct VI can decompose to an aldehyde, an acid and a hydrogen molecule: -+ VI RCHO + RC~O + H 'OH 2 (2. 19) -26- The VI oxidation potential of VI is very large. In the presence of acid, VI decomposes to an ester by ionic mechanism. Secondary hydroperoxides react with peroxy radicals; ketones and hydroxy radicals are formed: R CHR + 2 1 I OOH R0 , 2 R CR 1 2 I OOH R cR 1 2 + + R CR 1 2 + R0 H 2 I (2.20) OOH + ·OH (2.21) II 0 Acids catalyze the decomposition of secondary hydroperoxides to ketones and water. R CHR + 1 2 I o--o-+l H+ -+- l lCH R~ H+ I + R CH R 1 2 I+ 0 ()---{)-tj t + RC R + 1 2 H0 2 ( 2. 22) H+ II 0 The ketones react with another hydroperoxide molecule forming a molecule of hydroxy peroxide (VII): + +- R1 .,....OH '--c (2.23) R~ 'o--OR VII VII is a very strong oxidizer. Peroxy radicals with a special structure can rearrange. bond is broken and the HC structure is changed. A C-C Tertiary peroxy radicals -27- and a-phenyl peroxy radicals are two important examples of this class of reactions: ( 2 . 24) The decomposition of tetralyl hydroperoxide is an interesting example of such a decomposition : H iH RH ~ O· H 90 00· ct) +·OH + RH CO2 ~ 00 ROz· + R· ?1 -00 +·OH , (2.25) The reactions of peroxides and peroxy radicals with sulfur compounds are discussed in Chapter 3. -28- 2.4.2 Oxidation of Secondary Alcohols and Ketones Alcohols and ketones are relatively stable compounds and may accumulate in the reaction system. activates the a However, the oxygen in the molecule and B hydrogens, so that their abstraction becomes a very facile reaction. Therefore, as long as the concentration of alco- hols and ketones remains small, they will accumulate, but when their concentration increases, in the oxidation. listed below. they will start to play an active role The most important reactions that may take place are Examples were taken from .the oxidation of cyclohexanol and cyclohexanone. Oxidation of Secondary Alcohols 1. The oxidation is initiated by molecular oxygen by reaction 2.9. k03 For cyclohexanol the Arrhenius form of the rate constant is = 8.3 exp(-16000/RT) lit 2;mo~e 2/sec. 2. Oxygen absorption is rapid. The HA propagation reaction (2. 11) proceeds exclusively to the hydrogens on the carbon a to the oxygen. Arrhenius form of the rate of propagation for cyclohexanol is k2 = l.l • 10 7exp(-ll900/RT) lit/mole/sec 3. Terminat"-wn takes place exclusively via 2.13, however, because of the presence of the hydroxyl, it is slightly modified. For cyclo- hexanol it is: O 2 OH ~ O 0-0· =O + O O (H OOH + 02 (2.26) k6= 5·10 6exp(-2200/RT) lit/mole/sec. The hydroperoxy alcohol (I) is in equilibrium with H2o2 and cyclohexanone. The rate constant is -29- (2. 27) I I is a strong oxidizer and can cause branching: + ·OH (2.28) k35= 2.2 • 10 4exp(-16200/RT) sec- 1. The decomposition is much faster in the presence of transition metals, acids or on The rate constant is surfaces. The rate of decomposition via reaction (2.28) is very large; accumulation of I can result in an explosion. Oxidation of Ketones The main products of the oxidation of ketones are a-ketohydroperoxides, lactones and mono and dibasic acids. tion of the f3 hydrogens. Initiation is by abstrac- The acti ~ation energy for the abstraction of hydrogens f3 to the keto-qxygen is approximately the same as that for the hydrocarbon, in other words, the carbonyl group does not significantly change the bond energy of the neighboring hydrogens. Once some a- ketohydroperoxide is formed, it becomes the major source of free radicals. The rate constant for the decomposition of a-hydroperoxycyclo- hexanone is ki= 6.0 • 10 7exp(-20400/RT) sec -1 . , Typical chain reactions which yield mono (II) -and dibasic (III) acids proceed according to: (2.29) The peroxy acid (IV) is a very strong oxidizer and can very rapidly oxidize nucleophiles like sulfur or nitrogen compounds. 2.5 Oxidation of a Hydrocarbon in the Presence of a Homogeneous Catalyst The rate data of Table 2.5 clearly indicate that the rate limiting step in the propagation is the hydrogen abstraction, reaction (2.11). The rate of oxidation will therefore be enhanced by any catalyst that can abstract hydrogen at a larger rate than RO 2• One such catalyst is HB r. In the presence of HBr, the following sequence of reactions takes place: (2.30) Br· + RH + HB r + R• ( 2. 31) The net result of reactions (2.30) and (2.31) is reaction (2. ll), however much larger rates may be obtained, provided that the HBr -31concentration is not too small , HBr enhances the rate of initiation by the reaction: (2. 32) Reaction (2.32) is reversible , so in the presence of excessive peroxy or hydroperoxy radical termination may occur. The rate of the forward reaction (2 . 32) is given by k = 3.16 • 10 14exp(37700/RT) (R-2) and that of the reverse reaction by k-=- 3.63· 10 14exp(-2750/RT) . The contribution of reaction (2. 32) to the rate of initiation in systems with a weak hydrogen - carbon bond, or systems which contain oxidation products , may be negligible; however, it may be important in the initial stages of the oxidation . The rate of hydrogen abstraction by Br· (2 . 31) is smaller than that by ·OH, and therefore Br· is more selective than the hydroxy or alkoxy radicals. Thus specific isomers can be obtained by oxidizing hydrocarbons in the presence of HBr· . Rust and Vaughan (R-1) obtained stable hydroperoxi des from branched hydrocarbons, e.g., t-butyl hydroperoxi de from isobutane and ketones from hydrocarbons with secondary hydrogens. They concluded that the hydrogen with the smallest bond energy will be abstrated by the Br· . Barnett et al. {B-12) examined the oxidation of toluene in the presence of HBr and Lin and Kehat (L -2) investigated the oxidation of toluene and benzene in the presence of HBr , HCl and water. All the studies indicate that the Br· is selec- tive to the hydrogen with the smallest bond energy; and therefore -32oxidation will take place at this site. The presence of HBr will affect the oxidation of hydrocarbons in the following ways: 1. Catalyze the propagation by reaction (2.30). 2. Catalyze the initiation by reaction (2.32). 3. Cause tennination in a special range of parameters, via reactions (2.33), (2.34) or (2.35): (2.33) 4. RO· 2 +Br·+ . olefin+ o2 + HBr (2.34) R· + Br· RBr (2.35) ~ Decompose the hydroperoxide to ketone via an ionic mechanism: (2.36) 5. Inhibit the oxidation when the ratio of HBr/0 2 is too large by competing with the propagation reaction (2. 10): (2.10) R· + HBr + RH+ Br· (2.37) For a given hydrocarbon, with a given hydrogen-carbon bond energy, a region of the parameters T, [0 2]/[HBr] and [RH]/[HBr] can be defined, in which catalysis will be observed. 6. Dehydrate alcohols by an ionic mechanism: -33[HBr] ROH where REN is an olefin. ➔ (2.38) Reaction 2.38 complicates the oxidation because oxidation of olefins is much more rapid than oxidation of paraffins. -34- Chapter 3 OXIDATION OF ORGANIC SULFUR COMPOUNDS BY PEROXIDES Oxidation of organic sulfur compounds (SC) was investigated by many researchers. The work was reviewed by Behrman and Edwards {B-2), Barnard et al. (B-3) and by Davies (D-5). More recently, Curci and Edwards (C-1) reviewed the oxidation of nucleophiles by peroxides, and the effect of the solvent on the oxidation of sulfur compounds. The oxidation of thiols (mercaptans) was reviewed very recently by Capozzi and Modena , {C-2). Oxidation of thiols will not be discussed here be- cause no work has been done on thiols in this study. 3. 1 Literature Survey of the Rate Data on the Oxidation of Sulfur Compounds by Peroxides The data that have been published on the oxidation of SC by peroxides are summarized in Tables 3.1 and 3.2. was covered. Only low-conversion oxidation Data on the oxidation of sulfides in less well defined systems have been omitted; so were data on systems where inorganic salts were included (e.g.,(D-2)). Oxidation of sulfides which contain double bonds and oxidation of sulfides in solvents with double bonds were also not included. 3.2 Chemistry and Mechanism of Oxidation of Sulfur Compounds by Peroxides Curci and Edwards (C-1) have recently reviewed the mechanism of oxidation of sulfur compounds. clusions are given here. Therefore only generalizations and con- II II II II _ II II II II Di benzothiophene II II II II II II Di phenyl sulfide II II II II II II phenyl methyl sulfide II II Di-t-butyl sulfide II II Di-t-butyl sulfide II II Di-sec butyl sulfide II II II Di-n-butyl sulfide II toluene peracetic Tetramethylene II II benzene II II benzene+ 1ON acetic acid II II II II II II II So 1vent Acid Sulfur Compound Table 3.1 8.6 9.6 3.2 1.47 0. 83 5.65 o. 097 2nd 1:1 2nd 1:1 2nd 1: 1 2nd l : l -70 2nd 1:1 -20.0 40 30 2nd 1 : 1 2nd 1:1 2nd 1:1 2nd 1:1 20 15 2nd 1: 1 10 0. 0168 0.0091 14 . 7 14 . 7 14.7 9.6 o. 880 0. 0049 9.6 9.6 9.6 9. 6 9. 6 8. 6 0.483 0. 268 1. 7 2nd 1:1 -30.0 2nd 1:1 0. 53 2nd 1:1 -39 . 1 0 0. 24 2nd 1:1 -49 -50 -70 -70 4 2nd 1: 1 -70 E kcal/mole k** Rate* Temperature oc Rate Constants for the Oxidation of Sulfur Compounds by Peracids F-1 F-1 F-1 F-1 8.572 8. 572 8. 572 7. 166 7. 166 7. 166 F-1 F-1 F- 1 F-1 F-1 F-1 8. 338 F-1 8.338 8. 338 F-1 8.338 9. 164 9. 164 F-1 F-1 F-1 1og A Ref . I w I (J1 II II II II II II II II II II II tetramethylthiophene 2,5 dimethylthiophene 2 methylthiophene 3 methylthiophene II thi ophene II II II II benzothiophene II II II II II II II benzene benzene + 1ON acetic acid II II II II benzene+ 1ON acetic acid II benzene peraceti c II Sol vent Acid II di benzothiophene Sul fur Compound Continued 40 40 40 40 40 40 40 60 50 2nd 1: 1 2nd 1: 1 2nd 1: 1 2nd 1:1 2nd 1:1 2nd 1: 1 2nd 1:1 2nd 1:1 2nd 1:1 2nd 1:1 2nd l: 1 30 40 2nd 1:1 2nd 1:1 60 15 2nd 1 : 1 50 Temperature Rate* Tab 1e 3. 1 14.8 o. 142 5.8596 6 • 10- 5 F-1 4. 1 • 10- 4 3.4 • 10- 4 F-1 F-1 F-1 F-1 F-1 F-1 F-1 F-1 F-1 F-1 F-1 F-1 F-1 0.00186 0.0368 5. 8596 5.8596 5. 8596 5. 8596 5.8596 5. 8596 5. 8596 8.572 8. 572 0. 0017 0.0026 14. 8 14 . 8 14.8 14. 8 14.7 14.7 E log A Ref. 1O kcal/mole 0. 0725 0.0329 0. 0157 0.00433 0. 0694 0.0357 k** I w I O'\ II II II II II II II p-methyl perbenzoi c II II II II II perbenzoi c II II II II toluene pe roxybenzoi c II II II 2nd l:l 0.70 0.29 2nd 1:1 -30 -20 o. 11 2nd l:l -40 0.79 2nd 1:1 -20 II II 0.38 2nd l:l -30 0. 15 2nd l : l 4. 17 2nd 1:1 -45 -40 2.50 2nd 1:1 -55 II i -propyl alcohol 1. 29 2nd l: l -65 0.95 2nd 1:1 -15 II II II 0. 44 2nd 1:1 -25 II II II o. 19 2nd 1:1 II II -35 2.44 2nd l : l -45 II II II i-propyl alcohol 1. 32 2nd 1:1 -55 11. 3 11. 3 11. 3 9. 6 9. 6 9.6 5.2 5. 2 5.2 9. 9 9.9 9.9 6.6 6.6 6.6 0. 61 2nd 1:1 -65 II toluene p-methoxy perbenzoic E kcal/mole k** Rate* Law Temperature II Solvent Acid - Continued II p-p' dichlorobenzyl sulfide Sul fur Compound Table 3. 1 9. 5979 9.5979 9. 5979 8. 2516 8.2516 8. 2516 5. 646 5. 646 5.646 8.382 8. 382 8.382 6. 6881 6. 6881 6. 6881 0-1 0-1 0-1 0-1 0-1 0-1 0-1 0-1 0-1 0-1 0-1 0-1 0-1 0-1 0- 1 l og 10 A Ref. I I w ....., II 3 methoxythionaphthene sul foxi de 3 methylthionaphthene sulfoxide thi anaphthene sul foxi de II 3 methoxyth i onaphthene 3 methylthi onaphthene th i onaphthene II dioxane-H 20 II II II II II 25 30 30 30 25 dioxane-H 2o CH 2c1 2 30 0.0029 0.050 0.032 0.08 0.004 0.0057 0. 112 30 l.60 0.88 0.61 2nd 1:1 2nd 1 : 1 0.43 30 -35 -45 -55 2nd 1: 1 6. 9 6. 9 6.9 9.6 o. 89 2nd 1:1 -25 9.6 0.45 2nd 1: l -35 9.6 kcal/mole E 2nd 1': 1 o. 17 k** -45 oc Rate* Law - Continued Temperature II II II II II II CHll2 II II perbenzoic II II II II i-propyl alcohol · So 1vent II p-nitroperbenzoic II II II II p-ch 1oroperbenzoi c Acid II p,p' dichlorobenzyl sulfide Sulfur Compound Tab 1e 3. l 6.5578 6.5578 6. 5578 8.4687 8. 4687 8. 4687 log *10 A G-l K-1 K-1 K-1 G-1 K-1 K-1 K-1 0-1 0-1 0-1 0-1 0-1 0-1 Ref. I w I ex, II II II di oxane-H 20 Sol vent 25 25 0. 0037 0.0396 0.0032 25 k* 5.0 Rate** Law 25 Temperature oc - Continued E log* A kcal/mole lO For second order: lit/mole/sec. + From different supplier. Used as received. ** For first order: sec -l * Overall order, then order with respect to the sulfide and the oxidizer, respectively. II II di benzothi ophene dibenzothiophene s ul foxi de II perbenzoi c Acid di phenyl sulfoxi de diphenyl sulfide Sulfur Compound Table 3.1 G- l G-l G- l G- l Ref. I \.0 I w II II II II II II cyclohexene hydroperoxi de II II II II t-butyl hydroperoxide II cyclohexyl methyl sulfide 50 50 methanol ethylene glycol ethylene glycol methanol 50 50 50 50 n-butano 1 t-butano 1 50 t-butanol 2.2 x l0 -4 14 . l 2. 3 X lQ -3 12.7 l.l9 x l0 -4 15.1 l.65 x lo- 3 12 . 6 1. 27 X 1Q- 2 11. Q 2nd 1:1 2nd l: l 2nd 1:1 2nd 1:1 2nd l:l 2nd l:l 13 . 04 13.04 13 . 04 13.04 l . 43 x 10 - 5 17 . 5 l.71 x 10- 4 15 . 0 6.2 6. 2 6.2 6. 2 o. 13 8.078 8.078 8. 078 8. 078 B-4 B-4 B-4 B-4 B-4 B-4 B- 4 H-1 H-1 H-1 H-1 H-1 H- l F-1 E log* A Ref . kcal/mole lO 2nd 1:1 1st l :O 100 II II II 1st l :0 85 II II 1st 1:0 75 II 1st 1:0 . 50 1st 1:0 white oil, HtO, and acetic a id II k* 2nd 1:1 5 X 10- 6 lstl :O 9. 4 x l0 -2 l 00 100 40 Temperature Rate** oc Law II II H202 H202 acet i c acid H20,acetic acid, and white oil Solve nt II di benzoth i ophene sul foxi de dibenzothiophene H202 t-butyl hydroperoxide dibenzothiophene II Peroxide Sulfur Compound Table 3.2 Rate Constants for the Oxidation of Sulfur Compounds by Hydroperoxides and Peroxides I I .p. 0 Continued 11 II II II II II II II t-butyl peroxide II II cyclohexyl methyl sulfide II II II propi onitri l di isopropyl ether t-butano l ethyl alcohol ace ton it ri l II i-propyl alcohol 2nd 1:1 49.8 50 49.8 49.8 49.8 2nd 1:1 2nd l : l 2nd l : l 2nd l : 1 2nd l: l 2nd l: l 39.8 49.8 2nd 1:1 2, 3,0 . 9:l : 4 2,4,5,0:45:2 2.4 0:4 : 2 2, 7 0:7:2 30 ..2 50 50 cyclohexane benzene 50 cyclohexane II + air 50 II H202 II II p,p' dichlorobenzyl sulfide II II II t-butyl hydro peroxide benzene 7. 94 76 0-2 7. 6 x lO -5 17 . 2 l.9l xl0 -4 17 . 2 2. 02 xl0 -4 H-2 0-2 1.6 x l0- 5 9. 4 x l0 -6 0-2 0-2 0-2 l.8 x l0 -5 0. 7 X lQ -5 17 . 7 0-2 7. 9476 3.1 x l0 -5 17. 2 7. 9476 0-2 B-5 B- 5 B- 5 B-5 B-5 B-5 13. 4 6.5 8.5 2, 9 0. 9:2 50 II cyclohexene II 10.9 3rd 1:2 50 benzene E log* A Ref . 1O kcal /mole cyclohexene hydroperoxi de k* cyclohexyl methyl sulfide Rate** Law Sol vent Peroxide Temperature oc - Sul fur Compound Table 3.2 I I ..... +=- p-methoxy phenylene methyl sulfide p-methyl phenylene methyl s ul fide m-methyl phenyl ene methyl sulfide phenyl methyl sulfide p-chlorophenylene methyl sulfide phenyl benzyl sulfide II II II II II II II II II II II 94% c H 0H; 6% 2 5 H 0; 0.lN HCl0 2 4 methanol II II II II H202 diphenyl sulfide m-chlorophenylene methyl sulfide H202 t-butyl peroxide methanol t-butanol cyclohexyl peroxide II Solvent Peroxide p-n it rophenyl ene methyl II II cyclohexyl methyl sulfide Sul fur Compound - 25 25 25 25 25 25 25 25 25 50 50 50 oc 2nd 1: 1 2nd 1:1 2nd 1:1 2nd 1:1 2nd 1: l 2nd 1:1 2nd 1: l 2nd 1:1 2nd 1:1 2nd l : 1 2nd l: l 2nd l: l Rate** Law Continued Temperature Table 3 . 2 ,o- 3 7, 84 X 1Q- 3 6.34 X lQ- 3 4, 38 X lQ- 3 3, 73 X lQ- 3 2.27 X lQ- 3 2 . 04 x 13.2 13. 5 13 . 5 14 . 4 14.4 3 . 8 X lQ- 4 1. 41 X lQ- 4 14.4 X 14 . 4 lQ- 4 3. 2 o- 4 13. l l .0l x 10- 3 1. 58 x 1 15. 5 7. 253 M- 1 M-1 M-1 M-1 M-1 M-1 M-1 M-1 M-1 H-2 H-2 H-2 E 1og* A Ref. kcal/mole lO 7, 9 X lQ- 5 k* I I N ~ II II II II II II II II H202 II II II II t-butyl hydroperoxi de II II acetic acid II 25 D 0 + HC10 2 4 dioxane 25 45.9 25 25 25 H20 + HC10 4 020 2nd l : l 45.9 II 2nd l: l 3rd l: 2 3rd l: 2 2nd 1:1 2nd 1: 1 2nd l : l 2nd l: l 25 H20 D-2 • 2. 43 10-2 D-2 10-6 0-2 D-2 l o- 3 16.3 -6. 77 0-2 10-5 D-2 D-2 14. l l o- 4 D-2 10-3 14. l D-2 D-2 0-2 D-2 2.46 10-6 4.07 2.03 9.6 l. 91 6.42 l. 35 l. 54 4.5 l o- 4 13. 5 ,o-3 2nd l: l 25 II II II 13. 5 10-4 2nd l : l 9.7 D20 II II 13.5 10-3 5.32 2nd 1:1 34 . 2 II II II 13. 5 2.58 ,o-3 2nd l: l 25 II II II 13 . 5 0-2 M-1 M-1 7.253 l o- 2 13.2 10-4 2.03 ,o-3 E log* A Ref. kcal/mole lO 3.2 2nd l: l 25 3.73 k* 2nd 1:1 2nd l : l Rate** Law 25 oc Temperature - Contin ued 0 H0 2 II 94% c2H50H; 6% H20; O. lN HCl04 Solvent II H202 Peroxide II benzyl methyl sulfide di-n-butyl sulfide thi oxane Sulfur Compound Table 3.2 I ~ I w 2nd 1: l 19.3 1. 87 X 1Q-5 -4. 808 -4. 808 -5.901 -6. 120 -6. 120 -6. 120 -6.338 -6.338 -4.808 -4. 808 D-2 D-2 D-2 D-2 D-2 D-2 D-2 D-2 D-2 D-2 logi 0A Ref. For second order reaction mole/lit/sec. ** First the overall rate dependence, then the dependence on the sulfur compound and the peroxide, respectively. * For first order reaction sec -1 . 45.9 II II II 19.3 X ,o- 6 2nd 1 : 1 2.2 25 II II 13. 5 2. 58 x l 0-3 2nd 1:1 25 H20 N-methyl acetami de II 13.5 II 2nd l: l 25 D20 II II 14. 2 14.2 19.3 2.38 X 10 -3 1. 54 X 10 -3 2nd l: l 49.5 II II II 2nd l:l ethylene glycol 25 II II l 5. l 3. 27 x l 0-4 4.96xl0 -4 2nd l : l 45.9 II II II 15 .1 6.16 X lQ- 5 2nd l : l 25 methanol 2nd 1 : l 45.9 II 19.3 9. 45 x l 0-6 4.0 X lQ -5 2nd 1:1 25 E kcal/mole k* Rate** Law Temperature co II II t-butanol Solvent II H202 Peroxide - Continued II Thioxane Sul fur Compound Table 3.2 I I ~ ~ -45- l. Sulfur compounds are oxidized by a cyclic peroxy complex which contains a hydr.operoxide molecule and a protic molecule (HA). t R-0--0--+! + HA + s . . . . RI + '- RI ( 3. l ) ROH + HA + OSR21 (3.2) This mechanism was first proposed by Overberger and Cummins (0-1 and 0-2). Many other studies confirmed its validity, e.g. (B-2), (B-3). If excess protic solvent is available in the solution, then the solvent will form a complex with the hydroperoxide molecule, e.g., in alcoholic media: RO-O-H + R 0H 1 ~ R--0 :\ H o I I (3.3) RQ .. H In aprotic solvent the hydroperoxide will dimerize: 2ROOH "t- The chemistry of hydroperoxides was reviewed by Hiatt(H-3). (3.4) Peracids may either dimerize, (3.5) -46- or form an intramolecular complex, (3.6) The chemistry of peracids was reviewed by Swern (S-1). The mechanism of oxidation of sulfur compounds by peracids was discussed by Modena and Miola (M-1). 2. The solvent determines the rate equation as follows: A) In protic solvent the rate equation will be first order in each of the peroxide and the sulfur compounds, since the complex which actually oxidizes the sulfur compound contains only one peroxide molecule and one solvent molecule (B-5). B) In aprotic solvent, the rate equa- tion will be second order in the peroxide, and first order in the sulfur compound. The second order dependence on the peroxide indicates that the peroxide dimer is the specie which actually does the oxidation {B-5, B-6). C) The rate equation will usually be first order with respect to the peroxy acid, since the function of the cyclic complex can be performed by the intramolecular complex. 3. In this study, it has been observed that oxidation of sulfur compounds in a nonprotic medium results in many more products than oxidation in a protic medium. A possible explanation is that the peroxidic dimer has a larger oxidation potential than complex of peroxides with nonprotic molecules. 4. The rate of oxidation of sulfur compounds, with sulfur in the a-position to n electrons, is a few orders of magnitude smaller than -47that of compounds in which strong n-n electron interaction is impossible (F-1, B-4, B-5). 5. Acids catalyze the rate of oxidation but salts do not affect it (D-2, C-1). 6. Peroxides attack the sulfur atom faster than they attack ole- finic bonds (B-6, B-7, B-8, B-9, B-10, H-2). 7. At temperatures below 70°C the rate of oxidation of sulfur compounds does not change due to the presence of air or of free radical inhibitors (B-4, B-5, 0-1, 0-2) . Such behavior is consistent with the assumed mechanism, which does not include free radicals. However, free radicals react with the sulfur at higher temperatures (see part 11). 8. The protic nature of the solvent affects the free energy of the oxidation both by altering the mechanism of oxidation and by interacting with the sulfoxide (0-3) . Since sulfoxides are weak bases they can form complexes with protic solvent. The complex has low free energy and thus causes the rate of oxidation of sulfides to sulfoxides to increase and of sulfoxides to sulfones to decrease: [OJ -+ AH R [ "-- R/ R SOHA] -:_ 'so + HA R'. . (3. 7) (3.8) Reaction (3.8) shows that protic solvent affects then-electrons of the sulfur in a similar way to an bond in a position. The reduction in the availability of electrons on the sulfur causes the rate of oxidation to decrease. -48- 9. The rate of oxidation of aliphatic sulfides to the corres- ponding sulfoxides is much larger than the rate of oxidation of the sulfoxides to the sulfones. Stronger oxidizers are required to oxidize the sulfoxides to the sulfones than to oxidize the sulfides to the sulfoxide (S-2, H-4, H-5). The rate of oxidation of aryl sulfides and condensed thiophenes to the corresponding sulfoxides and sulfones is much smaller than that for aliphatic sulfides. The ratio of the rates of oxidation of aryl sulfoxides to sulfones and that of oxidation of the aryl sulfides to the sulfoxides is about one, but occasionally it is greater than one (G-1 ). 10. Sulfoxides were found to inhibit the oxidation of hydrocar- bons (0-4, C-3, B-11) and to decompose peroxides. Inhibition can be due to several processes; the important ones are: A. oxidation of the sulfoxide to sulfone by the peroxides: ( 3. 9) The consumption of peroxides by this reaction reduces the rate of branching and therefore the rate of oxidation. B. A reaction between decomposition products of the sulfoxide and oxy- gen or peroxides (B-11): CR 1 R11 II CH SR 111 + [ 2 I J (3.10) OH II I l2 R SOSR 111 II 111 +l H0 2 2 -49The sulfenic acid (I) dimerizes to the thiolsulphinate II, which then reacts rapidly with oxidizers to form thiolsulphonates III (C-2): R111 SOSR + RO 2H ➔ R111 SO 2SR + ROH· 111 111 ( 3. 11) The olefinic fragment (III) may react with oxygen or peroxide and cause additional inhibition. 11. Peroxy radicals at temperatures above 7O°C may oxidize sulfur compounds fn two ways: A. By direct radical attack on the n electrons of the sulfur: (3.12) B. By attack on the a or S hydrogens: H Ro • + 2 l-s- ➔ Ro 2H + -t-s- (3.13) o) The presence of the sulfur greatly stabilizes radical I. Radical I absorbs oxygen very rapidly and reacts in a similar way to other organic radicals (see Chapter 2). The result is the production of a and S hydroxy-sulfides and a and 8 keto-sul fi des. Sul foxi des react with peroxy radi ca 1s in the same way to produce a and S hydroxy-sul foxi des and a and S keto-sulfoxides. It is believed that the a hydroxy sul- foxide is the most stable specie because of the possibility of intramolecular cylic complex, II, which stabilizes the molecule via hydrogen bonding: -50OH 0 I I II 0-H ,, --c O ' 0 --c-s- ~ (3.14) 's;.,-1 12. a-hydroxy sulfides decompose to the corresponding mercaptan (I) and aldehyde (II) [B-3J : OH I I R'--C-S-R ➔ 0 HS-R + R'--C~ '-H (3.15) H II I Further oxidation will convert the aldehyde to the acid, and the mercaptan will react according to the following sequence: I I II [OJ+ RSH ➔ RSSR + H 0 2 [OJ _. ! ~ RSSR 0 [OJ -l➔ IV II II () [OJ+ RSSR V ➔ RS0 H + [RSOHJ 3 VI (3.16) VII The disulfide (III) is ea~ily oxidized to the thiasulfoxide IV. IV is oxidized to the thiasulfone V; V is oxidized to the sulfonic acid VI and to the sulfinic acid VII. VII is unstable, and will either rearrange and decompose or absorb a molecule of oxygen and become sulfonic acid (C-2). -51- 13. The reactions of HBr with sulfides and sulfoxides are complex and depend significantly on the temperature and on the solvent. The following is a list of reactions that were quoted in the literature for a system which contains aliphatic sulfides, sulfoxides, HBr and 1. Formation of sulfonium ion (R-4): (3.17) 2. Cleavage of the C-S bond (very slow (R-3): (3.18) Note that Br is incorporated into the organic structure. 