|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 FATE AND TRANSPORT OF ARSENIC IN CONSTRUCTION AND DEMOLITION DEBRIS LANDFILLS AND SOILS UNDERNEATH By JIANYE ZHANG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011
2 2011 Jianye Zhang
3 To my family
4 ACKNOWLEDGMENTS I would like to thank my advisor, Dr. Timothy Townsend, for his excellent guidance, encouragement patience, and support during my Ph.D. research. He has always provided me opportunities to learn new things in the area of solid and hazardous waste management as well as other environmental fields. He always encouraged me to think as an engineer and a sci entist. I also would like to express my gratitude to my committee members: Dr. Michael Annable, Dr. Jean Claude Bonzongo, Dr. Yong Cai, and Dr. Dean Rhue. They have provided me opportunities for discussion and have given me many helpful suggestions during my experiments. I wish I could have more time to learn from them. In addition, I would like to thank Dr. Cai for his generous help in arsenic and sulfur analysis with ICP MS in his lab at Florida International University. I would like to thank all group m embers for their great help and numerous advices Particularly, I want to thank Dr. Hwidong Kim, for his valuable suggestions and great support in every aspect of my research. The help from Dr. Jae Hac Ko and Dr. Brajesh Dubey is also appreciated. I also w ish to thank Dr. Ligang Hu at Florida International University for his assistance in conducting arsenic metal analysis. Special thanks go to my Chinese community in the group: Dr. Yu Wang, Mr. Yongqiang Yang, Ms. Jenny Hou and visiting professor Dr. Wendy Zeng. They are all great friends and we had a lot of pleasant moments together. From the bottom of my heart I owe a debt of gratitude to my families both in China and in the USA Without their encouragement, understanding and support I could have achieve d nothing in my study.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURE S ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 11 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 Arsenic Leaching from Chromated Copper Arsenate (CCA) treated Wood in Construction and Demolition (C&D) Debris Landfills ................................ ........... 15 Proble m Statement ................................ ................................ ................................ 17 Objectives and Research Approaches ................................ ................................ .... 20 Objective 1 ................................ ................................ ................................ ....... 20 Objective 2 ................................ ................................ ................................ ....... 21 Objective 3 ................................ ................................ ................................ ....... 21 Objective 4 ................................ ................................ ................................ ....... 21 Outline of the Disser tation ................................ ................................ ....................... 22 2 IDENTIFICATION AND QUANTIFICATION OF THIOARSENIC SPECIES IN CONSTRUCTION AND DEMOLITION DEBRIS LANDFILL LEACHATE ............... 23 In troduction ................................ ................................ ................................ ............. 23 Materials and Methods ................................ ................................ ............................ 27 Synthesis of Thioarsenic Compounds ................................ .............................. 27 Characterization of Thioarsenic Compounds ................................ .................... 28 Ion chromatography ................................ ................................ ................... 29 Mass spectrometry ................................ ................................ ..................... 30 Formation and Stability of Thioarsenic Anions ................................ ................. 31 Quantification of Thioarsenates in Ion Chromatography ................................ .. 32 Thioarsenate Analysis in C&D Debris Landfill Leachate ................................ .. 32 Results and Discussion ................................ ................................ ........................... 33 Characterization of Thioarsenic Compound s ................................ .................... 33 Ion chromatography ................................ ................................ ................... 33 Molecular mass spectrometry ................................ ................................ .... 34 Elemental mass spectrometry ................................ ................................ .... 37 Factors Influencing Thioarsenic Formation ................................ ....................... 38 Thioarsenate speciation change over time ................................ ................ 38 Concentration effect on thioarsenate formation ................................ ......... 39
6 Stability of T hioarsenic S pecies ................................ ................................ ........ 39 Quantification of Thioarsenates ................................ ................................ ........ 41 Thioarsenate Analysis in C&D Debris Landfill Leachate ................................ .. 45 Summary ................................ ................................ ................................ ................ 46 3 ARSENIC LEACHING AND SPECIATION IN C&D DEBRIS LANDFILLS AFFECTED BY DIFFERENT AMOUNTS OF GYPSUM DRYWALL ...................... 61 Introduction ................................ ................................ ................................ ............. 61 Materials and Methods ................................ ................................ ............................ 62 Waste Composition ................................ ................................ .......................... 62 Lysimeter Construction and Waste Loading ................................ ..................... 63 Lysimeter Operation ................................ ................................ ......................... 64 Leachate Analysis ................................ ................................ ............................ 64 Leachate Arsenic Speciation Anal ysis ................................ .............................. 65 Comparison of Preservation Methods ................................ .............................. 66 Arsenic Solubility in Sulfidic Solution: Batch Test ................................ ............. 66 Results and Discussion ................................ ................................ ........................... 67 Leachate P arameters over T ime ................................ ................................ ...... 67 Leaching of Arsenic from Landfills ................................ ................................ .... 69 Leaching of Iron from Landfills ................................ ................................ ......... 71 Comparison of Preservation Methods for Arsenic Analysis .............................. 71 Arsenic S peciation in Landfill Leachate ................................ ............................ 72 Thioarsenic speciation using ion chromatography ................................ ..... 72 Arsenic speciation using arsenic sieve cartridges ................................ ...... 73 Summary ................................ ................................ ................................ ................ 74 4 ADSORPTION OF THIOARSENIC ANIONS ON IRON OXIDE COATED SAND ... 86 Introduction ................................ ................................ ................................ ............. 86 Materials and Methods ................................ ................................ ............................ 91 Synthesis ................................ ................................ ................................ .......... 91 Mineral Characterization ................................ ................................ ................... 92 Adsorption K inetics ................................ ................................ ........................... 92 Adsorption I sotherms ................................ ................................ ....................... 93 Arsenic A dsorption under S ulfidic C ondition ................................ ..................... 94 Re sults and Discussion ................................ ................................ ........................... 94 Adsorptio n of Arsenic on Hematite coated Sand ................................ .............. 94 Effect of pH on Adsorption ................................ ................................ ................ 96 Summary ................................ ................................ ................................ .............. 102 5 ARSENIC RETENTION IN HEMATITE COATED SAND AFFECTED BY DIFFERENT LEVELS OF SULFIDE IN C&D DEBRIS LANDFILL LEACHATE .... 108 Introduction ................................ ................................ ................................ ........... 108 Materials and Methods ................................ ................................ .......................... 111 Sand C oating and I ron A nalysis ................................ ................................ ..... 111
7 Column S etup ................................ ................................ ................................ 112 Column O peration ................................ ................................ .......................... 113 Effluent A nalysis ................................ ................................ ............................. 114 XPS Characterization of C oated S and ................................ ........................... 114 Results and Discussion ................................ ................................ ......................... 115 General T rend of E ffluent P arameters ................................ ............................ 115 Iron M obiliza tion ................................ ................................ ............................. 116 A rsenic Removal on Hematite coated Sand ................................ ................... 117 Surface A nalysis of C olumn C oated S and by XPS ................................ ......... 119 Summary ................................ ................................ ................................ .............. 121 6 SUMMARY AND CONCLUSIONS ................................ ................................ ........ 133 Summary ................................ ................................ ................................ .............. 133 Conclusions ................................ ................................ ................................ .......... 136 Future Work ................................ ................................ ................................ .......... 138 APPENDIX A ADDITIONAL MATERIAL FOR THIOARSENATE IDENTIFICATION ................... 141 B ADDITIONAL MATERIAL FOR SIMULATED LANDFILL EXPERIMENT .............. 145 C ADDITIONAL MATERIAL FOR ADSORPTION EXPERIMENT ............................ 146 D ADDITIONAL MATERIAL FOR HEMATITE COATED SAND COLUMNS EXPERIMENT ................................ ................................ ................................ ...... 148 LIST OF REFERENCES ................................ ................................ ............................. 151 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 159
8 LIST OF TABLES Table page 2 1 Elemental composition of thioarsenic compounds ................................ .............. 58 2 2 pKa values of thi oarsenic acids and arsenic acid ................................ ............... 59 2 3 Determination of repeatability by replicate injections of the same sample .......... 60 3 1 Waste composition in C&D debris blank lysimeter ................................ ............. 85 3 2 Ion chromatograph conditions for speciation analysis ................................ ........ 85 4 1 Adsorption parameters of arsenic species on hematite coated sand ............... 107 5 1 Nomenclature of sand columns ................................ ................................ ........ 131 5 2 XPS r eference binding energies of iron, oxygen, and sulfur ............................. 131 5 3 XPS peak assignment and percentages of iron, oxygen, and sulfur of control sand, LS CS sand, and HS CS sand ................................ ................................ 132 A 1 Determination of injection volume ................................ ................................ ..... 144 A 2 Concentrations and corresponding peak areas of thioarsenates ...................... 144
9 LIS T OF FIGURES Figure page 2 1 Ion chromatogram of thioarsenic synthesis mixture: Isocratic elution ................ 48 2 2 Ion chromatogr am of thioarsenic synthesis mixture: Gradient elution.. .............. 48 2 3 Mass spectrum of monothioarsenate compounds (Fraction 1).. ......................... 49 2 4 M ass spectrum of ion chromatographic Fraction 2.. ................................ ........... 50 2 5 Mass spectrum of ion chromatographic Fraction 3.. ................................ ........... 51 2 6 Reaction kinetics of thioarsenates formation ................................ .................... 52 2 7 Concentration effect on thioarsenates formation.. ................................ .............. 53 2 8 Stability of thioarsenates in model reactions. ................................ ................... 54 2 9 Calibration curves of thioarsenates and arsenate.. ................................ ............. 55 2 10 Linearity curves of thioarsenates.. ................................ ................................ ...... 56 2 11 IC chromatograms of C&D debris landfill leachate.. ................................ ........... 57 3 1 Leachate parameters over time.. ................................ ................................ ........ 77 3 2 Arsenic concentration in leachate over time.. ................................ ..................... 78 3 3 The relationship between arsenic and sulfide in leachate from all landfills except Control.. ................................ ................................ ................................ .. 79 3 4 Arsenic solubility in sulfide solutions at pH 7: batch test.. ................................ ... 80 3 5 The relationship between iron and sulfide in leachate from all simulated landf ills.. ................................ ................................ ................................ ............. 81 3 6 Comparison of preservation methods for arsenic analysis.. ............................... 82 3 7 Thioarsenic speciation in leachate by ion chromatog raphy.. .............................. 83 3 8 Arsenic speciation by cartridge and ion chromatography.. ................................ 84 4 1 Adsorption isotherms of arsenite, arsenate and mono thioarsenate on hematite coated sand .. ................................ ................................ ..................... 103 4 2 Adsorption isotherms of arsenite, arsenate, and monothioarsenatre at different pH.. ................................ ................................ ................................ ..... 104
10 4 3 Adsorption isotherms of arsenite in sulfide solution.. ................................ ........ 105 4 4 Sulfide change and thioarsenate formation during adsorption at pH 7.. ........... 106 5 1 Parameters of influent and effluent in Phase 1 at continuous flow.. ................. 123 5 2 Ferrous and sulfide in influent.. ................................ ................................ ......... 124 5 3 Ferrous and sulfide of influent and effluent in Phase 2.. ................................ ... 125 5 4 Sulfide and ferrous concentration in influent and effluent. ............................... 126 5 5 Leachate and effluent parameters: total arsenic.. ................................ ............. 127 5 6 Relationship between influent sulfide level and arsenic removal efficiency of hematite coated sand columns.. ................................ ................................ ....... 128 5 7 Solid phase arsenic and iron content in hematite coated sand columns after column operation.. ................................ ................................ ............................ 128 5 8 XPS narrow scan O1s and Fe2p3 spe ctra of hematite coated sand. Binding energies in these spectra are uncorrected.. ................................ ..................... 129 5 9 XPS narrow scan S2p spectra of HS CS sand.. ................................ ............... 130 A 1 Ion chromatographic separation and mass spectrometric analysis of anions ... 141 A 2 A diagram of ASRS working principle.. ................................ ............................. 142 A 3 Synthesized thioarsenic compound: monothioarsenate. ................................ .. 143 A 4 Ion chromatogram of synthesized thioarsenic compound. ................................ 143 B 1 Eh pH diagram of As S O system.. ................................ ................................ .. 145 C 1 XRD spectrum of hematite.. ................................ ................................ ............. 146 C 2 Adsorption kinetics of arsenate on coated sand at pH5 .. ................................ .. 146 C 3 Color change with increasing arsenic concentration in arsenite adsorption experiment in sulfide solution.. ................................ ................................ ......... 147 D 1 Colu mn experiment setup.. ................................ ................................ ............... 148 D 2 Appearance of coated sand columns after operation.. ................................ ..... 149 D 3 Cleaned and dried sand from operated col umns for XPS analysis. ................. 150
11 LIST OF ABBREVIATION S BE Bonding Energy C&D construction and demolition CCA chromated copper arsenate COD chemical oxygen demand DO dissolved oxygen EPA Environmental Protection Agency ESI TOF e lectrospray ionization time of flight HPLC high performance liquid chromatography IC Ion chromatography ICP AES inductively coupled plasma atomic emission spectrometry ICP MS inductively coupled plasma mass spectrometry IRD iron reductive dissolution K d d istribution coefficient K F Freundlich adsorption constant K L Langmuir adsorption constant MA methyl arsenic acid MCL maximum contaminant level MCLG maximum contaminant level goal MSW municipal solid waste ORP oxidation reduction potential PVC Polyvinyl chl oride RCRA Resource Conservation and Recovery Act R f retardation factor SPE solid phase extraction
12 TDS total dissolved solids VOC volatile organic compounds XPS X ray photoelectron spectroscopy XRD X ray diffractometer
13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FATE AND TRANSPORT OF ARSENIC IN CONSTRUCTION AND DEMOLITION DEBRIS LANDFILLS AND SOILS UNDERNEATH By Jia nye Zhang December 2011 Chair: Timothy Townsend Major: Environmental Engineering Sciences The focus of this dissertation was on the leaching, speciation and transport of arsenic, particular ly inorganic arsenic, in construction and demolition debris (C&D ) landfills. With the aid of mass spectrometry for confirming the identities of thioarsenate components, a relatively simple gradient elution method using ion chromatograph was developed to detect anionic thioarsenate species in the leachate of C&D landfil ls. The method was shown to have good precision with relative standard deviation values around 5.0%. Linearity ranges were also established. The effect of sulfide levels on arsenic leaching and speciation was evaluated by analyzing leachate from laboratory simulated C&D debris landfills with various percentages of drywall disposals. Generally, sulfide concentrations were found to increase with higher percentage s of drywall in landfills. However, arsenic leaching was neither positively nor negatively cor rela shape relationship was found, which is probably controlled by the precipitation of arsenic sulfide minerals at relatively low sulfide levels and the formation of thioarsenates at higher sulfide levels. The conventionally used acid preservation method was proven to be inappropriate for C&D debris landfill leachate due to its loss of
14 recovery in total arsenic analysis. Thioarsenates were shown to be air sensitive, especially tetrathioarsenate. Preserving leach ate with minim al air exposure is the most appropriate way for thioarsenate analysis. The adsorption of monothioarsenate on hematite coated sand was similar to that of arsenate, in terms of adsorption capacities and pH dependence. Both arsenate and monothio arsenate have decreasing adsorption capacities with increasing pH. The similarity between the two was possibly due to their similar structures which have only one atom difference. The adsorption of arsenite on hematite coated sand in sulfide solutions was shown to depend strong ly on pH. At pH 5, the precipitation of arsenic sulfide caused the high fraction of arsenic removal from solution. At pH 7 and pH 10 the formation of iron sulfide minerals could possibly account for the slow increase in adsorption w ithin low arsenic concentration ranges. Hematite coated sand columns were used to investigate the transport and removal of arsenic drained from C&D debris landfill leachate. Sulfide levels in the leachate influent affected arsenic removal and iron dissolut ion on the coated sand. Lower arsenic removal was found when influent had higher level of sulfide. In contrast, more iron dissolution occurred with a higher content of sulfide in the influent. The low arsenic removal efficiency of the coated sand column w h en the level of sulfide in the influent was high has been suggested to be due to the iron dissolution and/or the formation of iron sulfide minerals, which was confirmed by X ray photoelectron spectroscopy results.
15 CHAPTER 1 INTRODUCTION Arsenic Leaching from Chromated Copper Arsenate (CCA) treated Wood in Construction and Demolition ( C&D ) Debris Landfills Arsenic, due to its high toxicity, has been used extensively in the wood preservation industry in the United States. The yearly industrial consumption o f arsenic in the U.S. between 2000 and 2004 was estimated to be around 20,000 metric tons, out of which the majority was used in producing wood preservative chromated copper arsenate (CCA). For example, in the U.S., 90% of the 19,600 metric tons of arsenic compounds consumed were used to preserve wood in 2002 (Reese, 2002) After 2004, the consu out of using arsenic in preserving residential wood products. However, in 2007, 50% of the 5,280 metric tons of arsenic compounds consumed were still used in wood preservation (Brooks, 2008) The use of CCA as preservatives significantly increased the service life of woo d products. However, potential arsenic contamination from CCA treated wood during service is of concern. Both risk assessment analysis from EPA (2001) and laboratory study (Khan et al., 2006b) showed that arsenic leaching from CCA treated wood is a health concern. Due t o this CCA treated wood was voluntarily phased out for residential use by manufacturers from 2004 (FR 2003) On the other hand, the disposal of CCA treated wood will continue for a long period. It has been estimated that 10 million m 3 of CCA treated wood will be disposed of in 2010 in the U.S. (Jambeck et al., 2007) In Florida, the amount of disposed CCA treated wood was estimated to be from 0.4 to 0.7 million m 3 per year between 2000 and 2030. Most of the CCA treated wood has been disposed of in landfills, especially in
16 construction and demolition (C&D) debr is landfills, since CCA treated wood waste is often managed as C&D debris when discarded. CCA treated wood is excluded from the definition of hazardous waste under RCRA (CFR 2003), so CCA treated wood can be disposed of in landfills without any restrictions even when standa rd leaching test s show the exceed a nce of arsenic (Townsend et al., 2005; Townsend et al., 2004) In the U.S., m bottom liners or leachate collec tion systems (Clark et al., 2006) In such unlined C&D debris landfills, the leachate produced will infiltrate into the soil underneath, and eventually into groundwater. The contamination of groundwater by arsenic will impact the quality of drinking water. In the long term, the exposure to dri nking water contaminated with elevated levels of arsenic can cause various diseases, including several kinds of cancers, diabetes, vascular disease, hypertension, neurological disorders, reproductive problems, and skin damage (Hopenhayn, 2006) T he U.S. EPA lowered the drinking water standard (maximum contaminant level, MCL) of arsenic from (FR 2001) and the new standard has been in effect since January 23, 2006. The max imum contaminant level goal (MCLG) for arsenic in drinking treated wood and the fact that most of the CCA treated wood is disposed of in C&D landfills, it is important to investigate the fate and transport of arsenic in unlined C&D debris landfills and the soil underneath. Arsenic leaching from CCA treated wood in C&D debris landfills was not considered to cause a potential health issue in the past, because it was thought that arsenic (arsenate or arsenite) could be immobilized either by precipitation or adsorption.
