a :5“ s 3: i “awn"..wEu :: r .4. «9&qu Era...“ (3"... ’1) u :1 1.40125.» 51...; if): 1., 3...... . .wfla - ”'2; LI I Ear—RY 2030! Michigan State a University . This is to certify that the dissertation entitled CATALYZED HYDROGEN PEROXIDE PROPAGATIONS: STUDY OF HYDROGEN PEROXIDE DECOMPOSITION, HYDROXYL RADICAL PRODUCTION, AND BYPRODUCT FORMATION FROM THE REMEDIATION OF AROMATIC HYDROCARBONS. presented by Carlos A. Sanlley Pagan has been accepted towards fulfillment of the requirements for the Doctoral degree in Environmental Enfleering IL/M A; d madfflfiv Majo rofessor’ 5 Signature Q; 7g 07,, 07 OO 9 Date MSU is an Affirmative Action/Equal Opportunity Employer -.-_n-._c-.—.---n-.-—--.-o--a--o—p—n— - PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 5/08 K‘IProleccaPreleIRCIDateDue indd CATALYZED HYDROGEN PEROXIDE PROPAGATIONS: STUDY OF HYDROGEN PEROXIDE DECOMPOSITION, HYDROXY L RADICAL PRODUCTION, AND BYPRODUCT FORMATION FROM THE REMEDIATION OF AROMATIC HYDROCARBONS. By Carlos A. Sanlley Pagan A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Environmental Engineering 2009 ABSTRACT CATALYZED HYDROGEN PEROXIDE PROPAGATIONS: STUDY OF HYDROGEN PEROXIDE DECOMPOSITION, HYDROXY L RADICAL PRODUCTION, AND BYPRODUCT FORMATION FROM THE REMEDIATION OF AROMATIC HYDROCARBONS By Carlos A. Sanlley Pagan Catalyzed hydrogen peroxide propagations have been studied in order to determine the role of hydrogen peroxide decomposition and hydroxyl radical production during in situ remediation, as well as the byproducts obtained from the oxidation of aromatic hydrocarbons. Three types of catalyzed systems have been studied: ferrous catalyzed, ferric catalyzed, and goethite catalyzed reactions. Hydrogen peroxide decomposition appears to be dependent only hydrogen peroxide concentration regardless of the catalyzer used. The rate of hydrogen peroxide decomposition is higher for reactions catalyzed with ferrous ion as opposed to ferric ion, while maintaining a similar yield for oxygen production. Goethite catalyzed reactions are slower, with a 24-hour reaction period required to achieve peak yield, versus a few minutes for soluble iron reactions. Hydroxyl radical production from goethite catalyzed systems appears to be out competed by hydrogen peroxide decomposition, resulting in an increase in hydroxyl radical yield after 24 hours of reaction. Reactions at pH 3 appear to produce more hydroxyl radicals than reactions at pH 7 under the same experimental conditions. The total yield of hydroxyl radicals could not be determined due to the production of byproducts in the reactions, some of which have been identified. Hydroxyl radical production from soluble iron reactions was estimated using kinetic modeling. Pyrene was removed from both aqueous and slurry systems using catalyzed hydrogen peroxide. Pyrene removal was greater in slurry systems than in aqueous systems, possibly because of limitations due to the water solubility of pyrene. Ferric catalyzed systems achieved better removal rates than ferrous catalyzed reactions at the same experimental conditions. Goethite catalyzed reactions appear to be the most efficient achieving similar removal rates while using lower hydrogen peroxide concentrations. Byproducts were found for all reactions; 12 of these products have been identified. The byproducts range from hydroxylation reactions to ring cleaved products. By the use of parameter estimation software we have found that the aqueous solubility, and therefore 'their mobility, of some of these byproducts is orders of magnitude larger than that of pyrene; causing a potential threat to the environment. Byproducts found in this study are similar to those found from gaseous ozonation of pyrene. Dedication To my family and friends iv Acknowledgements I have come from afar and withstood many ordeals to reach this summit, but one constant is that I was never alone. I would like to thank through this section the many individuals and organizations that helped me achieve this feat. First and foremost I would like to thank Dr. Susan Masten for her encouragement, help, and understanding throughout my studies at Michigan State University. Dr. Masten was a teacher, a friend, family and a mentor, for that I am eternally grateful. I would also like to thank the members of my research committee: Dr. Phanikumar Mantha, Dr. Volodymyr Tarabara, and Dr. Brian Teppen for their help and contributions to my work. I was very fortunate to receive the support of LASPAU and the William H. Fulbright Scholarship program, which enabled me to study in the United States and fulfill one of my life dreams. Once I reached Michigan State University I required financial support and received it from the East and West Michigan Chapters of the Air and Waste Management Association. My research funding was obtained from McMaster University, and NIEHS Superfund Grant 2P42ESO4911-05; thank you for your support. There were many people involved in my life at MSU to which I would like to show my appreciation by naming them here: Lori Hasse, Laura Taylor, Mary Mroz and Linda Steinman; Joseph Nguyen and Yiliang Pan, Dr. Shuguang Li, Dr. Rick Lyles, Dr. Neeraj Buch and Dr. Sayed Hashsham, from the CEE department; Dr. Paul Jones and Patrick Bradley fiom Department of Food Toxicology; Shawn McElmurry, the unofficial member of my committee; Dr. Daniel Jones, Beverly Chamberlain and Li-Jun from the Mass Spectrometry Facility; and Dr. Simon Davies from Agricultural Engineering. I have to give great thanks to Dr. Mackenzie Davis for being an inspiration, a mentor and a friend throughout my years at MSU. I have learned a great deal from you and I am still enjoying the fi'uits of our many conversations. To Dr. Roger Wallace, thank you for being my fi'iend, teacher, and collaborator. Your poise, knowledge and presence have always been greatly appreciated. Finally, I would like to thank Geneva, without whom I would not have made it this far. I am thankful for your support, your caring and your belief in me. Your light gave me strength and guided me through the days. Thank you, you are and always will be the love of my life. vi TABLE OF CONTENTS List of Tables ................................................................................................................... x List of Figures ............................................................................................................... xii Objectives and Hypotheses ................................................................................................. 1 Objectives ....................................................................................................................... 1 Hypotheses ...................................................................................................................... 2 Chapter 1. Background ...................................................................................................... 4 Polycyclic Aromatic Hydrocarbons ............................................................................ 4 In-Situ Chemical Oxidation ........................................................................................ 6 Permanganate .......................................................................................................... 6 Ozone ...................................................................................................................... 7 Persulfate ................................................................................................................. 7 Catalyzed Hydrogen Peroxide ................................................................................ 9 Conclusion ................................................................................................................ 14 References ................................................................................................................. 15 Chapter 2. Hydrogen Peroxide Decomposition .............................................................. 20 Introduction ................................................................................................................... 20 Decomposition of hydrogen peroxide from reaction with mineral Iron: Goethite 22 Method and Materials ................................................................................................... 23 Preliminary Experiments .......................................................................................... 23 Gravimetric Tests: Oxygen Evolution ...................................................................... 24 Effect of a chelating agent on Oxygen Evolution ..................................................... 25 Effect of aquifer Material on Hydrogen Peroxide Decomposition ........................... 26 Results ........................................................................................................................... 26 Goethite ..................................................................................................................... 35 Discussion ..................................................................................................................... 39 Soluble Iron ............................................................................................................... 39 Temperature .............................................................................................................. 41 Ottawa Sand Slurries ................................................................................................. 43 Soluble iron ........................................................................................................... 43 EDTA experiments ................................................................................................... 43 Goethite ..................................................................................................................... 44 Conclusions ................................................................................................................... 48 References ..................................................................................................................... 50 Chapter 3. Hydroxyl Radical Production ........................................................................ 55 Introduction ................................................................................................................... 55 Method and Materials: .................................................................................................. 57 Materials ................................................................................................................... 57 Soluble Iron Experiments ......................................................................................... 57 Goethite experiments: ............................................................................................... 5 8 Salicylic Acid Determination .................................................................................... 59 Time dependent SA reactions ................................................................................... 59 Byproduct Identification ........................................................................................... 60 GC/MS .................................................................................................................. 6O LC/MS ................................................................................................................... 60 Atomic Absorption .................................................................................................... 61 Results: .......................................................................................................................... 62 Soluble Iron ............................................................................................................... 62 Goethite ..................................................................................................................... 63 Time Dependent Reactions ....................................................................................... 72 Byproducts ................................................................................................................ 73 GC/MS .................................................................................................................. 75 Discussion ..................................................................................................................... 76 Sorption of SA to Goethite ....................................................................................... 76 Salicylic Acid Degradation by Goethite Catalyzed CHPs ........................................ 77 Byproducts ................................................................................................................ 82 Conclusions ................................................................................................................... 95 References ..................................................................................................................... 97 Chapter 4. Oxidation of Pyrene in Soil ......................................................................... 107 Pyrene ......................................................................................................................... 107 Method and Materials ................................................................................................. 108 Determining Pyrene Concentration ......................................................................... 109 Results ......................................................................................................................... 111 Organic Matter Soil: ............................................................................................... 112 Discussion: .................................................................................................................. 1 16 Ottawa Sand: ........................................................................................................... 117 MSU soil: ................................................................................................................ 119 Sodium Dodecyl Sulfate ......................................................................................... 120 MGP Site Soil ......................................................................................................... 122 Conclusions ................................................................................................................. 125 References ................................................................................................................... 126 Chapter 5. Byproduct formation from the oxidation of Pyrene by Catalyzed Hydrogen Peroxide Propagations .................................................................................................... 128 Introduction ................................................................................................................. 128 Method and Materials ................................................................................................. 131 Byproduct Structure Identification ......................................................................... 133 viii Sample preparation ............................................................................................. 133 Results ......................................................................................................................... 134 Byproduct Formation .............................................................................................. 136 Discussion ................................................................................................................... 143 Byproduct formation between different Catalyzers ................................................ 144 Comparison to PAH byproducts found fiom Fenton’s Reagent Reactions ............ 148 Comparison to PAH Byproducts from other Remediation Techniques .................. 149 Electrolytic Aeration of pyrene ........................................................................... 149 Photolysis of Pyrene ........................................................................................... 150 Biodegradation of Pyrene and other PAHs ......................................................... 150 Ozonation of Pyrene ........................................................................................... 152 Environmental Impact of PAH remediation ........................................................... 154 Mobility ............................................................................................................... 154 Toxicity ............................................................................................................... 155 Conclusions: ................................................................................................................ 1 57 References ................................................................................................................... 161 Chapter 6. Modeling ...................................................................................................... 167 Introduction ................................................................................................................. 167 Kinetic Model ............................................................................................................. 167 Temperature ............................................................................................................ 170 Modeling Iron Species and Concentration .............................................................. 172 Oxidation of Ferrous ion by Oxygen ...................................................................... 173 RMSE ...................................................................................................................... 174 Results ......................................................................................................................... 175 Discussion ................................................................................................................... 1 84 Conclusions ................................................................................................................. 1 86 References ................................................................................................................... 187 Chapter 7. Conclusions ................................................................................................. 189 Appendix A: Pyrene Oxidation Byproducts ................................................................... 195 Appendix B: Salicylic Acid Byproducts ......................................................................... 220 Appendix C: HPLC Chromatograrns .............................................................................. 238 Appendix D: MATLAB Code for Kinetic Model: ......................................................... 248 ix List of Tables Table 2-1: Regression analysis for oxygen produced during CHP reactions. The terms a, b, c, and (I refer to the coefficients in equation 13. CI is the 95% confidence interval for the preceding variable. ........................................................................................ 33 Table 2-2: Values for coefficients for equation 11 obtained from normalized solution temperature fits. ........................................................................................................ 33 Table 2-3: Coefficients and 95% confidence intervals for Oxygen production from Goethite experiments for an exponential regression fit. ........................................... 37 Table 2-4: Coefficients found form linear regression of data from gravimetric experiments during Geo'thite catalyzed CHPs. ......................................................... 38 Table 3-1: Results for the removal of 100 ppm of salicylic acid (SA) from solution using Goethite catalyzed reactions at pH 7. Shown are the amounts of Goethite weighed for each trial and its subsequent concentration, the total surface area in each sample vial the amount of salicylic acid found in each sample, the amount of salicylic acid sorbed onto the mineral and the percent removal of salicylic acid from the sample (corrected by sorption). The final column is an estimate of hydroxyl radicals produced calculated from the amount of salicylic acid removed after 72 hours ...... 66 Table 3-2: Results for the removal of 100 ppm of salicylic acid (SA) from solution using Goethite catalyzed reactions at pH 3. Shown are the amounts of Goethite weighed for each trial and its subsequent concentration, the total surface area in each sample vial the amount of salicylic acid found in each sample, the amount of salicylic acid sorbed onto the mineral and the percent removal of salicylic acid from the sample (corrected by sorption). The final column is an estimate of hydroxyl radicals produced calculated fiom the amount of salicylic acid removed after 72 hours. ..... 67 Table 4-1: Physicochemical properties of Pyrene1 ........................................................ 107 Table 4-2 Results from oxidation of 50 grams of MGP soil with 120,000 ppm of hydrogen peroxide with and without soluble iron addition. LOD: Limit of Detection ................................................................................................................................. 115 Table 5-1: Table showing m/z found in the combined extracts from the different CHP tested. ...................................................................................................................... 136 Table 5-2: Byproducts identified by LC/MS/MS with a list of major m/z peaks found in the spectra ............................................................................................................... 137 Table 5-3: Estimated parameters for each one of the byproducts found as well as the SMILES characters used to represent them in the parameter estimation software. 158 Table 5-4: Byprodcuts found from CHP, gaseous ozonation and bioremediation remediation of pyrene. ............................................................................................ 159 Table 6-1: Best fit for normalized temperature for soluble iron catalyzed reactions. Molar concentrations of iron and hydrogen peroxide are given along with the coefficients for the regression line and the R2. ........................................................................... 171 Table 6-2: Rate Constants for Reaction of Fe2+ Species with 02 from Morel 199344. .. 174 Table 6-3: Values for the hydrogen peroxide decomposition rate constant determined from the optimization of the objective function in equation 63 for reactions corrected and not corrected for temperature. .......................................................................... 17 7 xi List of Figures Figure 2-1: Volume of Oxygen produced during a CHP with high end hydrogen peroxide concentration (0.5M) catalyzed by fen'ic ion (0.01 M). Measured by water displacement on an inverted cylinder. ...................................................................... 27 Figure 2-2: Hydrogen peroxide concentration back calculated from the volume of oxygen collected in the inverted cylinder shown in Figure 1. .................................. 28 Figure 2-3: Moles of oxygen produced during reactions catalyzed with ferrous and ferric soluble iron for varying hydrogen peroxide concentrations. The graph shows that oxygen production is independent of catalyzer concentration .................................. 29 Figure 2-4: Relation between regression coefficients and the initial hydrogen peroxide concentration for oxygen production for soluble iron catalyzed reactions ............... 30 Figure 2-5: Change in temperature of solution during reaction normalized by initial temperature for three different hydrogen peroxide concentrations catalyzed by 0. 3 M Ferric ion. ............................................................................................................. 31 Figure 2-6: Comparison between solution temperature for ferrous and ferric catalyzed systems, normalized by the initial temperature ......................................................... 32 Figure 2-7: Relation between regression coefficients and the initial hydrogen peroxide concentration for normalized temperature for soluble iron catalyzed reactions ....... 34 Figure 2-8: Comparison of Oxygen production between reactions with and without Ottawa sand and with and without EDTA. ............................................................... 35 Figure 2-9: Oxygen production form Goethite catalyzed reactions ................................. 36 Figure 3-1: Pathway for the reaction of hydroxyl radicals with salicylic acid and its products 54. ................................................................................................................ 56 Figure 3-2: Plot of the change in hydrogen peroxide concentration as a function of time for ferric ion concentration of 0.06 M. Hydrogen peroxide was measured by iodometric titration. The initial concentration of hydrogen peroxide is shown in the legend ........................................................................................................................ 62 Figure 3-3: Response surface obtained from equation 1 for the removal of salicylic acid at pH 7 using goethite catalyzed reactions. The x-axis represents H from equation 2 and the y-axis represents F from equation 3. The color scheme represents the region at which a specific removal efficiency can be found and is delineated by the isoresponse lines shown in the figure. ...................................................................... 68 xii Figure 3-4: Response surface obtained from equation 4 for the removal of salicylic acid at pH 3 using goethite catalyzed reactions. The x-axis represents H from equation 2 and the y—axis represents F from equation 3. The color scheme represents the region at which a specific removal efficiency can be found and is delineated by the isoresponse lines shown in the figure. ...................................................................... 70 Figure 3-5: Plot of SA concentration over time for different hydrogen peroxide and Goethite concentrations for pH 3 reactions. Initial SA concentration of 100 ppm... 72 Figure 3-6: Plot of SA concentration over time for different hydrogen peroxide and Goethite concentrations for pH 7 reactions. Initial SA concentration of 100 ppm... 73 Figure 3-7: Oxidation of salicylic acid 100 ppm over time for pH 3 and pH 7 reactions with 0.0263 g/L goethite and an initial concentration of hydrogen peroxide of 0.13M. ....................................................................................................................... 74 Figure 3-8: Comparison of SA removal in pH3 and pH 7 samples at different SA concentrations. .......................................................................................................... 78 Figure 3-9: Typical chromatographic spectra from LC/MS analysis in negative ion mode for pH 3 reactions, showing the total ion count for negative ion data acquisitions. And the peaks for m/z 153- found in the samples. ................................................... 83 Figure 3-10: Mass spectra for pH 3 samples showing m/z 153' and 169', that are believed to be dihydroxy and trihydroxy benzoic acids. ......................................................... 85 Figure 3-11: Chromatographic peak of m/z 169" believed to be THBA, as compared to the total ion count on the LC/MS for pH 3 samples ................................................. 87 Figure 3-12: LC/MS spectra for retention time 5.17minutes for pH 7 samples, showing m/z 153' and 169'; which are believed to be DHBA And THBA ............................. 89 Figure 3-13: Chromatographic peak of byproduct m/z 153' shown here to compare its abundance versus total ion counts, in positive and negative ion mode for pH 7 sample during LC/MS analysis. It is possible to see from the first scan that three isomers may exist ...................................................................................................... 91 Figure 3-14: Possible structure for the byproduct m/z 169' found in the LC/MS spectra in Figures 3-8 and Figure 3-10 ...................................................................................... 93 Figure 3-15: GC/MS identified byproducts found from the oxidation of salicylic acid by Goethite catalyzed CHPs for both pH 3 and pH 7. . ................................................. 94 Figure 4-1: Structure of model PAH, Pyrene showing double bonds and fused benzene rings ......................................................................................................................... 107 xiii Figure 4-2: Results fiom the Oxidation of Pyrene in Ottawa Sand using CHP. The graph shows the percent oxidation achieved at the different hydrogen peroxide concentrations used under Ferrous and Goethite catalyzed systems. ..................... 111 Figure 4-3: Summary of results in soil containing organic matter. Percent of pyrene oxidation Vs hydrogen peroxide concentrations using both soluble F e2+ and natural minerals of the Soil. Results shown are averages of triplicates. ............................. 113 Figure 4-4 : Pyrene oxidation in soil containing organic matter catalyzed with soluble Fe2+ (Percent Oxidation vs. Hydrogen Peroxide Concentration) ............................ 114 Figure 4-5: Pyrene oxidation in soil containing organic matter catalyzed with mineral iron (Percent Oxidation vs. Hydrogen Peroxide Concentration). ........................... 114 Figure 4-6: Contaminated soil being extracted from geo-probes during sampling at the Grand Ledge, MI former MGP site ......................................................................... 123 Figure 5-1: Percent removal of pyrene from CHPs for solution and Ottawa sand reactions. The initial iron catalyst concentration was held at 0.06 M, while the initial hydrogen peroxide concentration was 0.11 M. ....................................................... 135 Figure 5-2: LC/MS Electrospray mass spectra for m/z 233 obtained from the oxidation of Pyrene in solution by Goethite. This m/z was found in the hexane extract. ........... 138 Figure 5-3: MS/MS spectra showing fragment ions from m/z 233, obtained from the hexane extract of Goethite catalyzed CHPs ............................................................ 139 Figure 5-4: Byproducts identified from the oxidation of Pyrene by CHPs in water and soil Slurries, using mineral and soluble iron catalyzed reactions. *3 isomers of the structure were found, but not readily distinguished. ............................................... 141 Figure 5-5: Proposed pathway for hydroxyl radical degradation of Pyrene by CHPs... 142 Figure 5-6: a) Pyrene molecule with numbered carbons and b) showing p-orbitals around the center region of the molecule. ........................................................................... 147 Figure 6-1: Plot showing the result from the optimization of the kinetic model described by equations 11-17 for an initial hydrogen peroxide concentration of 0.1 M and ferric ion concentration of 0.06 M.The lines represent the simulated concentrations obtained by the model and the A represent the hydrogen peroxide concentration obtained from iodometric titration. ......................................................................... 17 6 Figure 6-2: Plot showing the result from the optimization of the kinetic model described by equations 11-17 for an initial hydrogen peroxide concentration of 0.57 M and ferric ion concentration of 0.07 M. The lines represent the simulated concentrations xiv obtained by the model and the A represent the hydrogen peroxide concentration obtained from iodometric titration .......................................................................... 177 Figure 6-3: Graph showing the change in concentration for Ferrous and Ferric ions during CHP reactions with initial hydrogen peroxide concentration of 0.1 M and ferric ion initial concentration of 0.07 M for reactions modeled with equations 11-17 with temperature correction.The graph illustrates the cycling of ferric ion to ferrous ion and back. ........................................................................................................... 178 Figure 6-4: Graphs showing the concentrations of hydroxyl radical, perhydroxyl radical, and Superoxide formed during CHP reactions for initial hydrogen peroxide concentration of 0.1 M and initial ferric ion concentration of 0.07 M. Model is described by equations 11-17 with temperature correction. ................................... 180 Figure 6-5: Comparison of the oxygen production calculated by the kinetic model and experimental data collected for hydrogen peroxide concentration of 0.57 M and ferric ion concentration of 0.07 M .......................................................................... 182 XV Objectives and Hypotheses Objectives Chemical oxidation, used as an in-situ remediation technique, is a highly complex system that is commonly treated as a black box. This research attempts to shed light on the mechanisms that occur in catalyzed hydrogen peroxide systems while creating tools that can be applied for estimating the outcomes from this treatment. To complement these tools, byproduct identification fiom the oxidation of aromatic compounds during such reactions has been performed. The fate of these byproducts in the environment has also been estimated. The following objectives have been studied: 1) The conditions under which hydrogen peroxide decomposition inhibits hydroxyl radical production in catalyzed hydrogen peroxide reactions have been determined. 2) A kinetic model for hydroxyl radical production during catalyzed hydrogen peroxide reactions has been developed. 3) Byproducts formed from oxidation of pyrene in aqueous and soil matrices have been identified for the following oxidative treatments: a) Ferrous ion (Fey) catalyzed hydrogen peroxide. (Classic Fenton’s Reagent) b) Ferric ion (Fey) catalyzed hydrogen peroxide. (Modified F enton’s Reagent I) 4) 0) Mineral iron (Goethite) catalyzed hydrogen peroxide. (Modified Fenton’s Reagent II) A model that can be used to determine the potential fate of the identified by- products in the environment has been developed. Hypotheses The following hypotheses for this proposed study have been tested: 1) Based on the known theories of hydroxyl radical interaction with aromatic 2) species, in-situ chemical oxidation of polycyclic aromatic hydrocarbons in aqueous systems using catalyzed hydrogen peroxide, Fenton’s Reagent, will produce hydroxylated, quinonic, hydroquinonic and ring-cleaved by-products, whose physical-chernical properties (e.g., water solubility) will differ from the parent compound. Based on the known theories of hydroxyl radical interaction with aromatic species, in-situ chemical oxidation of polycyclic aromatic hydrocarbons in soil matrices using catalyzed hydrogen peroxide, F enton’s Reagent, will produce hydroxylated, quinonic, hydroquinonic and ring-cleaved by-products, whose physical-chemical properties will differ from the parent compound. The extent of byproduct formation will be greater than that of the aqueous system. 