3. Adduct formation with Br 2 (R-3): (3.19) In the presence of water, hydrolysis occurs and sulfoxide is fonned: ?i \ RSBr R + H20 t 2 RSR + 2HBr (3.20) The kinetics of this reaction were studied in detail by Miotti et al. (M-4) who examined several aryl-methyl sulfides at 25°C and in a mixed organic-aqueous medium. Under these conditions they state that the reaction is "practically irreversible". reversible (H-8). The reaction with iodine is However, several authors have used the reverse -52- reaction as a method to determine sulfoxide by titrating the bromine which is liberated (A-8, p. 33) (3.21) but reduction with HI was generally performed as an analytical method. The conclusion is that at room temperature Br 2 would probably oxidize sulfides but at higher temperatures the reverse reaction may become more important. A particularly interesting case is the behavior in a system which contains water dispersed in aprotic solvent. The solubility of HBr in water is very large, and it tends to be extracted to the aqueous phase. Sulfoxides are surface-active materials and concentrate on the interphase between the water and the organic medium. Since the sulfides are soluble essentially only in the organic medium and have no surface activity, they will remain evenly distributed in the bulk of the organic solvent. Since large concentrations of sulfoxide and HBr are realized on the boundary of the two phases, reaction (13.21) will dominate and the sulfoxide will be reduced back to the sulfide. -53- Chapter 4 THERMODYNAMIC ANALYSIS AND THE LOSS OF HEATING VALUE DUE TO OXIDATION Two kinds of thermodynamic analysis were made : l. Calculations of the equilibrium composition that the system will try to reach. 2. Calculation of the enthalpy losses due to oxidation. The results of the calculations of the equilibrium concentrations have a limited value because under the conditions of the reaction, the system was far from equilibrium . However, conclusions may be derived on the general trends of the reactions The method for calculating the equilibrium composition was detailed by Attar (A-6, p. 20); so only a brief discussion of the results will be given. A general method for calculating enthalpy changes is discussed first and then applied to the calculation of the heat loss due to the co-oxidation of Tetralin and butyl sulfide. The actual use of the method is demonstrated in Chapter 7, where the loss of heating value as a function of the reaction time and of the sulfide conversion was calculated and plotted. 4. l Calculation of the Equilibrium Concentration A variation of the steepest-descent method of White et al (W-6) was used. A detailed description of the method and of the modifica- tions that were made was given in a previous report (A-6). results are: The main -541. In the presence of limited amounts of oxygen, thE:! large organic molecules tend to crack to carbon and hydrogen. 2. The oxygen ends essentially as water, although small amounts of CO and CO 2 are also formed. 3. The sulfur ends as H2S. 4. The HBr does not change. Small perturbations in the values of the free energy or of the temperature did not affect the equilibrium composition. At temperatures higher than 500°C, organic reactants may reach equilibrium; indeed it is known that H2, C, and H2S are formed when hydrocarbons are heated or pyrolized unaerobically. However, at tem- peratures below 200°C, the kinetics of the reactions controls. 4.2 The Trade-Off of Energy vs. Desulfurization The penalty for the conversion of the sulfur compounds to the corresponding sulfoxides and sulfones is the loss of heating value of the fuel due to its oxidation. We therefore want to know how much heat- ing value is lost for a given conversion of the sulfur compound. The product distribution is controlled by the kinetics of the reaction, but since the enthalpy of the system depends on its composition, once the composition is known, it is possible to evaluate the enthalpy loss for a given conversion. The method is first discussed for a general system and then applied to the analysis of the oxidation of sulfur compound in a hydrocarbon medium. -55- 4.2. l A Method for the Evaluation of Enthalpy Changes due to a Complex Reaction The classical method for the evaluation of enthalpy changes of a chemical system requires detailed knowledge of the chemical composition and the corresponding thermodynamic data. Such information is difficult to derive for a complex mixture and therefore is rarely available. A very good estimate of the enthalpy change can be ob- tained, however, if the chemical functional group distribution (FGD) of the system is known. The FGD can be obtained experimentally much more easily than the detailed composition of the system. Furthermore, thermodynamic data for specific compounds are not required. The following assumptions are used in the derivation : system is isothermic and isobaric. are ideal mixtures. l) The 2) The reactants and the products 3) The enthalpy of a molecule is the sum of the enthalpies of the functional groups that make it. Let there be in the system m functional groups and let the jth group appear n . . times in the ;th component. Jl Denote the specific enthalpy of the jth group h. , then the enthalpy of the molecule, H., J l. will be m I h.J nJl.. ( 4. l) j=l If the concentration of the i th component is c.l ' then the enthalpy of the system per unit volume, HS ' is m nm n m n = = h .n . . C. = = c. I h .n . . l C.H. I l . J J l Ii Ij . J Jl l Ij h.c. HS . l l J J l i J (4.2) -56- c. denotes the concentration of the jth functional group in the sysJ tern. When the concentration vector changes, the enthalpy will change by 6HR. When all the concentrations and specific enthalpies are known the change in the enthalpy can be calculated from: .6HR = Ii C1• Ij h .n. •Ifinal - Ii C.1 Ij h .n .. initial J Jl J Jl n = I(c.f. . . 1 )H., i 1 1na 1 -c 1. . 1n1t1a But when only the FGD is known, m L'I HR = I h. C. j J J final ( 4. 3) 6HR can be calculated from m I hJ.c.J j initial Ij hJ.(c.J final c. initial) = m = J (4.4) The two methods are equivalent provided that the group additivity rule for the enthalpy of formation holds. The proposed method has two major advantages: 1. The functional group distribution can be determined experimentally much more readi.ly than the exact composition of numerous components. 2. Only the m functional groups that actually change need to be considered. 4.2.2 Application to a Reacting System Consider a system in which a chemical reaction takes place. Denote the chemical composition vector C(t) (n elements), and the FGD vector -57- c(t) ( m elements). If the system is modeled by the set of differ- ential equations ( 4. 5) then by integrating the initial value problem (4.5), one obtains c(t). c(t) is uniquely determined by from it. c(t) and can be readily calculated Substitution of c(t) in equations (4.4) give: t-HR(t) m = ~ hj(cj(t) - cj(o)) = ~. (I(t) c(o)) (4.6) 6HR can be plotted vs. the time and vs. the conversion of a particular component, vs. the production of a particular functional group, or vs. any other desirable parameter. tem results. A diagram for the energy of the sys- In Chapter 7 such graphs are presented in which the enthalpy loss due to the oxidation was plotted vs. the reaction time and vs. the sulfide conversion. 4.2.3 Case Study: The Enthalpy Loss in the Co-oxidation of Tetralin and Dibutyl-Sulfide 1. Definition of the Problem: Derive an algorithm to evaluate the loss of heating value of a mixture of Tetralin (TRLN) and dibutyl sulfide (DBS) due to oxidation. 2. Definition of the Initial Condition: Let the initial condition of the system be: 5.059 mole/liter TRLN and 0.1424 mole/liter DBS. Let the reaction be isothermal at 140°C and the oxygen concentration of 5.34 • 10- 3mole/1 iter. -58- 3. Definition of the Course of the Oxidation Let us assume that a model with N reactions can describe the sys tem. The major products of the reaction according to the model are: tetratone (CO), tetralol (ROH), l ,2-dihydronaphthalene (REN), tetralyl hydroperoxide (R0 2H), tetralin di-ketone (DICO), water (H 20), hydrogen peroxide (H 202), butyl sulfoxide (SO), butyl sulfone (S0 2), a-ketobutyl sulfoxide (COSO), a-ketobutyl sulfide (COS) and a-hydroxy butyl sulfoxide (SOOH). Note that the carbon structure of neither the TRLN nor the OBS was broken. 4. Analysis of the Functional Group Distribution and the Enthalpy Loss The analysis can be simplified if the molecules that did not react would be neglected. l The following subscripts are introduced: for TRLN, 2 for CO, 3 for ROH, 4 for REN, 5 for R0 2H, 6 for DICO, 7 for H 0, 8 for H o , 10 for DBS (or S), 11 for SO, 12 for so 2 , 13 22 2 for COSO, 14 for COS, and 15 for SOOH. Since the solutions are assumed to be ideal, their volumes do not change due to the reaction. The decrease in the concentration of TRLN is therefore given by 6 1:::.c 1 = I c.l ( 4. 7) 2 The decrease in the concentration of the sulfide is ( 4. 8) Only the groups that change in the reaction need be considered in the calculation. For example, one molecule of Tetralin (I) reacted to produce one molecule of tetralone. The net change in enthalpy is -59 - the enthalpy of a carbonyl group minus the enthalpy of a methylene group . 00 -----► .. oo 0 II I Note that the mechanism of the reactions does not influence the thermodynamic calculation. Only the final products do . The nomenclature of reference (B-14) is used, the data for the enthalpies of formation were taken from the tables in reference (B-14, p. 178-215) . The results are summarized in Table 4.1. Only the FG that changed were con- sidered, so In the case of Tetralin or its derivatives, only the carbons at positions l, 2, and 4 were considered. In the case of butyl sulfide and its derivatives, only the sulfur and the two a carbon atoms were considered. The sum of the enthalpies of formation for the groups that changed was denoted t I h.n ... j t J Jl The difference between Ij hJ.n Jl.. and its value for the original hydrocarbon was denoted .6.H 1. . To calculate the heat loss at any desired composition, equations (4.3) or (4 . 4) were used . Plots of the loss of heating value as a function of time and sulfide conversion are given in Chapter 7. 2 l 2 l 7. 46 9. 18 -36.9 -4.07 -15.6 7. 72 -13.25 -7.83 -68.42 -30.91 Cd-Cb,H Cd-C,H CO-C,Cb 0-0,C 0-0,H *c-c, (SO) ,H2 SO-C 2 C-C(S0 2)H 2 SOz-Cz C0-5,C l -37.4 0-C,H l l -8.90 C-Cb,0,C,H l l l l l l 2 l l 1· -4. 13 l S0 2 C-(C) 2 (H) 2 l so l DICO l R0 2H -3 . 96 REN C-Cb(H) 2C ROH co Enthalpy of Formation Group l l coso 1 cos l l l SOOH l 2 RH TRLN Table 4.1 The Change in the Functional Groups in the Co-Oxidation of Tetralin and Butyl Sulfide at 140°C and the Enthalpy of Formation of the Oxidation Products (Nomenclature and data taken from Ref. (B-14) and Tab le 7. l ) s BUS I I O"> 0 -4.81 12.57 -56.90 -31.30 C-C(H) 2s S-C 2 HzD . H202 R02H DICO · so so2 coso cos l SOOH RH TRLN S 2 BUS 0 0 12.68 -36.66 -77.93 2. 19 -84.08 -23.92 -23.15 0.11 -12.05 2.95 REN -32.94 -42.34 24.73 -24.61 -65.88 -0.76 -87.03 -26.87 -26.10 -2 . 84 -44.99 -54.39 ROH Estimated by the author. **6H; is the change in the enthalpy due to the formation of one molecule of oxidation product. **6H/ca l /mole J tI. h1.n 1..J -18.39 ·s . 64 Ehthal PY of Fonnat.i on CO *co-c,so *c-(OH),(SO )~C,H Group Table 4.1 (continued) 7' I O'l -62- Chapter 5 EXPERIMENTAL The experimental work had three goals: l) to establish whether catalysis of the hydrogen-abstraction step enhances the oxidation of the sulfur compounds, 2) to find the type of sulfur compounds which are produced and to infer from it the oxidation mechanism, and 3) to find limited amounts of kinetics data for selected systems. The work was exploratory. Therefore the qualitative answers were regarded as important as the quantitative ones. In the discussion of the experimental work there are four parts: l) An elaboration on the choice of the experimental system. 2) Description of the chemical reactor and the experimental procedure. 3) Description of the analytical instruments and analytical procedures. 4) Estimation of the possible mass transfer in the reactor. 5. l On the Choice of the Experimental System Choice of the experimental system involved type of reactor and analytical methods, the chemical system, and the range of parameters to be examined. 5.1. l The Reactor A batch reactor was chosen because long reaction periods were anticipated and because it can be easily designed to minimize the surface/volume ratio and thus minimize surface effects. In practice, oxygen and HBr were bubbled through the liquid solution of the hydrocarbon; however, because the rate of oxidation was rather slow, the liquid assumed rapid equilibrium with the gas and therefore - 63- modeling of the system as a batch reactor holds . (See also Section 5. 4). 5.1 .2 The Chemical System Two approaches to the choice of the chemi cal system were possible: 1) work on oil fraction; 2) work on model sulfu r compounds in a medium which simulates oil. Work on oil would provide information on the oxidation of a real fuel; however, it would have only limited value in the understanding of the mechanism of the reaction and to extrapolation of the data beyond the region that was empirically tested . Enormous analytical difficul- ties were anticipated for work with oil. Work with mode 1 compounds a 11 owed the determination of the mechanism of reaction of a single sulfur compound or of a homolog group. The analytical problemswere simpler and the system was much better defined. Extrapolation of the work on model sulfur compounds to actual fuel, however, might be difficult. The latter approach was chosen because it is more basic and may provide information on the mechanism of a class of reactions. The choice of four chemical components was required: the sulfur compound , the hydrocarbon carrier, the solvent, and the catalyst. 5.1 .2.1 The Choice of Sulfur Compound Di-n-butyl sulfide (DBS) and di-benzothiophene (DBT) were chosen for the preliminary study . The criteria for the choice were: 1. DBT and DBS resemble chemically the largest two groups of sulfur compounds that were found in gas oil and heavy crudes. -64- 2. Limited amounts of data were available on the rate of oxida- tion of DBT and DBS by organic peroxides . 3. DBT and DBS as well as some of their oxidation products are available conmercially. 4. DBT and DBS have a symmetric structure. A smaller number of products are possible with symmetric structures and therefore easier analytical problems were anticipated . The two major differences between the behavior toward oxidation of DBS and DBT are: that the sulfur in.DBS is connected to two electron releasing groups which tend to increase its nucleophility . in DBT is connected to two aromatic rings. The TT The sulfur electrons resonate with then-electrons of the sulfur and cause a reduction in the electron density. The oxidation potential which is required to oxidize DBT is much larger than that for DBS (E 112 (DBT) =-2.432 Vin DMF vs. Ag/AgCl electrode (G-2); E (DBS) = 0.77 Vin 0.1 N HCl in CH 30H vs. 112 Ag/AgCl electrode (D-5). The second major difference is that DBS contains aliphatic hydrogens, which can be abstracted much more readily than the aromatic hydrogens of DBT. As a result, one may expect the selectivity to be larger when DBT is oxidized compared with that when DBS is oxidized. 5.1.2.2 The Choice of the Hydrocarbon Medium and Solvent The hydrocarbon medium serves two functions: It is a source· of free radicals, R,and is a solvent for the reagents and products. In order to be able to infer from the results the behavior of an oil, a hydrocarbon with a benzene ring was desired, because most of the -65molecules of a heavy oil contain at least one. No clear criterion is available by which to evaluate the adequacy of alternative hydrocarbons to simulate fuel. The criterion that was adopted was the relative rate of hydrogen abstraction, beca~se the hydrogen abstraction was the rate limiting step of the oxidation for hydrocarbons. Two hydrocarbons were chosen with extreme reactivities toward hydrogen abstraction. The idea was that the reactivity of fuel will be an intermediate value between the two . The two hydrocarbons that were used were toluene and Tetralin; and the relative reactivity of the two toward hydrogen abstraction is 1:300 (H-6). The reactivity of all the common hydrocarbons is between the reactivity of these two. The properties of the hydrocarbons were modified by adding solvents with different proticity, e.g., bromobenzene (non-protic) and 1-octanol (protic). Bromobenzene and 0-dichlorobenzene are inert to oxi- dation below 200°C, because they do not have active hydrogens. The reactivity of the solvent will be seen when the hydrocarbon is non-reactive, as with toluene. However, if the hydrocarbon is much more reactive than the solvent, the reactions of the solvent will be ·of marginal importance. 5. 1.2.3 The Choice of the Catalyst The properties of the desired catalyst are reactivity toward hydrogen abstraction and specificity to the hydrogen abstraction step. In order to eliminate complications from mass transfer and sur- face effects, a homogeneous catalyst was preferable. Because information on the reactivity and specificity of HBr was available (see Section 2.5 ), -66- it was chosen to start with. HBr is a reactive catalyst for hydrogen abstraction; however, it also catalyzes the decomposition of peroxides, produces olefin by dehydration of alcohols, and isomerizes alkyl benzenes. The Br radicals form Br 2 which is even more reactive than HBr and further complicates the system . HBr can reduce sulfoxides to bi- valent sulfur, and in this way counteract the goal of obtaining oxidized sulfur compounds. The dangerous properties of HBr and its high cor- rosivity complicated the work and imposed safety constraints on the operation. HBr proved a useful catalyst to demonstrate the principle of the thesis; however, a more specific one is recommended for future studies. 6.2 The Reactor and the Experimental Procedure 5.2 . 1 Mechanical Design of the Experimental System Figure 5.1 is a flow diagram of the experimental system. A de- tailed description is given only for the non-standard items. The reactor was constructed from a standard 500 ml 3-necked pyrex flask to which two additional 24/40 ground glass points were symmetrically connected. Figure 5.2 shows the reactor. The central joint was a ground joint 34/40 into which a pyrex stirrer with Teflon blades was connected (Corning type 28-52). The other four necks were used for 1) a liquid sampling device (Fig. 5.3), 2) gas bubbler composed of a standard 20 rm, porous sintered glass type Mthrough which the oxygencatalyst mixture was dispersed, 3) a thermowell in which a thermistor (YSI Part No. 44004) and a copper- constantan 24G thennocouple were inserted, 4) a reflax condenser . The gas that came from the condenser was passed through a gas-sampling valve and vented. Gas samples were analyzed by GC for selected experiments . F-2 I I § t/ @] TPC1 - T Rsl 1--7-, Table 5. l . Figure 5. l Schematic Flow Diagram of the Experimental System. N2 C-2 02 IvAR~ - - 1Mo T- • • The nomenclature is detailed in GIC GCV I 0-, I '-J -68Table 5. l List of Major Parts Designation Function Material BB Gas bubbler Sintered Glass C-1 Oxygen cylinder C-2 Nitrogen cylinder C-3 HBr cylinder cw Coding water l .i ne tygon F-1 Oxygen flowmeter SS316 F-2 Nitrogen flowmeter Brass F-3 HB r fl owmeter Monel GC Gas chromatograph GCV Gas chromatograph sampling valve HM Heating mantel Glass Col 500W, 500 ml MOT Motor l /10 HP PI-1 Oxygen pressure indicator Matheson PN-63-3115 PI-2 Nitrogen pressure indicator Matheson PN-63-3115 PI-3 HBr pressure indicator R-1 Oxygen pressure regulator Matheson 8-350 R-2 Nitrogen pressure regulator Matheson 8-850 R-3 HBr pressure regulator Matheson B15-330 RCT Reactor Remarks Medium porosity 20 mm¢ See Figure 5.4 Carl, Basic SS316 Monel Pyrex Carl , 8 ports Matheson PN-63-3312 500 ml - 69- Table 5. 1 (cont'd) Designation Remarks Function Pyrex Pyrex, Teflon See Figure 5. 3 Glass, Corning Model 7272 SMP Sampling device STR Stirrer TI Temperature indicator TS Temperature sensing probe TR Temperature regulator TRC Thermocouple (ion-constantan) TRS Thennistor V-1 o2 Regulator valve Brass V-2 Brass V-3 N2 Regulator valve HBr Regulator valve V-4 HBr Vent valve Monel Swagelok, M-1GS4 V-5 o2 Vent toggle valve Brass Swagelok, B-OG52 V-6 N Vent toggle valve 2 Brass Swagelok, B-OG52 V-7 o2 Flow toggle valve Brass Swagelok, B-OG52 V- 8 N2 Flow toggle valve Outlet gas back pressure valve Brass Swagelok, B-OG52 Monel Swagelok, M-2JR WT-1 Water trap Brass WT-2 Water trap Brass V-9 Bailey Instr. Model BAT-7 Glass Versa-Term, Model 2156 YSP, Model 44004 Brass -70- SIDE VIEW Thermowell,......... TOP VIEW Material: Pyrex Uni ts: mm Figure 5.2. Reactor for Liquid Phase Oxidation of Hydrocarbons - 71- Rubber pipet filler - - Container for cooling mixture 3-way Teflon stopcock, 3 mm~- - Sample bottle - Ground joint 24/40 - Figure 5. 3 Sampling Device for Liquid Samples -72- Swagelok connection Packing.._ Aluminum pl ate - ----411111111■ _pressure gauge reducer -T -t:·.:J~: (]- -Swagelok connection Manomete ri c U tube -. . . . . "-.metering valve Figure 5•.4 A fixed-bed pressure drop volumetric fl owmeter. The manometeric pressure difference is proportional to the volumetric flow rate. Used to measure accurately flow rates of the order of 10 ml/hr. Leads A and Bare connected by a tube packed with glass beads 0.00095 11 through which the gas flows. -73- The reactor was heated with a Glass-Col 500 ml mantle and con- trolled with a Versa-therm controller, Model No. 2156. The tempera- ture was monitored with an electronic thermometer, Bailey Inst. Model Bat 7. Oxygen, hydrogen bromide and nitrogen were fed to the reactor from cylinders of compressed gas. standard gas regulators. The pressure was regulated with The rate of flow was monitored with a flow- meter which was specially designed for this purpose. one of the flowmeters. Figure 5.4 shows The rate of flow was measured by monitoring the pressure drop that it created across a fixed bed of small glass beads. The range of flow measurement was determined by the length of the bed of beads, the length of the manometer tubes, and the density of the fluid in them. 5.2.2 Experimental Procedure Each experiment was conducted as follows: l. All glass and Teflon parts which came in contact with the reacting solution were deactivated. (The deactivation procedure is described in Appendix B). This deactivation resulted in the etherification of the active OH groups on the glass surface. It was found neces- sary to repeat the treatment after every three experiments. 2. The reaction components were purified {Appendix B) and stored under nitrogen. 