17 Arsenic was thought to be able to form precipitate with sulfide under C&D debris landfill conditions. C&D debris landfills are usually characte rized by high hydrogen sulfide concentr ation (up to 3000 ppm by volume) in landfill gas and high sulfide concentration (greater than 100 mg/L) (Dubey, 2005; Jambeck, 2004) in landfill leachate. Labo rato ry C&D debris landfill studies found that the leachate usually has near neutral pH (~6 to 7) and low ORP values (~ 300 mV to 500 mV) (Dubey, 2005; Jambeck, 200 4) According to the Eh pH diagram in aqueous phase by Brookins (1988) arsenic sulfur minerals such as orpiment (As 2 S 3 ) or realgar (AsS) could form under those conditions. Additionally, it is believed that a rsenic can be immobilized by soils underneath unlined C&D debris landfills. It is well known that both arsenate and arsenite can be adsorbed onto iron oxide minerals, which are ubiquitous in soils. Arsenic levels in groundwater may not necessarily b e high when the soil arsenic content is high, due to the adsorption of arsenic on iron minerals in soils or the formation of iron arsenic coprecipitate such as Scorododite. Based on the above 2 aspects of reasoning, it would be expected that arsenic leache d out from CCA treated wood could be immobilized in C&D debris landfills, either by forming arsenic sulfide precipitates or by adsorption on soils underneath landfills. So there would be no significant arsenic leaching from C&D debris landfills. However, r esults from recent studies on C&D debris landfills showed that this is not necessarily true. Problem Statement Recent studies have shown evident arsenic leaching in C&D debris landfills and impact on groundwater. In laboratory simulated C&D debris landfill studies, arsenic
18 concentration in leachate was increased due to the co disposal of CCA treated wood (Dubey, 2005; Jambeck, 2004; Jang, 2000; Weber et al., 2002) The average arsenic concentration s (0.1 2.5 mg/L) in the leachate from the experimental laboratory C&D debris leaching columns (lysimeters) w ere much higher than the U.S. drinking water standard (1 G roundwater from monitoring wells in some C&D debris landfills in Florida has also shown the increa sed arsenic concentration s (57 and 40 g/L in one landfill site) due to the disposal of CCA treated wood in unlined landfills (Khan et al., 2006a) Even though the total number (1~3 out of 21) of arsenic impacted sites currently seems to not be large considering the continuously increasing disposal of CCA treated wood and the time needed for arsenic transport in landfills and soils, significant impact will possibly be seen in the near future (Jambeck, 2004) The somewhat une xpected results of arsenic leaching in C&D debris lysimeters are very likely related to the formation of anionic arsenic sulfide (thioarsenic) species. Under high sulfide concentration, orpiment solubility has been shown to increase, possibly due to the fo rmation of thioarsenic anions (Eary, 1992; Wilkin et al., 2003) Both theoretical and experimental results have suggested the existence of thioarsenic anions under reducing conditions with high sulfide concentratio n (Helz et al., 1995; Lee, 2005; Stauder et al., 2005; Wallschlager and Stadey, 2007; Wilkin et al., 2003) It is reasonable to presume that thioarsenic anions are formed in the leachate of C&D debris landfills, wh ich is usually characterized by high sulfide concentration and low redox potential. The formation of thioarsenic anions was also suggested to enhance the mobili ty of arsenic due to their poor adsorption on iron oxide minerals (Lee, 2005)
19 On the other hand, studies (Lovley, 1991; Lovley et al., 2 004) have indicated that iron oxide minerals, in which iron is in its oxidized form (Fe 3+ Ferric ion), can be reductively dissolved under anaerobic biological reducing conditions. High levels of arsenic in groundwater have been found to be associated wit h the reductive dissolution of iron oxide minerals in soils. Biological reducing conditions will develop when the oxygen in soils decrease s to a low level, which may be caused by a disruption of natural oxygen diffusion. A landfill may play a role cutting off or dramatically decreasing oxygen diffusion rate in soils. Iron oxide minerals have been shown to be reductively dissolved in the presence of organic matter under biological reducing condition. This process is called iron reductive dissolution (IRD). A s IRD proceeds, ferric iron is converted into ferrous iron and is released into the soil aqueous phase. Numerous studies have found that arsenic species previously bound to iron oxide minerals can be released into soil solutions and increase the ir level in groundwater (Cummings et al., 1999; Nickson et al., 2000) On the other hand, ferrous Fe(II) can react with sulfide under anaerobic condition to form precipitate, such as pyrite (FeS 2 ) or troilite (FeS), which have been reported to adsorb arsenic (Bostick and Fendorf, 2003) Arsenic can also co precipitate with iron under sulfidic condition to form precipitate such as arsenopyrite. The formation of precipitate redu ces the aqueous arsenic concentration. As arsenic containing leachate infiltrate s through soils underneath a C&D debris landfill the mobility of arsenic is thus affected by both sulfide in the leachate and iron minerals in the soil. As seen in the discuss ion so far i t is critical to investigate the speciation and adsorption behavior of the potentially formed thioarsenic anions in C&D debris landfills leachate. The interrelation among arsenic, sulfide and iron minerals regulates arsenic
20 retention and mobil ity. The mobilized arsenic may eventually enter into groundwater. Due to the serious health problems arsenic may cause even at very low concentrations, it is important to study how arsenic in C&D debris landfill leachate retains and mobilizes in the soils (particularly iron containing soils) underneath landfills. Objectives and Research Approaches The goal of this dissertation was to explore the fate and transport of arsenic in C&D debris landfills. As seen above, the co disposal of gypsum drywall in C&D de bris landfills may significantly affect the speciation and mobility of arsenic in rather complicated fashions. Therefore, the specific objectives are: Develop ing a speciation methodology and i dentify thioarsenic anions in C&D debris landfill leachate. The method was calibrated and validated. Inv estigating the effec t of gypsum drywall disposal on arsenic leaching and speciation in laboratory C&D debris landfills. Comparing the adsorption isotherms and partition coefficients between thioarsenates and arsenate /arsenite on hematite coated quartz sand. Investigat ing arsenic removal in hematite coated sand columns affected by leachate with different sulfide levels under C&D debris landfills. Objective 1 The first objective was to d evelop a speciation methodology t o identify thioarsenic anions in C&D debris landfill leachate. The m onothioarsenate compound was synthesized according to relevant literature. Molecular and elemental mass spectrometry techniques were used to characterize and confirm the structure s of th io arsenate anions A n ion chromatographic method for the identification and quantification of thioarsenates was developed and validated The formation and stability of thioarsneate anions was also investigated.
21 Objective 2 The second objective was to i nvesti gate the impact of gypsum drywall percentage on arsenic leaching and speciation in laboratory C&D debris landfills. Leaching columns (lysimeters) simulating C&D debris landfills were built and loaded with laboratory made C&D debris waste. Various amount s o f gypsum drywall w ere used in lysimeters to evaluate the impact of the percentage of gypsum drywall on arsenic speciation and concentration in C&D debris landfill leachate. Objective 3 The third objective was to c ompare the adsorption isotherms and partit ion coefficients between thioarsenates and arsenate/arsenite on iron oxide coated quartz sand. Batch experiment s were conducted using hematite coated sand. Adsorption experiment s were conducted with pH values of 5, 7, and 10 Adsorption isotherms and parti tion coefficients were obtained and compared. In the case when the thioarsenate model compounds were unstable or hard to synthesize or purify, an alternative way of studying adsorption was pursued in which arsenite in sulfide solutions was used to simulate condition s where the sulfide concentration is high and thioarsenic species will potentially form. Objective 4 The fourth objective was to i nvestigate arsenic removal in the hematite coated sand columns affected by different sulfide levels in the leachate under C&D debris landfills. Four transparent acrylic columns, each with a n inner diameter of 2.5 cm and a length of 30 cm were packed with hematite coated sand Leachate drained directly from the laboratory C&D debris landfills in Objective 2 was pumped t hrough the sand columns. Effluent from the columns w as collected in a nitrogen atmosphere and
22 analyzed for total arsenic and iron concentration s The relation ship between iron, sulfide, and arsenic in the effluent w as studied. The oxidation state change of iron on coated sand before and after the operation of the columns was analyzed. Outline of the Dissertation This dissertation consists of 6 chapters. The first chapter is the introduction, which describes the overall background, the motivation and objecti ves and approaches of this research. From chapter 2 through chapter 5, each chapter covers an individual topic which aims to answer the questions in each objective. Chapter 6 gives the summary and conclusion s Chapter 1: Introduction Chapter 2: Identificat ion and quantifica tion of thioarsenic species in C & D debris l andfill leachate Chapter 3: Arse nic leaching and speciation in C & D debris landfills affected by different amounts of gypsum drywall Chapter 4: Adsorption of arsenic and thioarsenic species on h ematite coated sand Chapter 5: Arsenic and iron mobilization in hematite coated sand affected by different levels of sulfide in C & D debris landfill leachate Chapter 6: Summary and c onclusion s
23 CHAPTER 2 IDENTIFICATION AND Q UANTIFICATION OF THI OARSENIC SP ECIES IN C ONSTRUCTION AND D EMOLITION DEBRIS LANDFILL LEAC HATE Introduction The research described in t his chapter aimed to develop a simple chromatographic speciation methodology for determination of th ioarsenic anions in laboratory construction and demoli tion (C &D ) debris landfill leachate. The underlying hypothesis is that under the conditions (biological reducing condition with high concentrations of sulfide) prevalent in C&D debris landfills, thioarsenic anions will form in the leachate and can be detec ted by ion chromatography. C&D debris landfills are usually characterized by a high sulfide concentration in leachate and high hydrogen sulfide concentration in landfill gas. The high concentration of sulfide w as thought to be one of the most important fac tors that would contribute to the immobilization of arsenic species in C&D debris landfills, presumably by forming insoluble arsenic sulfide compounds, as the result of the reaction between arsenate/arsenite and sulfide (Brookins, 1988) However, several recent studies (Stauder et al., 2005; Wallschlager and Stadey, 2007; Wilkin et al., 2003) have shown the formation of soluble arsenic sulfide (thioarsenic) anions, from both labo ratory chemical reactions and real groundwater samples at contaminated sites. On the other hand, theoretical calculations showed that the formation of thioarsenic from the reaction between arsenate/arsenite and sulfide is thermodynamically plausible. Meanw hile, although there is no enough leachate quality data from C&D debris landfills, several laboratory simulated C&D debris landfills (lysimeters) studies have shown the evident leaching of arsenic; and the arsenic concentration in leachate was positively r elated with the amount of CCA treated wood in the C&D debris.
24 In natural water systems, the most common arsenic species are arsenite (dissociated anion of arsenious acid H 3 AsO 3 ) and arsenate (dissociated anion of arsenic acid H 3 AsO 4 ). In reducing conditio ns, arsenite is more thermodynamically stable; while arsenate is more stable in oxidizing conditions. The acid dissociation constants for arsenic acid are: pK a1 = 2.20, pK a2 = 6.97, pK a3 = 11.53; while for the arsenious acid, they are: pK a1 = 9.22, pK a2 = 12.13, pK a3 = 13.40. Under usual soil and groundwater envornmental pH (~5 8), arsenate exists as the dissociated forms H 2 AsO 4 and HAsO 4 2 ; while arsenite exists mainly as the neutral form H 3 AsO 3 The toxicities of arsenite and arsenate are very different in that the former is much more toxic than the latter. As a result, the speciation analysis is of critical importance in evaluating the impact of arsenic mobil ization in natural environment. The arsenic speciation in simulated C&D landfills leachate was co nducted using high performance liquid chromatography (HPLC) coupled with ICP MS (Jambeck, 2004; Khan et al., 2006a) It was found that As(V) was the main species, followed by methyl arsenic acid ( MA ) and As(III). While the analysis of municipal solid waste (MSW ) lysimeters leachate showed that As(III) predominated in most of the experiment. Both lysimeters were in anaerobic reducing conditions characterized by low ORP values of leachate. Two major differences were the acidic pH (<5.0) in the leachate from the MS W lysimeter and higher sulfide concentration in the leachate from the C&D debris lysimeter. The prevalence of the more oxidized form As(V) in the C&D debris lysimete r leachate seemed to be contradictory to what was expected under reducing conditions. The p otential existence of thioarsenic anions may provide more explanations to previous leaching and speciation results.
25 The assertion of the existence of thioarsenic species is primarily based on thermodynamic considerations of the solubility measurements of solutions in equilibrium with orpiment (As 2 S 3 ) (Eary, 1992; Webster, 1990) Molecular orbital calculations and Raman spectroscopic data have also confirmed their existence ( Helz et al., 1995; Tossell, 2001; Wood et al., 2002) However, natural geochemical conditions are far different than those with saturated arsenic sulfide under lab conditions R esearch on the identification of thioarsenic species in natural sulfate reduci ng conditions is still very rare. Thermodynamic properties of thioarsenic species have been studied using both experimental and theoretical methods. Using molecular orbital theory, Helz et al. (1995) calculated the bond distances, vibrational frequencies, gas phase energetic and proton affinities for various thioarsenite anions and molecules. It was shown that in gas phase, both monomeric and t rimeric thioarsenite were favorable in the conversion from As(OH) 3 to thioarsenite. However, monomeric thioarsenites, As(SH) 2 and AsS 2 (SH) should be prevalent in sulfidic natural waters undersaturated with As 2 S 3 solid phase. Experimental and theoretical calculation results by Tossell (2001) and Clarke and Helz (2000) showed the existence of a very stable thioarsenite AsS(OH)(SH) copper(I) complex. While the corresponding As(V) complex is much less stable. The theoretical results, together with previous solubility and spectroscopy stud ies seem to support the existence of thioarsenites over thioarsenates. However, recent speciatio n study using ion chromatography and mass spectrometry favored the formation of thioarsenate in sulfidic waters (Schwedt and Rieckhoff, 1996; Stauder et al., 2005; Wallschlager and Stadey, 2007) The benefit of spe ctroscopic methods is that they can monitor oxidation
26 state change; however, they cannot monitor individual species in a complex system. By using chromatographic methods, individual species can be separated and identified. Traditionally, there are two typ es of arsenic speciation methods: chromatographic method and hydride generation method. In hydride generation method, which is the foundation of the EPA Method 1632, arsenic is converted to volatile arsenic hydride (AsH 3 ). Either atomic absorption spectros copy (AAS) or inductively coupled plasma mass spectrometry (ICP MS) is used to measure arsenic after conversion. In chromatographic method, high performance liquid chromatography (HPLC) is used to separate different arsenic species, including arsenite, ars enate, and several organic arsenic species. Each separated arsenic species is then measured by elemental analysis method such as ICP MS. With neither method however, thioarsenic species cannot be detected specifically. Recently, i on chromatography on line coupled with inductively coupled plasma mass spectrometry (IC ICP MS) has been used in identifying thioarsenic species. IC is used to separate thioarsenic anions, which are assumed to have different retention ability in resin columns. On line ICP MS is us ed as a sensitive elemental detector to detect the separated anions. Schwedt and Rieckhoff (1996) firs tly separated monothioarsenates AsO 3 S 3 and arsenate AsO 4 3 using ion chromatography. Stauder et al. (2005) used IC ICP MS analyzed groundwater at an arsenic contamina ted site and found significant amount of thioarsenate besides arsenite and arsenate. They proposed the disproportion mechanism to explain the formation of As(V) thioarsenates in reducing conditions. Wallschlager and Stadey (2007) also identified thioarsenate anions in sulfide arsenit e mixtures using IC ICP MS. They also confirmed the formula of
27 thioarsenate anions by using molecular mass spectrometry (ESI TOF MS) to obtain the accurate molecular weight. However, they were not able to identify thioarsenate species in real environmental samples due to high matrix interferences. In this chapter ion chromatography was used together with both molecular and elemental mass spectrometry techniques to identify thioarsenic species. Thioarsenate model compounds as IC standards were synthesized a nd characterized by mass spectrometry. A simple gradient elution ion chromatography method was developed with preliminary quantification being conducted. Materials and Methods Synthesis of Thioarsenic Compounds Monothioarsenate Na 3 AsO 3 2 O : The procedure was modified from literature (Schwedt and Rieckhoff, 1996) A Schlenk round bottom 2 neck flask wa s charged with 5 g (0.05 mol As) As 2 O 3 and 6 g (0.15 mol) NaOH. 20 ml de ionized water is added to the flask and mixture is stirred to dissolve and bubbled with nitrogen. Then 1.6 0 g (0.05mol) sulfur wa s added to the flask. The flask is then flushed with nitrogen and kept un der nitrogen atmosphere. The solution wa s heated to 100 o C and stirred and refluxed for 2 hours. The color of the mixture turned brown, and then yellow to almost colorless. The mixture wa s then filtered in ho t to remove excess sulfur The filtrate was crys tallized by methanol diffusion method in which the low density solvent methanol was added slowly on top of the aqueous filtrate layer Methanol, in which thioarsenate has lower solubility than in water diffused into aqueous layer and c olorless needle sha ped crystals were obtained. Dithioarsenate Na 3 AsO 2 S 2 2 O : A Schlenk round bottom 2 neck flask wa s charged with 5 g (0.05 mol As) As 2 O 3 and 6 g (0.15 mo l) NaOH. 20 ml de ionized water
28 wa s added to the flask and mixture wa s stirred to dissolve and bubbled with nitrogen. Then 3.20 g (0.10 mol) sulfur wa s added to the flask. The flask wa s then flushed with nitrogen and put under nit rogen atmosphere. The solution wa s heated to 70 o C and stirred for 2 day s. The mixture wa s then filtered in hot to remove excess sulfur and the filtrate slowly cooled to 4 o C. Trithioarsenate Na 3 AsOS 3 2 O : (Stauder et al., 2005) A Schlenk flask wa s ch arged with 5 g (0.05 mol As) As 2 O 3 and 6 g (0.15 mol) NaOH. 20 ml de ionized water wa s added to the flask and mixture is stirred to dissolve and bubbled with nitrogen. Then 4.80 g (0. 15 mol) sulfur wa s added to the flask. The flask wa s then flushed with nitrogen and put under nitrogen atmosphere. The solution wa s heated to 100 o C a nd stirred for 1 hour. The mixture wa s then filtered in hot to remove excess sulfur and the filtrate is slowly cooled to 4 o C. Tetrathioarsenate Na 3 AsS 4 2 O : (Staude r et al., 2005) A Schlenk flask wa s charged with 9 g (0.04 mol S) Na 2 S and 1.21 g (0.038 mol S) sulfur. 10 ml de ionized water is added to the flask and mixture wa s stirred under warm heating to dissolve and bubbled with nitrogen. Then 25 ml de ionized wa ter and 2.66 g (0.025 mol) AsS wa s added to the flask. The flask wa s then flushed with nitrogen and put under nit rogen atmosphere. The solution wa s heated to 100 o C and stirred for 30 minutes. The mixture wa s then filtered in hot the filtrate is slowly coo led to 4 o C. Characterization of Thioarsenic Compounds Several synthesis experiments have been tried. For the synthesis of monothioarsenate, pure crystal was obtained after crystallization For other thioarsenates, no pure compounds were successfully synth esized and purified. In this case, the mixture solution after synthesis was diluted and thioarsenate components
29 were separated using ion chromatograph. For the solutions of either pure crystals or synthesis mixture, e luted fractions corresponding to each p eak in the chromatogram were collected and further characterized using mass spectrometer. Here the ion chromatograph was used to separate each potential thioarsenate anion, and the mass spectrometer was used as the tool for molecular formula identification by measuring their exact molecular weight or the ratio of arsenic and sulfur The assumption was that different thioarsenate anions have different retention ability on the column and each IC fraction represents one potential thioarsenate compound. Ion chr omatography The instrument used was a DX 500 ion chromatograph (Dionex, Sunnyvale, CA) equipped with GP40 gradient pump and CD20 conductivity detector. The data acquisition software was PeakNet 5.2. T he work flow of IC was shown in Figure A 1 in Appendix A Sample solutions from synthesized thioarsenic compounds (or synthesis mixture, in the case where separation or crystallization of pure compounds was unsuccessful) were injected into anion separation column (IonPac AS16, 4 mm 250 mm, Dionex, Sunnyvale, CA). A guard column (IonPac AG16, 4 mm 50 mm, Dionex, Sunnyvale, CA) was connected in the eluent line before the separation column to prevent sample contaminants from entering the separation column. Sodium hydroxide was used as the eluent. An anion suppr essor (Ultra II ASRS, Dionex, Sunnyvale, CA) was used to suppress the background conductivity by converting anions into their protonated species (a diagram explaining the role of ASRS is shown in Figure A 2 in Appendix A ). External water was supplied to AS RS for suppressor regeneration. Protonated species were then detected by DS3 conductivity detector (Dionex,
30 Sunnyvale, CA). After the separated anions ( protonat ed) species passed through the detector, fractions were collected and further analyzed by either molecular or elemental mass spectrometry, which will be described below. Ion chromatography was run using both isocratic and gradient elution. In the isocratic elution, 35 mM NaOH was used as the only eluent. In the gradient elution, 35 mM NaOH and 70 mM NaOH were used to achieve both good separation and shorter elution time. The detailed ion chromatograph instrument parameters are listed in Table 2 1 Mass spectrometry Molecular mass spectrometry: To determine the formulae of thioarsenic compounds, the pu re compound w as dissolved in water (or the synthesis mixture was diluted) and the solution was injected into IC. T he IC fractions were collected at the corresponding retention times and the samples were analyzed at the Mass Spectrometry Service Lab in the Chemistry Department at the University of Florida Time of Flight (TOF) mass spectrometry (Agilent 6210 Accurate Mass TOF LC/MS system) was used to accurately measure the formula weight of thioarsenic anions using negative ion mode. Time of flight mass spe ctrometry (TOF MS) was used because it can provide accurate mass and direct information on molecular weight. The electrospray ionization (ESI) was chosen since it offers less fragmentation and makes spectra simple to interpret. The parameters for the mass spectrometer are listed in Table 2 2 The isotopic masses of potential thioarsenic anions were compared to the observed masses in the mass spectra in terms of the exact molecular weight and mass accuracy The calculation of mass accuracy (error) is shown i n Appendix A.