3) 4) 5) 6) The catalyzer selected to promote hydroxyl radical formation during F enton system reactions will affect the structure of the by-product formed and/or the rate of by-product formation. A mathematical model can be constructed in order to predict the type of byproduct formed under specified matrix conditions; specifically what will be the dominant species in the system. The environmental fate of the byproducts that result from the chemical oxidation of pyrene using Fenton’s Reagent may be predicted by computational modeling; using the properties of the by-products formed (i.e., mobility, sorptive behavior) or the properties of model compounds with similar structures. Hydrogen peroxide decomposition into oxygen, a parallel competitive reaction occurring in Fenton systems during both homogeneous and heterogeneous catalysis, inhibits maximum hydroxyl radical production. Oxygen production from hydrogen peroxide decomposition during F enton reactions can be quantified and may be used as a tool to predict hydroxyl radical production during in-situ chemical reactions. Chapter 1. Background Polycyclic Aromatic Hydrocarbons Polycyclic Aromatic Hydrocarbons (PAHs) are a group of common environmental contaminants comprised of 2 or more fused benzene rings. These compounds are formed during the incomplete combustion of organic matter and fossil fuels, and are found in petroleum derived products, such as diesel, kerosene, and gasolinel’ 2. PAHs are characterized as being recalcitrant. The mobility of these compounds is limited due to their low water solubility (log water solubility at 25°C ranging from -3.61 to -8.22) 3; with solubility decreasing as their molecular weight or hydrophobic surface area increases3. PAHs also have a strong affinity toward soil organic matter (log K0“, from 3.36 to 6.5)4. As a result, PAHs are stable in the environment for long periods of time. The recalcitrant nature of the PAHs as well as their mutagenic/carcinogenic potential, makes them potential health risks and therefore they are listed as priority pollutants 5 by the US. Environmental Protection Agency (U.S.EPA). The US. EPA states that at least 8 out the 16 PAHs listed as priority pollutants may produce skin, liver and/or lung cancer in humans. PAHs are associated with industrial activities where fossil fuels are used, leading to their presence in highly industrialized areas. The formation of PAHs during the combustion of fossil fuels and their subsequent release into the atmosphere has allowed for their widespread distribution in air, water, soil and sediments 2. The presence of PAHs in the atmosphere varies due to seasonal fluxes as fossil fuel consumption increases during winter months. The contaminant may move to a different medium by wet or dry deposition from the atmosphere. PAHs may also reach the environment through leaching from wastes, leaking underground storage tanks or direct contact of petroleum derived products with soil or water 1. Manufactured Gas Plant (MGP) sites, for example, have been found to contain PAH contamination 6; occurring from mismanagement of coal tar wastes, as well as direct dumping of these wastes into soil (US. EPA). Current efforts for the remediation of these compounds in soil and water include: soil 1, 2, 7, 8 . . Ex-srtu washing, in-situ and ex—situ chemical oxidation and biodegradation treatments involve the pumping of water from the contaminated site or the excavation of the contaminated soil/aquifer material in order to remove the contaminant; the US. EPA promotes the use of incineration for soils extracted from sites that contain PAH contamination. In-situ treatment may include the use of chemical agents to neutralize or oxidize the contaminant in place. Bioremediation, which can be performed both ex-situ and in-situ, includes bioaugmentation or biostimulation of bacteria or fungi that degrade PAHs. Bioremediation may degrade the these chemicals by utilizing them as a carbon source for the bacteria or the contaminant may be cometabolized 2. Sims and Overcash 9 (1983) found that one, two, and three ring PAHs are believed to be acutely toxic and that four and five ring PAHs may be genotoxic. Using Gap Junction Intercellular Communication (GJIC), the epigenetic toxicity of some PAHs and their derivatives was studied by Luster-Teasley et al. 10, Hemer et a1. (ref), and Upharn et a1. 11. The studies found that GJIC was inhibited by both the PAHs and their ozonated derivatives; with some derivatives creating greater inhibition than the parent compound. ln-Situ Chemical Oxidation In-Situ Chemical Oxidation or ISCO has evolved as an effective process to remediate organic chemicals form soils, especially in places were infrastructure impedes other treatment techniques 12. ISCO involves the addition of a strong oxidant; often ozone, catalyzed hydrogen peroxide, potassium permanganate or persulfate; into the soil matrix. The goal of chemical oxidation is to change the valence state of the chemical species, ultimately breaking carbon bonds and forming carbon dioxide and water. The effectiveness of these remediation schemes is dependent on the site characteristics, distribution of the oxidant in the matrix, the target compound, and the oxidant used 12’ 13. Permanganate There are two forms of permanganate used for ISCO: potassium permanganate (M04) l3, 14 or sodium permanganate (NaMnO4) . Sodium permanganate is sold as a 40% solution, while potassium permanganate is sold in bulk as a solid. Permanganate is known to oxidize a wide variety of chemicals, such as phenols, cresols, and cyanides; and is believed to react more effectively with olefins than with aliphatic compounds. It is especially reactive with double bonds. Permanganate is commonly used to remediate water contaminated with chlorinated chemicals, such as trichloroethylene: C2C13H + 2KMn04 —> 2C02(aq) + 2Mn02(s) + 2KCl + HCl (1) Permanganate produces manganese dioxide as a product from the oxidation reactions and in most cases the dioxide will coat soil grains and reduce soil matrix permeability. Gates- Anderson, 2001 15 used permanganate as an oxidizer for PAHs and found that the system could achieve the similar results to those of catalyzed hydrogen peroxide (about 90% disappearance). Ozone Ozone is a very reactive and unstable species of oxygen. It is used as an oxidizer and a disinfectant. Ozone is a gas with a half-life of 30 seconds in distilled water at 20 °C. The solubility of ozone in water is three times that of oxygen at the same temperature or about 24 ppm at 20 °C. The use of ozone as a remediation agent requires an on-site generator to produce the gas, which entails high capital costs for treatment. Ozone gas has been studied as a remediation agent for PAHs in both water and soil; and the byproducts obtained from this system have been also identified and studied 10’ 16-20. Persulfate Sodium persulfate (N azszog) has been recently employed as a chemical oxidant for in- situ remediation. The persulfate ion is a slow reacting oxidant considered more powerful than hydrogen peroxide, whose reactivity is a function of pH and concentration 2' Neutral 2- — 1 5208 +21L120—>2HSO4 +202 (2) Dilute acid (pH 3-7) 2 — + — S208 + 2H20 + H —> 2HSO4 + H 202 (3) Strong acid 2 — + — - S208 + 2H20 + H —> H504 + HSO5 (4) Alkaline S 02‘+0H“+H+—>H50‘+Soz‘+-1-0 (5) 28 4 4 22 The oxidative strength of persulfate may be increased with the addition of heat or ferrous salts (although copper, silver or manganese can be used). The increase in strength is believed to result from the formation of sulfate radicals (804'), which are oxidation potency comparable to the hydroxyl radical. The half life of the sulfate radical is believed to be approximately 4 seconds at 40°C. Persulfate is activated at high pH, where it is capable of degrading chlorinated compounds and has been used in acid conditions to remove VOC’s from groundwater. The system does not have the capability of degrading chlorinated solvents in acid conditions. The catalytic effect of metals in the persulfate system decreases over time and the distance from the injection point, limiting the application of this oxidizer. Catalyzed Hydrogen Peroxide In 1894, H.J. Fenton reported that ferrous ion promoted the oxidation of malic acid by hydrogen peroxide. This discovery jurnpstarted the study of the chemical system that now bears his name, Fenton’s Reagent. It was not until 1934 that Haber and Weiss proposed that the addition of the ferrous ion catalyzes the decomposition of hydrogen peroxide, producing a highly oxidative radical species known as the hydroxyl radical 22. The catalytic reaction and the subsequent reactions that occur in the presence of an oxidizable 23, 24 substrate are shown below H202 + Fe2+ —> Fe3 ++‘0H+ -0H (6) -0H+Fe2+—)0H_+Fe3+ (7) -0H+RH—->H20+R- (8) R - +Fe3 + —-) Fe2 ++'Pr0ducts' (9) 2R-—>RR (10) R+Fe2++H+—)RH+Fe3+ (11) -0H+H202 —>H20+-00H (12) 2-0H—>H202 (13) H202 +Fe3 + —>Fe2+ + H+ + ~00H (14) F enton’s Reagent is defined as a dilute hydrogen peroxide solution catalyzed by ferrous iron, A dilute hydrogen peroxide solution is considered to be less than 300 ppm (0.03%) of the oxidizer. In in-situ applications, these dilute conditions may not be achieved and higher concentrations (4-20%) of peroxide are typically used. Adding to this discrepancy, the ferric ion produced during the reaction may precipitate out of solution at pH values greater than 5, which prompted the addition of chelating agents into the system to maintain iron solubility. Watts et al. have proposed the use of the term Catalyzed Hydrogen Peroxide Propagations (CHP) as a collective title for the new Fenton hybrid systems 25’ 26. The contaminant removal efficiency using these CHPs is varied and dependent on the chemical properties of the contaminant, the matrix and the CHP formulation used. The use of the original Fenton’s composition is commonly referred to as Classical Fenton’s Reagent. Hydroxyl radicals are highly reactive, nonspecific oxidizing agents”. Walling et al.28 studied the reaction of the hydroxyl radical with several substituted benzene compounds and found that reactions would either cause hydroxylation of the aromatic ring forming phenolic compounds, leading to ring cleavage, or cause a dimerization of the molecule forming a biphenyl group. The hydroxylation of the aromatics was found to have a first- order dependence on the concentration of the compound and that ring deactivating groups such as bromide and chloride have a positive effect on the hydroxylation 29. Casero et 6130 found that the ratio of ferrous to ferric ions (or excess hydrogen peroxide) affected the intermediates and end products formed during the oxidation of aromatic amines. If the iron species is predominantly in its ferric state the formation of dimers was favored, while the predominance of ferrous ion led to ring cleavage byproducts. 10 Not all hydroxyl radicals formed by CHPs may be free in solution, some are caged or become complexed before they interact with the target molecule 31. Walling found that high concentrations of hydroxyl radical scavengers did not completely prevent the oxidation of mandelic acid (a-hydroxy benzeneacetic acid), and therefore concluded that some of the hydroxyl radicals were not entering the bulk phase of the solution, but still reacting with the target compound. Hydrogen peroxide decomposition may be catalyzed by polyvalent metal ions other than the ferrous ion to create hydroxyl radical formation. Copper (CuZI), titanium (Ti3+), and manganese (Mn2+, Mn”) all produce Fenton-like chemistry when used as catalyzers 23’ 24’ 28, 31, 32 33 ’ . Fenton systems may also be catalyzed by other iron species, such as ferric ion 33-3 24, 36 mineral iron 5 and chelated iron . The studies performed by Watts 33 and by Pignatello 37 utilize the ferric ion as the main catalyzer in CHP reactions. Ferric ion creates reactions that differ from the classical system by the formation of a greater amount of perhydroxyl ions (HOz-or -OOH). The ferric ion catalyzed system reactions are described by Walling, 1974, as: H202 + Fe3 + —> Fe2 + + H+ + 00H (15) H202 + Fe2 + —> Fe3 ++_0H + -0H (16) This ferric ion catalyzed system was found to be dependent on the stability of the carbonium ions formed during oxidation of an organic compound 24’ 29. Pignatello also POintS out that due to the rapid oxidation of the ferrous ion to ferric ion (95% oxidation in 11 2.8 minutes in a 2:/1 mM ferrous : hydrogen peroxide ion solution) most of the actual studies performed on the degradation of organics may have been actually ferric ion catalyzed and not ferrous ion catalyzed 37. This same study found that the ferric system resulted in the mineralization of 2,4 dichlorphenoxyacetic acid (2,4-D) while with the ferrous ion system, 2,4-D reacted more rapidly, but mineralization was not achieved. Mineral iron catalyzed reactions utilize the natural iron species found in soils as a 35 . . . 33-35, 38 catalyzer . Goethite 18 one of the most common minerals used as catalysts . These minerals usually contain a mixture of both ferric and ferrous ions and thus the reactions that occur are a combination of ferric/ferrous systems. For insoluble minerals Kwan et a1. 38 proposed Fenton’s Reagent initiation reactions that occur on the mineral surface: 5Fe3++H 0 HEFe3+H 0 (17) 2 2 2 2 aFe3+H 0 —)H0 +H++-:-Fez+ (18) 2 2 2 -=- Fe2 + +H202 —>~0H+ arr-+2 Fe3 + (19) Since the reaction is dependent on the extent of contact between the surface and hydrogen peroxide, and the binding of hydrogen peroxide to iron atoms on the surface; it is reasonable to assume that these reactions will be slower than those that contain soluble iron. 12 Fenton chemistry has been found to be pH dependent, being most effective at pH 2-3 37' 39; with its oxidizing efficiency decreasing as the pH increases. For this reason, chelated 24, 36, 40, 4] iron species such as ferrous/ferric-EDTA are sometimes used to create F enton- like reactions at higher pH. The chelated species are expected to prevent the precipitation of the iron from the solution, and thus avoid the pH adjustment that would be necessary for the reaction to take place effectively. No published study has looked at byproduct or intermediate product formation fi'om Fenton’s Reagent systems catalyzed by mineral iron. Some studies lacked effort in determining byproducts; Chen et al. 42 stated no byproducts were found in GC analysis fi'om the oxidation of n-hexadecane, 2—methylnaphthalene, and Diesel Fuel after 24 hour oxidation; but only looked at low mass products in the GC. The extraction of these byproducts was performed with a Sohxlet method, which uses hexane as the extracting solvent. Hexane would only extract non-polar compounds and if hydroxylation is the principle means of attack for hydroxyl radicals, then polar compounds would be expected to a major fraction of the products formed. A study performed by Bowers et al.43’ 44 on the oxidation 2,4-diclorophenol and dinitro- ortho-cresol with Fenton’s Reagent concluded that not all the hydrogen peroxide was used to convert the organic matter into carbon dioxide, but that the parent compounds were altered leaving byproducts in the solution. 13 Conclusion Polycyclic aromatic hydrocarbons are widespread environmental contaminants. Their carcinogenic potential has made them priority pollutants in the US. and therefore contaminated areas must be treated in order to minimize potential health risks. In situ chemical oxidation (ISCO) is one of the most common techniques used for treatment of PAH contaminated sites. Of the four ISCO chemicals used, Fenton’s reagent is the most common because of its availability, cost, and versatility. Recent work has shown that PAH remediation with ozone produces byproducts that are in some instances more soluble and hazardous than the parent compound. The following work has looked at the potential problems that arise from the use of Fenton’s Reagent as remediation tool and the potential byproducts that area formed from the chemical oxidation of PAHs by this technique. 14 References l. Kostecki, P. T.; Calabrese, E. J ., Petroleum Contaminated Soils: Remediation Techniques, Environmental Fate, and Risk Assessment. 3rd ed.; Lewis Publishers, Inc.: Chelsea, MI, 1989; Vol. 1. 2. Juhasz, A. L.; Naidu, R., Bioremediation of High Molecular Weight Polycyclic Aromatic Hydrocarbons: A Review of Microbial Degradation of Benzo[a]pyrene. International Biodeterioration & Biodegradation 2000, 45, 57-88. 3. Schwarzenbach, R. P.; Gschwend, P. M.; Irnboden, D. M., Environmental Organic Chemistry. lst ed.; Wilkey-Interscience, John Wiley & Sons Inc: New York, 1993. 4. Chiou, C. T.; McGroddy, S. E.; Kile, D. E., Partition Characteristics of Polycyclic Aromatic Hydrocarbons on Soils and Sediments. Environmental Science and T echnolog 1998, 32, (2), 264-269. 5. Keith, L. H.; Telliard, W. A., Priority Pollutants: Prospective View. Environmental Science and Technology 1979, 13, (4), 416-424. 6. Srivastava, V. J .; Kelley, R. L.; Paterek, J. R.; Hayes, T. D.; Nelson, G. L.; Golchin, J., A Field Scale Demonstration of Novel Bioremediation Process for MGP Sites. Applied Biochemistry and Biotechnology 1994, 45, (6), 741-756. 7. Heitkarnp, M. A.; Freeman, J. P.; Cemeglia, C. E., Napthalene Biodegradation in Environmental Microcosms: Estimates of Degradation Rates and Characterization of Metabolites. Applied and Environmental Microbiology 1987, 53, (1), 129-136. 8- Nam, K.; Rodriguez, W.; Kukor, J. J., Enhanced Degradation of Polycyclic Aromatic Hydrocarbons by Biodegradation Combined with a Modified Fenton Reaction. Chemosphere 2001, 45, (2001), 11-20. 9- . Sims, R. C.; Overcash, M. R., Fate of Polycyclic Aromatic Compounds (PNAs) in Sou-Plant Systems. Residue Reviews 1983, 88, 1-68. 15 10. Luster-Teasley, S. L.; Yao, J. J .; Hemer, H. H.; Trosko, J. E.; Masten, S. J., Ozonation of Chrysene: Evaluation of By-Product Mixtures and Identification of Toxic Constituent. Environmental Science and Technology 2002, 36, (5), 869-876. 11. Upham, B. L.; Masten, S. J .; Lockwood, B. R.; Trosko, J. E., Nongenotoxic Effects of Polycyclic Aromatic Hydrocarbons and Their Ozonation By-Products on Intercellular Communication of Rat Liver epithelial Cells. Fundamental and Applied Toxicology 1994, 23, 470-475. 12. Yin, Y.; Allen, H. E., In Situ Chemical Treatment. July 1999 13. Group, I. T. a. R. C. W. Technology Overview Dense Non-Aqueous Phase Liquids (DNPL): Review of Emerging Characterization and Remediation Technologies. http://www.itrcweb.org/DNAPL-1 .pdf 14. Gates-Anderson, D. D.; Siegrest, R. L.; Cline, S. R., Comparison of Potassium Permanganate and Hydrogen Peroxide as Chemical Oxidants for Organically Contaminated Soils. Journal of Environmental Engineering 2001, 337-347. 15. Beltran, F. J .; Ovejeor, G.; Rivas, J ., Oxidation of Polycyclic Aromatic Hydrocarbons in Water.4. Ozone Combined with Hydrogen Peroxide. Industrial Engineering and Chemistry Research 1996, 35, (3), 891-898. 16. Luster-Teasley, S. The Use of Gaseous Ozone to remediate Pyrene Contaminated Soils: A study of By-Product Production, Environmental effects on Remediation Efforts, and Scale-Up Volume I & II. Dissertation for the Degree of Ph. D., Michigan State University, East Lansing, MI, 2003. 17. Yao, J .-J .; Huang, Z.-H.; Masten, S. J ., Ozonation of Benz[a]anthracene: Pathway and Product Identification. Water Research 1998, 32, (11), 323 5-3244. 18. Yao, J .-J .; Huang, Z.-H.; Masten, S. J ., The Ozonation of Pyrene: Pathway and Product Identification. Water Research 1998, 32, (10), 3001-3012. 19. Benitez, F. J .; Beltran-Heredia, J .; Acero, J. L.; Rubio, F. J ., Chemical Decomposition of 2,4,6-Thrichlorophenol by Ozone, Fenton's Reagent and UV Radiation. Industrial Engineering and Chemistry Research 1999, 38, (1999), 1341 -1 349. 16 20. Haber, F .; Weiss, J ., Catalytic Decomposition of Hydrogen Peroxide by Iron Salts. Proceedings of The Royal Society, London 1934, 147, 332-351. 21. Jones, C. W., Applications of Hydrogen Peroxide. The Royal Society of Chemistry: Cambridge, 1999. 22. Walling, C., F enton's Reagent Revisited. Accounts of Chemical Research 1974, 8, 1 25-13 1 . 23. Watts, R. J.; T eel, A. L., Chemistry of Modified Fenton's Reagent (Catalyzed H202 Propagations-CHP) for In Situ Soil and Groundwater Remediation. Journal of Environmental Engineering 2005, 131, (4), 612-622. 24. Watts, R. J .; Teel, A. L., Treatment of Contaminated Soils and Groundwater Using ISCO. Practice Periodical of Hazardous, Toxic, and Radioactive Waste Management 2006, 10, (1), 2-9. 25. Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. Critical Review Constants for Reactions of Hydrated Electrons Hydrogen Atoms and Hydroxyl Radicals (oOH/O-) in Aqueous Solution; 1988. 26. Walling, C.; Johnson, R. A., Fenton's Reagent V. Hydroxylation and Side-Chain Cleavage of Aromatics. Journal of the American Chemical Society 1974, 97, (2), 363- 367. 27. Augusti, R.; Dias, A. 0.; Rocha, L. L.; Lago, R. M., Kinetics and Mechanism of Benzene Derivative Degradation with Fenton's Reagent in Aqueous Medium Studied by MIMS. Journal of Physical Chemistry 1998, 102, (52), 10723-10727. 28. Casero, I.; Sicilia, D.; Rubio, S.; Perez-Bendito, D., Chemical Degradation of Aromatic Amines By Fenton's Reagent. Water Research 1997, 31, (8), 1985-1995. 29. Walling, C.; Amarnath, K., Oxidation of Mandelic Acid by Fenton's Reagent. Journal of the American Chemical Society 1982, 104, (5), 1 185-1 189. 17 30. Walling, C.; El-Taliawi, Fenton's Reagent III. Addidtion of Hydroxyl Radicals to Acetylenes and Redox Reactions of Vynil Radicals. Journal of the American Chemical Society 1973, 95, (3), 848-850. 31. Teel, A. L.; Warberg, C. R.; Atkinson, D. A.; Watts, R. J ., Comparison of Mineral and Soluble Iron F enton's Catalysts for the Treatment of Trichloroethylene. Water Research 2001, 35, (4), 977-984. 32. Kong, S.-H.; Watts, R. J .; Choi, J .-H., Treatment of Petroleum-Contaminated Soils Using Iron Mineral Catalyzed Hydrogen Peroxide. Chemosphere 1998, 37, (8), 1473-1482. 33. Watts, R. J .; Stanton, P. C.; Howsawkeng, J .; Teel, A. L., Mineralization of a Sorbed Polycyclic Aromatic Hydrocarbon in Two Soils using Catalyzed Hydrogen Peroxide. Water Research 2002, 36, 4283-4292. 34. Kakarla, P. K.; Andrews, T.; Greenberg, R. S.; Zervas, D. 8., Modified Fenton's Processes For Effective In-Situ Chemical Oxidation-Laboratory and Field Evaluation. Remediation 2002, Autumn, 23-36. 35. Pignatello, J. J ., Dark and Photoassisted Fe3+ Catalyzed Degradation of Chlorophenoxy Herbicides by Hydrogen Peroxide. Environmental Science and Technology 1992, 26, (5), 944-951. 36. Kwan, W. P.; Voelker, B. M., Rates of Hydroxyl Radical Generation and Organic Compound Oxidation in Mineral-Catalyzed Fenton-like Systems. Environmental Science and Technology 2003, 37, (6), 1150-1 158. 37. Tyre, B. W.; Watts, R. J .; Miller, G. C., Treatment of Four Biorefractory Contaminants in Soils Using Catalyzed Hydrogen Peroxide. Journal of Environmental Quality 1991, 20, (October-December), 832-838. 38. Pignatello, J. J .; Baehr, K., Ferric Complexes as Catalysts for "Fenton" Degradation of 2,4-D and Metolachlor in Soil. Journal of Environmental Quality 1994, 23, (March-April), 365-370. 18 39. Yamazaki, I.; Piette, L., EPR Spin Trapping Study on the Oxidizing Species Formed in the Reaction of The Ferrous Ion with Hydrogen Peroxide. Journal of the American Chemical Society 1991, 113, (20), 7588-7593. 40. Chen, C. T.; Tafuri, A. N.; Rahrnan, M.; Foerst, M. B., Chemical Oxidation Treatment of Petroleum Contaminated Soil Using Fenton's Reagent. Journal of Environmental Science and Health 1998, A33, (6), 987-1008. 41. Bowers, A. R.; Gaddipati, P.; Eckenfelder Jr., W. W.; Monsen, R. M., Treatment of Toxic or Refractory Wastewaters with Hydrogen Peroxide. Water Science and Technology 1989, 21, 477-486. 42. Morel, F. M.; Hering, J. G., Principles and Applications of Aquatic Chemistry. John Wiley & Sons, Inc.: New York, 1993; p 588. 19 Chapter 2. Hydrogen Peroxide Decomposition Introduction Fenton’s Reagent is a complex set of reactions in which a dilute solution of hydrogen peroxide is catalyzed by polyvalent metals, usually iron, to form hydroxyl radicals. The catalyst is recycled back its original state through a series of complex reactions. Fenton’s Reagent has seen a large increase in demand as an in situ remediation technique but with modifications to the catalyzer and oxidant used. The modified Fenton Reagent systems are becoming known as Catalyzed Hydrogen Peroxide Propagations (CHP). CHP systems utilize higher hydrogen peroxide concentrations than those used in traditional F enton’s reactions; this increase in concentration is usually done as a means of compensating for the nonselective nature of hydroxyl radicals and the amount of oxidizable material present in project sites. The typical reactions in a Fenton Reagent system are: Fe2+ +H202 —)Fe3+ +0H‘ +-0H kl=76M'ls'1 (1) Fe3 + + H202 —> Fe2 + + H+ + H0, . kg: 0.02 M's" (2) H202 +-0H—>H20+H02- k3 =2.7x 107M'1s'l (3) Fe2+ +-0H—>Fe3+ +0H‘ k4=4.3x 108M"s'1 (4) Fe3+ +H02-—)Fe2+ + 02 + H+ k5 =10x104 M's" (5) Fe2+ +H02 -+H+ —)Fe3+ +H202 k6=1.2 x106M“s" (6) 20 H0 -—>-0‘+H+ k7=8x105M“s“ (7) 2 2 Fe3 + + 02— —>Fe2 + + 02 k8 = 1.5x 108 M‘s" (8) Fe2+ + .0; + 2H+ —) Fe3 + + H202 k9=10x107 M's“ (9) Hydrogen peroxide decomposes to oxygen over time; this process is accelerated by the presence of soluble iron and other polyvalent metals, as well as by certain minerals that may be present in the soil. The kinetic rate constant for hydrogen peroxide decomposition to oxygen is rarely found in the literature, nor is the equation ever incorporated in the Fenton’s Reagent system. 1 H202 —)H20+502 k10—? (10) Studies of Fenton’s Reagent reactions avoid the issue of hydrogen peroxide decomposition by using low end concentrations of both hydrogen peroxide and soluble iron (1 uM-lmM), where oxygen evolution is minimal. In field applications much greater peroxide concentrations are commonly used. Typically solutions of 4%-15% hydrogen peroxide injected into the soil or surface waterZI. Oxygen production in these systems would be significant wherever soluble iron species exceeds mM concentrations or where mineral iron is found, and thus the limiting reactant in CHPs, hydrogen peroxide, would decrease at a faster rate than expected, resulting in a decrease in the removal efficiency of the system. 21 Another factor that affects the remediation process is that hydrogen peroxide decomposition is an exothermic reaction releasing 98 kJ/mole of energy; this energy is transformed to heat, which then increases the temperature of the solution accelerating all reactions in the system including kinetic, sorptive and decomposition reactions. Decomposition of hydrogen peroxide from reaction with mineral Iron: Goethite Hydrogen peroxide decomposition on mineral surfaces is easily seen due to the formation of oxygen bubbles on the surface and the diffusion of these throughout the liquid phase. This decomposition rate was found to be zero order by Watts“, which is the expected reaction order for surface catalyzed reactions. Lin and Gurol46 found a relationship for hydrogen peroxide decomposition on goethite to follow the empirical equation: kd = ko(FeOOH)n (11) F eOOH is the concentration of Goethite (g/L) ko has an empirical value of 0.037, dimensionless. kd is the decomposition rate constant in 1/(Ls/mole) In a subsequent paper Lin and Gurol47 give a value for a hydrogen peroxide decomposition rate, kd, on goethite that does not match the value the authors would get from equation 11. The reason for this discrepancy may be due to the equation being incomplete. Using the goethite concentration in g/L would not work because the surface 22 area (SA) and number of available kinetic sites (S) would differ on samples of goethite not obtained from the same manufacturing process, where the size and shape of the particles will vary from batch to batch. The equation found is unique to the goethite used for that particular experiment. For that reason, the goethite concentration (g/ L) should be substituted for the effective surface area of the goethite (m2/ g) in the sample. Method and Materials Chemicals: All chemicals used were A.C.S reagent grade unless stated otherwise. Ferrous perchlorate, ferric perchlorate, Ethyldiamnie acetic acid disodium salt (EDTA), salicylic acid, potassium iodide, bypyridine, goethite (or-FeOOH, 30-50 mesh), and perchloric acid were obtained from Sigma-Aldrich (St. Louis, Mo). Stabilized hydrogen peroxide 30% solution was obtained from Fisher Scientific (Hampton, NH). The hydrogen peroxide concentrations greater than luM were determined by iodometric titration with sodium thiosulfate48. Concentrations of hydrogen peroxide less than 1 uM were titrated by peroxidase catalyzed oxidation of N,N-diethyl-p-phenylenediamine (DPD) as described by Bader et a1.49 and modified by Voelker and Sulzberger50 to minimize interference by ferrous and ferric ions. Preliminary Experiments Experiments were designed to determine the decomposition rate of hydrogen peroxide to oxygen in CHP systems and how the rate of decomposition varies with the concentration of catalyzer, Fe2+, Fe3+, and/or goethite. The tests were performed by measuring water 23 displacement on an inverted graduated cylinder. Borosilicate vials (40 mL) were completely filled with the desired catalyzer solution and capped with a Teflon liner to which the desired amount of hydrogen peroxide was injected through the liner. The liner was cut and fit with a Teflon tube with a gas sparger at the other end. The sparger was placed inside an inverted graduated cylinder filled with water held inverted inside a tub of water (Tonicelli Experiment). As oxygen was produced in the 40 mL vial it flowed through the tube and into the graduated cylinder the gas then displaced water and a volume reading was recorded form the cylinder. The volume of water displaced was then used to back calculate the hydrogen peroxide concentration using the ideal gas law to determine the moles of oxygen produced from the volume reading, and the rate of oxygen production versus hydrogen decomposition of 2 to one to determine the hydrogen peroxide concentration. _fl2_=lfl (12) dt 2 dt Gravimetric Tests: Oxygen Evolution Experiments were conducted to measure the change in mass that occurred in the reactive samples over time using an analytical scale. CHP reactions were carried out in 40 mL borosilicate vials containing 25 mL of an aqueous catalyzer solution. Before initiating the reaction the weight of the vials was recorded. To start the reaction a previously weighed and titrated hydrogen peroxide solution was added to the catalyzer mixture and allowed to react for 2 minutes before the first measurement was taken. The 2 minute time frame was given to allow the aqueous solution to become saturated with the oxygen produced, 24 thus creating steady state conditions for mass loss in the system. For this data to be of significance we have to assmne that the only mass leaving the system would be that of the oxygen produced during the reaction. The mass difference over time was used to calculate the moles of oxygen produced by: Moles of Oxygen = mass (13) 32 g 02 /mole Because the decomposition of hydrogen peroxide to oxygen is an exothermic reaction the temperature of the reaction was monitored and recorded throughout the experiments with the use of temperature probe attached to a Jenway 4330 pH/Conductivity meter. The pH of the solution was also monitored during reactions, but no buffer was added. The pH of the samples was adjusted using 0.1 M perchlorate acid or 0.1M sodium hydroxide solutions. The experiments were carried out for a wide range of catalyzer and oxidant concentrations (10'5 to 1M), for the three catalyzer systems used: ferric ion, ferrous ion and Goethite. Goethite experiments were carried out at pH 7, were Watts et al.45 determined was the pH with the greatest oxygen production. Effect of a chelating agent on Oxygen Evolution CHP reactions often use the addition of a complexing agent to solubilize the ferrous or ferric ion at pH values greater than 3. For this reason, EDTA was added as desired and the effect this complexing agent had on oxygen evolution was recorded. EDTA (0.1 M) 25 was added to the reaction vials containing the soluble iron species and allowed to interact with the iron for 20 minutes before the start of the experiments. Effect of aquifer Material on Hydrogen Peroxide Decomposition To determine the effect that aquifer material might have on the production of oxygen during CHP reactions, the experiments were also carried out as soil Slurries using 20 g of Ottawa sand. Experiments were conducted with the soil either partially or fully saturated with water/catalyst solution. Results from all experiments are averages of at least triplicates. Results Preliminary tests using water displacement as an oxygen production measure show that the rate of decomposition of hydrogen peroxide may be lst order for CHPs using soluble ferric ion as seen in Figure 2-1 and Figure 2-2. Water displacement experiments were replaced with gravimetric tests to improve on precision and accuracy. The displacement of water by oxygen showed a region during the reaction phase in which the oxygen production followed a linear trend. This time frame was then selected for further experiments with soluble iron catalyst. 26 120 f Volume Produced [mL] 0 10 20 30 40 50 60 70 Time [min] Figure 2-1: Volume of Oxygen produced during a CHP with high end hydrogen peroxide concentration (0.5M) catalyzed by ferric ion (0.01M). Volume of air was measured by water displacement on an inverted cylinder. 27 .0 or N . y = 0.0942x + 0.1898 i E ; R2=0.9765 * 1-3 0,5 1 .................. s ........ ......... — ..... a. .— ................. ; .............. . 3 1 g o_4 a..- Q. 0 a E. 5 .3, 0.3 - ----------- s 5 g 0.2 - ......................... 0.1 _ o i ; .~ ; ; 0 o 5 1o 15 20 25 30 Tlme [mln] [—-— Hydrogen Peroxide [mole/L] - i- Ln (Hydrogen Peroxide)] Figure 2-2: Hydrogen peroxide concentration over time, back calculated from the volume of oxygen collected in the inverted cylinder shown in Figure 1. The figure also shows the linear relationship between log concentration and time for the reaction The change in mass in the reaction vials was measured for several hydrogen peroxide and soluble iron concentrations as seen in Figure 2-3. The data was corrected for evaporation by weighing blanks; 40 ml borosilicate vials containing water, catalyzer solution; and diluted hydrogen peroxide were used as evaporation blanks. The change in mass that occurred in these vials, if any, over the experimental time frame was subtracted from the change in mass of the reactions. 28 30 Time [mln] + H202 [0.62] Ferric [0.06] + H202 [0.62] Ferric [0.34] + H202 [0.11] Ferric [0.06] w H202 [0.11] Ferric [0.34] + H202 [0.11] Ferrous [0.06] + H202 [0.62] Ferrous [0.06] -+— H202 [0.11] Ferrous [0.34] Figure 2-3: Moles of oxygen produced during reactions catalyzed with ferrous and ferric soluble iron for varying hydrogen peroxide concentrations. The graph shows that oxygen production is independent of catalyzer concentration. The oxygen production curves were generated from the data collected from soluble iron gravimetric tests. The best fit for the curves and the coefficients found from these fits are found in Table 2-1. The best fit for the curves was found using equation 14. Moles 02 produced(t) = a-eb't +c-ed 't (14) The coefficients from Table 2-1 were then graphed versus the initial hydrogen peroxide concentration to try and establish a correlation between the concentrations of oxidant and catalyzer used, as shown in Figure 2-4. 29 0.0150 1 g ' 0.0250: . e ' : iY=O-?17X'0-001 3 . . 0.0200 1 . y=-0.026x+0.019 E , 00100 i R =0-998 l , 0.0150 1’ R2=0.697 c: 1 . . .o . . 0.0050 I 00100 J o ' i ; . , . 0.0050 1 . , p 0.0000 ram m" ____s.. "'4 * 't’ 3 .1 0.0000 ' r— ,_._.__ 3"“ " i r " ‘7 0.00 0.20 0.40 0.60 0.80 i 0.00 0.20 0.40 0.60 0.80 ‘ Hydrogen Peroxide [M] , Hydrogen Peroxide [M] 0 — 3 : : 0.000 ~ , - : . \. l 3 y , : y=-0.562x-0.159; . —0.005 I ‘ ‘ -O.200 '1’ ' R2=O.948 ' 3 o §y=-1E-02x+1E-04\.° : —c g . 5 '0-01 R2=9E-01 of 2 '0400 T i g . i ‘ ' i . .0 -0.015 t i ' T I -0.600 1 ' - . 0.00 0.20 0.40 0.60 0.80 0.00 0.20 0.40 0.60 0.80 5 Hydrogen Peroxide [M] 3 Hydrogen Peroxide [M] Figure 2-4: Relation between regression coefficients and the initial hydrogen peroxide concentration for oxygen production for soluble iron catalyzed reactions The temperature of the solution during the reaction period was also monitored and a plot of normalized temperature over the experimental time for ferric catalyzed systems can be seen in Figure 2-5. Only three of the trials are shown for clarity. A comparison for the temperature profiles between ferrous and ferric catalyzed reactions is seen in Figure 2-6. ‘30 2.11 .:::,5 , ; i I c _ 11111. .. ii 111.. 1.01.... S 10 15 20 25 30 35 Time [min] _..- H2020.73M ~---H2020.34M ] [ -+-H202 0.11M Figure 2-5: Temperature of the solution was measured during reactions. Here the temperature data is normalized by initial temperature for initial hydrogen peroxide concentrations of 0.11, 0.34 and 0.73 M. The ferric ion concentration used in all experiments was 0.3 M at a pH ranging between 2-3. 31 1.6 . e 1.5 , ' . a. i 1.4 . _.__o __ . __ : i' _ . _ . I i. l i * ~ i ‘ i 01.3 . ~ i+¥+¥ E I -' *‘ 1.21J .' I a". r T I; Jerri-ea 1101 iii—i iii-11H i~iiiiii 1 I 0.9 1’ l 0.3 1- - e _, — - -v— W v -- a - - 0 5 10 15 20 25 30 35 Time min +Ferrous0.06M H202 0.13M [ ]_'_ Ferrous 0.06M H202 0.70M ——A—— Ferric 0.06M H202 0.11M - -- Ferric 0.06M H202 0.62M Figure 2-6: Comparison between solution temperature for ferrous and ferric catalyzed systems, normalized by the initial temperature at 2-3 pH range. As with the oxygen production, the normalized temperature curves were fit through regression analysis, and the values for the best fit equation are shown in Table 2-2. The best fit was found using the equation: —=a-e ' +c-e ' (15) 32 .000 000.0 0000- 0~00 000.0- 080.0 008.9 000.0 0000 00.0 00.0 ~35 .000 80.0 000.0- ~80 -0¢ ~800 008.? 