3. The reactants were mixed in the reactor, under nitrogen atmosphere and a sample of the cold mixture was taken. 4. The mixture was heated to the desired temperature while HBr and nitrogen at the desired rates were flowing through. The N2 rate -74was adjusted to be the same as the desired oxygen flow during the oxidation. 5. One to two hours after the mixture reached temperature, a sample of the hot liquid was withdrawn and the nitrogen flow was switched to oxygen at the same rate. The moment of turning on the oxygen was denoted time 0. 6. Samples were withdrawn every fifteen minutes in the first two hours, and every thirty minutes in the following three and a half hours. 7. The reaction was stopped after five and a half hours by turning off the heat, the oxygen, and the HBr flow, The reactor was cooled in an ice bath to about 60°C and the solution was vacuum pumped out into a specially designed cold container. 8. The UV absorbance of the samples was measured at the desired wavelengths immediately after the reaction, using the first sample as a reference. 9. The samples were analyzed qualitatively and quantitatively in the following days. 5.3 Methods of Chemical Analysis The discussion of the qualitative methods is separated from that of the quantitative ones. 5. 3. l Methods of Qualitative Analysis One of the biggest problems that has been encountered was the qualitative analysis of the reaction products. The difficulty was due to l) the complexity of the multicomponent mixtures involved 2) the -75availability of small and dilute samples only, 3) the low conversion that was used, and 4) the low sensitivity of the equipment used for qualitative analysis (e.g., IR and NMR). Traditionally, qualitative analysis is done using IR, NMR, MS and UV techniques. However, in order to utilize these techniques, a rela- tively large amount of a pure material is required. Separation and purification was done by chromatography; however, except for a few of the major components, it was impossible to isolate sufficient amounts of pure components to carry on a meaningful analysis. First, the non-standard methods are described,and then a list of standard techniques is given. 5.3.1.l IR Techniques IR analysis gives information on the functional groups that are in the sample. A compilation of the IR absorption bands of sulfur functional groups is given in Table 5.2 (L-1): Table 5.2 Group IR Absorption Bands for Sulfur Functional Groups Band, cm -1 Band, cm -1 Group -0-SO 2-0- 1230-1150, 1440-1350 -S=O 0 II -S--01+ 1060-1040 .J-QR 1135-1125 -S0 2-SR RG-S--OR 1215-1150 Aryl sulfides 732-710 -S0 2- 1160-1140, 1300-1350 Primary sulfides 660-630 -o-so 2- 1200-1145, 1350-1300 ~ 1090 so 3H, so 3 1210-1150, 1060-1030, and 650 1150-1110, 1350-1330 Secondary sulfides 630-600 Tertiary sulfides 600-570 -76- Perkin-Elmer IR equipment with two kinds of cells from Bames Engineering Company was used : l. NaCl cell, Model 0003-504 for water free samples. 2. BaF 2 cell, Model 0004-035 for moist samples. The BaF 2 windows allow only modest sensitivity because they absorb up to 70% of the light. They are not useful for wavenumbers smaller than 950 cm- l. 5.3. 1.2 GC-MS Computer GC is a powerful separation technique,and MS is an excellent identification method for organic compounds. When the two methods are com- bined,it is often possible to identify components of mixtures which cannot be done in any other way. The output from a GC column is fed into a mass spectrometer after most of the carrier gas is separated. MS scans are taken at the desired frequency,and the results are stored on a tape. At the end of the chromatogram, the data are retrieved from the tape and analyzed as required. In this study the method was used for the identification of three kinds of components: l) bromine containing compounds, 2) sulfur containing compounds and 3) hydrocarbon oxidation products. The method for the identification of Br compounds was developed in this study. The method of Ref. G-4 for identifying sulfur compounds was attempted. Identification of hydrocarbon products is a standard technique, and will not be detailed here. Identification of Bromine Compounds The identification of bromine compounds was based on two facts: • 1•s pres en t 1•n na t ure ,n .- t wo 1•so t opes, - Br 79 an d Br Sl , w,'th a l ) bromrne -77- relative abundance of 100:98, respectively; 2) organic bromine compounds form a Br+ ion. To identify a bromine compound, three signals are simultaneously monitored as a function of time : the total current of ions , the ion current due to m/e = 79 and that due to m/e = 81 . Whenever a peak was seen in the total ion current , the s i ze of the signals from the m/e = 79 and the m/e = 81 were compared. When their value was the same, it was assumed that that component contained bromine . Bromobenzene and its impurity , bromoethyl benzene were clearly identified by this method . Butyl bromide was also seen, but no tetralyl bromide was found . Only m/e below 250 was scanned. Identification of the Sulfur Compounds MS of bivalent sulfur compounds shows a peak at m/ e = 47, which corresponds to H2C = SH+ (S-6), and at m/e = 45, which corresponds to CHS+ (G-4). However, fragments with the former m/e are common to many organic compounds, and the abundance of the latter is rather small and not sufficiently specific unless a high-resolution spectrometer is used. The natural abundance of sulfur isotopes is s34 ;s 32 = 4.4/100, so if the total concentration of sulfur compound is small, the m/e = m/e + 2 peak cannot be seen. Tetra and hexavalent organosulfur compounds were in such small concentration, and so poorly resolved, that none were seen where they were expected . Obviously, the MS is less sensitive to sulfur than the FPD, or the oxidized compounds did not reach the ion source because they -78- adsorbed to a surface in the duct between the GC and the MS. The GC-MS was carried out on the home-made unit of the Chemistry Department, and on a Finnigan Model 3200El unit at the Jet Propulsion Laboratory. The standard techniques that were used for qualitative analysis are listed in Table 5.3. 5.3.2 Methods of Quantitative Analysis Most of the quantitative analysis was done by gas chromatography. The methods that were developed in this study are detailed in Table 5.4. Other methods that were used are liquid chromatography, polarography, titrametry and calorimetry. Standard determination of the heat of combustion was done using a Parr Adiabatic Calorimeter. The non-standard methods are described briefly in the following sections. 5.3.2. 1 Determination of H2o2 and Hydroperoxides by Polarography (POL-1) The method of Willits et al. (W-1) ~as used with a small modification. Platinum wire was used as an anode instead of a mercury pool. The solvent used was 0.3 LiCl in 50:50 c6H6: CH 30H. Al ml sample was diluted with nine ml solvent and the polarogram was recorded from Oto -2V. E for hydrogen peroxide is -0.2V and for tetralyl hydroperoxide 112 -0. 75V. Each mole/liter of the original sample creates 4.0064 • 10 3 µa current, both for the H2o2 and the tetralyl hydroperoxide. The methods of Drushel (D-1) were used for the analysis of sulfur compounds by polarography. Notation UV Ultraviolet spectrometry Carry Model 14M Polarography, Sargent, Model IV 8 POL IR Infrared spectroscopy Perkin-Elmer Model 257 7 9 NMR Nuclear Magnetic Resonance. Varian A-60. Comparison of retention times or GC-RT volumes. Hewlett-Packard Mod. 5700 Comparison of retention times on GC-RPOL columns with different polarity. Liquid chromatography. Spectra LC Physics Mod . 3500B Mass spectrometry of separated comGC-MS pounds at the outlet of a GC (Homemade, Chemistry Dept, Caltech). Mass spectrometry. DuPont Mod.21-4926 MS Method and Equipment Ref. Identification of compounds by comM-3, paring them to known compounds. E-3 Identifying homolog members when some E-3 of the homologs are available. Identifying compounds by their reK-3 tention volume. • Identifying minor components which Sect. could be separated ea~ily from the 5.3. 1.2 major ones Identifying a compound and its S-6 structure. Identifying a compound by the NMR S-6 chemical shift of its hydrogens or carbons. Identifying the chemical functional S-6 groups in a compound by the absorption of IR light. Identifying the unsaturated functional S-6 groups by absorption of UV light. Identifying functional groups which Z-1 can go redox reactions (to support identification by other methods). D-3 Application Methods of Qualitative Analysis 6 5 4 3 2 No. Table 5.3 .....,I I I..O WC Wet Chemistry 10 Application Many chemicals or groups can be identified by their reactions with specific chemicals. The reagents differ from one problem to another . continued Ref. Note : The prefix R prior to the notation means that the analysis was done on a sample which was concentrated by vacuum distillation. The prefix S means analysis of a component which has been separated from the mixture. Notation Method No. Table 5.3 - I OJ 0 I 6 3 3 3 6 6 Sulfur Containing Oxidation products of Butyl Sulfide GC-3 Sulfur Containing Oxidation products of Butyl Sulfide GC-4 Butyl Sulfide, tetrolin and oxidation products GC-5 Butyl Sulfide, toluene and bromobenzene GC-6 Butyl sulfide, oxidation products GC-7 Tetralin, octanol oxidation products of Butyl Sulfide GC-8 1/8 1/8 1/8 1/8 1/8 1/8 Q Gas Chrom Q Gas Chrom w 80/100 80/100 Chromosorb 60/80 w Chromosorb 60/80 w 80/100 80/100 Chromosorb 60/80 Q Gas Chrom Q Gas Chrom 1/8 6 Tetralone and tetralol from Tetralin GC-2 10 10 DEGS , DEGS 18 20 He 50 N2 15 8 min@ N2 17 70°C at 10° /min to 200°c 60-200°C N 19 @8° /min. 2 110 160 80-200°C N2 38 @4°c1 min. 20 N2 19 60-200°C N 19 2 min. @ l0°C 0 / 200 N2 19 10 10 Carbowax 20M Carbowax 20M Carbowax 20M DEGS DEGS 120 10 DEGS Gas Chrom 80/100 Support Q 1/8 Diameter (in. ) 6 Length (ft) CONDITIONS Station- Temp. Carrier Support Stationary -ary Program Flow oc Phase size mesh Phase% ml/min Methods of Analysis by Gas Chromatography COLUMN 1,4 Dihydronaphthalene from Tetralin GC-1 Separation of; and Notation Tab 1e 5. 4 I _, I CX) Length (ft) 3 2 Separation of; and notation Tetralin, DBT, GC-9 Toluene, DBT, 0-Dichlorobenzene 1/8 1/8 Diameter (in.) Gas Chrom. Q z Gas Chrom. Support 80/100 60/80 OV-1 20M Carbowax l 20 StationSupport Stationary ary size mesh Phase Phase% continued COLUMN Table 5.4 - He 30 min 100-180° N2 20 @ l5°C/ 215 CONDITIONS Temp. Carrier Program Flow °C ml/min I I 00 N -83- 5.3 . 2.2 Acidometric Titration of Organic Acids The method is applicable to the determination of dilute (C < ,o- 3 mol/lit) organic acids. The potential of a glass electrode is monitored vs. a double-junction reference electrode . Sat urated LiCl in CH 30H is used in the outer junction cell . 1-2 ml organic solution i s dissolved in 10 ml isopropanol and 5 ml 0.3 M LiCl solution in 70:30 volume ratio of a mixture of 95% c2H50H, 5% i-C 3H70H and water are added. titrate the acid. 0. 01 NaOH in H20 is used to The titration should be done very slowly in a well mixed cell, because the reactions of higher molecular weight acids are ,, slow . Radiometer automatic titrator was used for these measurements . 5.3.2.3 Liquid Chromatography Reverse-phase isocratic liquid chromatography was used for verification of the determination of tetralone and tetralyl hydroperoxide. 70% CH 30H in H20 was used and the absorbance was monitored at 254 nm and at 310 nm. Spectra Physics liquid chromatograph Model 3500B, equipped with Perkins-Elmer detector LC-55 was used for these detenninations. 5.3.3 Analytical Procedure The following qualitative analysis measurements were made on samples from all the experiments: 1. IR and R-IR 2. 3. WC for elementary sulfur GC-RT and GC-RPOL. On selected experiments GC-MS, MS, NMR and R-NMR, UV and POL were made. The following quantitative measurements were made on the samples of a 11 -84- the experiments: UV at 310 and 475 nm and often at other wavelengths in the visible range and GC measurements for sulfur compounds and for the hydrocarbon products. The determination of most of the samples was made more than twice. Selected samples were analyzed by polarography, titrametry, and calorimetry. The results of the analysis are given in Chapter 6. 5.4 Estimation of the Possible Mass Transfer in the Reactor In kinetics measurements it is required that the observed rate ., of change of concentrations be controlled by the rate of reaction and not by the rate of mass transfer. The dimensionless number Ms gives the ratio of the rate of mass transfer per unit volume to the rate of the fastest oxidation step. ( 5. l) where k.R, transfer, is the mass transfer coefficient, A the total area for mass V the volume of the solution, k2 the rate of constant for the fastest oxidation reaction, [0 2] an estimate of the oxygen concentration in the solution, and [R·] is an estimate of the free radical concentration in the solution. k.R, was estimated from equation (5.2) (C-5): DO 2/3 2 ) 113 k.R, = 0.3l(g 1!.) (\Jp (5.2) where µ and p are the viscosity and the diffusivity of the solvent and D is the diffusivity of oxygen in it. µ was calculated as if 02 p was estimated from a nomograph (P-1). the mixture were ideal. -85- D0 was estimated from the formula of Wilke and Chang (W-2): 2 D = 7.4 • 10 -8 T(xM) l 12 µ (5.3) V0. 6 where M is the molecular weight of the solvent, T the absolute ternperature, µ the viscosity cp, V the molecular volume of the solvent, and x the association factor of the solvent. x =l for non-protic 3 The values that were obtained are: p = 0. 9 gm/cm, 2 -4 2 2 Substitution in (5.2) µ = 2 • 10- poise, D = 9.2 • 10 cm /sec materials. 0 yields 2 ki = 0. 106 mole/cm/sec. The dispenser produced bubbles approxi- mately 0. 7 mm 2 10.8 cm. , the total area for mass transfer was approximately The volume of the liquid was approximately 300 cm 3. Had the oxygen been in equilibrium with the gas phase, the oxygen concentration would be about 6 • 10- 3 mole/liter. The concentration of R· is approximately 10- 10 mole/liter when there is sulfur compound in the solution, and 10- 7 when there is no sulfur compound. (These values were estimated by the computer integration procedure, see Chapter 7). k2 is approximately 7 • 106 liter/mole/sec. The smallest value of Ms is approximately: M = s 3 0.106 • 10.8 • 6 • 10= 5.45 . 300 • 7 • 10 6 • 6 • l O- 3 • 10- l O (5.4) Mass transfer of oxygen is possible at a much larger rate than the rate of oxygen consumption; therefore, one may assume that the solution is in equilibrium with the oxygen in the gas phase; and that the reactor is a batch reactor with a constant oxygen concentration. -86- Chapter 6 EXPERIMENTAL RESULTS The results of nine experiments are given in this chapter. The analysis and discussion of the results are made in Chapter 7. For each of the experiments three kinds of data are listed: l. the experimental conditions, 2. qualitative results, 3. quantitative results . and When standard qualitative or quantitative analysis techniques were used, no details are given. Methods which were developed for this study are discussed in Chapter 5 or Appendix A. Experimental errors are possible in the characterization of the experimental conditions, in the qualitative or in the quantitative analysis. The experimental conditions are known and were controlled with an error smaller than: l. 0.5% in the weights of the hydrocarbon and the solvent, 2. 0.03% in the weights of the sulfur compound, 3. 0.2°C in the temperature. The large error was due to cooling down of the reactor content when a sample was withdrawn. 4. 5 mm Hg (estimate) in the pressure, 5. 20 sec in the time (period of sampling), 6. 10 mm in the HBr partial pressure, 7. 70 mm Hg in the o2 parti a1 pressure for the first 30 min of the experiment, and 30 mm Hg after that. The identification of a chemical compound is rarely certain. When known and standard material is available for comparison, the chances for a -87- correct identification are relatively large; however, when a standard is not available, or when the compound is not pure or is a minor component of a complex mixture, the certainty of the identification decreases . Most of the sulfur con t aining products are minor compon- ents of a complex mixture, and no standard is available for comparison. The preparation of such standards involves considerable handling of a and B halosulfides, which are very dangerous poisons, similar to mustard gas. Because of lack of adequate facilities the compounds were not synthesized i ndependently . Synthesis of standards was done when it was not too dangerous (e.g., butyl sulfonic acid and l,2dihydronaphthalene). In the tables which detail the qualitative results, the last column gives the author's relative confidence in the certainty of the identification in a scale from Oto 100. The quantitative analysis of sulfur compounds by gas-chromatography has been detailed by the author in two papers, which have been submitted for publication (A-4,A-5). Systematic errors of up to 25% are possible for the sulfur compounds for which cal i bration standards were not available. The error in GC data with an FID detector was usually below 3% for non-routine analysis and less than 1% for routine analysis. Such an accuracy was not sufficient for the major components, e.g., the hydrocarbon and the solvent. The total consumption of hydrocarbon was of the order of the 1% possible error of the analysis. Thus material balance on the hydrocarbon-derived species was impossible. However, the major products of the oxidation of the hydrocarbon were determined -88and provide a measure of the oxidation . The accuracy of polarographic and tetrametric methods is estimated as 20 and 2%, respectively . In the tables of the quantitative results, t he method which was used for the determination is written under the component name. The time interval between the end of the exper iment and the analysis affect the quantitative results. The solutions that are formed keep reacting even when frozen . Therefore only a limited nurrber of reliable determinations can be made on each sample. The result of the analysis of the reaction product is markedly different from that of the last sample, although the time difference between them is small . The product changes faster than the sample, even under nitrogen atmosphere because it was kept in a large bottle with a small er surface/volume ratio . 6. 1 Experiment l 6. 1. l Experimental Conditions Hydrocarbon: Tetralin Sul fur compound: None Solvent: Bromobenzene 218 grams Total Pressure: 100 grams Oxygen (0 2 ), P = 591 mm Hg, CO= 5.3xlo- 3mol/lit 02 e 2 Hydrogen bromide (HBr), PHBr= 149 mm Hg, 3 CeHBr= l.3xl0- mol/lit 739.8 mm Hg Temperature: 140 ± o. 2 °C Duration of run: 336 min Oxidizer: Catalyst: -89- 6.1.2 Qualitative Results of Experiment l Table 6. l Qualitative Results of Experiment l Compound Method Support Evidence Confidence of the Identification and Notes Tetra lone GC-RT IR,POL, VV Tetralol GC-RT IR , NMR 100 compared with synthet i c material 100 compared with synthet i c materi a1 90 compared with syn thet i c materi a1 100 iodometric method 1,2-Dihydronaphthalene GC-RT IR Tet~alyl hydroperoxide POL WC Hydrogen peroxide POL 90 Water 100 Comments on the Qualitative Analysis 1. 475 mm). Pale yellow color developed during the oxidation (Amax= It was possible to extract some of the colored material by an aqueous 2N NaOH solution. identification. Its quantity was very small, however, for At least two components were found in it using MS. One compound had two Br atoms; the material had shown a strong OH absorbance band in the IR.By comparison with the literature (R-3) it is believed that the compounds are Y~O-phenoxybutyric acid (I) which exists in equilibrium with its phenol . (II}. ,9 01::i 4 H+ I - -- 0 rt'OH ► II -90The yellow color is due to the phenol-quinone equilibria Br III III is the result of the reaction of Br2 with the olefin-acidic product, IV, of the oxidation of 1,2-dihydrohaphthalene. Br 2 _is the dominant termination product of Br• . ct° /.0 rArc!~ f~o + ~ H 001 o B~-► Br IV &.oo H Br II I Low eV MS of the solid residue (Appendix A) gives two strong peaks at m/e = 223 and at m/e = 297 . Both contain Br and may corres- pond to III which lost CO 2 or two H20 molecules . 2. Standard iodometric methods had shown very small color indi- cation of the presence of peroxide in the product. The polarographic method of Willits et al. (W-1) had shown the presence of hydroperoxide and hydrogen peroxide (E 112 = -0.73 volts respectively, using DME and Pt electrodes). E112 = -0.2 volts, However, as the concentraand tion of tetralone increases, its half wave decreases due to the hydroperoxide (for tetralone E112 = -1.2 volts). -913. The water that was formed made the material turbid . The turbidity increased when the samples we re cooled . 6.1 . 3 Quantitative Results of Ex~eriment l Table 6.2 Quantitative Results of Experiment 1 The method of analysis of each specie is written under the component name. 1,2-Dihydro- Tetra lone Sample Time Tetrali n naphthalene mo1 /1 it Hydrogen_perox!de No. (min) moa6~it mo 1~~~t x , 02 mol/l1tp0~g1 Ge: 11 o2 0. 034 5. 109 0.00 0 0 1 15 5.014 0.212 0.0 2 30 5.059 0. 186 0.352 0.028 3. 45 0. 138 0.490 0.075 4.943 0. 239 4 60 0. 426 0.262 5 75 4.925 0. 640 0.175 o. 211 6 90 4.995 0. 750 o. 175 105 0. 270 7 4. 891 0.780 0. 150 o. 71 120 1. 26 8 0. 412 9 150 0.62 1. 18 0.524 10 180 4.750 1. 63 1. 30 0.599 4.750 1. 34 1. 55 11 210 0. 849 1. 98 1. 56 12 4.950 1. 20 240 1. 96 1. 77 4.440 1. 27 270 . 13 1. 53 2 .10 1. 57 14 300 4.260 1. 71 2.23 1. 55 15 330 product 336 2. 76 2.20 1. 64 Comments on the Quantitative Analysis The concentration of tetralyl hydroperoxide was 0.00, 8. 24·10- 5, and 9.0·10- 5 mole/liter at times 0, 15 and 30 min, respectively. The 1. tetralone that is formed did not allow the determination of hydroperoxide after 30 min. -922. Approximately l0- 5mole of acid was formed (about 3.3·10- 5 mole/liter). The acid was very weak and exhibited very slow rate of reaction with base. Potentiometric titration in organic media (Section was used for its determination . The heat of combustion of 5.3.2.2) the first sample was 8504.5 cal/gram and that of the product was 8483.1 cal/gram. 6.2 Experiment 2 6. 2. l Experimen ta 1 Conditions Hydrocarbon: None ~ulfur compound: Dibutyl sulfide 6.23 grams Sol vent: Bromobenzene 436 Oxidizer: Total pressure: Oxygen (0 2 ), P = 619 mm Hg, ce = 5.6·10- 3mol/lit 02 02 Hydrogen bromide (HBr), PHBr= 120 mm Hg, 3 CeHBr= 1. l·l0- mol/lit 738. 8 mm Hg Temperature: 140 ± 0. 5 °C Duration of run: 334 min Catalyst: grams 6.2.2 Qualitative Results of Experiment 2 Table 6.3 Qualitative Results of Experiment 2 Compound Method Support Evidence Butyl sulfoxide GC-RT 100 Butyl sul fone GC-RT GC-RPOL,S-NMR IR GC-RPOL a-ketobutyl sulfoxide GC-RPOL IR, R-IR 80 a-ketobutyl sulfide GC-RPOL IR,R-IR 70 a-hydroxy butyl sulfide GC-RPOL IR,S-IR 70 Butyl di sulfide GC-RT Confidence of the Identification and Notes 85 40 Because of the low confidence, it was denoted A in Table 6-5. -93- Comments on the Qualitative Analysis l. Intense brown-yellow color was formed in the reaction. See comment l in 6. l for the explanation. 