31 Elemental mass spectrometry: To determine the arsenic and sulfur ratio in the thioarsenic anionic species, eluted IC fractions were collected at University of Florida Solid and Hazardous Waste Lab (and diluted if necessary) and analyzed with ICP MS (Elan DRCe, Perkin Elmer, Shelton, CT) at Bioinorganic and Environmental Analytical Facility at Florida International University (Miami, FL) for total arsenic and total sulfur simultaneously. To eliminate the interference on As + due to ArCl + the D ynamic Reaction Cell (DRC) system was used and oxygen was employed as the reaction gas. The monitored ion for As and S were therefore AsO + and SO + respectively. Glutathione was used as the standard for sulfur calibration. Arsenic standard solution was pur chased from Fisher. The linear range of As and S were 10 200 ppb and 100 2000 ppb, respectively. Formation and Stability of Thioarsenic Anions To explore the formation potential of thioarsenic anions in aqueous sulfidic system, model reactions were conduct ed by mixing sulfide solutions with arsenite or arsenate solutions. All the solutions were made in an anaerobic chamber in which the oxygen level was less than 0.1% v/v, as measured by a GEM 500 gas analyzer (LandTec, Colton, CA). Deoxygenated nanopure wat er, which was obtained by boiling water and cooling to room temperature while bubbling with ultra high purity (> 99.999%) nitrogen gas, was used to make all the solutions. Sulfide solution was made by diluting 2000 mg/L (actual concentration was 1910 mg/L) sulfide standard in sodium phosphate buffer. The final pH of the sulfide solution was 7.3. Arsenate or arsenite stock solution was spiked into the sulfide solution to achieve the desired concentration of arsenic in the mixture. The mixture was sealed in 4 0 mL VOC vials in the anaerobic chamber and wrapped with aluminum foil to avoid light exposure. The mixed samples were kept in the
32 chamber under nitrogen for future IC analysis. The samples for IC analysis were prepared inside the chamber and analyzed imme diately. The effects of air exposure and temperature on the stability of thioarsenic were also investigated. Since pure thioarsenic compounds (except monothioarsenate) were not available, thioarsenic anion speciation transformation in the model reaction be tween sulfide and arsenite was monitored over time for 3 different scenarios: exposed to air at room temperature, exposed to air in a refrigerator, and under a nitrogen atmosphere. Quantifi cation of T hioarsenates in Ion Chromatography Since on ly monothioa rsenate pure crystal was obtained, no other reference standards were available for the thioarsenates quantification according to normal method Alternatively, the quantification (or semi quantification) of thioarsenates used the method as described below. The thioarsenates synthesis mixture with unknown concentrations of thioarsenates was separated by IC. Fractions corresponding to each thioarsenate were collected, diluted to certain volume, and analyzed for arsenic concentration by ICP MS. Using the initia l injection volume, the arsenic or thioarsenates concentration in the original sample was calculated. The injection volume which was unknown due to the replacement of the original injection loop by a new one, was determined by analyzing the arsenic concen tration in the collected arsenate fraction with known initial injected concentration of arsenic. The method was vali dated in terms of precision, specificity, and linearity. System suitability was evaluated based on some chromatographic parameters. Thioarse nate Analysis in C&D Debris Landfill Leachate Leachate from laboratory simulated C&D debris landfills was collected and sealed in VOC vials. Le
33 particulates. Filtered leachate was further treated with C18 (Bond Elut, Varian Inc., Palo Alto, CA) solid phase extraction (SPE) cartridge to remove organic compounds to prevent the contami nation of the IC separation column due to strong retention of organic compounds on the resins. Leachate was then analyzed as soon as possible after pretreatment. Results and Discussion Characterization of Thioarsenic Compounds Ion chromatography Isocratic elution: The crystal obtained from the synthesis of monothioarsenate was dissolved in water and subjected to IC analysis. It was found that t he crystal is almost a pure compound (a picture of the crystals i s shown in Figure A 3 of Appendix A ) : in the chrom atogram there was only one major peak ( Figure A 4 in Appendix A ). The analysis of this fraction performed afterwards by mass spectrometr y revealed that this peak accounts for mon othioarsenate anion (discussed in the mass spectrometry section below). When r etention time s w ere compared, arsenate eluted at a retention time a little after 6 minutes; while monothioarsenate eluted a f t er almost 10 minutes. There is a tiny peak around 20 minutes, which represents dithioarsenate, as will be discussed later. Since th ere were no other pure thio arsenate compounds available, the mixture of the synthesis was diluted and analyzed by IC. A typical chromatogram is shown in Figure 2 1 Besides the first 2 major peaks (which represent sulfide and arsenate, respectively), there were 4 observable peaks with the last being eluted at after 60 minutes. The fractions corresponding to the 4 peaks were named Fraction 1 through Fraction 4. This indicates that the mixture is composed of several components besides the original reactants a rsenite (and arsenate) and sulfur (sulfide). The a rsenite signal is
34 negligible due to the low conductivity of arsenous acid (H 3 AsO 3 ) The long retention time also indicates that those anions have a stronger affinity with the column compared with that of co mmon inorganic anions such as nitrate or halides. Fraction 1 through Fraction 4 were collected individually and analyzed with mass spectrometry to confirm their formula weights and structure. Gradient elution: In the gradient elution, a low concentration ( 35 mM) eluent was used initially to prevent the peak overlap of monothioarsenate with other common inorganic anions. To separate other thioarsenates which eluted after 20 minutes in isocratic elution, a high concentration (70 mM) eluent was used. Peak broa dening for later eluted components (especially for Fraction 3 and Fraction 4) was reduced and the total elution time was less ( Figure 2 2 ). Molecular mass spectrometry Fraction 1: Figure 2 3 shows the mass spectr um of the synthesized monothioarsenate comp ound. A peak at m/z of 156.8936 Da is considered to be the negatively charged diprotic monothioarsenate ion (H 2 AsO 3 S ) which has the theoretical mass of 156.8946 Da. The accuracy ( error of only 6.4 ppm ) in this case indicates a good confirmation of the for mula identity. The most intense peak at m/z of 138.8833 Da is then attributed to mono charged negative ion AsO 2 S which results from the dissociation reaction below: H 2 AsO 3 S AsO 2 S + H 2 O As an important note, the major fraction as shown in Figure A 4 and the Fraction 1 in Figure 2 1 and Figure 2 2 show similar mass spectra, confirming that the component of Fraction 1 is actually monothioarsenate.
35 Fraction 2: The mass spectr um of F raction 2 was shown in Figure 2 4 The peak at an m/z of 172.8719 Da is from the mono charged anion H 2 AsO 2 S 2 (theoretical mass 172.8718 Da). After the loss of one molecule of water, this becomes AsOS 2 (theoretical mass 154.8617 Da) which appears w ith the strongest peak at m/z 154.8616. The perfect mass match confirms that the Fraction 2 was dithioarsenate. If only nominal mass is considered the peak at an m/z of 172.8 Da could also be from H 2 AsS 3 (theoretical mass 172.8540 Da), in which one sulfu r atom substitutes 2 oxygen atoms in H 2 AsO 2 S 2 However, when considering the accuracy, one sees that the error is much higher (103 ppm). So the possibility of being trithioarsenite can be excluded. Fraction 3: Following the interpretation of the Fraction 1 and 2, a reasonable guess is that the components of the other 2 fractions (Fraction 3 and 4) would be trithio and tetra thioarsenates, respectively. In the mass spectrum of Fraction 3 ( Figure .2 5 ), however, the expected trithioarsenate molecular ion (H 2 AsS 3 O theoretic mass 188.8489 Da) peak was not found. The most abundant ion appears at m/z 170.8380 Da, which matches the theoretical mass of AsS 3 (170.8384 Da). The mono charged negative ion AsS 3 is very likely produced from the dissociation of diprot ic trithioarsenate (H 2 AsS 3 O ) or tetrathioarsenate (H 2 AsS 4 ) by losing either one water molecule (H 2 O) or one hydrogen sulfide molecule (H 2 S), even though there was no molecular ion of either H 2 AsS 3 O or H 2 AsS 4 observed: H 2 AsS 3 O AsS 3 + H 2 O H 2 AsS 4 AsS 3 + H 2 S The failure to observe the molecular ion of either H 2 AsS 3 O or H 2 AsS 4 could be due to the instability of these two anionic species during sample preparation or when samples
36 were analyzed by mass spectrometer. The operat ing conditions of the mass spectrometer might also not optimized to analyze the samples. Fraction 4: The mass spectrum (not shown) of this fraction showed a similar pattern to that of trithioarsenate. No molecular ion of tetrathioarsenate (H 2 AsS 4 theoret ical mass 204.8261 Da) was observed. Instead, the putative fragment ion AsS 3 (theoretical mass 170.8384 Da) was the most abundant peak observed. As will be seen in Chapter 3, the species of Fraction 4 is very unstable in air, even at relativ ely low temper ature (4 o C to 25 o C ). Due to this instability, it is hard to prepare samples and analyze them using the current ly adopted mass spectrometric method. Extreme care has to be taken to guarantee a minim al exposure of the sample to air. For example, deoxygenat ed nanopure water may be required as the water for suppressor regeneration and to make low oxygen eluent. The collection of the fraction is better done under the protection of a nitrogen atmosphere. Samples may be prepared shortly before the mass spectrome try analysis, in which milder conditions will be needed to prevent any significant degradation of the species. Wallschl ger and Stadey (2007) also investigated thioarsenates identity by using tandem mass spectrometry, in which method the molecular ion was selected and further fragmen ted to observe fragment ions from the selected molecular ion. This method can confirm the molecular structure based on fragmentation pathways in addition to the accurate m/z values. In their study, the mixture of sulfide and arsenite were analyzed by on li ne coupled ion chromatography mass spectrometry. According to their mass spectra results, formulae of both monothioarsenates and dithioarsenates were confirmed, since the accuracy (error) values were within 5 ppm. They also concluded
37 that the other two thi oarsenates were tritio and tetrathioarsenates. However, the m/z values of the molecular ions as well as the fragment ions for both trithio and al mass. The observed molecular ion m/z values they provided were 188.8627 Da for trithioarsenate (theoretical mass 188.8489 Da) and 204.8330 Da for tetrathioarsenate (theoretical mass 204.8261 Da), respectively. The errors for these two formulae assignment were as high as 73.1 ppm and 33.7 ppm, respective ly. Elemental mass spectrometry Eluted fractions collected from IC separations were also analyzed by ICP MS. The objective was to obtain the atomic ratio between sulfur and arsenic in thioarsenates by measuring the two elements simultaneously. As seen in T able 2 3 in Fraction 1 and Fraction 2, the S/As ratio matched well with what was expected. The S/As ratio is around 1 in Fraction 1 and it is around 2 in Fraction 2. These ratios, combin ed with the molecular mass spectrometry results, confirm that the com ponents in Fraction 1 and Fraction 2 were monothioarsenate and dithioarsenate, respectively. For Fraction 3 and Fraction 4, however, the measured S/As ratios deviate from the expected values with high variance. As mentioned earlier and as will be discussed later, trithioarsenate and (especially) tetrathioarsenate are not stable. During the time period from sample collection to sample analysis, significant speciation conversion may occur, which may affect the elemental analysis result. For example, sample ac idification prior to elemental analysis may cause the precipitation of As 2 S 3 (Stauder 2005 ), which will inevitably affect the final results. On line coupled IC ICPMS was used to confirm the structure of thioarsenates (Stauder et al., 2005; Wallschlager and Stadey, 2007) Their results showed a good
38 match between the measured and the expected values of S/As ratios. Since on line systems were used (thioarsenates were separated by IC and S/As concentrations measured di rectly by ICP MS without fraction collection), the potential of speciation conversion was eliminated or limited to the minimum. In summary, ion chromatography proved capab le to separat e all the thioarsenate anions. The components of the four fractions (Fra ction 1 through Fraction 4) are monothio through tetrathio substituted arsenates, according to the current and literature results. Factors Influencing Thioarsenic Formation Thioarsenate speciation change over time Both arsenite and arsenate react with su lfide and produce thioarsenate anions under the conditions investigated. However, arsenite showed faster kinetics than arsenate. As shown in Figure 2 6a the distribution of different thioarsenates in arsenite after mixing. However, for the arsenate and sulfide system, the speciation distribution between 2 days and 5 days is very different (Figure 2 6b ). Additional thioarsenates were even forming after 60 days. This indicates the conversion from arsenate to thio arsenates is much slower than the conversion from arsenite to thioarsenates. If arsenite is the reaction intermediate in the reaction between sulfide and arsenate, the slower kinetics may be due to the conversion from arsenate to arsenite. It is known that sulfide has a great tendency to reduce arsenate to arsenite (Nordstrom and Archer, 2003) With the fast reaction kinetics observed (Rochette et al., 2000) especially at acidic pH, the pathway via arsenite in the thioarsenate formation reaction between arsenate and sulfide seems
39 questionable. In fact, mechanisms involving simple ligand exchanges are more popu lar (see references in Rochette 2000). H 2 AsO 4 + H 2 2 AsO 3 S + H 2 O On the other hand, the mechanism accounting for the conversion from arsenite to thioarsenate is more interesting and still not well accepted. Stauder et al. (2005) proposed a disproportionation mechanism in which As(III) is oxidized to As (V) and reduced to As(0): 5H 3 AsO 3 + 3H 2 2 AsO 3 S + 6H 2 O + 3H 2 O The detailed discussion of this mechanism is beyond the scope of this research. References (Helz 2008, Rochette 2000, Stauder 2005) may be referred to for further information. Concentration effect on thioarsenate formation The relative concentrations of sulfide and arsenic (arsenite or arsenate) determined the percentages of different thioarsenates in the model reaction systems. When arsenic concentrations are the same, a higher concentration of sulfide produced more thioarsenates which have higher S/As ratios. As seen in Figure 2 7 trithioarsena te was the major component when arsenite and sulfide were 0.1 mM and 1 mM, respectively. However, when sulfide was 10 mM, the major species was tetrathioarsenate. This trend is similar to what was observed by Satuder et al. (2005), Wallschlager (2007), and Wilkin (2003), even though in the latter case thioarsenites rather than thioarsenates were believed to form. Stability of T hioarsenic S pecies As observed in this investigation as well as in the literature, thioarsenates are not stable. The stability of th ioarsenates was investigated by monitoring speciation change
40 over time in model reactions as well as in C&D debris landfill leachate. Two factors, air exposure and temperature, were investigated. The samples used were the mixtures of sulfide and arsenite, which were made under a nitrogen atmosphere in a glove box. The mixtures were kept in a glove box for 40 days to reach equilibrium before stability monitoring. The hypothesis was that there would be no further speciation change if the mixtures were stored under a nitrogen atmosphere for 40 days after initial mixing. Figure 2 8 compares the stability of thioarsenates (or speciation transformation) under 3 scenarios: under N 2 at room temperature (r.t.), exposure to air at room temperature, and exposure to air when refrigerated. At day 0 (started under N 2 and r.t.), the dominant species were trithioarsenate and tetrathioarsenate, with the latter being the major component. A minor amount of monothioarsenate was also present. After 1 day of exposure to air at roo m temperature, most tetrathioarsenate was converted to trithioarsenate and monothioarsenate. Even for the refrigerated sample, 1 day of exposure to air caused a clear conversion from tetrathioarsenate to trithioarsenate and monothioarsenate. After 7 days e xposure to air, both at room temperature and under refrigeration, tetrathioarsenate completely disappeared and monothioarsenate levels were also significantly diminished. For the sample stored under nitrogen, even after 14 days at room temperature, there w as no evident species transformation and the IC chromatogram was almost identical to the original one at day 0. However, a slight shift in retention time was observed, which might be caused due to the replacement of new eluent and minor operation condition variation. After 14 days in air, the amount of trithioarsenate was significantly less than that after 1 day in air.
41 Quantification of T hioarsenates Calibration Curves: The thioarsenate synthesis mixture with unknown concentrations of thioarsenates was sep arated by IC. Fractions corresponding to each thioarse nate were collected, diluted to a certain volume, and analyzed for arsenic concentration by ICP MS. Based upon the initial injection volume, the arsenic or thioarsenates concentration in the original sa mple was calculated. The injection volume was determined by analyzing the arsenic concentration in the collected arsenate fraction with known initial injected concentration of arsenic. Table A 1 show s all the data used in determining injection volume. Usin g the calculated injection volume, the concentrations of the various forms of arsenic in the original injected thioarsenate solution were calculated and listed in Table A 2. Also listed in the table a re the peak areas of individual thioarsenate s in each in jection. The average concentration of a specific thioarsenate in 3 injections was used as the nominal sample concentration. Calibration was done by using the two point method with the assumption that the curve passes through the origin. Calibration curves for thioarsenates are shown in Figure 2 9 The horizontal axis represents the arsenic concentration, and the vertical axis represents the peak areas in IC chromatograms. In the IC analysis, the conductivity detector measures the electrolytic conductivity o f cations or anions of the analyte components. As described in previous section, the analyte anions are converted to their protonated neutral form after they pass through ion suppressor. For the anions of a strong electrolyte, such as chloride or nitrate, their protonated forms (HCl or HNO 3 ) dissociate completely. However, for weak electrolyte anions, such as arsenate or thioarseantes, the percentages of ionic form s are determined by the dissociation constants of the
42 corresponding acids. At neutral pH, for example, arsenic acid dissociates and forms more anions (arsenate) than that of arsenous acid, of which the anion is arsenite. This actually explains why arsenite is hard to detect in IC using conductivity detector. At the same arsenic molar concentration, arsenite has much lower conductivity response than that of arsenate. Therefore, the slope of the calibration curve may to some extent reflect the dissociation ability of the corresponding thioarsenic acid, assuming that the molar ionic conductivity of all the thioarsenate anions and arsenate anion are similar. According to the calculated or measured pK a values (Helz and Tossell, 2008) in Table 2 4 monothioarsenic acid (H 3 AsSO 3 ) is a slightly weaker acid (with a higher pK a ) than arsenic acid. All other 3 thioarsenic acids were predicted to be strong acids with negative pK a1 values. In this research, sim ilar trend was observed for the acidity of thioarsenic acid, i.e., much higher values of slopes (conductivity responses) from the calibration curves were found for dithioarsenate throught tetrathioarsenate, which implied that di tri and tetra thioarsen ic acids may be stronger acids that arsenic acid. The fact that thioarsenic species have stronger acidities than arsenic species was explained by their differences in proton affinity (Helz and Tossell, 2008) A s ulfur atom has less proton affinity than an oxygen atom. Hydrogen atoms in either thioarsenic acid or arsenic acid molecules are preferentially attached to oxygen atoms (forming hydroxyls rather than thiols). The magnitude of pKa values are determined by the number of sulfur and oxygen atoms in the molecule. For example, in each of these anions, HAsS 2 O 2 2 HAsSO 3 2 and HAsO 4 2 there is an unpro tonated oxygen atom. The association of the second proton (to the second oxygen atom) should occur at almost
43 the same pH as the first This means that all three acids, H 3 AsS 2 O 2 H 3 AsSO 3 and H 3 AsO 4 have almost the same pKa 2 Accuracy: Accuracy measures t he closeness of the agreement between the value found by a given method and the actual value as represented by the reference standard. Since there are no pure synthesized standards for the compounds analyzed in this investigation (except for monothioarsena seen above, quantification was based on the match between IC peak areas and the elemental analysis results from ICP MS. Therefore, the accuracy of ICP MS analysis, including all the steps from fraction collection to I CP MS calibration, determined the accuracy of this method. In addition, the original synthesis mixture that was used to obtain each fraction may have b e en too concentrated thus possibly overloading the column and leading to peak tailing. As a result, the original concentration may be out of the linear dynamic range. This will be discussed further below. Precision: The precision is defined as the agreement among individual analysis results from repeated analyses. The precision can be measured by three metho ds: repeatability, intermediate precision, and reproducibility. Repeatability represents the ability to give the same results during a short period of time (for example, same day) under identical conditions. Intermediate precision deals with variations of the results between d ifferent days of analysis, analysts, or equipments. Reproducibility refers to the agreement between results from different laboratories. In this experiment, repeatability based on 6 consecutive analysis runs at 2 concentration levels i s shown in Table 2 5 Peak areas of only three thioarsenates were listed, since no tetrathioarsenate was observed in these 2 samples. Relative standard deviation (RSD) for all components was
44 around 0.05, which is not very low in terms of quantitative analy sis. The relatively high prepared at the beginning of analysis Specificity: Specificity refers to the ability to accurately measure the specific analyte of interest in th e presence of other components which may be expected to exist in the sample. For identification purposes, specificity may be demonstrated by separat ing an analyte from other components in the sample. In the current gradient method, all thioarsenate anions are eluted after 9 minutes, which is longer than many common inorganic anions. Phosphate has a retention time at around 11.5 minute, which is longer than that of monothioarsenate but shorter than that of other thioarsenate anions. For C&D landfill leachate specifically, baseline separations from major anions such as sulfate and sulfide can be achieved for these thioarsenates, especially for di through tetra thioarsenate anions. Th is method, however, is not suitable for the analysis of common inorganic anio ns, such as bromide, chloride, floride, nitrate, sulfate, or sulfide, since these anions are all eluted quickly (before 8 minutes) and peak overlap occurs. Linearity and Range: Linearity represents the ability of the method to give the results that are pro portionally related with analyte concentration within a certain range. Since pure standards were not available, the linearity was tested using 4 sets of diluted solutions from synthesis mixture As shown in Figure 2 10 within the concentration ranges test ed, trithioarsenate and dithioarsenate showed good linearity with R 2 greater than 0.99. For monothioarsenate, however, the most concentrated solution showed a large deviation from the prediction based upon extrapolating low er concentration value s.