8-00.0 -00 00.0 00.0 03...» _ 039:0“. 500 000.0 0000- 0000 ~00? 008.0 008.? 08.0 0000 0~.0 0.0 00-03:» —000 80 -00- ~80 00~0- ~800 008.9 08.0 ~0~0 00.0 00 03:0 ~000 000.0 0~0.0- 000.0 000.0- 008.0 -000- 0000 0000 0~.0 00.0 03.: .000 000.0 0000- .0~0.0 000.0- 0800 0000.0- 000.0 000.0 00 ~00 035 .000 000.0 000.? 08.0 000.? ~800 008.9 80.0 000.0 00.0 00.0 035 .000 000.0 000.? ~80 000.9 0080 0.080. 03 ~00 00.0 ~00 ~35 .000 000.0 000.9 08.0 000.9 ~800 08.0. 08.0 80 00.0 00.0 03.: _ 0.0.00. _ w. . 0o _ 0 . 00 r e r 00 . 0 . 00 . 0 P 0... _ N0000 . 00¢ 00300090000 0020300 Bum—0:000: Got 0000.050 : 002005.00 00.0 0000065000 000 0000.03 "Nun 030,—. B .000 000.0 :0? 3000.0 0008.0. 0~80 0080 00-0000 0080 00.0 ~00 ~33 .80 000.0 RN? 09000.0 0008.0- 0~80 0080 00-0000 0080 00.0 00.0 035 _ 0:230. .000 000.0- 0000- 00-0000 0008.0- «~80 0080 8-0000 0080 00.0 00.0 0320 80 08.0 03.? 00-0000 0080. 0080 080 00-0000 080 000 00.0 035 .80 0000. 0000. 00-0000 2000.0- 008.0 ~80 00-0000 0000.0 00.0 80 ~35 .000 ~30 000.? 090000 00080- 0080 0080 00-0000 0080 00.0 .000 03.: _ 2:0". _ ~00 _ 0o . 0 _ 0o . 0 . 0o _ 0 _ 0o _ 0 _ 0.0 . N0000 _ 030.00., $030000. 00.0 .08 3200:. 000000-0080 $3 00.0 00 .0 .2 00200050 5 00000005060 00.0 00 00.000 0 0:0 .0 .0. .0 08000 00:. 000300000 0:00 wage 0000680. 000308 00.0 00030000 08000003”. “TN 033—. The coefficients from Table 2-2 were then graphed versus the initial hydrogen peroxide concentration to determine if there was any correlation between them, as shown in Figure 2-7. 2-000 7 z j e , 0.0000 ,. 1.500%: ‘ I I ; ' L I . ‘ -0.0050 1 “’ 1-000 , y=0.940x+0.981 . , J n . ' ' 0-500 I R2=0.973 I : -0.0100 } y=-0-2017X+0-001 - __ - _; 3 ~ '| ; R =0.990 000" ‘ ' E ‘ 0.0150 *- ' ‘ ' ‘ 0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 Hydrogen Peroxide [M] , Hydrogen Peroxide [M] O'OOO ‘ y=-0.922x+0.012} 0000 o -0.400 3 2 --"~’ '- =- - - , . 3 4000,)- 2é812x 0.303 -0.600 3 f g ? - I R =0.419 -0300 + I : ; 3 g 0.000 + - - 0 0.2 04 ()6 0,8 3 o 0.2 0.4 0.6 0.8 Hydrogen Peroxide [M] : Hydrogen Peroxide [M] Figure 2-7: Relation between regression coefficients and the initial hydrogen peroxide concentration for normalized temperature for soluble iron catalyzed reactions. The results for the experiments with Ottawa sand Slurries and EDTA experiments are shown together in Figure 2-8 34 ‘ “ " ,yieiirrrriiiii-iirii-ii - W +- kr.’ _---_—»-~F- Hag-Hm” W“ - - 0 T T l I T O 5 1 0 15 20 25 30 Time [min] + H202 [0.11] Ferric [0.06] ---—- H202 [0.62] Ferric [0.06] + 209 Ottawa -- H202 [0.11] Ferric [0.06] + 5mL EDTA [0.1] - 9- H202 [0.62] Ferric [0.06] Figure 2-8: Comparison of Oxygen production between reactions with and without Ottawa sand and with and without EDTA. Goethite Results from the gravimetric experiments for hydrogen peroxide catalyzed by goethite can be found in Figure 2-8. Fits for the regression analysis on data can be found in Table 2-3 for an exponential regression using equation 14. Table 2-4 shows the coefficients for the same data fit through linear regression. Moles of 02 Produced = a-x+b (16) 35 0.0035 0 :0 - 0003 .- ........................................................................................................................ II 00025 ............................................................................................................... fl & . g 0.002- ---------------- f -------------------------------------------------------- ‘H i : -' 1 : : O 1 : ; § 00015 i-------~-~--- ---------------------- . ---------- ;,0/i ------ o - 5 | :TI/ : 2 /’l . 1 a 1 'I“ O I f LL 1 I fii I ‘T T 0 100 200 300 400 500 600 700 800 Time [min] -._ H202 0.086 Goe 4.74 g/L —-— H202 0.018 Goe 4.74 g/L _.— H202 0.086 Goe 0.82 g/L -~—H202 0.018 Goe 0.82g/L +H202 0.098 Goe 2.79g/L - 0- H202 0.0085 Goc 2.78g/L -8— H202 0.052 Goe 5.54 g/L -—— H202 0.051 Goe 0.028 g/L —-——— H202 0.051 Goe 2.78 g/L Figure 2-9: Oxygen production form Goethite catalyzed reactions 36 33.3 $333 3933- 33.33. 333.3- 333 333.3 33.33 3.33 33 333.3 333.3 33.3 33.3 3-33- 3-33- 3.33.3 3.33.3. 333. 3.33 3-33 3-33.3 m:- 3333 33.3 333 3.3.3..» 333.3 3-33.3 3-33- 3-33 3-33.3- 3-33 3-33 3-33.3 3-33.3 3.3 333.3 32.3 33.3 33.3 333 3.333. 333 3-33- 3-33 3.33 3-33 3-33 33.3 333 32.3 33.3 :35 33.3 33.33 333.3- 333 3-33- 333 3-33 3-33- 333 3.3 33.33 333.3 333.3 .333 3.33.... 3.33.3. 333 3.33- 3-33 3-33 3-33 3-33 2.3 333.3 3.333 333.3 3.35 .333 3-33 3-33- 3-33 333- 3-33 3-33 3-33 3-33 33 333.3 33.3 23.3 .3333 3-33 3-33- 3-33 3-33.3- 3-33 3-33 3-33 3-33 :3 333.3 33.3 :33 33.3 3-33.3 3-33- 3.33 3.33.3- 333.3 3-333 3-33.3 3-33.3 N3 383 33.3 33.3 :33 333.3 3-33.3 3-33.3- 3-33.3 3-33- 33.333 3-33 3-33. 3-33 33.3 R33 333 33.3 33.3 3-33.3 333.3. 3-33 3.33- 3-33.3 3-33 3-33.3 3-33.3 3.3 33.3 233.3 33.3 33.3 3-33.3 333-- 3-33 3.3.3.3. 3333 3-33 3-33 3-3: 33.3 33.3 333.3 33.3 3.: 33.3 3-33.3 3-33- 3-333 3-333- 3333 3-33.3 3-33 3.33 33.3 R33 3333 33.3 333.3 3-33.3 3-33- 333 3.33.3. 3333 3-33. 3-33 3.3.3.3 33.3 33.3 333.3 33.3 .333 3-33.3 3333. 3.33.3 3.33.3. 3.33 333.3 3.333 3-33.3 33 33.3 333.3 33.3 :33 _ H: 6 3 6 g 6 3 U . r523 2+3»... 3.3380 :2. SN: .33 3532on 335:0me =3 Sm 3288598 85300 an 30:3:on 9&sz 8m £3335 3:33:83 .33 3:3 36205300 "mun 333,—- 37 33 5333 3-333 3.33 3-3: «33 3333 333 ~33 33 3-333 3.33 8-333 3-333 .33 3333 333 ~33 3.3-3.35 33 3-333 3.33.3 8.33 3-333 33 333 3333 333 33 3-333 3.33 33-333 3-333 33 3333 3333 333 33 3.33 3-333 8333 3-33 333 333 3333 333 :35 .333 3-333 3-333 3-333 333 333 3333 333 .333 33 3.33.3 3.33.3 8.33 3-333 333 333 333 333 33 3-32 33: 3-333 3-333 33 3333 333 333 3.33 33 3.33.3 3.33 8-333 33.333 33.3 333 333 333 33 3-333 3333 8.33 33.333 333 333 333 333 33 3-333 3.33 3333 3-333 .33 3333 333 333 as.» 33 3-32 3333 8-333 3-333 33 3333 3333 3333 33 3-333 3-333 8-333 3-333 33 .3333 3333 :33 33 3-3: 3.33 3.333 3.3333 33 3333 3333 3333 sat» 33 3-33.3 3-333 8-333 3-333 .33 3333 :33 333 33 3-333 3.33 33-333 3333 .33 3333 3333 333 33 3-333 3.33 33333 3-333 33 3333 3333 333 3.3.: 33 3-333 3-333 8-333 3-333 33 333 333 3333 33 3-333 3-333 8-333 3-333 33 .333 333 3333 33 3-333 3.3333 33-333 33-333 33 333 .333 3333 ~33 33 3.33 333.3 3.333 3.33 33 R33 333 333 33 3-333 3-333 3.333. 3.33 33 333 333 333 33 3-333 3.33 8-333 3-333 33 333 333 333 3.3.: _ m. .u a .u a :5. <3 .2. $3 .2 83 .2. SE .32.... 3350 BSA—333 823300 macaw 33583-698 353.53% 88% 3333 mo 3232on 30:: 8.38 258 33385030 31m «Ea-m. 38 Discussion Soluble Iron Water displacement experiments show, as seen in Figure 2-1 and Figure 2-2, that hydrogen peroxide decomposition in the presence of soluble iron may be first order. Since the decomposition of hydrogen peroxide is exothermic, the temperature of the solution increases as hydrogen peroxide decomposes, thus the rate of decomposition is accelerated. Because of this issue, the water displacement experiment was replaced with a gravimetric test that did not depend on temperature and the precision of the measurements could be improved. As seen in Figure 2-3 the gravimetric data collected show that hydrogen peroxide decomposition with soluble iron is independent of the catalyzer concentration and the oxidation state of the soluble iron. Results from reactions containing 0.06 M and 0.34 M ferric ion concentration and 0.11 M hydrogen peroxide concentration show no significant difference within 95% confidence for the rate of oxygen production. Figure 2-3 also shows that hydrogen peroxide concentration appears to control the rate of oxygen production, and that higher hydrogen peroxide concentrations produce oxygen at a higher rate for the same catalyzer concentration. It can be seen in this figure that the moles of oxygen produced by an initial hydrogen peroxide concentration of 0.62 M is significantly higher than those produced from an initial hydrogen peroxide 0.11 M concentration for either ferric ion concentrations of 0.06 or 0.34 M. This finding is also consistent with a first order rate expression with respect to hydrogen peroxide decomposition. 39 Oxygen production from ferrous ion catalyzed systems was not significantly different from ferric ion catalyzed systems for the same catalyzer and hydrogen peroxide concentrations. The reason for this is that the Fenton system initiation reaction, equation 1, is faster than the decomposition reaction; thus most, if not all, of the ferrous ions react with hydrogen peroxide to form ferric ion and hydroxyl radicals, before the hydrogen peroxide decomposes. This leads to reactions being catalyzed mostly by ferric ion, even if, initially, iron ions were in the ferrous state. The data obtained from oxygen production with both ferrous and ferric ions, was fitted with a regression line to better understand the process. The best fit curve for the data is a double exponential expression shown in equation 13. The coefficients from this fit were then graphed versus the initial hydrogen peroxide and iron concentrations to determine if any correlation between them existed. No correlation was found between iron concentration and any of the coefficients, but there appeared to be a strong correlation between hydrogen peroxide concentration and the coefficients, seen in Figure 2-4, this finding is consistent with what has been described in the previous paragraph about Figure 2-3. The value of coefficient “a” for this system increases as hydrogen peroxide concentration in what appears to be a linear trend. Inversely coefficient “b” decreases linearly as hydrogen peroxide concentration is increased. These results appear to indicate that the first term of the equation and coefficient “a” are related to the decomposition of hydrogen peroxide into oxygen as it reacts with the catalyzer; while the second term of the equation may be dependent on either the release 40 of oxygen from solution; or the reaction of hydrogen peroxide with the ration of iron catalyzer that is cycling back to its initial state. Temperature The reason why the best fit for the oxygen production is an exponential equation may be because of the increase in solution temperature that results as hydrogen peroxide decomposes. The temperature of the solutions was monitored throughout the experiments, in order to correct the rate constant obtained from oxygen production for hydrogen peroxide decomposition. Figure 2-5 shows the normalized temperature profiles between three different initial hydrogen peroxide concentrations catalyzed with 0.3 M ferric ion. As expected the higher hydrogen peroxide concentration produces the highest temperature increase in the solution. The normalized temperature profiles show that the maximum reaction temperature is reached early in the reaction period, before 5 minutes, after which the solution is slowly cooled down by the release of oxygen, which absorbs energy; and convection processes, due to the difference between ambient temperature and solution temperature. Normalized temperature profiles for the reactions between ferrous and ferric catalyzed systems appear to differ only in their initial increase in temperature. Ferrous catalyzed reactions reach maximum temperature earlier than ferric catalyzed systems, shown in Figure 2-6. This difference in temperature increase causes ferrous reactions to be more violent than ferric catalyzed reactions, making ferrous catalyzed systems more dangerous to work with. This rate of temperature increase also leads to an increase in the initial rate 41 of hydrogen peroxide decomposition and sometimes forces the reactions to become diffusion limited when certain concentrations of hydrogen peroxide and the catalyzer are mixed. The solutions become diffusion limited when the reactions occur faster than the hydrogen peroxide can mix in the solution. By testing certain hydrogen peroxide and ferrous catalyzer ratios, it was found that concentrations higher than 0.5M ferrous ion in solution combined with concentrations 8% or higher of hydrogen peroxide produce diffusion limited reactions. Ferrous and ferric catalyzed systems only differ, in terms of hydrogen peroxide decomposition, by the steepness of the temperature elevation at the start of the experiment. This could mean that hydrogen peroxide is preferentially decomposed when the ferrous ion is present, explaining why the ferric ion reactions were observed to be less violent than those catalyzed by ferrous ion. Ferric catalyzed systems require that the reaction shown in equation 2 occur before hydrogen peroxide decomposes or catalyzes to hydroxyl radical. The decomposition of hydrogen peroxide may only be occurring when the iron cycling process produces ferrous ion in the solution. Since the ending point for reactions in both systems appears to be the same, this point may be irrelevant when comparing oxygen production. The temperature profiles from the reactions were fit with a regression analysis like those performed for oxygen production. The best fit for this regression was found with equation 14. This appears to indicate that the oxygen production and the temperature of the solution are directly related. As before, the coefficients from this regression analysis were graphed versus catalyzer and initial hydrogen peroxide concentration, and found that the 42 latter appears to have strong correlation to the coefficients presented in Table 2-4. This is expected as well because the cause for temperature increase is the decomposition of hydrogen peroxide. Ottawa Sand Slurries Soluble iron Ottawa sand was saturated with water by adding 4.5 mL of water/catalyzer solution to 20 gm of sand. Under completely saturated conditions reactions in Ottawa sand reactions with soluble iron catalysts were found not to be significantly different fi'om that observed in solutions without Ottawa sand, as shown in Figure 2-8. This means that under these conditions the sand did not interfere or interact with the CHP reactions and did not affect oxygen production. Results from reactions with unsaturated Ottawa sand could not be reproduced, because of oxygen was not released freely into the atmosphere during reaction. Throughout the experiments with unsaturated sand, air bubbles were seen forming inside the sand matrix, and it appeared that the sand particles prevented the release of the oxygen being produced. EDTA experiments Oxygen production in CHP reactions decreased with the addition of EDTA, and continued to decrease as the EDTA concentration increased. At EDTA concentrations less than 2.5 mM, no significant effect on oxygen production was observed when compared to reactions conducted in the absence of EDTA, as seen in Figure 2-8. When the concentration of EDTA reached 2.5 mM, oxygen production decreased. The oxygen 43 production continued to decrease as more EDTA was added, with oxygen production becoming negligible when the EDTA concentration reached 25 mM. Figure 2-8 shows the oxygen production for a solution containing 12.5 mM of EDTA. The oxygen production rate found may be only due to uncomplexed iron reacting with the hydrogen peroxide. Further experiments and modeling iron speciation under these conditions may shed more light on this phenomenon. Pignatello and Baehr 40 have shown that hydroxyl radicals are generated during Fenton reactions catalyzed by ferric-complexes. Ferric complexes are used as a substitute catalyzer because these do not precipitate with an increase in pH. By using ferric complexes we may be able to carry out CHPs while using high hydrogen peroxide concentrations without the risk of losing a significant amount of the oxidant due to oxygen decomposition. Inadvertently we may be also able to diminish the decomposition of hydrogen peroxide, allowing for the use of higher concentrations of the oxidant. Goethite Oxygen production from goethite catalyzed CHPs appears to be dependent on the mineral concentration, and thus surface area, as shown in Figure 2-9. The rate of oxygen production, over a 24 hour reaction time, for all samples goes through an initial exponential production stage and later settles into a linear production stage. The initial production stage occurs within the first hour of commencement of the reaction. The main difference between the different catalyzer to oxidant ratios tested appears to be this initial stage of the reaction. 44 A regression analysis of the data in Table 2-4 was conducted to determine the relationship between the oxidant concentration and catalyzer concentration. The coefficients obtained from the regression analysis along with their 95% confidence interval are given in Tables 2-3 and 2-4. Table 2-3 is a compilation of the coefficients from an exponential fit using equation 13. The first term of the equation refers to the initial surge in oxygen production. Coefficient “b” in Table 2-3 is the exponent of the first term which may be related to the hydrogen peroxide decomposition rate constant (km), if we compare the two terms in equation 13 as two first order reaction rate equations in integrated form. By the same analogy, the second term of the equation may refer to the stabilization of the reaction where the oxygen production is now limited by the number of sites interacting in the mineral surface. Since the decomposition reaction and the hydroxyl radical reaction both occur at the mineral surface, the reactions compete for the active sites available at the mineral surface. The coefficients for the first term of equation 13, in Table 2-3, may be dependent on both hydrogen peroxide and goethite concentration, not just goethite concentration. By analyzing the data fiom Trials 1 and 2 we can see that a decrease in hydrogen peroxide concentration from 0.086M to 0.018M, with the same goethite concentration, resulted in a decrease in the value of “a” by a factor of 2. Meanwhile the value of “b” increases by the same factor for these same reactions. This finding suggests that low concentrations of hydrogen peroxide decompose at a faster rate than higher concentrations for the same goethite concentration. 45 Watts et. a1“5 and Lin and Gurol46 found that the decomposition of hydrogen peroxide on goethite follows zero order kinetics. If we ignore the first 2 data points for each experiment, the data appears to follow a linear trend. By fitting a straight line to the data we obtain the regression coefficients and their respective, 95% confidence intervals, as shown in Table 2-4. From this data we can see that the rate constant for oxygen production, for all goethite and hydrogen peroxide concentrations use d in this study, is relatively constant with a value of 1.62 x 10'6 i 2.25 x 10'7 M min'1 (2.70 x 10'8 :t 3.75 x 10'9 M s") for 24 hours of reaction. The elimination of the first two data points from the analysis may be justified by the fact that the solution required this time to equilibrate or to reach oxygen saturation at which point the oxygen production and desorption should reach steady state conditions. Lin and Gurol’s equation for hydrogen decomposition from goethite catalysis, does not appear to hold true for our experiments, because the value of the decomposition constant obtained fiom equation 11, would vary for each of our experiments due to the different goethite concentrations used. The decomposition constant according to Lin and Gurol, would range from 0.029 to 0.178 M s'1 for the range of goethite concentrations used. These results are obtained by substituting the value of FeOOH in equation 11. In our experiments we observed a decomposition rate which is 2 to 3 orders of magnitude less than the rate constants found by Lin and Gurol. It should be noted that the value we found only takes into account the decomposition of hydrogen peroxide to oxygen and does not take into account the catalysis of hydrogen peroxide into hydroxyl radicals, which is also a major hydrogen peroxide sink in the system. Since Lin and Gurol determined their rate 46 constant by hydrogen peroxide titration, they have incorporated the hydroxyl radical formation into their rate constant. The data collected for oxygen production from all the experiments can be used to generate kinetic models for the CHPs tested. These models are developed in Chapter 6. 47 Conclusions The decomposition of hydrogen peroxide into oxygen is an important hydrogen peroxide sink during CHP reactions, especially when high concentrations of hydrogen peroxide are used. This reaction should be added to any models that account for hydrogen peroxide loss. There is a limit to the concentrations that can be used for CHPs that are catalyzed by soluble iron. Solutions having high concentrations of hydrogen peroxide (28% or 2.5x10' 3 M) and high soluble ferrous ion concentrations (0.5 M) become diffusion limited toward hydrogen peroxide, thus creating a boundary to which these chemicals can be mixed to form an effective remediation tool The empirical relationship determined in this study can be used to predict the decomposition rate of hydrogen peroxide or the formation rate of oxygen in an in-situ environment. The relationship can help in creating kinetic models and to better assess the chemical concentrations that can be used in CHP remediation systems. The rate of oxygen production was found to be independent of the concentration and oxidative state of soluble iron used. Hydrogen peroxide decomposition was found to be first order. The temperature of the reaction can be determined from the empirical equations obtained in this study, which could also help for in situ remediation modeling by enabling us to correct for the effect of temperature during the reactions. 48 The use of a complexing agent like EDTA would help minimize the loss of hydrogen peroxide by decomposition when soluble iron catalysts are used. Oxygen production is not affected by Ottawa sand in completely saturated conditions, when soluble iron is used as the catalyzer. If the solution is not saturated the reactions suffer from channeling and surface-bubble interactions that affect the reproducibility of the experiments. Goethite catalyzed systems appear to have two oxygen production reaction rates; an initial rate which occurs during the first hour of reaction, and a secondary linear rate, which was found to be constant for all hydrogen peroxide and goethite concentrations studied. The rate constant was found to be 2.70 x 10'8 M s". 49 References l. Kostecki, P. T.; Calabrese, E. J ., Petroleum Contaminated Soils: Remediation Techniques, Environmental Fate, and Risk Assesment. 3rd ed.; Lewis Publishers, Inc.: Chelsea, MI, 1989; Vol. 1. 2. Juhasz, A. L.; Naidu, R., Bioremediation of High Molecular Weight Polycyclic Aromatic Hydrocarbons: A Review of Microbial Degradation of Benzo[a]pyrene. International Biodeterioration & Biodegradation 2000, 45, 57-88. 3. Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M., Environmental Organic Chemistry. lst ed.; Wilkey-Interscience, John Wiley & Sons Inc: New York, 1 993. 4. Chiou, C. T.; McGroddy, S. E.; Kile, D. E., Partition Characteristics of Polycyclic Aromatic Hydrocarbons on Soils and Sediments. Environmental Science and Technology 1998, 32, (2), 264-269. 5. Keith, L. H.; Telliard, W. A., Priority Pollutants: Prospective View. Environmental Science and Technology 1979, 13, (4), 416-424. 6. Srivastava, V. J .; Kelley, R. L.; Paterek, J. R.; Hayes, T. D.; Nelson, G. L.; Golchin, J ., A Field Scale Demonstration of Novel Bioremediation Process for MGP Sites. Applied Biochemistry and Biotechnology 1994, 45, (6), 741-756. 7. Heitkamp, M. A.; Freeman, J. P.; Cemeglia, C. E., Naphthalene Biodegradation in Environmental Microcosms: Estimates of Degradation Rates and Characterization of Metabolites. Applied and Environmental Microbiology 1987, 53, (1), 129-136. 8. Nam, K.; Rodriguez, W.; Kukor, J. J ., Enhanced Degradation of Polycyclic Aromatic Hydrocarbons by Biodegradation Combined with a Modified Fenton Reaction. Chemosphere 2001, 45, (2001), 11-20. 9. Sims, R. C.; Overcash, M. R., Fate of Polycyclic Aromatic Compounds (PNAs) in Soil-Plant Systems. Residue Reviews 1983, 88, 1-68. 50 10. Luster-Teasley, S. L.; Yao, J. J .; Hemer, H. H.; Trosko, J. E.; Masten, S. J ., Ozonation of Chrysene: Evaluation of By-Product Mixtures and Identification of Toxic Constituent. Environmental Science and Technology 2002, 36, (5), 869-876. 11. Upham, B. L.; Masten, S. J .; Lockwood, B. R.; Trosko, J. E., Nongenotoxic Effects of Polycyclic Aromatic Hydrocarbons and Their Ozonation By-Products on Intercellular Communication of Rat Liver epithelial Cells. Fundamental and Applied Toxicology 1994, 23, 470-475. 12. Yin, Y.; Allen, H. E., In Situ Chemical Treatment. July 1999, 1999. 13. Group, I. T. a. R. C. W. Technology Overview Dense Non-Aqueous Phase Liquids (DNPL): Review of Emerging Characterization and Remediation Technologies. http://www.itrcweb.org/DNAPL-1.rfif 14. Carve], D. D.; Cartwright, R. T. In Measuring in situ F enton's application success, Proceedings of the Third International Conference on Remediation of Chlorinated and Recalcitrant Compounds, Monterrey, CA, United States of America, May 20-23 2002, 2002; Press, B., Ed. Battelle Press: Monterrey, CA, United States of America, 2002; pp 1173-1178. 15. Gates-Anderson, D. D.; Siegrest, R. L.; Cline, S. R., Comparison of Potassium Permanganate and Hydrogen Peroxide as Chemical Oxidants for Organically Contaminated Soils. Journal of Environmental Engineering 2001, 337-347. 16. Beltran, F. J .; Ovejeor, G.; Rivas, J ., Oxidation of Polycyclic Aromatic Hydrocarbons in Water.4. Ozone Combined with Hydrogen Peroxide. Industrial Engineering and Chemistry Research 1996, 35, (3), 891-898. 17. Luster-Teasley, S. The Use of Gaseous Ozone to remediate Pyrene Contaminated Soils: A study of By-Product Production, Environmental effects on Remediation Efforts, and Scale-Up Volume I & II. Dissertation for the Degree of Ph. D., Michigan State University, East Lansing, MI, 2003. 18. Yao, J .-J .; Huang, Z.-H.; Masten, S. J ., Ozonation of Benz[a]anthracene: Pathway and Product Identification. Water Research 1998, 32, (11), 3235-3244. 19. Yao, J .-J .; Huang, Z.-H.; Masten, S. J ., The Ozonation of Pyrene: Pathway and Product Identification. Water Research 1998, 32, (10), 3001-3012. 51 20. Benitez, F. J .; Beltran-Heredia, J .; Acero, J. L.; Rubio, F. J ., Chemical Decomposition of 2,4,6—Thrichlorophenol by Ozone, Fenton's Reagent and UV Radiation. Industrial Engineering and Chemistry Research 1999, 38, (1999), 1341-1349. 21 . Council, I. T. R. Technical and Regulatory Guidelines for In-Situ Chemical Oxidation of Contaminated Soil and Groundwater 2nd ed. ISCO-2; Interstate Technology & Regulatory Council: Washington, DC, 2005. 22. Haber, F.; Weiss, J ., Catalytic Decomposition of Hydrogen Peroxide by Iron Salts. Proceedings of The Royal Society, London 1934, 147, 332-351. 23. Jones, C. W., Applications of Hydrogen Peroxide. The Royal Society of Chemistry: Cambridge, 1999. 24. Walling, C., Fenton's Reagent Revisited. Accounts of Chemical Research 1974, 8, 125-13 1 . 25. Watts, R. J .; Teel, A. L., Chemistry of Modified Fenton's Reagent (Catalyzed H202 Propagations-CHP) for In Situ Soil and Groundwater Remediation. Journal of Environmental Engineering 2005, 131, (4), 612-622. 26. Watts, R. J .; Teel, A. L., Treatment of Contaminated Soils and Groundwater Using ISCO. Practice Periodical of Hazardous, Toxic, and Radioactive Waste Management 2006, 10, (1), 2-9. 27. Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. Critical Review Constants for Reactions of Hydrated Electrons Hydrogen Atoms and Hydroxyl Radicals (oOH/O-) in Aqueous Solution; 1988. 28. Walling, C.; Johnson, R. A., Fenton's Reagent V. Hydroxylation and Side-Chain Cleavage of Aromatics. Journal of the American Chemical Society 1974, 97, (2), 363- 367. 29. Augusti, R.; Dias, A. O.; Rocha, L. L.; Lago, R. M., Kinetics and Mechanism of Benzene Derivative Degradation with Fenton's Reagent in Aqueous Medium Studied by MIMS. Journal of Physical Chemistry 1998, 102, (52), 10723-10727. 30. Casero, 1.; Sicilia, D.; Rubio, S.; Perez-Bendito, D., Chemical Degradation of Aromatic Amines By Fenton's Reagent. Water Research 1997, 31, (8), 1985-1995. 52 31. Walling, C.; Amamath, K., Oxidation of Mandelic Acid by F enton's Reagent. Journal of the American Chemical Society 1982, 104, (5), 1 185-1189. 32. Walling, C.; El-Taliawi, Fenton's Reagent III. Addidtion of Hydroxyl Radicals to Acetylenes and Redox Reactions of Vynil Radicals. Journal of the American Chemical Society 1973, 95, (3), 848-850. 33. Teel, A. L.; Warberg, C. R.; Atkinson, D. A.; Watts, R. J ., Comparison of Mineral and Soluble Iron Fenton's Catalysts for the Treatment of Trichloroethylene. Water Research 2001, 35, (4), 977-984. 34. Kong, S.-H.; Watts, R. J .; Choi, J .-H., Treatment of Petroleum-Contaminated Soils Using Iron Mineral Catalyzed Hydrogen Peroxide. Chemosphere 1998, 37, (8), 1473-1482. 35. ' Watts, R. J .; Stanton, P. C.; Howsawkeng, J .; Teel, A. L., Mineralization of a Sorbed Polycyclic Aromatic Hydrocarbon in Two Soils using Catalyzed Hydrogen Peroxide. Water Research 2002, 36, 4283-4292. 36. Kakarla, P. K.; Andrews, T.; Greenberg, R. S.; Zervas, D. 8., Modified Fenton's Processes For Effective In-Situ Chemical Oxidation-Laboratory and Field Evaluation. Remediation 2002, Autumn, 23-36. 37. Pignatello, J. J ., Dark and Photoassisted Fe3+ Catalyzed Degradation of Chlorophenoxy Herbicides by Hydrogen Peroxide. Environmental Science and Technology 1992, 26, (5), 944-951. 38. Kwan, W. P.; Voelker, B. M., Rates of Hydroxyl Radical Generation and Organic Compound Oxidation in Mineral-Catalyzed Fenton-like Systems. Environmental Science and Technology 2003, 37, (6), 1 150-1158. 39. Tyre, B. W.; Watts, R. J .; Miller, G. C., Treatment of Four Biorefractory Contaminants in Soils Using Catalyzed Hydrogen Peroxide. Journal of Environmental Quality 1991, 20, (October-December), 832-838. 40. Pignatello, J. J .; Baehr, K., Ferric Complexes as Catalysts for "Fenton" Degradation of 2,4-D and Metolachlor in Soil. Journal of Environmental Quality 1994, 23, (March-April), 365-370. 53 41. Yamazaki, 1.; Piette, L., EPR Spin Trapping Study on the Oxidizing Species Formed in the Reaction of The Ferrous Ion with Hydrogen Peroxide. Journal of the American Chemical Society 1991, 113, (20), 7588-7593. 42. Chen, C. T.; Tafuri, A. N.; Rahrnan, M.; Foerst, M. B., Chemical Oxidation Treatment of Petroleum Contaminated Soil Using F enton's Reagent. Journal of Environmental Science and Health 199.8, A33, (6), 987-1008. 43. Bowers, A. R.; Gaddipati, P.; Eckenfelder Jr., W. W.; Monsen, R. M., Treatment of Toxic or Refractory Wastewaters with Hydrogen Peroxide. Water Science and Technology 1989, 21, 477-486. 44. Morel, F. M.; Hering, J. G., Principles and Applications of Aquatic Chemistry. John Wiley & Sons, Inc.: New York, 1993; p 588. 45. Watts, R. J .; Foget, M. K.; Kong, S.-H.; Teel, A. L., Hydrogen Peroxide Decomposition in Model Subsurface Systems. Journal of Hazardous Materials 1999, 69, (2), 229-243. 46. Lin, S.-S.; Gurol, M. D., Heterogeneous Catalytic Oxidation of Organic Compounds by Hydrogen Peroxide. Water Science and Technology 1996, 34, (9), 57-64. 47. Lin, S.-S.; Gurol, M. D., Catalytic Decomposition of Hydrogen Peroxide on Iron Oxide: Kinetics, Mechanism, and Implications. Environmental Science and Technology 1998, 32, (10), 1417-1423. 48. Christian, G. D., Analytical Chemistry. 5th ed.; John Wiley & Sons, Inc.: New York, 1994; p 812. 49. Bader, H.; Sturzenneger, V.; Hoigne, J ., Photometric Method for the Determination of Low Concentrations of Hydrogen Peroxide by the Peroxidase Catalyzed Oxidation of N,N-Diethy1-p-Phenylenediamine (DPD). Water Research 1988, 22, (9), 1 109-1 1 15. 50. Voelker, B. M.; Sulzberger, B., Effects of Fulvic acid on Fe(II) Oxidation by Hydrogen Peroxide. Environmental Science and Technology 1996, 30, (4), 1106-1114. 54 Chapter 3. Hydroxyl Radical Production Introduction Hydroxyl radical ('OH) production experiments using catalyzed hydrogen peroxide propagations were performed in order to obtain the optimum concentrations of catalyzer and reagent required to generate a maximum yield of radicals. Hydroxyl radicals can be determined by indirect measurement, in which a probe compound reacts with the 'OH and the concentration of the probe is measured and related to the concentration of OH 1 ) radicals. Probe compounds are chemicals with high rates of reaction (109-10IO M'1 s' with CH, which form byproducts that are easily measured after reaction. Chemicals such 52,53 - d 54 56’ as benzoic acid 51, n—propanol 51, dimethyl sulfoxide (DMSO) , salicylic aci D-phenylalanine 57, terephthalate 58, 3-cloro benzoic acid 59, 4-hydroxy benzoic acid 60 and formic acid 61 have been used to quantify the ‘OH. The change in concentration of the probe over time can be correlated to the ‘OH concentration if we assume that only hydroxyl radicals react with the probe. Salicylic Acid (SA) has been found to be a very selective "OH probe 54, making it the frontrunner for 'OH determination. The selectivity of SA is due to the effect the hydroxyl and carboxyl substitutions have on an aromatic ring. The hydroxyl group is in the ortho position with respect to the carboxyl group in the SA molecule. The position of these structures forces hydroxyl radicals to act on carbons 3 and 5 of the ring, thus creating 2,3 55 and 2,5 dihydroxy benzoic acids as the initial products of the reaction of hydroxyl radical and SA. These dihydroxy benzoic acids along with catechol have been found as byproducts of the reaction54’ 62. The reaction rate for the hydroxylation of the aromatic ring in SA was found to be 2.7 x 1010 L mol'1 8’1 by Amphlett and Adams 63 using pulse radiolysis. O O OH OH 'OH OH Salicylic Acid HO O O O OH OH HO OH OH 2,3-Dihydroxy Benzoic Acid 2,5-Dihydroxy Benzoic Acid 'OH {:02 OH OH Catechol Figure 3-1: Pathway for the reaction of hydroxyl radicals with salicylic acid and its 54 products . 56 Method and Materials: Materials Chemicals: All chemicals used were A.C.S reagent grade unless stated otherwise. Ferrous perchlorate, ferric perchlorate, salicylic acid, potassium iodide, bypyridine, goethite (a- F eOOH, 30-50 mesh), and perchloric acid were obtained fiom Sigma-Aldrich (St. Louis, MO). Stabilized hydrogen peroxide 30% solution was obtained from Fisher Scientific (Hampton, NH). Soluble Iron Experiments Experiments were designed to determine hydroxyl radical concentrations during soluble iron catalyzed reactions in both aqueous solutions and in Ottawa sand soil Slurries. No suitable hydroxyl radical probe was found for soluble iron reactions (ferric and ferrous) at the concentrations used. Most of the above mentioned probes either complexed with iron in solution (formic acid, SA, 4-hydroxybenzoic acid, propanol) or were not sufficiently soluble (benzoic acid) to obtain a measurable amount in solution. Finally DMSO could not be used because the iron concentration caused interference during analysis. Hydrogen peroxide concentrations were measured during these reactions, using the DPD49 method modified to decrease iron interference”, and the data collected was used to calibrate a kinetic model. The kinetic model was then used to estimate of hydroxyl radical concentration for the reactions. These results are presented in Chapter 6. 57 Goethite experiments: Goethite catalyzed hydrogen peroxide propagations were carried out in solution and in aqueous suspensions of Ottawa sand. Salicylic acid was used as the 'OH probe. Central Composite designs, first employed by Watts35 and described by Diamond“, were used to obtain the optimal oxidizer to catalyzer ratio for SA removal. Concentrations of hydrogen peroxide greater than 3% were used. Stock solutions of salicylic acid were made fresh each week. Solution reactions were carried out at pH 3 and pH 7. All solutions and suspensions were buffered using a 10 mM phosphate. A total volume of 20 mL of solution was used for all OH/goethite experiments in aqueous solution (no Ottawa sand). Prior to use, goethite was cleaned by rinsing the mineral with DI water to remove any iron dust from the surface; the mineral was then dried in an oven at 105 °C and later placed at 550 °C for 2 hours to remove any organic material present. Surface area for the goethite was measured using BET analysis and found to be 74.5 m2/ gm. For Ottawa sand, 20 grams of sand were used. Goethite was added to each vial while being shaken in a Mistral multimixer to help distribute the mineral throughout the sand. 4.5 mL of water containing the desired SA concentration was added to each vial. The system was allowed to equilibrate for 24 hours in a dark hood prior to the start of reactions. After the addition of hydrogen peroxide, the vials were mixed gently (at 85 cycles/hour for the duration of the experiment) in an orbital shaker. 