2. Several new bands in the IR region were formed, their list and explanation are given in Tab . e 6. 4 below . Table 6.4 New IR Bands Formed in Experiment 2 Band cm -1 Observation 2860, 2920, 2945 small intensity consumption of aliphatic hydrogens 1715, 1750 most prominent 1420, 1360, 1218 _small intensity 1230-1130, 1420-1350 seen in R-IR formation of C=O groups 0 so 2 , RO-S--OR (?) -0-S0 2-0(?) 1050-1070 3. Interpretation R-IR Twenty-five new sulfur-containing compounds were detected in the product by the sulfur-specific FPO. The concentration of seven of them is estimated to be above ,o- 6mole/liter. 4. Comments 3, Section 6.1.2 are applicable. 6.2 . 3 Quantitative Results for Experiment 2 -94Table 6.5 Quantitative Results for Major Components, Experiment 2 The concentrations were multiplied by 10+6 throughout. All sulfur compounds were detennined by method GC-4. -Keto a-Hydroxy a- Hydroxy Butyl Butyl Butyl Butyl Butyl Sample Time Sulfoxide Sulfone Sulfide Sulfide Sulfoxide A B No. (min) mol/lit mol/lit mol/lit mol/lit mol/lit mol/lit mol/lit 0 0 8.8 i 15 8.8 0.68 2.3 2 30 14.8 2.22 2.3 3 45 19.0 o. 68 2.70 2.3 4 60 56.0 2. 76 6.20 3.96 5 75 82.0 5.40 2.7 7.00 2.8 6 90 106 . 0.68 15.0 4.6 9.20 3.36 7 105 190 l. 54 18.6 5.8 4.6 2.02 8 120 304 2. 76 28.0 9.2 l. 34 4.6 2.3 9 135 544 12.6 54.0 19.0 3.90 5.8 3. 16 10 176 2686 21.0 158. 25.0 3.90 6.2 3.48 11 206 5880 31.6 118. 19.0 3.40 6.2 2.6 12 235 5361 48.0 490. 38.0 10.4 7.8 3.36 13 265 6399 54.0 1040. 47.0 14.8 11. 0 3.7 14 300 9793 89. 0 1053. 64.0 16.0 14.8 9.2 2.3 Comments on the Quantitative Analtsis l. The calibration curve for butyl sulfoxide was used in the quanti - tation of all the sulfur compounds except for the butyl sulfone. Because of differences in the response factors for different sulfur functional -95groups, the set of results for each component may have a systematic error of up to 25%. The differences in the response factor for oxidized sulfur compounds were discussed in reference (S-8) . 6.3 Experiment 3 6.3. l Experimental Conditions Hydrocarbon: Tetra l in 218 grams Sulfur compound: Dibutyl sulfide 6. 23 grams Solvent: Bromobenzene Oxidizer: Catalyst: 100 grams Oxygen (0 2 ), P0 = 629 mm Hg, CO= 5.7xlo- 3mol/lit 2 e 2 Hydrogen bromide (HBr), PHBr= 0.0 mm Hg, ceHBr= 0.0 mol/lit Total pressure: 629 mm Hg Temperature: 140 ± O. 3 °c Duration of Run: 345 min 6.3.2 Qualitative Results for Experiment 3 Table 6.6 Qualitative Results of Experiment 3 Support Confidence of the Compound Method Evidence Identification and Notes Butyl sul foxi de GC-RT 100 GC-RPOL GC-RPOL, S-NMR,IR IR,R-IR a-Keto butyl sulfide a-Hydroxy butyl sulfoxide GC-RPOL . IR,R-IR 80 IR,UV,POL 100 GC-RT Tetra lone 70 Comments on the Qualitative Analysis l. Pale yellow color was formed in the reaction. in Section 6.1.2. See comment l -966.3.3 Quantitative Results for Experiment 3 Table 6.7 Quantitative Results for Major Components, Experiment 3 Method of determination for each component appears under its name. Butyl - Hydroxy -Keto Sulfoxide Butyl Sul foxi de Butyl Sulfide Sample Time Tetralin mol /1 it mol/lit mol/lit X 103 X lQ5 No. (min) mol/1 iter X 106 GC-8 GC-4 GC-4 GC-4 0.4 5.514 0 0 0.54 l 15 5. 577 2. l o. 74 2 30 5.482 4.6 4.9 3 45 6.8 l. 18 4 60 5.444 12. 6.8 1. 78 5 85 5.698 17. 10. 4.2 6 110 5. 723 13. 6.8 7 140 5. 720 60. 20. 8 190 5.420 60. 40. 9 230 5.616 19.0 210. 79. 10 270 5.467 18.0 1700. 98. 11 310 5.583 18. 0 1100. 120. 12 340 5. 227 16.0 96Q. 99. p 345 19.0 2000. 280. 19.5 Comments on the Quantitative Analysis 1. Comment l in Section 6.2.3 is applicable. -976.4 6.4.l Experiment 4 Experimental Conditions Hydrocarbon: Tetral in Sulfur compound: Dibutyl sulfide Solvent: Bromobenzene Oxidizer: Catalyst: 218 grams 12.46 grams 100 grams Oxygen (0 2 ), P =669 mm Hg, ce = 6xlo- 3mol/lit 02 02 Hydrogen bromide (HBr), PHBr= 71 mm Hg, _ 4 CeHBr= 6.4xlO mol/lit Total pressure: 739.8 mm Hg. Temperature: 140 o. 3 °C Duration of run: 341 min 6.4.2 Qualitative Results for Experiment 4 Table 6.8 Qualitative Results of Experiment 4 Compound Support Method Evidence Butyl sulfoxide GC-RT Confidence of the Identification and Notes 100 Butyl sul fone GC-RPOL,S-NMR, IR GC-RT GC-RPOL a-Hydroxy butyl sulfoxide GC-RPOL GC-RPOL, R-IR 80 a-Keto butyl sulfide GC-RPOL IR,R-IR 70 a-Hydroxy butyl sulfide GC-RPOL IR,R-IR 70 a-Keto butyl sulfoxide GC-RPOL IR,R-IR 70 Butyl disulfide GC-RT 50 Butyraldehyde GC-MS 90 85 -98- Comments on the Qualitative Analysis . l. IR bands were found which can be due to -so 2-s- and to --O~S0 2-{}-- as detailed in comment 2 , Section 6.2.2 . ·2. IR examination of the GAS above the liquid shows: 1) absorption band at 1640 cm- 1 ; absorption at this wavenumber is due to double bond, possibly 1- butene; 2) absorpt ion band at 3400-3500 cm-l, due to hydroxyl, possibly butyl alcohol; butyl aldehyde. GC-MS. 3) absorption band at 1740 cm -1 , possibly due to Butyraldehyde was found also in the liquid solution by Butyraldehyde is produced by the decomposition of a-hydroxy- sulfide . After a day in the freezer the only bands left were for --OH and for double bond 3. (much larger in size than in the original sample). Mass spectrometric examination of the gas phase above the reac- tor shows: a) The presence of very small amounts of hydrogen (H ) 2 (Confidence : 60%). b) A strong molecular ion at m/e = 90, possibly due to butyl mercaptan. An ion at 92 about 5% in height of the ion at 90, supports the identification of butyl mercaptan because the isotopic ratio of s34 to s32 is 4.4 to 100. Furthermore, fragments at 90-34 = 56, and at 34 are found. The ion at 34 may be due to H2s+ Mercaptans are known to lose H2S ( S-6). c) Comment 1, Section 6. 1.2 is applicable. 6.4.3 Quantitative Results for Experiment 4 The data are presented in Tables 6.9 and 6. 10. The components that -99start to be important after 100 min are given in Table 6.8. Table 6.9 Quantitative Results for Major Components, Experiment 4 The concentration of all the components was multipied by 10+ 4. The method of determination for each component is written under the component name. Sample Time (min) No. 0 l 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 0 6 12 18 24 30 36 42 48 54 60 90 120 150 180 210 240 270 300 330 Tetralol mol/1 it GC-2 -Hydroxy -Keto Butyl Butyl Butyl Butyl Tetralone Sulfoxide Sul fone Sulfoxide Sulfide mol/lit mol /1 it mol/lit mol/lit mol /1 it GC-3 GC-3 GC-2 GC-3 GC-3 0.026 0.28 0.33 2. 15 1.0 3.7 0.8 4.5 0.96 4.7 2. l o. 179 4.45 2. 66 6. 39 8.00 8. 21 8.31 7.48 9.28 2.06 4.06 3.98 4.96 3.33 3.68 l. 66 2.9 2. l l. 5 3.0 2.5 2. l 1.0 23. 32. 59. 82. 86. 95.5 96. 100. 83. 6.0 10. 16. 24. 27. 27. 21. 17. 11. l. 7 0.6 0.7 0.88 l. 2 1.0 1.2 1.0 0. l 0.2 l. 55 3.4 5.5 14.2 22.2 25.0 25.3 25.0 The heat of combustion of sample O was 8687.9 cal/gram and that of sample 19 was 8594.0 cal/gram. 12.3 34 . 5 330 19 7.6 36.5 13. 5 • 300 18 6.7 34.0 9.5 270 17 4. 3 26.5 240 16 1.95 2. 6 0.42 1. 36 l. 20 1. 2 l. 45 1.2 0.53 0.79 *in the first 120 min it increases gradually from about l0- 5mole/liter to approximately 8-10- 5 mole/liter. 14 . 0 5.6 l. 6 1.8 16.3 2.6 210 15 1. 9 0. 78 7.0 0.34 180 14 0.9 0.34 2.5 0.34 150 13 0. 8 0.7 o. 18 120 12 Sample Time (min) No. a-Hydroxy a-Keto Butyl Sulfide* Butyl Sul foxi de mol/lit mol/lit RT=l425-1435 sec RT=l715-1750 sec B' mol /1 it RT=980-1000 sec A' mol/lit RT=840-850 sec B" mol/1 it RT=l000-1030 sec Table 6.10 Quantitative Results, Experiment 4. Components that are produced after a Method of determination for each component appears under its 1ong time. name. RT is retention time in the GC, using method GC-3; the concentrations were multiplied by 10 4 throughout. I _. I 0 0 -101Comments on the Quantitative Analysis l. Comment l, Section 6. 2.3 is applicable. 6.5 Experiment 5 6.5. l Experimental Conditions Hydrocarbon: Tetralin 218 grams Sul fur compound: Di butyl sulfide 6.23 grams Solvent: Bromobenzene Oxidizer: Catalyst: 100 grams Oxygen (0 2 ), P0 =626 mm Hg, C O =5.6xlo- 3mol/lit 2 e 2 Hydrogen bromide (HBr), PHBr=ll4 mm Hg,_ 3 CeHBr=l.OxlO mol/lit Total pressure: 739.3 mm Hg Temperature: 140 ± O. 3°C Duration of run: 375 min 6.5.2 Qualitative Results for Experiment 5 Table 6. 11 Qualitative Results, Experiment 5 Support Confidence of the Method Evidence Identification and Notes Compound Sulfur (elementary) WC Butyl Sul foxi de GC-RT Butyl Sul fone GC-RT , Ref. yellow 60 Acetone-KOH (U-1) col or IR,S-NMR, 100 GC-RPOL GC-RPOL 100 Butyl Sulfonic Acid WC S-NMR 60 a-Hydroxy Butyl Sulfoxide GC-RPOL IR,R-IR 80 a-Keto Butyl Sulfide GC-RPOL IR,R-IR 70 a-Keto Butyl Sul foxi de GC-RPOL IR,R-IR 70 -102Table 6.11 (continued) Compound Method Butyl Disulfide GC- RT Support Evidence Confidence of the Identification and Notes 50 a-Hydroxy Butyl Sulfide GC- RP0L IR Tetralol GC- MS IR l 00 Tetralin Diketone GC-MS IR 80 Butyl Bromide GC-MS Butanone GC-MS IR 100 3-0-phenyl Propionic Acid GC-MS IR 70 100 See Section 5.3. 1.2 Comments on the Qualitative Resu l ts l. Comments l, 2, and 3, Section 6.1 . 2 are applicable. 2. Comments 1, 2, and 3, Section 6. 2.2 are applicable. 3. Butyl bromide was identified by GC-MS (see 5.3. 1.2 for the details of the method.) No other Br-containing compounds which boil below 200°C were found. Hm"lever, MS analysis of crystalline solid that was obtained by method WC-1 contained two BR atoms/molecule. 6.5.3 Quantitative Results, Experiment 5 The concentration of the compounds which become significant above 10- 4 mole/liter only after 120 minutes are given in Table 6.13. 0 15 30 45 60 75 90 105 l 00 150 180 210 240 270 300 330 360 0 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Time (min) Sample No. RT is 0. 1424 0. 1463 0. 1439 o. 1366 0. 1355 0. 1349 0. 1343 0. 1356 0. 1340 0. 1370 o. 1317 0.1113 0.0829 o. 1069 o. 1131 o. 1132 0. l 078 0. 1 0.3 0. 6 2. 2 5. 2 9.6 18. 14 . 2 12.2 4.0 4.2 .2.8 2.2 7.8 12.4 2. 4 1. 9 5.9 13.4 19.6 62 . 335. 680. 860. 1440 . 1040. 6. 68 6.27 9.35 4.08 6. 68 6. 27 7. 98 7.98 12 . 9 20.5 47 . 5 716 6860 12860 21090 24360 26470 2. 49 3. 15 3.60 4.70 8. 31 14.9 21.4 32.l 33.6 34.9 32 .9 29.5 24.6 1. 92 1.90 2. 72 1. 36 2. 72 4. 2 3. 42 11. 8 17. 6 18. 4 18. 4 45. 66. 70 . 90. 81. 87. Butyl a-Hydroxy Butyl Butyl a-Keto Butyl a- Keto Butyl Sul fone 5 Sulfoxide Butyl Sul foxJde Sulfide Sulfoxide 5 Sulfide mo l / l it mo l / l i t x 10 mo 1/1 it x 10 mo l /1 it x ,o-6 mo 1/1 it x 103 mol /1 it x 10 RT= 1936 RT= 1305 RT= 1145 RT= 2320 RT= 247 RT= 1274 GC-4 GC-4 GC -4 GC-4 GC-5 GC-4 Method of detennination for each component appears under its name . the retention time in seconds. I __, I w 0 -104Tab 1e 6 .13 Quantitative Results, Experiment~ The concentration of components which become important after 120 min. The analysis was done by method GC-4. RT is the retention time in seconds. a-Hydroxy Butyl Sulfi~e mo l /1 it x 10 RT= 1317 A mo l / l i t x l O4 RT = 748 B mo 1/ 1it x 1o4 RT= 830 0.5 0.92 0.33 150 1. 32 1. 12 0.47 10 180 2.06 l. 33 0. 50 11 '·210 5.8 2.34 0.94 Sample No. Time (min) 8 120 9 \ 12 240 6.6 4.80 2 .16 13 270 8. 7 18.8 7.8 Comments on the Quantitative Analysis 1. Comments 1 and 2, Section 6. 1.3 are applicable. 2. Comment 1, Section 6.2.3 is applicable. 6. 6 Experiment 6 6.6. 1 Experimental Conditions Hydrocarbon: Toluene 60. 0 grams Sul fur compound: Dibutyl sulfide 6.23 grams Sol vent: Bromobenzene 332.4 grams Oxidizer: Oxygen (o2 ) , P0 = 636 mm Hg, CO =5.6 xlO mol/l1t 2 e 2 Hydrogen bromide (HBr), PHBr= 103 mm Hg, _ . 4 CeHBr=9.0x10 mol/l1t 739.3 mm Hg Catalyst: Total Pressure: -3 . -105Temperature: 136 . 5 ± 0. 1 °C Duration of run: 336 min 6.6.2 Qualitative Results for Experiment 6 Table 6.14 Qualitative Results of Experiment 6 Compound Method Support Evidence Butyl Sul foxi de GC-RT IR Butyl Sulfone GC-RT Confidence of the Identification and Notes 95 80 a-Keto Butyl Sulfoxide GC-RPOL IR,R-IR 80 GC-RPOL IR,R-IR 70 a-Hydroxy Butyl Sulfide GC-RPOL IR,R-IR 70 a-Hydroxy Butyl Sulfoxide Benzaldehyde GC-RPOL IR,R-IR 80 GC-RT IR 70 a-Keto Butyl Sulfide Comments on the Qualitative Results 1. Comments l, 2 and 3, Section 6. 1.2 are applicable. 2. Comments 2 and 3, Section 6.2.2 are applicable. 6.6.3 Quantitative Results for Experiment 6 The products that are formed after 105 min are given in Table 6.15. -106Table 6. 15 Quantitative Results, Experiment 6 The determination of the toluene and butyl sulfide was done by method GC-9. The sulfur compounds were determined by method GC-6. Butyl Butyl a-Keto Butyl Butyl Sulfoxide Sul fone Sul foxi de Sample Time Toluene Sulfide mol /1 it mol/lit mol/lit X 104 X 104 X 104 (min) mol/lit mol/lit No. 0 l 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0 15 30 45 60 75 90 105 120 150 180 210 240 270 300 330 2.30 2. 19 2. 15 2.15 2. 22 2 .18 2.01 2. 13 2. 12 2 .11 2. 17 2.05 2. 10 2.07 o. 156 o. 144 o. 145 0.57 2.3 0. 150 0. 139 0. 140 o. 142 o. 140 0. 141 0. 141 o. 136 o. 135 0. 140 o. 131 o. 128 6.8 1.04 3.4 7.7 8.4 10.3 10. 5 3.85 4.75 6.35 7.7 3.3 10.3 10.0 10.2 9.2 10.0 1.0 6. 1 4.4 4.8 4. 7 4. l 0.49 1.01 4.5 6.3 16.5 19.8 18.5 l 0. 1 15. 2 28. 5 29.5 36.5 -106aTable 6. 16 Quantitative Results, Experiment 6. The concentration of the sulfur compounds which become important after 105 min. The determination of the sulfur compounds was by method GC-6. RT is the retention time in seconds . a-Hydroxy A'' B" a-Keto Butyl a-Hydroxy Butyl Sulfoxide mol/lit moi/lit x ,o4 x 104 Sample Time mo~/~f!d:104 B~~{j1~~~f~~i mol /1 it x ,o4 RT= 910 RT= 1105 RT = 1920 RT= 375 RT= 845 No. (min) 0.33 7 105 8 120 0.33 9 150 0.57 10 180 11 210 1.01 1.01 12 240 2.00 13 270 14 15 o. 61 1. 87 o. 42 2.5 0.6 0.73 ~-7 1.03 1. 62 1.013 3.5 3.03 3.3 1. 74 2.01 4.4 1.08 300 3.3 1. 75 1. 72 4.6 4. 75 330 3.65 2. l 2.00 5.4 2. 35 0.43 0.95 Comments on the Quantitative Results 6.7 6.7.l 1. Comment 1, Section 6.2.3 is applicable. 2. Comment 2, Section 6.1.3 is applicable. Experiment 7 Experimental ·conditions Hydrocarbon: Tetra l in 218 grams Sulfur compound: Dibutyl sulfide 6.23 grams Solvent: 1-0ctanol Oxidizer: 80.8 grams Oxygen (0 2 ), P =609 mm Hg, ce =5.5xl0- 3mol/lit 02 02 -107- 6.7.2 Catalyst: Hydrogen bromide (HBr), PHBr=l28 mm Hg, CeHBr=l. 15xl0- 3mol/lit Total pressure: 736 . 7 mm Hg Temperature: 140 ±0 . 3 °c Duration of run: 330 min Qualitative Results fo r Experiment 7 Table 6. 17 Qualitative Results, Experiment 7 Compound Method Butyl Sulfoxide GC-RT a-Keto Butyl Sulfoxide GC-RP0L Support Evidence Confidence of the Identification and Notes IR,R-IR, S-NMR IR,R-IR 100 80 Comments on the Qualitative Results 1. The yellow color that appeared was considerably lighter than that seen in experiments 2 to 6. 2. Fewer sulfur-containing compounds were produced than i n experiments 2 to 6. There were only two major ones. 6.7.3 Quantitative Results for Experiment 7 Table 6. 18 Quantitative Results, Experiment 7 The determination of the sulfoxides was done by method GC-8 . The tetralone was determined by UV; the sulfid~ by method GC-5. Tetra lone Butyl Butyl a-Keto mol/lit Sulfide Sulfoxide Butyl Sulfoxide Sample Time (min) No . xu~o3 mo~e-~t mol6t~~ x 103 moltl~~x,06 0 0 1 15 30 2 0.108 0. 1292 0.1167 0. 1143 0. l o. 707 10. 12 4. l -108- Sample Time (min) No. 3 4 5 6 7 8 '9 10 11 12 13 14 15 Table 6.13 (continued) Tetra lone Butyl Butyl a- Keto Sulfide mol /1 it Sulfoxide Butyl Sulfoxide X lQ3 mol/lit mol/lit x 103 mol/lit x 106 45 60 75 90 105 120 150 180 210 240 270 300 330 2.95 3.0 3. l 3.9 3.95 5.0 7.0 10. 5 11. 3 14.8 15. 2 19.2 22.5 o. 1103 o. 1127 0.0987 0.0991 0 .0897 0.0996 o. 1073 0. l 003 0.1011 0.0961 15. 4 18.4 21. 4 27.3 27.8 29. l 7.3 6.2 4.3 14. 32. 24.4 28.0 27.2 26.4 25.4 29.6 32. 6 20.6 20.8 20.7 6.8 Exeeriment 8 6.8. l Exee ri mental Conditions Hydrocarbon: Tetralin Sulfur compound: Dibenzothiophene Solvent: Bromobenzene 218 grams 6.813 grams Total pressure: 100 grams Oxygen (0 2 ), P = 620 mm Hg, ce =6.4xl0- 3mol/lit 02 02 Hydrogen bromide (HBr), PHBr=ll7 mm Hg, _ 3 . CeHB = l.2xlO mol/lit 737.2 mm Hg •r Temperature: 159 . 9 ± 0. 2 ° C Duration of run: 335 min Oxidizer: Catalyst: -109- 6.8.2 Qualitative Results for Experiment 8 Table 6. 19 Qualitative Results, Experiment 8 Compound Method Support Evidence Confidence of the Identification and Notes Tetra lone GC-RT IR,S-IR,LC 100 Tetralol GC-RT IR,LC 70 Corrments on the Qualitative Results 1. No sulfur-containing compound was found in the product. No IR bands which can be attributed unambiguously to oxidized sulfur compounds were found. Strong carbonyl band (1690 cm- 1) is evident and a very weak -OH stretch band (3500 cm -1 ) can be recognized. 2. Comments 1 and 3, Section 6. 1.2 are applicable. 6.8.3 Quantitive Results for Experiment 8 Table 6.20 Qtiantitative Results, Experiment 8 Sample No. Time (min) 0 l 2 3 4 5 6 7 8 9 0 15 30 45 60 75 90 105 120 150 Tetra lone mo l / l it x 10 3 GC-2 2.265 2. 577 2.926 5.837 6.863 9.296 9.036 8. 100 8.954 DBT mol /1 it GC-9 o. 1133 0. 1189 o. 1090 o. 1204 0. 1058 0.1084 0.1119 0. l 061 0. 1098 0. 1146 Table 6.20 -110(continued) Sample No. Time (min) Tetra lone mo l / l i t x l O3 DBT mo l /1 it 10 11 12 13 14 15 180 210 240 270 300 330 335 8. 975 8.580 11. 93 12.06 15.20 17.75 20.60 0. 1056 0. 12 71 o. 1149 0.1167 0. 1186 o. 1155 0 .1147 p Comment~ on the Quantitative Analysis l. Comment 2, Section 6. 1.3 is applicable. 2. Within the error of the analysis the concentration of DBT does not change throughout the run; however, there is a very slight trend toward decrease in the concentration with time. 6.9 6.9. l Experiment 9 Experimental Conditions Hydrocarbon: Tetralin Sulfur compound: Dibenzothiophene Solvent: 1-0ctanol 218 grams 6.813 grams Total pressure: 80.8 grams Ox-ygen (0 ), P0 =624 mm Hg, C O =6.4xlo- 3mol/lit 2 2 e 2 Hydrogen bromide (HBr), PHB =114 mm Hg, r . -3 CeHBr=l.2 xlO mol/lit 738.2 mm Hg Temperature: 160 ±0.2 °C Duration of run: 339 min Oxidizer: Catalyst: -111 6.9.2 Qualita~ive Results for Experiment 9 Table 6.21 Qualitative Results, Experiment 9 Compound Method Support Evidence Tetra lone GC-RT IR,S-IR,LC 100 Tetralol GC-RT IR,S-IR,LC 70 Confidence of the Identification and Notes Comments on the Qualitative Results 1. Comments 1 and 2, Section 6.8.2 are applicable . 6.9.3 Quantitative Results for Experiment 9 Table 6.22 Quantitative Resu l ts, Experiment 9 (The concentrations were determined by method GC-9.) Sample No. Time (min) Tetra lone mol/d~!{ 10 3 DBT irol/lit GC-9 0 1 2 0 15 30 45 60 75 90 105 120 150 180 210 240 o. 784 o. 1122 3 4 5 6 7 8 9 10 11 12 0.784 1.27 3.00 4.41 6.20 6.94 7.43 7. 31 8.07 9.21 10.96 13. 21 o. 1102 o. 1024 o. 1080 o. 1051 0.0986 0.1013 0. 0985 0.1040 0.0996 0. l 007 0. 1047 -112Table 6.22 ( continued) Sample No. Time (min) Tetra lone mo l /1 it 10 3 DBT mol/lit 13 14 15 270 300 330 339 14. 14 16.61 17. 66 17. 17 0.0999 0. 0995 0.0992 0. 1039 p Comments on the Quantitative Results 1. Comments l and 2, Section 6.8.3 are applicable. -113- Chapter 7 ANALYSIS AND DISCUSSION OF THE RESULTS The two goals of the analysis of the results were: 1) to investigate the reaction mechanism, and 2) to obtain approximate values for the rate constants of the most important reactions which take place in the system. Several other topics are discussed in this chapter, such as the loss of heating value and the effect of the protic nature of the solvent on the oxidation. The system is extremely complex and many approximations were made in order to reach some conclusions. Many of the approximations are analyzed; however, some common approximations are not detailed, for the sake of brevity. 7.1 The Structure and the Logic of the Discussion In the analysis of kinetic data, the most important steps of the mechanism detennine the reaction products. Therefore, the qualitative results were used to investigate the reaction mechanism. The quanti- tative data were used to estimate the rate constants of the reactions and to study the change in enthalpy of various reactions. The discussion of the results consists of four parts: 1) introduction, in which the tools that were used in the analysis are examined, 2) computer simulation of the system and a comparison of the data with predictions of the model, 3) semi-quantitative discussion of the effects of the catalyst, the solvent, and the hydrocarbon, and 4) semi-quantitative discussion of the loss of heating value vs desulfurization. -114- Errors in the identification of species and other systematic errors, plus random experimental errors in the quant i tative analysis, were not taken into account in the simulation because of the size of the system involved. In the oxidation of a hydrocarbon many products are formed . Typical products are alcohols, ketones, esters, olefins, and acids . The reactivity of these species toward oxidation is much larger than that of the original hydrocarbon; therefore the course of oxidation changes soon after small amounts of intermediates accumulate . For example , small amounts of acids have enormous effect on the rates of decomposition of peroxides and on their oxidation potential . Within the constraints of the work it was impossible to take the effect of products of advanced oxidation into account. Therefore, the results of the simulation agree with the data for the first 100 min of the oxidation and do a less adequate job after long periods of oxidation . when the concentration of secondary and tertiary products become too large . 7.2 System Approach to the Simulation of the Chemical System The simulation of a complex system is a difficult problem and cannot always be done satisfactorily. Modeling from first principles is usually complex and needs in general more variables than empirical models; however, such models have a better chance of representing a system for an extended range of values of the variables. Empirical and semi-empirical models would, in general, give a better description of a system in a limited range of the variables and would need a smaller -115- number of parameters; however , extrapolation of such models is more risky and may lead to erroneous results. The area of simulation of complex reaction systems has recently been revi ewed by Weekman (W-4). Modeling of a chemical system , taking into account all the elementary reaction steps , indeed needs many parameters and involves numerous differential equations . However, i t has a better chance of representing the system correctly when the interaction with other chemical systems are also considered . The following logic was applied in the modeling of the reactions in systems which contain a hydrocarbon, a sulfur compound , oxygen and a catalyst: l. Define subsystems and model them from first principles, using elementary reaction steps as much as possible, e. g., subsystem NOCAT contains a hydrocarbon, oxygen and sulfur compound only. 2. Verify the adequacy of the model by comparing the theoretical predictions with the experimental data. Data for the subsystem were used to modify the parameters of the model. 3. Combine the submodels to represent a larger system. 