45 This ind icates that high concentration causes column overloading and peak broadening, which affects the accuracy of quantification. For trithioarsenate and dithioarsenate, the linear range can be estimated to be within 0.1 to 1.8 mg As/L, and 0.2 to 5.0 mg As/L, r espectively. Thioarsenate Analysis in C&D Debris Landfill Leachate Due to the disposal of gypsum drywall and CCA treated wood in C&D debris landfills, leachate produced usually has high concentration s of sulfide and a significant amount of arsenic (Dubey 2 005). Arsenic speciation in landfill leachate has been previously investigated using HPLC coupled with ICPMS (Khan et al., 2006a; Ponthieu et al., 2007) It was found that As(V) was the dominant arsenic species in the leachate of C&D debris landfill with CCA treated wood (Khan et al., 2006a) However, that study did n o t distingu ish among all possible As(V) species, such as arsenate and various thioarsenate anions. Figure 2 11 shows chromatograms of the leachate collected from laboratory simulated C&D debris landfills which contain both gypsum drywall (1 wt%, 12.4 wt%) and CCA tre ated wood (10 wt%). Besides arsenate, several other arsenic containing entities, e.g., mono di and trithioarsenate, have been found in the leachate. Trithioarsenate is the major component among all thioarsenates. Due to the low level of thioarsenic con centrations in the eluted fractions from the leachate, no further mass spectrometric identification was conducted to confirm the structures of the species in leachate. However, leachate was spiked with arsenite and there was significant increase in trithio arsenate peak area after spike, which confirmed that the separated species was due to arsenic.
46 Summary High sulfide levels and significant arsenic leaching in C&D debris landfills necessitate the speciation analysis of arsenic, including potential thioars enic species, which according to the literature, can be formed via the reaction between arsenite/arsenate and sulfide under reducing conditions. To this end the research described in this chapter aimed to develop a simple method based on ion chromatograp hy. Due to the lack of commercial ly available compounds, sodium thioarsenates were synthesized and a pure sodium monothioarsenate crystal was obtained, which was analyzed using molecular mass spectrometry thus confirming its identity. For other thioarsena te anions, i.e., di tri and tetra thioarsenate, the synthesis mixture of thioarsenate was injected into the ion chromatograph and thioarsenate components were collected as separate fractions, which were then analyzed by molecular mass spectrometry. The molecular formulae of thioarsenates were confirmed. A gradient elution ion chromatographic method using sodium hydroxide as a mobile phase and a conductivity detector was established. Compared to the isocratic elution method which required more than 60 min utes for an analysis run, the gradient elution method was able to separate thioarsenate anions within 25 minutes with better peak shapes and enough resolution All thioarsenates were eluted after common inorganic anions, including chloride, bromide, nitrat e, sulfate, sulfide, and phosphate. Monothioarsenate was eluted first at ~10 minute; and tetrathioarsenate was eluted last at ~20 minute. Due to the unavailability of 3 thioarsenate compounds, the method was calibrated based on the arsenic concentration in each fraction obtained from ICP MS and the injection volume. Two point calibration method was used, assuming a linear curve
47 passing through the origin. Higher slopes found in the calibration curves suggested stronger acidities than arsenic acid for all bu t monothioarsenic acid. The method was validated by performing precision test and linearity test. Relatively good precision with RSD around 0.05 was obtained at 2 concentrations Linearity was good for di and tri thioarsenate with R 2 values greater than 0 .99 and a linear range from 0.2 to 5.0 mg As/L and 0.1 to 1.8 mg As/L, respectively For monothioarsenate, severe peak tailing was found when the tested concentration was high, as indicated by the significant deviation of the last point in the linearity c urve. The linear range was from 0.1 to 2.0 mg As/L with an R 2 of 0.91. Using this ion chromatograph y method, simulated landfill leachate was tested and thioarsenate anions were found to be present It was found that their concentrations were positively cor related with sulfide concentrations in leachate Th e method described above was developed to identify thioars e nic species. However, d ue to the low dissociation constant o f arsenous acid, it can not be used to detect arsenite effectively. In order to detect arsenite, other detectors such as electrochemical detector should be used to increase the analyte response. From the chromatographic separation point of view, the method cannot identify organic arsenic species. Currently, there is no existing method that i s capable of detecting arsenite, arsenate, thioarsenates, and organic arsenic species simultaneously. One possible solution to complete separation of all species may be to us e muti dimensional separation techniques, such as 2 D liquid chromatography, in wh ich 2 columns with different physico chemical properties are used
48 Figure 2 1. Ion chromatogram of thioarsenic synthesis mixture: Isocratic elution using 35 mM NaOH There are 4 peaks (1, 2, 3, 4) after the sulfide and arsenate peaks at the beginning. Figure 2 2 Ion chromatogram of thioarsenic synthesis mixture: Gradient elution Four These 4 peaks were confirmed by mass spectrometry results to be due to monothioarsenate through tetrathioarsenate.
49 Fig ure 2 3 Mass spectrum of monothioarsenate compounds (Fraction 1) Molecular ion (H 2 A s SO 3 ) is the second most intensive peak; fragment ion (AsO 2 S ) is the most abundant peak. The observed masses match the theoretical masses very well. The inset table sho ws the theoretical masses and the observed masses of H 2 A s SO 3 and AsO 2 S
50 Figure 2 4 Mass spectrum of ion chromatographic Fraction 2 Molecular ion (H 2 A s S 2 O 2 ) is the second most intensive peak; fragment ion (AsOS 2 ) is the most abundant peak. The obs erved masses match the theoretical masses very well. Fraction 2 is confirmed to be dithioarsenate.
51 Figure 2 5 Mass spectrum of ion chromatographic Fraction 3 Molecular ion (H 3 A s S 3 O ) is not observed; fragment ion (AsS 3 ) is the most abundant peak. T he observed mass of the fragment ion match the theoretical masses very well. Fraction 3 is confirmed to be trithioarsenate.
52 Figure 2 6. Reaction kinetics of thioarsenates formation: (a) 0.1 mM arsenate and 10 mM sulfide, pH 7; (b) 0.1 mM arsenite and 10 mM sulfide, pH 7. Arsenite shows faster kinetics than arsenate.
53 Figure 2 7. Concentration effect on thioarsenates formation. (a): 0.1 mM arsenite; (b): 10 mM sulfide; (c): 1 mM sulfide + 0.1 mM arsenite; (d): 10 mM sulfide + 0.01 mM arsenite; (e): 10 m M sulfide + 0.1 arsenite. At the sulfide to arsenite ratio currently used, monothioarsenate and dithioarsenate only show as minor peaks.
54 Figure 2 8. Stability of thioarsenates in model reactions. Chromatograms show the peak intensity changes over time at room temperature/in refrigerator and in air or nitrogen atmosphere. The peak shift at 17 minute at 14 days is probably due to the change of eluent to fresh one. Thioarsenates are shown to be unstable to air exposure.
55 Figure 2 9 Calibration curve s of thioarsenates and arsenate Due to the lack of available standards, thioarsenate synthesis mixture was injected into IC and each thioarsente fraction was collected to determine absolute arsenic concentration by ICP MS. The slopes for di tri and te tra thioarsenate calibration curves are much larger than that of arsenate implying stronger acidities of their corresponding thioarsenic acids One point calibration was used
56 Figure 2 10 Linearity curves of thioarsenates (a): monothioarsenate; (b) : dithioarsenate; (c): trithioarsenate. Thioarsenate synthesis mixture was diluted into 4 different concentration levels and replicates were analyzed with IC. Horizontal axis represents the relative concentration; vertical axis represents arsenic concentra tion obtained from IC chromatograms. The plot for tetrathioarsenate is not available due to low concentration after dilution. In the plot of monothioarsenate, the significant deviation at the last point indicates severe peak tailing at high concentration l evels.
57 Figure 2 11 IC chromatogram s of C&D debris landfill leachate The leachate from DW12 lysimeter has significantly more trithioarsenate than other lysimeters. The baseline increase between 12 minute and 17 minute is due to the eluent change in the gradient elution. Control : No CCA wood, 12.4% drywall; DW1 : 10% CCA wood, 1% drywall; DW12 : 10% CCA wood, 12.4% drywall.
58 Table 2 1 Ion chromatograph conditions Item Details Columns IonPac AS16/AG16, 4 mm 250 mm (Dionex) Detector CD20 DS3 1 ( Dionex) Pump GP40 gradient pump (Dionex) Autosampler AS40 (Dionex) Eluent 35 mM NaOH, 70 mM NaOH Gradient 0 11.5 min: 35 mM 11.5 15.5min: 35 mM 70 mM 15.5 35 min: 70 mM 35 38 min: 70 mM 35 mM 38 40 min: 35 mM Anion suppression ASRS Ultra II 4 mm (Dionex) Regeneration mode External water addition, 10 mL/min Suppression current 300 mA Sample volume 5 mL Injection volume 70 L Retention times of thioarsenates AsSO 3 3 : 9.38 min AsS 2 O 2 3 : 15.08 min AsS 3 O 3 : 17.68 m in AsS 4 3 : 20.65 min
59 Table 2 2. Typical mass spectrometry conditions Item Details Instrument Aglient 6210 TOF MS Ionization mode Electrospray, Negative mode Fragmentor voltage 175 V Capillary voltage 4000 V Skimmer voltage 65 V Dry ing gas temperature 350 o C Drywall gas flow 10 L/min Injection volume 2 L Table 2 3 Elemental composition of thioarsenic compounds Fraction 1 2 3 4 S/As 0.96 (0.85/1.07/0.95) 2.02 (2.08/2.26/1.73) 3.43 (5.33/1.53) 3.08 (3.95/2.21) Table 2 4 pKa values of thioarsenic acids and arsenic acid (in Helz and Tossel 2008) Acid Dissociation Equation pKa Major Species at pH 7 H 3 A s SO 3 1) H 3 A s SO 3 = H 2 A s SO 3 + H + 2) H 2 A s SO 3 = HA s SO 3 2 + H + 3) HA s SO 3 2 = A s SO 3 3 + H + 3.3 7.2 11 H 2 A s SO 3 HA s SO 3 2 H 3 A s S 2 O 2 1) H 3 A s S 2 O 2 = H 2 A s S 2 O 2 + H + 2) H 2 A s S 2 O 2 = HA s S 2 O 2 2 + H + 3) HA s S 2 O 2 2 = A s S 2 O 2 3 + H + 2.4 7.1 10.8 H 2 A s S 2 O 2 HA s S 2 O 3 2 H 3 A s S 3 O 1) H 3 A s S 3 O = H 2 A s S 3 O + H + 2) H 2 A s S 3 O = HA s S 3 O 2 + H + 3) HA s S 3 O 2 = A s S 3 O 3 + H + 1.7 1.5 10.8 HA s S 3 O 2 H 3 A s S 4 1) H 3 A s S 4 = H 2 A s S 4 + H + 2) H 2 A s S 4 = HA s S 4 2 + H + 3) HA s S 4 2 = A s S 4 3 + H + 2.3 1.5 5.2 A s S 4 3 H 3 A s O 4 1) H 3 A s O 4 = H 2 A s O 4 + H + 2) H 2 A s O 4 = HA s O 4 2 + H + 3) HA s O 4 2 = A s O 4 3 + H + 2.3 6.99 11.8 H 2 A s O 4 HA s O 4 2
60 Table 2 5 Determination of repeatability by replicate injections of the same sample Inje ction Sample 1 (High concentration) Peak Area Sample 2 (Low concentration) Peak Area Monothioarsenate Dithioarsenate Trithioarsenate Monothioarsenate Dithioarsenate Trithioarsenate 1 107864.4 78209.6 54326.6 23992.0 23470.0 24511.0 2 112795.8 81852.9 5 7259.2 24630.4 21653.6 28022.0 3 109036.8 82874.7 58688.6 22204.8 23456.8 25663.0 4 120308.8 84564.9 61348.1 22483.6 20685.1 25716.0 5 118563.9 84943.1 59278.0 23392.8 22055.3 26532.8 6 110404.5 88276.1 60935.0 22956.1 23117.0 25986.5 Mean 113162.4 83 453.6 58639.3 23276.6 22406.3 26071.9 Standard Deviation 5159.0 3379.6 2589.3 920.7 1130.7 1162.1 RSD 0.046 0.040 0.044 0.040 0.050 0.045
61 CHAPTER 3 ARSENIC LEACHING AND SPECIATION IN C&D DE BRIS LANDFILLS AFFEC TED BY DIFFERENT AMOUNTS OF GYPSUM DRYWA LL Introduction As sulfide production in C&D debris landfills has been shown to be due to the degradation of sulfate dissolved from gypsum drywall, common sense would indicate that the drywall percentage in C&D debris landfill would affect the production a nd concentration of sulfide, which would consequen tly affect the speciation of arsenic. Previous laboratory C&D debris landfills studies have shown that the levels of arsenic in leachate positively correlate with the amount of CCA treated wood present in t he waste (Dubey, 2005; Jambeck, 200 4; Jang, 2000) Arsenic speciation study using HPLC ICP/MS indicated that the major form of arsenic in leachate from C&D debris landfills was inorganic As(V) while in MSW landfill leachate inorganic As(III) was the major form (Khan et al., 2006a) Both arsenate and arsenite were shown to react with sulfide under reducing condition s and form various thioarsenic species, including mono di tri and tetra thio species (Stauder et al., 2005; Wallschlager and Stadey, 2007; Wilkin et al., 2003) The identity of the dominating species is dependent on the sulfide concentration. With higher sulfide concentration, higher thio substituted species are formed. The objective of this research was to investigate how the percentage of gypsum drywall in C&D debris landfills affect s the production of sulfide and consequently the leaching and t he speciation of arsenic in leachate. C&D debris lysimeters loaded with lab made waste w ere built and operated under simulated landfill conditions. Leachate was collected and arsenic leaching and speciation was analyzed using the methodology developed in C hapter 2. The hypothesis for this experiment wa s that sulfide levels in
62 leachate are related to the percentage of gypsum drywall present in C&D debris and that different sulfide levels in leachate w ould cause differences in arsenic leaching and speciation. Materials and Methods Waste Composition The composition of the waste in lab oratory simulated landfills was based on an EPA C&D debris characterization report (EPA, 1998) and pre vious C&D debris lysimeter studies in the group (Jang 2000; Jambeck 2004; Dubey 2005). In all lysimeters, the major components included wood (either treated or untreated), concrete, drywall, cardboard, and asphalt roofing. Minor components included copper wire, steel sheet, insulation, etc. The detailed waste composition is listed in Table 3 1. In previous studies (Jang 2000; Jambeck 2004; Dubey 2005), three different percentages (0.1 wt %, 5 wt %, and 10 wt %) of CCA treated wood have been used. The more CCA treated wood was used, the more arsenic was leached. In this experiment, 10 wt % of the waste used in experimental lysimeters consisted of CCA treated wood. Five lysimeters were constructed: Control and 4 experimental ones: DW0, DW1, DW6, and DW12 In all l ysimeters, the percentages of two components vary: CCA treated wood and gypsum drywall. In the Control lysimeter, no CCA treated wood was used and gypsum drywall percentage (12.4 wt %) was the same as that in DW12 lysimeter. In the DW0, DW1, DW6, and DW12 l ysimeters, gypsum drywall accounts for 0 wt %, 1 wt %, 6 wt %, and 12.4 wt % of the total waste, respectively. Glass cullets were added in DW 0, DW1 and DW6 lysimeters, in which the sum of glass cullets and drywall accounted for 12.4 wt % of the total waste. The purpose of using glass cullets in the waste was to keep the percentage of all the components the same while varying only
63 the percentage of gypsum drywall. Also glass has a similar density (2400 2800 kg/m 3 ) to gypsum (2300 kg/m 3 ) and can be considered an inert material in terms of reactivity. All the waste components were size reduced according to previous C&D lysimeter studies. Size reduced components were homogenized in a big plastic tub. The saw dust from the CCA treated wood was used to analyze for ar senic content. Lysimeter Construction and Waste Loading Five lysimeters were constructed and loaded with simulated C&D debris waste in the laboratory. Each lysimeter consisted of 2 caps and one Schedule 40 PVC pipe with a diameter of 15 cm and a length of 1.2 m. A leachate collection port was installed in the center of the bottom cap. In the top cap, a water addition port was installed. Each port Malden, MA, USA). Before a dding the waste, acid rinsed river rock and pea gravel were prepared by soaking them in 1 M HNO 3 overnight, after which they were rinsed in de ionized water u ntil no further pH change was seen A 6 cm layer of river rock was placed in the bottom of each ly simeter, on top of which a 6 cm layer of pea gravel added These 2 layers acted as a drainage layer for leachate collection. The mixed and size reduced waste was added into the columns in 5 lifts. After each lift, the waste was compacted inside the pipe to achieve an average bulk density of about 300 kg/m 3 which is similar to that used in previous studies. The waste layer was approximately 1.2 m in height. On the top of the waste layer, a layer of geonet was placed under another layer of geofabric. Finally a layer of 15 cm glass chips was placed on top of the geofabric. The glass chip layer is designed to assure even water distribution during the addition of
64 water into the lysimeters. A coil of perforated tube is put on top of the glass chip layer and used for water addition. Lysimeter Operation Lysimeters were flushed with nitrogen gas for 30 minutes before the addition of water. Deionized water was added from the top using a peristaltic pump ( Manostat Carter 8/3 Multi Channel Cassette Pump Thermo Scienti fic). During the first 100 days, 1000 mL water was added bi weekly; after that, water was added at a rate of 100 mL/day except during weekends Each time the addition of water lasted for 25 minutes at a flow rate of 4 mL/minute. The water addition rate sim ulated the natural precipitation rate of 11.5 cm/month (4.5 inch/month). Leachate Analysis Leachate was collected bi weekly from the drainage ports located at the bottom of the lysimeters. Samples for elemental analysis were preserved with nitric acid and stored at 4 until acid digestion. Samples used to determine other parameters (pH, conductivity, DO, ORP) were either analyzed right after collection or collected and sealed in amber VOC vials and analyzed within 24 hours after sample collection. Leachate pH was tested using an Orion 5 Star portable meter (Thermo Scientific, Beverly, MA, USA). Leachate oxidation reduction potential (ORP) was tested using an ORPTestr 10 portable meter ( OAKTON Instruments Vermon Hills, IL, USA). Chemical oxygen demand (COD) was measured using the dichromate digestion method (Hach Method 8000) with a Hach DR/4000 spectrophotometer (Hach Company, Loveland, CO, USA). Sulfide was measured using methylene blue method (Hach Method 8131,
65 equivalent to Standard Method 4500 S 2 D) al so with a Hach DR/4000 spectrophotometer. Blanks, replicates, and calibration s were conducted as appropriate. Analysis for total metals was conducted using inductively coupled plasma atomic emission spectrometry (ICP AES) (Trace Analyzer, Thermo Electron C orporation) according to EPA SW846 Method 6010B after leachate samples were digested following EPA SW846 Method 3010A. L eachate A rsenic S peciation A nalysis Before thioarsenic speciation analysis using the ion chromatograph, leachate was first passed throug leachate was further treated with a C18 (Bond Elut, Varian Inc., Palo Alto, CA) solid phase extraction (SPE) cartridge to remove organic compounds and prevent the contamination of the IC sepa ration column with the organic compounds retained in the resins. Leachate was then analyzed within 2 hours after pretreatment, following the speciation method developed in Chapter 2. Briefly, gradient elution was applied by using 35 mM and 70 mM NaOH as th e eluents in a Dionex DX 500 Ion Chromatograph equipped with CD20 conductivity and GP40 gradient pump modules. The detailed IC analysis condition s are listed in Table 3 2. The concentration s of thioarsen ates w ere determined semi quantitatively. Besides thi oarsenic speciation, arsenic cartridges (MetalSoft Center, Piscataway, NJ) were also used to s peciate As(III) and As(V). About 30 mL leachate or chemical solution (diluted if necessary) was injected manually through a cartridge with a plastic syringe at a flow rate around 30 50 mL/min The first 5 mL filtered solution was discarded. The collected filtered solution and the original solution were analyzed for