58 The vials were covered to prevent photo-catalyzed reactions from occurring. pH was monitored during the reaction and any pH adjustments required were performed by adding drop wise 1M perchloric acid solution or IM sodium hydroxide solution. Salicylic Acid Determination Salicylic acid concentrations were measured by RP-HPLC using an isocratic mobile phase consisting of 60% acetonitrile and 40% water at a 0.5 mL per minute flow rate, with a total runtime of 15 minutes. A Waters 2487 Dual absorbance UV detector set at 234 nm was used for detection along with a Waters MODEL fluorescence detector set at excitation and emission wavelengths of 310 and 400 nm, respectively. The detection limit for the UV detector was 0.59 ppm, while the detection limit with the fluorescence detector was 2.2 ppb. Time dependent SA reactions Experiments were performed to determine how SA concentrations changed over time and to determine the rate of hydroxyl radical production. The experiments were carried out using the same methodology as described earlier, while 100 uL samples were taken at 5 minutes, 30 minutes, 4 hours, 24 hours and 72 hours after the addition of hydrogen peroxide to the system. 59 Byproduct Identification GC/MS Samples were decanted after 72 hours of reaction time, in order to remove the solution phase from the goethite. No additional reagents were added to quench reactions, because the reaction will not take place if goethite was not present in the vial. The samples collected were placed under nitrogen gas until they were completely dry. After drying 50 uL of BSTFA (O-bis(trimethylsilyl)acetamide) with 1% TMS (trimethylsilyl), and 15 uL of pyridine were added to each sample and placed in an oven at 100°C for 2 hours to allow derivatization to occur. Five micro liters (SuL) of the solution were then injected into the GC/MS fitted with a DB-5 capillary column. LC/MS A Shimadzu LC-20AD HPLC with a Waters Z-spray electrospray ion source was used in this study. Data acquisitions were performed in both positive and negative ion modes. A 150 mm length 1mm diameter, 5p. Thermo Hypersil-Keystone BetaBasic-18 PIONEER column was used. The LC used a gradient program with acetonitrile and water containing 0.15% formic acid as mobile phase and a flow rate of 0.1 mL per minute. The initial phase of the gradient program contained 10% acetonitrile, which was increased linearly to 60% over 30 minutes, after which the acetonitrile concentration is increased to 100% over ten minutes. At this time, the concentration of acetonitrile was decreased to 60% over ten minutes, and finally the concentration was decreased immediately to 10% and held at this concentration for 5 minutes. The program had a total run time of 60 minutes. 60 The LC/MS detector was set with a cone voltage of 40 V and a capillary voltage of 3.17 kV. The desolvation and cone gas (N2) had flow rates of 400 L/hr and 10 L/hr, respectively. To extract SA from Ottawa sand, 10 mL of water were added to the slurry and mixed thoroughly for 4 hours before sampling. The sample was filtered through a 22 um membrane filter to remove any suspended material. Atomic Absorption Atomic absorption analysis was performed on samples after 72 hours of reaction time. A Perkin Elmer AAnalyst 400 using flame ionization and hollow cathode iron lamp (DL3uNxFeE6) was used for the analysis at a 280 nm wavelength. A 1 mL sample of the water was passed through the flame, the difference in absorption was measured and the concentration determined using a previously created calibration curve. The calibration curve standards ranged from 0.5 to 25 pg of iron, and were checked before each run. The data collected proved that less than 3 pg of iron, the detection limit for our method, were dissolved in solution after 72 hours of reaction at pH 3. 61 Results: Soluble Iron The concentration of hydrogen peroxide over time for soluble iron reactions was measured, an example of which can be seen in Figure 3-2. This data was then used in combination with a kinetic model to obtain a value for the rate constant for the decomposition of hydrogen peroxide, kd. Hydrogen Peroxide [moles/L] Time [min] Figure 3-2: Plot of the change in hydrogen peroxide concentration as a function of time for initial ferric ion concentration of 0.06 M. Hydrogen peroxide was measured by iodometric titration. The initial concentration of hydrogen peroxide is shown in the legend 62 Goethite During experimentation reaction blanks were run in parallel to account for the loss of salicylic acid by sorption onto the mineral. All concentrations have been adjusted by this value. At pH 7, in absence of hydrogen peroxide, 16.3 :1: 1.98% salicylic acid sorbed onto 8.89 g/.L of goethite, while at pH 3, 38.7 d: 3.2% was lost to sorption. Salicylic acid removal efficiency appeared to differ between pH 3 and pH 7. A comparison of SA removal between pH 3 and 7, for different SA concentrations (and the same oxidant/catalyzer ratio) can be seen in Figure 3-8. For an initial SA concentration of 10 ppm, the removal efficiencies were found to be 56 :1: 2% and 64 d: 0.5% respectively; while at an SA concentration of 100 ppm SA was removed 57 i 1% and 16 i 2% by the same oxidizer/catalyzer ratio. The concentrations of hydrogen peroxide studied ranged from 0.001 to 0.85 M, while goethite was held between 2 x104 and 0.1 M as ferric ion. Table 3-1 shows the data for aqueous phase experiments at pH 7 for the oxidation of 100 ppm SA. The trial number in the table represents the points required for the Central Composite Design analysis. This table also details the hydrogen peroxide and iron concentrations used for each trial along with the weighed amount of mineral and the corrected values of iron concentration present. The Results section of Table 3-1 gives the concentration of SA found in each vial, the amount of SA lost due to absorption into the mineral and the percent removal achieved after 72 hours of reaction time. The estimated moles of hydroxyl radicals 63 produced over the course of the experiment are shown in the last column of Table 3-1. This value was estimated from the concentration of salicylic acid removed; assuming one mole of SA is removed by one mole of 'OH. The percent removal shown in Table 3-1 was then used to generate a regression equation to help predict the potential removal of SA from solution at pH 7, as described by the central composite design method 64 This regression equation was found to be: Z(H,F) = 0.16 - 0.084H + 0.039F — 0.0022H2 (1) H and F are dependent on the hydrogen peroxide and goethite concentration as ferric ion, and can be found by: H o -0.43 :1 2—%)J30 (2) : [Fe]-0.05 (3) — 0.0353 Using this regression line a contour plot, shown in Figure 3-3, for the response of the system to different hydrogen peroxide and goethite concentrations was made. The iso- response lines show the decimal value of SA removal and represent the boundaries at which a combination of hydrogen peroxide and goethite will achieve a given removal. 64 The data for pH 3 are recorded in a similar manner, as shown in Table 3-2. The regression obtained from this data is shown in Equation 4, and contour plot generated fi'om the equation can be seen as Figure 3-4. The regression equation for pH 3 data was found to be B(H, F) = 0.57 - 0.105H + 0.06F - 0.067112 — 0.086F2 + 0.051HF (4) H and F are determined using equations 2 and 3. This regression analysis can be better expressed as a function of surface area of goethite. Since the reactive properties of a mineral are dependent of the number of features on its surface”, this implies that the total surface area of the mineral is more important than the mass of mineral present. In order to do this we can substitute F in equation 1 for F,. F Surface Area—6.65 s' —467 (5) 65 q GEvofifig x 32: \ EM om 66 .EwmoB 3 :9: $3 mm 0:50.00 loco fl x “Bob doom—:03 couobaoonoo 6802.80 85800 as 6m as xsmsk 3&5... £9: as Q: owed Nomod «mood Smod mvd 9. com—ow; £69 dos o.§ omod Emod ooood Smod mvd NP vo.m r 5 $9. do.» mos hood dovod good Fomod med 3 comic; $9. was ode wood Nomod moood Fomod mvd or mommvd £62. dos ooh Bod Fomod Sood Pomod med o vomové .x. K vod dds omod Noood voood woood mvd o moflomd {so Rd? ton «mug. Foord onto oooFd med 5 oo+mood *6 do.» n. 3. «mod oomod doood Smod o _.ood o vomoo... $N~ 3N mdo ovod momod «mood Smod mod m momvwd exam odd moo Bo; ode oomod o3od m...o v voflvoé 33 5d eds 354 5. Ed Edd ow Pod mud n mommoé $m omdr vdo men. 2. smood mmmpd vmood 9d N womb; $9.0. med? odo mom. 3 mmood mEFd vmood mud F E: 30.59... 75.3 9020» "Edna FE. «0.0.2 .30... 35:51 8 mm 00.5000 QB cos 0% .x. x 300.00% 0.8.: 0000—00—00 000000000000 00000 00 000.0000 00.9. P .0 $00 00.0? mdw Nmod oomod owwod Fomod 00.0 .m— oomovd $8 8.9 m. 3 «N00 oomod mwwod Pomod med 9 mom—00.0 00000 0:00 0:30:00 .00 0000000 05 000: 0000—00—00 00000000 202000 308003: .00 00000000 00 m: 0538 :00: 0.2. @0098 3: 00000008 03008 05 .00: 0:00 0:30:00 .00 0300000 0000000 05 000 :000000 05 0000 00:08 0000 030:3 .00 0000000 05 “03.000 :000 0: 0000.: 0:00 0:30:0m .00 0000000 05 :03 03003 :000 0: 0000 000,080 :89 0:0 00000000000 000000300 0: 000 :05 :000 00.: 00:30? 00:000w .00 3000000 05 000 032m .m mm 00 00000000 0003000 000.00% M0000 00:28 .00: A4<- ‘Goe 0.089gmH202 0.43M +Blank Goe0.15gm Figure 3-6: Plot of SA concentration over time for different hydrogen peroxide and Goethite concentrations for pH 7 reactions. Initial SA concentration of 100 ppm Byproducts Byproducts from the reaction of hydroxyl radicals with salicylic acid have been found in the samples during and after the 72 hr time frames. During experiments where SA concentration was measured over time, the formation of these byproducts was observed at pH 3 and 7, with the use of an HPLC/UV detector as described earlier. Samples were analyzed by LC/MS and GC/MS to determine the products resulting fi'om the reaction of hydroxyl radical and SA. The formation of these byproducts can be seen for both pH regimes in Figure 3-7 73 pH3 pH7 ..' r '7 , Salicylic Acid ... Salicylic Acid --4 5min Unknown _.1 Unknown DHBA ‘5 .. . A III-1 n.- .l I! n: r} M 3,. _ _ _/\\ k g __ 7 v i I J 'p g‘lw‘iwlwi'i ‘ 3'.r.;7.'.1.'1.-"".L"" L"l'i"l'f‘l 1'1 I'lr-L‘I'Ejk'l' 1 -- - ___,.__ Unknown H 7 -: pH 3 ‘ f. p -—1 30min Salicylic Acid 5. 30min l' S 1' l‘ A d 1 .._ i aicy 1C cu n. . . THBA "j Unknown DHBA "1 1' in: N N J ._: '5‘ l {‘i “1 ‘ A \ "._________,.' Lad; ‘~. .1 .... L ""1"i‘l"i‘i'i i‘1'-c'7»?t -\ 1r 1‘1 1'1 l'i‘1'1 1'7'3’1'3.’er" . Unknow In- ‘ .pfl|'il-. Unknown 7’ H 7 .4 pH 3 —: . P --1 4 hr Salicylic Acid "f 4 hr li _ . . __‘ / n. _ Salicylic Acnd 4 A ‘ ’ DH 0" I J ”J l . J ~~ - - 4' - ,_ J —— ‘ ——-~ —————~‘ —— l i i l l I i i l 1 ' 1 T 7 3 ‘n l -1.._..., . H 3 _j ......2_4 hr pH 7 '1 24 hr . p ,_1 2 . . , .1 S l' I' A 'd ‘ Unknown Salicylic Acid "' Unknown. alcylc c. "T I .4 . l "v I 7 v ' [.1 I.- I ..3 .r- -____ — ‘ ‘——‘ — r r s l i x .1. t ‘— 1 l 1'7 i T A ~- PH 3 ~ “"7‘2hr pH 7 4 72 hr ”.1 . . . .. 1 Salicylic Acud 1 _ Unknown ”n Unknown 1 . . ".1 Salicylic Acid f .g 1 ... A i ~ q__—_,_J er . _ l i’w F I o l I 6 S " l l 1 l i 8 ‘5 Figure 3-7: Oxidation of salicylic acid 100 ppm over time for pH 3 (left column) and pH 7 (right column) reactions with 0.0263 g/L goethite and an initial hydrogen peroxide concentration of 0.13M, analyzed through HPLC/UV at 234 nm. Shown in the figure are peaks formed during the 72 hours of reaction time. Dihydroxy benzoic acid (DHBA) and trihydroxy benzoic acid (THBA) peaks have been identified. 74 Results from LC/MS data are shown in Figures 3-9 thru 3-13. Samples from LC/MS analysis had prominent peaks in the negative ion mode and show that dihydroxy (m/z 153) and trihydroxy (m/z 169) benzoic acids are formed and were present in both pH 3 and 7 experiments. Peaks 99-, 109-, represent the loss of C00' and 'OH from the molecular ion (M-l). GC/MS GC/MS analysis of the samples revealed different byproducts than those observed from the LC/MS results. The products found were more oxidized versions of SA and its known products. The structures of these byproducts can be seen in Figure 3-15. 75 Discussion Sorption of SA to Goethite Mineral catalyzed systems, like all heterogeneous catalyzed systems, require that both the oxidant and the contaminant are in close proximity to active sites on the mineral surface“. Because of this, before any reaction can occur, the compound must sorb to the reactive surface. In both systems SA was found to sorb to the mineral, although it sorbed to a greater extent at pH 3 (38%) than at pH 7 (16%). We may be able to explain this result by the findings from Evanko and Dzombak66, who found that alterations to the functional group of carboxylic acids influenced their sorption to goethite. He et. al67 found that SA complexes with goethite. A change in pH will influence both the SA’s functional group and the charge on the mineral surface. The pKa for SA is 2.97, which is close to the pH of 3 used for reactions. As such, about half of the SA found in solution would be present in its protonated form and the other half would be unprotonated. On the contrary, at pH 7, about 99.99% of the SA would be unprotonated as the salicylate ion, which may not sorb as readily to the mineral surface. This conclusion contradicts the findings by Kamik et a168, who found that there was no significant sorption of SA to a hematite coated membrane at pH range of 2.5-3, but a high sorption of SA at pH of 7. A pH 3 solution could also affect the mineral surface by dissolving some of the iron and obstructing a portion of the reaction sites with dust; precipitating iron phosphate that sometimes formed during experimentation; or inversely by activating sites that were not 76 available to react. Because the disappearance of SA was greater at pH 3 than pH 7, site activation appears more likely. Salicylic Acid Degradation by Goethite Catalyzed CHPs The removal efficiency of SA by goethite catalyzed CHPs increased as the initial SA concentration was increased, at both pH 3 and 7. Figure 3-8 shows a comparison of SA removal for initial SA concentrations of 10 and 100 ppm, at pH 3 and 7, at the same initial hydrogen peroxide and goethite concentration. Since SA removal depends on the diffusion of the contaminant to the mineral surface, an increase in concentration would allow more SA to be in proximity to the mineral, ergo an increase in removal is expected. Another explanation might be that SA degradation occurs according to a pseudo first order reaction on the mineral surface. The mineral surface should not change appreciably over the course of the experiment, leading to a pseudo-first order hypothesis. 77 ..... n! nil. h ' z .93 I L5: 70.0 60.0—»~--~~w ,_ ,, ,, , , . rrrrrrrrrr - rrrrrrrrrrr l-H 50.0 _, ------------------------ n , , , rrrrrrrrr 4s .0 O % Removal m .0 0 1—0-1‘ 10.0 - 0.0 H202 0.43 M Goe 4.46 g/L llpna lOppm IpH7 10 ppm DpH3 100ppm UpH7 100 ppml Figure 3-8: Comparison of SA removal in pH 3 and pH 7 samples at different SA concentrations. The results from Table 3-2 show that at pH 3, CHP resulted in similar removal efficiencies for SA after 72 hours for all goethite and hydrogen peroxide combinations used. On the contrary, at pH 7, the efficacy of SA removal varied for the different goethite and hydrogen peroxide combinations used. This may be due to the saturation of the goethite surface by the oxidant and thus the active sites on the mineral produced hydroxyl radicals to its fullest extent. The linear regression fits for oxygen production in goethite catalyzed systems at pH 7, found in Chapter 2, appear to corroborate this conclusion; where oxygen production appeared to reach a constant rate after 1 hour of reaction time for all hydrogen peroxide and goethite concentrations tested. Another explanation is that the byproducts produced are consuming hydroxyl radicals. The true 78 amount of 'OH formed from these experiments cannot be measured unless the concentration of one or more byproducts is measured as well; in part because competing reactions would occur as byproducts are formed, consuming 'OH before they interact with SA. The data obtained from SA removal experiments was used to generate a response surface for SA degradation for a wide range of hydrogen peroxide and goethite concentrations. This response surface can be a useful tool for estimating the desired range of oxidant to catalyzer ratio necessary to remove a contaminant. This data becomes especially useful for field applications where a range of concentrations is more applicable than a lone single value for the required ratio of oxidant to catalyzer. From Figure 3-8 the maximum SA removal achieved at pH 7 was 30% ([SA]() =100 ppm) for a 0.43 M hydrogen peroxide concentration. On the other hand, at pH 3, 60% removal of SA was achieved at initial concentrations of 100 ppm and 0.43 M for SA and H202, respectively. The maximum SA removal at pH 7, 30%, was achieved with concentrations of 0.77-0.85 M hydrogen peroxide and goethite concentrations between 0.021 and 1.78 g/L (2.27 x 10'4 M and 0.02 M as Fe, respectively). In contrast, at pH 3, a hydrogen peroxide concentration between 0.51 and 0.77 M with a goethite concentration between 2.48 to 4.9 g/L (0.035 and 0.055 M as Fe) was required to obtain maximum percent removal (60%). Reactions at pH 3 achieve SA removal that is 200% higher than reactions at pH 7 (60% removal vs. 30% removal, respectively) under the same experimental conditions and range of concentrations. To achieve this maximum removal, pH 3 reactions were carried 79 out at the same hydrogen peroxide concentrations as pH 7 reactions; but reactions at pH 7 require higher goethite concentrations. The use of surface area as a surrogate for active reaction sites on the surface of a mineral can also be used to estimate hydroxyl radical production. According to Kwan and Voelker 38, the amount of hydroxyl radicals produced by goethite over a period of time can be estimated by the surface area of the mineral and the hydrogen peroxide concentration: VOH = 10‘7 [Surface Area] [H202] (6) VOH = hydroxyl radical production over time. From this equation we can see that hydrogen peroxide concentration will be the limiting factor in determining 'OH production. Since hydrogen peroxide decomposes on the surface of goethite and produces oxygen, the rate of decomposition needs to be known in order to estimate 'OH production Because Kwan and Voelker used concentrations of hydrogen peroxide of 5 mM, they may have been able to ignore the decomposition of hydrogen peroxide to oxygen at the mineral surface; at least there is no mention of this reaction in their paper, and it’s difficult to know if the hydrogen peroxide concentration in equation 6 takes this reaction into account. According to the authors, at a constant goethite concentration, the steady state hydroxyl radical concentration does not depend on hydrogen peroxide concentration, allowing them to use a fixed hydrogen peroxide value. It is our belief that hydroxyl radical production will vary with time because the hydrogen peroxide concentration varies with time. 80 By using the data presented in Table 3-2 and 3-3, showing the number of moles of hydroxyl radicals produced over the 72 hour time frame, and assuming the OH radical production was constant throughout that time we can determine that the rate of OH radical produced in our system at pH 3 is approximately 1 x 10'9 M 3'], while at pH 7 the production rate was found to be 4 x 10'10 M 3']. Watts et al45 found that the rate of OH radical production for goethite in soil Slurries at pH 7 was 2.94 x 10'8 M s", which is 2 orders of magnitude greater than our rate. If we take into account the hydroxyl radicals that are consumed when forming byproducts, and we add this mass to the calculated production rates from our study, we may find that the hydroxyl radical production rate in our experiments is similar or higher than the rates found by other authors. Figure 3-5 and 3-6 show the removal of SA over the 72 hours for several hydrogen peroxide and goethite concentrations. It can be observed in both figures that SA removal increases significantly for most reactions after 24 hours in both pH 3 and pH 7 systems. Since the majority of the oxygen production occurs in the first 24 hours (Chapter 2) we may conclude that hydrogen peroxide decomposition is affecting hydroxyl radical production. The figures also show that reactions containing the same hydrogen peroxide concentration but different goethite concentrations have similar rates of SA removal at pH 7 but reactions at pH 3 depend on both hydrogen peroxide and goethite concentrations. In Chapter 2 we found that oxygen is produced at a rate of 2.70 x 10'8 i 3.75 x 10'9 M s'l . Using this decomposition rate and the data for hydroxyl radical production presented 81 elsewhere in this chapter we may be able to hypothesize the mechanism by which hydrogen peroxide reacts at the mineral surface. Byproducts Previous studies have shown that the byproducts formed from the reactions of hydroxyl radicals with SA include 2,3 hydroxybenzoic acid (2,3-DHBA) and 2,5 dihydroxybenzoic acid (2,5-DHBA) and catechol“. During the oxidation of SA with CHP, several byproducts were formed. Using both LC and GC/MS analysis, some of these byproducts were identified. LC/MS analysis showed the presence of DHBA as well as a trihydroxy benzoic acid (THBA) after 72 hours of reaction. The spectra for these byproducts are seen in Figures 7-11. Although LC/MS spectra did not provide extensive information about the THBA found, a possible structure has been developed and is shown in Figure 3-14. 82 Figure 3-9: Typical chromatographic spectra from LC/MS analysis in negative ion mode for pH 3 reactions, showing the total ion count for negative ion data acquisitions. And the peaks for m/z 153- found in the samples. 83 828.8 8.8 8.8 8.3 8.8 8.8 8.8 8.8. 8.8 8.8 8.9 8.20, L. .- 21w! , e . . . 2 . , . 4,.-121 .. -. - --2..-ll. .- - !2-7? . , . e .1. i - u - . 1 ll,» ..h.\\i~,}/./ i\.\1 -\ .- 1 [Fix-- . ll. - llll. .- l. I it I .l - l .. i l -.- . .. 11 .J- .1i./. 2 awn-8 82:. «N: ; 2.H 7 2 2 2 2 2 2 2 2 2 a % 22 a ._2 28; 2r 2 22 .2 85a ._ o: wmr. 69 -mm 228m ”m 2.8 3880 8.28 . 2 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 89 89 86 :?.€... 0.52.4 2112:4321... 4.1.1.2 1.132..-.3:23:21 . 2a.... .21. i2 2.222.111: .1 . :2.222.42...2J22,2_.§2i. j:r§$1¥...j%222.2i21.22.2832 ... .2 .. . 2 . . . 22, 2.2 2.32.2.2“. 2.22. .0 2.22222 . {.3 8.8 3.; . 2 8.8 22- 22. 222 _ 2222222222222 2222, 2.22. 2 222,22 _ 2. $383 39. . . . 2 2 a: . 3 mm a 1 . 2.2.22. 2 2 «on 8.? am 2 _ 2 8m :2, 2 2 ~32. __ 2 2 2 2 2 _ 2 2 2 2 2 2 . 2 2 as. .__ 2_ 2 «:22 2 2 2 r 2 2 2 2 2 22 _ 2.2 88 222 . 888 8.0 .2. - .8. .$ 25.-6 ”m v8 now-«.80 01a. 20¢ o=>o=um 84 Figure 3-10: Mass spectra obtained from pH 3 samples showing m/z 153' and 169', which are believed to be dihydroxy and trihydroxy benzoic acids. 85 com 0mm ovN 0mm CNN 9N CON oar 8.. at 02 on 2. oer on— our 0:. Cor om on on 00 N\E pub—pn.>P>>k>h>>>>>kk>>b¢bo.»>2>p~u>p»HprrLbhbbbPPhphp0>>b>>rPLir>I>>».n.p—>>>»»>>h>Tn>r~>>>>1~abbpbhpfl>->»>>»>~>Pkrphw. >JPP>L>—rbbrirtbkhb>>>bprhpwpuu»—>2>>>—rn>-2—asflbbppp»-Phi-L>1—>F>»b»_>»wfi~_>>Wh.>>>>w>>hb4>2,>>>P«HP»>>H QOMN 2 2 . 2 . 2 3:2 2 no: 2 $5 . 2 2 . \ . \ m8 _ 2 03.: wOww \ _ \2 _ 9: 2:8. . 2 . 2 th 2 w 2 2 2 2 _/ QNOF 32 . 1% V Qmmw Y 8QO @229 i 59 -mm 2.8 ”N 2280284832824 528: so 2832 n: ".8 8880 NSNLGEIQN MIR 22 0.30:6 86 Figure 3-11: Chromatographic peak of m/z 169' shown on the upper part of the figure believed to be trihydroxy benzoic acid (THBA), compared to the total ion count (TIC), lower part of the figure, for pH 3 samples by LC/MS analysis. The intensity of the THBA peak is 2 orders of magnitude less than the total intensity of ions detected signifying that it may not be the most prominent by product formed. 87 8.8 8.3 88 OEFJ — - m _ 1P 1— 23:3 ”._h .3 .56 N 8.8 8.8 8.8 p h FLIP Lk _ — «381% we. 3 8.3 [8.822 mom? of 1mm :50” ”N sceniaiunu “22232808; 22.3 .23“ m3 - % 2 2 / mo; 2 2 2 _ 8, 52 "62.2 “9.52.230 8.0 2 2 agéfi mp. m K m /_.¢ n I %I r... vac-5888 «in .23 2.26.28 88 Figure 3-12: LC/MS spectra for peaks eluding at a retention time of 5.17minutes ( from LC/MS column) for pH 7 samples. In the spectra m/z 153' and 169' are visible; these peaks are believed to be dihydroxy benzoic acid (DHBA) and tn'hydroxy benzoic acid (THBA). 89 8 8 8 8m 8m 8m 8m 88 SN 08 8. 8. 8. 8. 8. o... 8. 8. o.. 8. 8 8 N\E :...Z:..:._ZI_:ZT»...£IT.IFIL4. TIL.ZL .IFI..ZIF.I.liffi...m.rL.tt.$i.Ttripwiif2.“... TILKITI»? t..:.::..ILiifiLLI .FIFI»; iliffvtfixftrrt» .IIJJ. é _ r a; Sr m3. 8.8 28. Q3. 3.. . / 3o. . ,% o8. r 88... m8. u .8. .mm ._Sm “a E§888t8$885 so :89 8. So 8880 seeming“ 5.. 22 9.32.8 90 Figure 3-13: Chromatographic peak of byproduct having m/z 153' shown here to compare its abundance versus total ion counts, in positive and negative ion mode for pH 7 sample during. LC/MS analysis. It is possible to see from the first scan that three isomers may exist. 91 8.8 8.8 8.8 8 9. 8 9. 8.8 8 8 8.3 8.8 8.9 8 o. 86 OEPFP»..._..L>~>»»— ..>|£F>»..— r» > p _ »..~> — L». pulpL» p p... prr brbb by .-h b > - rLl. . .P»¥L— cos now. an “and 3...” D? OWB r. . ma. . T 8. 3.8 8.: 8.3. . .V-.VF r 8.... \. . So. 89... one. 2.... +mw 50m H. m8 3880 8.8 8 on 8.8 8 n». 8.9. 8.8 8.8 8 mm 8 on 8.9 86. 8 n Frk. _»»rrrpp>>pbyhrhu.fi_ pp f. L>>»[|[—»>.>p>uL+L .._.. _ _>»>.FL».._ p >—p> >~>t>>~ .b»... h pt». > _> r .lillilllllll/ llllnill llzllal still - till. all--- ; lllfh 2r), fill.-- 1- ll! 1.;) xii/N! ./l\\.}//....\ H. N z/ _ 8 .m cm 3 ., i M. _ .. _ ._ _ i ., _.. _W .__8 . l 0/0 v . __ ;. SN _ ._ worm; _ o: 3.... 1 8. -mm cmom N moo 8880 8.8 8.8 8.8 8.9. 8.9. 8.8 8 on 8.8 8.8 8 m. 8 o. 86 —W»¢\U»_...bpp...w..4.1._.»._h0>..4—‘P .11prpul.{_fi,».l_ .h...»1> p?» _4. 1i_].|.lc.lu_.h>» _ r1.“4.”>1l»_‘h-.._.>>»W¢.A.WU:—d..1Lth....-—>|.l»y_phylr o 2.... . I %I $08.. mm. 86 I .8. .mm snow “N 08 3880 scouiaiu 5.. 23 232.8 92 rpm COOH OH Figure 3-14: Possible structure for the byproduct m/z 169' found in the LC/MS spectra in' Figures 8 and Figure 10. The presence of these byproducts indicate that some of the hydroxyl radicals produced did not selectively interact with SA in solution, instead reacting with any possible species found in solution, including other byproducts. As a result, the use of SA removal efficiencies to determine OH radical production may result in an under-prediction of the actual amount of radicals produced. GC/MS analysis revealed byproducts that had a higher degree of oxidation than the byproducts found by LC/MS. Figure 3-15 shows the structure of the byproducts found from GC/MS. Some of the byproducts found are more volatile and have lower molecular weights than the parent compound, which explains why they were found by GC and not LC analysis. The formation of these byproducts show that the actual yield of hydroxyl radical production may be significantly higher than that estimated from SA removal alone. The formation of hydroxyl and dihydroxy acetic acid requires the reaction of SA with multiple radical species to obtain that degree of oxidation. 93 lie 60 fl? 3wa O HO OH OH OH OH O O Maleic Acid Malic Acid :9, HO :0, HO OH HO OH 2-hydroxy Acetic Acid 2, 2 di-hydroxy acetic acid Figure 3-15: GC/MS identified byproducts found from the oxidation of salicylic acid by Goethite catalyzed CHPs for both pH 3 and pH 7. In Figure 3-7 one can see the presence of a large peak labeled “unknown” throughout the experimentation. Although multiple attempts were made to determine the structure of this compound, the exact nature of the byproduct could not be determined. A UV spectral purity analysis of the peak was performed using a diode array detector set to scan between 200 and 300 nm. The analysis showed that the peak may be mixture of byproducts as its initial (at the time it starts to show in the chromatogram), central (apex of the chromatogram) and spectra from the peak tail, do not match. Although we were not able to separate the peaks with our system, we may have identified the structure of the byproducts that form the peak inadvertently with the GC/ MS analysis. 94 Cor: l1 Conclusions At the high iron concentrations employed in this study, it was not possible to quantify the hydroxyl radical concentration for soluble iron experiments. Unfortunately we were not able to identify a hydroxyl radical probe that did not complex with the soluble iron or was able to be maintained in solution at the pH required. Hydroxyl radicals can be measured with SA as a radical probe in mineral iron catalyzed reactions. The mass of hydroxyl radicals produced during the reactions was found to be at least 2 x10"6 moles for reactions at pH 3. At pH 7,.the mass of hydroxyl radicals produced was found to be between 1 x 10'11 to 2 x10'6 moles after 72 hours. Reactions at pH 7 were less efficient at removing SA than reactions carried out at pH 3. It is our belief that the protonation of SA at lower pH may be the key to this difference. Byproducts are formed during CHP reactions at both pH 3 and pH 7 Six different byproducts (Dihydroxy benzoic acid, trihydroxybenzoic acid, malic acid, maleic acid, 2 hydroxy acetic acid, and 2,2 dihydroxy] acetic acid) were identified using LC and GC/MS analysis, some of which are more oxidized forms than those found in the literature. One of the byproducts found 2,3 DHBA is a known byproduct of the reaction of hydroxyl radicals with salicylic acid. In order to determine the actual amount of hydroxyl radicals produced we need to quantify all the byproducts formed and determine the concentration of at least one them. 95 In mineral catalyzed reactions the oxygen production rate appears out-compete hydroxyl radical production rate during the first 24 hours of reaction. By comparing the hydroxyl radical time dependent reactions from this chapter to the oxygen production experiments from Chapter 2, we found that the major oxygen production surge occurs in the first 24 hours of reaction time, while hydroxyl radical production increases notably after 24 hours of reaction, 96 References 1. Kostecki, P. T.; Calabrese, E. J ., Petroleum Contaminated Soils: Remediation Techniques, Environmental Fate, and Risk Assesment. 3rd ed.; Lewis Publishers, Inc.: Chelsea, MI, 1989; Vol. 1. 2. Juhasz, A. L.; Naidu, R., Bioremediation of High Molecular Weight Polycyclic Aromatic Hydrocarbons: A Review of Microbial Degradation of Benzo[a]pyrene. International Biodeterioration & Biodegradation 2000, 45, 57-88. 3. Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M., Environmental Organic Chemistry. lst ed.; Wilkey-Interscience, John Wiley & Sons Inc: New York, 1993. 4. Chiou, C. T.; McGroddy, S. E.; Kile, D. E., Partition Characteristics of Polycyclic Aromatic Hydrocarbons on Soils and Sediments. Environmental Science and Technology 1998, 32, (2), 264-269. 5. Keith, L. H.; Telliard, W. A., Priority Pollutants: Prospective View. Environmental Science and Technology 1979, 13, (4), 416-424. 6. Srivastava, V. J .; Kelley, R. L.; Paterek, J. R.; Hayes, T. D.; Nelson, G. L.; Golchin, J ., A Field Scale Demonstration of Novel Bioremediation Process for MGP Sites. Applied Biochemistry and Biotechnology 1994, 45, (6), 741-756. 7. Heitkamp, M. A.; Freeman, J. P.; Cemeglia, C. E., Naphthalene Biodegradation in Environmental Microcosms: Estimates of Degradation Rates and Characterization of Metabolites. Applied and Environmental Microbiology 1987, 53, (1), 129-136. 8. Nam, K.; Rodriguez, W.; Kukor, J. J ., Enhanced Degradation of Polycyclic Aromatic Hydrocarbons by Biodegradation Combined with a Modified Fenton Reaction. Chemosphere 2001, 45, (2001), 11-20. 9. Sims, R. C.; Overcash, M. R., Fate of Polycyclic Aromatic Compounds (PNAs) in Soil-Plant Systems. Residue Reviews 1983, 88, 1-68. 97 10. Luster-Teasley, S. L.; Yao, J. J .; Hemer, H. H.; Trosko, J. E.; Masten, S. J ., Ozonation of Chrysene: Evaluation of By-Product Mixtures and Identification of Toxic Constituent. Environmental Science and Technology 2002, 36, (5), 869-876. 11. Upham, B. L.; Masten, S. J .; Lockwood, B. R.; Trosko, J. E., Nongenotoxic Effects of Polycyclic Aromatic Hydrocarbons and Their Ozonation By—Products on Intercellular Communication of Rat Liver epithelial Cells. Fundamental and Applied Toxicology 1994, 23, 470-475. 12. Yin, Y.; Allen, H. E., In Situ Chemical Treatment. July 1999, 1999. 13. Group, I. T. a. R. C. W. Technology Overview Dense Non-Aqueous Phase Liquids (DNPL): Review of Emerging Characterization and Remediation Technologies. http://www.itrcweborg/DNAPL-l .pdf 14. Carve], D. D.; Cartwright, R. T. In Measuring in situ Fenton’s application success, Proceedings of the Third International Conference on Remediation of Chlorinated and Recalcitrant Compounds, Monterrey, CA, United States of America, May 20-23 2002, 2002; Press, 3., Ed. Battelle Press: Monterrey, CA, United States of America, 2002; pp 1173-1178. 15. Gates-Anderson, D. D.; Siegrest, R. L.; Cline, S. R., Comparison of Potassium Permanganate and Hydrogen Peroxide as Chemical Oxidants for Organically Contaminated Soils. Journal of Environmental Engineering 2001, 337—347. 16. Beltran, F. J .; Ovejeor, G.; Rivas, J ., Oxidation of Polycyclic Aromatic Hydrocarbons in Water.4. Ozone Combined with Hydrogen Peroxide. Industrial Engineering and Chemistry Research 1996, 35, (3), 891-898. 17. Luster-Teasley, S. The Use of Gaseous Ozone to remediate Pyrene Contaminated Soils: A study of By—Product Production, Environmental effects on Remediation Efforts, and Scale-Up Volume I & II. Dissertation for the Degree of Ph. D., Michigan State University, East Lansing, MI, 2003. 18. Yao, J .-J .; Huang, Z.-H.; Masten, S. J ., Ozonation of Benz[a]anthracene: Pathway and Product Identification. Water Research 1998, 32, (11), 3235-3244. 19. Yao, J .-J .; Huang, Z.-H.; Masten, S. J ., The Ozonation of Pyrene: Pathway and Product Identification. Water Research 1998, 32, (10), 3001-3012. 98 20. Benitez, F. J .; Beltran-Heredia, J .; Acero, J. L.; Rubio, F. J ., Chemical Decomposition of 2,4,6-Th1ichlorophenol by Ozone, Fenton's Reagent and UV Radiation. Industrial Engineering and Chemistry Research 1999, 38, (1999), 1341-1349. 21. Council, I. T. R. Technical and Regulatory Guidelines for In-Situ Chemical Oxidation of Contaminated Soil and Groundwater 2nd ed. ISCO-2; Interstate Technology & Regulatory Council: Washington, DC, 2005 . 22. Haber, F.; Weiss, J ., Catalytic Decomposition of Hydrogen Peroxide by Iron Salts. Proceedings of T he Royal Society, London 1934, 147, 332-351. 23. Jones, C. W., Applications of Hydrogen Peroxide. The Royal Society of Chemistry: Cambridge, 1999. 24. Walling, C., Fenton's Reagent Revisited. Accounts of Chemical Research 1974, 8, 125-13 1 . 25. Watts, R. J .; Teel, A. L., Chemistry of Modified F enton's Reagent (Catalyzed H202 Propagations-CHP) for In Situ Soil and Groundwater Remediation. Journal of Environmental Engineering 2005, 131, (4), 612-622. 26. Watts, R. J .; Teel, A. L., Treatment of Contaminated Soils and Groundwater Using ISCO. Practice Periodical of Hazardous, Toxic, and Radioactive Waste Management 2006, 10, (1), 2-9. 27. Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. Critical Review Constants for Reactions of Hydrated Electrons Hydrogen Atoms and Hydroxyl Radicals (oOH/O-) in Aqueous Solution; 1988. 28. Walling, C.; Johnson, R. A., Fenton's Reagent V. Hydroxylation and Side-Chain Cleavage of Aromatics. Journal of the American Chemical Society 1974, 97, (2), 363- 367. 29. Augusti, R.; Dias, A. O.; Rocha, L. L.; Lago, R. M., Kinetics and Mechanism of Benzene Derivative Degradation with F enton's Reagent in Aqueous Medium Studied by MIMS. Journal of Physical Chemistry 1998, 102, (52), 10723-10727. 30. Casero, 1.; Sicilia, D.; Rubio, S.; Perez-Bendito, D., Chemical Degradation of Aromatic Amines By Fenton's Reagent. Water Research 1997, 31, (8), 1985—1995. 99 31. Walling, C.; Amamath, K., Oxidation of Mandelic Acid by Fenton's Reagent. Journal of the American Chemical Society 1982, 104, (5), 1 185-1 189. 32. Walling, C.; El-Taliawi, Fenton's Reagent III. Addidtion of Hydroxyl Radicals to Acetylenes and Redox Reactions of Vynil Radicals. Journal of the American Chemical Society 1973, 95, (3), 848-850. 33. Teel, A. L.; Warberg, C. R.; Atkinson, D. A.; Watts, R. J ., Comparison of Mineral and Soluble Iron Fenton's Catalysts for the Treatment of Trichloroethylene. Water Research 2001, 35, (4), 977-984. 34. Kong, S.-H.; Watts, R. J .; Choi, J .-H., Treatment of Petroleum-Contaminated Soils Using Iron Mineral Catalyzed Hydrogen Peroxide. Chemosphere 1998, 37, (8), 1473-1482. 