4. Check the predictions of the combined model vs the experimental data and modify it as required. Although the approach is straightforward, its implementation is difficult. Rate constant data for a few of the elementary steps were available from independent sources. study. The rest were estimated in this Systematic and random errors in the experimental data compli- cate the task of the evaluation of the adequacy of the mechanism and the values that are obtained for the rate constants. -116Moreover, each subsystem is nonlinear and stiff, and additional nonlinearity is introduced when the subsystems are combined. The non- linearity due to the combining of the subsystems is presented by the new 11 coupling 11 reactions, which are not required when the subsystems are considered separately. Two subsystems are first desc ri bed and then combined to describe the general reaction system. The nomenclature used is the . same a~ in the computer program, and is given in Table 7.1. Figure l shows the reactions in the subsystem hydrocarbon-0 2-HBr (denoted NOSC). The heavy lines describe reaction paths; the broken lines are connection lines of species that take part in several reactions. Data for sev- eral of the reactions in NOSC were found in the literature, the values of the rest of the rate constants were estimated theoretically and modified as required to fit the experimental data. Figure 2 shows the reactions in the subsystem hydrocarbon-0 2sulfide (denoted NOSC). Since many of the elementary steps of NOSC were known, a detailed mechanism was used in the modeling of NOSC. However, in order to keep the nurrt>er of equations down, several rate steps in the mechanism of NOCAT were combined. The details of the approximations are given in the next section~ the logic of the lumping is given in Section 7.6 . 7. 3 The Interactions of the Sulfur Compounds in the Subsystem NOCAT Sulfur compounds (SC) like butyl sulfide (BuS) interact with hydrocarbon-oxygen systems in three ways: 1) lose a hydrogen to a free radical, usually to R0 2· , which is the most abundant one; 2) reduce R0 2 - to RO • and form sulfoxide ; 3) react with the cyclic peroxy -117- complex and form sulfoxide and alcohol. The interaction of R0 2• with the hydrogens of the SC is very similar to that of the reaction of a hydrocarbon with R0 • (see Sec2 tion 2.2). The result of the initial oxi dation is the formation of an alcohol from the s ulfide (e . g., a-hydroxy butyl sulfide). A sim,lar mechanism converts the hydroxy sulfide to the corresponding ketone, e.g . , a-keto butyl sulfide . a- HYdroxy sulfides and a-keto sulfides are unstable and may oxidize or decompose . The compound a-hydroxy sulfide is a very unstable material, and it decomposes rather rapidly to an aldehyde and a mercaptan. The rate determining step for the oxidation of the side chain is the hydrogen abstraction reaction: ( 7. l ) Therefore, the rate of (7 . l) determines the rate of production of c3H7-0H-S-C 4H7 (denoted COHS). Since the a-hydrogen is extremely re- active to abstraction, the a-hydroxy sulfide is rapidly converted to the a-keto sulfide (denoted COS). If the decomposition reaction of COHS to aldehyde and mercaptan is neglected, then the rate of hydrogen abstraction from the sulfide determines the rate of production of COS. As a result, one may write a single equation for the rate of formation of COS from the sulfide: ~ -C-S- + RO • I 2 H (7 . 2) -118Table 7.1 Name Abbreviations of Species Names Used Specie Name Specie ALO Aldehyde, c3H7cf~ PRAC HBR-R02H complex BR Bromine radical R Tetralyl radical BR2 Bromine molecule REN l ,2 Dihydronaphthalene CIO Butyl alkoxyl radical RH Hydrocarbon CO Ketone (tetralone) RO Tetralin alkoxyl radical C02S a-peroxy butyl radical R02 Tetralin hydroperoxy radical C03H a-hydroxy hydro peroxide ROH Tetralol COHS a-hydroxy sulfide R02H Tetralyl hydroperoxide COS a-keto sulfide S Butyl sulfide, or sulfur compound COSO a-keto sulfoxide SO Sulfoxide CS a-butyl sulfide radical S02 Sulfone DICO Tetralin diketone S00 a-peroxy sulfide FROH a-tetralol radical SOOH a-hydroxy sulfoxide HBR Hydrogen bromide SOS H20 Water Butyl hydroxy mercaptyl radical H202 Hydrogen peroxide 02 Oxygen -119- 4 I I ~ ROH + o 2 + CO H202. f II kHBr).lr •~ Rr I fgo3~-teHO+ed I REN +H20 iHsri~- - - ~ - - _J -----► CO+H20 Figure 7. l The Oxidation of a Hydrocarbon in the Presence of HBr. The nomenclature is explained in Table 7.1. - RO• J oll l '----1.. COHS + H2O + R· A· \ R \ \ - 17 t - •- • [OJ [ ( '---- RSH+ALD • OXIDATION l [ol► J UNSTABLE SPECIE CATALYST NOT s1MuLATE'o RSSR +H 2 O ~ R~SR ~R~SR [\RS03H+IBS08] I_ ..L. T . o" It\ (.U - - <}- - RSSR + H20 Figure 7.2 The Interaction of an Aliphatic Sulfide with the System Hydrocarbon -02-HBr. The nomenclature is explained in Table 7. l. A· is the sum of all the radicals . -121- The rate constant for the overall reaction will be approximately the rate of abstraction of hydrogen from S. If the concentration of o2 and RH do not vary much during the experiment, their value can be incorporated into the rate constant . This modification simplifies the mathematical manipulation of the set of reactions. Similar arguments were applied to derive the rate limiting equations for the formation of the a-hydroxy sulfoxide (SOOH) from the sulfoxide (SO) by reaction H 01··0 I II ( 7. 3): ROH+ -C-S2 (7.3) I H Note that the a-hydrogen of SOOH has larger bond energy than that of the a-hydrogen in COHS because the formation of the cyclic intramolecular complex. The rate of reaction of SOOH to the a-keto sulfoxide (COSO) is of comparable order to that of the reaction of SO to SOOH, and therefore they had to be separated. observed experimentally, This stabilization effect is SOOH accumulates in the initial period of the reaction while COHS does not . At room temperature the rate of reaction (7.4) is very small, and therefore the oxidation of sulfide is not influenced by inhibitors of free radical reactions. I I I I RO . +-c-s-c- ➔ 2 0 I II I RO· +-C-S-CI I (7.4) However, at higher temperatures, and when the concentration of R0 2• is relatively large, its effect may be sizable. The product of reaction (7.4), the sulfoxide SO, does not react with R0 2• to the sulfone so 2 . -1227.4 On the Coupling Reactions of Subsystems NOSC and NOCAT The coupling reactions between subsystems NOSC and NOCAT are : l. Hydrogen abstraction from S by Br : S + Br · 2. CS+ HBr + ( 7. 5) Reduction by SO of HBr to water and Br 2SO + 2HBr S + H2O + Br 2 + ( 7. 6) 3. Oxidation of the SC by the peroxi de-HBr complex, PRAC, e. g., S + PRAC + SO + ROH + HBr • Reactions (7. 5) and (7 .6 ) are undesired reactions. (7 . 7) Reaction (7.5) ini tiates a sequence of reactions whi ch ends in the oxidation of the hydrocarbon part of the SC, and therefore in loss of energy . Equation (7.6) shows that HBr can reduce the . sulfoxide back to the sulfide. It is therefore the exact opposite of the desired reaction. Moreover, Br is formed which reacts with the hydrocarbon to the organic bromide. 2 The net effect of this reaction is the formation of brominated hydrocarbon and the loss of catalyst. This adverse effect is discussed further in Section 7.8. 7.5 Computer Simulation of the Reaction System The methodology of the simulation was as follows: l. Assume a reaction mechanism . 2. Obtain a set of rate constants for the individual steps of the mechanism . -1233. Integrate the differential equations with the initial conditions of the relevant experiment. 4. Plot and compare the curve with the experimental data. 5. Correct the mechanism, the values of the rate constants, or modify the initial conditions to obtain a better fit. 6. Repeat steps 3-5 until a good match is obtained. 7. Do steps l-6 separately for subsystems NOSC and NOCAT. 8. Combine the results of step, add coupling reactions and repeat steps 1-6 for the combined system CATSC. Because of the size of the program and the limited resources avail- able, the match between the experimental data and the computer prediction was done manually. Introduction of automatic search for the rate con- stants which fit the data best is recommended for future studies. 7.6 Simplifications Used in the Simulation and Their Justification Technical limitations do not allow the use of models with too many differential equations or species. Moreover, too large models can- not be easily utilized and therefore their value is limited. In order to obtain smaller systems which will still be useful, it is necessary to reduce the number of differential equations that are used. One way to reduce the size of the model is by lumping reactions together. A general discussion of lumping was made by Aris and Gavalas (A-9) and by Wei and Kuo (W-5). Only a brief presentation will be given here. Two kinds of lumping are distinguished: 1) lumping in series; and 2) lumping in parallel . Lumping in series can be used when a series of -124- reactions are considered in which one step is clearly rate limiting, e. g., in the sequence of reactions A+ B + C + D. reaction A+ B is limiting If the rate of the the rate of conversfon of A to D, the reactions can be lumped to the single reaction A+ D, with the rate constant of the reaction A+ B. By the use of such a simplification, two components and two reactions were eliminated. Such a simplification may lead to erroneous results if components are added which react with the eliminated species or if the conditions of the reaction change. When many compounds which belong to the same homolog group react in the same way, with the same reagent, instead of considering the concentration of each specie separately, the sum of their concentrations can be treated as a single specie. For example, a mixture of esters of fatty acids which are hydrolyzed in a basic solution can be treated as a single ester. A special case of parallel lumping involves lumping of functional groups. For example, all the primary hydrogens can be lumped together, or all the olefinic bonds can be lumped into a single II reactant". Such a lumping improved the mode 1ing of the hydrogenation of hydrocarbons tremendously (J-1 ). Another special case of parallel lumping is the treatment of many free radicals which react in the same way, as a single specie. This approach is implicit in the classic treatment of oxidation when the SS approximation is used . The advantage of lumping in parallel is a great reduction in the number of species in the system; however, the mathematical treatment of a system of compounds, some of which belong to several functional groups, may be difficult . The activation energy of reactions -125- of similar functional groups may be very close. However, the frequency factor of homologs with a wide distribution of molecular weights may be very different. In this work, the secondary hydrogens of the butyl sulfide were lumped together in order that treatment of many isomers will not be required. 7.6. l Rate Data for Individual Reaction Steps The rate data that were used came from three sources: 1) from the literature, 2) from semi-empirical correlations, and 3) from the computations of the "best fit 11 for the experimental data. Whenever reliable data were available from the literature, they were used with no change. Data that were estimated were changed within the general range of values for rates of similar reactions. The rates of steps which were not found in the literature were manipulated as necessary to obtain the best fit. The rate data taken from the literature are given in Table 7.2. The following methods were used to estimate rate constants: l) Benson's Method for the frequency factor of a reaction in the liquid phase. Benson (B-15, p. 506) derived an approximate formula to predict the frequency factor for liquid phase reactions if their value in the gas phase is known. This method can be summarized as follows: If the reaction is first order the frequency factor is the same for the gas and for the liquid. If the reaction is of m order, then -126Table 7.2 Rate Constants for Elementary Steps in the Oxidation of Tetralin by Molecular Oxygen The value of the frequency factor or (lit/mole) 2/sec, as required . k0 is in sec-l , lit/sec/mole , l og 10 E Reaction kcal/ mole 2RH + o2 ➔ 2R· + H202 2RH + o2 + 2R· + H2o2 k 0 ko Ref. 20.7 3. 4·10 3 3. 53 D- 7 23.8 2. 4 • 10 5. 38 C-4 24.4 9. 3 · 10 8 8. 97 37. 7 2. 4 • 10 13 13. 38 E-1 , p.75 R-2 5 Notes Used in t he program Adapted to liquid by Benson's method 7.83 B-13 R0 2 - + RH ➔ R0 2H+ R· 4.5 6. 76 • 10 7 2.5 • 10 4 4.4 B-13 R0 2 • + RH ➔ R0 H+ R· 2 8. 3 5.62 • 10 6 6.75 H-6 2R0 • ➔ CO+ ROH+ o 2 2 2R0 2, ➔ Non Radicals 6.61 • 10 6 9 4. 3 8. 51 • 10 6.82 H-6 9.93 H-7 2RO 2 • ➔ Non Radicals 0 2 • 10 7 7.30 B- 13 In tetralin 2R0 • ➔ CO+ ROH+ 0 2 2 2R0 2• ➔ Non Radicals 2. 6 5. 5 • l 0 8 8.74 B-13 In CH 30H 0.4 4.2 • 10 7 7.62 E-1, p. 321 B-13 Used in the simulation Negligible compared with 2R0 2• + non-radicals in the presence of excess o2 Adapted to liquid by Benson's method R• + o ➔ R0 2 2 0 0 6.85 2Br· ➔ Br 2 0 5.19 10 16 16.58 K-2 Used in the program -127- 10 2m-2 rn- 1 me For binary elementary steps Aliq/Agas ~ 20. Benson's approximation is valid if the activation energy in the gas is the same as in the liquid, and if the solvent is non-protic. 2. Pola_yni - Hammet Approximation for the Activation Energy of Reactions of Homologs The Polayni approximation can be used to estimate the rate and frequency factor for merrt>ers of a homolog group which reacts with the same reagent. Data for a few members of the group have to be known. The method can be summarized as follows: 1 ) For a homolog group that reacts in the same way, 1og A = liSt /R a: Ea where A is the frequency factor of the reaction, tist is the entropy of formation of the ac- tivated complex, and Ea is the activation energy of the reaction. 2) The activation energy for members of the same homolog group, which reacts with the same reagent, is linearly related to the heat of their reaction. 3) The heat of the reactioll.S.an be predicted from bond energies since it is a thermodynamic property. 7.6.2 Initial Values and the Concentration of o2 and HBr in the Solution The initial values for the major components were determined by -128the way which the $Olution was prepared. Minor components, e.g. , REN, ROH, CO, etc. were either determined experimentally or given a value below the detection level of the analysis method. The data for the solubility of o2 in Tetralin weretaken from reference (C-4) and fitted into a modified form of Henry's law . The equation obtained is: (7.8) PO is the partial pressure of o2 in mm Hg, [0 2] is the concen2 tration of o2 in the liquid i n mole/ l ite r, T is the absolute tempera- where ture in oK. Equation ( 7.-8) describes the data of Carlsson and Robb (C-4) with accuracy better than 2%. The data were for mm Hg and l00°C < t < 140°C. 130 < P0 < 700 2 However, it is believed the equation can be ' safely used in the temperature range of 80 to 160°C. The numbers yielded by the equation are of the order of 10- 3-10- 2 mole/liter and are consistent with the data that were reported in Ref. (E-1), p. 94, for various hydrocarbons. The solubility of o2 in nonpolar hydrocarbons does not vary sharply with the temperature. Therefore, the values that are ob- tained from equation (7.8) were used to give an estimate for the initial value of [0 2] and [HBr]. The concentration of o2 and HBr was assumed constant throughout the reaction period and equal to the equilibrium concentration. This assumption is reasonably accurate for o2 because fast mass transfer from the gas to the liquid is possible (see Section 5.4). The partial pressure of HBr was naturally considerably smaller than that of the oxygen, but can provide adequate mass transfer to -129- maintain equilibrium concentration for the catalyst. However, after a sufficient amount of oxidation products is formed, reactions in which HBr is rapidly consumed become important. If a large concentration of HBr is used initially, then the concentration of oxidation products • will become significant after a short period, and the rate of consumption of HBr may be of comparable magnitude to the rate of HBr mass transfer. fails. In this case, the assumption of constant HBr concentration The conclusion is that the adequacy of the assumption that the concentration of HBr is constant is a function of the concentration of the oxidation products and of the initial concentration of HBr that was used. 7.6.3 On the Computer Programs and the Manipulation of the Data The complexity of the sys tern warranted mechanical manipulation .of the data by the computei!- Handling the problem was split into two programs, INTREACT and FILE. INTREACT is composed of 1. An integration module, capable of integrating stiff differential equations. ( Geer's method was used.) 2. An input module which transforms a given reaction mechanism into the form of differential equations.sets the initial conditions, and accepts the control parameters for the integration and the 1/0. 3. An output module which stores the values of the concentration, the instantaneous Jacobians and of a few other parameters on disc files for further manipulation. It is also possible to obtain a printout of the rate of the individual reaction steps, the concentrations of the species, -130- and the rate of change of individual species at any instant of the reaction. Program INTREACT is a modification of program MODKIN which was developed by Whitney (W-3).* The differences between MODKIN and INTREACT are: 1) INTREACT produces a disc file of the data which allows further manipulation of the results of the integration, e.g., calculations of the eigenvalues of the instantaneous Jacobian, algebraic manipulation of the data, etc. 2) INTREACT accepts the rate data either in the fonn of activation energy, frequency factor and temperature, or in the form of given constants and the temperature. Program FILEt was built to manipulate data. FILE performs the following operati \ ns: l) Plots on the Calcomp in a variety of ways data that were produced by INTREACT or experimental data. 2) Performs a variety of algebraic operations on data files (experimental or from INTREACT) and p·lots the results. 3) Calculates the eigenvalues of matrices that were produced by INTREACT and plots the resultant complex numbers, or their functions, as a function of the reaction time, as a function of the concentration of selected components, or vs other parameters. (This part has not yet been tested.) *The author wishes to thank Mr. S. Sanders for letting me copy his deck and helping me get acquainted with MODKIN. tThe author wishes to thank Mr. Glenn Wood for his help in developing this program. -131- 7.6.4 The Steady State Assumption in the Simulation Its Limitations and Application Modeling the kinetics of a chemical reaction amounts to the integration of a set of differential equations with given initial conditions. For a complex system the mathematical treatment becomes unmanageable because the differential equations are often stiff, coupled and nonlinear. The steady state (SS) approximation can reduce the number of differential equations and thus simplify the mathematical operations. Nonetheless, it is an approximation which should be used with caution. The SS can be applied to every intermediate which can be consumed at a much larger rate than it is formed~ If the ratio of the two rates is large, the absolute concentration of the intermediate will be small, and hopefully its rate of change will be virtually zero. A typical ap- plication is to reactive intermediates like free radicals. In the fol- lowing example the validity of the SS in the initial period of the simple autoxidation reaction is discussed. The reaction mechanism can be described by: 2RH + o 2 R· + o2 ko 2R· + H202 + kl R0 2• + k2 R0 • + RH 2 + k6 2R0 , 2 + R0 H + R· 2 ROH+ CO+ o 2 When large excess of o2 and RH exist, the concentrations do not * If the intermediate can be formed by a branching reaction, its concentration may oscillate. The SS approximation fails in this case. -132change appreciably; thus the rate of initiation w0 in the beginning of the reaction is approximately constant: (7.9) The rate of change of [R0 2-J is The rate of change of [R•] is (7.11) The sum of equations (7. 10) and (7. 11) is (7. 12) Since k1 is very large and the concentration of oxygen is then [R0 2•J >> [R·], and equation (7.12) becomes also large, (7.13) Let [R0 2 -] \ = 0 then: t=O t = The SS value of [R02·J is of its SS value, t 0 _99 , is (7.14) ✓wo/k6 and the time required to achieve 99% -133t = log 199 = 2 •65 l (7.15) ✓k 6 w 0 2~ ForTetralin at 140°C, [0 J'\.,10- 2mole/liter~ 2 1og 199 ta. 99 = _ _ ___,_..__ _ _ _ _ _ = 2. 07 sec 2/ 6 . 35 • 10 - 8 • 2. 56 • 10 7 The computer simulation shows that R0 2- reaches its SS value after 50 min (Figure 7.3). The SS gives valid approximations if the intermediate is rapidly consumed, slowly produced, and autocatalysis does not occur. However, if the autocatalysis is important, the concentration of the free radicals may oscillate. Apparently the branching is not negligible in the case of autoxidation, and therefore the use of the SS approximation is undesirable. Five models for the kinetics of oxidation of a hydrocarbon with molecular oxygen were developed using the SS. The models are: Auto- oxidation of a pure hydrocarbon, Model 1 with a homogeneous catalyst, Model 2 with a retardant which reacts via a non-radical mechanism, Model 3 with retardant that reacts via a free radical mechanism and a combination of Models 2 and 3. The SS models are too simple to enable the quantitation of the numerous sulfur containing products, therefore numerical integration was used to account for them. A full discussion of the SS models is given by Attar (A-7). Algorithms for the numerical integration of stiff differential equations can handle problems with a limited stiffness only. It was found that if the two reactions k(140°C) = 2.68 10-? (7.16) -134and k(l40°C) = 3. 79 • 10 16 (7 . 17) are incorporated as part of the set of differential equations, an extremely unstable system results. When the SS i s applied to the Br· radical, the system becomes stable and manageable, because its stiffness is drastically reduced . The SS is valid i n this case because the Br· the ratio of the rates of (7. 17) and (7 . 16) is very large, .is not produced by an autocatalytic step, and the time required for the [Br·] to achieve 99% of its SS value is approximately 5 • 10- 5 sec . Since the rate of production and the rate of consumption of Br· depend only on the concentration of HBr and of o2 , and these are constant, once the [Br·] achieves its SS value it maintains it . throughout the reaction period. Therefore in the computer simulation of the system, the Br· radical was treated as a reagent with a constant concentration [Br·]= . ✓km[0 2 J[HBr] - ~ - - - = constant kt where km is the rate constant of reaction (7. 16) and constant of reaction (7.17). (7.18) kt is the rate The SS approximation fails when sufficient oxidation products are formed because in this case the rate of reduction of HBr becomes of comparable magnitude to the rate of mass transfer of HBr. -135- 7.7 The Results of the Computer Simulation The following programs are presented: 1 . EXPERIMENT 1 , which simulates subsystem NOSC (denoted EXP 1) 2 . EXPERIMENT 3, which simulates subsys t em NOCAT (denoted EXP 3) 3. EXPERIMENT 4, which simulat es CATSC, a system combined from NOCAT and NOSC (denoted EXP 4). The computer listing of the reactions that were integrated in EXP 1 is given in Table 7.3. in Table 7. 