66 total arsenic concentration. The concentration in the filtered solution represents arsenite (As(III)) concentration. Comparison of Preservation Methods To find out the effect of sulfide on the recovery rate in total arsenic analysis, two sets of experiments were conducted. In one set, 8.0 mg/L and 50 mg/L sulfide solutions were each spiked with 1 mg/L and 5 mg/L arsenic (reference solution made by dissolving As 2 O 3 in HNO 3 ). Half of these samples w ere then adjusted to a pH of over 10 by adding 0.1 mL 50% wt NaOH. To eliminate the formation of As 2 S 3 precipitation or to remove sulfide, 5 mL 30% (v/v) H 2 O 2 sol ution was added. The samples were let stand for 2 hours before adding 1 mL concentrated trace metal grade HNO 3 to bring the pH below 2 again. The samples were refrigerated before further digestion a ccording to regular EPA method. T he other half of the samp les were not treated using the method above. Instead, they were only preserved with nitric acid and refrigerated before further digestion in accord ance with the regular EPA method. In the second set of experiment s leachate collected from Control lysimete r, which wa s supposed to have the minimum amount of arsenic, was used as the matrix and spiked with arsenic. The arsenic spiked leachate samples were then treated using H 2 O 2 as described above. For comparison, half of the arsenic spiked samples were subjec ted to digestion without treatment with H 2 O 2 Arsenic Solubility in Sulfidic Solution: Batch Test Arsenite and various concentrations of sulfide solutions were pH adjusted to neutral with sodium phosphate and they were mixed in glove box under nitrogen atm osphere. The mixture was sealed in VOC vials and wrapped with aluminum foil and left in glove box for a week before filtration with 0.45 m syringe filter. The filtered
67 solutions were analyzed for sulfide, total arsenic, thioarsenate and arsenic catridge s peciation. Results and Discussion Leachate P arameters over T ime Most parameters were found to have r elatively large fluctuations i n the early phase of lysimeter operation (Figure 3 1), followed by steady trends in the later phase. The fluctuations might be lysimeters as well as the less regular water addition schedule during the first 100 days of operation. There was an overall trend of increasing pH for all the lysimeters and the pH values were stabilized between about 6.2 and 6.8, which matches the results from previous studies (Dubey, 2005; Jang and Townsend, 2003) A closer look and comparison of the pH values after 150 days indicates a subtle correlation between the percentages of drywall and the pH of leachate in C&D debris lysimeters. The pH was lower for DW0 and DW1 lysimeters (which had 0% and 1% of drywall) than that for Control DW6 and DW12 lysimeters (which had either 6% or 12.4% of drywall). Considering the facts that the dissolution o f calcium sulfate is a big contributor of dissolved solids in C&D debris landfills (Jang and Townsend, 2003) and the weak acidities of both HSO 4 and HS (biologically reduced from HSO 4 ), it is not surpri sing to see the impact of drywall percentages on leachate pH. The presence of H 2 S ( pK a1 7.0) and HS is probably making C&D landfills a buffered system controlled by sulfide, sulfate, and overall alkalinity The effect of drywall percentage is also seen on conductivity. The conductivity values of leachate from Control, DW1, DW6 and DW12 lysimeters were significantly higher than those of the DW 0 lysimeter, which had no drywall in the waste. When
68 comparing all the experiment al lysimeters, it was seen that hig her drywall percentage solids (TDS), but according to the positive relation between TDS and conductivity, these conductivity results confirm the aforementioned conclusion t hat the dissolution of calcium sulfate contributes largely to the total dissolved solids (TDS) (Jang and Townsend, 2003) CCA in treated wood also seems to contribute to the conductivity. Comparing Control (no CCA treated wood) and DW12 (with CCA treated wood) lysimeters, the latter had higher conductivity, even though both lysimeters had the same percentage of drywall. Similar findings were observed in a previous study on treated wood (Dubey, 2005) The water addition frequenc y was again found to affect this leachate parameter. The overall conductivity was increased after the first 100 days, during the time period in which water was added less frequently. This indicates that the smaller water travel rates may help release more ions. ORP stands for oxidation reduction potential, which is a measure of the oxidizing or reducing ability of a solution. While ORP is used to monitor the oxidizing or reducing capability of a compound, it is more often used to monitor the dominant reacti on or microorganism activity in an aqueous system in environmental studies. In previous studies (Jang 2000, Jambeck 2004, Dubey 2005), it was found that sulfate reducing bacteria (SRB) activity dominates in C&D debris lysimeters. This experiment showed the effect of drywall percentages on ORP values. With the general trends of decreasing ORP for all lysimeters, Control lysimeter showed the most negative ORP throughout the experiment For the experimental lysimeters DW1, DW6 and DW12 it took about 50 days f or ORP to decrease from positive values to about 200 mV. The ORP was
69 observed to be more negative for lysimeters with a higher percentage of drywall. Generally DW0 lysimeter had the highest ORP and the ORP values for Control and DW12 lysimeters (which ha d the same drywall percentage as in previous studies) observed in this experiment were a little higher (less negative) than those in previous studies (Jang 2000, Jambeck 2004, Dubey 2005), which might be caused by the relatively lower temperature or water addition rate employed in this experiment. Sulfide in C&D debris landfills is produced from sulfate reduction by sulfate reducing bacteria. As Figure 3 1 ( D ) shows, the percentage of drywall affects the sulfide concentration in leachate. DW0 and DW1 lysime ters, in which drywall percentage is either 0% or 1%, had much lower sulfide concentration than other lysimeters. While Control lysimeter had the highest and rather steady level of sulfide all the time, there seems to have been an initial lag period for DW 6 and DW12 lysimeters. This observation of increasing sulfide concentration is also reflected in the decreasing ORP values for DW6 and DW12 lysimeters in Figure 3 1 ( C ). The obvious difference in the trends of ORP and sulfide concentrations between Control and DW12 lysimeters may be due to the toxic effect of CCA dissolved from treated wood, which inhibited the activity of microorganisms during the initial lag period. Similar to the ORP results above, the observed sulfide concentrations in this experiment w ere lower than previous studies (Jang 2000, Jambeck 2004, Dubey 2005). Again, temperature difference may be a major reason causing this difference. Another reason may be due to the smaller sizes of the lysimeters in this experiment. Leaching of Arsenic fro m Landfills Figure 3 2 shows the arsenic concentration in the leachate over time. Except for the almost negligible concentration (below 0.01 mg/L) in the Control lysimeter, there
70 was no obvious difference among DW0, DW1 and DW12 lysimeters. This was somewh at unexpected, since it was originally hypothesized that sulfide levels would have either positive or negative impact on the leaching of arsenic. However, previous results (Jambeck, 2004) by the sulfate reducing bacteria activity. To make it more complicated, DW6 lysimeter shows lower arsenic leach ing than that in either DW1 or DW12 As discussed above, sulfide concentration is positively dependent on the percentages of drywall in the 3 experimental lysimeters. When sulfide and arsenic concentrations from all experimental lysimeters were plotted, as shown in Figure 3 3, the data points appeared shape. At the lower sulfide concentration end, arsenic concentration decreases with increasing sulfide concentration; on the other end, however, arsenic concentration increases with increasing su lfide concentration. Data points from DW6 lysimeter mostly cluster in the middle range of sulfide concentration. In the batch test where arsenite was mixed with different concentrations of sulfide solution at neutral pH, the final arsenic concentration in the filtered part also showed similar trend as seen in Figure 3 4. As the E h pH diagram (Figure B 1 in Appendix B ) implies, when sulfide is present, arsenic can form precipitate such as As 2 S 3 (orpiment) or AsS (realgar). As E h becomes lower, soluble thioar senic species are becoming more stable. Research conducted by Wilkin et al. (2003) showed that the formation of mononuclear thioarsenic species accounts for the increased solubility of orpiment with increasing sulfide conce ntration Eary (1992) The sulfide concentration (10 4 M, about 3 mg/L) at which orpiment has the lowest solubility was on the same level of sulfide concentration (1000 g/L, i.e., 1 mg/L)
71 i n leachate when lowest arsenic concentration was measured. It is not clear to what extent the formation of orpiment or other arsenic sulfides affects the arsenic leaching, but it is reasonable to say that arsenic sulfides and thioarsenic formation are cont ributing factors in controlling arsenic leaching in C&D debris landfills. Leaching of Iron from Landfills As expected, the leaching of iron in C&D debris landfills is also affected by sulfide levels. Figure 3 5 shows a negative relationship between leachat e iron concentration and sulfide concentration in all simulated landfills. The negative relationship is probably controlled by the precipitation reaction of iron sulfide. As suggested in previous studies ( Dubey 2005; Jambeck 2004), high concentration of su lfide in leachate was also the main factor contributing to the low copper concentrations due to the formation of copper sulfide precipitates. However, the negative relationship should be considered as a result regulated by the equilibrium between all rea cting components. As will be see n in Chapter 5, if iron is in excess, high levels of sulfide also mobilize iron significantly. Comparison of Preservation Methods for Arsenic Analysis It has been known that arsenite/arsenate can be precipitated by sulfide u nder acidic condition. Therefore, conventional acid preservation of C&D debris landfill leachate, which often has high concentration of sulfide, may not be appropriate. Figure 3 6 shows the effect of preservation method on arsenic recovery in several scena rios. In chemical solutions with 2 different levels of sulfide, 1 mg/L arsenite was spiked and solutions were acidified with nitric acid As seen in Figure 3 6 (a), with either low (8 mg/L) or high (50 mg/L) levels of sulfide present, significant loss of a rsenic was observed when using acid preservation (HUT and LUT). However, when the solutions
72 were first treated with hydrogen peroxide under basic condition before acidification, no loss of arsenic occurred (HT and LT) When arsenite was spiked in leachate collected from Control lysimeter (Figure 3 6 (b)), which has negligible arsenic, hydrogen peroxide treated sample (TS1) again showed smaller variation than acid preserved one (UTS1) even though ANOVA test age recovery Leachate collected from DW6 lysimeter was also tested using 3 treatment conditions ( Figure 3 6 (c)): hydrogen peroxide treatment (T), acid preservation (UTP), without any treatment (UTNP). The hydrogen peroxide treated sample (T) showed the b est recovery and the least variation. Even though the acid preserved sample (UTP) showed better precision than the sample without any treatment (UTNP), there was no significant difference in the average recovery. A rsenic S peciation in Landfill Leachate Thi oarsenic speciation using ion chromatography As the percentage of gypsum drywall in C&D debris lysimeters affected the levels of sulfide, it is expected that there would be difference s in arsenic speciation. The relationship between sulfide levels and the amount of thioarsenates is shown in Figure 3 7. The data were obtained from samples collected during 8 conse c utive sampling eve nts (Day 144 through Day 247). The average s ulfide levels in DW1 were about 100 g/L, comparing with over 10,000 g/L and 20,000 g/L in DW6 and DW12, respectively. The relative magnitude of sulfide levels matches the percentages of drywall initially used in landfills. As previously illustrated in Figure 3 3, total arsenic concentration in leachate was neither positively n or negativ ely cor related with sulfide levels. Here the lowest arsenic concentration among the 3 landfills was seen for DW6.
73 The total amount of thioarsenates in leachate was obtained from IC analyses. DW12 had the highest absolute level and the highest percentage (l ess than 10%) of thioarsenates, followed by DW 6 with less than 4% of arsenic being thioarsenates. There were no thioarsenates detected in DW1 leachate. These results, together with the data in Figure 3 3, again suggest that arsenic leaching and speciation are controlled by both precipitation and thioarsenic formation reactions between arsenic and sulfide. Arsenic speciation using arsenic sieve cartridges Due to the low conductivity of arsenous acid (H 3 AsO 3 ), the developed IC method is not sensitive enough t o detect arsenite. Arsenic cartridges (MetalSoft Center, Piscataway, NJ) were used to analyze arsenic speciation. These disposable cartridges have been shown to be able to remove arsenate while not retaining arsenite on the adsorbent in the cartridge. By m easuring the concentration of arsenic in the filtrate, either arsenite or arsenate concentration can be calculated. As seen in the top part of Figure 3 8, total arsenic in leachate was again neither positively nor negatively correlated with sulfide concent rations, as mentioned earlier. I n DW1 leachate, which had the lowest level of sulfide, 76.7% of total arsenic was arsenite, assuming only arsenite and arsenate were present in the leachate. Much higher fractions of arsenite were seen in DW6 and DW12, which is reasonable, since more reducing conditions existed in those 2 landfills. However, a little seen in DW6 leachate than that in DW12 leachate, even though DW6 leachate had higher concentrations of sulfide (which is diff erent than the case in Figure 3 7, where DW12 had higher sulfide concentrations than in DW6) difference between the total arsenic concentration in leachate and arsenite concentration in cartridge effluent, was ~0.14 mg As/L.
74 In order to explain it, the relationship between each thioarsenate component and sulfide concentration was plotted in Figure 3 8B. As expected, there was no thioarsenate detected in the leachate of DW1. For DW6 and DW12 leachate, monothioarsenate and trithioarsenate were detected. The h igher concentration of sulfide (and total arsenic) in DW6 leachate was accompanied by the formation of more thioarsenates than what was found in DW 12 leachate. The average amount of monothioarsenate found in DW6 cartridge analysis as shown in Figure 3 8A. As will be seen in Chapter 4, monothioarsenate has similar ads orption characteristics to arsenate, possibly due to the similar molecular structure ; they are identical except for one atom. This means that monothioarsenate if present, may also be retained on the cartridge just as arsenate is n DW6 leachate, as shown in Figure 3 8A, is most likely monothioarsenate, which was formed under high sulfide concentration. The value of more than 100% arsenite in DW12 may be from analysis errors which occurred during the sample dilution process H owever this at least indicated that arsenite fraction in DW12 leachate was close to unit y Summary In order to investigate the effect of sulfide levels on arsenic leaching and speciation, laboratory s imulated C&D debris landfills with varying percentages of dry wall were constructed and leachate was analyzed. It was shown that drywall percentage s in landfills contributed positively to sulfide levels. For arsenic leaching, however, the level of sulfide plays an interesting role. At both low and high levels of sulf ide, arsenic the
75 sulfide level was in a certain range of approximately 10 00 g/L. It was speculated that the formation of arsenic sulfide minerals and thioarsenic anionic spec ies account for at least part of the observation. Within a relatively low range, the increase in sulfide concentration helps immobilize arsenic by forming arsenic sulfide precipitates. When sulfide level is high enough, however, the increase in sulfide con centration leads to more arsenic leaching by forming thioarsenic species. Choosing an appropriate method for preservation of C&D debris landfill leachate is critical to prevent the low recovery of arsenic, as shown in results from either chemical solutions or leachate samples. Pale yellow precipitates, which are most likely insoluble arsenic sulfide compounds, were observed when arsenite in sulfide solutions were preserved with nitric acid. By treating the solutions with hydrogen peroxide before acidificati on, much better recovery of arsenic was achieved. Conventional acid preservation method may cause significant loss of arsenic and large variation in analysis results. Using the IC method developed in Chapter, 2, C&D debris landfill leachate was analyzed fo r thioarsenates. The relative percentage s of thioarsenate components even though not higher than 10% of total arsenic, were shown to be positively dependen t on the sulfide concentration. Arsenic speciation analyses using disposable cartridges were also co nducted. The cartridges have been used to distinguish arsenite, which is not adsorbed by the adsorbent in the cartridge, from arsenate. It was found that higher range. T he mass balance of arsenic species based on ion chromatography and c artridge analysi s
76 monothioarsenate, which is probably transformed from arsenite under high sulfide concentrations. Arsenate and mono thioarsenate are suggested to have similar adsorption behavior on the arsenic cartridges.
77 Figure 3 1 Leachate parameters over time. (A ) pH; ( B ) conductivity; ( C ) ORP ; (D ) sulfide. The first 9 liters of water was added weekly ; water was then added da ily after that. Control : No CCA wood, 12.4% drywall; DW0 : 10% CCA wood, no drywall; DW1 : 10% CCA wood, 1% drywall; DW6 : 10% CCA wood, 6% drywall; DW12 : 10% CCA wood, 12.4% drywall.
78 Figure 3 2. Arsenic concentration in leachate over time The first 9 l iters of water was added weekly; water was then added daily after that. Control : No CCA wood, 12.4% drywall; DW0 : 10% CCA wood, no drywall; DW1 : 10% CCA wood, 1% drywall; DW6 : 10% CCA wood, 6% drywall; DW12 : 10% CCA wood, 12.4% drywall.
79 Figure 3 3. Th e relationship between arsenic and sulfide in leachate from all landfills except Control Data for Control was not included, since there was no CCA treated wood in Control landfill DW0 : 10% CCA wood, no drywall; DW1 : 10% CCA wood, 1% drywall; DW6 : 10% CCA wood, 6% drywall; DW12 : 10% CCA wood, 12.4% drywall.
80 Figure 3 4. Arsenic solubility in sulfide solutions at pH 7 : batch test Sulfide solutions were made and pH adjusted in nitrogen glove box. Arsenit e stock solution was spiked to sulfide solutions to reach total arsenic concentration of ~18.4 mg As/L. Dashed line represents the initial total arsenic concentration. Error bars represent standard deviations of triplicate results. The data clearly shows a valley in arsenic concentration with various sulfide concentrations. Dashed line: Original c oncentration
81 Figure 3 5. The relationship between iron and sulfide in leachate from all simulated landfills. Control : No CCA wood, 12.4% drywall; DW1 : 10% CCA wood, no drywall; DW1 : 10% CCA wood, 1% drywall; DW6 : 10% CCA wood, 6% drywall; DW12 : 10% CCA wood, 12.4% drywall.
82 Figure 3 6 Comparison of preservation methods for arsenic analysis a) 1 mg/L arsenite in sulfide solution. HT : 50 mg/L sulfide + H 2 O 2 treatment; HUT : 50 mg/L sulfide + no H 2 O 2 treatment; LT : 8 mg/L sulfide + H 2 O 2 treatment; LUT : 8 mg/L sulfide + no H 2 O 2 treatment. b) 1 mg/L arsenite spiked in leachate from Blank lysimeter. UTNS : no H 2 O 2 treatment + no spike; TNS : H 2 O 2 treatment + no spike; TS1 : H 2 O 2 treatment + spike; UTS1 : n o H 2 O 2 treatment + spike. c) leachate from DW6 lysimeter. T : H 2 O 2 treatment; UTNP : no H 2 O 2 treatment + no acid preservation; UTP : no H 2 O 2 treatment + acid preservation. Error bars represent the standard deviation of triplicate results.
83 Figure 3 7. Thioa rsenic speciation in leachate by ion chromatography The data were obtained from samples collected during 8 conse c utive sampling events (Day 144 through Day 247). Total arsenic (As_total) was measured by ICP AES after sample digestion. Thioarsenates were m easured using IC. Error bars DW1 : 10% CCA wood, 1% drywall; DW6 : 10% CCA wood, 6% drywall; DW12 : 10% CCA wood, 12.4% drywall. ND : non detected.
84 Figure 3 8. Arsenic speciation by cartridge and ion ch romatography. (A) Arsenite (AsIII) analysis by disposable cartridges; ( B ) Thioarsenate analysis by ion chromatography. The data were obtained from s amples collected during 3 consec utive sampling events. Total arsenic (As_total) was measured by ICP AES afte r sample digestion. The percentages in (A) are the ratios of arsenite (AsIII) to total arsenic (As_total). Error bars represent standard deviations of 3 DW1 : 10% CCA wood, 1% drywall; DW6 : 10% CCA wood, 6% drywall; DW12 : 10% CCA wood, 12. 4% drywall.