35. Watts, R. J .; Stanton, P. C.; Howsawkeng, J .; Teel, A. L., Mineralization of a Sorbed Polycyclic Aromatic Hydrocarbon in Two Soils using Catalyzed Hydrogen Peroxide. Water Research 2002, 36, 4283-4292. 36. Kakarla, P. K.; Andrews, T.; Greenberg, R. S.; Zervas, D. S., Modified F enton's Processes For Effective In-Situ Chemical Oxidation-Laboratory and Field Evaluation. Remediation 2002, Autumn, 23-36. 37. Pignatello, J. J ., Dark and Photoassisted Fe3+ Catalyzed Degradation of Chlorophenoxy Herbicides by Hydrogen Peroxide. Environmental Science and Technology 1992, 26, (5), 944-951. 38. Kwan, W. P.; Voelker, B. M., Rates of Hydroxyl Radical Generation and Organic Compound Oxidation in Mineral-Catalyzed Fenton-like Systems. Environmental Science and Technology 2003, 37, (6), 1 150-1158. 39. Tyre, B. W.; Watts, R. J .; Miller, G. C., Treatment of Four Biorefractory Contaminants in Soils Using Catalyzed Hydrogen Peroxide. Journal of Environmental Quality 1991, 20, (October-December), 832-838. 40. Pignatello, J. J .; Baehr, K., Ferric Complexes as Catalysts for "Fenton" Degradation of 2,4-D and Metolachlor in Soil. Journal of Environmental Quality 1994, 23, (March-April), 365-370. 100 41. Yamazaki, I.; Piette, L., EPR Spin Trapping Study on the Oxidizing Species Formed in the Reaction of The Ferrous Ion with Hydrogen Peroxide. Journal of the American Chemical Society 1991, 113, (20), 7588-7593. 42. Chen, C. T.; Tafuri, A. N.; Rahrnan, M.; Foerst, M. B., Chemical Oxidation Treatment of Petroleum Contaminated Soil Using Fenton's Reagent. Journal of Environmental Science and Health 1998, A33, (6), 987-1008. 43. Bowers, A. R.; Gaddipati, P.; Eckenfelder Jr., W. W.; Monsen, R. M., Treatment of Toxic or Refractory Wastewaters with Hydrogen Peroxide. Water Science and Technology 1989, 21, 477-486. 44. Morel, F. M.; Hering, J. G., Principles and Applications of Aquatic Chemistry. John Wiley & Sons, Inc.: New York, 1993; p 588. 45. Watts, R. J .; Foget, M. K.; Kong, S.-H.; Teel, A. L., Hydrogen Peroxide Decomposition in Model Subsurface Systems. Journal of Hazardous Materials 1999, 69, (2), 229-243. 46. Lin, S.-S.; Gurol, M. D., Heterogeneous Catalytic Oxidation of Organic Compounds by Hydrogen Peroxide. Water Science and Technology 1996, 34, (9), 57-64. 47. Lin, S.-S.; Gurol, M. D., Catalytic Decomposition of Hydrogen Peroxide on Iron Oxide: Kinetics, Mechanism, and Implications. Environmental Science and Technology 1998, 32, (10), 1417-1423. 48. Christian, G. D., Analytical Chemistry. 5th ed.; John Wiley & Sons, Inc.: New York, 1994; p 812. 49. Bader, H.; Sturzenneger, V.; Hoigne, J ., Photometric Method for the Determination of Low Concentrations of Hydrogen Peroxide by the Peroxidase Catalyzed Oxidation of N,N-Diethyl-p-Phenylenediamine (DPD). Water Research 1988, 22, (9), l 1 09-1 1 1 5. 50. Voelker, B. M.; Sulzberger, B., Effects of Fulvic acid on Fe(II) Oxidation by Hydrogen Peroxide. Environmental Science and Technology 1996, 30, (4), 1106-1114. 101 51. Lindsey, M. E.; Tarr, M. A., Quantitation of Hydroxyl radicals During Fenton Oxidation following a single Addition of Iron and Peroxide. Chemosphere 2000, 41 , (2000), 409-417. 52. Yang, X.-F.; Guo, X.-Q., Study of Nitroxide—Linked Naphthalene as a Fluorescence Probe for Hydroxyl Radicals. Analytica Chimica Acta 2001, 434, 169-177. 53. Steiner, M. G.; Babbs, C. F., Quantitation of the Hydroxyl Radical By Reaction with Dimethyl Sulfoxide. Archives of Biochemistry and Biophysics 1990, 278, (2), 478- 481. 54. Punchard, N. A.; Kelly, F. J ., Free Radicals: A Practical Approach. Oxford University Press: New York, 1996; p 310. 55. Yang, X.; Zhan, M.-j.; Kong, L.-r.; Wang, L.-s., Determination of Hydroxyl Radicals with Salicylic Acid in Aqueous Nitrate and Nitrite Solutions. Journal of Environmental Sciences 2004, 16, (4), 687-689. 56. Diez, L.; Livertoux, M.-H.; Stark, A.-A.; Wellman-Rousseau, M.; Leroy, P., High Performance Liquid Chromatographic Assay of Hydroxyl Free Radical using Salicylic Acid Hydroxylation During In Vitro Experiments Involving Thiols. Journal of Chromatography B 2001, 763, 185-193. 57. Bailey, S. M.; Fauconnet, A.-L.; Reinke, L. A., Comparison of Salicylate and D- phenylalanine for detection of Hydroxyl Radicals in Chemical and Biological Reactions. Redox Report 1997, 3, (1 ), 17-22. 58. Saran, M.; Summer, K. H., Assaying for Hydroxyl Radicals: Hydroxylated Terephthalate is Superior Marker than Hydroxylated Benzoate. Free Radical Research 1999, 31, 429-436. 59. Elovitz, M. 8.; Von Guten, U., Hydroxyl Radical/Ozone Ratios During Ozonation Process 1. Rct Concept. Ozone Science &Engineering 1999, 21, 239-260. 60. Duesterberg, C. K.; Waite, T. D., Kinetic Modeling of the Oxidation of p- Hydroxybenzoic Acid by Fenton's Reagent: Implications of the Role of Quinones in the Redox Cycling of Iron. Environ. Sci. T echnol. 2007, 41, (1 1), 4103-4110. 102 61. Duesterberg, C. K.; Cooper, W. J.; Waite, T. D., Fenton-Mediated Oxidation in the Presence and Absence of Oxygen. Environ. Sci. T echnol. 2005, 39, (13), 5052-5058. 62. Albarran, G.; Schuler, R. H., Concerted Effects in the Reaction of OH Radicals with Aromatics: Radiolytic Oxidation of Salicylic Acid. Radiation Physics and Chemistry 2003, 67, 279-285. 63. Amphlett, C. B.; Adams, G. E.; Michael, B. D., Advances in Chemistry Series 1968, 81, 231-250. 64. Diamond, W. J ., Practical Experimental Designs for Engineers and Scientists. Lifetime Learning Publications, Wadsworth Inc.: Belmont, 1989. 65. Steinfield, J ., Chemical Kinetics and Dynamics. 2nd ed.; Prentice Hall: Upper Saddle River, N.J., 1999. 66. Evanko, C. R.; Dzombak, D. A., Influence of Structural Features on Sorption of NOM-analogue Organic Acids to Goethite. Environmental Science Technology 1998, 32, 2846-2855. 67. He, J .; Ma, W.; Song, W.; Zhao, J .; Qian, X.; Zhang, 8.; Yu, J. C., Photoreaction of Aromatic Compounds at Goethite/H20 interface in the Presence of H202: Evidence for Organic-Goethite Surface Complex. Water Research 2005, 39, 119-128. 68. Karnik, B. S.; Davies, S. H.; Baumann, M. J .; Masten, S. J ., Use of Salicylic Acid as a Model Compound to Investigate Hydroxyl Radical Reaction in Ozonation- Membrane Filtration Hybrid Process. Journal of Environmental Engineering Science 2007, 6, (5). 69. Huling, S.; Pivetz, B. E. Engineering Issue Paper: In-Situ Chemical Oxidation; 600R06072; Environmental Protection Agency: Ada, Oklahoma, July 28 2006, 2006; p 60. 70. Belkin, S.; Steiber, M.; Tiehm, a.; Frimmel, F. H.; Abeliovich, A.; Werner, P.; Ulitzur, S., Toxicity and Genotoxicity Enhancement during Polycyclic Aromatic Hydrocarbon Degradation. Environmental Toxicology and Water Quality 1994, 9, (4), 303 -309. 103 71. Bogan, B. W.; Trbovic, V., Effect of Sequestration on PAH degradability with Fenton's Reagent: Roles of Total Organic Carbon, Humin and Soil Porosity. Journal of Hazardous Materials 2003, B100, 285-300. 72. Bogan, B. W.; Trbovic, V.; Paterek, R. J ., Inclusion of Vegetable oils in Fenton's Chemistry for Remediation of PAH-Contaminated Soils. Chemosphere 2003, 50, 15-21. 73. Kawahara, F. K.; Davila, B.; Al-Abed, S. R.; Vesper, S. J .; Ireland, J. C.; Rock, S., Polycyclic Aromatic Hydrocarbon (PAH) release from Soil during Treatment with Fenton's Reagent. Chemosphere 1995, 31, (9), 4131-4142. 74. Saxe, J. K.; Allen, H. E.; Nicol, G. R., Fenton Oxidation of Polycyclic Aromatic Hydrocarbons after Surfactant-Enhanced Soil Washing. Environmental Engineering Science 2000, 17, 233-244. 75. Flotron, V.; Delteil, C.; Berrnond, A.; Camel, V., Remediation of Matrices Contaminated by Polycyclic Aromatic Hydrocarbons: Use of Fenton's Reagent. Polycyclic Aromatic Compounds 2003, 23, 353-3 76. 76. Martens, D. A.; T, F. j. W., Enhanced Degradation of Polycyclic Aromatic Hydrocarbons in Soil Treated with Advance Oxidative Process- F enton's Reagent. Journal of Soil Contamination 1995, 4, (2), 1-14. ‘ 77. Huang, H.-H.; Lu, M.-C.; Chen, J .-N., Catalytic Decomposition of Hydrogen Peroxide and 2-Chlorophenol with Iron Oxides. Water Research 2001, 35, (9), 2291- 2299. 78. Sedlak, D. L.; Andren, A. W., Oxidation of Chlorobenzene with Fenton's Reagent. Environmental Science and Technology 1991, 25, (4), 777-782. 79. Beltran, F. J .; Gonzalez, M.; Rivas, F. J .; Alvarez, P., Fenton Reagent Advanced Oxidation of polycyclic Aromatic Hydrocarbons in Water. Water Air and Soil Pollution 1998, 105, (1998), 685-700. 80. Lindsey, M. E.; Tarr, M. A., Inhibited Hydroxyl radical Degradation of Aromatic Hydrocarbons in the presence of Dissolved Fulvic Acid. Water Research 2000, 34, (8), 23 85-23 89. 104 81. Lee, B.-D.; Nakai, S.; Hosomi, M., Application of Fenton Oxidation To Remediate Polycyclic Aromatic Hydrocarbons-Contaminated Soil. Journal of Chemical Engineering of Japan 2002, 35, (6), 582-586. 82. Walling, C., Intermediates in the Reactions of F enton Type Reagents. Accounts of Chemical Research 1998, 31, (4), 155-157. 83. Masten, S. J. In Use of In-Situ Ozonation For the Removal of VOCs and PAHs from Unsaturated Soils, Proceedings of the Symposium on Soil Venting, Houston, Texas, 1991; US. Environmental Protection Agency: Houston, Texas, 1991; pp 29-45. 84. Liang, Y. In Biodegradation of Pyrene in Soil Microcosms: Identification of a Toxic Intermediate, EWRI 2005: Impacts of Global Climate Change, Anchorage, Alaska, 2005; Environmental and Water Resources Institute: Anchorage, Alaska, 2005. 85. Chen, R.; Pignatello, J. J ., Role of Quinone Intermediates as Electron Shuttles in Fenton and Photoassited Fenton Oxidations of Aromatic Compounds. Environmental Science and Technology 1997, 31, (8), 2399-2406. 86. Goel, R. K.; Flora, J. R. V.; Ferry, J ., Mechanisms for Naphthalene Removal during Electrolytic Aeration. Water Research 2003, 37, 891-901. 87. Zeng, K.; Hwang, H.-M.; Fu, P. P.; Yu, H., Identification of l-Hydroxypyrene Photoproducts and Study of the Effect of Humic Substances on its Photolysis. Polycyclic Aromatic Compounds 2002, 22, 459-467. 88. Heitkamp, M. A.; Freeman, J. P.; Miller, D. W.; Cemeglia, C. E., Pyrene Degradation by Mycobacterium sp.:Identification of Ring Oxidation and Ring Fission Products. Applied and Environmental Microbiology 1988, 54, (10), 2556-2565. 89. Alexander, M., Biodegradation and Bioremediation. second ed.; Academic Press: New York, 1999. 90. Liang, Y.; Gardner, D. R.; Miller, C. D.; Chen, D.; Anderson, A.; Weimer, B. C.; Sims, R. C., Study of Biochemical Pathways and Enzymes Involved in Pyrene Degradation by Mycobacterium sp. Strain KMS. Applied and Environmental Microbiology 2006, 72, (12), 7821-7828. 105 91. Hemer, H. A.; Trosko, J. E.; Masten, S. J ., The Epigenetic Toxicity of Pyrene and Related Ozonation Byproducts Containing an Aldehyde Functional Group. Environmental Science and Technology 2001, 35, (17), 3576-3583. 92. Satoh, A. Y.; Trosko, J. E.; Masten, S. J ., Epigenc Toxicity of Hydroxylated Biphenyls and Hydroxylated Polychlorinated Biphenyls on Normal Rat Liver Epithelial Cells. Environmental Science and Technology 2003, 37, (12), 2727-2733. 93. Sedlak, D. L. The Abiotic Reactions of Polychlorintaed Biphenyls (PCBs). Doctoral Dissertation, University of Wisconsin, Madison, 1992. 94. Hauser, B.; Schrader, G.; Bahadir, M., Dependence of Genotoxicity of Benzo[a]pyrene Suspensions in MUtatox Test on Dissolved Concentration and S9 Addition. Ecotoxicology and Environmental Safety 1997, 38, (3), 224-226. 95. Lambert, M.; Kremer, S.; Anke, H., Antimicrobial, Phytotoxic, Nematicidal, Cytotoxic, and Mutagenic Activities of l-Hydroxypyrene, the Initial Metabolite in Pyrene Metabolism by the Basidiomycete Crinipellis Stipitaria. Bulletin of Environmental Contamination and Toxicology 1995, 55, (2), 251-257. 106 Chapter 4. Oxidation of Pyrene in Soil Pyrene Pyrene was chosen as the model PAH for this study because of the information available regarding byproduct formation from the remediation of the contaminant through multiple techniques. Pyrene is a symmetrical molecule as seen in Figure 4-1. A list of chemical and physical properties for pyrene is shown in Table 4-1. Pyrene Figure 4-1: Structure of model PAH, Pyrene showing double bonds and fused benzene rings. Table 4-1: Physicochemical properties of Pyrenel Property Value/Units Molecular Weight 202 gm/mol Melting Point 149° C Boiling Point 360° C Water Solubility 0.14 mg/L Vapor Pressure @ 20° C 9.16E-5 Pa Log Kow 5.32 Kom 104'5 kg om/L 107 Method and Materials Chemicals: All chemicals were ACS grade unless stated otherwise. Pyrene (98% purity) and ferrous ammonium sulfate were purchased from Sigma Chemical Company.(St. Louis, MO). HPLC grade acetonitrile was obtained from J .T. Baker Company (MG Scientific Inc. Pleasant Prairie, WI). Sodium dodecyl sulfate (lauryl) was obtained from Pierce Chemical Company (Perbio) (Rockford, 1].). Hydrogen Peroxide 30% was purchased from Fisher Company (Pittsburgh, PA). The initial concentrations of hydrogen peroxide and soluble iron used in this study were determined by comparison to the available literature on PAH oxidation using Fenton’s reagent 2'8. The experiments were set-up as soil slurry or suspension type systems. The reaction time for our experiment was also taken from comparisons between the literatures, where 24 hours was the conventional reaction time. An experiment involving 60,000 ppm of hydrogen peroxide was carried out for 72 hours and found that at 95% confidence level, the removal was not different from a 24 reaction at the same concentration. All authors 2'8 adjusted the pH before adding the hydrogen peroxide in order to create ideal conditions for F enton’s chemistry to occur. All reactions were run in borosilicate glass vials using a single addition of hydrogen peroxide to start the reaction. The vials were cleaned in a 1% hydrochloric acid wash, rinsed in ultrapure water and allowed to dry; to remove organics, the vials were placed in an oven at 550°C for 2 hours. 108 CHP reactions using soluble iron were allowed to continue for 24 hours and were quenched by adding sodium hydroxide 1M solution to precipitate iron. Reactions using goethite as the catalyzer were allowed to react for 72 hours and quenched by decanting the solution from the catalyzer. Organic matter soil was collected from the MSU campus during construction of the Spartan Child Development Center. The soil was sieved through a 40 mesh strainer to remove any debris and rocks. The soil was then analyzed for organic matter content, which was found to be 0.8%. For Ottawa sand and organic matter soil experiments a pyrene stock solution was added drop wise to vials in a multishaker containing Ottawa sand. The vials were then placed in a dark fume hood for 24 hours to allow the solvent to evaporate. The initial pyrene concentration was held constant at 300 mg/kg sand. Determining Pyrene Concentration Pyrene was extracted from water or soil by adding 10 ml of hexane and placing the vials in on a tumble shaker for 24 hours. The concentration of pyrene in the hexane was then measured. If dilution was necessary, aliquots were taken from hexane extract and added to a 2 mL auto sampler vial containing lmL of hexane. This dilution was repeated until an absorbance similar to the absorbance of a 10 mg/L calibration standard was achieved. A Perkin-Elmer 200 series RP-HPLC coupled with UV and fluorescence detectors was used to determine the pyrene concentration. A Supelco C18 5p 25 cm column was used 109 for the separation, using water and acetonitrile as the LC mobile phase with a flow rate of 1mL per minute. HPLC program used an initial acetonitrile: water ratio of 25% held for 3 minutes; a linear gradient over ten minutes increased acetonitrile to 90% concentration and held for 10 minutes, after which the acetonitrile is reduced linearly over a 5 minute span to 25%; the total run time was 28 minutes. Under these conditions, pyrene eluded at 19.44 i 0.14 min. A Waters 2487 Dual wavelength absorbance UV detector measured absorbance at 254 nm, while a Waters fluorescence detector with excitation wavelength of 295 nm and an emission wavelength 375 were used to detect pyrene. The fluorescence detector was used to increase the sensibility of the experiments for pyrene concentrations lower than 1 ppm. Manufactured Gas Plant Soil Samples were collected from a former manufactured gas plant site in Grand Ledge, Michigan. Three bore holes were dug and the contaminated soil extracted was analyzed for PAH contamination. The HPLC method used for the analysis of the MGP soil allowed for the resolution of 16 PAH peaks. This method has an initial a 25% acetonitrile concentration which was increased using a linear gradient over a period of 3 minutes to 90%. This condition was then linearly increased to 100% acetonitrile mobile phase and held for 20 minutes. Finally the acetonitrile concentration was linearly decreased back to the original 25% over an 8 minute time frame. 110 Results A maximum removal efficacy of 88.1i1.8% was achieved for Ottawa sand with an initial hydrogen peroxide concentration of 300 g/kg and 3 g/kg ferrous ammonium sulfate. Without the addition of ferrous ammonium sulfate, only 46.1 i- 5.8% removal of pyrene was obtained. Contaminated soil from a former Manufactured Gas Plant site containing 16 polycyclic aromatic hydrocarbons was treated under the same conditions as the previous soils. Removal efficiencies were less than 80% and accompanied by high experimental errors. Experiments carried out with sodium dodecyl sulfate (SDS), added to enhance the solubility of pyrene, resulted in less removal than those observed in experiments under the same conditions but without SDS. 100% 90% -F V i 80%—7 i ,7 ,,,——, 70% ~ 7, ,, 60% < 50% - 40% «» 30% 4 20% a 10% -> 0% Percent Removal H202 Concentration 120,000 ppm (10% Fe) I 120,000 ppm (1% fe) 8120.000 ppm (no Fe) III 300,000 ppm (soluble Fe) B 300,000 ppm (no soluble Fe) I 34000 ppm (Goe 1009m/kg) 568.000 ppm (Goe 1009m/kg) E] 68,000 ppm (Goe 200 gm/kg) Figure 4-2: Results from the Oxidation of Pyrene in Ottawa Sand using CHP. The graph shows the percent oxidation achieved at the different hydrogen peroxide concentrations used under Ferrous and Goethite catalyzed systems. lll Organic Matter Soil: Pyrene contaminated MSU soil was allowed to react using different concentrations of hydrogen peroxide as shown in Figure 4-3. Pyrene concentrations decreased as hydrogen peroxide concentration increased reaching a maximum oxidation level of 75.9 i 19.8% with soluble iron addition and 79.9 i 13.6% oxidation with no soluble iron added. The results are separated for each CHP condition; Figure 4-4 for soluble iron catalyzed reactions and Figure 4-5 for reactions catalyzed by mineral iron. 112 5% 80.08 on. or; on. on. o: can 085$ Eaa ooodw $9; Eaa ooodm 0&9; on. meowgcoocoo «on: can 80.8 once; on. do: «sebum 0528 we 008er 05 E can 822% 2: FEB Eon £8 sz mag—9:: genome com ommmow ocieom comet»: .w> Baguio 0:23 we E859 06 £52? Saw—m 2:. ”m-.. 953,.— Eaa Sod ouoc¥_lil_ on. E3 cor Eng ow Eng to on o: men-op; on. ouoc_ on. M « : :lw l .Fm —& $0 FM , {cow : , s8 - s8 - s8 y $om uomwo wowed - $05 , $8 -nxbm 860—. 113 100% 90% 80% 70% 60% 50% 40% 30% 20% 1 0% 0% Percent Oxidation H202 Concentration 1Oppm I 60 ppm 5100 ppm I 6.000 pm 1320.000 ppm 930. 00 ppm B 60,000 ppm B120 000 ppm lill 300, 00 ppm Figure 4-4 . Pyrene oxidation in soil containing organic matter catalyzed with soluble ferrous iron (percent oxidation vs. hydrogen peroxide concentration). I no 1 00% 90% 80% 70% 60% 50% 40% 30% 20% 1 0% 0% Percent Oxidation H202 Concentration El 0.1 ppm I 60 ppm 5 100 ppm I 6.000 ppm El 20,000 ppm 830,000 ppm 360.000 ppm $120,000 ppm B 300,000 ppm Figure 4-5: Pyrene oxidation in soil containing organic matter catalyzed with mineral iron (percent oxidation vs. hydrogen peroxide concentration). 114 Table 4-2 Results from oxidation of 50 grams of MGP soil with 120,000 ppm of hydrogen peroxide with and without soluble iron addition. Sample 1 taken at a depth of 10 feet, while Sample 3 was taken at a depth of 8 feet. ..€2tseétri“€%ain’i— ' cinemas...” Wands" teem? 2592' we 9‘ 99!“ fitment]. H262 was Reduced arhp Extracted before ,- . i] , of mi" : _' _ ,, H292:h‘@tlrtplitfi 1: 53.»: . ~ :1 _ 7, , Naphthalene < LOD - < LOD - Acenaphthylene 4,539 Sample 1 1670* (Fe Added) 63 i 32% Acenaphthene 4566* Sample 1 1591* (Fe Added) 65 i 70% 2510* (no Fe Added) 45 i 26% Fluorene 1093 Sample 3 776* (no Fe Added) 29 t 14% Phenanthrene 1567 Sample 1 1132 (Fe Added) 27.8 i 30% 1261 (no Fe Added) 19.5 i 27% Anthracene 542 Sample 1 396 (Fe Added) 27 i 29% 455 (no Fe Added) 16 :l: 36% 1175 Sample 3 259 (Fe Added) 78 i 49% 921 (no Fe Added) 22 i 7% Fluoranthene 3351 Sample 3 1279 (Fe Added) 61.8 :1: 17% 2813 (no Fe Added) 16 i 20% Pyrene 4938 Sample 1 3945 (Fe Added) 24 i 32% 4068 (no Fe Added) 18 i 27% 10927 Sample 3 2891 (Fe Added) 73 j: 12% 6389 (no Fe Added) 41 :l: 19% Chrysene 263 Sample 1 151 (Fe Added) 42 i 39% 229 (no Fe Added) 23 i 36% 579 Sample 3 98.37 (Fe added) 83 i 16% 329 (no Fe Added) 43 i 39% Benzo(b) fluoranthene < LOD - < LOD - Benzo(k)flouranthene 837 Sample 3 215 (Fe Added) 74.3 i 7% 532 (no Fe Added) 36.4 :t 19% Benzo(a)pyrene < LOD - < LOD - Indeno(l,2,3-cd)pyrene < LOD - < LOD - Dibenzo(a,h)anthracen < LOD - < LOD - e Benzo(g,h,l)perylene < LOD - < LOD - “ Estimated from values < LOQ and > LOD 115 R: (A) '1‘ m Discussion: Reactions that occurred in soluble iron-containing suspensions appeared to react differently when compared to that in suspensions that did not contain added soluble iron. White-grayish fumes, most likely NH3 emitted from the vials containing ferrous ammonium sulfate during reaction, which were visible during the first 30 minutes of the reaction. These same fumes were visible for a shorter period of time in experiments with Ottawa sand containing added ferrous iron. Reactions with soluble iron were also violent, and exothermic and resulted significant increase in the temperature of the container; while no noticeable change in temperature was noted in suspensions that did not contain soluble iron. A dark-red compound was formed during the reaction in suspensions containing ferrous iron; this compound seemed to act as a pH-buffer. The pH after 24 hours of reaction time was found to remain less than 3 in suspensions where the compound was formed, while in suspensions where the compound was not visible, the pH increased to greater than 4 after the same time period. The addition of 20 m1 of a 2 N sodium hydroxide solution were required to raise the pH to neutral conditions in the suspensions where the compound was visible; on the other hand, just a few drops of the same solution were required to reach neutral pH in suspensions lacking the red compound. The temperature rose during the soluble iron catalyzed reactions, reaching up to 90°C (measured by placing a thermometer in the flask while the reaction was taking place); although the temperature generally remained at ~70°C. Because a simple glass 116 thermometer was used, these measurements may have been recorded in the liquid-bubble interface, and the temperature in the bulk of the soil may not have been as high. Nevertheless the released energy from the reactions created an increase in temperature that may affect the oxidation process. The increase in temperature at the liquid—bubble interface will likely have large effects in field applications, especially when dealing with compounds whose flash point may be exceeded during the reaction. Temperature monitoring during field scale applications should be performed. Reactions in Ottawa sand had one noticeable difference to those of MSU soil and soil obtained from the MGP site; the reactions in the low organic content soils seemed faster than reactions in the other soils. Bubbling was visible in the regular soil samples for hours; while this bubbling was visible in Ottawa sand for only a few minutes. Ottawa Sand: The presence of soluble iron in Ottawa sand samples increased the contaminant removal efficiency. A possible explanation is the formation or lack thereof, of hydroxyl radicals in the system. Suspensions with no added iron would require the interaction of hydrogen peroxide with trace metals found in the sand to produce hydroxyl radicals. Alternatively pyrene may be oxidized by hydrogen peroxide, a weak oxidant when not catalyzed 6’ 9. On the other hand, samples with added soluble iron were able to produce a Fenton reaction, thereby producing sufficient hydroxyl radicals to oxidize the contaminant. 117 When the initial oxidant concentration was increased to 300,000 ppm from 120,000 ppm, a 40% increase in the removal efficiency of the contaminant (88.1%, compared to 62.8%) was achieved when soluble iron was present. Meanwhile, at the same concentration, in suspensions with no added iron, removal was much less (46.1%). If we assume that these soils contain few hydroxyl radical scavengers, the hydroxyl radicals formed should have few obstacles to prevent contaminant removal, thus these hydrogen peroxide concentrations should be able to oxidize the contaminant present. If hydroxyl radical formation is inefficient, contaminant removal is not likely to occur. From the results, we can determine that Ottawa sand in the absence of added soluble iron did not create the required conditions for pyrene oxidation. It appears that in Ottawa sand suspensions without added soluble iron, hydrogen peroxide is the major oxidizing compound. Oxidation by hydrogen peroxide is a much slower reaction than oxidations by hydroxyl radicals. Although some reaction occurred in these suspensions, a large portion of the hydrogen peroxide must not have reacted, resulting in lower removal efficiencies. This was evident by the release of gas during agitation after the given reaction time had elapsed. Oxidation reactions of pyrene in Ottawa sand produced consistent results due to the very controlled nature of the reactions taking place. With very few radical scavengers present, the possible side reactions appear to be minimal, therefore allowing straight Fenton’s chemistry to occur. 118 MSU soil: From Figure 4-3 we may conclude that the addition of soluble iron did not increase the removal efficiencies in the MSU soil samples, in contrast to what was observed with the Ottawa sand. The major difference between these two types of soil was the amount of hydroxyl radical scavengers present. MSU soil had components (including organic matter and metals, including mineral iron) that may have influenced the reactions and had some apparent effect on the pyrene removal efficiency. Suspensions with soluble iron could have reacted in the following manner: the iron in solution made it possible for the hydrogen peroxide to generate hydroxyl radicals and convert all ferrous ions present, into ferric ions. The reactions at this point become ferric ion catalyzed and depend on how fast the ferric ion cycles back to its ferrous state to produce new hydroxyl radicals 4’ 10. The initial surge of hydroxyl radicals formed would have reacted with any scavengers encountered, including other hydroxyl radicals, limiting the effectiveness of the Fenton chemistry on pyrene removal. The exothermic nature of the reactions also could have caused a loss of hydrogen peroxide, forcing any unreacted hydrogen peroxide to decompose into water and oxygen. The cumulative effect of these factors may be represented in the lower removal of pyrene when compared to suspensions with the same oxidant concentration in the absence of added soluble iron. Reactions in suspensions where ferrous ammonium sulfate was not added should and did behave differently because of their dependence on the availability of the iron mineral present in the soil to react with the hydrogen peroxide. The mineral iron in the soil slowly 119 A-M; dissolved, allowing the hydroxyl radical to slowly form and interact with scavengers and with the contaminants. This effect apparently helped the process; a less violent reaction occurred, allowing a smaller net loss of hydrogen peroxide from the solution by decomposition or volatilization; reducing the extent to which hydroxyl radicals reacted with each other. These effects resulted in improved oxidation in suspensions that had no added iron. Due to the heterogeneous nature of soil samples, the analytical error increased, as seen in Figure 4-4 and Figure 4-5, in samples from the MSU soil. This increase in error could be due to the side reactions that are more likely to predominate in more complex soils. Hydroxyl radical scavengers and soil organic matter present interact with the oxidation reactions, making it difficult for radicals to attack the contaminant directly. The number of side reactions that may occur in the field could be too numerous to model or account for, therefore high analytical errors can be expected. Sodium Dodecyl Sulfate In experiments containing SDS a great amount of foam was produced specially in those suspensions that contained soluble iron, and if carelessly attended, the foam would spill out of the containers. The addition of SDS to both Ottawa sand resulted in a decrease in the removal of the contaminant and an increase in the experimental error. The experiments yielded 16-59% 120 removal and errors between 17-54%. The soil samples that had been treated with SDS before oxidation clumped up and became difficult to dry. Experiments with SDS in MSU soil yielded less oxidation of pyrene than observed in experiments without SDS. The oxidations resulted in a 32.2 i 21% removal when soluble iron was added and 45.5 i 2.1% with mineral iron was used as catalyzer. It appears that the violent nature of the reactions with added soluble iron in combination with the added SDS, affected the ability to obtain consistent results. Surfactants are known to affect the surface tension of the fluid, if gas is being released and bubbles formed, the material trapped in these bubbles maybe taken out of the solution and not allowed to react, thus affecting the reproducibility of these reactions. SDS was added as a means to enhance the partitioning of pyrene to partition into solution and increase the oxidation efficiency of the hydroxyl radical, but results obtained do not show this, since no improvement was seen in the oxidation. Martens et al.2 found a 400- fold increase in pyrene oxidation (from 8% to 55%) after the addition of 5 mM SDS solution to his samples. Our results do not confirm Marten’s results. SDS in our system appeared to be detrimental to the oxidation process, acting as a hydroxyl radical scavenger. The rate constant for the reaction of hydroxyl radical with dodecylsulfate ions in aqueous solutions (8.2 x 109) is comparable to that of known 11 hydroxyl radical sinks, such as bicarbonate (8.6 x 108) and carbonate (3.9 x 108) ions This may explain the decrease in removal efficiency found in our samples. Furthermore, 121 the product of the rate constant (k) and the initial molar concentration of our components can be calculated; using the rate constant for pyrene (the documented range of PAH-OH radical rate constants is 5 x 109 to 5 x 1010) 11 and the solubility of pyrene at 25°C (6.76 x 10'7) ': Reaction with SDS: kSDg[dodecylsulfate] = 8.2 x 109 [5 x 10'6] = 4.1 x 104 ,, Reaction with Pyrene kpy,[pyrene] = 5 x 109 [6.76 x 10”] = 3.38 x 103 or using the higher reaction rate for hydroxyl radicals with PAHs kpyr[pyrene] = 5 x 10‘°[6.76 x 10”] = 3.38 x 104 We can conclude from these equations that hydroxyl radical reactions with SDS in aqueous solutions are very competitive when compared to the reactions between pyrene and hydroxyl radicals. This competition could yield results like the ones obtained, although they do not explain Martens’ results. MGP Site Soil The soils obtained from the Grand Ledge former MGP site were difficult to work with. The samples were heterogeneous, containing: gravel, sand, brick, tar, and water (Figure 4-6). The samples were homogenized to the greatest extent possible. 122 Figure 4—6: Contaminated soil being extracted from geo-probes during sampling at the Grand Ledge, MI former MGP site The results of experiments with MGP soil are shown in Table 4-2. No quantifiable data could be obtained using our method from Sample 2 (4-6 11 depth). This data was then discarded from further analysis/discussion. Results from the oxidation of the MGP soil, as shown in Table 4-2, were variable and were accompanied by large errors. The degree of contamination and the complexity of the matrix involved in the reactions lowered the degree of precision that could be obtained. This uncertainty was also present in the analyses performed on the soil by Consumers Energy, where concentrations of some PAHs were noticeably different between two sets of analyses performed in a 7 month time frame. Naphthalene concentration, for example, in Sample 1 was found to be 2.2 g/kg of soil based on analysis performed on November 2002; this same location was re-sampled in June 2003 and the naphthalene concentration was 4.4 g/kg of soil. Similarly pyrene was first reported at 820 mg/kg of soil and was later reported at 1,400 mg/kg for the same location. 123 Given the fact that the MGP site has been closed since the 19303, such an increase in concentration is unlikely to result from new contamination or the mobility of the contaminants. The site’s complexity affects the accuracy of any experiments performed on soil obtained from it. From the little information we can extract from the results, it seems that the addition of soluble iron assisted the oxidation process. Although iron was present in the soils excavated, it appears that the iron was not available to assist Fenton reactions. The inefficiency of oxidation is also clear; no contaminant was removed to a 90% level. With errors found in the system, removal efficiencies were poor. 124 Conclusions PAHs can be removed from soil using catalyzed hydrogen peroxide reactions. Pyrene was removed from several soil matrices with removal efficiency decreasing with increasing matrix complexity. Reactions in Ottawa sand achieved the highest level of removal (88.1i1.8%), followed by reactions in soil containing organic matter (75.9i19.8%) and finally MGP site soil (61 .8i17% for fluoranthene). Temperature becomes increasingly important in hydrogen peroxide catalyzed systems. As noted here the temperature of the matrix may increase substantially during reaction, this may cause an acceleration of all the reactions involved, including hydrogen peroxide decomposition. Catalyzed hydrogen peroxide systems do not appear to be the most sensible solution when dealing with highly contaminated soils like those found in MGP sites. The non- specificity of the hydroxyl radical plus the strictly aqueous environment required the use of higher chemical doses than might be expected from stochiometry. 125 References l. Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M., Environmental Organic Chemistry. 1st ed.; Wilkey-Interscience, John Wiley & Sons Inc: New York, 1993. 2. Martens, D. A.; T, F. j. W., Enhanced Degradation of Polycyclic Aromatic Hydrocarbons in Soil Treated with Advance Oxidative Process- Fenton's Reagent. Journal of Soil Contamination 1995, 4, (2), 1-14. 3. Kong, S.-H.; Watts, R. J .; Choi, J .-H., Treatment of Petroleum-Contaminated Soils Using Iron Mineral Catalyzed Hydrogen Peroxide. Chemosphere 1998, 37, (8), 1473-1482. 4. Watts, R. J .; Stanton, P. C.; Howsawkeng, J .; Teel, A. L., Mineralization of a Sorbed Polycyclic Aromatic Hydrocarbon in Two Soils using Catalyzed Hydrogen Peroxide. Water Research 2002, 36, 4283-4292. 5. Kawahara, F. K.; Davila, B.; Al-Abed, S. R.; Vesper, S. J .; Ireland, J. C.; Rock, 8., Polynuclear Aromatic Hydrocarbon (PAH) release from Soil during Treatment with F enton's Reagent. Chemosphere 1995, 31, (9), 4131-4142. 6. Beltran, F. J .; Gonzalez, M.; Rivas, F. J .; Alvarez, P., Fenton Reagent Advanced Oxidation of polynuclear Aromatic Hydrocarbons in Water. Water Air and Soil Pollution 1998, 105, (1998), 685-700. 7. Kakarla, P. K.; Andrews, T.; Greenberg, R. S.; Zervas, D. 8., Modified Fenton's Processes For Effective In-Situ Chemical Oxidation-Laboratory and Field Evaluation. Remediation 2002, Autumn, 23-36. 8. Nam, K.; Rodriguez, W.; Kukor, J. J ., Enhanced Degradation of Polycyclic Aromatic Hydrocarbons by Biodegradation Combined with a Modified Fenton Reaction. Chemosphere 2001, 45, (2001), 11-20. 9. Tyre, B. W.; Watts, R. J .; Miller, G. C., Treatment of Four Biorefractory Contaminants in Soils Using Catalyzed Hydrogen Peroxide. Journal of Environmental Quality 1991, 20, (October-December), 832-83 8. 126 10. Pignatello, J. J .; Baehr, K., Ferric Complexes as Catalysts for "Fenton" Degradation of 2,4-D and Metolachlor in Soil. Journal of Environmental Quality 1994, 23, (March-April), 365-370. 11. Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. Critical Review Constants for Reactions of Hydrated Electrons Hydrogen Atoms and Hydroxyl Radicals (oOH/O-) in Aqueous Solution; 1988. 127 Chapter 5. Byproduct formation from the oxidation of Pyrene by Catalyzed Hydrogen Peroxide Propagations Introduction In recent years F enton’s Reagent and its hybrid successor, Catalyzed Hydrogen Peroxide Propagations (CHP), have seen a surge in their use as in-situ remediation technologies" 2. CHPs consist of the addition of 30% hydrogen peroxide, accompanied by a metal ion catalyzer, usually an iron species3’ 4. Fenton’s reagent on the other hand consists of a dilute hydrogen peroxide solution (~0.03% solution), catalyzed by ferrous iron (soluble salt). Catalyzed hydrogen peroxide systems could be an ideal technique for in-situ remediation of contaminated sites: the system produces a strong, albeit, non-specific, oxidizing agent; the system is aqueous, allowing for its application to both soil and natural water systems; the catalyzer is readily found in natural soils; and unreacted hydrogen peroxide decomposes naturally to water and oxygen, usually saturating the matrix, which helps the proliferation of natural biota after remediation. These reasons have allowed CHP to be used for the remediation of sites containing: PAHs 5'13, -22 halogenated compoundsM'lg, petroleum derived products20 , and other hazardous chemicals. The system has also been used in combination with other remediation processes such as bioremediation and is a common part of advanced oxidation processes 23'25. A number of consulting companies utilize CHPs as their primary method of in-situ site remediation. 128 The oxidation efficiency of this technique is based on the hydroxyl radical formation. 26,127 Lindsey et a1. studied the effect of fulvic acid and humic acid on hydroxyl radical formation from Fenton’s Reagent, and found that hydroxyl radical concentration was four times lower in natural water samples when compared to samples with no dissolved organic matter. We can extrapolate from this that the formation of hydroxyl radicals in soil samples will decrease as well, due in part to the presence of soil organic matter, as well as metals (copper, manganese, mineral iron), carbonate and bicarbonate, ammonium and nitrite and any other oxidizable species in the soil. The expected outcome is that partially oxidized products will exist after treatment. Nam et a1 13, stated: “very little study has been done on whether partially oxidized organic compounds pose fewer hazards than the parent compound.” Fenton remediation was performed on soil contaminated with 3-,4-, and 5-ring PAHs by Lee et al. 28. PAHs were extracted from the matrix using ethanol, and Fenton’s Reagent was added to the ethanol extract. Several PAH diones were identified using GC/MS from this experiments, but it is unclear how the Fenton’s Reagent system behaves in an ethanol matrix and if the byproducts come from the interactions with organic radicals formed from ethanol or from Fenton oxidation directly. 29-31 Walling et a1. were the first to critically analyze Fenton’s Reagent and hydroxyl radical interactions with known contaminants. Walling found that hydroxylation of aromatic compounds with hydroxyl radicals was common. Walling theorized the reaction pathways and reaction rates for these compounds, for Fenton systems that were catalyzed 129 by it 115 its his ,. iili if? KC: H10 app PA] by ferrous, ferric and cuprous ions. Augusti et a1. 32 found through membrane introduced mass spectrometry (MIMS) that the oxidation/mineralization of benzene derivatives with Fenton’s Reagent proceeded via hydroxylation producing phenolic, hydroquinonic (p- hydroxy phenol) and quinonic (1,4 dioxy-benzene) intermediates. Similar functional groups were found after the oxidation of PAHs by Fenton’s Reagent. Luster-Teasley 33, Upham et al.34, and Yao et. a1.35 identified byproducts obtained from the ozonation of pyrene in soils. The byproducts found had biphenyl and anthracene backbone structures containing multiple, hydroxyl, carboxylic, and aldehyde groups. These byproducts, especially those with bay regions, were found to be more toxic than the original contaminant, leading to the question of what would be created in field applications if the contaminant were not completely mineralized during treatment. Although studies have been performed to measure the effectiveness of CHPs and FR in rernediating polycyclic aromatic hydrocarbons (e.g., see Kakarla et. al.36, Beltran et. a123., Kong et a121 ., Watts et al.9), the end products from these reactions are unclear. If the only factor measured is the disappearance of a compound, then there may be a risk of increased toxicity in the environment afier remediation, due to byproduct formation. In this study we have identified the byproducts formed from CHP reactions with pyrene as a model PAH, as well as their formation rate and estimated the potential fate of these byproducts in the environment. The knowledge gained from these studies can be readily applied and should identify potential hazards that may arise during or after remediation of PAHs is performed. 130 Method and Materials A single addition of hydrogen peroxide was used for all experiments. All reactions were carried out in borosilicate glass vials. The vials were cleaned in a 1% hydrochloric acid wash, rinsed in ultrapure water and allowed to dry. To remove organics, the vials were placed in an oven at 550 °C for 2 hours. . CHP reactions using soluble iron were allowed to react for 24 hours and quenched by adding sodium hydroxide (1M) to precipitate iron. Reactions using goethite as a catalyzer were allowed to react for 72 hours and quenched by decanting the supernatant solution from the catalyzer. Goethite used in the experiments was rinsed under a deionized water stream to remove iron dust from the surface and dried in an oven at 105°C. Finally, the goethite was cleansed of any organic contamination by placing it in an oven at 550°C for two hours, prior to use. Pyrene oxidation was carried out in two systems, water and Ottawa sand. For reactions in water, acetonitrile was used as a co-solvent to enhance the solubility of pyrene. The volumetric amount of acetonitrile in the solution was not allowed to exceed 3% (or about 1 ml in 20 ml of solution) because of reactions that occurred between the solvent and hydroxyl radicals. The effect acetonitrile had on pyrene solubility can be calculated by the Morris equation 37 131 sat Log mzx Csat =(alogK0w+beC) (1) W where, C33; is saturated concentration of the solute in the mixture C fiat is the saturated concentration of the solute in water K0W is the water-octanol partition coefficient for the solute, fc is the volume fraction of co-solvent, and a and b are coefficients unique to the solvent. For acetonitrile, these coefficients are 0.9 and 0.83 respectively3. According to this equation, the solubility of pyrene should increase by 48% to about 0.21 mg/L. Initial pyrene concentrations were 100 mg/L, resulting in a suspension with the water. A pyrene stock solution was prepared by dissolving 0.5 grams of the compound in 200 mL of acetonitrile. This solution was then diluted as appropriate. Ottawa sand was rinsed with acidic ultrapure water and allowed to dry prior to use. These experiments were carried out in the absence of acetonitrile. The pyrene stock solution was added drop by drop to a vial containing 20 grams (dry weight) of sand while being 132 mixed by a Mistral Multimixer (Model 4600, Lab-Line Instruments Inc.). The vials were placed in a dark hood for 24 hours to allow the solvent to evaporate before commencing the experiments. The initial pyrene concentration was held at 300 mg/kg of sand. Ottawa sand reactions were carried out as aqueous slunies. The soil (20 g dry weight) was saturated by adding 4.5 mL of water, which contained iron at a concentration ranging fiom 0.001 to 0.06 moles/kg of soil. For reactions with mineral iron, goethite was added to the dry sand while being mixed, to allow homogenization of the mineral in the matrix, before the addition of water. Byproduct Structure Identification Sample preparation Samples from the oxidation experiments were placed under a gentle nitrogen gas stream until completely dry. Hexane (5 mL) was added and the vial was placed on an orbital shaker (Lab-Line, Model 3590) for 1 hr. The hexane was then decanted and 5 mL of acetonitrile was added to the vial and placed on the shaker. This process was repeated with 5 mL of 0.1 M sodium hydroxide solution followed by 5mL of 0.1 M perchloric acid solution. The extracts were then taken for LC/MS analysis. A Shimadzu LC-20AD HPLC was used coupled with a Waters Quattro micro API mass spectrometer. The LC used a 5n Thermo Hypersil-Keystone BetaBasic-18 PIONEER column with dimensions of 150 by 1mm. The LC used a gradient program with 133 acetonitrile and water, containing 0.15% formic acid, as the mobile phase and the flow rate was 0.1 mL/min. At the start, the ratio of acetonitrile to water was 10:90, the ratio was increased linearly to 60:40 over 30 minutes, and then continually increased to 100% acetonitrile over ten minutes. At this time the concentration of acetonitrile was decreased to 60% over ten minutes, after which the concentration was decreased immediately to 10%, with this concentration held for 5 minutes. This program had a total run time of 60 minutes. The LC/MS used a Waters Z-spray electrospray ion source with acquisitions performed in both positive and negative ion modes with a cone voltage of 40 V and a capillary voltage of 3.17 kV. The desolvation and cone gas (N2) had flow rates of 400 W and 10 L/hr respectively. For MS/MS analysis, the collision energy was held between 25 and 35 eV. Results CHPs catalyzed by ferrous ion are more violent and exothermic than those catalyzed by ferric ion, because the rate of hydrogen peroxide decomposition into oxygen is faster when ferrous ion is present, leading to a rapid loss of hydrogen peroxide from the system (as described in Chapter 2). Pyrene was oxidized by CHPs in both soil slunies and water systems. Reactions that occurred in solution achieved only moderate removals as compared to those in soil slunies described in Chapter 3. As seen in Figure 5-1, reactions with soluble ferrous iron resulting in 9.4% removal; while reactions with ferric ion achieved 15.2% removal in 134 aqueous solutions. Pyrene removal increased for both of these ions in Ottawa sand slurry reactions, with ferrous ion catalyzed reactions achieving 36.1 % removal and ferric ion catalyzed reactions resulting in 49.8% removal. 0%: ::: \ ,.- V \ Ferrous Ottawa Ferric Ottawa Figure 5-1: Percent removal of pyrene from CHPs for solution and Ottawa sand reactions. The initial iron catalyst concentration was held at 0.06 M, while the initial hydrogen peroxide concentration was 0.11 M. Goethite catalyzed reactions also displayed the trend found from the reactions with soluble iron, higher pyrene removal in Ottawa sand experiments as compared to reactions in aqueous systems. For reactions with a Goethite equivalent of 0.06 M ferric ion, the pyrene concentration was found to be below our analytical detection limit for Ottawa sand Slurries, while reactions in aqueous solutions resulted in less than 1% removal. Byproduct Formation Byproducts from the oxidation of pyrene with both soluble and mineral iron catalyzed CHPs have been found and identified. Several compounds having different mass to charge ratios (m/z) were found in hexane and acetonitrile extracts, while no byproducts were found in the water extracts. The full range of compounds found is listed in Table 5-1. MS/MS analysis was performed in order to obtain information about the structure of the compounds formed. Figure 5-2 and Figure 5-3 show an example of this analysis the electrospray mass spectra and MS/MS spectra for m/z 233 which is believe to be a pyrene-quinone structure. The complete set of electrospray and MS/MS spectra for the compounds found can be seen in Appendix A. Table 5-1: Table showing m/z found in the combined extracts from the different CHP tested. 233, 207, 235, 315, 275, 247, 234, 251, 221, 267 219 257, 255, 249, 251, 331, -485 , 204, 353, 271, -329, -285, -315 -286 136 :5: . .= ante Table 5-2: Byproducts identified by LC/MS/MS with a list of major m/z peaks found in the spectra ‘6 .,:. . - .-,,- ,-.- " 1'"; .‘t 'i’l '5 l ‘ 219, 201,191 FI 219ES+ FII 217ES- 271/216, 189 FIII 235ES+ 235, 217, 207, 192, 179, 164 FIV 251ES+ 251, 233, 205 FV 233ES+ 233, 205, 177/176, 151 FVI 249ES+ 249, 221,203,192/ 193,165 FVII 221ES+ 220, 1777/176, 165, 151, 77 FVIII 267ES+ 267, 249, 239, 221, 193, 165 FIX 247ES+ 247, 218, 201, 190/189 FX 204ES+ 204, 203, 202, 176, 151 FXI 223ES+ 223, 205, 177/176, 152/151, 98, 63 Pyrene 203ES+ 203, 202, 201 137 A858 0938: of E 988 83 NE £5. 625.com E .8328 E 6:23 we 596me 2: 80¢ 35890 mmm NE e8 88on 32: 358585 mg“: "aim 95w?— oom 8m 8m SN 8 SN 0% SN own SN 8w 8. o3 o: 8. 9: 9; one om. N): .,.uWW-..pm_:-0.L.:...0..V.:.,..i...i—.0.1.PIVJVW.ov._:u».—-utv.ouvun.4.» LT...../_.o.w._.o.:..—.\..L:vpr:”munmbrp:kptnw:;VJ.n.:n0......mpunppi:3L:.V?»n$-I.LPJV—.I..L::1....or:6 8m 8 new 8m 08 238538 new :8 8% 8m ofimemua 8m 8e 22: 35:3 8. one 7.3 8%? 845 at.“ new a: emu r 1% Ewen 8N . 8. +8 88 a 2838::38:58:90:so22.582 so 9:85 Swot no“ on mOmvime nu coozgopfio 138 ammo cog—See ozfioom mo Hombxo 0538: 05 Set «confine .mmm NE 80¢ meow Connemara meg/2% «:0on £392 "mm PSME SN ova 8N cum 9“ 8m 8F 2: at 2: one 9; one our o: 2: 8 on S 8 8 NE. .. . .21.. . . ...._._.. n . ..... . ...1. . . . . . . . . . . . . . . . _. ._ ._ _ _ _ ._.. N x. f \ a f / _ f _ / \ / 3N gamma 53mm imam \_r 378— Em _ m2 853 9; one RF 852.5969 mm/mm 8 5 E 8% a 8 No SN vow or; new a: $4 MNNMN 33 ._oz NONIEU e: O O 83.4 E +mm8~ do 96593 a foo e Afimmumfllmiémtmv: :5 £845 8. 8018290 mam: 32E 8: an moo.me 2 canamuotao 139 Electrospray ionization produces both negative and positive ions; negatively charged ions have been highlighted in Table l. Electrospray ionization creates adducts with positive ions such as [M+H]+ formed by cation attachment, , while negative ions are often formed by deprotonation to give [M-H]'. Any m/z found would require the addition or subtraction of 1 Dalton to establish its molecular mass. Not all m/z found could be identified as byproducts from the reactions because they did not produce sufficient ions from the electrospray process to allow identification. For other structures, such as m/z 233 ([M+H]+ of pyrenequinone), insufficient information was available to identify the specific structural isomer, but molecular ion and fragment ion masses in MS/MS spectra were consistent with its assignment as pyrenequinone. There is as well the possibility that some byproducts may have been lost during the drying of the sample with nitrogen gas, or were irreversibly sorbed to the walls of the reaction vials 1]. The structures for the byproducts identified can be found in Figure 5-4 and in the first column in Table 5-4. 140 OH 0 HO OH F1 0 FIB i FIIIA l I FIIIB CHO CHO CIOHIOO C16HIOO C16H1002 C16H1002 Mol. Wt.: 218.25 Mol. Wt.: 218.25 Mol. Wt.: 234.25 Mol. Wt.: 234.25 o FIV CHO COOH WA. 0 C16H1003 CI6H802 C16H802 C16H803 Mol- Wt-z 250.25 M01. Wt.: 232.23 M01. Wt.: 232.23 M61. Wt.: 248.23 FVII O FVIII FIX H H o CHO CHO OH HO FX 0 O O .. O Q CHO CHO CH0 HO CISHSOZ C16H1004 C13111005 C16H12 Mol. Wt.: 220.22 Mol. Wt.: 266.25 M01- Wt: 246.22 M01. Wt: 204.27 i NH FXI CHO OH FXII N|H FXIII i 2 CISHIOOZ C18H13NO M lcégliléfi 27 M01. Wt: 222.24 M61. Wt: 259.30 0- t-- - Figure 5-4: Byproducts identified from the oxidation of pyrene by CHPs in water and soil Slurries, using mineral and soluble iron catalyzed reactions. *3 isomers of the structure were found, but not readily distinguished. 141 -2e'l -2e' keto-enol MW 234 0“ H01 0 W0". ”W 23“ * earrangement H. -2e -2H CHO CHO MW 2 48 000 M. 234 ago 0 eel -2H+i MW 250 CHO COOH -2e' 1 rearrangement HJ‘O i-2H+ COOH 090 MW 266 OHC CHO Figure 5-5: Proposed pathway for hydroxyl radical degradation of pyrene by CHPs 142 Discussion A greater extent of removal was obtained in the presence of Ottawa sand as compared to aqueous solutions. Pyrene solubility may have had a major role in this result, because although the addition of acetonitrile increased its solubility, the amount of acetonitrile added may have been insufficient to ensure that pyrene was solubilized and available to interact with the hydroxyl radicals produced, instead hydrogen peroxide may have decomposed and the excess hydroxyl radicals reacted with the other species in the reaction (ferric ion, ferrous ion, and hydrogen peroxide). The fact that pyrene was present in a greater amount in a suspension and not in the solution, would have prevented most of the pyrene, from reacting with the oxidant. It is thought that pyrene was more available to react in Ottawa sand slunies because it may have sorbed to sand particles allowing any hydroxyl radicals formed nearby to interact directly with the contaminant. These results are consistent with findings from Masten”, where PAH removal by ozonation was greater in soils slunies than in aqueous solutions. Pyrene oxidation byproducts were found for both soluble and mineral iron catalyzed CHPs. The byproducts tentatively identified range from pyrene substituted with a hydroxyl group to quinone to ring cleavage products. All of the byproducts found appear to have been formed from the reaction of hydroxyl radicals with pyrene and are thus oxidized versions of the contaminant. The byproducts found were formed from a single addition of hydrogen peroxide, which could have limited the degree and/or extent of oxidation that was achieved, in comparison the continuous addition of hydrogen peroxide 143 would likely have resulted in the formation of a continuous supply of radicals into the system, producing more highly degraded products. Figure 5-5 shows the proposed pathway for hydroxyl radical interaction with pyrene. The pathway was based on the assumption that hydroxyl radicals were the only species present in the system capable of oxidizing the contaminant. Numerous studies have found that hydroxyl radicals interact with aromatic compounds by hydroxylating the ring. This is the first step in the mechanism, from there on the nonspecific nature of the hydroxyl radicals can form multiple byproducts and isomers, leading to phenolic (FI), ketonic (FIB), quinonic (FVA and FVB), carboxylic (FIV); and ultimately, biphenyl structures are formed (FVIII and FIX). Byproduct formation between different Catalyzers The byproduct formation varied under the three catalyzers used, as seen in Table 5-1. As a general trend we can conclude that an increase in hydrogen peroxide concentration (oxidant) increased the extent of oxidation of products found in solutions catalyzed by goethite. For the three concentration schemes used only the combination of 0.38 M hydrogen peroxide with a 0.06 M (as ferric ion) goethite concentration achieved significant byproduct formation. Under these concentrations we were able to find 16 different products (m/z), while under concentrations of hydrogen peroxide (0.01 and 0.11 M) combined with goethite concentrations of 0.06 and 0.001M only two products were found m/z 241 and 219. The difference in the number of products found can be attributed to the nature of goethite catalyzed reactions. Heterogeneous catalysis of hydrogen 144 peroxide is a surface reaction that requires the diffusion of hydrogen peroxide and pyrene to the goethite surface. If undissolved pyrene covered the surface of the catalyzer it may have impeded the diffusion of hydrogen peroxide to the active sites, and thus less of the oxidation reaction may have taken place; an increase in oxidant concentration appeared to force its way into the catalyzer surface thus increasing oxidation. Both ferrous and ferric catalyzed systems created byproducts, but the byproducts identified in ferric catalyzed reactions appeared to be more oxidized than those from ferrous systems. As a result, byproducts FVIII and FIX, corresponding to the two byproducts with the highest degree of oxidation, were found in fenic systems. In ferrous systems the most oxidized byproducts found were FIV and FVI, which are a few oxidizing steps behind the ultimate products found by ferric catalysis, as seen in Figure 3. An increase in hydrogen peroxide and iron concentration in both systems resulted in less byproduct formation. This decrease in byproduct formation may be due to an increase in hydrogen peroxide decomposition at these concentrations of iron, with ferrous systems having the highest decomposition rate between the two. This result is consistent with the results fiom oxygen decomposition described in Chapter 2. Ottawa sand experiments resulted in fewer byproducts than those found in the aqueous solution experiments. This result may be due to several factors: the byproducts may have been sorbed to the soil particles; the impurities extracted from the sand may have masked byproducts and thus we could not separate them during analysis; or the extraction method 145 used may not have been the optimal. Although any of these reasons may be true, byproduct m/z 233 (pyrenequinone) was produced by all three catalyzers in this system. Pyrenequinone (FV, m/z 233) was found throughout all the experiments, making it the most likely byproduct to be found in remediation of pyrene contaminated sites. The formation of pyrenequinone has special significance because: (i) pyrenequinone is known to be toxic”, (ii) its presence validates that oxidation is indeed occuning; and (iii) Chen and Pignatello40 found that the presence of quinone intermediates helped catalyze CHP reactions by converting the ferric ion into ferrous ion. Many different isomers of pyrenequinone may be formed and the byproducts shown in Figure l (FVA and FVB) represent two of the most stable isomers that may be formed. Unfortunately, insufficient information was obtained from the MS/MS data to distinguish between the possible isomers. A sample LC/MS chromatogram showing the ion abundance peak for m/z 233 (pyrenequinone or FVA or FVB) versus the total ion count, and its molecular ion peak is shown in Figure 5-2. Figure 5-3 shows the MS/MS product ion mass spectrum obtained from this m/z with the fragment ions formed arising fiom successive losses of CO (28 Da) that are consistent with a quinonic structure on the pyrene ring. The reason structures F VA and FVB were selected as the most probable isomers is because hydroxyl radicals are most likely to react with carbons 4,5 and 11, 12, which have a greater electron activity present. The increased electron activity in the area is due to the tt-tt bonds located here, Which create a larger electron cloud, and thus making these 146 carbons more likely to interact with radicals. The byproducts found from the ozonation of pyrene33 confirm that these are the most reactive bonds. Figure 5-6: a) Pyrene molecule with numbered carbons and b) showing p-orbitals around the center region of the molecule. Byproduct FX is an interesting compound because it is a product of a reduction reaction rather than an oxidation reaction. The most likely scenario is that of pyrene reacting with excess hydrogen peroxide or with superoxide to reduce one of the double bonds forming hydro and di-hydro pyrene. These compounds were readily found in all CHP reactions investigated. Byproduct FX was easy to find because it fluoresced at the same excitation and emission wavelengths used for pyrene but had a different retention time compared to that of pyrene. Byproducts F X11 and FXIII are products resulting fi'om the reaction of acetonitrile, a co- solvent in aqueous reactions, with hydrogen peroxide or hydroxyl radicals. These byproducts, although unexpected, may be of significance because of the multitude of contaminants and chemical species present in the natural environment. Coal tar contaminated soils contain benzene, naphthalene and aliphatic compounds that may be involved in the reactions to form a multitude of byproducts. The structures found 147 illustrate that other radical species may be formed during CHP reactions and these radicals can react with the contaminant of interest to form other byproducts. A compound with m/z 208 was found, but could not be positively identified. Enough information was present in the mass spectra to determine that the structure was comprised of an acetarnide group attached to a highly oxidized bi-phenyl structure. The acetamide group was formed from the reaction of hydroxyl radicals with acetonitrile (co-solvent) making it a similar reaction to the one that formed byproducts FXII and FXIII. Comparison to PAH byproducts found from Fenton’s Reagent Reactions Beltran et al. 23‘ 25 identified byproducts from Fenton’s Reagent reactions and hydrogen peroxide/ozone reactions with PAHs (fluorene, phenanthrene and acenaphthene) in water by GC/MS. Beltran et al.23 reported 11 different byproducts including phenolic and quinonic derivatives; but researchers were unable to identify 16 other byproducts. These byproducts found by Beltran are similar to those found in this study. From the 11 identified products three were a form of the hydroxylated PAH (9-phenanthrenol, 9- fluorenol, l-napthol 2-,ethyl), one was a ketone (9-fluorenone) and 2 were biphenyl structures (o-hydroxybiphenyl and 2(H)-l-benzopyran-2-one,3,4-dihydro-). These byproducts are similar to the structures identified for byproducts FI and F111 (hydroxylated), FIB and FV (ketonic or quinonic), FVIII and FIX (biphenyl). These structures are also consistent with the proposed pathway, shown in Figure 5-5, of degradation by hydroxyl radicals, where an initial hydroxylation leads to phenolic 148 compounds, and a second reaction leading to ketonic structures. The final products of these reactions are substituted biphenyls which were found by Beltran and this study. Beltran unfortunately does not go into discussion about the byproducts formed, but states that a smaller quantity of byproducts was found from oxidations using Fenton’s Reagent than with oxidation using ozone, but was not sure why this happened. The major difference between ozone and CHP reactions is that the former can react directly with the contaminant while the latter goes thru a series of transformations to form the radicals that attack the molecule. Ozone is also capable of forming hydroxyl radicals; the result is then that more of the contaminant reacts in the ozone system, when compared to a Fenton’s reaction, for the same dose of oxidant. Comparison to PAH Byproducts from other Remediation Techniques. PAH byproducts from other remediation techniques have been identified. Byproducts 39, 43, 44 have been found from, electrolytic aeration“, photolysis”, , biodegradation and ozonation33‘ 35‘ 45. Electrolytic Aeration of pyrene Goel41 has shown found that naphthalene is degraded by electrolytic aeration in a similar manner as pyrene, in the proposed pathway, starting with phenolic structures and leading to ketonic and quinonic byproducts. Goel indentified 1,4 naphthaquinone and, l-napthol as byproducts. 149 Phat 121‘s“: in a 1101.. PTO? for iii 26: Photolysis of Pyrene 1.42 used l-hydroxypyrene as a model PAH to study the irradiation of UV light Zeng et a in a water/acetonitrile system. The authors identified two pyrenequinone byproducts along with a pyrene quinone dimer. The byproducts in this study are consistent with the proposed mechanism shown in Figure 5-5. The only difference between the byproducts found in this study and the byproducts found by Zeng is that the actual location of the quinonic groups was not determined in this study but both of the structures found by Zeng can be found in the proposed mechanism as possible results from the oxidation of pyrene by hydroxyl radicals. Biodegradation of Pyrene and other PAHs Byproducts (metabolites) from the biodegradation of pyrene by Mycobacterium sp. have been identified by Heitkamp44 and are shown in the third column of Table 2. The degradation of PAHs by aerobic bacteria follows hydroxyl radical interaction through enzymatic catalysis; oxygen from the atmosphere is transformed by a di-oxygenase enzyme into hydroxyl radicals (and sometimes, superoxide) that attack the contaminant 46’ 47. These metabolites are similar to byproducts found by CHP oxidation. As seen in Table 2, hydroxypyrene is found in both systems (structures FI and BI). Pyrenone (FIB), which is found in CHP treatment, is one oxidative step away from this structure. Phenanthrene carboxylic acid (BIV), also found after ozonation (OIV), has a similar structure to byproducts FIIIB, FIV and FVII found in CHPs. Byproduct FIIB and F IV are one oxidative step away from becoming phenanthrene carboxylic acid, while FVII is a 150 tautomerization of this product. This implies that both biodegradation and CHP treatment follow similar schemes for the oxidation of PAHs. An important aspect of bioremediation that should be taken into consideration is that enzymatic processes are selective, while hydroxyl radicals generated fiom CHP treatments are not. Metabolites from biodegradation and byproducts from CHPs should differ because of this, since hydroxyl radicals in a CHP reaction would not selectively form one structure over another. The selective nature of enzymatic processes allows them to become more efficient oxidizing systems than those that are non-selective. Hydroxyl radicals generated during CHP reactions will react with any species present in solution, instead of just reacting with previously oxidized species. Liang39‘ 48 found 4,5 pyrene quinone (BVI) from the biodegradation of pyrene in soil microcosms containing Mycobactrium sp. This byproduct was also found in the CHP remediation of pyrene in water and Ottawa sand (FV). The byproduct found by Liang indicates that biodegradation in soils is similar to that of the PAHs in solution and the metabolites found by Heitkamp should also be present in the Liang experiments. This finding supports the proposed pathway, because the structure produced is a product of hydroxyl radical reaction with the most reactive bonds of the molecule as stated earlier in this paper. The biodegradation of naphthalene and benzanthracene yielded similar byproducts as compared to those obtained from the bioremediation and CHP treatment of pyrene. 151 Salicylic acid, catechol and naphthol have been found from the oxidation of napthalene“, while two dihydrodiols of benzanthracene were found from bezanthrazene degradation. Both of these studies confirm that biodegradation of PAHs follows the phenolic, ketonic/quinonic pathway that has been proposed. Ozonation of Pyrene Most of the information on byproduct formation from the remediation of PAHs comes from chemical oxidation using gaseous ozone. Studies by Luster-Teasley et a1. 33 and Yao et a1. 35’ 49 give an extensive look into the byproducts formed from the oxidation of PAHs with gaseous ozone . The studies found that the ozonation of pyrene resulted in the oxidation of specific bonds forming aldehyde, and carboxylic acid functional groups and later ring cleavage byproducts, in which biphenyl structures are seen and bay regions are formed, as shown in the second column of Table 3. Most of the byproducts found from ozonation are more oxidized than those from CHPs in this study. Two byproducts found in this study, the substituted biphenyls (FVIII and FIX), are similar to those formed from ozonation (OV and OXII). In general, CHP oxidations formed precursor byproducts to the substituted biphenyls found by Luster-Teasley and Yao, but more of these biphenyl structures would likely have been formed if more hydrogen peroxide had been added. As seen in Table 3, byproducts FIIB, FIV, FVII and FVIII are homologous to byproducts 0111, ON, OVI and OVII found from ozonation. The difference in byproducts produced in the two systems, CHP and ozonation, may be explained by the reaction pathway. Ozonation results in two types of reactions, direct 152 (molecular) ozone reactions and hydroxyl radical reactions. CHP reacts only through hydroxyl radical interactions and thus for similar oxidant doses, reaction byproducts found from ozone remediation will have a greater degree of oxidation than those from CHP reactions. If we would continue to add hydrogen peroxide we may achieve the same degree of oxidation in CHP systems as those found in ozone systems, byproducts FVIII and FIX may be indicators of this possibility. PAHs partition to the non-polar region of a matrix 50; in soil this region is the soil organic matter. If the molecules are bound to the soil organic matter and do not solubilize well to the water phase, then a strict aqueous system like CHP, may not reach the contaminant to the extent required for mineralization. Ozone when used in its gaseous state may reach the contaminant more freely because it would not require water as a means of transport to the contaminant. As Luster-Teasley found, the humidity of the soil affected the rate of oxidation, the rate of pyrene disappearance decreased as the percent humidity increased obtaining differences of 6% and 12% for soil conditions of 5% and 10% humidity respectively, when compared to dry soil 33. The experiments also found that an increase in pH accompanied by an increase of humidity in the matrix resulted in a decrease of up to 40% in the extent of oxidation (found at pH 6-8). These results would indicate that ozone may not produce complete oxidation of the contaminant if the contaminant is in an aqueous phase, unless the ozone dosage is increased, and that byproducts will most likely be present after the treatment is finished. 153 Environmental Impact of PAH remediation Mobility The hydroxylation of contaminants may cause an increase in the mobility of the molecule, by increasing the polarity of the molecule and the means to form hydrogen bonds with water. By using the SPARC estimation software available from the University of Georgia, Chemdraw Ultra (CambridgeSoft) and EPISuite available from EPA, some physical chemical properties for the byproducts found were estimated and seen in Table 5-3. In general an increase in mobility or aqueous solubility could cause a concern for in-situ chemical oxidation (ISCO) remediation. Since the agent is pumped into a groundwater system per se, then the fluids in the matrix will be forced into motion. If the contaminant becomes more soluble; then the end result is the spreading of the contaminant and the amplification of the treatment zone and potential threat. If the structure of the byproducts is known and the conditions that create these byproducts are also known, then determining the environmental fate of these byproducts during and after remediation should be accomplished. Parameters were estimated for two different pyrenequinone structures as well as the two isomers for F 111 since we could nOt determine the correct structure fiom our data. The aqueous solubility of the byproducts FIB and FIX are about one order of magnitude greater than that of pyrene. Byproducts FVII, F VIII and FXI have aqueous solubilities that are least 2 orders of magnitude greater than that of pyrene. All byproducts with 154 exception of compounds FIIIA and F X have a lower K0W than pyrene. All byproducts also show a lower K0m value than pyrene with the exception of FX, dihydropyrene. These findings suggest that most byproducts will have a tendency partition to the aqueous phase as opposed to an organic matter or solvent (perhaps a DNAPL) as pyrene would. Byproducts FVII, FVIII and FXI are very soluble, which means their mobility in the environment is likely to be greater than that of pyrene. The SPARC estimated data (water solubility, Kow, diffusion coefficient) are very dependent on the melting point used. If we assume that all byproducts have the same melting point as pyrene then the solubilities of all compounds would increase by at least one order of magnitude. The melting points used for these estimations, were all estimated from EPISuite, which relies on functional group contribution to estimate the melting points. This method may have its limitations; byproducts FIIIA, FVI, and FIX all have the same estimated melting point even though they are different in structure. Byproducts, such as FVI which has isomers, will have different melting points depending on the isomer and as such, the estimated data may not be accurate. Toxicity l.45 and Hemer et al.51 found that byproducts formed from the Luster-Teasley et a oxidation of PAHs with gaseous ozone inhibited gap junction intercellular communication (GJIC) of rat-liver epithelial cells, leading to the conclusion that some daughter products formed during remediation were in fact more toxic than the parent 155 compounds. They also concluded that byproducts that contained a bay region (like biphenyl groups) were more toxic than compounds that did not have this structure present. Little is known about the toxicity of byproducts formed from Fenton’s Reagent reactions. As Bowers et al.52 described intermediates from the remediation of wastewater contaminated with refractory chemicals may be more toxic than the parent compound if the reaction does not completely mineralize the contaminant. Satoh et a1.53 studied the epigenetic toxicity of hydroxylated biphenyls and hydroxylated polychlorinated biphenyls (PCB), all potential products from Fenton’s Reagent and CHP reactions with biphenyls and PCBS, and found that the hydroxylated products were in most cases equal or more toxic than the parent compound. Sedlak54 found and identified several hydroxylated PCBs alter the oxidation of chlorobenzene with Fenton’s Reagent. The PCBS were apparently of hydroxylated chlorobenzenes formed during reaction. The formation of these could increase the toxicity of the solutions once reactions are terminated. We have found 11 byproducts relating to the oxidation of pyrene by CHP reactions. Some of these compounds are known to be toxic. Byproduct F I (1-hydroxypyrenc) is known to be acutely toxic and genotoxic”, it is commonly used as a biomarker for exposure to PAH contaminations. This first byproduct is also known to be toxic to numerous to other organisms“ . Products FIIIB, FIV, FVIII, FIX and FXI all contain a bay region with an aldehyde group, which is known to increase the toxicity of the molecule“. Byproducts that have quinonic structures like FIB, FVA, FVB and FVI may be toxic to other forms of life, including bacteria”. 