4. The listing of the reactions used in EXP 3 is given The coupling reactions that were added to combine EXP 1 and EXP 3 are given in Table 7.5. The coupled system, CATSC, was very large and technical limitations did not allow inclusion of all the reactions that may occur . As was anticipated, the theoretical predic- tions do not match the data of EXP 4 very well. Of particular interest is the reaction between the sulfoxide and the HBr, which reduces the sulfoxide back to the sulfide: The rate constant of this reaction is large and is extremely sensitive to the concentrations of HBr and H o. Thus the concentration of HBr determines 2 not only the rate of oxidation of the hydrocarbon and the sulfur compound, but also the rate of reduction of the sulfoxide back to the sulfide: s RH,o 2 HBr I • so t RH,O 2 HBr > ► HBr ► so2 s i i ENERGY LOGlu(KO) Ro CCNSTo LIST OF REACTIONS ---- -------·- --···•· ·------ ·-- - - -- - - - -- 1 Ri R l:l 0 RC 1.00 t-0 loOO H202 loOO __ CO PROCUCTS HBR = l.,JO H20 loOO BR = 1.00 1-1202 loOO BR 1-202 CJ = loOO CICO 2o00 H20 H202 H,u2 = loOO H8R 2o00 H20 Hu HJ.;. 1-iBR HBR R02 02 RCH HC2 Ri.J = l .,QO f<C2 l oOO 02 = l a OO R at:Llo...0..0_.6.R = 1.,00 ROH loOO R Rrl -= lo00 l-'2.02 lo00 R RJ, = loOO FROH loOC R02H F";JH = 1 oOO CO loOO H02 R..Jt... = lo0-0 RCH loOO CO 1 o00 C2 Rj" rl = 11100 PRAL ~.!" RH R [2 tiBB, HO2 lo 00 t,Q 1~00 H20 R02H lo OO R H 202 C.,i =__J._._Q_Q_ CC3 H HBR R.J t1 = lo0O REN loOO H20 loOO HBR HeR R"'~ H_ =. l e00 CC lmOO H20 . . l .oOO HBR RC2H R.. '- H = loOO CC loOO ROH loiJO 02 BR RH = loOO R loOO HBR CJ,;> = 2o 00 H = 2 o. 00 Ru~H = 1 .. 00 H2O2 H,~2 = loO-0 RH = loOO R'"'~ RH RE A.,, ANTS Reactions and Rate Constants for the Subsystem RH-0 2-HBr (NOSC) Oo~40:: 0 1 Ool35E-07 02 Oo<;50E _01 Oc.e23E-Ol 2 Oe200~ 02 Oo 8S7E 01 Coll4F-03 3 . Oo244E 02 0.., <;84E 0 1 Oo216E-O5 4 Oo2S3C: 02 C.,R30F Cl Oo 5 7 5E Cl 5 Oe228E 02 Cot 3 lf-0_2 _________________ _ ___ __ 6 ___ 0o0_________ - Oo220E _0 1 7 Ool55E 02 OoS60-: 01 () 0 2 5) F. 02 Co 205 ~ 0_2 8 Oo l Z_OE O 2 o.. 1421: 02 9 Oo ll61: 02 Oo833E 01 C o l49E 0 3 Do 113E 1)2 0o550~ Cl Oo245f 09 10 00783!: 01 <1oct6E C8 11 Oo'-' 12 0 9 800F 00 __ Oo560i:: 1J l __ Col : tJ!;__c_t_ 13 C.,450F. Cl Do 820:: 0 l rlot58E C6 0 0 23 0E O 1 14 Oo820E 0 1 Oo<;clE 07 15 Co750E Cl Oo880f 01 Ooc77E 05 H: Oo 700f 1J l OoCI Oo lCfiE 08 17 ,) o 4 CO E OC Oo 7 6 2 t Jl r:lo 2~6E C'E ___ _ ___18 ____ Col 40 E .. 02 ___ Oo E50i: 01 -- Ool23E 02 1Jo516E ,:)c,470!: 0 1 19 Oo 820'.= 0 l 06 Do 700E () 1 Oo lCOE O 8 20 IJoO (10 0 Oo450E U l 21 Do316E 05 H202 22 OoO Oo c00t O 1 ColCO E 07 HRR r .. -- -- -- 1-- --0 o 2 3 8E O2 I Table 7.3 I __, I CT\ w --- -·- ·- Ro CCI\STo I 26 24 25 22 23 21 20 0 091 0:: Dl Oo800f 01 Co c5 0 E 01 Oo455 E (l l Ooc50:: 0 1 Co140E 02 Oo180E 02 n.c Ool62~ 02 Oo S60 f: 0 1 Oo l40E J 2 Oo2 00 E 02 Co2<'0E C2 '1o! 73E 0 2 Oo516E 06 Oo 29 0 E 02 Col5 0 E (2 Ooc;40:: 0 1 _18 __ 0 0 186!: _02 ___ Oo940= , ) l 19 Oo 220E C2 Oo 130:: 0 2 Oo355E 05 Oo2t OE 04 Oo4l3E 06 Ool23E 02 Do375E 00 Oo 363E 05 Oo228E 02 0 0 278,E 01 Co?t Of 00 Oo516E 06 Oo82 0f: 0 1 Do470f 01 ' 16 17 Oo 256E 0 8 Oo316E 0 .5 Co 4 70 E C1 ,J o762E 0 1 Oo450E 01 Oo 820E O l Ou C8. 06 07 03 05 Oo_lOO E 0 8 Oo l49F. J o676E Ooc58E 'Jo SclE Oo677E O~ t31E-02 Oo228E 02 Ooll4 E-03 Oo_2 l6E-05 _ 14 15 o.n Oo783E 01 Oo820E 0 1 Oo 8331: 0 1 Oo-;84E 01 00 515!_: 01 -Oo 220E u 1 Oo8g7E 01 Oo820~ 0 1 01 00880:: 01 -. 0 9 700 E O 1 0l 01 02 Co244E 02 Co2 S3E 02 Co 8 30~ 0 l o. = 2o0 0 R l oOO _ loOO = 2o0.0 J:W = lo OC R0 leOO = l eJ_:) HC2 lQOO = loOO RC2H loOO = _l_.Jl_O____ill:3J:L_ RCZH RJ~H = loOO CC loOO CZ R = lo 00 R0.2 R..; RH = loOO RCH loOO HC2 R.ri = loOO H202 loOO ROH RJt:: = loO C FROH loOO Rn _C,nH R.J ~ H H202 H2..J2 R.,;, RH _H.2_Q2__C.1 RH RE Al.. i ANTS R R l<OH R HO HO _ CO _ H202 loOO C2 1.o_o H-20 PROCUCTS s S..i H,u2 He;, .., s~JH 9 lc,00 R02H 1 oOO RO 1~00 H20 loOO t-20 ROH ----leOO lo 0 0 H20 loOO t-i20 = loOO SCOH loOO R = 1.00 cos □ 1.00 R lo OO H2C le00 H2 0 = 1 .00 cos loOO R = lo OO CG$0 laOO CO = loOO SC SIJ 2 = 1 0 0 SO2 RC2H RC2H S -= loOO SC s Ru~ = loOO SC s R02 R02 C03H C...1~ R02 R02H CZ _l-AOJLJi Etli<li-=---1-LO.O. CO R02 RJc. = loOO RCH leOO CO loOO 02 H202 h202 CJ = loOO CICO 2o 00 H20 RH Hu = l oOO R loOO H20 H..., ,_ RH = loCO R lo OO H202 CC3H S = loOO SO loOO CO loOO H20 CC3H SIJ = loOO SG2 l jp_OO CO 1-e__Q_Q__ r.2D 02 LIST CF REACTIONS LOGl O(KO) --- -··- 002 38!: 02 Oo540E 0 1 Oot 35E- 07 Oo 200E Q2 . Oo<i50f; OJ . 82~~- o L ______ . ---- -----------. - - - __ 6 __ c .. o _ 7 0 01 161: '1 o !) 8 9 Co 4 5C• ~ 10 0 o23 0E 11 Oo 75 0 E J 2 Oo0 13 Oo4 00 E 5 4 l 2 3 1-- ----- i FNERGY • ·· Table 7.4 Reactions and Rate Constants for the Subsystem Hydrocarbon-0 2-Sulfide (NOCAT) I __, I w ......., 0 . 200E 02 0.830E 01 co~~ sT. 0.298E-Ol R. 0.260~ 02 0.128~ 03 0.939F 03 o.3goE 03 ·-----•-·- ·-- -- ----------- --•--- · --------- --- -·--·------ -•·•·- ____ _,_. _ ___ --"~·------- ··-- -· .... -- RH HAR BR ~H = 1.00 C02S = cs so = 1.00 S = 1.00 H20 ~ 1 .. 00 RR2 ·- · ··-··-· 1.00 HBR 1. 00 H8~ 1 .00 HBQ, PRODUCTS CJ2S 1.00 SOOH 1.00 R CO2$ = 1.00 ClO 1.00 sos CO2S = 1.G0 SOOH 1.00 R S0'lH = 1.00 SO'J 1.0 0 HBR s D9- = -1 .oa sos- --1 .. 0-9 .---c 1-8-- ----- - . ClO 1.00 ALO 1 .. 00 R = 1.00 so -5 --- -- -=----h.-OB ---€-5 -- -l ··•-0-0- -HBR- 1.00 ROH 1.00 ROH -= 1.00 COSO 1.00 ROH = 1.00 S02 - ·- ----- ··--- - - --- Hf\R -- - ---- - · -----· ~2 0. ?50E 01 ·• ----4-3- ------0-. 2 OOf ---0-2-- -- -O.--9--5tt-f:--- O-l- --- --0.- ,(22.-3.:-0--l--44 0.830E 01 ·J .575E 01 0.228f 02 45 o. 0 0. 7-90f 01 0. 794E 08 0.120E 02 0.650E 01 0.900~ 01 O.ll6E 02 COS SO S REACTANTS PR AC PRAC PR AC , 36 0.169E 02 0.103E 02 0.23uE 02 L--- 3-7--- -0.-&0BE--Ol---- 0.---1--G-2-E- -B-2------ 0--. S2--'S~----G6----- ------------ - - -----BR - 38 0.0 0.800E 01 O.lOOE 09 02 RH 39 0.830F 01 0.575£ 01 0.228E 02 40 41 o.SOOE 01 o.lBOE 02 0.171E 02 34 35 LriGlO(KO} ~NF.P CY LIST OF REACT1CNS Table 7.5 Coupling Reactions, Model CATSC l co I _, w -139When water is formed, it separates as a suspension and extracts into it the HBr. Such effects were not included in the model, and therefore we have the poor agreement in CATSC. (See Section 7.8 and Chapt.3 for details . ) The following is a list of general comments on the simulation of a reaction in a complex multi-comp onent system: 1. The reactions in a multi-component system can be simulated using a theoretical model as a base. However, all the phenomena that occur should be accounted for in order to obtain a good match of the data. 2. Several steps of the mechanism can be lumped together and used with the rate constant of the rate-limiting step. This approach simplifies the numerical treatment of the model; however, it may lead to erroneous results when subsystems are coupled. 3. The steady state approximation can be used to reduce the stiffness of the system. 4. The adequacy of the SS approximation is questionable when autocatalytic steps exist. The concentration of free radicals may oscillate for a given system, when the initial values and/or the rate constants are changed. The oscillations can be traced back to free radicals that take part in more than two reactions. Full discussion of this observation is beyond the scope of this thesis, and will not be given here. In order to avoid the uncertainty of the SS approximation, all the free radicals except for Br• were treated as components of the system. -140- The implications of the results of the simulation to the system Tetralin-o -BuS - HBr are: 2 1) The mechanism of oxidation that was prop osed in Chapters 2 and 3 can be used to adequately simulate the reactions in the subsystems NOCAT and NOSC. 2) The presentation of the data by the model is adequate for the first hundred minutes or so, but it fails for more advanced times. Better presentation could have been obtained if the reactions of secondary products and the effects of the separation of H o were in2 cluded; however, a very large number of reactions and components would have been required in this case. 3) The concentration of HBr and Br• which are assumed affect the results very drastically. Small perturbations in the assumed initial concentration of ROH and CO change drastically the distribution of products in NOSC. 4) The role of HBr is dual at low concentrations (<~5 • 10- 4 mole/liter) and its effect is mainly as a catalyst to the oxidation of both the hydrocarbon and the sulfide. However, when HBr is present at larger concentrations, the oxidation products that are formed oxidize it to Br 2 and water. The last reaction is very sensitive to the actual concentration of HBr and H o. (See the discussion of the role 2 of HBr as a catalyst, Section 7.8). 5) Simulation of CATSC without the reaction SO+ 2HBr shows that more than 99.9% of the sulfur compound would have been (7.19) -141converted to the sulfoxide and sulfone . This implies that catalysis of the hydrogen abstraction step is a very efficient way to selectively oxidize the sulfide, but a catalyst other t han HBr should be used. 6) The failure of CATSC to represen t the data properly is believed to be 1:ecause it does not include the effects of water. Water extracts the HBr and the surface active sulfoxides, and reaction (7.19) is tremendously enhanced, since it goes via an ionic mechanism. Only representative samples of the results of the simulation are given. The list of figures and the most important results from them are given in Table 7. 6: Table 7. 6 A List of Figures from the Simulation with Explanation of Their Significance • * F 1gure No. Simulate Experiment No. Conclusion 7.3 1 The most important free radicals reach SS only after about 50 min. Note also that the assumption [ROz•) >> [R•) is valid. 7.4 1 The major oxidation products of the hydrocarbon are simulated within the experimental error in the first 'v 100 min. 7.5 1 Water is formed in a relatively large amount. The concentration of the oxidizers ROzH and HzOz increases and then levels, and remains about constant. !he concentration_of H o2 _ is not simulated very well~ 4 however, it is believed that a large experimental error in the polarographic method used did introduce some error. Neglecting the redox reaction between HzOiR0 2H and HBr could also have been the source of error. -142- Table 7. 6 (continued) Figure No . * Simulate Experiment No. Conclusions 7.6 3 The conversion to ke tone in the oxidation of tetralin in the absen ce of HBr and in the presence of S is considerab ly smaller. More alcohol was formed in 3 than in 1 , but considerably less ketone, the more advanced oxidation product . The simulation fails again for larger times. 7.7 3 The simulation of the production of SO is good in the first 100 min. 7. 8 3 The concentration of oxidizing species is considerably smaller in the absence of HBr . Compare Figure 7.5. 7. 9 3 The agreement of the theoretical curve with the data of production of secondary products which contain sulfur is good, although "lumped" rate constants were used in their simulation. *The scale of some of the components was blown up in order to show the behavior better. Better fit of the data can be obtained if all of the -rate constants are calculated to fit the empirical data. Fewer parameters are required but the: applicability .o.f . the model is limited to conditions close to those for which the correlations were made. The theoretical prediction of the concentration of sulfide in EXP 5 is presented in Figure 7.10. The equations which were used to obtain Figure 7.10 are given in Table 7. 7. Attempts to model the data of EXP 4 from first principles without taking the s eparation of water into account, and without the use of more ,1o200E Oo244E Q0 293E 1 0 830E OoO ? 6 7 Oo540~ Oo950f: Oa8G7~ Oo~84~ 0 0 575~ -00220~ ---[ OG 1 01 01 01 01 01 01 Ro CONST o Oo635E-07 Oo823E-01 Ooll4E-03 Oo216E-05 Oo228E 02 Oof31E-02 tit KO ) 02 I Oo155E 02 Oo960E bl - -0~250E 02· 8 1 o2 0 5E 02 Ool20~ 0 2 Ool42E 02 ~- ~ ~ll6f n~ Ooa33~ 01 O~l49E 03 [ 10 Oo550~ 01 0oll3= 0 2 Oo245E 09 11 ~on 00783= O! Oo676E 08 12 n 00 Oa560~ 0 1 Ool50E 06 0 BOO ~ I.._. _ _ _ ··-- ··--·- - - - - -- -- -- - - - - - , ' 13 '1o45f\1: <.n 00820:: 0 1 11 o658E 06 !4 Oo~30~ 01 Oo820E 01 Co961E 07 15 Oa7src 01 Oo880~ 01 Oo677E 05 16 ) 0 1) 0o 700::: \')l Oo lOOE ,')8 17 Oo4Cl0~ 00 t)o7621: 01 Q.,256E 08 18 0al40f 02 On850E 01 Ool23E 02 _ _ _ 19 ___ r'i; 410f. -(IT Oo 8'20;: 01 Oo 516E 06 --- -2~ 0o0 Oo700~ Dl OolOOF 08 21 0o0 0o450~ 0 1 Oo316E 05 H202 22 DoO Oo600~ 01 ColOOE G7 HBR 23 0 ol8fE 02 Oo940~ 01 no360E 00 24 ~ ol?OE 02 Oo800~ 01 O.,lOlE 00 _ _ _ _ 25____ 0ol86E 02 Oo940F: 01 Oo36i')E 00 26 Ool70~ 02 Oo850~ 01 Oo319E 00 ~7 Do220E 02 O~ljO~ 0 2 Oo228f 02 28 lolSOE 02 Oo830~ 01 Oo230E 01 29 r1~f93i:: 02 Oo960~ 01 Oo823E 00 30 0a200E 02 Oal32~ 02 Oo413E 03 31 Oo200~ O? Ool52= 0 2 Oo413E 05 32 Ool60~ 01 Oo820~ 01 Oo226F 08 33 -- 6~"ft5F. 02 Oo<ilO~ 0 1 Oo233E 01 3 4 5 i1 o2_3_EH'. 02 l 02 02 02 01 - - - ••• E N-E P Gy -- RJ;j loOO loOO so = loOO H2 □ 2 loOO BR = loOO CIC □ 2oOO H20 = loOO HBR 2o00 H20 = lo10 SO loOO CO = loOO SO loOO qoH = loOO S02 loOO CO loJO soz loDO QOH = loOO COS loOO R = loOO COS □ lo OO ROH C.J;, = lo r)O COSO T;cfo · c □ -SJ : loOO SOOH loOO R SJJH = loOO COSO loOO R H,J2 = loOO so loOO H20 H~J2 loOO soz loOO HZO H202 H202 C03H PRAC Cil3H PQAC H02 PRAC CC3H Q02 RQ2 S = = HJ, CJ H,~2 S S S.J SJ S CJ~ HB~ loOO 02 loOO H8R l~QO HBR loOO 02 loOO H20 PROO UC T S 02 H20 HBR H20 HBR H2C HBR -1.~00 H20 loOO H20 loOO H2G loOO loOO loOO loOO loOO loOO loOO - =---=l-'o-,O =-o=---H -2 -=---o -=-- --,l -o---cc 0--:::0~ 8--,,-R_ _ _ __ -H-J- = loOO loOO loOO loOO = loOO = = = = = loOO = loOO = loOO = loOO = loOQ = loOO = loOO = = loOO = loOO loOO H202 lo(10 co RO loOO HO H02 lo OO HO RC2H loOO R C03H - --REN loJO H20 CO loOO H2Q CO loOO ROH R loOO HBR R02 R02H loOO BR RCH loOO R H202 loOO R FROH loJO R02H CO loOO H02 ROH loOO CO PRAC = 2o00 R = 2o00 t-lC HBR- RJ~H R02H RJtH BR R;1 02 R HBR RJ, RH RJ H02 Ri ROH RJ, C2 FrL;H R02 RJ, HBR RJ,H HeR HBR R;1 C.J,H RJtH H202 H~J2 RH RJ, H202 CJ RH RE A; ; A NT S Table 7.7 Reactions and Rate Constants for the Fit of Experiment 5 ' ~ t • 0 0 0 ·o 0 0 N 0 0 0 ug u z oo w zD I- CI:::r O:o I- 0 0 z 0 I~ 0 0 ::Co w _, I H'l'OA (')CAA B(')N • TE TAAL! N 21 B GRAMS SULFUR C(')t<P(') UNO N(') NE CATALYS T• HBR PHBA=l49MM HG (')XIOIZER• (')XYGEN P02=591MM HG SOLVENT• BROMOBENZENE 100 GRAMS EQUILIBRIUM CONCENTAETJON OF• (')2•5.3 E-3H/L EQUILIBRIUM CONCENTAETJON OF•HBR•l .3 E-3H/L TEMPERATURE• 140 OG C EXPERIMENT NO. l TIME - MINUTES SC!l.E FACTOR • 10 2 Figure 7.3 Free Radicals vs .• Time, Experiment 1 L..------------------'-----------------~-----------------~------------------' 0.500 1.500 0.0 1.000 2.000 '-.o· _, . ' i_J 0 (D 0 0 0 "' 0 0 0 0 0 I ....... I ..i:,. ~ 0 0 - • 0 0 0 s 0 0 N 0 0 0 <Tl u z oo UO w zO I- cr a a::§? I- 0 .0 I--=' U: ~ zo I ~8 1 w _, '-.. 0 _, <D _a 1-0 w a:: c-- 0 0 0 0 0 a, 0 0) 0 0 - ~ 0 0 ". " y I .ODO " . Figure 7.4 " y " 2 .000 TIMF - MINIJTFS . V . " z V . y Major oxidation products " HYDROCARBON• TETRRLIN 21B GRAMS SULF UR COMPO UNO • NO NE CATALYST• HBR PHBR=l49MM HG OXIDIZER• OXYGEN P02=591HM HG SOLVENT• BROHOBENZENE 100 GARMS EQUILIBRIUM CQNCENTRETIQN OF• 02•5 .3 E-3H/L EQUILIBRIUM CONCENTRETION OF•HBR•l.3 E-3H/L TEMPERATURE• 140 OG C EXPERIMENT NCJ. co 3.000 . " REN TRLN lv . = E -zl z = E -1,. E SCALE FACTOR = IO 2 ~.000 u, I I __, ..i::,. 0 0 0 ~ s D 0 D D N D u~ OD z u w zO (ID a:D ;:: ,_ ": D zo I 11> D D ::Ea .....J w .....J~ '-.O >-o -o w a: r- D D 0 (X) D D 0 D D a> - C! D D 0.0 .__ • 1.000 • Figure 7.5 • • 2.000 TlME - MINlJTFS • • • Water and oxidizing species. • HYDROCARBON• TETRALIN 218 GRAMS SULFUR COMPOUND• NONE CATALYST• HBR PHBR=l49MM HG OXIDIZER• OXYGEN P02=591MM HG SOLVENT• BAOHOBENZENE 100 GRAMS EQUILIBRIUM CONCENTRETION OF• 02•5.3 E-3H/L EQUILIBRIUM CONCENTRETION OF •HBR•l.3 E-3M/L URE• 140 DG C EXPERIMENT NO. • • H202 = E -3 ■ 3.000 -- • SC!l.E FRCfCfl • 10 2 ij_OOQ ......I O"I I ~ 0 D 0 D D :=: D D "! D 0 0 u~ oo z u uJ zO er: I- "'• <Io 1-0 D - z 0 I"; Lo D _j uJ '- 0 _j~ -□ 1-0 uJ er: r-- 0 0 0 ~ 0 0 0 D D cr, - ~ D D 0.0 '""'- y Figure 7. 6 l .000 :r y 2.000 TIME - MINUTES y Major oxidation products y HYOR (j[AR 8(jN • TE TRAU N 218 GRAMS SULF UR CDMP(jUNO • BU TYL SULFIDE 6 .23 GRAMS CATALYST• N(jNE (jXfOIZER• (jXYGEN P(j2=629MM HG SOLVE.NT• BROHOBE.NZE.NE. 100 GRAMS EQUfLIBRIUM CONCENTRETION OF• 02•5.7 E-3M/L EQUfLIBRfUM CONCENTRETION OF•HBR•O.O E-3M/L EMPERATURE • 140 OG C EXPERIMENT NCJ. 3 y = E 3.000 TRLN = E co l v Ox SCIU fRCTOO = 10 2 ij _OQO I _, I ~ .j::> w er: a 0 0 0 0 ~ 0 0 0 N a 0 u~ 00 z u w er: ::, I- • zO a:a 1-0 D - z I"? ::Co ....J D w ,o -□ ....J~ 1-0 0 r- a a 0 a, 0 a "'0 a a - a a ~ 0.0 -1.000 Figure 7. 7 HYDROCARBON• TETRALIN 218 GRAMS SULFUR COMPOUND• BUTYL SULFIDE 6.23 GRAMS CATALYST I NONE OXIDIZER• OXYGEN P02;629MM HG SOLVENT• BAClMClBENZENE 100 GRAMS EQUILIBRIUM CClNCENTRETION OF• Cl2•5.7 E-3M/L EQUILIBRIUM CClNCENTAET!ClN ClF•HBA•O.O E-3M/L EMPE AATU RE • 140 OG C EXPE RI ME NT NCl . 3 2.000 3 .000 • V-1,. 2 ; E - = E - SC!t.f fRCTOA = 10 Major Sulfur-Containing Oxidation Products TIME - MINUTES so 502 SO ij , OQ0 I __, I 00 ~ 0 D u, "' 0 D D ~ D 0 N D D 0 OD uD z u UJ >-- =-: zo CI:D >-er □ ;: zo I :Co 0 __J UJ '-. O __J~ >--o -□ UJ a: .,..0 0 0 (Xl 0 0 0 a, 0 0 - ~ 0 0 0 .0 ,__ N □. 3 Figure 7. 8 1 .000 2.000 TIME - MINUTES Water and Oxidizing Species HYDROCARBON• TETRALIN 218 GRAMS SULFUR COMPOUND • BUTYL SULFIDE 6.23 GARMS CATALYST I NONE OXIDIZER • OXYGEN P02=629MM HG SOLVENT• BROMOBENZENE 100 GRAMS EQUILIBRIUM CCJNCENTRET!ClN OF• 02•5.7 E-3M/L EQUILIBRIUM CCJNCENTRETJCJN ClF •HBR•O.O E-3M/L EMPE RRTU RE • 140 DG C EXPERIMENT 3 .000 ~.000 I "'1.0" I _, 0 I HY09 □ CRRB □ N• N □. 3 TETRRLIN 218 GRAMS SULFU9 c □ HPOUNO• BUTYL SULFIDE 6.23 GRAMS CATALYST I N □ NE □ XIO!ZER, □ XYGEN PC2=629MM HG SCLVENT• BRCMCBENZENE 100 GARMS EQUILIBRIUM CQNCENTRETIQN OF• 02•5.7 E-3M/L EQU IL IBRIUM CCNCENTRETION OF •HBR•O.O E-3M/L EMPE RRTU RE• 140 OG C ::XPERIMENT 0.0 Figure 7.9 1.000 2 .000 TIME - MINUTES z • z z • 3.000 i • SCIU FFUOR • 10 2 SOOH C E - SCI Second Generation of Sulfur-Containing Oxidation Products ~~ i ~ ~ D 0 0 0 N 0 0 ! -::: t- u~I Oc : w '=; zO a: a a:;;! I- • l- El zo 0 I Ln LO ~□, _J co ' ~:1 ~! C, en 0 0 0 a, 0 0 0 0 ~. 000 I __, I 0 <.Tl w a: 0 lJ"! 0 0 ~ 0 ~ 0 0 0 0 N 0 u~ oo z u w zO a: :7 f-- • <Io f --0 - z 0 I 2:o 0 _J w '-. O _J~ >-o _o r-:0 0 0 0 0 0 OJ 0 en 0 0 - ~ 0 0 0.0 .___ TIME - MINUTES 2.000 e • e Figure 7.10 Sulfide and Hydrocarbon vs Time. simultaneously. 1.000 HYDROCARBON• TETRALIN 218 GRAMS SULFUR COMPOUND• BUTYL SULFIDE 6.23 GRAMS CATALYST• HBR PHBR=l 14MM HG OXIDIZER• OXYGEN P02=626MM HG SOLVENT• BROMOBENZENE 100 GRAMS EQUILIBRIUM CONCENTRETION OF • 02•5.6 E-3M/L EQUILIBRIUM CONCENTRETION OF•HBR•l.O E-3M/L EMPEAATUAE• 140 OG C EXPERIMENT N □. 5 3.000 . l • SCALE FRC TOA • IO E 2 The rate-constants were fitted e . TRLN ij,000 I __, __, (J'1 -152than 45 steps, gave results which deviated by up to two orders of magnitude, where secondary sulfur-containing products were concerned. The deviation is because when NOCAT was developed, many steps were lumped together in order to reduce the number of differential equations. Reactions that may occur between lumped species are not recognized when the coupled model CATSC is built; thus the picture is distorted.Attempts to model CATSC with several different lumpings were unsuccessful. The most important conclusions from the simulation of the oxidation with molecular oxygen of butyl sulfide in Tetralin at 140°C with HBr as a catalyst are: 1) It is possible to obtain models which will describe the subsystems RH-0 2-s (NOCAT) and RH-0 -HAr (NOSC) within the accuracy of the exper2 imental data. 2) When one model is considered, e.g., NOCAT, it is possible to lump several reactions together, and to use the rate constant of the ratelimiting step as the rate of the lumped reaction; however, when models are combined, lumped reactions should be "delumped", and considered in the structure of the new model. In particular, species that take part in many reactions cannot be lumped. 3) HBr catalyzes the oxidation of hydrocarbon and sulfides retard it. 4) More ketone is formed when HBr is present, while in the presence of aliphatic sulfide, the yield of the alcohol is increased considerably. A discussion of the use of simulation to predict changes in enthalpy was given in Chapter 4. Examples are given in Section 7.11 -153- 7.8 Evaluation of HBr as a Catalyst Catalysts affect the rate of a reaction in two ways : 1) by introducing a new and faster mechanism to effect the conversion of a selected specie; 2) by enhancing the rate-l i miting step of the prevailing mechanism. In complex systems, where many species compete for a given amount of reactant, catalysis is bound to change also the distribution of the products. The role of HBr in the system hydrocarbon-oxygen-sulfide-HBr is much more complex than catalysis of the hydrogen abstraction, which is the rate limiting step of the oxidation . HBr can act as an acid, can share its protons and form covalent complexes, and can reduce oxidizers like peroxides and sulfoxides. The dominant effects are a strong func- tion of the concentration of HBr, the concentration of oxidizers in the solution, and the amount of water in the solution. In the following sections a semi-empirical discussion of the role of HBr is given. The main object is to define a region of parameters like concentrations, conversions, etc., in which HBr could be used efficiently as a catalyst. 7.8.1 The Effect of HBr on the Overall Rate of Oxidation The major product of the oxidation of Tetralin is a-tetralone, CO , The extinction coefficient of CO at 310 nm is much larger than that of ROH and REN; therefore, the absorbance of UV light at 310 nm can be used as a gross measure of the overall rate of oxidation of the hydrocarbons. Aliphatic sulfides, sulfoxides and sulfones do not interfere at this wavelength. -154- 1.0 Abso bance at 310 nm _, 10 EXP 3 □ EXP 4 o EXP 5 6. 10 0 2 3 4 5 TIME - HRS Figure 7. 11 The Absorbance at 310 nm vs the Time, Experiments 3, 4, and 5. 6 -155- Two conclusions are drawn from Figure 7.11, which shows the absorbance at 310 nnn vs. the reaction time: 1) HBr increases the overall rate of oxidation; 2) HBr causes the appearance of an induction period. In Table 7.8 the rate data are Table 7.8 summarized: The Initial Conditions and the Effect of [HBr] on the Rate of Oxidation of Tetralin. Experiments 3, 4, and 5. Induction period (minutes) Rate of Formation of ketone (hr-1) Experiment [HBr] 3 mode/liter x 10 [S] mode/liter 3 0 0.1424 4 0.64 0.2848 48 0. 7223 5 1. 34 0.1424 78 1.5463 0.2908 Since the absorbance is linearly related to the ketone concentration, it can be seen that after the induction period ketone is formed according to a first order rate law. The rate of ketone formation was calculated assuming an overall behavior as a first order reaction. The ratio of the first order rate constants for Experiments 5 and 4 is 2.14, very close to the ratio of concentration of 2.094. When these observations are coupled with the actual values of concentration of tetralone, the following rate equation describes the formation of tetralone: d[CO] ~ -~= dt where t (7. 