85 Table 3 1. Waste composition in C&D debris blank lysimeter Component Percentage Source Untreated wood a Concrete Asphalt roofing Gypsum drywall b Cardboar d Aluminum sheet Copper wire Steel bar Steel s heet Insulation a : in all but the Control lysimeter, CCA treated wood accounts for 10 %; the total wood accounts for 33.6%. b : gysum drywall accoun ts for 12.4% in the Control and DW12 lysimeters. In other lysimeters, glass will be used to partially substitute gypsum drywall to achieve 12.4%. Table 3 2. Ion chromatograph conditions for speciation analysis Item Details Columns IonPac AS16/AG16 4 mm 250 mm (Dionex, Sunnyvale, CA) Detector CD20 DS3 1 (Dionex, Sunnyvale, CA) Pump GP40 gradient pump (Dionex, Sunnyvale, CA) Autosampler AS40 (Dionex, Sunnyvale, CA) Eluent 35 mM NaOH, 70 mM NaOH Gradient 0 11.5 min: 35 mM 11.5 15.5min: 35 mM 70 mM 15.5 35 min: 70 mM 35 38 min: 70 mM 35 mM 38 40 min: 35 mM Anion suppression ASRS Ultra II 4 mm (Dionex, Sunnyvale, CA) Regeneration mode External water addition, 10 mL/min Suppression current 300 mA Sample volume 5 mL Injection volume 70 L Retention times of thioarsenates AsSO 3 3 : 9.38 min AsS 2 O 2 3 : 15.08 min AsS 3 O 3 : 17.68 min AsS 4 3 : 20.65 min
86 CHAPTER 4 ADSORPTION OF THIOAR SENIC ANIONS ON IRON OXIDE COATED SAND Introduction The adsorption and desorption between contaminants and soi l components determine how contaminants move in groundwater. Understanding the adsorption and transport of contaminants in soils and groundwater is necessary in order to evaluat e the risks associated with contamination, and if possible, to develop appropri ate remediation plans. Computer models are usually employed to simulate the mobility of soluble contaminants in soils and groundwater. The movement of the contaminant is controlled by hydraulic properties of the soils as well as the interactions between th e contaminant and the soils. In many models, a retardation factor ( R f ) or distribution coefficient ( K d ) is used to describe the interactions between the contaminant and the soil. Specifically, the retardation factor is a dynamic parameter which quantifies directly the capability of the soil matrix to retard contaminants migration relative to water movement; and the distribution coefficient is a thermodynamic term which describes the partitioning of the solute between solid and liquid phases in equilibrium (Lal and Shukla, 2004) The retardation factor / distribution coefficient for a specific contaminant includes all the interactions between that contaminant and the soil, such as adsorption and pre cipitation. Due to these interactions, the retardation factor / distribution coefficient is affected by many variables, such as temperature, pH, redox potential, organic matter, ionic strength, and etc. There are many ways to obtain the retardation factor or distribution coefficient; however, column flow through experiments and laboratory batch experiments are the most common methods. The r etardation factor can be calculated directly from the column experiment as the ratio of groundwater (tracer) transport
87 ve locity and contaminant transport velocity. The d istribution coefficient is often obtained from batch experiments. In batch experiment s soil and contaminant solution with known concentration are mixed and shaken for a certain amount of time to reach equil ibrium. Distribution (or partition) coefficient K d which quantitatively describes the partition ing of contaminant between soil and solution phases, is defined as the ratio of contaminant concentration in solid phase to solution phase at equilibrium : (4 1) Theoretically, the distribution coefficient is a thermodynamic term whose use presupposes the system to be reversible, independent of contaminant concentration, and able to reach equilibrium within a reasonable amount of time. However, these assumptions are almost impossible to achieve for real systems. For example, equilibrium is hardly reached and distribution between solid and solution is not linear (dependent on contaminant concentration). Despite the obvious deviation s from thermodynamic assumptions, batch experiments still have been extensively used in soil adsorption studies. For non linear adsorption, in which the distribution coefficient is dependent on contaminant concentration, isotherm adsor ption models are often used to describe a concentrations are usually varied while keeping all other conditions the same. Contaminant concentrations adsorbed on soils are then plo t ted against the equilibrium concentration of contaminant in solution and a n isotherm adsorption curve is thus
88 obtained. Two adsorption models are usually used to fit the isotherm curves: Langmuir model (Langmuir, 1918) and Freundlich model (Freundlich 1926) In the Langmuir model, all the adsorption sites on solid surface s are considered to be equivalent and the adsorption is assumed to be one layer coverage. So saturation adsorption exists when all the sites are covered. The Langmuir Model is: (4 2) Where C solid = contaminant adsorbed on solid phase (mg/kg) C solid,max = maximum concentration of contaminant adsorbed on solid phase (mg/kg) = equilibrium contaminant concentration in solution (mg/L) = Langmuir adsorption constant A linear regression form of the Langmuir equation is : (4 3) By plotting versus and can be obtained from the slope and the intercept respectively When i s small, K d is equal to In the Freundlich model, contaminants can be adsorbed on solid surfaces in multiple layers and the adsorption sites may not be the same. The Freundlich equation is: (4 4) Where K F = Freundlich adsorption constant N = dimensionless constant A linear regression form of the Freundlich equation is
89 (4 5) K F and N can be obtained from t he slope and intercept by plotting versus When N equals 1, K F equals K d The adsorption and transport of inorganic arsenic species, including both arsenate and arsenite, have been investigated extensively using laboratory batch experiments or continuous column flow through tests. Natural soils (Ghosh and Bhattacharyya, 2004; Jiang et al., 2005; Williams et al., 2003; Zhang and Selim, 2005) iron mineral coated sand (Herbel and Fendorf, 2006; Kocar et al., 2006) and synthesized iron oxide minerals (Dixit and Hering, 2003; Pierce and Moore, 1982; Raven et al., 1998) have been used as the experimental units. Distribution or partition coefficient s ( K d ) and retardation factors ( R f ) have been obtained and various models on arsenic adsorption and transport have been developed. Studies have show n that arsenic adsorption on soils is positiv ely cor related with the percentages of iron oxide minerals and clay content in soils (Manning and Goldberg, 1997b; Wauchope and McDowell, 1984) Arsenic adsorption on pure minerals, including iron oxide and clay minerals, has been studied extensively (Chakraborty et al., 2007; Luengo et al., 2007; Manning et al., 1998; Masue et al., 2007; Raven et al., 1998) Iron oxide minerals are ubiquitous in natural soils. There are many types of iron oxide compounds; among which hematite is one of the m ost commonly encountered especially in sub tropical or tropical regions (Cornell and Schwertmann, 2003) Batch experiments have been used to study arsenite and arsenate adsorption on pure iron oxide minerals, including ferridhydrite (Moldovan and Hendry, 2005; Pierce and Moore, 1980; Raven et al., 1998) goethite (Antelo et al., 2005; Dixit and Hering, 2003; Manning
90 et al., 1998) and hematite (Redman et al., 2002) Iron oxide coated sand were also used in dynamic colum n experiments to investigate the effect of anaerobic microbial activity on arsenic mobilization (Herbel and Fendorf, 2006; Kocar et al., 2006) The adsorption of arsenite and arsenate on ferrihydrite has been studi ed (Pierce and Moore, 1980, 1982; Raven et al., 1998) It was found that at low arsenic concentrations, the adsorption followed the Langmuir isotherm; while at higher arsenic concentration, the adsorption increased with increasing concentration. The adsorption of both species was also strongly pH dependent. At low pH (< ~8 9), arsenate adsorption slightly changed or decreased with increasing pH; comparatively, arsenite adsorption increased. At higher pH (> 10), both species adsorbed less with increasing pH, with arsenate affected much more by pH increase. The pH dependence was generally explained by the changes of surface charge and arsenic species. With increasing pH, the surface charge of ferrihydrite becomes more negative. The lower pKa values of arsenic acid (pKa 1 = 2.20, pKa 2 = 6.97, pKa 3 = 11.53) (Flis et al., 1959) makes arsenate more easily affected by pH at lower pH than arsenite, which has higher pKa values for its corresponding arsenious acid (pKa 1 = 9.22, pKa 2 = 12.13, pKa 3 = 13.40) (Britton and Jackson, 1934) Arsenic adsorption and trans port in sulfidic systems, or the adsorption and A m odel developed by Lee (2005) has indicated that thioarsenite anions might be less prone to adsorb on the surface of iro n minerals than either arsenate or arsenite, and consequently these could enhance the mobility of arsenic species. Based on the results of this model, it should be anticipated that thioarsenite will have higher mobility in a soil environment, where iron mi nerals are the major component affecting arsenic transport.
91 It is necessary to evaluate the adsorption and transport of thioarsenic species, or arsenic species in sulfidic systems, on soils and in groundwater. This research aim ed to provide some preliminar y information on the adsorption of thioarsenic species on some iron oxide coated quartz sand by using laboratory batch experiments. Materials and Methods Synthesis The t hioarsenic compound (monothioarsenate Na 3 AsSO 3 ) was made according to the previously e stablished methods described in Chapte r 2. Hematite powder was purchased from Strem Chemicals ( Newburyport, MA ). Hematite coated quartz sand was made following to the relevant literature. Hematite coated sand : The coating procedure was modified from the li terature (Scheidegger et al., 1993; Schwertmann and Cornell, 2000) 1.2 g hematite was weighed into a 2 50 mL polyethylene bottle, which wa s then capped and rotated for 24 hours at 30 rpm to obtain a homogeneous suspension 100 g sand was then added into t he suspension and the bottle was rotated for another 24 hours at 30 r pm. The sand added had previously been soaked with 1 M HNO 3 for 2 hours and rinsed with de ionized water until the measured conductivity was lower than 5 S/cm. After rotation, sand particles, which were coated with hematite, settled quickly and free (unat tached) hematite particles were separated by pouring off the upper layer suspension. Coated sand was rinsed with de ionized water, a 0.01 M pH 3 NaNO 3 solution, and then with de ionized water again. The purpose of using a solution of pH 3 wa s to remove wea kly bound hematite and its aggregates. The coated sand wa s air dried at room temperature in a ventilation hood for at least 7 days The coated sand wa s digested according to the EPA SW 846 Method 3050B. The amount of iron coating wa s measured by an ICP AES
92 (Trace Analyzer, Thermo) according to the EPA Method 6010B. The actual iron coverage was 0. 44 % by weight Mineral C haracterization The mineral structure of the purchased hematite w as analyzed and verified by an X ray diffractometer (Philips APD 3720 XRD ) (the XRD spectrum is shown in Figure C 1 ) Total iron and arsenic contents were analyzed using an ICP AES (Tracer Analyzer, Thermo Elect r onics) following the EPA Method 6010B after digestion according to the EPA Method 3050B. Adsorption K inetics All the s teps of the adsorption experiment s (except the step of rotation) were conducted in a glove box under nitrogen atmosphere with oxygen level of less than 0.1% as measured by a GEM 500 Gas Analyzer ( LandTec ) All the solutions were made with deoxygenated nan opure wate r The adsorption kinetics was only evaluated at pH 7. An initial arsenic concen tration of 0.1 mM/L (7.5 mg/L) wa s chosen. Hematite coated sand of 1 .000 g was weighed and put in a 40 mL VOC vial, followed by the addition of 18.0 mL 0. 0 1 M NaNO 3 s olution T he suspension pH was then adjusted to the desired value using either dilute HNO 3 or dilute NaOH. An aliquot of 1 mL 150 mg/L arsenic stock solution (wit h desired pH already adjusted) wa s spiked to the suspension. The final solution volume was adj usted to 20 mL. The purpose of using 0. 0 1 M NaNO 3 was to keep all the samples at somewhat constant ionic strength. The vials were sealed and rotated for various amount s of time. T he suspension was then centrifuged for 10 minutes at 10,000 g after 30 minute s, 1 hour, 2 hours, 4 hours, 12 hours, 24 hours, and 48 hours of rotation followed by filtration The filtrate w as preserved with concentrated trace metal grade nitric acid for total
93 arsenic and iron analysis using ICP AES according to the EPA Method 6010B. Arsenic adsorbed on sand was calculated based on the difference in concentration of the supernatant befor e and after adsorption tests. All the kinetic tests were run in triplicate. Thioarsenic speciation ana lyses were carried out for the adsorption of thioarsenic on sand to check whether thioarsenic species ha d significantly transformed to other species. The purpose of the blanks was to eliminate the effect of the wall of reaction vessels. Controls w ere run w ithout spiking arsenic solution. Another set of controls wer e run using pure quartz sand instead of coated sand. Adsorption I sotherms The equilibrium time used for the adsorption isotherm experiment s was 24 hour s based on preliminary adsorption kinetics te st. Batch experiments to obtain adsorption isotherms w ere run under 3 different pH values: 5, 7, and 10 At each pH value, n ominal concentrations of 0.75 mg/L, 1.5 mg/L, 3.75 mg/L, 7.5 mg/L, 15 mg/L and 37.5 mg/L of arsenic solutions were used in the adsor ption The experimental setup in the adsorption isotherm measurements wa s the same as that in the kinetic test Adsorption isotherm s were obtained by plotting arsenic concentration s on the solid phase against arsenic concentration in the supernatant after the reaction. Langmuir or Freundlich models were used to fit the adsorption data using statistical patches included in SigmaPlot (Version 11.0, Systat Software, Inc., San Jose, CA ) Two parameter Power equation and Modified Hype rbola I equation were used for the Freundlich model and the Langmuir model, respectively. Adsorption constants and values for goodness of fit ( R 2 ) were obtained.
94 Arsenic A dsorption under S ulfidic C ondition Since only mono th ioarsenate was synthesized succes sfully and the reaction between sulfide and arsenite produces thioarsenates quickly, adsorption experiments of arsenite on iron oxides were carried out in sulfide solution to observe the general adsorption behavior of arsenic in sulfidic systems The expe rimental setup w as similar to those in the adsorption kinetics experiments described above. However, arsenite was spiked in 5 mM sulfide solution and p H was adjusted using dilute nitric acid and dilute sodium hydroxide Re sults and Discussion Adsorption o f A rsenic on H ematite coated S and T he structure of hematite was confirmed by the XRD spectrum as shown in Figure C 1. H owever, no analysis was performed to further confirm the identity of hematite after it was coated on sand. The total iron coverage analyz ed by ICP AES was 0. 44 wt% The purpose of the adsorption kinetics experiment was to find an appropriate reaction time for later isotherm measurement s Figure C 2 shows the kinetics of 7.5 mg/L arsenic (arsenate) adsorbed on coated sand at pH 5. E quilibriu m was reached approximately between 12 hours and 24 hours Although there might be differences in kinetics among other pH or arsenic species (Raven 1998) no further evaluation was conducted and an equilibration period of 24 hours was chosen for adsorption isotherms experiments based on the results of this research and the relatively fast reaction kinetics between arsenite/arsenate and iron oxide minerals described in prior rese arch (Gupta et al., 2005; Pierce and M oore, 1982; Raven et al., 1998)
95 A dsorption isotherms of arsenite, arsenate, and monothioarsenate on hematite coated sand are shown in Figure 4 1 B oth the Langmuir (solid line) and Freundlich model s (dashed line) were used to fit the data points. F or ar senite, the Langmuir model fit s better than the Freundlich model, as seen in both Figure 4 1 and the higher R 2 in Table 4 1. F or arsenate and monothioarsenate, however, the Langmuir model was better only at pH 10. T his may indicate that a similarity exists between arsenate and monothioarsenate in terms of adsorption behavior. This similarity was not unexpected since monothioarsenate differs with arsenate only by one sulfur atom. A s there are still other oxygen atoms in monothioarsenate that can form simila r surface complexes as that of arsenate, it is reasonable to anticipate similar adsorption behavior. All the adsorption isotherms in Figure 4 1 exhibit clear nonlinear behavior which w as also characterized by low values of N in Freundlich model (Table 4 1 ). This observed nonlinearity matched results of arsenite/arsenate adsorption studies by others (Zhang 2005, Manning 1997, Williams 2003, Raven 1998, Pierce 1982). The constant N reflects the extent of heterogeneity of surface sorption sites in terms of so lute retention affinities. The low values of N suggest that the sorption sites on the coated sand were heterogeneous. The constant N also indicates the concentration dependence of the sorption process. Due to the distribution of sorption sites with differe nt binding strength, sorption only occurs on those with highest binding energy when concentration is low. A relatively small value of N means good adsorption over almost the entire range of concentrations, whereas a relatively high value of N means the ads orption is somewhat suppressed at low concentrations. The relatively higher values of N at pH 10, especially for that of arsenate (Table 4 1), indicates th e suppression of strong binding sites at pH
96 10. This may be due to the increasing negative charge on the matrix surface, which repels the negative charges of adsorbing anions. The suppression of strong binding sites at pH 10 is also implied by the much smaller K L values in the Langmuir model, which is a measure of the binding energy of sorption sites. The re are also significant difference s between the K L values at pH 5 and pH 7 for arsenate and monothioarsenate. Much higher K L values at pH 5 suggest stronger binding which is possibly due to the interaction between the positive surface charge and the negati ve charge of adsorbing anions. Effect of pH on A dsorption A s pH can affect the surface charge of iron oxide minerals as well as the charge state of free aqueous species, it is expected that adsorption will be different at different pH values. Figure 4 2 sh ows the isotherms of arsenite, arsenate, and monoarsenate at pH 5, pH 7, and pH 10. From the Langmuir fit curves, saturated monolayer adsorption capacity values C solid, max were obtained. For arsenite, the adsorption plateau at pH 7 was greater than eith er that at pH 5 or pH 10. F or arsenate or monothioarsenate, however, the maximum adsorption was seen at pH 5, and adsorption capacity decreased with increasing pH values. This observation was comparable to what was found for arsenite and arsenate adsorptio n on ferrihydrite (Raven et al., 1998) or hematite ( Mamindy Pajany 2009) B oth the surface charge of iron oxide minerals and the charge state of arsenic species need to be considered when explaining the different adsorption behavior between arsenite and arsenate/monothioarsenate. F or arsenite, almost all species exist in neutral form at pH 5, which is neither attracted n or repelled from the iron oxide surface by electrostatic forces. A t pH 10, the surface charge of iron oxide is negative. A f air amount of arsenite also exist s as ne gative ions, so they are
97 repelled by the mineral surface. A t pH 7, however, the surface of iron oxide bears positive charge, while some arsenite exists as anions. The attraction between them promotes adsorption of arsenite at neutral pH. Monothioarsenate s hows similar pKa values as that of arsenate (see discussion in Chapter 2). A t pH 5, the surface of iron oxide minerals is positively charged; while the majority of monothioarsenate or arsenate is anionic. W hen pH is increased, the surface of iron oxide has less positive charge and more negative charge, which repels more against monothioarsenate or arsenate anions. T he adsorption of these two arsenic species therefore shows a decreasing trend with increasing pH. T he pKa values of other thioarsenate species w ere shown to be even lower than that of arsenate, which means similar trend with pH is expected in terms of their adsorption on iron oxide minerals. The adsorption of arsenite and arsenate on iron oxide surfaces is usually described using an inner sphere c omplexation or outer sphere complexation model. The inner sphere complexation model involves the ligand exchange and the removal of water molecules during the process. In the outer sphere complexation model the adsorbing anions attach to the surface via h ydrogen bonding interaction. On the surface of hematite, it was found that arsenate adsorbs through the form ation of inner sphere complexes; whereas arsenite forms both inner and outer sphere complexes (Ona Nguema 2005; Goldberg 2001) Inner sphere comple xes are suggested to have stronger binding. From the adsorption isotherms (Figure 4 1 ) and adsorption constants obtained (Table 4 1), monothioarsenate behaves similar to that of arsenate in adsorption. This is reasonable from the complexation point of view since the substitution of one single oxygen atom in arsenate by sulfur atom may not affect either
98 complexation process. In other thioarsenate anions, however, with the increasing number of sulfur atoms replacing oxygen atoms, the binding strength of the complexes, or even the formation of those complexes, may be significantly affected. This is especially the case for tri and tetra thioarsenate in which there is only one or no oxygen atoms, and the formation of bi dentate inner sphere complexes on the su rface of iron oxides is impossible. On the other hand, lower pK a values (stronger acidity) for tri and tetra thioarseniate (see discussion in Chapter 2) mean higher proportion of dissociated anions at the same pH, which may improve the electrostatic inter action between positive charged iron oxide surface and thioarsenates in an acidic environment. Effect of S ulfide on A dsorption Since there were no other pure thioarsenate compounds available, the effect of sulfide on adsorption of arsenite on to coated san d was investigated. Figure 4 3 shows the adsorption isotherm s T he adsorption data does not follow the Langmuir model while the Freundlich model fit s the data well only for experiments carried out at pH 7 and pH 10 The isotherms are closer to linear than those of in non sulfide solutions as can be seen from the higher N values of 0.47 and 0.71 at pH 7 and pH 10, respectively. As discussed above, the high values of N indicate that the adsorption was suppressed at low arsenic concentrations. This was seen f rom the slow increase of the isotherms at the low concentration compared to sharp increases in isotherms of arsenic in non sulfide solutions When K F values are compared, t he adsorption capacity of arsenite in sulfide solutions was also lower than that in non sulfide solutions at pH 7 and pH 10 The initial intention was to see how thioarsenate anions behave differently in terms of adsorption on hematite coated sand. A lthough there was an increasing absolute amount (and possibly lower fraction) of thioarse nate species formed when the arsenite
99 concentration was increased (Figure 4 4) the formation of thioarsenate alone may not be the only reason for the initial ly slow increase of isotherms the low adsorption capacity and the close to linear relationship. I nterestingly, a color change was observed for the adsorption suspension after rotation W ith low er arsenite concentrations, a black colored substance was formed, which was unstable and disappeared when exposed to air. W ith higher arsenite concentrations, l ess or no ne of this black substance was observed (Figure C 3 ). T he substance was very likely iron sulfide minerals, which can form from the transformation of hematite minerals under high sulfide concentrations (Neal et al., 2001) I n this adsor ption experiment, at lower arsenite concentratio ns the high sulfide concentration in solution caused the formation of iron sulfide. W hen the arsenite concentration increased, however, the formation of thioarsenate s would consume sulfide to such a level that it would decrease or prevent the formation of iron sulfide A s a result, less or no black precipitate was observed for higher level s of arsenite solutions. Figure 4 4 shows the increase in the amount of thioarsenate along with the decrease of sulfide in solutions Even when started with almost the same level of s ulfide (~70 mg/L) the formation of thioarsenate consumed sulfide gradually, especially for the one with the highest initial arsenite concentration. Another reason that may explain the consumption of sulfide in solution is the formation of arsenic sulfide precipitates, such as orpiment (As 2 S 3 ) or realgar (AsS). T he formation of iron sulfide minerals and the relatively low adsorption when arsenite concentration was low indicate that arsenic adsorption on iron sulfide minerals may not be as favorable as on ot her iron oxide s such as hematite in this case. Iron sulfides have been shown to be able to remove arsenite (Wolthers 2005, Bostick and
100 Fendorf 2003) by forming a FeAsS like surface precipitates (Bostick and Fendorf 2003). However, the removal through the f ormation of FeAsS like phases occurs only when sulfide concentration is low. The addition of sulfide below orpiment saturation was shown to inhibit arsenite adsorption on iron sulfides. With the addition of 5mM sulfide at pH 7, the adsorbed arsenic was abo ut 1 mg As /g Fe at 5 mg As /L solution concentration in 1g/L FeS 2 (Bostick and Fendorf 2003) which was similar to the observation of the initial low adsorption in this experiment (about 2 8 mg As/g Fe, assuming an iron content of 0. 44 % in coated sand) The inhibition of arsenic adsorption upon addition of sulfide was suggested to be due to the competiti ve adsorption from sulfide or the formation of particle un reactive soluble arsenic sulfide (thioarsenic) species. A series of adsorption experiments with va rious sulfide levels may be conducted to further explore the effect of sulfide levels on adsorption. Under even higher sulfide levels, precipitation of arsenic sulfide minerals such as orpiment controls the solubility of arsenic. Macroscopic adsorption e xperiments showed that FeS isotherms at pH 7 and pH 9 fitted the Langmuir model at low arsenic activity (less than 15 M), but the relationship became close to linear relationship at higher arsenic activity which causes the formation of FeAsS like phase a nd changes the adsorption mechanism. As seen in many cases of arsenic adsorption on iron oxides minerals, the adsorption of arsenite on hematite coated sand in sulfide solution was also dependent on pH. At pH 7 and pH 10, a similar amount of absorption was observed, though there was a little more absorption for pH 7 when solution arsenic concentration was lower ( Figure 4 3 ). Similar pH dependence was also observed by Han et al. (2011). In order to find better
101 solutions for arsenite removal under anoxic reducing conditions, Han et al. (2011) synthesized FeS coated sand and investigated its adsorption characteristics. Adsorption isotherms were fit using the Langmuir model and a rsenic removal capacities obtained were 41.6, 10.7 and 12.7 mg As/g FeS at pH 5, 7, and 9, respectively. The significant increase of arsenic removal at pH 5 was thought to be possibly due to the formation of arsenic sulfide minerals such as realgar (Gallegos et al. 2007 2008) or orpiment (Han 2009) due to the reaction between arsenite and sulfide which is from the dissolution of FeS at low pH. In the present research, the data well. Rather, the Freundlich model was used to fit d ata at pH 7 and pH 10 with values of R 2 equal to 0.9827 and 0.9675, respectively. The obvious deviation from the Langmuir model is possibly due to the development of 3 dimensional structure s including surface precipitates, such as FeAsS as suggested by Bo stick and Fendorf (2003). At all pH values, the formation of arsenite sulfide precipitate may contribute to the removal (adsorption) of arsenite. However, the formation of arsenite sulfide is more favored under acidic condition s, as evidenc ed by the decre asing solubility of orpiment with decreasing pH (Clarke and Helz 2000) It means that the contribution to adsorption from arsenic sulfide formation varies with pH. According to experimental results and a prediction model (Eary 1992, Wilkin 2003) of amorpho us orpiment solubility in sulfidic solutions, in 5 mM sulfide solution, arsenic is under saturated below 30 mg/L at pH 7 or pH 10. At pH 5, however, arsenic is over saturate d at concentrations as low as 0.75 mg/L The is otherm data at pH 5 in Figure 4 3 sh ows a sharp increase after arsenic solution concentration exceeds approximately 1 mg/L possibly as the result of the precipitation of arsenic sulfide.