156 Conclusions: The remediation of PAHs using CHP systems in both aqueous solutions and soil Slurries produce byproducts. Eleven byproducts have been identified from the chemical oxidation of pyrene in these systems. The byproducts found are similar in structure to those found in from bioremediation and ozonation of pyrene in the same systems. Three different catalyzers were used for the CHP reactions in the two matrices. Byproducts were found in all of the systems, therefore we may conclude that byproducts are produced whenever CHPs are used for the remediation of PAHs. The byproducts found are consistent with hydroxyl radical interaction with aromatics. A proposed pathway that may lead to the products found has been developed. Parameter estimation software has been used to estimate the potential fate of the byproducts found. The results show a general trend of decreasing Kow for all byproducts when compared to that of pyrene. Aqueous solubility some cases increased by 3 orders of magnitude, leading to the belief that the byproducts formed may be more mobile that the parent compound. Some of the byproducts found are known to be toxic or to interrupt GJIC communication. This leads us to believe that CHP reactions yield products whose toxicity is greater than that of the parent compound. 157 nil—d F-m PF-HUP—d bF-nJTnnJLP—Uih Avd “damn-d IothauJudinnwn—HJ Tn...-———2x .aJn—d z-H ——:..J\(I I-w ud—n-unv.‘ turd-duvAV--F->F~ Un—n.‘Av MVP-av -.d-..-:iv .nnm‘ .I.--o.vhaiv~——-w--wfih masz-w—n-~[.~ I": v4 U‘Afiuw ~ .. o a. - ll ‘ 33c «£5 3895.5 88m 333030 0 «ED 335805 89a 33880 e Ofimm mam: @03580 0 mum 88% 33880 a BEmEm wean 33880 3 .0338 noueafimo 53888 05 E Eofi E828: 3 e8: 8.633323 mums—Em 05 me :03 me venom muoavoazn one no one some 8m 88383.39 33::3m "mum oEah 158 Table 5-4: Byproducts found from CHP, gaseous ozonation and bioremediation remediation of pyrene. Bioremediation by CHP Gaseous Ozonation My e ob a cterium sp. OH OH FI B1 0 CHO FIB H OH i i cuo CH0 FIII Q BIII CHO COOH 159 Table 5-4 Continued CHP Gaseous Ozonation Bioremediation by Mycobacterium sp. O O HO FVI CHO COOH 3% < I CHO CHO OVI € CH0 CHO FVIII HO HOOC % OVIII OH HO i CHO HO FIX CH0 CH0 i CH0 COOH 01 X H H at FX CHO OH CO FXI 160 References 1. Council, I. T. R. Technical and Regulatory Guidelines for In-Situ Chemical Oxidation of Contaminated Soil and Groundwater 2nd ed. ISCO—Z; Interstate Technology & Regulatory Council: Washington, DC, 2005. 2. Huling, S.; Pivetz, B. E. Engineering Issue Paper: In-Situ Chemical Oxidation; 600R06072; Environmental Protection Agency: Ada, Oklahoma, July 28 2006, 2006; p 60. 3. Watts, R. J .; Teel, A. L., Chemistry of Modified F enton's Reagent (Catalyzed H202 Propagations-CHP) for In Situ Soil and Groundwater Remediation. Journal of Environmental Engineering 2005, 131, (4), 612—622. 4. Watts, R. J.; Teel, A. L., Treatment of Contaminated Soils and Groundwater Using ISCO. Practice Periodical of Hazardous, Toxic, and Radioactive Waste Management 2006, 10, (1), 2-9. 5. Belkin, 8.; Steiber, M.; Tiehm, a.; Frimmel, F. H.; Abeliovich, A.; Werner, P.; Ulitzur, S., Toxicity and Genotoxicity Enhancement during Polycyclic Aromatic Hydrocarbon Degradation. Environmental Toxicology and Water Quality 1994, 9, (4), 303-309. 6. Bogan, B. W.; Trbovic, V., Effect of Sequestration on PAH degradability with F enton's Reagent: Roles of Total Organic Carbon, Humin and Soil Porosity. Journal of Hazardous Materials 2003, B100, 285-300. 7. Bogan, B. W.; Trbovic, V.; Paterek, R. J ., Inclusion of Vegetable oils in Fenton's Chemistry for Remediation of PAH-Contaminated Soils. Chemosphere 2003, 50, 15-21. 8. Kawahara, F. K.; Davila, B.; Al-Abed, S. R.; Vesper, S. J.; Ireland, I. C.; Rock, 8., Polynuclear Aromatic Hydrocarbon (PAH) release from Soil during Treatment with Fenton's Reagent. Chemosphere 1995, 31, (9), 4131-4142. 9. Watts, R. J .; Stanton, P. C.; Howsawkeng, J.; Teel, A. L., Mineralization of a Sorbed Polycyclic Aromatic Hydrocarbon in Two Soils using Catalyzed Hydrogen Peroxide. Water Research 2002, 36, 4283-4292. 161 10. Saxe, J. K.; Allen, H. E.; Nicol, G. R., Fenton Oxidation of Polycyclic Aromatic Hydrocarbons after Surfactant-Enhanced Soil Washing. Environmental Engineering Science 2000, 17, 233-244. 11. Flotron, V.; Delteil, C.; Berrnond, A.; Camel, V., Remediation of Matrices Contaminated by Polycyclic Aromatic Hydrocarbons: Use of Fenton's Reagent. Polycyclic Aromatic Compounds 2003, 23, 353-376. 12. Martens, D. A.; T, F. j. W., Enhanced Degradation of Polycyclic Aromatic Hydrocarbons in Soil Treated with Advance Oxidative Process- Fenton's Reagent. Journal of Soil Contamination 1995, 4, (2), 1-14. 13. Nam, K.; Rodriguez, W.; Kukor, J. J ., Enhanced Degradation of Polycyclic Aromatic Hydrocarbons by Biodegradation Combined with a Modified Fenton Reaction. Chemosphere 2001, 45, (2001), 11-20. 14. Carvel, D. D.; Cartwright, R. T. In Measuring in situ Fenton's application success, Proceedings of the Third International Conference on Remediation of Chlorinated and Recalcitrant Compounds, Monterrey, CA, United States of America, May 20-23 2002, 2002; Press, B., Ed. Battelle Press: Monterrey, CA, United States of America, 2002; pp 1173-1178. 15. Benitez, F. J .; Beltran-Heredia, J .; Acero, J. L; Rubio, F. J ., Chemical Decomposition of 2,4,6-Thrichlorophenol by Ozone, Fenton's Reagent and UV Radiation. Industrial Engineering and Chemistry Research 1999, 38, (1999), 1341-1349. 16. Huang, H.-H.; Lu, M.-C.; Chen, J .-N., Catalytic Decomposition of Hydrogen Peroxide and 2-Chlorophenol with Iron Oxides. Water Research 2001, 35, (9), 2291- 2299. 17. Pignatello, J. J ., Dark and Photoassisted Fe3+ Catalyzed Degradation of Chlorophenoxy Herbicides by Hydrogen Peroxide. Environmental Science and Technology 1992, 26, (5), 944-951. 18. Sedlak, D. L.; Andren, A. W., Oxidation of Chlorobenzene with F enton's Reagent. Environmental Science and Technology 1991, 25, (4), 777-782. 19. T eel, A. L.; Warberg, C. R.; Atkinson, D. A.; Watts, R. J ., Comparison of Mineral and Soluble Iron Fenton's Catalysts for the Treatment of Trichloroethylene. Water Research 2001, 35, (4), 977-984. 162 20. Chen, C. T.; Tafuri, A. N.; Rahman, M.; F oerst, M. B., Chemical Oxidation Treatment of Petroleum Contaminated Soil Using Fenton's Reagent. Journal of Environmental Science and Health 1998, A33, (6), 987-1008. 21. Kong, S.-H.; Watts, R. J.; Choi, J .-H., Treatment of Petroleum-Contaminated Soils Using Iron Mineral Catalyzed Hydrogen Peroxide. Chemosphere 1998, 37, (8), 1473-1482. 22. Kostecki, P. T.; Calabrese, E. J ., Petroleum Contaminated Soils: Remediation Techniques, Environmental Fate, and Risk Assesment. 3rd ed.; Lewis Publishers, Inc.: Chelsea, MI, 1989; Vol. 1. 23. Beltran, F. J .; Gonzalez, M.; Rivas, F. J .; Alvarez, P., Fenton Reagent Advanced Oxidation of polynuclear Aromatic Hydrocarbons in Water. Water Air and Soil Pollution 1998, 105, (1998), 685-700. 24. Srivastava, V. J .; Kelley, R. L.; Paterek, J. R.; Hayes, T. D.; Nelson, G. L.; Golchin, J ., A Field Scale Demonstration of Novel Bioremediation Process for MGP Sites. Applied Biochemistry and Biotechnology 1994, 45, (6), 741-756. 25. Beltran, F. J .; Ovejeor, G.; Rivas, J ., Oxidation of Polynuclear Aromatic Hydrocarbons in Water.4. Ozone Combined with Hydrogen Peroxide. Industrial Engineering and Chemistry Research 1996, 35, (3), 891-898. 26. Lindsey, M. E.; Tarr, M. A., Quantitation of Hydroxyl radicals During Fenton Oxidation following a single Addition of Iron and Peroxide. Chemosphere 2000, 41 , (2000), 409-417. 27. Lindsey, M. E.; Tarr, M. A., Inhibited Hydroxyl radical Degradation of Aromatic Hydrocarbons in the presence of Dissolved Fulvic Acid. Water Research 2000, 34, (8), 2385-2389. 28. Lee, B.-D.; Nakai, S.; Hosomi, M., Application of Fenton Oxidation To Remediate Polycyclic Aromatic Hydrocarbons-Contaminated Soil. Journal of Chemical Engineering of Japan 2002, 35, (6), 582-586. 29. Walling, C., Fenton's Reagent Revisited. Accounts of Chemical Research 1974, 8, 125-131. 163 30. Walling, C., Intermediates in the Reactions of Fenton Type Reagents. Accounts of Chemical Research 1998, 31, (4), 155-157. 31. Walling, C.; Johnson, R. A., Fenton's Reagent V. Hydroxylation and Side-Chain Cleavage of Aromatics. Journal of the American Chemical Society 1974, 97, (2), 363- 367. 32. Augusti, R.; Dias, A. 0.; Rocha, L. L.; Lago, R. M., Kinetics and Mechanism of Benzene Derivative Degradation with F enton's Reagent in Aqueous Medium Studied by MIMS. Journal of Physical Chemistry 1998, 102, (52), 10723-10727. 33. Luster-Teasley, S. The Use of Gaseous Ozone to remediate Pyrene Contaminated Soils: A study of By-Product Production, Environmental effects on Remediation Efforts, and Scale-Up Volume I & II. Dissertation for the Degree of Ph. D., Michigan State University, East Lansing, MI, 2003. 34. Upham, B. L.; Masten, S. J .; Lockwood, B. R.; Trosko, J. E., Nongenotoxic Effects of Polycyclic Aromatic Hydrocarbons and Their Ozonation By—Products on Intercellular Communication of Rat Liver epithelial Cells. Fundamental and Applied Toxicology 1994, 23, 470-475. 35. Yao, J .-J .; Huang, Z.-H.; Masten, S. J ., The Ozonation of Pyrene: Pathway and Product Identification. Water Research 1998, 32, (10), 3001-3012. 36. Kakarla, P. K.; Andrews, T.; Greenberg, R. S.; Zervas, D. 8., Modified Fenton's Processes For Effective In-Situ Chemical Oxidation-Laboratory and Field Evaluation. Remediation 2002, Autumn, 23-36. 37. Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M., Environmental Organic Chemistry. lst ed.; Wilkey—Interscience, John Wiley & Sons Inc: New York, 1993. 38. Masten, S. J. In Use of In-Situ Ozonation For the Removal of VOCs and PAHs fiom Unsaturated Soils, Proceedings of the Symposium on Soil Venting, Houston, Texas, 1991; US. Environmental Protection Agency: Houston, Texas, 1991; pp 29-45. 39. Liang, Y. In Biodegradation of Pyrene in Soil Microcosms: Identification of a Toxic Intermediate, EWRI 2005: Impacts of Global Climate Change, Anchorage, Alaska, 2005; Environmental and Water Resources Institute: Anchorage, Alaska, 2005. 164 40. Chen, R.; Pignatello, J. J ., Role of Quinone Intermediates as Electron Shuttles in F enton and Photoassited Fenton Oxidations of Aromatic Compounds. Environmental Science and Technology 1997, 31, (8), 2399-2406. 41. Goel, R. K.; Flora, J. R. V.; Ferry, J ., Mechanisms for Naphthalene Removal during Electrolytic Aeration. Water Research 2003, 37, 891-901. 42. Zeng, K.; Hwang, H.-M.; Fu, P. P.; Yu, H., Identification of 1-Hydroxypyrene Photoproducts and Study of the Effect of Humic Substances on its Photolysis. Polycyclic Aromatic Compounds 2002, 22, 459-467. 43. Heitkamp, M. A.; Freeman, J. P.; Cemeglia, C. E., Naphthalene Biodegradation in Environmental Microcosms: Estimates of Degradation Rates and Characterization of Metabolites. Applied and Environmental Microbiology 1987, 53, (1), 129-136. 44. Heitkamp, M. A.; Freeman, J. P.; Miller, D. W.; Cemeglia, C. E., Pyrene Degradation by Mycobacterium sp.:Identification of Ring Oxidation and Ring Fission Products. Applied and Environmental Microbiology 1988, 54, (10), 2556-2565. 45. Luster-Teasley, S. L.; Yao, J. J .; Hemer, H. H.; Trosko, J. E.; Masten, S. J ., Ozonation of Chrysene: Evaluation of By—Product Mixtures and Identification of Toxic Constituent. Environmental Science and Technology 2002, 36, (5), 869-876. 46. Juhasz, A. L.; Naidu, R., Bioremediation of High Molecular Weight Polycyclic Aromatic Hydrocarbons: A Review of Microbial Degradation of Benzo[a]pyrene. International Biodeterioration & Biodegradation 2000, 45, 57-88. 47. Alexander, M., Biodegradation and Bioremediation. second ed.; Academic Press: New York, 1999. 48. Liang, Y.; Gardner, D. R.; Miller, C. D.; Chen, D.; Anderson, A.; Weimer, B. C.; Sims, R. C., Study of Biochemical Pathways and Enzymes Involved in Pyrene Degradation by Mycobacterium sp. Strain KMS. Applied and Environmental Microbiology 2006, 72, (12), 7821-7828. 49. Yao, J.-J.; Huang, Z.-H.; Masten, S. J ., Ozonation of Benz[a]anthracene: Pathway and Product Identification. Water Research 1998, 32, (11), 3235-3244. 165 50. Chiou, C. T.; McGroddy, S. E.; Kile, D. E., Partition Characteristics of Polycyclic Aromatic Hydrocarbons on Soils and Sediments. Environmental Science and Technology 1998, 32, (2), 264-269. 51. Hemer, H. A.; Trosko, J. E.; Masten, S. J ., The Epigenetic Toxicity of Pyrene and Related Ozonation Byproducts Containing an Aldehyde Functional Group. Environmental Science and Technology 2001, 35, (17), 3576-3583. 52. Bowers, A. R.; Gaddipati, P.; Eckenfelder Jr., W. W.; Monsen, R. M., Treatment of Toxic or Refractory Wastewaters with Hydrogen Peroxide. Water Science and Technology 1989, 21 , 477-486. 53. Satoh, A. Y.; Trosko, J. E.; Masten, S. J., Epigenc Toxicity of Hydroxylated Biphenyls and Hydroxylated Polychlorinated Biphenyls on Normal Rat Liver Epithelial Cells. Environmental Science and Technology 2003, 37, (12), 2727-2733. 54. Sedlak, D. L. The Abiotic Reactions of Polychlorintaed Biphenyls (PCBs). Doctoral Dissertation, University of Wisconsin, Madison, 1992. 55. Hauser, B.; Schrader, G.; Bahadir, M., Dependence of Genotoxicity of Benzo[a]pyrene Suspensions in MUtatox Test on Dissolved Concentration and S9 Addition. Ecotoxicology and Environmental Safety 1997, 38, (3), 224-226. 56. Lambert, M.; Kremer, S.; Anke, H., Antimicrobial, Phytotoxic, Nematicidal, Cytotoxic, and Mutagenic Activities of l-Hydroxypyrene, the Initial Metabolite in Pyrene Metabolism by the Basidiomycete Crinipellis Stipitaria. Bulletin of Environmental Contamination and Toxicology 1995, 55, (2), 251-257. 166 Chapter 6. Modeling Introduction The modeling of CHP systems is a difficult task because of the complexity of the chemical reactions that take place in the environment during remediation and the fact that advection and dispersion become important for incompletely mixed systems. Usually 5-9 Fenton systems are modeled using kinetic reactions“4 and completely mixed conditions are assumed. Unfortunately when remediation of soil and groundwater is performed, completely mixed systems are rarely if ever achieved. Kinetic modeling is required to determine the optimal catalyzer and oxidant ratios for field applications. This chapter will deal with building three models for the CHP systems tested. The first will be a completely mixed kinetic model, the second a diffusive-reaction model and finally an advection-dispersion model will be created for remediation of aromatics in porous media. Kinetic Model A kinetic model that describes the concentrations of multiple species during the reaction for each of the catalyzed hydrogen peroxide propagations used has been developed. This model was developed in order to verify the rate of hydrogen peroxide decomposition to oxygen in Fenton reactions and thus determine how much hydroxyl radicals are being produced from these reactions. 167 The key to modeling the behavior of Fenton’s Reagent reactions is to be able to accurately predict the rate of OH radical production. This prediction can be approximated by modeling the concentrations of hydrogen peroxide and catalyzer with time in the system, as well as knowing the possible hydroxyl radical sinks present. Using known catalyzed hydrogen peroxide reaction and rate constants from the papers by Walling lo; ll,12 Pignatello and Rivas 4; the concentrations of hydrogen peroxide, ferrous and ferric ions may be determined using the following equations for classic Fenton’s system: Fe2+ +H202 —-)Fe3+ +0H‘ +-0H k1=76 M‘s" (1) Fe3 + + H202 —> Fe2 + + H + + H02 - kg: 0.02 M‘s“ (2) H202 + -0H-—>H20+H02- k3=2.7x107M'ls'l (3) Fe2+ +-OH—-)Fe3+ +0H' k4=4.3x108M'1s'1 (4) Fe3+ +H02-—>Fe2+ +02 +H+ k5=1x104M"s" (5) Fe2 + + H02 ~+H+ -> Fe3 + + H202 k6=1.2 x106 M's" (6) H02. —) 02' + H + k7 = 8x105 M's“ (7) Fe3 + + 02— ——> Fe2 + + 02 k8 = 1.5x 108 M's" (8) Fe2 + + '02— + 2H+ —+ Fe3 + + 51202 k9 =1x107 M‘s" (9) 168 From these reactions we can write an equation for the concentration of hydrogen peroxide, ferrous ion, ferric ion, hydroxyl radical, superoxide radical, and perhydroxyl radical. If we add the hydrogen peroxide decomposition reaction we can also calculate oxygen production from the system. l H202—)H20+502 kd=? (10) Hydrogen Peroxide: dCHZOZ kC k C k C ——=- + + d, ( 1 F92+ 2 Fe3+ d) H202 (11) + (k6CH02- + k9 C 02-. )CFeZ + Ferrous ion: dC —F—92:——(k C +k C +k C )C dt 1 H202 6 H02- 9 02—. Fe2+ (12) +(k C +k C +k C )C 2 H202 5 H02- 8 0;. Fe3+ Ferric Ion: dC —Ee—3:—(k C +k C +k C )C ' _ 1 H 0 6 HO - 9 — 2+ ‘1’ 2 2 2 02 - Fe (13) —(k C +k C +k C )C Hydroxyl Radical chH k C C (k C k3C )C (14) = — + dt 1 1762+ H202 4 Fe3+ H202 0H Perhydroxyl Radical dCHO dt =sz Fe3+ C11202 -(k5CFe3+ +k6CFe2+ +k7)CH02 (15) 169 Superoxide dCOZ— = — k C k C - 16 dt k7CH02 ( 8 Fe3+ + 9 F82+)C02 ( ) Oxygen dC02 z - 17 dt deCHZOZ +k5CFe3+ C1102 +k5C 1783+ Co2 ( ) Similar equations may be written for the byproducts found during these reactions. The addition of sinks such as bicarbonate (HCO3') and carbonate (CO32') ions may be added to the model to create a more accurate model. Temperature The temperature of the reaction solution is not constant in CHP reactions because hydrogen peroxide decomposition to oxygen is an exothermic reaction, releasing 98 kJ/mole of heat. Since the reaction constants are temperature dependent, the rate constants in our model need to be corrected throughout the experiment. Temperature of the reactions can be calculated using the empirical equations obtained from regression analysis in Chapter 2. —=a'e' +c-e' (18) 170 171 _80 800 8:- m~00 08.0- 080.0 28.9 «.80 $3 8.0 00.0 ~33 _80 .83. 8.2- ~80 ~20- 88.0 380. 00-8.0 -3 8.0 :0 =2: _ Soto“. .80 0:0 «.80- 08.0 ~00? 88.0 008.? 08.0 :2 0~.0 #0 8.03: 8.0 8~0 -0.~- ~80 8a? 880 88.0- .80 ~03 30 2.0 0.8: 000 800 80.0- 080 0.8.0- 0080 -~0.0- :00 83 0~.0 8.0 0.0.: 000 0:0 002- 08.0 80.0- 88.0 0000.0- 200 83 80 ~00 23¢ 80 80.0 000.0- 08.0 80.0- ~800 800.0- 88.0 800 8.0 :0 «.3: ~00 80.0 000.0- 2.00 080- :80 .0080- ~80 ~80 8.0 ~00 ~35 00.0 08.0 840- 08.0 82- 88.0 28.0. 80 ~03 8.0 :0 :5: 3:9. RJ 6 _ 0 _ U _ 0 _ U _ 0 _ 6 _ a _ 0.0 _ NOE _ .~m 2: 98 on: soamohmou 05 .8% $220508 05 ~33 wee? 52m 20 02x83 sumac»: 98 no: mo 053055050 5202 .N SE05 5 Contact £83008 350—800 no: 0528 .08 2305950 383800: 05 co 803050 508830 05 com "7w asap. The value of K is then corrected by the equation: KT =K206T‘20 (19) Where KT is the kinetic constant for temperature, T K20 is the kinetic constant at 20 °C 9 is a constant that is estimated from the activation energy of the reaction and was found to be 1.058 T is the new temperature for the reaction in °C Modeling Iron Species and Concentration The speciation of iron in solution is very difficult to ascertain especially during Fenton Reagent reactions, where both ferric and ferrous ions catalyze the reactions and are the basis of the cycling process. Morell3 describes the speciation of iron as a basis of the hydrogen ion concentration. [Fe3+] =103.2[H+]3 (20) [FeOH2+]=101[H+]2 (21) [Fe(OH)‘2*] = 10‘2'5 [11+] (22) [Fe(OHflii=10—18'4[15i+]_1 (23) [Fe2(0H)§+1=103‘5[H+14 (24) 172 [Fe3 (0H)Z+] = 103'3[H+ )5 (25) Oxidation of Ferrous ion by Oxygen Ferrous ion oxidizes to ferric ions in the presence of oxygen. Since CHP reactions produce oxygen, this reaction may become important for modeling purposes. The equation for this oxidation is: ._. k fO[Fe2+] +k f1[FeOH+] +k fO[Fe2+] (26) k obs [ Fe 2+ ] Or —1 2 2 K1KW+k .3sz 1+K1Kw +flZKw IH+1 f 2 [17+]2 [19+] [H+]2 (27) kobs = ka +kf1 K1 and [32 are the formation constants of the mono and dihydroxy species of ferrous ion in water and kf is the rate constant for the reaction of oxygen with the different iron species in solution. The values for kf, kfo kn and kn are found in Table 2 The equation can also be written as: dC 2+ 3 ' __13:_ = —kf [Fe2+][02] = kfoiFe2+1[021 + k}1[Fe0H+1[021+ k}2[Fe<0H>21[021 (28) 173 Table 6—2: Rate Constants for Reaction of F e2+ Species with 02 from Morel 199313. Pseudo First Order Fez? Rate Constant (s' Second-Order gate Species 1)“ Constant (M s )*** Fe2+ 1.0 x 10‘8 7.9 x 10'6 Feon+ 3.2 x 10'2 25 Fe(OH); 1.0 x 104 7.9 x 10‘5 0 F62 +1. 6.3 x 10'3 5.0 * E Fe 2+ denotes iron on Goethite surface ** For a constant P02 =1 atm *** Derived from first order rate constants using a Henry’s constant of 10’ 2'9 M atm’l RMSE To determine kd for soluble iron reactions, the kinetic reactions described by equations 11-17 were solved using a stiff ordinary differential equation solver in Matlab (ODE23s), that solves a fourth order Runga-Kutta method, while optimizing the objective function: _ 2 RMSE—\[Z(CSl-m —C0bs) (63) where Csim is the concentration of hydrogen peroxide calculated by the model CobS is the concentration of hydrogen peroxide obtained from iodometric titrations shown in Chapter 3. 174 301 ll For optimization, the PsearchTool in Matlab running a genetic algorithm was used, with starting point 0.001 xlO'3 M s'1 and minimum and maximum boundaries of 1x 10'8 to 10 M s]. Results The results from the optimization of the kinetic model are in are shown in Figure 6-1 and in Figure 6-2. The model parameters obtained are then shown in Table 4. The optimization tool was run for both systems, with and without temperature correction. Figures 6-3 and 6-4 show the change in concentration and the production of the other species involved in the reaction: ferrous and ferric ion, OH radical, perhydroxyl radical, and superoxide. These were found not be substantially different between the different oxidant/catalyzer reactions tested. A comparison of the oxygen production predicted by the kinetic model and the data collected in the lab is shown in Figure 6-5. 175 Hydrogen Peroxide 0.12 l l l T l Temperature Corrected . - - - - Not Temperature Corrected 0.1 — ~ 0.06 _ q Molar Concentration 0.04 — 0.02 - 0 -500 0 500 1000 1500 2000 2500 Time (see) Figure 6-1: Plot showing the result from the optimization of the kinetic model described by equations 11-17 for an initial hydrogen peroxide concentration of 0.1 M and ferric ion concentration of 0.06 M.The lines represent the simulated concentrations obtained by the model and the A represent the hydrogen peroxide concentration obtained from iodometric finafion. 176 Hydrogen Peroxide 0.6 1 Ti 1 l l — Temperature Corrected 0 5 Not Temperature Corrected 0.3 _ 0.2 - Molar Concentration C e -01 L l l l J 1500 o 500 1000 1500 2000 2500 Time (sec) Figure 6—2: Plot showing the result from the optimization of the kinetic model described by equations 11-17 for an initial hydrogen peroxide concentration of 0.57 M and ferric ion concentration of 0.07 M. The lines represent the simulated concentrations obtained by the model and the A represent the hydrogen peroxide concentration obtained from iodometric titration. Table 6-3: Values for the hydrogen peroxide decomposition rate constant determined fiom the optimization of the objective function in equation 63 for reactions corrected and not corrected for temperature. I kd I I H202 Fe Temperature correcetd RMSE Not Temperuture corrected RMSE L 0.10 0.07 0.00194 0.014323126 0.00203 0.0136 I L 0.57 0.07 0.00841 0.081772679 0.9854 0.3199 | 177 Figure CHP r1: Intent Griffith Figure 6-3: Graph showing the change in concentration of ferrous and ferric ions during CHP reactions with initial hydrogen peroxide concentration of 0.1 M and ferric ion initial concentration of 0.07 M for reactions modeled with equations 11-17 with temperature correction. The gaph illustrates the cycling of ferric ion to ferrous ion and back. 178 Aoomv 9.0.... «or ~00 03 we vow sow . _ _ 4 _ NO 0 0209.00 229an8. 1| 1 1 no.0 1 1 00.0 1 \ 1 00.0 illlt _ _ _ l/ _ No.0 0E8“. 69$ 90.... 9 ~00 00F ~-o_. row row a 0 _ ill- 0 O 1 1 N m m 1 1 v m / m 1 1 e M 882.00 9398th '1 g _ _ _ _ 0 0:95“. or x [w] uouenueouoo 179 Figure 6—4: Graphs showing the simulated concentrations of hydroxyl radical, perhydroxyl radical, and superoxide formed during CHP reactions for initial hydrogen peroxide concentration of 0.1 M and initial ferric ion concentration of 0.07 M. The model is described by equations 11-17 with temperature correction. 180 80$ 05:. mow 09 0-9 9.9 .1111 1 o 1 .N m 9 w 1 L v w m - .0 M 889.00 2:098:00... i w eExQoasm 0-00 x l 00 Aoomv oEF 808 $6.... 1 or 09 0-9 079 now 09 0-9 079 4 a o 0 \1 o 1 . N m 1 md m e e w w - .0. .0. . . F .0. m 0 m 1 1 o .IWl 1 /\ 1 m.—. M 1 . m i . N .0059”. 306.353. $-00 x .0053”. $0662.. or x NF- 53111 11111 trace Figure 6-5: Comparison of the oxygen production calculated by the kinetic model and experimental data collected for hydrogen peroxide concentration of 0.57 M and ferric ion concentration of 0.07 M 182 “I .5! 80F CV mm 8 mm ON .12. or m o 0 _ ._ . W 1_ ,- 0 _ _ 2 _ _ _. _ _ , , M, m _ _ W ' 11 111% 1iifl ..... 1 1111 llllll iii 01111 1;-.11111 11Noo.o 7 A _ _ _ ._ i , 0 _ _ M. 2 2 2 _ _ H m . _ . 1 1111.41 ll_: 11- m- 1 ...il 111.11 11 111111 all .111¢ooo _ _ z w w _ .1 _ _ 2 _ V _ _ _ W _ v _ o 1 ,- 11_1- 111_ .111 ,1 1_1 1 1 1 111 11, 1|11L- W11- - . _ _ . 0 . _ _ .1 88 _ o _ _ 2 v 0 2 .3. .- + o r e o _ e o o o LeeOQOOOOOteooooooeoveoooeo _ _ _ w _ _ 1|1 .1011. 1L... 111 .v ......... a ...... 1+1 1111111 fill-11 11111806 _ _ m 5:08:8me _ fl W Enos. 9 W‘ W A A 111 11. fi 1| 11+-- 111 1- | 1 1 _ .-m111_.||- 0111. -l15.o _ _ M _ _ _ _ _ 0 2 m _ _ _ _ . _ _ b _ _ _ 0 _ 2 ~50 ueBAxo 10 com" 183 Discussion The optimization of the kinetic model yielded rate constants for the decomposition of hydrogen peroxide in ferric ion catalyzed reactions that appear to be dependent on the hydrogen peroxide concentration. For solutions with initial hydrogen peroxide concentration of 0.1 M, the rate constant was found to be 0.0194 M 3.1; while for an initial hydrogen peroxide concentration of 0.57 M the RMSE was found to be one order of magnitude higher, from 0.3199 to 0.0817. This can be explained by the fact that samples with 0.1M initial hydrogen peroxide concentration did not have a significant increase in the temperature of the solution (from 19 to 22 °C, Chapter 2) but samples with 0.6 M did result in a significant increase in temperature, i.e., from 19 to 30 °C. The value of the rate constant may vary because of the errors that accompany the measurement of hydrogen peroxide in these reactions and the effects that a drastic change in temperature would have on the reactions. Since titration requires time to terminate, the CHP may still be reacting during the titration, adding error to the measurement. Also, a change in temperature of the solution results in a change in sample concentration, due to the change in volume of a liquid and the kinetic energy of the ions, and may lead to additional errors. Still we believe the data collected and the optimization of the model is a good way of fine tuning our system. Figure 6-5 shows a comparison of the oxygen production predicted by the model and the data collected in the laboratory from gravimetric tests (Chapter 2). The models is under predicting the amount of oxygen produced, according to the data collected. There may be 184 several explanations for this: the data after 5 minutes of reaction time may be the result of a supersaturated solution releasing oxygen in order to equilibrate with the environment. Another explanation is that the model requires the incorporation of the oxygen mass transfer coefficient and therefore, the model should be calibrated to the oxygen production. From Figure 6-3 we can see that the ferrous ion concentration in solution during the reactions is insignificant, when compared to the ferric ion concentration (5 orders of magnitude less), so we can conclude that most Fenton Reagent reactions are ferric catalyzed and not ferrous catalyzed. This condition may show that ferrous ion concentration is the limiting factor in hydroxyl radical production and the recycling is of ferric ion to ferrous ion is the rate limiting step of the process. This issue becomes even more important when high hydrogen peroxide concentrations are used because if we include the oxidation of ferrous ion by molecular oxygen, as shown in equations 26 thru 28, the ferrous ion concentration will decrease even more. More robust analysis is being performed on the data presented in this chapter to determine if this is indeed correct, the results will be published in a future paper along with the data already shown 185 Conclusions A kinetic model was constructed for the CHP reactions tested, in which temperature was used to correct the rate constants for the reaction. Values for the hydrogen peroxide decomposition rate constant were determined using the data from iodometric titration of hydrogen peroxide in Chapter 3. Although the values varied for different hydrogen peroxide concentrations, the order of magnitude remained constant at 10‘3 min'lfor all samples. These models help create a more comprehensive representation of catalyzed hydrogen peroxide behavior under different environmental conditions, and should allow us to estimate concentrations of the major components of this system during remediation. The present model can be expanded to encompass the concentration of the target contaminant and to incorporate the role of oxygen in the oxidation of the ferrous ion. 186 References 1. Gallard, H.; De Laat, J ., Kinetic Modeling of Fe(III)/l-1202 Oxidation Reactions in Dilute Aqueous Solution Using Atrazine as Model Organic Compound. Water Research 2000, 34, (12), 3107-3116. 2. Beltran, F. J .; Gonzalez, M.; Rivas, F. J .; Alvarez, P., F enton Reagent Advanced Oxidation of polynuclear Aromatic Hydrocarbons in Water. Water Air and Soil Pollution 1998, 105, (1998), 685-700. 3. Beltran, F. J .; Ovejeor, G.; Rivas, J ., Oxidation of Polynuclear Aromatic Hydrocarbons in Water.4. Ozone Combined with Hydrogen Peroxide. Industrial Engineering and Chemistry Research 1996, 35, (3), 891-898. 4. Rivas, F. J .; Beltran, F. J .; Frades, J .; Buxeda, P., Oxidation of p-Hydroxybenzoic Acid by Fenton's Reagent. Water Resources 2001, 35, (2), 387-396. 5. Augusti, R.; Dias, A. O.; Rocha, L. L.; Lago, R. M., Kinetics and Mechanism of Benzene Derivative Degradation with Fenton's Reagent in Aqueous Medium Studied by MIMS. Journal of Physical Chemistry 1998, 102, (52), 10723-10727. 6. Duesterberg, C. K.; Cooper, W. J .; Waite, T. D., Fenton-Mediated Oxidation in the Presence and Absence of Oxygen. Environ. Sci. Techno]. 2005, 39, (13), 5052-5058. 7. Duesterberg, C. K.; Waite, T. D., Process Optimization of Fenton Oxidation Using Kinetic Modeling. Environ. Sci. Techno]. 2006, 40, (13), 4189-4195. 8. Duesterberg, C. K.; Waite, T. D., Kinetic Modeling of the Oxidation of p- Hydroxybenzoic Acid by F enton's Reagent: Implications of the Role of Quinones in the Redox Cycling of Iron. Environ. Sci. T echno]. 2007, 41 , (1 1), 4103-4110. 9. Godala, M.; Nowicki, L., Determining the Kinetics of a Chemical Reaction from Calorimetric Measurements and the Amount of Gaseous Products. Inzynieria Chemiczna I Procesowa (Chemical Engineering and Process) 2005, 26, (2), 363-376. 10. Walling, C., Fenton's Reagent Revisited. Accounts of Chemical Research 1974, 8, 125-131. 187 11. Pignatello, J. J ., Dark and Photoassisted F e3+ Catalyzed Degradation of Chlorophenoxy Herbicides by Hydrogen Peroxide. Environmental Science and Technology 1992, 26, (5), 944-951. 12. Pignatello, J. J .; Baehr, K., Ferric Complexes as Catalysts for "Fenton" Degradation of 2,4-D and Metolachlor in Soil. Journal of Environmental Quality 1994, 23, (March-April), 365-370. 13. Morel, F. M.; Hering, J. G., Principles and Applications of Aquatic Chemistry. John Wiley & Sons, Inc.: New York, 1993; p 588. 188 Chapter 7. Conclusions F enton’s Reagent systems or catalyzed hydrogen peroxide propagations are a complex set of reactions of which have been extensively studied. Still there are many unanswered questions. The complexity of the system leads most users to approaching it as a black box. The studies described in this work will add some clarity to the use of this reagent, especially where high end hydrogen peroxide concentrations are used or where oxygen production may become a hindrance. In this study we have found that the decomposition of hydrogen peroxide to oxygen is an important oxidant sink during CHP reactions, especially when high concentrations of hydrogen peroxide are used. This reaction should be incorporated into models to account for hydrogen peroxide loss. The empirical relations determined in this study can be used for predicting the decomposition rate of hydrogen peroxide or the formation rate of oxygen in an in-situ environment. There may be a limit to the maximum oxidant and catalyzer concentrations that can be used; when high concentrations of hydrogen peroxide (28% or 2.5x10-3 M) and soluble ferrous ion (0.5 M) meet, the reactions become diffusion limited. 189 The rate of oxygen production was found to be independent of the concentration and oxidative state of soluble iron catalyzer when no chelating agents are added; the relation appears to be first order with respect to hydrogen peroxide concentration. The temperature of the reaction may be estimated using the empirical equations derived in this study. There are many variables that may have impact on the reactions. The use of a complexing agent such as EDTA would help minimize the loss of hydrogen peroxide through its decomposition to oxygen when soluble iron catalysts are used. Completely saturated Ottawa sand does not affect oxygen production in soluble iron catalyzed reactions. The increase in solution temperature leads to an increase in the rate of reactions and thus increases the rate at which oxygen decomposes. A change in pH of the solution during reaction may lead to the loss of catalyzer due to the formation of insoluble salts. Goethite catalyzed systems appear to have two rates of oxygen production; an initial rate which occurs during the first hour of reaction and a secondary linear rate which was found to be constant. Oxygen production during the first 24 hours of reaction appears to inhibit hydroxyl radical production during that time span. 190 Hydroxyl radicals can be measured with SA as a radical probe for mineral iron catalyzed reactions. In goethite systems reactions at pH 7 were less efficient at removing SA than reactions carried out at pH 3. The ionization of the SA at pH 3 appears to have some effect on the hydroxyl radical reactions. Byproducts are formed during CHP reactions for both pH 3 and pH 7 experiments. Six different byproducts were identified using LC and GC/MS analysis, some of which are more oxidized forms than those found in the literature. One of the byproducts found 2,3 DHBA is a known byproduct of the reaction. In order to make a more accurate determination of the actual amount of hydroxyl radicals produced we need to determine the concentration of at least one of these products. We found that PAHs can be removed from soil using catalyzed hydrogen peroxide reactions. Pyrene was removed from several soil matrices with the removal efficiency decreasing with increasing matrix complexity. Reactions in Ottawa sand achieved the highest level of removal, followed by reactions in soil containing organic matter. Temperature became increasingly important in the system. The temperature of the matrix may increase substantially during reaction, and may cause an acceleration of all the reactions involved, including hydrogen peroxide decomposition. 