20) k [HBr] [RH] C is in hr, k C = 0.608(lit/mole)/hr at 140°C. A very crude estimate of the Arrhenius parameters A and E a of the rate constant can be obtained if the effect of the DBT on the rate of -156- Cone ntration of Sulf ne mole/liter EXP4 ° EXP5 6 2 3 4 5 6 TIME - HRS Figure 7. 12 Variation of the Concentration of Butyl-Sulforte vs Time Experiments 4 and 5. 3. No sulfone was found in Experiment -157- ketone formation in Experiment 8 is neglected. (Since the concentration of DBT did not change , this assumption is valid . ) In this case, the initial rate of ketone formation is about 1. 265 ± 0. 3 hr-l and the Arrhenius coefficients are E a = 13.02 kcal/mole . 6 A= 4.74 • 10 (lit/mole)/hr and The error in A and E can be up to 20%, but the values of A and E have the r ight magnitude . Note that the rate constant was not presented as the sum of the rates of CO production with catalyst and without it . The presentation as a sum of rates is required when isolated steps of the mechanism are considered. However, relation (7 . 20) is an empirical rate law in which many steps were lumped. The induction peridd that was observed when HBr is present is because the HBr catalyzes the decomposition of RO H and of H o . 2 2 2 by the reaction of HBr and o 2 Initiqtion is very slow, and it is believed that most of the initiation is by decomposition of RO H: 2 (7. 21) The RO H 2 produced when o 2 is introduced to the system reacts as fol- lows: HBr I H 20 + Br 2 + ROH RO H + HBr t 2 PRAC + CO + H2O + HBr (7 . 22) When sufficient PRAC is produced, equilibrium is achieved and an approximately constant [RO H] is obtained . 2 initiates the reaction . The RO H decomposes and 2 This mechanism contradicts the assumption that [Br] is at SS from the very beginning; however, the computer simulation -158- indicates that this is a valid working assumption. Moreover, the con- centration of R0 H does not vary much after the initial period. 2 7. 8. 2 The Effect of HBr on the Rate and Select i vity of the Oxidation of Aliphatic Sulfides in Tetralin HBr influences the rate and the s e lectivity of oxidation of butyl sulfide in the following ways: 1. It increases the overall rate of oxidation, as discussed in Section 7.8.1 . Th e increased rate of oxidation r es ults in an increased level of oxidants like R0 H and H o , and in turn in an increased rate of oxi2 2 2 dation of S by the homolytic mechanism (Chapter 3) . 2. It increases the activity of the radicals in the system. The free radicals react with Sin two ways: to oxidize the sulfur, H H 0 I I II RO • + P r-C-S-B u -+ PrC-S-Bu + RO · 2 I I H (7.23) H and to abstract hydrogen from the sulfide, H I R0 • + PrC-S-Bu-+ ROH+ PrC-S-Bu 2 I 2 I H and H I H I Br· + Pr-C-S-Bu I (7.24) -+ (7. 25) HBr + P~-S-Bu I H H The a-butyl sulfide radical I (denoted CS) reacts very rapidly with oxygen and ends finally as the corresponding ketone (COS), alcohol (COHS), a -keto sulfoxide (COSO) or a-hydroxy sulfoxide. OH I 0 II Pr-C-S-Bu I H Pr-C-S-Bu COHS cos OH 0 \ II Pr-C-S-Bu I 0 0 Pr-C11,-J!-Bu H SOOH coso -159The oxidation of the hydrogen on the carbon is an undesired process which results only in loss of energy . Such processes are para- sitic and cause a decrease in th e se l ectivity toward the desired products, the sulfoxides and sulfones . 3. It increases the degree of ox i dation of th e sulfur compound. When no HBr is present, the sulfur i s not oxidized beyond the sulfoxide; however, sulfone (S0 ) and other sulfur-containing, highly oxidized com2 pounds are obtained when HBr is present. The advanced oxidation of the sulfur can be due to one of the following : a) the fonnation with HBr of an oxidizing complex which has a large oxidation potential; b) the formation at a large rate of secondary and tertiary oxidation products, e.g., peroxy acids, which have much larger oxidation potential than the primary peroxides. The data do not provide a clear answer as to which mechanism dominates. Figure 7.11 shows the concentration of sulfone as a function of time for experiments 4 and 5. The so 2 appears approximately at the same time the induction period ends; this may support the theory that so 2 is formed by the reaction: SO + PRAC ➔ so 2 + HBr + ROH (See Section 7.8.1 where the induction period is discussed.) (7.26) However, other advanced oxidation products may have been the oxidizers, e.g., the hydroxy-hydroperoxide (denoted co H) which is formed only when sufficient 3 CO and H o are present: 2 2 co + (7.27) -160- co 3H is believed to have an oxidation potential similar to that of peracids . 7.8.3 Ionic Reactions and Their Rela tion to the HBr One of the products of the oxidation is water which, when present in large concentrations , can change th e course of the reaction. If water is produced at a slow rate, it evaporates out of the liquid; however, when a large concent rat io n of catalyst is present, the rate of production of water may be so large that it separates from the organic phase. The presence of water in this case shows as a turbidity in the system . Turbidity appeared in experiment 5 after about two hours. When water separates, the following changes in the course of the reaction occur: 1) the HBr is extracted from the organic phase to the aqueous phase; 2) the solubility of HBr in H 0 is so large that the 2 maximum mass transfer rate possible is still too small to replenish it; 3) the oxidized sulfur compounds e.g., sulfoxides, which have a strong surface activity, concentrate on the interphase between the organic and the aqueous phase; 4) ionic reactions can occur in the aqueous phase and on its surface. The most important ones are believed to be: a) Formation of acid-base complex (I) with the sulfoxide + - + SO+ HBr [S-0--H--Br] (7.28) I b) Reduction of the complex by an ionic reaction with an HBr molecule (7.29) The Br 2 which is formed reacts with the Tetralin or with REN. In both -161- cases the hydrocarbon is partially brominat ed . c) Oxidation of I by a peroxide-HBr complex ~ [SCr-HB r ] + ~- --O'o I> S0 Br.H/ 2 + 2HBr + ROH (7.30) Reactions (7 . 29) and (7 . 30) are parall e l ; howev er , (7.29) reduces the sulfoxide into unoxidi ze d s ulf i de , whil e (7.30) is the most beneficial reaction, since it converts the sulfox ide to s ulfone . Wh en the concentrat ion of HBr is large, water will be formed at a larger rate and will sepa rate a fter shorter t i me. However, if only a small amount of water is formed, it may evaporate with the gas stream and never r e ach the concentration r e quired to separate. · Fig ure 7.13 shows the concentration of SO in the solution in expe riments 3, 4, and 5. was not used . The concentration is the largest in EXP 3j where HBr It is smaller in EXP 4, where th e concentration of HBr was small("-' 6.4 • 10 -4 mole/liter). It is even smaller in EXP 5, where the HBr concentration was l arge("-' 1.34 • 10- 3 mole/liter). The SO con- centration becomes undetectable shortly after the water separates in the form of turbidity. A myriad of other reactions. can occur in the aqueous phase; however, the data available are not sufficient to evaluate their effects. 7.8.4 The Potential of HBr as a Catalyst for the Selective Oxidation of Alt.2.hatic Sulfides to Sulfones and Sulfoxides The most desired product of the oxidation is the sulfone, since sulfones decompose to hydrocarbons and so . 2 Less desired products are the -162- Concen Sulf mol of EXP. 3 □ EXP. 4 6 EXP. 5 ° ' 6 0 Figure 7.13 2 3 4 TIME - hrs Variation of the Concentration of Sulfoxide as a Function of Ti me , Experiments 3 , 4, and 5 6 -163- sulfoxides, which form a sulfide and sulfone molecule when pyrolyzed under mild conditions: (7.31) c H --CH CH + S0 1 + c H 2 5 2 2 4 10 gas (7. 32) Here so denotes gaseous sulfur dioxide. 2 The evaluation of HBr is made based on the following functions: 1. The "desulfurization selectivity", I/JDS, which is a measure of the fraction of sulfur compounds that can be desulfurized by a mild pytolysis: [S0 ]+0.5[S0]+0.5 [COSO] + O.S[SOOH] 2 (7.33) [S0 ] +[SO]+ [COSO] + [SOOH] + [COHS] +[COS]+ [S] 2 Define the numerator as DSO and the denominator as TS, then DSO TS = -- = 2. DSO [S] (7.34) 0 The "relative desulfurization selectivity", 1µrDS, which is a measure of the selectivity of the oxidation to products which can be desulfurized. DSO DSO I/JDS = TS - [ S] - [S] - [S] (7.35) 0 3. The "relative desulfurization selectivity for sulfone and sul- foxide", 1/JrDSO' which is a measure of the conversion of sulfide to desulfurizable species without loss of energy to reactions on the side -164- 10 EXP 3 D t.4=10 EXP 46 M=,o"' EXP 50 M=/.0 o, 4 2 5 6 TIME - HRS Figure 7.14 The Variation of the Relative Mechanistic Selectivity as a Function of the Time, Experiments 3, 4, and 5 -165- chain ljJrDSO = 4. [ so ] + 0. 5 [so] [SO ] + O.5[SO] 2 TS - [S] 2 [SJ 0 - [S] (7.36) The "selectivity of incorporatio n of oxygen into desulfuriz- able species", ijJDOS, which is a measure of the selectivity of the , oxidation to the sul f ur atom: ijJODS = DSO S[O]system (7.37) ijJODS is closely related to the energy penalty of the desulfurization . Since the latter is discussed in Section 7. 11, it will not be discussed here . 5. The "mechanistic selectivity", ijJMPR' which relates the concen- tration of sulfoxides and sulfones to the oth er oxidation products of the sulfur (7. 38) If it is assumed that SO and so 2 are produced predominantly by reac- tions with peroxides, while all the other compounds are products of free radical reactions, then ijJMPR is a measure of the relative importance of the two mechanisms . The author is aware that species like COSO and COOH could have formed also by a combined free radical and hemolytic mechanism. However, it is believed that basically ijJMPR represents correctly the relative importance of the two mechanisms. -166- 10 EXP D EXP 6 EXP 0 Rela Desu Sele -I 0 10 1/Jrds 0 0 2 3 4 5 6 TIME - HRS Figure 7.15 The Variation of the Relative Desulfurization Selectivity as a Function of the Time, Experiments 3, 4, and 5 -167In order to eliminate redundancy, and because our main interest is the mechanism of the reaction, only wrds and WMPR will be discussed . Figure 7.14 shows WMPR vs. the time of the reac tion for experiments 3, 4, and 5. The most important conclusions that can be made are: 1. Oxidation of the sulfide proceeds via a non-radical mechanism when no catalyst is used, however , when HBr is added, a large fraction of the sulfide is consumed by a free radical mechanism. Oxidation by a free radical mechanism proceeds not only on the sulfur but on the carbons as well. This observation was anticipated, since the function of adding HBr was to increase the activity of the free radicals. 2. The activity of the free radicals is a very strong function of the concentration of HBr. 4 Increasing the HBr concentration from ~ 6.4 • 10- mole/liter to 1.3 • 10 -3 mole/liter results in a decrease of more than an order of magnitude in the mechanistic selectivity. 3. In experiments 3, 4 and 5 the selectivity increases first and then decreases. The increased selectivity in experiment 5, which is observed after three hours, is attributed to the separation of colloidal water and a change in the reaction mechanism. vity is due to the formation of so 2 This increased selecti- which was not detected at earlier times. 4. An optimal residence time can be found for which maximum selec- tivity is obtained. This time is a function of the HBr .concentration. The relative desulfurization selectivity, WrDS' is of interest when the main goal is to estimate the optimal residence time for desulfurization. Figure 7.15 shows the variation of WrDS vs. the time, for mole/liter [S02] + [SO] I~ ,o- 0 7?stJ I / 7 180 3009' ,.o ~MPR x M IO EXP3 □ I EXP 4 l!l. I EXP5 o 10 M ./ 230!Sl 1206 / / / / / / / / / / ""/ / / / / / / 7 / / / / / / / / / / / Relative Mechanistic Selectivity / / / / D 1100 4 60llJ/ ~//J / j/ / / '>""''-'\ 345~' .~~- ?·u, ✓ • I The Sum of Concentrations of Sulfoxides and Sulfones vs the 1500" Figure 7.16 ,o-l- ,o-· 10 l IO""'I-- ,,;.,_ / "' 00 I I t-' -169experiments 3, 4 and 5. The most important conclusions from the plot are: 1. An optimal concentration of HBr exists f or which the desul- furization selectivity is maxi mized. This v alue is larger than 4 0.0 and smaller than 6.4 • 10- mol e /lite r. 2. The selectivity is a ve r y strong function of [HBr]. crease in [HBr] above 6 . 4• 10- 4 An in- mole/liter increases the overall rate of reaction of free radicals and therefore of Wrns· However, if so 2 is the only desired product [HBr] concentrations larger than 10- 3 should be used. The optimal conditions are a function not only of the selectivity but also of the reactivity . Figure 7.16 shows the phase plane diagram of [SO]+ [S0 ) vs. WMPR. 2 The time was used as a parameter; the arrows indicate the course of progress of the reaction. The most important conclusions from Figure 7.16 are: 1) At small conversions, considerably larger selectivity is obtained when no catalyst is used. 2) When catal- yst is used, the general trend of the selectivity is to increase with the increase in the conversion, but when no catalyst is used, the selectivity decreases with the conversion. A region in the phase plane exists in which larger conversions and selectivities can be obtained in the presence of a catalyst. was shaded. Mapping of this area into the subspace of [HBr]-time can be used to obtain the optimum operation conditions. 6.4 • 10- 4 This area When [HBr] ~ mole/liter at times larger than about 65 min but smaller than about 315 min, better selectivities and conversions will be obtained -170due to the presence of HBr. 1.8 • 10- 3 Note that~= 15 and [SO]+ is obtained after about 65 min when [HBr] [so 2 ] = = 6.4 • l0- 3mole/ liter, and after about 300 min in the absence of HBr. 7.8.5 Summary of the Evaluation of HBr as a Catalyst to the Selective Oxidation of Aliphatic Sulfides The following effects were observed due to the presence of HBr in the system: 1. The oxidation proceeds after a short induction period. 2. The rate of formation of ketone after the induction period can be correlated with the equation: E d[CO] dt k 0 The approximate values of k (lit/mole)/hr and 3. Rt [HBr][RH] e 0 and E around 140-160°C are A= 4.74• 10 6 E = 13.02 kcal/mole. HBr enhances the rate of formation of water and when the [HBr] is large, turbidity appears. When colloidal water is present, SO is rapidly reduced back to the sulfide. 4. so 2 is formed only when HBr is present. 5. HBr increases the total activity of free radical attack on the hydrocarbon part of the aliphatic sulfide. 6. A range of parameters exist for which better selectivities and larger conversion of the sulfide to the sulfoxides and sulfones will be obtained due to the presence of HBr., e.g., when [HBr] = 6.4• 10 -4 mole/liter the same selectivity and conversions are obtained as when HBr is not present after about 310 minutes. -171- It is believed that if a very small concentration of HBr is used, the reduction of sulfoxide back to sulfide will be minimized, since its dependence on [HBr] is at least second order. The HBr indeed enhances the rate of oxidation, howeve r, since it also reduces the sulfoxides to the sulfides when the goal is the oxidation of sulfur compounds. A more suit able catalyst for the hydrogen abstraction should be used. The use of other catalysts may help to eliminate the experimental error due to the extraction of HBr into colloidal water and due to the tendency of HBr to take part in ionic reactions. 7.9 The Hydrocarbon Medium and Its Effect on the Oxidation of Butyl Sulfide Three experiments are compared: EXP 2 with no hydrocarbon, EXP 4 with Tetralin, and experiment 6 with toluene. The initial conditions given in Table 7.9, the temperature of the reaction was 140°C except in experiment 6, where it was 136.5°C. Table 7.9 The Initial Conditions of Experiments 2, 4, and 6 Experiment Hydrocarbon No. Hydrocarbon Concentration =CRI-I mole/liter Sulfide CRH Cone. = cs Cs mole/liter HBr 02 Cone. cone. mol/lit mol/lit X 103 X 103 2 Butyl Sulfide 0.1423 0.1423 1 1.1 5.6 4 Tetralin 5 . 395 0.2784 19. 4 0.64 6.0 6 Toluene 2 .177 0.1424 15.3 0.9 5.6 The main goal of the comparison is to try to determine how the system would behave if other hydrocarbons are used instead of tetralin and -172toluene. The reader is reminded at this point that the rates of hydro - gen abstraction (HA) from virtually all the aryl hydrocarbons is between the rate of HA from Tetralin and toluene, and th a t the rate of HA is the rate limiting step of the oxidation . Th e kine tic results of the three experiments for the behavior of two of the products is compared for SO which is formed mainly by a hemolytic ox idation, and for COSO in the formation of which free radicals play a key role. 7.9.1 The Kinetics of SO and so 2 Formation in Different Hydrocarbons Figure 7.17 shows the concentration of SO vs. the time. The most important conclusions from the data are : 1. The concentration of SO in the first 2-3 hours can be corre- lated according to first order kinetics, after which it levels. 2. The curves for the first 2 hours can be almost superimposed when the ratio of hydrocarbon to sulfide is similar (EXP 4 and 6). The ratio of the concentrations of [SO] in the presence of a hydrocarbon to that when neat sulfide is reacted variesin the range of 10-30, but for the most part it is 10-20. 3. The concentration of SO when it reaches a steady value is about 9.8 • 10- 3 mole/liter when Tetralin is used, and about 10when toluene is used. 3 mole/liter The ratio of the two is 9.8, which is considerably larger than can be accounted for by the concentration of the original hydrocarbons. The following arguments can provide a simple explanation of the phenomenon. From Chapter 3 it is known that a larger oxidation poten- tial is required to oxidize sulfoxide to sulfone than to oxidize sulfide to sulfoxide. Secondly, if it is assumed that both toluene and tetralyl -173- Conce tration of S lfoxide m e/liter EXP 2 o EXP 4 6 EXP 6 □ 0 2 3 4 5 TIME - HRS Figure 7t l7. The Effect of the Hydrocarbon Medium on the Formation of Sulfoxide. 6 -174hydroperoxides can oxidize the sulfide to the sulfoxide, then the similarity of the initial rate of formation of SO is clear . However, since larger oxidation potential is required to convert the SO to the sulfone, a difference in the rate of removal of SO may be observed when oxidizers with different oxidation potential are used . is Tetralyl hydroperoxide more stable than tolyl hydrope r oxide, and therefore is a weaker oxi- dizer. The arguments that were presented can be used to derive the fol- lowing rate equations: Let the rate of change of [SO] be controlled by equation (7.39) : d[SO] dt where P (7 . 39) is an unspecified oxidizing specie, k f and k the lumped rates of formation and consumption of SO. that [S] and P are respectively C Let us also assume are approximately constants*, so [SO] TsT = kf k -k pt (1 - e C ) (7.40) ➔ 00 (7. 41) C antl after long enough time, [SO] Ts1 t If two different hydroperoxy species oxidize the sulfur compounds, then (7.42) *From the computer simulation [RO 2H], [RO 2 •], and [H 2o2 ] reach approximately constant level shortly after the beginning of the reaction . The effect of the different [HBr] was neglected. -175Several approximations lead to an interesting corollary: 1. Assume that the rate of formation of [SO] is a weak function of the hydrocarbon, namely, kf Then 1 [S0] [so] 1 2 [S] 2 (7. 43) [ S] l The similarity of the initial part of the [SO] accumulation curves in the presence of toluene and Tetralin support this assumption. 2. The elementary step in the oxidation of Sor of SO which re- quires the largest energy is the breaking of the oxygen-oxygen bond of the peroxide(~ 32 kcal/mole). The other steps involve hydrogen bonds or tautomerization (3-8 kcal/mole). The strength of the 0-0 bond is determined by the structure of R·; more stable R· will form more stable peroxides (B-16). However, when R· is more stable, the energy of the bond R-H will be smaller. If the previous arguments are used with the Polanyi relation for hydrogen abstraction with a= 0.86 then (7. 44) A and A were taken to be approximately the same. 1 2 Substitution of numbers leads to 6H - 6H = 6H - 6H . = 1.54 kcal/mole 1 toluene tetralin 2 Considering all the approximations that were made, the value is surprisingly close to what one would anticipate for the difference between the two benzylic hydrogens: -176- EXP 2 o EXP 4 6 EXP6 □ [so ] 2 mole/1 10 3 TIME - HRS Figure 7.18 4 5 The Concentration of Sulfones vs Time, Experiments 2, 4 and 6 -177- toluene 'tetralin No value was found in the literature for the bond energy of the tetralyl hydrogen. For the benzylic hydrogen of toluene, The value of 6H 1 6H 2 ~ 85 kcal/mole. is therefore approximately 83.5 kcal/mole. The accumulation of so 2 is shown in Figure 7.18. The most impor- tant observation is that the concentration of sulfone does not continue to increase, but after sufficiently long time it decreas~s. crease in the concentration of so 2 The de- is attributed to the accumulation of secondary oxidation products with a much larger oxidation potential. The secondary products can oxidize the sulfones to the sulfonic acids, etc. Large enough concentration of oxidation products is never realized in experiment 2 where the only hydrocarbon source is the sulfide itself. Note that the so 2 is oxidized faster in toluene than in Tetralin, and that the value of [S0 ] reaches its maximum after about 210 min in 2 toluene and after about 375 min in tetralin. These observations are con- sistent with the previous assumption that stronger oxidizers are formed in toluene. 7.9.2 The Kinetics of COSO Formation in Different Hydrocarbons The principle that more reactive species will be less selective is clearly demonstrated when the accumulation of COSO in different media is considered. -178- 1cJr---r---r-~-,---,--,---T"'-.--------- □ □ □ □ □ □ □ I 0 o/ EXP2 EXP4 EXP6 0 I 0 0 2 3 4 5 TIME - HRS Figure 7.19 The Concentration of a-Keto Butyl Sulfoxide vs Time, Experiments 2, 4 and 6 6 -179Figure 7,19 shows the concentration of COSO vs. time in experiments 2, 4, and 6. Since COSO is a product of free radical reactions, its concentration provides information on the reactivity of free radicals. The most important conclusions that were derived from Figure 7.19 are: 1. COSO accumulates at the earlier stages of the reaction accord- ing to a first order rate law. k 2. = 1. 50 hr 2 -1 k 4 = The rate constants at 140°C are 0.494 hr -1 k 6 = 3.58 hr -1 The slope of the curve is 7. 25 times larger when the hydrocarbon is toluene than when it is Tetralin. The reason is again the large activity of the benzyl radical, which results from the larger C--tl bond energy in toluene. The more reactive benzyl radicals (or benzyl peroxy radicals) will not differentiate between the hydrogens and abstract all of them, while the more stable tetralyl radicals will preferentially abstract the weaker hydrogen on the Tetralin. 