102 Summary I n this chapter, adsorption isotherms were measured and compared for arsenite, arsenate, and mon othioarsenate at different pH values M onothioarsenate adsorption showed similar pattern as that of arsenate, which was different than that of arsenite. A s both monothioarsenate and arsenate have similar acidities and molecular structure, it is reasonable that they behave similarly in terms of adsorption on iron oxides according to the surface complexation model of arsenic on iron oxides For other thioarsenates, especially for tri and tetra thioarsenate anions, however, the formation of surface complexes, such as inner sphere bi dentate bi nuclear or bi dentate mono nuclear complexes, may be difficult or inhibited. I n consequence, the adsorption behavior of tri or tetra thioarsenate on iron oxides may be very different than that of arsenate. Thermodynamic calculation s and modeling are probably necessary in order to evaluate the adsorption process of thioarsenate anions. The effect of sulfide on arsenic adsorption was investigated by conducting arsenite adsorption batch experiments on hematite coated sand. A rsenite adsorption on coated sand in high sulfide solutions showed a slow increase in adsorption at the beginning compared with the overall sharp increase in adsorption isotherms of arsenic in non sulfide solutions. A t low arsenic concentrations, the for mation of iron sulfide minerals was suggested by the black suspension and this may have been inhibit ing the adsorption of arsenite, in comparison to adsorption on pure iron oxide minerals. The formation of FeAsS like phases may also contribute to arsenic r emoval, especially at high arsenic concentrations. Arsenic sulfide precipitation controls the mobile concentration of arsenic depending strongly on solution pH. In an acidic environment, the precipitation of arsenic sulfide minerals plays more roles in imm obilizing arsenic.
103 Figure 4 1 Adsorption isotherms of arsenite, arsenate and monothioarsenate on hematite coated sand Dotted lines represent Freundlich fit of the data; solid lines represent Langmuir fit of the data. Two direction error bars repre sent the standard deviations of triplicate results. For arsenite, Langmuir model fits better at all pH values. For arsenate and monothioarsenate, Langmuir model fits better than Freundlich model only at pH 10.
104 Figure 4 2 Adsorption isotherms of arse nite, arsenate, and monothioarsenatre at different pH Among Langmuir and Freundlich models, only the better fitted curves are plotted. Long dash lines represent fitted curve at pH 10; dotted lines represent fitted curve at pH 7; solid lines represent fitt ed curve at pH 5. Two direction error bars represent the standard deviations of triplicate results.
105 Figure 4 3 Adsorption isotherm s of arsenite in sulfide solution Initial sulfide concentration is approximately 5 mM. The inset figure shows the enlarg ed part of ad sorption data at pH 5 and pH 7. Dotted line and solid line in the inset figure represent the fitted curves at pH7 and pH 5, according to Freundlich model.
106 Figure 4 4 Sulfide change and thioarsenate formation during adsorption at pH 7 Nom inal initial sulfide concentration is 100 mg/L. Initial arsenite concentration is approximately 18.4 mg As/L. For each pair of columns, the left one represents sulfide concentration of the sulfide arsenite mixture before rotation; the right one represents sulfide concentration of the sulfide arsenite mixture after 24 hours of rotation.
107 Table 4 1. Adsorption parameters of arsenic species on hematite coated sand AsIII pH K L C solid,max r 2 K F N r 2 5 6.06211.1713 25.5200 0.9599 18.69481.1066 0.12260.02 47 0.8854 7 7.86911.1325 31.8477 0.9794 23.83131.8567 0.11290.0319 0.8002 10 0.98190.1493 25.1522 0.9800 12.73411.4958 0.21740.0456 0.8793 AsV pH K L C solid,max r 2 K F N r 2 5 1142.7609.7 29.6337 0.6637 26.79830.3404 0.06050.0044 0.9838 7 6.638 33.1800 11.9829 0.6211 9.65270.2056 0.08640.0091 0.9589 10 0.29470.0109 6.5955 0.9991 2.11490.3363 0.31040.0549 0.9157 AsS pH K L C solid,max r 2 K F N r 2 5 38.815426.0123 20.4670 0.7679 16.73620.7549 0.11040.6647 0.9343 7 1.64282.7590 21.3055 0 .1385 14.73920.6647 0.13150.0187 0.9499 10 1.30950.3438 14.4610 0.9481 7.63661.2895 0.22770.0779 0.7386 AsIII in Sulfide pH K L C solid,max r 2 K F N r 2 7 6.25240.6534 0.47290.0453 0.9827 10 3.43750.8934 0.70640.1057 0.9675
108 CHAPT ER 5 ARSENIC RETENTION IN HEMATITE COATED SAND AFFECTED BY DIFFERENT LEVELS OF SULFIDE IN C&D DEBRIS LANDFILL LEACHATE Introduction Iron minerals, usually iron oxide minerals, are normally considered the sink of arsenic in soils. Arsenite or arsenate can b e adsorbed on the surface of soils by forming surface complexes with iron oxide minerals (Catalano et al., 2008; Manning et al., 1998; Manning and Goldberg, 1997a; Wang and Mulligan, 2008) The adsorption of arseni c on iron oxide minerals prevents the natural water systems from being contaminated by arsenic in many cases. In fact, iron oxides or iron oxide coated materials have been used as remediation agents for arsenic removal. Arsenic can also form co precipitate s such as scorodite (FeAsO 4 2H 2 O) with iron minerals. On the other hand, significant arsenic mobilization still occurs in certain environmental scenarios where iron oxide minerals exist. The iron reductive dissolution (IRD) process has been suggested as a major contribution to arsenic mobilization in iron containing soils (Cummings et al., 1999; Harvey et al., 2002; Hering and Kneebone, 2001; Langner and Inskeep, 2000; Nickson et al., 2000) In an IRD process, th e supply of oxygen is restricted and an anaerobic reducing condition develops as anaerobes thrive. Ferric iron, which is usually the form of iron in iron oxide minerals, acts as an oxygen acceptor and is reduced to ferrous iron, which is mobile in aqueous phases Arsenic, which was previously bound to the iron oxide minerals, is therefore released to the solution phase and mobilized. Under natural conditions, the consumption of oxygen by microorganisms is balanced by the diffusion of air into soils. However this natural balance can be disrupted by many a nthropogenic activities, such as the construction of a landfill, which
109 will eliminate or greatly decrease the rate of oxygen diffusion into the soils and promote the evolution of anerobic condition s In unli ned C&D debris landfills, leachate can also penetrate into the soil and promote the formation of anaerobic reducing conditions. In Florida, elevated iron concentration s ha ve been found in groundwater at several landfill sites, including both C&D debris and MSW landfills (Rhue et al., 2008) even before the waste placement. This indicates that the source of the elevated iron is not from the landfill leachate, but from the native soils underneath the landfill. It was also reported that arsenic mobilization at a landfill is possibly due to the reduction of iron oxides minerals (deLemos et al., 2006) Arsenic concentration s in groundwater monitoring wells w ere found to be positively corre lated with dissolved iron concentration s in one landfill (Stollenwerk and Colman., 2004) It has also been suggested that the reducing condition s in the soils near a landfill site increased the mobilization of naturally occurring arsenic (Keimowitz et al., 2005) not necessarily by the dissolution of iron minerals in soils. In C&D debris landfills, one feature is that the anerobic and highly reduc tive condition s often lead to high sulfide concentration s in the leachate. In the soils underneath unlined C&D debris landfills, the interaction s among arseni c, sulfide, and iron are expected. Arsenic sulfide minerals (orpiment and realgar) were suggested to be formed in landfills when the sulfide concentration is high (Hounslow, 1980) X ray absorption spectroscopy (XAS) was used to directly detect the speciation in an arsenic impacted aquifer sediment Their results showed that the dissolved arsenic concentration was controlled by the formation of a realgar like arsenic sulfide at arsenic concentration s On the other hand, iron sulfide minerals
110 can form under conditions of high sulfide concentration s when iron materials are present (Bostick and Fendorf, 2003) Gammonsa and Frandsen (2001) studied the fate of several metals in H 2 S rich wetland s They proposed that both ferrihydrite and goethite can be t ransformed to amorphous iron sulfide (FeS) and ferrihydrite the less crystalline one, can form precipitate at lower sulfide concentrations: Fe(OH) 3 (ferrihydrite) + H 2 S(aq) + H + + e 2 O logK = 14.84 FeOOH(goethite) + H 2 S(aq) + H + + e eS(am) + 2H 2 O logK = 8.95 These iron sulfide minerals, including pyrite (FeS 2 ), mackinawite (FeS), arsenopyrite (FeAsS), are usually considered as arsenic adsorbents. Adsorption study (Wolthers et al., 200 5) showed that both arsenate and arsenite can be adsorbed onto mackinawite (FeS), with arsenate adsorbed more strongly than arsenite. As a result, the formation of either arsenic or iron sulfide minerals will immobilize arsenic especially arsenite, which is not easily adsorbed by soils a t normal pH The immobilization mechanisms may include adsorption, co precipitation, or precipitation. However, recent studies showed that arsenic may be mobilized under high sulfide conditions by forming thioarsenic spec ies (Rochette et al., 2000; Stauder et al., 2005) Since high sulfide concentration s are often seen in the leachate from C&D debris landfills and d ue to the fact that a rsenic is leached from C&D debris landfills, it will be interesting to see how arsenic mobility in iron containing soils is affected by various sulfide concentrations. The objective of this experiment wa s to investigate how the retention of arsenic ( from C&D debris landfills leachate ) on iron oxide m inerals is affected by different sulfide levels under anaerobic reducing conditions. Hematite coated sand was packed
111 into c olumns through which leachate from laboratory C&D debris landfills w as pumped Effluent was analyzed for both ferrous ir on and arsen ic. The hypothesis wa s that sulfide in C&D debris landfill leachate w ould impact the mobilization of both iron and arsenic in the soils under the landfills. Materials and Methods Sand C oating and I ron A nalysis Hematite coated sand: The coating procedure wa s modified from literature (Scheidegger et al., 1993; Schwertmann and Cornell, 2000) 1.2 g hematite was weighed into a 2 50 mL polyethylene bottle, which wa s then capped and rotated for 24 hours at 30 rpm to obtain a homogeneous suspension 100 g sand was then added into t he suspension and the bottle was rotated for anothe r 24 hours at 30 rpm. The sand was previously soaked with 1 M HNO 3 for 2 hours and rinsed with deionized water until the measured conductivity was lower than 5 S/cm. After rotation, sand particles, which were coated with iron oxides, settled quickly and f ree (unattached) hematite or goethite particles were separated by pouring off the upper layer suspension. Coated sand was rinsed with de ionized water, 0.01 M pH 3 NaNO 3 solution, and de ionized water again. The purpose of using a solution of pH 3 wa s to r emove weakly bound hematite and its aggregates. The coated sand wa s air dried at room temperature in a ventilation hood for at least 7 days The coated sand wa s digested according to EPA SW 846 Method 3050B. The amount of iron coated on the sand wa s measur ed by ICP AES (Trace Analyzer, Thermo) according to EPA Method 6010B. The actual iron coverage was 0. 44 % by weight Total iron and arsenic in coated sand: The analysis of total iron and arsenic content s was based on the EPA SW 846 Method 3050B. A clean ac id rinsed
112 Erlenmeyer flask wa s charged with 2.00 g coated sand An aliquot of 10 mL 1:1 (v/v) trace metal grade nitric acid wa s added to the flask and covered with a ridged watch glass. The flask wa s put onto a hot plate and heated without boiling. After 3 0 minutes, another 5 mL of nitric acid wa s added to the flask. The flask wa s heated for 2 hours and then removed from the hotplate. A volume of 5 mL 30% (v /v) hydrogen peroxide solution wa s added into the flask and the flask wa s then heated for another 2 h ours. After that, 10 mL of trace metal grade ( 37% by volume) hydrochloric acid wa s added and the flask wa s heated for 10 minutes. Then the digested mixture wa s filtered through a Whatman 41 filter paper and the filtrate wa s adjusted to 50 mL using de ioniz ed water in a volumet ric flask. The sample solution wa s then analyzed for metals (iron and arsenic) using an inductively coupled plasma atomic emission spectrophotometer (ICP AES) (Thermo Electron Corporation, Trace Analyzer) according to the EPA Method 60 10B. Column S etup The columns were made with clear PVC pipe s (Schedule 40) with an inner diameter of 2.5 cm (Harvel, Easton, PA) By using a clear column, the potential color change of the sand could be observed during the experiment. Two sets of sand colu mns were built. In the first set leachate from C&D debris lysimeter DW12 ( see Chapter 3) with 12.4% drywall in the waste was used as the influent In another set leachate from lysimeter DW1 with 1% drywall in the waste was used. In both sets, there was a control column and an experiment al column. The control column was packed with acid rinsed uncoated sand, while the experiment al column was pa cked with hematite coated sand.
113 Each column was 30 cm long, capped with a rubber stopper at both ends. At each end of the column, glasswool was placed between the rubber stopper and the sand layer to prevent particles from entering the fluid system. The height of the sand layer wa s 25 cm. The sand (approximately 208 g in each column) was packed in a wet addition manne r as described below. De ionized water was first added into the column up to about half height of the column. Sand was then added from the top of the column while tapping the column gently to help settle the sand and t o avoid significant air spaces. The p o re volume was determined by measuring the volume of the water needed to saturate the sand. Table 5 1 shows the nomenclature of the columns. Column O peration The columns were first operated using a continuous saturated up flow mode (Phase 1) The columns we re first flushed with de ionized water to obtain background iron and arsenic concentrations. After that, leachate was used as the influent. As shown in Figure D 1 in Appendix D leachate was collected directly from a lysimeter by draining it gravimetricall y into a flask under a nitrogen atmosphere. The leachate was then pumped into both the uncoated and coated sand column s in parallel by using a peristaltic pump which was equipped with microbore PVC tubing sets with a inner diameter of 0. 25 m m. The average influent injection rate was one pore volume per day, which was about 50 mL/ day Meanwhile, leachate in the flask was collected through the sampling port for further analysis, which included pH, DO, ORP, sulfide, ferrous iron, and metals. The inset in Figu re D 1 shows the description of the ports for the leachate collection In the second phase (Phase 2) of operation leachate was pumped into each column within 30 minutes. The leachate in the column was left undisturbed for 24 hours before effluent collecti on and next cycle of leachate injection.