191 The remediation of PAHs using CHP systems in both aqueous solutions and soil Slurries was found to produce byproducts. Eleven byproducts have been identified from the chemical oxidation of pyrene in these systems. The byproducts found are consistent with hydroxyl radical interaction with aromatics and are similar to those found in bioremediation and ozonation of pyrene under similar conditions. Three different catalyzers were used for the CHP reactions in the two matrices. Byproducts were found in all of the systems, therefore we may conclude that byproducts are produced whenever CHPs are used for remediation of PAHs. Very little is known about these byproducts after they are formed, because of this we investigated the fate of these byproducts during and after remediation. Parameter estimation software was used to estimate the chemical properties of the byproducts found. As a general trend Kow decreased for all byproducts when compared to pyrene, leading us to conclude that the byproducts are more mobile than the parent compound; aqueous solubility some cases increases by 3 orders of magnitude. This increase in mobility is a potential health risk as some of the byproducts found are known to be toxic or to interrupt GJIC communication. 192 Incorporating all this information into a useful tool lead us to create a kinetic model for the CHP reactions tested, in which temperature data collected was used to correct the rate constants for the reaction. Values for the rate constant of hydrogen peroxide decomposition in soluble iron systems were determined using the data from iodometric titration of hydrogen peroxide in Chapter 3. This model helps create a more comprehensive representation of catalyzed hydrogen peroxide behavior under different environmental conditions, and should allow us to estimate concentrations of the major components of this system during remediation. The next step is to expand the model to encompass the concentration of the target contaminant and to incorporate the role of oxygen in the oxidation of the ferrous ion. 193 Appendix A 194 Appendix A: Pyrene Oxidation Byproducts LC/MS DATA: Sample of LC/MS chromatogram of m/z 233 vs. TIC from Pyrene oxidation using CHPs ........................................................................................ 197 Sample of LC/MS chromatogram of m/z 260 vs. TIC from Pyrene oxidation using CHPs ........................................................................................ 198 Sample of LC/MS chromatogram of m/z 217 vs. TIC from Pyrene oxidation using CHPs ....................................................................................... 199 Sample of LC/MS chromatogram of m/z 265 vs. TIC from Pyrene oxidation using CHPs ....................................................................................... 200 Sample of LC/MS chromatogram of m/z 205 vs. TIC from Pyrene oxidation using CHPs ....................................................................................... 201 Sample of LC/MS chromatogram of m/z 204 vs. TIC fi'om Pyrene oxidation using CHPs ....................................................................................... 202 Sample of LC/MS mass spectrum for peak at 27.91 minute retention time from soluble iron hexane extract .............................................................. 203 Sample of LC/MS mass spectrum for peak at 13.17 minute retention time from soluble iron hexane extract .............................................................. 204 Sample of LC/MS mass spectrum for peak at 1.58 minute retention time from soluble iron hexane extract .............................................................. 205 Sample of LC/MS mass spectrum for peak at 1.57 minute retention time from soluble iron acetonitrile extract206 Sample of LC/MS mass spectrum for peak at 1.81 minute retention time from soluble iron acetonitrile extract ......................................................... 207 Sample of LC/MS mass spectrum for peak at 22.15 minute retention time from goethite catalyzed hexane extract ...................................................... 208 Sample of LC/MS mass spectrum for peak at 31.23 minute retention time from goethite catalyzed hexane extract ...................................................... 209 195 MS/MS DATA: Sample of MS/MS spectrum for m/z 218 from soluble iron catalyzed systems..210 Sample of MS/MS spectrum for m/z 235 from soluble iron catalyzed systems..211 Sample of MS/MS spectrum for m/z 267 from soluble iron catalyzed systems..212 Sample of MS/MS spectrum for m/z 247 from soluble iron catalyzed systems..213 Sample of MS/MS spectrum for m/z 204 from mineral iron catalyzed systems.214 Sample of MS/MS spectrum for m/z 221 from mineral iron catalyzed systems 215 Sample of MS/MS spectrum for m/z 260 from mineral iron catalyzed systems.216 Sample of MS/MS spectrum for m/z 223 from mineral iron catalyzed systems.217 Sample of MS/MS spectrum for m/z 219 from mineral iron catalyzed systems.218 196 o 00. 00 00.00 00.00 00.8. 00.04 00.00 00.00 00.00 00.0~ 00.8 00.8 00.0 EFF . . . 1 . 1 1 1 F 0 1 F 1 i . . F 1 1 . v . . 0 . F . t 0 . F . . 1 r 1 1 _ 1 . vi 11p 0 ..F .0... 000.00.00.00... .0; 000.0; 0, . .0 FF. .0 ..0, .__; 0.1.... F. ..F. FFF; . .__,01..._ FFLFFF FF. .. F..F.;F.FF... .; F. .00.... F: ... F F. . F... .0.0...;.;;.. :FFFFF... ..F... 07.70.}... 0.3.... ,. 0:0 , FF: FF. . FF, 0.;... ..FF.F........F....FF0_.._;_.F;.01;FF FFFIFMFF}... FF: ..FFF...FF.F..F.F... .. .F;;_ F;; % 030 F... F. . 8.3. ..F 30. 00 00 F.F F0 8 F0 8 L. 8 F. F 88. 0.10.8. 8.00 8cm 8an 2. 3.8.3 3. v.0. F om on F; F F mm. mv F 80.00 F;_ FFFF; 0...... F004 F. :00 0; FFFFF 00 04 FF;;;. 8. 04 ;; 04 04 80.04 F. 00.0.. _ o F. mm 0 ~.m 108 +00 0000 8 80 800:0 00.00 00.00 00.00 00. 8. 00.04 8.00 00.00 00.00 00.8 00.8 00.8 00.0 1 .1 11.41 111 l . .. 4 1 11.1. .1 44 .114. 1 . l 41 11. 1 114.14%4WF1QM 911.111 %;11$$L.fi.fim4jnmuwlpséb FFJE£44L~§J¢E Fz«<.1;. I J 1. 4.41 1.1111 1.1114 . 11.1 . .1 F1 1.: . 4 F 4.1. < r . 1.1 4.. 1.4 1.41.... O N . 00.00 E 8 F F 04 8.0”; a 8 F ; ; . ; F. - ..x... 00.8 F 008.0 0 mmm Shaw 1 GOP +00 0000 8 .00 00086 Snob com on xmxéu 9% can—532.50 197 o 8.8 8.00 8.00 8.3 8.9. 8.8 8.8 8.0m 8.8 8.0— 8.9 8.0 EFF . F . 1 1 1 1 1 _ 1 . 1 . 1 1 1 . 0 . 0 . F 1 TL 0 o F .0... 0...... 0......00.....0000..0.... 0} .000... 0 . ; _ . . . 0. 0.0. . . .0. 0 0.0.00... .. . ; .00.... ;.F....F00.0.0.... F... F . ... 0 . {F .00....0 00...: ..F. 21... ..F}. 0 F :00. .00.. 00.0 F: 2....wa 50);... FF 0.0.: , 0 0... 0.0 .0... 0.0.0... 0. F000; .00. ......F..._. .0 ..F... .0. ;. 0 00.00 F. ._.0 . .0; F0... 0; ..0. f1. F00 .9 F003,... .00F 2...... 1.0;. . 1 o/ . . . v 00.0.0 00R _. 00.00 ..F F F 00.3 00. 00. 00.00 t. 000.. ..F . . 0 0. . 0000 F 0 F...;.0,;B.00. .F 2.00; F . 00.00. _.0.;.0 0000. 00.00.-._ 00.0.. rm.mv . m . “~2me mm? 1000 +mm cmow ”r So @009 F0 8.0m 006m oo.om 00.90 00.9» oohm oodm oodw oodN 00.0.: 00.9 006 1 0 . . . 1 10 . .0 1 101 14 1 4.141. 11.1.41‘0414‘”? a“ ‘0 £ 1 . . .1 . . . w 1. . 1 10 1 . . F b 1 o 0...}. .00.... F .. F . ; .F F F F F, %I F. F. ,; F F 0000.0 . 0 CON hvdm 8.. +ww cmom HF 0001000005 00000.. com 00 032.0“. E 080003003 198 oodN 00.9 00.3 oodw 00.: ooumw 00.3. oodw God? 00.: GEE. Tr » . . . _ . r p + . y ‘0 _ o . ... .. .. 0/ 5 a.» .a ”...... _ .... .. .p ... 3.... ,......_.V.....,.. .......__.._.__.................... . .... ... .. . . .. .. w _ _ _ ......V... ...,...... .__ ._. . . 220m.m . p u.... ....» _.,......,.......... 7...... ._ 0mm mm” m». _ ...... a ...)...v .... 0.... ...._.__T _._... .....H ...; .... .. V... ...... d. ON OF o¢.m : .. . . .__ V .__ ..._ ......L. <.: ._ a: . __ 3.2 a .. . _ ) .... ._...V J ..._. .j .m , 1. _. 1.... ... ...... ..... .... ... . ...............; _ .7 42...... f .. ._V. ......... 4. .. ......» ...2. .... ..._ . ...,.. .__ . . . . Na OF "N OF ...... .... .... ......V... V........_... .... ..V. . m V... ..., . .. .__... ...V . ...)...3. . ........ ..... :.. ...... ..... _._.__,_.. :... .... .... . .... ...... ..V... V. .. .. .. ._V. ... .._. ..NMWta... mwNwmm Nw Mr Nw 2.0m. . . OWE... ......_....V..,........._ ...... .. ... ...__ ...” ...V. ....V.. 18.9 ...... _ .... w 9 hwwwp.22:.3 ._. ms 2. .. .... 50.n— ” . .. ._ .. . n . . . _ . . . . 88¢ 8 8 8.2 v2: ... 85 a :2.: 8.99. 2 E E e 8.3 3 n. 0:. +mm cmom ”F oodN oodw r r . . V A. .. .7 V ...V... _ -. .... r . ..z . V.., .. . > — I. u. .- . _ r A P ‘ . a .__. _. . ... 3dr ”mm—V ON.m_. .3me :N +wm cmow ; _ 90.9. V ...4..«....u . ..r .x J...... ) y. :.. ... . .... ..H J... .... . .. .... . v3: 3 2 .3” t 8.5? 8.: 8.9 .»4V.»>. ...».m V. Iva. .. < . . ‘1 V .....l L V... ... fl: \ . . .. «.... FWD. mm 2 mm. m: 5.38.... roow 818820 00.“: code _ 00%: u . .. .. A A L » . . 0 . . 0 p .0 . 0 . . > 0 . y _ . 0 . p r _ . _ . o 1...}... . ‘ 3 3.. Av" {r2076}! ( n . 0‘1§.r.l«p.r; .. I..)J..i..>< (1 3...! ... 05V :..: 13,?355. 9?... V f 0...}. 0 . 00.0.5014 _ ... J 4 0 . . ‘ 0. 0 0.0.0 ..........00.. 0. . I.“ 7 P... ‘ r . _ . .. ‘ 0/0 #0:; a}... 0... 0.00.0.0 .00.... 0.0.... _ . .00. 0...: ...... 0.000.. .. .._ _ . V 2.00. $203.10.. 0000....)r0000: ... 0.4000. .._. 00. 00 0o 00 3 0. :..0. .. on. 00 ..F; . . mm. m _ .._ .. .._. 0000.0 .2 o: 00..]. 80 .mm cmow nN N00 0000.0. 00 00.00 00.00 . 00.00 00.0v 00.000 00.00 00.00 . 00. 0N... 00.0N 0.0.0F 00.00. 00.0 . 91.—0 nVal..w. . l.. . . .. . . .4 .. 0. ...1». -. ...4. .....f..,..}.4.,.4. d. I4 V lit—3 I. J u 1.0 ..., 141- 0.00-...). ..a09‘4413 . To . 0.0.00 . 00.00030 000.0 3.00 3].: 000.000030000005 3.8.3002 00: ..000030 00.0 0 . - %I . 0 ._ 0 0m0N0 03 00.3 1 r80 .00 :80 UN N00 00000 F0 3020 .80 8 220.9. «00 083008.08 200 2.8.8 8.8 8.8 8.9. 8.8 8.8 8.8 8.mN 8.8 8.9 8.2 8..., _IP L £ _ p y p > L p . . y » p . p r L r » > » L 5 y L . p r r r w > r . N LI. r L . . > L .31.; ‘..._L_._¢1..?:} .L I 5.: x ‘- ....- «1.5L: .._. ....L... :... 4.3.9.... ...}... s. 3...... . . L . ......1. «13;: . .... . ......L. L . .. .‘L‘: L. R o m¢‘. : V..} .\ .._-Li N.. on ... .....LLL. L... 3...... L. .L L..V.LL.L.. V LL)... ......y...‘ 55...... ... 5...... B N... 3.8 :3 .... .LL... L ...... .....L._....L... . o... a 3. NN . . 2 mm NNL. 5 . 8. ..N L. .... .... ..L .33.... L... 8 8 E NNL . B 3 1...}... .. ...xL L V «N..... L Nm 8 8.3.5841 3. N.. .... .LL . L 3N... N. No no. 3.2 8 Na 8 L mommN Nm 9. NEN 8 o: ..N. 88 u r F +3 cmow ”L .8 8285 code oo.mm oodm 00.9 oodv oo.mm oodm 00. mm 00.0N 09m— oodw oo.m K 1 r 1 . . v ..11. 1.1: 1.. V y . :44”! .1Liarstfl.‘7....1..s.i(.a.T....l$.L-Hew....r.? ::va‘j"?.¥fi€Jy .T....¢.\_ .1451. . J... . 5... L...3Jm$y_.m/4.1‘.(ltdr Q..\\fi£§ $1 .....345. . . 34.1 u . q 1 .41L131cnr1xr- t..._1A..... .o on. 3 L 8 3 . . on. «N Mm. m. _ L. L.__ __ L. 9.8 V _ L L V 88 L .._ . .. L ..V V L 3.2 L , % R8 SNN. mmNN 88... . _ 8N N..,NN +mm cwow up :..: 50-885 .82.. tom 8 30.5.... 2 canamzotuo 201 00.0w 8.0m oodm oodv 00.9 09mm 8.8 09mm 00.90. 8.3 00.9 oo.m ”Err r? . y, , » “py‘ » ‘p _ L»+L_ » O k :2.. {y if N. {3.13 _. {if J.;.Cu‘x.,..._\._4:i§¢. 4212.... ...e 51? .. #194,. . ... .i (:1 £ r? éfrxzxeiwim 2.2. 5.2 $9 3!. ,. r .. , f- J 3: ... . ..., . n _ n . F .V NV.NN . a A. v5.31... .._...r(¢.r_d _ r13: ck k3,. .: 76%.. 1a.?) «pt‘airitwffi...ekifiapfif L . . . OV.@N ‘ 1:24 gr .. ,_. . . J ... .__7 J . x_ 2mm . a ........i E on 8:. . «0% e . 2 on .on . . 8 z _ 31.3 «who 8.3 3. ow 5.39 9. mm 3 mm mm _ g y _ o/ I _ 7 ¥ . __ % _.. W. 93 wwwvd 0: Sb ‘8. +mm cmom “w moo @9840 8.8 8.3 8.0m 00.3 00.8 006m 8.0m oohm 8.3 8.3 9:: 00.0 R . n :::.{2 \ «a .It.1.}»l .yl)» . , _ . r F r L F » \»‘...\rl\r\.w\i\...l‘ . y . _ o t alflaukj {\ffl.5(.l£1.)\!. ,7 . figs‘.1#jlu.f1tj1 H}.{t ‘5...»ka 4* A _. . I . . «Nam «can H.; .‘ madam on R _; I 8.8 J. : 5.5 m. a : %. % A _ 33¢ vow mm; roow +wm cmom ”F 890285 68: Ba 8 30.22 a. =8=m3o=8 202 8v 2:. 8m 8m 8m own 8m 8m 8m ova 8— o? o: o9 2: 8 8 ~\E . L , .. .7." , L ; 4.411. , > _, } 179:4. » 4 o as? 239. 39.8 338 mmmmgg 2m 830387 2N , . _ 7 «.8 SN «8 _ new .82 8:5 E 8. 859 s. 8.78.83; 2 85 7 , 7 . 7 _ 7 Sword 7 8’7 km , _ 7 5: «mm _ 7 8m 7 , 7 7 7 7 77 7 % v 7 7 m8 7 7 7 r mom: a: u 8. 1mm 80m g 76m:3.8Ifiormvegmormmo: so 38.5 mm? Bo 88:0 .82.. no: 8 55.9. E coaamzotuo 203 8m 8m 8m 08 Sm 08 8F 2: . o: of 9.: oi 7 8’ 8P o: 2: 8 8 8 8 o N‘E .. . ..7. .. .7» . .. 7. .7LL p>r4.7. .7....‘ 7 ..7: 77 7. 7...“. .7 . 7.7.17.._.7rL.fi.7 4.. y .7. .7777 7 ..7 ..770.7L.».} a. .7 _ 11 . 7 4 4 838 87 888838 :«E New 3,: z: 8. 2.. 770:. 8; 77877 «737 $87 _7 ....7 m7: 7 7: 77 7 8 8 78 87 77 S7 787 8 777 7 _. _ 7 . k: 7 _ 7 7 7 _ 7 7 7.8: 87 m3 7 mm? 777 7_8— 7 82 77 7 7777 7 887 7 E. 7 7 7 _ «m. 7859. 7 7 7 S 8 7 77 _ 7 7 7 7 7 7 77 8F .. .7 7 7 _ 7 _ 7 7 7 a 7 7 7 7 7 7 8m 7 7 7 7 7 7 7 7 7 7 ND? _ 7 7 7 7 7 8 7 7 _ 7 7 7 7 7 7 7 7 7 7 8 7 7 7 7 7 7 7 7 _ 7 7 7 7 7 7 7 7 7 7 _ 7 . 7 7 7 7 7 7 7 . __ 7 7 77 7 7 7 7 7 7 77 7 7 7 7 V 0/0 7 7 7 7 7 7 7 7 mm; 7 7 7 7 7 _ 7 v. 7 7 7 7 2N 7 7 7 7 7 7 7 7 7 7 7 E 7 7 88s my 8. u 58 +8 88 ”F 77«888537888887 so 82.27 .8 So 8820 782. .8: 8 51.9. E 7.853278 204 00v ovv owv oov own own ova can con omN _r. h...7 .7. own ovw CNN cow om? omp ovp oww 00? cm on N): l.1L - .7. 7 7.....7. 7r. 4.. 1.7 77:. 7th.. rrr7. L.¥ ....Lfi 4171 mi... .m17.+...L rrl: A71 7 -.. . 7 .. - C 5v _ _ P. 87 m5 7 may _ 8N . 7.8m BN7 7 m2. 5 5n: . _ 3 _ 8 :.. 7 m9. «9 :3 .8 087 8 978 @884“ 888 7 78 8mm 7.57 _mmm 8m mg «a BF 2: 8F 8. . «$7877 8 2 _ 8 we _ 7 7 7 . 7 RF 7 _ Rpm 7 7 7 7 7 mvm 7 7 7 7 7 . 7 7 m2 7 __ 7 _ 7 77 7 7 7 . 7 mm 7 7 7 . 8 7 .M% 7 7 7 7 7 7 7 7 7 7 7 7 mom: a: n 8. .8 778.0, ”N 3878.888an :5 $8.: 8 8 88,7 5 38¢ to”. on xuzé... _.fi czamgtuo 205 can own can con 8m 8m 8m 8m 8m 2: o9 03 8F 2: 8 8 NxELL. 7 . -TT. . . ...7. -.r-.l.r- .7..1.«. % 7+ -. _.7 7 ...777L»7...7. .Ji. . .k; .. .7.4.4.l1 7 . .JerLX»--. r-o E” 688 8m .8 .8 m8 m8 m8 8 8~ 8~E~ 8m 8m 8m 8w 8 man BN7 8. 8;: at 829m! 5 .2 o: E 8. 8 8 F E 8 8788 8N 7 7 7 7 7 7 . 7 7 7 7 7 - % v8 7 7 7 7 7 7 7 7 7 7 889 cm .8? +3 88 ; 78:78:33-8ng so at“: 8 m8 885 .82.. .81 8 30.23 :7 7.8.7.2380 206 own can own own com com com ovm CNN oow om? cow P LL¥7V .> ovw omw ooF ow om NE . [F 7 {7 yrl. p43 .. r7 y s H . . _L7JI.># ” r r 4 8 8" In 888 8m m8 an 8m 8m 88 888 88m 888 m8 @8218 SN 27: «2 o? 8* 8... . 8N. vow cop th homo; +mm cmow ”— s.._.F7 . - LL>LI.|v.-7. O 2:887 8 a 7. 7‘ 7 7. 77.7 _ .. 7«4 7 A 359 8F c: 7 a 88 7 7 7 m: 7 7 7 7 h we 2: 77878887887833 so 728. c 8 «8 88:0 88: com 8 293". «u czamzotuo «j 207 ow... 00¢ 03‘ ON». 09. can own ova own com owN . 0mm ovw CNN . cow . om. . om. o3 om. 00. on . om Unfinil .va..mv$.. 37.7 78...”... .157 .__. . 7M8?” 7 777.17 7777.77 7.7.7 .14; ..77 1.17 ._77. 7 77 .77 .77 7 .77. 7 77 7 7. J7 o 8...: n3 .9. «~83 8.. E... . @868 www.mm 7777 77 88 788277 E 777 7 8. 77 MW.._7 .77 7. . 8 n8 8.; .n 0877 .63 _ _ 78 7 .7 _ . . 7 7 m8. _ 7 . 8. 78. _ . 38 8 8m 8m .8 , 7 8. 7 mm. . _ 8... 78... 7 7 .77_$_ 7. 7 . 8_ 8 _ .v. _ _ 7 _ VNN arm 7 raw 3? . mm ms 77 8m 9. 8. . E7 8 % 7 ._ 8 88+ 8N I 8. +87 88 7. 28988187 5.3888 so .8..va m8 .8 8.8.0 385 tom nu mOmuuxw: at can—530.30 208 own 05 com 8m 08 EN 8m 8a SN 98 RN o.~ 8m 8. 8. E. 8. 8. 8. 8. 8. o.. 8. 8 8 o. 8 NxE 7 7 7 7; 7 _ T . 7 7t. . I._7 FL? 7: .7 . L 77 71.7. _.IL, 7 477 7.7 7 7 O .8 ..m 388 28 «8:78 238 n8. «8.8 828 ”8 ma EN 8m 8. 8. 2.7L. 5. .o. .m. .38. 8.8.. m: 8. B88 88 .o 28 8 E «8 7 % 7 E8. .8 .. 8. +87 88 7. 2827.87.£98878..78..7 £088.57 8.. .8 8.8.0 .82.. com 8 mooim: 27 :stzotuo 209 o8 8m 08 o8 8. 8. 8. 8. 8. 8. 8. 8. o.. 8. 8 8 8 8 8 N\E .7. .7 7L“? . P1747 8‘47Wfl7wl7r 7.f»fi. 7 ¥7J .7. 7 7.0 . 7 ...LF. 77r «8 47 7 77 71....7 ‘ 4. 77 7 7 o 88 ~88: o8 887778.377 8.8128. 8.8. 8. .m. 8. ....ov. 8.8. 8.. 8. ... 8.5. 88.88 8 2 R 8 888 88 8 .8 788 ___ 8. 8.- 8. 7 77 3. _ 7 _ _7 7 7 SN 7 7. 8N % ~87 ..N oo. 8.88 .8 . 888 B 22880 8. 7:. .8%..5887 so $8.87 .8 80.88.85 . 7... 7 .77....77 L 77777 7. .+.h ... 7.. 7 7.77>777..7 77.7 77 LP». .7.». .r7..77 7 7T7L7 7: . . 7. 4 .7 7 7.. 4 7 1 7 . O 828 N87 7 SN .8 . . 5.77 8. .8. 8. 8. 8.8.8.58. 8. o... 8. 8.8. :..... 3. 8.8. 8 8 8 5 88 R o. 88 88 8 7m.N . 7 do. was . 7 cm. .88 7 7 _ 7 %l 77.77.. 7 7 7 ..N .8 8. «88 $88 .0 92880 ”m. NON 2.88.+.m.787-8.§..7 .5 88.87 R. 88688.0 mam: 88: Ba 8 “3:6". :7 cflamwotuo 210 omN OVN omN LL CNN oFN CON cm. car on. owe omw oer omr ONw o: cow 0a cm on 00 on N\E . .7 7 7 r7 77 777 7 7 7. 7 7 _ 7 7 7 1 7.7 1777. 7 . 77 777 . 7777 _ .77777777 7 7 7 . 7 O 877. 77 _ 77 _ 7 7 77 8. 8. 8. a... .38. «N. 8. 2.9.8. 8 88 8.» .8. m. 88 8 _ 8N . 77. n. 8.8 __vnm . 7 7 7 mm. .88 7 _ 7_ QFN 7 7 77 va _ _7 ..N 88 7 .8. 7 7 7 _ 78. 8. 7 7 3w 7 % 7 7 NON _ 7 7 7 7 7 7 $8.». 2. 8. 88mm 70 22.788 ”N 28.8818. .H.mm7-.t.7m8.7 .5 88.87 8.. «890880 .8... at on. 211 own o.~ omm emu . ovm emu omm o.~ com om. ow. o.. oo. 7 on. 7 ow. on. om. o.. co. co om o. om _ on ~>c 77.. 7 7_77 umw77.77._w v 7.77 777 77 7 7...: 477 7 -:.7nmm..7:77_u .7 7 . _ 7r77.77 7 mm.7.~ 7 .7 7. 7 777;...77. 7 7.1.7.7;r777 m 7 7 . . o 7 7 _wvm _ mam v-7 _. _ «a. . __7 77 7H 7o.~ 7mo~ 7 «w. new mom 7 7. 7 «on .mm 7 7 7 mm. mam 7o- 7 7 .7~ 7 7 7 7 7 mvm 7 _ 7 7 mum _ 7 - % mm. vmm... ..od [8. $38 .0 99.7960 7m 2.8qu $872898. 60 $8. .7 8 «81.8800 .86 8+. on. 22122 com on. ovm cm. 7 0.. com om. om. o.. oo. om. ov. on. om. o.. oo. om om o. oo on N\E 7 .L 7 777 . 77777. 7 7 7 .77 . 17-7IJ7477 7 fifill‘ 7Er. 7777 .7777 77 7 7 .7 77 . 7 .. 7.7%. 7 7 .7 .77 LL Fry 7‘} o 8.8. .8 7 .73.. 8.8. 7 N.. _ 7 77. 7. .8. . 8.8.8.8.8. 8. 8.8. N28.8.8.5 858 .8 2:8 .88 8.7.. 8 77.. no.7 mm. .; .7 ow. 7 .7 7 8.. . MVN 7NwN 77 778—7 av. 77cc. __ 7 _ _ 7 .m. 7 7 my. _mm. 7 _ co. 7 7 7 No.7 7 7 7 .7 7 . m.._ _ % .v. 7 7 7 .o. _ r wmovn m.~ oo. +8.77. .0 22.783 7. 2.8887. . 53.88. .7 Eu 3 ...87 .8. 807.888 .86 8+. on. 22123 CNN cow omN ocw 0mm ‘ ..-.F J 8N . 2N _ com on. 7 8. 2. cm. omr 0v. on. cm. 0.. 00. ca on on 00 cm N\E rL 7 _ 77. 7 L . 7 7 777 .7 777 7 7 LL 7 CL 7 0 8m SN BN7 7 3.8. 8.2. 777 2.8. E. .77.. 8. N... 8. aw. 73.8. 3.0.. 8 33 8 8 m. t 2% 5 .m an .m o8 2.772. .m. .8 .7. E73. 7 7 7.... 7 7 7 7 7 .7 7 77.78 77 77 0/0 7 77 77 SN 7 7 7 ..8 8. 77,33 +8.78 7.0 22.7978 7. £8.37 new .818390 mam: 88: Ba 8 30.29. 1 7.853.750 214 .3 8m .8 o; 08 8. 8. 2. 8. 8. 8. 8. om. o.. 8. 8 8 o. 8 8° ”‘57 7 rs. 7 7.71%} ‘ wblp777 7.7 r» 74777.7 »_7 7 J 4W£|7+.L 7 7. 7 7«7 7 7 7 77? 7 tr; L7 . 7 7 «“777 7 r 777.7..7 7 7 28m , _ 77 7 ....8. 8. 8. 0.. 9.8.8. 8 8 7 7 .. 8 88 8 mmmhmw QNN FNN CNN NON hmr mar mmr oar our 7 mkr 7 7 now wa 8F FNr m: 8 mm mm Fm as on QB rm .8 7 _ _ 7 8. a. 7 7 m... 7 t. __ 7 7 7 7 7 7 7 8. .m. 77 R. 7 % .88.. S. I .8. 7mm .mm .o 92880 ”m 7386353338 80 $8.87 8 .8 88.80 man: 88.. .8: 3 333: 2 :Samaotuo 215 EN 8m 8m 8m ~>c8~7717 777 7 7 8N «8 .8 8m 3m 8m $8.. +mm8m .0 227.88 7. CmN CNN CFN CCN CC. 7 CC. Chv CC. . cm. 0: 7 H 1 fi« «74 _ FMN nNN 8. 7 NON _7 _ .a. 7 8.8.3. ... 8. 8. 8. n... B. 8. 8. N$9.8. 5.88 88 8 _ o. m. 88 888 C r 9 CC F om CC on CC Cm. mu .8. 2 888+... 285...“.97 :5 38.87 8. .8688 .o mam: Snot tom an mooim: 9. 5393.30 216 8m 8m 8m _ o.~ . 8m 8. 8. E. 8. 8. o... 7 8. 8. 7 o.. 8. 8 8 . o. 8 8 mam8~tT5~8~ a? 2377.88.71 t 8. _ 77 8. .38. 8. 2.77... 8.85 8 8 8.7.. .o o 88 77 .m. 8. .7... 8.7 78.8.7 7 8. 8.8. 8 8 ..8 8 888 .8 SN 7 now 7 R: 8. 3.7.. on. momvom 3. 7 7 7 8. 77 77 77 77 7 77 75. 72. .8. 7 7 7 7 7% 77 7 7 woo...” 8. [8. +wmeN .0 82:03.80 “v 88787788783783 :6 :88: 8. 8:88.30 «in: 88.. Ba 8 333: 8 aggro". 217 Ema 8m SN 08 8. 8. E. 8. 8. o... 8. 8. o.. 8. 8 8 o. 8 8 7 7 7 7 4 . 47 7. .7 7. 7 7 . . 7 . 4 . _ o 7 7 7 777.777.7. 77 _ 7 7_ .7 _ . . 7 ___ ..7 7 7 7 7 7 8. __ 8m .8 7.EN momma 8. 8.7 8. .8. 2. .58. .o. S. 8. a: .3 R. 8. mm. a. .N. 0.. 8.5.. 8.8 a 8 888.8 e. 322888888 3 _ 77 :78. .7. 783 SN 7 7 7 7 7 7 7 77 7 7 8. arm 7 7 7 7 7 7 7 % 7 .8. 7 7 ,8. $8.. $88 .0 998:8 7. .m. 83.8.7 .5 88.87 .8. 8-8880 8... moo 218 Appendix B 219 Appendix B: Salicylic Acid Byproducts LC/MS DATA Typical chromatogram of salicylic acid byproducts obtained from Goethite catalyzed CHPs at pH 7 .................................................................. 221 Typical chromatogram of m/z 93 vs. TIC obtained from Goethite catalyzed CHPs at pH 7 ..................................................................................... 222 Typical chromatogram of m/z 153 vs. TIC obtained from Goethite catalyzed CHPs at pH 7 .............................................................................. 223 Typical chromatogram of m/z 169 vs. TIC obtained from Goethite catalyzed CHPs at pH 7 .............................................................................. 224 Mass Spectrum for m/z 169 obtained from Goethite catalyzed CHPs at pH 7...225 Typical chromatogram of salicylic acid byproducts obtained from Goethite catalyzed CHPs at pH 3 .................................................................. 226 Typical chromatogram of m/z 153 vs. TIC obtained from Goethite catalyzed CHPs at pH 3 .............................................................................. 227 Typical chromatogram of m/z 169 vs. TIC obtained from Goethite catalyzed CHPs at pH 3 .............................................................................. 228 Mass Spectrum for m/z 169 obtained from Goethite catalyzed CHPs at pH 3. . .229 GC/MS DATA Typical GC/MS chromatogram for salicylic acid samples after CHP Goethite catalyzed reactions ....................................................................... 230 Close up for peaks with retention time of 15.5 and 16. 2 minutes ................. 231 Close up for peaks with retention time of 18.9 and 19.20 minutes ................ 232 GC/MS mass spectrum for peak with retention time of 15.5 min .................. 233 GC/MS chromatogram showing to compounds eluding at the same retention time of 15.5 min ................................................................................ 234 GC/MS mass spectrum for peak with retention time of 16.2 min ................. 235 GC/MS mass spectrum for peak with retention time of 18.9 min ................. 236 220 0 09mm oo.om ooov oodv oodm oo.o.m oohm oo.oN oofi. oo.o. oo.m EC! . , p 7» . > . . . . > 7 L1} L1? r . . r . n » o _ .. .77 7 71 .7.. L J . .. 7 .7: 7...7......7........7......7...2..._....77...7.._...7......._7.:771.7.7717... .7 7777.77.77 77...... .... 7.7.7.7... 1...... 7.7.. 7 .. . . .A¢.1.5a7 .%.&.7_‘..#{...7.__. .7. .1744 7. 7 7 . J 7 m0 N.. .3 V.. .2?) 7.. :2... ...7.... 7........... .2. J .7738. ...7.)...7..7....J:$......7 .. 8 m .....m 2......77. 7.7.7. 7. 88 . . . .. ...7....1... . . :..... :75... 7 777......77 ._7 77 8.8. 0.38.8 .7 an“... if} . 7 $7388.85? 8.8 3.577.. 77 3.9.83 8.... 77 2. 9... om 7.7 72.9. 7 o/o 8 .887 8 8.... 8.... 7 7 8.... 8.3 B... 8.3 7 . 8... 88.. . mm .... r +wm cwow H. moo noowmoG oo.mm oo.om oo.mv oo.ov oohm oo.om oodw 8.0m ood. oo.o. oo.m r r r r . . s . 7 7 7 > y r . t 7 . . 7 . » . . . _ t H r _ > . H . r y n y . o 91.--- 1.31... ...f./ .... 1:11.1131.21.111.11 .3- -1- a l , (l- .- ..1 14 t l -1 --l. - ,t .t 2 7s7s771.--.,t7-31-71;}.,1.2 7.: i7. 7,. .1 / HER! 8.8 ..777 7 (:../r; 7 . .{?\~.. .4o7iw..7 7.. 7 Ix. 7 8 .... 88.. 7. 7 .77 77 __ _ 7. 7 7 7 7 7 .7 7 7 7 7 7 .7 7 7 7 7.. 7.7 7 7 ...7 77. 7 7 77.. 7 o/ol 77 77 .7 7 EN 7. 77 77 7 88.. 77. -mm cmow 7N moo noowmoO E... .22. 2.8.8 221 all as: 8.8 8.8 8.3 8.9. 8.8 8.8 8.2" 8.8 8.2 8.2 8...... . 7 L . . 7 . 7 - . 7 . . . . 7 . . . 7 . . 7 . . . L 7 7 7 o .__. 7. sp.— . . . .f.) . . 7 . ... . .. . ... . . .7 7 7. 7. . T ...7 777 7 V 7 . ...7.... 77.7777777777777 777.77.747.37... 7777...7....7...7177777..777 77777777.. 7777 7777777 7. 777 .777..77777..777..77.777. . .7. £777.77. 1 7 7.7“ ..f.x.3w.‘-.777.fi‘§. 5.773 7.3.7.1777. 7. _ 7 7 7 J 7 mmNpsQ: . 777.77.7‘7..7..... 7 7 a... 7. .77 .. 7, 77 .. . 7 .7. . _ 7 7 . . ... 77 77. . 7 a... 7 7.77 7.7777777777777777... 7 .77... 7......77...7777.7.77 . 7 7 7.: E 7 777777.. 7 7.77 77 7 . . . . .7. . 77 7.. $877 288...- - 8:. {777.777.7777 7 ._ «$7.; 7 . 7 7. . 7 moon . 8.3. .7 some; 8.8 ...-:.., 772.8 7 ..F... 8.3 _ 7. /o 5.53.8 7 23 8.: o 8.: 7 _ 78.3 2..... So on: . 78.... mom... 8.37 +8 .68 H. 8018.88 8.8 8.8 8.9. 7 8.9. 8.8 8.8 8.8 8.8. 8.2 8.2 88 11...2-.--+ .7 7 - 7 7 7 7 11:51-71. 111-11 7 11 : 7 o 1.1/.... .x.£.t.../. Rhitlxliiyfl $0.6 ...,777 7 ,7r71l-xxxr/J/ .7 ...-...7.? ...-.77 7 7. .,., 7 mo. 5 no we 7.7 77 7. 77 7 77.. 7 7 7 777 7 7 7 77 77 7 7 777 7 7 777 7 7 7 7 7 - %- 7 77 7 SN 77 _ 77 7 7 8.3 7 -mw 8% 7m 80 8880 :... 22.. 27.6778 221 oohm oo.om oofiv oo.ov 09mm coon oo.m~ oodN oo.mw oooF oo.m ”Emu—- . , r 1 1 . _ 1 , 1 . 1 . 1 > 1 1 1 _ L p 1 1 o t? (21.. «1.341 R... .1511. x191..- 113 l. modem x...... 1%.!) (II/ _.I x». .ri. .. , ...3' . ,. .2 A mo 5 8 o . ,. . _, 7 .., _v ,_, , ; _,..., ,. , % o 1 fl. i SN 9 or won! t o: 2.: u 8. .mm :mom ”N moo noonoO oo.mm oodm oodv oodv oohm coon oo.mN Doom com: 099 ooh » 1 1 h 1 . . 1 1 1 1 F W p . _ 1 1 1 .r 1 1 .1: 1 1 . 1 1 1 1 . . o _, “ sword mm -mw cmom N 59 8015880 5.. 23 232.8 222 09mm oodm 86v oodv 2: . _ . . 1 1 . 1 . .. _ 1 . . . coho 8.0m 00.90. 00.0N 8.3 Do or 09w F _ . 1 1 :1, .1 1 ‘5‘. , n 1 1 1. . 1 p . 1. o 1.1. .... .... 11.11.1111,. a... . ... _ We?“ Kt... . (141/1 8. _.m wmdv °/o _ ... _ 8.3 .. o: 9.: 1 8. -mm 88 “N m8 8880 8.8 8.8 8.9 . 223 8.9 802 8.0, 81 3o 1 181 .8 58.0. ..N moo 3380 En 23‘ 232.8 8.8 8.8 8.8 8.9. 8.8 8. on 8.8 8.8 8.8 8.3 8.2 8.... a we . F . > .1 . V _ . — V o . A . > P E. ‘(IlJJI’t/ Elli-[kit]. .1111 \|\lr}/\i.5|l.1IIID‘ Itl{.11o§1\4(/. {I}, C K/\\l a... . 1. W418 / 8.21. ./ I v, _. 1 ./. j .._ r 8.8 Mann.— ES 1 for -mm 518 8 m8 8880 8.8 8.3 88 8me . 8.8 88 88 88 8.8 . 8“,: 8.2 8.11. . o 3.353;," .. :..2§.1,§..§_:.....§._: 3.1.... 8 MM 3;“: Egg. :::; :2. {:..S. 15.81.23??? 8.88% _ . 4.3.15.8 8.23:: 8.: 8m we: 88 8 mm 8 . .3. v 3.8.3. 8 :.8 E. 8 B. 8 8 z. 8 5 m. .. 8.3m 8.8. . 8.3. 8.” 8.. . S. n. 8.8 B 8— 6.2 .9: %- 0‘0”. . nm.v— 8.8.. 2.8 omwwe . 2: 8. 8 8 I .8. :3m 8 88 8880 2.3 22‘ 2.2.6 224 8.. 8.. o8 .8 08 8... 8m 98 08 98 08 08 8. 8. o... 8. 8. 8 8 N): . 1Ifr 1. 1 . .171? 1% _ 1 .1. fr _.1 .1...1.1 . 1... . . 1.1 o 8.. 83.... 8.. 8.. m8 «8 8m 8.” 88.882 11.8.8 .8 8888 m8 «8 98 8889888888. 02...... 8. ...... 8. 8 .m ._c 8 8 a... 8. .. _ 8. 3.. o... _ . . _ v . V . % 8. 1 . 88.. 8. I 8. -mm 80m 8 ..83888.88888. .5 88.... 8m 8.. 8880 :... .22.. 232.8 225 oem..8 8.8 . . 8.8 . 8.8 . 8.8 . 8.8 88. 8.8 8.8 88.. g .88.. _ 88 o _ .._ _T... . .. _ _. ...... _. _. L _ _. 5 .. ._:._..... .1... _...._.. .._..3......_. ....3... ...... ......13 ...... . .___._ .._.........._....._ .._...S _ .. ...: ...... :..... ... 1%... . .... .... . _ . :.. .3: _. ...... ...... . ...__......: . :.. .... . _ ... .... . o ...... . ... _. ..._......._... . .__. 8.8. .88 .. 8.8 ’ .... _ 8.8 . . . I . b .2.. 8.8:. 88. 28.8 88 _ ._ .. 8 ... . _ ‘ g 8.8 8....-. 8.8 88. NoNv mod...” 8.... 88.. 8.8 . .18. 0: 8.. +8 88 ”. «818880 8.8 8.8 _ 8.8 .1 . 88 . 8.8 .8 8: 8.8 _ 8 8. 8.8. 8.8. 8H... . . .88 . 1. . o I. 1.. 1 . [111.1 11.11. 1.1....) {I»J\.\s~<..f.t.f. !\1II( 1 11L .. .1 11111 11l1 r 1 I11 II’III1 Ill I 1.11 1 r 1“; 1t- 11 111.1! 1 [I I111...___.\11..11 1ll111 1 .11l1.f1.,1113bl1/Jt11 ...lfll 11 .. u / 8.8 8 S 8 ... 1... .. .... .. . .__ . _ . . . o/ol . . ___. . _.. .8. .. _. A _ .._ _ .... 8.8.8 .. -8. o: 8. -8. 88 8 8018880 ...... .22 232.8 226 2:8 8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 c. ..\ -. 1i 1... . ...). .. 11111.1. , ....111- 1 . . . 1111---..-1,.I-.-.1-.[ . N11: --.1.1_-1.1.1- .. . .11-- L f _ I .11 o /,J Iii-IJC (\r\ .111I1..|\.\..!....11 ......8 1 [I 1.1;: 4 B 8 8.8 8.... x... 4 .. 4 4 8 4 4 .. 4 4 4 o .8 4 o/ .8. 4 44 8.. 88.. _8 _ r4 4 8 88 8.8 .8. . 0:. 8w -8. .88 80w 8 :8 8880 code 86m .098 8.9 8.9. oohm 8.8 00.3 oodm 8.9 oowow oo.m *A‘ir‘» ... . _W.‘W. mm. 1.. 1...... .. .1 ....fl . . .... :... . .. .4. . . ..4. .1 ...I.. 4. ....1 . .m.m184 #4.}? .1 :.....41. ...-.__ «14.44.? ‘qui‘lflfé _.xawivfi—d‘34 :4}.?E ‘gw4444144.4 fldq44$4r143¢4fi o . _ . 848 . 8. 8 88 ... 8 8 .8 8.8 8 8.8 .N ... 8.....8... _ 8 . a. . $8. .Nm .4. . 4 4 «mm 8.8 444 _ _8 8 No.4 4 88 4 _ 4 4.4 4 _ __ 8... 4 _4 8 _ %l N..... 4. 8.. .4 .4 . 88 44 .. 8.9 .__ 88.8 8.8 4 -mm 8m .N «8 S880 «:.. 20¢ 28.8.88 227 8...». 8?. _ 8.8 _ 89. _ 8H3 . 8.3 , 8H8 8.3 _ 848 _ 8H9 _ 8HoF _ 8.0 . a w .. -1z/z§<&\)/.\l\\\[lil.llql \ :1 I... ‘z/Ki 95\Is./\/I,./ F98 8.: a FF 1 x % / 8F , h F _. F ,_ Q22 LF F 0: 8F -mm 5&6 “m 39m F8880 8.8 . 8.3 , 8.8 8.3 8.9. 8mm 83m . 8.3 8.8 8.3 8.8 8a. 3:233 o F E; ._Figa ..F: £22,; JFFFFFEFF: ‘ a} £333 25 8.3. 89. 2:; , i; . {3‘1 3‘5]... 8F 8. 3 Fa. , Fm. Fm 8.8 . 2.9. 2.2“ on!“ 88 HIKE wF FF FFNF 89on . $~ Fawn 8mm 2L . 37: FF. $3.an an a 8.: .. F2“ No.8 mFFn %- ; Freon: h 38 . «Nam 9..an mop opém bow -mw enema F.8IF8N80 2.... E2 232.8 228 0mm. 0N.m oFm ....oom 8N omN CNN ooN CON: _.ovN omN .o.NN_. oFN DON. omFL :ow.F OFF _owF omF. 3F omF ONF OFF 2: 8.8 0F 8 Nmm .IL . F5. . ...... .F . :L F. _. ... ..F F. i . _ F. 4.11 O N 8NmFNn.FN 8N 8N 8N mFNNFN 8F _ . ...... ..8.F .8..F . a ..1 8F. .. ...FF. , _ m8 8an 8an 8N 8N8N 8N 8N _N...N FNN 85: _. 8F8F. FnF FNFNN; FE; 88:3 .. : _ OFF” . mFF .__ 88F» .. . __ . 8F. H _ . 8 OFF . . m FF. 8. . 8 . . 8 _ _ _ . . . _ . 2 w _ : 8F * ANS _ _ . . -% . i . . . , _ . 8F . . 88.8 8F I [SF -mm 88 UN ..mFFHmmYmanoeéowwmFV 50 89.3 8F 28 F888 2... 23.. 2.5.8 229 0.0.. 00er CONN? COMO? 8.0 8.0 b h P b n p n J. 1...J 1(4/j mE.N«@nv/0 .WONIO P NMON “MU—F 4.12sz . Ali-06...: {#90 P .F°&.0N .5900 .590? .hofom F5900 .FQON. .Nu0.¢.0m .F0+00 .mo+0 P 090?. 00.9.00. 00.700. 09°F. m90m. 00.700. 00CNUC3£< .QQOP. .mgoN. 1mo+mm. 230 .fillWET—I wE .NHFNU/D .MONO F. NMON ”0 Cu who+mm who+wv . .ho+mm who+mw .mo+me . ..w0+m.N. F. F .mo+mv.. .mo+mmF. mwo+mw.r QOCWUCDQIQ 231 AnumEE. 00.6.. OWLQF ONFdF oohmw OQMwF OOMQF OVMQF .oooooom .oooooov .ooooooo v v .oooooom .no+mF .Fo+mm.. .Fo+mv.. v 9». damn/D .MDND FNMON H”FE. mucmncand. 232 AuuNFHC ovm ONm 00m omN O©N OWN ONN DON 09. 00.. 0: ON_. 00F. 0m 00 CV th NmF NVN mt. _.OF 99. th. . .0 , mm N F _. F0 _. mm .OOOOON .OOOOOV .000000 .000000 .OOOOOOF .OOOOONF .OOOOOVF .Ooooowr .OOOOOw F mK. 3 mEdFm—uzn—dOhO _.NmoN H.EE 00mg": OF NEWS. ..F0 009024 mocmocsnd. 233 Inll-E F 0L..0P ONHOP Donn? OOWOP. Donor five—.0? 0*..0F. 0*.0? 00.0? 00.0? 0~.0? 0~.0? - p — - p r. p - rr - h n F > p p F p h - b b n h P h h D b P b h P P! D — * .F b h —* .P + hL7.o b P h > 000000 000000 F WE.NuN—U/0.WOK0F.N.N0~ $05.50” 0a 05.1m0v00.mmm C0. OOCNUCJna/l AI‘ECI ”fine? 05“”? noun... 00..”? 0.0—.0? 00..”... 0*».0? 0*..0... 00h”? ONPO? 0N..”... ONMO? <1 . . 0 000000 000000 F IE..u.—U’O.00L0 ?N00~ NAOF.0N0 cu 05.KN.OV 00.0“” C0. UQCfl-UCJQI‘ Ain't-LEI 0k..0—. ONHOF OOMOP coho... OOMOF coma? 0*.0 0*..09. 09.0? 0.0—.0? ONMO? ONMO? b b h h F b P b .F .F D b P b I b F F b h h F b b D I I I I .0 —> h P ~_*- hh rrle h h - ner> F 000000 000000 F. .E.-al—UJD.W050 funnel ”AOFKNd‘N 0a 05. P‘Nv 00.N‘Nu C0. '0 Elfin... 394‘ ..nIIUCC ...... whhmf Ohhmf 00.0? oohwr wwhmr 00mm? 0*HW—s OVMWV mmhm? 00mm? muhw? ONMW? h b b P h h 0 II!////1\\\\\\i| 000000 000000 F IE..u'sU/AU.00K|0P~00~ .AOk..mon 00 0h.NONv 00.0mm“ CO— -nvr-IFUCFJOAI‘ .HII'EF 05.0... 05.0? 00.0 00.0? 00.0? 00-0... 0'-0—\ 0'-°P. 00.0? 0n.0—. 0~.0? 0~.0? F- » h —L b b r —r h b h b — b p n P h h rL b h P b b - hL - h P \— .P h h h — h b P b F I p r h — h h F - — h F b L .—.o < 000000 000000 F .E.‘u.g.m0.\|0 PNmuO' ...th.~°? nu Oh. ?0?v 00.NOF C0. vocnU::Q{ 234 00m” OQN OQN —-n-—nrhh.—— 4n— n - «___: .mXWN ONN OmuN 1! vrm mmN ?QN .3-.. QON \IQN 1‘1 .— NO? ON? QR? 00? MO? 9?? 0V? mum”? ON? Qr? 00? 00? ?Q WE.fiuNU/0.MHONO?NMON H.C_E 00.0-0? 0“. N00.0? ...0 $00L0>< 0m MN 00 00 in IIN\ E CV 1333...... Pooooom 3 3 3 O .oooooor nooooomr .OOOOOON mooooomu .oooooom wooooomm wooooooe 3 Wooooomv 00CNUC39< 235 0“ ON‘ 00‘ 0000:00'0 0N”00”01N0I~0.~0"00Iu00f 0| PO‘P our 00.- 00 am sh.“ 00” nm‘ on." n . is r” .50 _ ”no 0E- mango .nOFO Tmnou 0‘“ ”on menu mo... Pho- h“? 0.1"? n--L-E \Irm.Q—u OH ‘00-”? to ”mmLfl\/( N0? 00000 f 00000“ 00000” 00000.‘ cocoon 000000 00000.“ 000000 000000 000000 nu 00000 f F 00000” a: 00000” p: 00000‘ .- ooooom .- 000000 '- wocmucjnd‘ 236 Appendix C 237 Appendix C: HPLC Chromatograms HPLC chromatograms of hexane and acetonitrile extracts from CHP reactions extracted from both soluble and mineral iron catalyzed reactions in aqueous matrices. Hexane Extracts Soluble iron catalyzed: chromatogram using ultraviolet detector (254nm) ....... 239 Soluble iron catalyzed: chromatogram using fluorescence detector (ext: 310nm, em: 400nm) ............................................................................... 240 Goethite catalyzed: chromatogram using ultraviolet detector (254nm) ........... 241 Goethite catalyzed: chromatogram using fluorescence detector (ext: 310nm, em: 400nm) ..................................................................................... 242 Acetonitrile Extracts Soluble iron catalyzed: chromatogram using ultraviolet detector (254nm) ...... 243 Soluble iron catalyzed: chromatogram using fluorescence detector (ext: 310nm, em: 400nm) ............................................................................... 244 Goethite catalyzed: chromatogram using ultraviolet detector (254nm) ........... 245 Goethite catalyzed: chromatogram using fluorescence detector (ext: 310nm, em: 400nm) ..................................................................................... 246 238 VN NN ON 0.. Dr V.. N9 0.. xxx/(N.. No.31. _. —. DON 00m” 00? 000 00“ 00h com 000 239 0N ¢N NN ON m.. 0., V.. NF 0., m 0 _.V N \S(W..N00mv.v2 00.. DON 000 00¢ 000 000 00h 000 000 240 0N VN NN 0N mr 0r VF NF 0r >2<~ l, N000??? CON 00¢ 000 000 000? OONr ooqr 000v nomv OOON 241 0N VN NN ON 0.. 0.. .Vv N.. 0.. >>>(m a. 000vrr 00V 000 000 000v 00H? 00Vr 00a? 000v 000N 245 vu NH 0N m. __.._____________._.____—___.___._pp_.__ 0. ¢. N. o _______._________._____—________ . m 0 v. N ______Fh.________________.p.._________.—_._____rr 0000.15,. nzufi nxum nxuv nxun nzuo nxuh nxum 246 Appendix D 247 Appendix D: MATLAB Code for Kinetic Model: This model was run using the data collected from the oxygen production data and the hydrogen peroxide titrations. The RMSE was obtained using the Genetic Algorithm function. Authors: Geneva Hulslander and Carlos Sanlley function [cpr,T] = FentonFerric2T(t,c); global c0 tspan kl T; cpr = zeros(7,1); % CO=[O.1,0,0.07,0,0,0,0]; %Trial 1 %decomposition of h202 in l/s %fenton reaction k's; k2=76; k3=2*1o‘-3; k4=1*1o‘-3; k5=5*1o*-7; k6=2*1o*5; k7=3.3*1o*7; k8=(1.2)*1o*6; k9=1*10‘7; V=2s*1o*-3; TO=19.5; theta=1.058; %T=20; %Not temperature corrected T=To*(1.102.*exp(-(0.0012./60).*t)-O.105.*exP(-(0.406./60).*t)); %Trial 1 %T=To*((1.612.*exp((-0.0094./60).*t))-(0.639.*exp((-O.539./60).*t))); %Trial 2 %kl=0.00194; %From optimization Trial 1 kl=0.0084l; %From optimization Trial 2 % hydrogen peroxide cpr(1)=-(kl.*theta.‘(T-20)+k2*theta.‘(T—ZO)*c(2)+k3*theta.‘(T- 20)*c(3)+k7*theta.‘(T-20)*c(4))*c(1)+k9*theta.‘(T-20)*c(6)*c(2); % Ferrous Ion cpr(2)=-k2.*theta.‘(T-20).*c(l).*c(2)+k3.*theta.‘(T- 20).*c(3).*c(l)+k4.*theta.‘(T-20).*c(3).*c(5)+k5.*theta.A(T- 20).*c(6).*c(3)-k8.*theta.‘(T-20).*c(2).*c(5)—k9.*theta.‘(T- 20).*c(2).*c(6); % ferric ion cpr(3)=-(-k2.*theta.‘(T-20).*c(1).*c(2)+k3.*theta.‘(T- 20).*c(3).*c(l)+k4.*theta.*(T-20).*c(3).*c(5)+k5.*theta.‘(T- 20).*c(6).*c(3)-k8.*theta.‘(T-20).*c(2).*c(5)-k9.*theta.*(T- 20).*c(2).*C(6)); % hydroxyl radical cpr(4)=k2.*theta.‘(T-20).*c(l).*c(2)-k7.*theta.A(T-20).*c(l).*c(4); % perhydroxyl radical cpr(5)=k3.*theta.*(T—20).*c(1).*c(3)-k4.*theta.‘(T-20).*c(3).*c(5)— k6.*theta.‘(T-20).*c(5)-k8.*theta.‘(T-20).*c(2).*c(5); % superoxide cpr(6)=k6.*theta.‘(T-20).*c(5)-k5.*theta‘(T-20).*c(6).*c(3)- k9.*theta.‘(T-20).*c(2).*c(6); O 5 Oxygen 248 cpr(7)=(2.*kl.*theta.‘(T-20).*c(1)+k4.*theta.‘(T- 20).*c(3).*c(5)+k5.*theta.‘(T-20).*c(3).*c(6)).*V; function [rmse]=fentonobj(kk) global c0 tspan k1 X; kl=kk(l); X=kk(2); tspan = [0:15:2100]; C0=[0.l,0,0.07,0,0,0,0]; %Trial 1 %CO=[O.5736,0,0.07,0,0,0,0]; %Trial 2 options=odeset('AbsTol',[1e-7,1e-12,1e-7,le-12,le-12,le-12,1e-7]); [t,c]=ode23s(@FentonFerricZT,tspan,c0, options); load work time=b(:,l)*60; simc=interpl(t,c(:,7),time); Obs=b(:,2); rmse=sqrt(sum((simc-obs).‘2)); plot(t,c(:,7),'-g');hold on %errorbar((a(:,l))*60,a(:,2),a(:,3)./2,'r“); errorbar((b(:,1))*60,b(:,2),b(:,3)./2,'bp'); title('Oxygen'); xlabel('Time (sec)'); ylabel('moles'); hold off o\° o\° o\° global c0 tspan kl; k=fminbnd(@fentonobj,le-6,1); 249