7.9.3 Summary of the Discussion of the Effect of the Hydrocarbon The hydrocarbon media influence both the oxidation potential available in the solution and the relative importance of the freeradical and the non-radical oxidation. The most important conclusions from the discussion are: 1. The oxidation of butyl sulfide to the sulfoxide is initially independent of the hydrocarbon, but the level of the equilibrium concentration it reaches is determined by the stability of the radical R·. When R· is more stable it will be possible to obtain larger concentrations of sulfoxides and sulfones. -1802. The bond energy of the tetralyl hydrogen is smaller than that of the toluyl radical by about 1.54 kcal/mole. 3. When the free radical R· is less stable, the medium is more reactive and more sulfur products of free- radical reactions are obtained. The rate of abstraction of a hydrogen from sulfide by a • 1 is • estimate • d to b eab out 7 . 25 • O._2784 = 14 . 2 times • b_enzy 1 ra d 1ca 0 1423 that by a tetralyl radical. The results show that as far as oxidation to the sulfoxide was concerned, no major difference could have been detected in the initial rate of oxidation when toluene or Tetralin were the hydrocarbons. The equilibrium concentration of sulfoxide was, however, about 10 times as large when Tetralin was used, and the sulfone concentration was 2-5 times. Since the reactivity of all the aryl hydrocarbons is intermedi- ate between that of toluene and tetralin, no major difference can be anticipated when these results are applied to a mixture of many aryl hydrocarbons, e.g., fuel. The parasitic side-chain oxidation of the sulfide by a freeradical mechanism change significantly from one hydrocarbon to the other, and may influence the final optimization of a process. 7.10 The Influence of the Protic Nature of the Solvent on the Rate of the Reaction Two experiments are compared, No. 4 and No. 7. The main differ- ence between the initial conditions of the two is the solvent used. In experiment 4 the non-protic solvent bromobenzene was used to modify the properties of the Tetralin. In experiment 7 1-octanol (0CT0L) was used to introduce protons to the system. The volume of OCT0L was taken -181to provide the same volumetric concentration of tetralin as in experiment 4. 7.10.1 The Effect of the Protic Nature of the Solvent on the Selectivity of the Oxidation Only four major SC products were detected in experiment 7, compared with some 25 compounds in the product of experiment 4. The major two components found in experiment 4 were SO and COSO. The selectivity of a reaction depends on the reactivity of the reagents; less reactive reagents are more selective. If this prin- ciple is applied to the oxidation of butyl sulfide, it can be inferred that the oxidant in experiment 7 was less reactive than the oxidant in experiment 4. A plausible explanation is that the oxidant in experiment 4 was a hydroperoxide dimer (I), while in 7 a complex of an octanol mole- cule and a hydroperoxide molecule (II) R I H····0 , 0 R-0-0... _H/ I a b II I Since two peroxy components are present in I, and only one hydroperoxide molecule in II, the oxidation potential of I can be expected to be larger than that of II. The oxidation potential of the peroxy- oxygens that are marked may be different than that of the peroxy- oxygens marked a b ; the fact that sulfone was not detected in experi- ment 7 supports the conclusion that larger oxidation potential was -182available in experiment 4 than in experiment 6, since larger oxidation potential is required to oxidize SO to so 2 than S to SO (see Chapter 3.) An interesting stipulation can be made on the mechanism of the oxidation in the two solvents. Figure 7.21 shows that the concentration of COSO in experiment 7 is considerably smaller than that of the COSO in experiment 4~ although the concentration of SO in EXP 7 is about three times that in EXP 4. cal In EXP 4 the whole gamut of products of free-radi- attack on the sulfide was found, while in EXP 7 only small concen- trations of COSO, a product of free-radical reactions, were detected; A plausible explanation is that the octanol solvates the free-radical R0 • and thus reduces its reactivity. 2 7.10.2 On the Interaction between Sulfoxides and Protic Solvents - Sulfoxides are bases and can interact with protons. In fact, the basicity of the sulfoxide bond was used to separate butyl sulfoxide from sulfone by ion exchanging (H-9). A crude estimate of the energy of the hydrogen bond of butyl sulfoxide and octanol can be obtained as follows: 1. The bond energy of the sulfoxide-octanol complex is the heat of reaction (7. 45): (7.45) 2. Reaction (7.43) is the difference of the following two reac- tions: SO+ 2HBr + S + H o + Br in bromobenzene 2 2 (7.46) + in octanol Lrn 0 (7.47) -183 ... • so coso EXP 4 EXP 7 • 0 Figure 7.20 6 o 10 M== 2 3 TIME - HRS . .A 4 5 The Concentration of Butyl Sulfoxide and a-KetoButyl-Sulfoxide vs Time, Experiments 4 and 7. 6 -184If the arguments that were used to develop equation (7.44) are used again, and the difference in the "activation energies" of reactions (7.44) and (7.45) are equated to the difference in the heat of the reactions, then = e -(6HB-L'iH )a./RT 0 (7. 48) steady value Let a= 1 then: 3 • 10 - 2 0.1292 6HB - L'iH0 0.2848 10-2 L'i¾"' -1.55 kcal/mole. 1. 987 • 413 "' e 1. 987 • 413 (7.49) The value is in the general range of energies of hydrogen bonds, 0-8 kcal/mole. The enthalpy of mixing of dimethyl sulfoxide and water at 0.1 mole fraction was found to be -0.352 . kcal/ mole (L-4). The value of the "enthalpy of interaction" of dimethyl sulfoxide and phenol is -6. 5 ± 0. 2 kcal/mole, and that of tetramethylene sulfone and phenol -4.9 ± 0.3 kcal/mole. Phenol is, however, much more acidic than octanol and therefore considerably stronger interaction can be anticipated. 7.10.3 Summary of the Effect of the Solvent on the Oxidation of Butyl Sulfide Protic solvents affect the oxidation of aliphatic sulfides in the presence of Tetralin, HBr and o in the following ways: 1) increase 2 the selectivity of the oxidation to the sulfoxide; 2) stabilize the sulfoxide by hydrogen bonding and thus allow more of it to remain in the solution; 3) reduce the reactivity of free radicals, in particular of -185R0 • by solvating it. 2 A crude estimate of the enthalpy of interaction between butyl sulfoxide and octanol is -1.55 kcal/mole. Considerably improved yield and larger rates of oxidation of sulfide to the sulfoxide can be obtained when the reaction is conducted in a protic medium, e.g., octanol. However, the increased selectivity comes with a smaller reactivity toward the production of sulfones. 7.11 Loss of Heating Value Due to the Oxidation Attempts to evaluate experimentally the loss of heating value by direct calorimetry has shown that the heat loss is within the experimental error. For example, the heat of combustion of the. initial mix- ture of experiment 4 was 8688 cal/gram and that of the product of the reaction was 8594 cal/gram, about 1.08% difference. The heat of com- bustion of the initial mixture of experiment 1 was 8504 cal/gram and that of the product was 8483 cal/gram, about 0.25% difference. The value of these results is questionable as a tool to evaluate the overall loss of heating value due to a very mild oxidation. However, for the purpose of this study it indicated that a very small fraction of the heating value was indeed lost. The method described in Chapter 4 was used to estimate the loss of heating value from the computer model. result. Figure 7.21 shows a typical The heat loss is very small, 2 cal/gr (2 kcal/liter) and is essentially negligible. The loss of heat increases when HBr or sulfide is added; however, the total value of lost heat remains negligible. a:: U.J 0 0 0 ;;P g 0 U) g 0 a, U.J CI: o • :co 1--- _J 0 (fl (fl a: u :.::: ...Jo ,"! _JO ~o 1--- 0 0 0 0.0 EXPERIMENT N □. 5 HERT L □ SS Figure 7.21 TIME - MINUTES 2.000 KCRL / LITER 3.000 SOU: ffl:TM • 10 2 The Loss of Heat of Combustion Due to the Oxidation, Experiment 5 1.000 HYDROCARBON• TETRALIN 218 GRAMS SULFUR COMPOUND• BUTYL SULFIDE 6.23 GRAMS CATALYST• HBR PHBR=llijMM HG OXIDIZER• OXYGEN P02=626MM HG ' SOLVENT• BROMOBENZENE JOO GRAMS EOUILIBRIUM CONCENTRETION OF• 02•5.6 E-3H/L EQUILIBRIUM CONCENTRETION OF•HBR•l.O E- 3H/L EHPERATURE• lijQ OG C ij_OOQ °'I I 1--' 00 -187- 60 IA ~ Los of Heating Vale Kcal/liter . 30 20 £ £ &~ 10 ~A A 0 0 10 20 30 % Sulfide that Reacted in Experiment 5 Figure 7.22 The Loss of Heating Value vs the Percent of Sulfide that Was Oxidized -188- Moreover, the data indicate that the loss of heat varies linearly with the sulfide that reacted. Although 15-30% of the sulfide were converted, the heat lost was very small. This indicates that the oxidation is relatively very selective to the sulfide, and that loss of heating value due to the oxidation will not play a major economic role if low conversion is maintained. 7.12 Oxidation of Dibenzothiophene A few exploratory experiments were made in order to see the appli- cability of the oxidation method to other classes of sulfur compounds e.g., to condensed thiophenes. No oxidized sulfur compounds were de- tected in the product, and the concentration of the thiophene d:id not change as a result of the oxidation. The type of solvent or hydrocar- bon used did not influence the sulfur compound. The area of selective oxidation of thiophenic compounds is of extreme importance, and is recommended for future study. -189- Chapter 8 SUMMARY AND RECOMMENDATIONS 8. l Summary The most important result of the study is the demonstration that selective oxidation of an aliphatic sulfide in a hydrocarbon medium can be achieved by conducting the oxidation with molecular oxygen, at a low temperature (e.g., 14O°C), and in the presence of a catalyst for the hydrogen abstraction step (e.g., HBr). When HBr is used to catalyze the rate of hydrogen abstraction, the oxidation rate is larger, the oxidation potential of the solution is larger, and the rate of oxidation of the side chain of the sulfide is also increased. Sulfone was found only when HBr was introduced into the system. The energy of the bond of the hydrogen that can be most easily abstracted, D(R----+1), determines the effect of the hydrocarbon on the oxidation of the sulfur compound. The larger D(R----+1) is, the more sulfur-containing products are formed, and the deeper and faster the oxidation of the side chain occurs. This observation is consistent with the rl)J e that more reactive species are less selective, since if the D(R----+1) is larger, R· and RO 2H will be less stable and stronger oxidizers. The rate of formation of sul foxi des and their equilibrium concentration is larger when a protic solvent (e.g., 1-octanol) is present, compared with when the solvent is aprotic (e.g., bromobenzene). Less than 1% loss of heating value is observed as a result of the oxidation. Since about 25% of the sulfides were oxidized, the selectivity of the process is satisfactory. -190- 8.2 Recommendations Knowledge in three areas is required in order to be able to evalu- ate a desulfurization process by selective oxidation: 1. Kinetic data on the oxidation of the sulfur compound. 2. Kinetic data on the pyrolysis of the oxidized sulfur compounds which were fonned. 3. The magnitude of the loss of heating value due to the oxidation. In this study, a novel method to selectively oxidize aliphatic sulfides was demonstrated. However, attempts to oxidize dibenzothio- phene by this method were unsuccessful. Because thiophenic compounds constitute a major fraction of the sulfur compounds in heavy oils and coal, it is imperative to devise a method to oxidize them. The oxida- tion potential that is required to oxidize aliphatic sulfoxide to the sulfone is approximately the same as that required to oxidize a thiophene to the corresponding sulfoxide and sulfone. Since butyl sulfone was found as a product of the oxidation of butyl sulfide, it is believed that a modification of the proposed method can be used to oxidize thiophenes to the sulfoxides and sulfones. The author recommends that the work be continued according to the following sequence: 1. Establish conditions in which it is possible to oxidize thiophenes to the corresponding sulfoxides and su~fones. 2. Derive kinetic data on the oxidation. 3. Derive kinetic data on the rate of decomposition of the oxidized sulfur compounds that were fonned. It is recommended not to use HBr as a catalyst because it complicates the -191- system by many side reactions. straction step should be found. Other catalysts for the hydrogen ab- -192References A-1 Attar, A., "Desulfurization of Heavy Fuel Oil by Oxidation", MS Thesis, Chemical Engineering Dept., Technion, Israel Institute of Technology, Haifa, Israel (1972). A-2 Attar, A., "Introduction to Fuel Desulfurization Internal Report, Chemical Eng. Dept., California Institute of Technology, Pasadena, Ca. 91125, Feb. 1973. Available from Professor W. H. Corcoran. 11 , A-3 Attar, A., "Sulfur Compounds in Coal", submitted for publication. A-4 Attar, A. and Corcoran, W. H., "Quantitative Evaluation of Chromatographic Data from Nonlinear Detectors and the Sulfur FlamePhotometric Detector", Part I, submitted for publication. A-5 Attar, A., Forgey, R., Horn, J., and Corcoran, W. H., ibid, Part II, submitted for publication. A-6 Attar, A., "Selective Oxidation of Organosul fur Compounds", Progress Report, Chemical Eng. Dept., California Institute of Technology, Pasadena, Ca., March 1973. \. A-7 Attar, A., "Selective Oxidation of Organosulfur Compounds", Interim Report, Caltech, Feb. 1976. A-8 Ashworth, M.R.F., The Determination of Sulfur-Containing Groups , Academic Press, London (1972). A-9 Aris, R., Gavalas, G. R., Philos. Trans. R. Soc. London 260, 357 (1966). 8-1 Barnard, D., Fabian, J. M., and Koch, H. P., J. Chem. Soc., 2442 (1949). 8-2 Bahrman, E. J. and Edwards, J. 0. in Progress in Physical Organic Chemistry, Vol. 4, p. 93, Streitwieser, A. and Taft, R.W. (eds.), Wiley-Interscience, New York (1967). -193B-3 Barnard, D., Bateman, L., and Cunneen, J. I., in Organic Sulfur Compounds, Vol. I, p. 229, Kharasch, N. (ed.), Pergamon Press, London ( 1961 ) . B-4 B-5 B-6 B-7 B-8 Bateman, L. and Hargrave, K. B., Proc. Roy. Soc. A224, 389 (1954). IDEM, IBID, A-224, 399 (1954). Barnard, D., J. Chem. Soc., 489 (1956). Bateman, L. and Shipley, F. W., J. Chem. Soc., 1996 (1955). Bateman, L., Cunneen, J. I., and Ford, J., J. Chem. Soc., 3056 (1956). B-9 IDEM, IBID, 1539 (1957). B-10 Barnard, D., Bateman, L., Cain, M. E., Colclough, T., and Cunneen, J. I. , J. Chem. Soc., 5333 ( 1961 ) . B-11 Bateman, L., Cain, M., Colclough, T. and Cunneen, J. I., J. Chem. Soc., 3570 (1962). B-12 Barnett, B., Bell, E. R., Dickey, F. W., Rust, F. F., and Vaughan, W. F., Ind. Eng. Chem. 11, 2613 (1949). B-13 Bamford, C. H. and Dewar, M.J.S., Proc. Roy. Soc. Al98, 252 (1949). B-14 Benson, S. W., Thermochemical Kinetics, Wiley, New York (1968). B-15 IDEM, The Foundation of Chemical Kinetics, McGraw-Hill, New York, (1961). B-16 Benson, S. W. and Shaw, R., "Thermochemistry of Organic Peroxides, Hydroperoxides, Polyoxides and Their Radicals", Vol. I, p. 105 in Organic Peroxides, D. Swern (ed.), Wiley, New York (1970). C-1 Curci, R. and Edwards, J. 0., in Organic Peroxides, Vol. I, p. 230, Swern, D. (ed.), Wiley-Interscience, New York (1970). C-2 Capozzi, G. and Modena, G. in The Chemistry of the Thiol Group, Vol. 2, p. 785, Patosi, S. (ed.), John Wiley and Sons, London (1974). -194C-3 Cain, M. E. and Cunneen, J. I., J . Chem. Soc., 2959 (1962). C-4 Carlsson, D. J. and Robb, J. C., Trans. Farad. Soc. 62, 3404 ( 19 66) . C-5 Calderband, P. H. and Moo-Young, M. B., Chem. Eng. Sci . .1§_, 35 ( 1961). D-1 Drushel, H. V., Miller, J. F., Hubis, W. and Clark, R. 0., Anal. Chi·m. Acta.!.§_, 394 (1956). 0-2 Dankleff, M.A.P., Curci, R., Edwards, J._O., and Pyum Hae-Yung, J. Am. Chem. Soc. 90, 3209 (1968). 0-3 Drago, R. S., Wayland, B., and Carlson, R. L., J. Am. Chem. Soc. 85, 3125 (1963). 0-4 Denison, G. H. Jr. and Condit, P. C., Ind. Eng. Chem. B_(ll), 1102 (1945). 0-5 Davies, A. G., Organic Peroxides, Butterworths, London (1961 ). 0-6 Drushel, H. V. in The Analytical Chemistry of Sulfur and Its Compounds, Part II, p. 49, Karchmer, J. H. (ed.), Wiley, New York ( 19 72). D-7 Denisov, E.T . , Kinetka i Kataliz i, 53 (1963). E-1 Emannual, N. M., Denisov, E. T., Maizus, Z. K., Liquid-Phase Oxidation of Hydrocarbons, Plenum, New York (1967). E-2 Eckstein, B. H., Scheraga, H. A., and Artsdalen, E.R.V., J. Chem. Phys. g_(l), 28 (1954) . E-3 Ettre, L. S. and Zlalkis, A. , The Practice of Gas Chromatography, lnterscience, New York (1967). F-1 Ford, J. F. and Young, V. 0., Prep. Div. Pet. Chem., ACS lQ_(2), C-111 (1965). -195- G-1 Greco, A., Modena, G., and Todesco, P. E., Gaz. Chim. Ital. 90, 671 (1960). G-2 Gerdil, R. and Lucken, E.A.C., J . Am. Chem. Soc. 88(4), 733 (1966). G-3 Germin, J. E., Intra-Science Chem. Report §_(3), 101 (1972). G-4 Gallegas, E. J., Anal. Chem. 47(7), 1150 (1975). H-1 Heimlich, B. N. and Wallace, T. J., Tetrahedron _g__g_, 3571 (1966). H-2 Hargrave, K. R., Proc. Roy . Soc. A235, 55 (1956). H-3 Hiatt, R., "Hydroperox_ i des", Chapt. I, p. 1 in Organic Peroxides, Vol. 2, Swern, D. (ed.), Wiley-Interscience, New York (1971). H-4 Hartough, H. D., Thiophene and Its Compounds, Interscience, New York ( 19 5 7) . H-5 Hartough, H. D. and Meisel, S. L., Compounds with Condensed Thiophene Rings, Interscience, New York (1954)~ H-6 Howard, J. A. and Ingold~ K. V., Can. J. Chem. 44, 1119 (1966). H-7 Howard, J. A. and Ingold, K. V., Adv. in Chem., Ser. 75, 258 (1968). H-8 Higuchi, T. and Bensch, K. H., J. Am. Chem. Soc. 68, 5486 (1966). H- 9 Horak, V. and Pecka, J., J. Chrom . .!i, 97 (1964). H-10 Horne, W. A. and McAfee, J., "Hydrogenation of Petroleum and Its Fractions", Chapt. 5 in Advances in Petrol. Chem. and Refinery, Vol. 3, Kobe, K. A. and McKetta, J. J. {eds.), lnterscience, New York {1960). I-1 1-2 Ingold, K. V. ' Chem. Rev.§_}_, 563 (1961 ). Ingold, K. V.' 7th World Petrol. Cong., Mexico City (1967), Vol. 7, p. 15. J- 1 Jaffe, S. B., Ind. Eng. Chem., Proc. Res. and Develop. ]l_, 34 (1974). -196K-1 Kucharczyk, N. and Horrak, V. , Collect Czech. Chem. Commun. 34 2417 (1969) (Eng. Trans.). K-2 Kondratiev, V. N., Rate Constants for Gas-Phase Reactions", U.S. Dept. of Commerce (1972). K-3 Kirkland, J. J. (ed.), Modern Practice of Liquid Chromatography, Wiley-Interscience, New York (1971) . K-4 Kerr, J. A. and Trotman-D i ckenson, A. F., Prog. React. Kinet. ]_, 107 (1961). K-5 Kirk, A. D. and Knox, J. H., Trans. Farad.Soc._§§_, 1296 (1960). L-1 Landis, P. S. and Sedlak, M. in The Analytical Chemistry of Sulfur and Its Compounds, Part II, p. 793, Karchmer, J. H. (ed.), WileyInterscience, New York (1972). L-2 Lin, A. and Kehat, E., Ind. Eng. Chem . , Prod. Res. and Develop . .§.(4), 436 (1969). 11 L-3 Lloyd, A.C., Int. J . Chem. Kin. VI, 169 (1974). L-4 Lawrence, H. C. and Pigott, S. P., J. Chem. Thermod. l, 221 (1971). M-1 Modena, G. and Miola, L., Gazz. Chim. Ital. 87, 1306 (1957). M-2 Medvedev, S. S. and Pad'yapol 'skaya, A., Zh. Fiz. Khem. 11., 719 (1939) (Ref. E-1, p. 109). M-3 McNair, H. M. and Bonelli, E. J., "Basic Gas Chromatography", Varian Aerograph, Palo-Alto, California (1969). M-4 Miotti, U., Modena, G, and Sedea, L., J. Chem. Soc. f, 802 (1970). M-5 McKinley, J. B., "The Hydrodesulfurization of Liquid Petroleum Fractions Chapt. 6, p. 405 in Catalysis, Vol. V, Emmett, P. H. (ed.), Reinhold, New York (1957). 11 , -197 0-1 Overberger, C. G. and Cummins, R. W., J . Am. Chem. Soc.~. 4250 ( 1953). 0-2 IDEM, IBID, ]2, 4783 (1953). P-1 Perry, R. H. (ed.), Chemical Eng ineer's Handbook, 4th Ed., McGrawHill, New York (1974), p. 3-197. R- 1 Rust, F. F. and Vaughan, W. E., Ind. Eng. Chem. 1.J...(11), 2595-2616 ( 1949) . R-2 Rosser, W. A. and Wire, H., J . Phys . Chem . .§1_, 1753 (1959) . R- 3 Robertson , A. and Waters, W. A. , J. Chem. Soc., 1574 (1948). R- 4 Reid, E. E. , Organic Chemistry of Bivalent Sulfur, Vol. III, Chemical Pub. Co. , New York (1960). S-1 Swern, D., "Organic Pero xy Acids", Chapt. VI in Organic Peroxides, Vol. I, Swern, D. (ed . ), Wiley- I nterscience, New York (1956). S-2 Suter, C. M., The Organic Chemistry of Sulfur, John Wiley, New York ( 19 44) . S- 3 Sheldon, R. A. and Kochi, J . K., Oxid . and Combust . Rev. i, 135 S-4 (1973). Swern, D. (ed.) Organic Peroxides , Vol. I, II and III, WileyInterscience, New York (1971) . S-5 Schuman, S. C. and Shalit, H., Catalysis Rev . .1_(2), 245 (1970). S-6 Silverstein, R. M., Bassler , G. C., and Morril, T. C., Spectrometric Identification of Organic Compounds, John Wiley and Sons, New York ( 196 7) . S-7 Schuit, G.C . A. and Gates, B. C., A.I.Ch.E.J . .12_(3), 417 (1973). S-8 Sugiyama, T. , Suzuki, Y., and Ta keuchi, T., J. Chrom. Sci. ll.(12), 639 (1973) . -198T-1 Thomas, J . R., J. Am. Chem . Soc. I]_, 246 (1955). T-2 Twigg, G. H., Disc. Farad. Soc . 14, 240 (1953). U-2 Urbanski, T., Talanta 1, 799 (1962 ) . W-1 Willits, C. D., Ricciuti, C., Knight, H.B., and Swern, D., Anal. Chem. 24(5), 785 (1952) . W-2 Wilke, C.R. and Chang, P., A.I.Ch.E.J. l, 264 (1955). W-3 Whitney, D. C., "Modkin: A Computer Program for the Automatic Construction of Chemical Kinetics Mechanisms and Solutions of Their Governing Equations 11 , Report R73-51 , prepared for EPA Research Center, Triangle Park, N. Carolina, 27711, Feb. (1974). W-4 Weekman, V. W. Jr., "Chemical Reaction Engineering Reviews", Hurlburt, H. M. (ed.), ACS, Adv. in Chemistry Series 148, p. 98 (1974). W-5 Wei, J. and Kuo, J.C.W., Ind . Eng. Chem. Fundam. ~, 114 (1969). W-6 White, W. B., Johnson, S. 751 (1958). Z-1 Zuman, P. Substitution Effects in Organic Polarography, Plenum, New York ( 19 6 7) . M. and Dantzig, G. B., J. Chem. Phys. -199- Appendix A Analytical Procedures A.l Separation of Acidic Oxidation Products for MS Extract 10 ml reaction product with 10 ml 2N NaOH. Separate the phases and acidify the aqueous phase with concentrated HCl. yellow color fades when the pH becomes acidic. Extract the acidic liquid with two 5 ml fractions of di-ethyl ether. Separate the phases, wash the ether phase with 2 x 2 ml distilled water. the ether phase and let it dry in an open beaker. separates and often crystals show. The Separate Yellow-brown liquid -200Appendix B Purification Procedure B.l Purification of Tetralin Extract 300 ml tetralin with 5 xlO ml concentrated H2so 4 and separate the red-yellow sulfuric acid fraction. If the fifth extrac- tion is still yellow, keep extracting with H2so 4 until the yellow color becomes barely visible. All the next stages of the experiment should be done under inert atmosphere. Wash the tetralin twice with 50 ml distilled water and twice with 20 ml O.lN NaOH. organic phase and dry it over MgS0 4 . Separate the Vacuum distill the tetralin (5 mm 76°C, l mm 64°C) and pass it through a column with freshly activated alumina. atmosphere. Vacuum distill it again and keep it under argon Keep the purified tetralin in a dark place. If the purification was done more than a week before use, distillation may be required. To check the purity use 1 µt samples and a Carbowax 20 column at 145°C. At range l • 10 3 better than 99.3% purity should be obtained. B.2 Deactivation of the Reactor Wall Mix hexamethyl disilazone, I, trichlorosilane, II, and dry pyridine, III, in the proportion 3:1:5. Apply on the dry surface of the reactor immediately after preparation. away from moisture. Exercise caution and keep Do not apply on rubber parts.