114 Effluent A nalysis The collected effluent was either analyzed right after collection or preserved for further analysis. The parameters that were measured right after collection included pH, ORP, sulfide, ferrous iro n, and arsenic speciation. Effluent was also preserved with nitric acid for analysis of metals after digestion The measurement of pH was conducted using an Orion 5 Star portable meter (Thermo Scientific, Beverly MA). Sulfide and ferrous iron concentratio ns were measured using a DR 4000 Spectrophotometer (Hach Company). Sulfide was measured using the USEPA methylene blue method (Hach Method 8131). COD measurement used the USEPA reactor digestion method (Hach Method 8000). Ferrous iron was measure d us ing th e Phenanthroline Method (Hach Method 8146). Arsenic speciation method developed in Chapter 2 was used to identify thioarsenic speciation in the leachate and in the effluent. Metal analysis followed the USEPA Method 6010B after samples were digested accordi ng to the USEPA Method 3050. XPS Characterization of C oated S and At the end of the column operation, coated sand at the bottom of each coated sand column was sampled and rinsed with de ionized water inside a glove box filled with nitrogen. The sand was the n dried with a nitrogen gas flow. X ray photoelectron spectroscopy (XPS) was performed on a 5100 XPS System ( Perkin Elmer Inc., Eden Prairie, MN ). The X ray source was a monochromatized Al K ray (1486.6 eV). The instrument was operated at a pressure of 10 8 10 9 torr. Sand samples were mounted onto the sampling stage using a carbon tape. Both survey and narrow scans were acquired. Survey scans were used to determine the range and approximate abundance
115 of elements on surface. Narrow scans were used to determine the chemical or oxidation state s of the elements. A free software AugerScan 3.2.0 Demo (RBD Instruments, Bend, OR) was used to fit the spectra of narrow scans after baseline subtraction Results and Discussion General T rend of E ffluent P a rameters During Phase 1 operation, the pH of the influent leachate ( Figure 5 1 ) follow ed a pattern similar to the one o bserved in Chapter 3, i.e., leachate with low er sulfide (LS) level h a s lower pH than that of high sulfide (HS) level leachate. This again indicate s that sulfide level may play a key role in controlling pH in C&D debris landfill leachate. The pH values of effluent, however, all increase, no matter what the influent sulfide level is No effort was expended to experimentally determine the reas on for the pH increase in the effluent. However, the slow oxidation or precipitation of some metal cations, such as Fe 2+ or Ca 2+ may contribute in part to pH increase. Changes in microbial activity may also play a role. ORP levels of influent also cor rela te with sulfide levels. The ORP value is lower for influent with high sulfide concentra tion. Compared to influent, ORP values for effluent increase and show more variability, indicating potential air intrusion into the system during overnight continuous fl ow. This was also evidenced by the near non detection of ferrous iron or sulfide in the effluent. High sulfide influent had higher con ductivity which was possibly due to the presence of more dissolved calcium and sulfate (or sulfide) ions as a result of t he high percentage of drywall in DW12 sulfide influent and effluent. For high sulfide influent, there were slight decreases in
116 conductivity. This may be due to the transformation or retention of sulfid e anions within the column or liquid lines. Iron M obilization Iron mobilization was only investigated in Phase 2 operation, where discontinuous flow mode was employed Ferrous concentrations in the influent, which was directly from the leachate of DW1 and DW12 lysimeters, were very different. Ferrous concentration in DW12 leachate was below 0.3 mg/L all the time compared with about 6 mg/L ferrous in DW1 leachate. The production of ferrous is favored by reducing conditions according to pe pH diagram. Howeve r, the more reducing condition in lysimeter DW12, as seen from the lower leachate ORP values in Figure 5 1 release more ferrous. This could be due to the high levels of sulfide in DW12, which can react with ferrous iron to form insoluble iron sulfi de minerals. The inverse correlation between sulfide and ferrous levels as shown in Figure 5 2 indicates that sulfide level is a n important controlling factor in iron mobilization in C&D debris landfills. In contrast to low ferrous levels which are limi ted by the high concentration of sulfide as mentioned above, ferrous concentration s in the effluent from both hematite coated sand columns increased (Figure 5 3 ) regardless of whether high level (HS CS) or low level (LS CS) sulfide influent was used The increase was more dramatic in the HS CS column, which had lower f errous concentration s and higher level s of sulfide in the influent (better seen in Figure 5 4 ). The mobilization of iron actually was not surprising, since the influent s for both columns were reducing, especially for the more reducing HS CS column. However, the result was somewhat contradictory to what was observed in the leachate as described above, where lower levels of ferrous were seen when sulfide level s were higher. This should simply me an that there is a balance or
117 equilibrium between the two factors sulfide and ferrous. Whenever one of them is prevalent, the other one will be limited. When using discontinuous flow, sulfide levels (Figure 5 3 ) from effluent in the uncoated sand column (H S U with the continuous flow mode in which case sulfide was almost non detectable after pass ing through the columns. This indicate s that the discontinuous flow mode was more effective in preventing air intrusion into the system. However, the sulfide level decreased dramatically after passing through coated sand column, as seen especially for HS CS column. The large difference in sulfide levels between coated sand and uncoated sand columns again suggested the formation of i ron sulfide minerals, which consumed much of sulfide in the influent. In terms of iron reduction, hematite is usually considered as a relatively stable iron oxide mineral compared with other minerals, such as ferrihydrite and lepidocrocite (Cornell and Schwertmann, 2003) The reductive dissolution of iron oxide minerals may occur either through chemical or microbial pathway s In the current experiment, oxidation reduction potentials in both DW1 and DW12 le achate were negative (reducing) enough to cause iron reduction. Meanwhile, microorganisms such as iron reducing bacteria or sulfate reducing bacteria may prosper within the operat ed s imulated landfills and may aid in iron reduction. Higher level of sulfide in the leachate of DW12 than that in DW1 may contribute more to ferrous production in the coated sand column. A rsenic Removal on Hematite coated Sand Arsenic speciation analysis was unable to d etect thioarsenic anions in any of the effluents. This was n ot surprising, as only a small amount of thioarsenic w ere present
118 even in the influent. Based on arsenic speciation analysis (see Chapter 3) using arsenic cartridge sieves, arsenite and organic arsenic species were predominant (over 70% 80%) in all leach ate samples. Total arsenic concentration change over time for both Phase 1 and Phase 2 is shown in Figure 5 6 The influent from DW1 (LS) had more arsenic than the influent from DW12 (HS). Uncoated sand columns (HS UCS, LS UCS) showed almost no ability to retain arsenic. Both coated sand columns (HS CS, LS CS) showed some ability to remov e arsenic. Better removal efficiency was seen for the column infiltrated with the low sulfide influent (LS CS). R emoval efficiency w as calculated based on the total arsenic in the influent and effluent ( Figure 5 6) As numerous studies have reported that iron sulfide can adsorb arsenic anions ( Bostick 2003, Han 2011) the formation of iron sulfide did not seem to provide better arsenic removal efficiency than that of hematit e in this experiment. This matched previous observation s in Chapter 4, where lower arsenic adsorption was found under sulfidic conditions when iron sulfide minerals could form. It should be noted that in both investigations by Bostick (2003) and Han (2011) the experimental systems did not have free sulfide. They simply explored the adsorption of arsenic on iron sulfide mineral inhibit ed the adsorption of arsenite on iron sulfide surface. Two possible reasons for the inhibition of arsenite adsorption were suggested to be competitive sorption from sulfide anions and suppressed mineral dissolution of iron sulfide. In th is study two of the reasons accounting for the lower ar senic removal can be propos ed. As the similar observation was shown in Chapter 4, the formation of iron sulfide may inhibit the ligand exchange between mineral surface and arsenite/arsenate
119 and consequently makes arsenic adsorption unfavorable. In the pres ence of sulfate reducing bacteria, hematite was shown to dissolve and form iron sulfide minerals on the surface of iron oxide (Neal et al., 2001) As high sulfide solutions were used in the adsorption study, it is very likely that iron sulfide was formed on the coated sand surface. Another reason for the low arsenic removal may be due to the dissolution of the hematite on which arsenic adsorption occurred. After the operation of columns was completed, t he column sand was collected, digested and analyzed for iron and arsenic content. As shown in Figure 5 7 even though the relative fraction is minor, iron dissolution decreased the iron content in the coated sand. The solid phase iron content on the HS CS sand was 50 80 mg/kg less than that of the LS C S sand. The ferrous concentration changes between influen t and effluent were ~6 mg/L and ~18 mg/L for the LS CS column and the HS CS column, respectively. For both the LS CS and HS CS columns, the sand from the top part of the column contains less iron than that in the bottom part. This was somewhat unexpected, since the influent would become less reducing as it moved down gradient and less iron would be reduced near the top of the column. However, as dissolved iron moved through the column, secondary phases may form as suggested by Benner et al ( (2002) ). The top part of the dissolved iron would just pass through the column without any chance to form secondary phases. Surface A nalysis of C olumn C oated S and by XPS As mentioned earlier the formation of iron sulfide minerals was presumably one contribut ing factor to several measured parameters including sulfide, ferrous and arsenic levels Similar to what was found in adsorption experiments in Chapter 4, a black substance was found in the column infiltra ted with high level s of sulfide influent
120 (HS CS) (Figure D 2 in Appendix D) The sand samples from the bottom of columns LS CS and HS CS were cleaned and dried ( Figure D 3 in Appendix D ) for XPS analysis. The objective was to determine the oxidation or che mical state of iron and sulfur on the surface of coated sand. Figure 5 8 shows the narrow scan spectra and their fitted spectra of three sand samples: control sand ( original coated sand ) LS CS sand and HS CS sand. The notation, such as O1s or Fe2p3, repre sents electron valence band from which electrons were excited under X ray and released and detected. The v ertical axis is the intensity and t he h orizontal axis is the binding energy (BE) of electrons of the specific valence band. Binding energy represents how strong ly an electron is bound to the valence band. For a specific element and a specific valence band, a change of the oxidation state or chemical surroundings can change the BE value. A h igher value of BE indicate s a more oxidized state and vice versa Binding energies of iron, oxygen, and sulfur species found in relevant literature ar e listed in Table 5 2. The fitted peak positions and potential species were listed in Table 5 3. Assignment s of the fitted peaks were compared with the reference values. Both Fe2p3/2 spectra of the control and LS CS samples can be fitted with one major peak at around 714 eV and a couple of other minor peaks at higher BE values. The major peak represent s a Fe2p3/2 peak in Fe 2 O 3 based on a compari son with reference values after calibration. The spectrum of HS CS sand, however, showed an almost 5 eV shift toward lower BE values with a major peak at 709.1 eV. The corrected BE value of the major peak is coincident with some iron sulfide minerals such as pyrite (FeS 2 ) and
121 macki nawite (FeS) (Neal et al., 2001) The second majo r peak at 711 eV and other minor peaks indicated that multiple iron species were present on the surface. In the O1s spectra, the major peak for the control sand was 536.4 eV, which was assigned to SiO 2 The 533.3 eV peak was assigned to Fe 2 O 3 as the second major peak. Tow other peaks at 534.7 eV and 535.9 eV were tentatively assigned to surface adsorbed hydroxyl groups and water molecules, respectively. In both the LS CS and HS CS sand, no SiO 2 signal was observed. In the LS CS spectrum, the second major pe ak was assigned to Fe 2 O 3 which appeared as a shou l der similar to that of the control sand. I n the HS CS spectrum, however, the peak shape was not similar to those seen in the control or LC CS. A second peak at 533 eV was used to fit the spectra, but it wa s less likely from Fe 2 O 3 O nly HS CS sand showed a S2p signal, and the spectrum of which is shown in Figure 5 9 T wo major peaks at 162.6 eV and 163.7 eV were possibly due to reduc ed sulfur compounds. S2p BE values for iron sulfide minerals have been repor ted to be around 161 eV to 163 eV. The observed BE values for HS CS cannot be from more oxidized forms such as sulfur or sulfate, which have higher values around 164 eV to 168 eV. No observable evidence indicates that arsenic iron co precipitates were form ed. Summary I ron mobilization in C&D debris landfills and in soils underneath could be affected by sulfide levels. T he relationship between sulfide and ferrous iron is controlled by the balance between the two entities. H igh levels of sulfide can either in hibit or promote mobilization of iron in landfill conditions depending on the relative amount of the available iron minerals If iron oxide minerals are limited, high levels of sulfide will limit the dissolution of iron. On the other hand, if iron oxide m inerals are in excess, high
122 levels of sulfide may significantly increase iron reductive dissolution. W hen evaluating the potential effect of C&D debris landfills on iron mobilization, the limiting factor must be determined first. T he black minerals which formed when leachate with high levels of sulfide passed through hematite coated sand were analyzed by XPS. The results indicated the black minerals were iron sulfides. N o t enough evidence was found to show that arsenic co precipitates with iron or sulfur. Besides the formation of solid phase sulfide minerals, other sulfur species, such as polysulfides or thiosulfate, were probably formed on hematite surface. These sulfur species can potentially increase the mobility of some heavy metals according to literat ure. The removal efficiency of arsenic from C&D debris landfills on hematite coated sand columns was calculated. Solid phase arsenic content analysi s indicated that arsenic uptake on coated sand was w ay below adsorption saturation. Better removal was seen when influent sulfide level was lower, even though the initial arsenic concentration was higher in the influent T he formation of iron sulfide minerals didn t seem to promote arsenic adsorption on coated sand; instead, it seemed to inhibit the retention of arsenic on hematite T his was similar to the results from adsorption experiments. The more reducing condition s of the influent with a higher level of sulfide caused more iron dissolution, which also possibly contributed to the lower efficiency in arsenic removal.
123 Figure 5 1 Parameters of influent and effluent in Phase 1 at continuous flow. (A) pH; (B) ORP; (C) Conductivity. LS : influent with low sulfide (DW1 leachate); LS CS : effluent from hematite coated sand column with low sulfide influent; LS UCS : effluent from uncoated sand column with low sulfide influent; HS : influent with high sulfide (DW12 leachate); HS CS : effluent from hematite coated sand column with high sulfide influent; HS UCS : effluent from uncoated sand column with low sulfide influent
124 Figure 5 2 Ferrous and sulfide in influent. The circle on the upper left corner includes influent from DW12 lysimeter; the circle on the bottom right corner includes influent from DW1 lysimeter.
125 Figure 5 3 Ferrous and sulfide of influent and eff luent in Phase 2 (A) Sulfide; (B) Ferrous. LS : influent with low sulfide (DW1 leachate); LS CS : effluent from hematite coated sand column with low sulfide influent; LS UCS : effluent from uncoated sand column with low sulfide influent; HS : influent with hi gh sulfide (DW12 leachate); HS CS : effluent from hematite coated sand column with high sulfide influent; HS UCS : effluent from uncoated sand column with low sulfide influent.
126 Figure 5 4 Sulfide and ferrous concentration in influent and effluent LS: influent with low sulfide (DW1 leachate); HS: influent with high sulfide (DW12 leachate); LS CS: effluent from hematite coated sand column with low sulfide influent; HS CS: effluent from hematite coated sand column with high sulfide influent.
127 Figure 5 5 Leachate and effluent parameters: total arsenic In Phase 1, the columns were operated in continuous flow mode with flow rate 1 pore volume per day. In Phase 2, the columns were operated in discontinuous flow mode with average contact time of 24 hours. LS : influent with low sulfide (DW1 leachate); LS CS : effluent from hematite coated sand column with low sulfide influent; LS UCS : effluent from uncoated sand column with low sulfide influent; HS : influent with high sulfide (DW12 leachate); HS CS : effluent from hematite coated sand column with high sulfide influent; HS UCS : effluent from uncoated sand column with low sulfide influent.
128 Figure 5 6. Relationship between influent sulfide level and arsenic removal efficiency of hematite coated sand columns. A rsenic removal efficiency is calculated based on the total arsenic amount in all the influent and effluent samples. Error bars of sulfide concentration represent standard deviations of all sulfide concentrations of all influent samples. LS CS : hematite coat ed sand column with low sulfide influent from DW1; HS CS : hematite coated sand column with high sulfide influent from DW12. Figure 5 7 Solid phase arsenic and iron content in hematite coated sand columns after column operation.LS CS: sand from hematite coated sand column with low sulfide influent (from DW1 leachate) ; HS CS: sand from hematite coated sand column with high sulfide influent (from DW12 leachate) Sand from both top and bottom parts (within about 2 cm from each end) are collected and analyze d by ICP AES. Error bars represent standard deviations of triplicate results.
129 Figure 5 8 XPS narrow scan O1s and Fe2p3 spectra of hematite coated sand Binding energies in these spectra are uncorrected. Component bands were fitted to spectra after base line subtraction. Assignments to component bands were compared to reference values (See Table 5 2 and Table 5 3) Control: original hematite coated sand; LS CS: sand from the bottom end of h ematite coated sand column with low sulfide influent (from DW1 lea chate) ; HS CS: sand from the bottom end of h ematite coated sand column with high sulfide influent (from DW12 leachate)
130 Figur e 5 9 XPS narrow scan S2p spectra of H S CS sand Binding energy in the spectrum is uncorrected. Component bands were fitted t o spectra after baseline subtraction. Assignments to component bands were compared to reference values (See Table 5 2 and Table 5 3). HS CS: sand from the bottom end of h ematite coated sand column with high sulfide influent (from DW12 leachate) No S2p ban d was detected in Control sand and LS CS sand.
131 Table 5 1. Nomenclature of sand columns Name LS CS LS UCS HS CS HS UCS Sand Used Coated Sand U n coated Sand Coated Sand U n coated Sand Leachate Used 1% drywall lysimeter 1% drywall lysimeter 12.4 drywall lysimeter 12.4 drywall lysimeter Table 5 2. XPS reference binding energies of iron, oxygen, and sulfur Element Species BE (eV) Reference Fe Fe 2 O 3 711.1 McIntyre and Zetaruk, 1977 710.4 Nefedov, 1977 711.6 Mills and Sullivan, 1983 FeS 2 707.25 B rion, 1980 Fe 0.89 S 708.5 Buckley and Woods, 1985 FeS 710.5 Carver, 1972 FeS 2 707.5 Buckley and Woods, 1987 O Fe 2 O 3 529.9 Mills and Sullivan, 1983 SiOH 531.3 Han, 2011 H 2 O 532.3 SiO 2 532.1 S Fe 0.89 S 161.6 Buckley and Woods, 1985 FeS 2 162.8 Buckley and Woods, 1987 FeS 160.8 Yu, 1990 FeS 2 162.4 van der Heide 1980 S n 2 163.3 Pratt 1994 S 2 O 3 2 164 Manocha and Park, 1977 S 0 164.6 Manocha and Park, 1977
132 Table 5 3. XPS peak assignment and percentages of iron, oxygen, and sulfur of control sand, LS CS sand, and HS CS sand Sample Fe O S BE (eV) Percent (%) Species BE (eV) Percent (%) Species BE (eV) Percent (%) Species Control 709.99 73.9 Fe 3+ O 529.42 26.2 Fe 2 O 3 711.66 6.8 530.83 11.2 SiOH 712.95 19.3 532.04 2.9 H 2 O 532.56 59.7 SiO 2 LS CS 709.81 66.3 Fe 3+ O 529.7 24.7 Fe 2 O 3 712.05 33.7 532.24 75.3 SiO 2 HS CS 707.19 40.4 Fe 2+ S 531.36 33.3 FeOOH/SO 4 2 160.64 34.9 S 2 709.08 30 532.8 66.7 SiO 2 161.82 47.1 S 2 2 710.82 7.4 Fe 3+ O 163 .02 12.2 S n 2 712.01 18.4 164.12 5.8 S 2 O 3 2 /S 0 714.29 3.7
133 CHAPTER 6 SUMMARY AND CONCLUSIONS Summary Arsenic, as a toxic and carcinogenic element, was used extensively prior to 2004 as a component of the preservative in manufacturing chro mated copper arsenate (CCA) treated wood. The estimated annually disposal of about 10 million m 3 CCA treated wood (Jambeck et al., 2007) most of which has been disposed of in construction and demolition (C&D) debris landfills, poses gre at health and environmental concern. Even though it has long been argued that high sulfide levels normally encountered in C&D debris landfills would immobilize arsenic by forming arsenic sulfide precipitates, previous investigations on simulated landfills indicate d that arsenic concentration in landfill leachate is significantly elevated due to the disposal of CCA treated wood. Meanwhile, recent theoretical and experimental results (Stauder et al., 2005; Wallschlager and Stadey, 2007; Wilkin et al., 2003) suggest ed that, in addition to insoluble arsenic sulfide minerals, soluble thioarsenic anions may be formed and in fact, predomina te under reducing conditions with high sulfide concentrations. The behavior of arsen ic in sulfidic conditions and the formation of thioarsenic species could affect the availability and mobility of arsenic in C&D debris landfill environment. However, no previous work has been carried out on the identification of thioarsenic species and how sulfide levels affect arsenic leaching in C&D debris landfill leachate. The overall objective of this dissertation was to investigate the formation and identification of thioarsenic species and how sulfide levels affect arsenic mobility in C&D debris land fill environment.
134 There were four parts of the experiment conducted. The first part aimed to develop a relatively simple chromatographic method to identify and quantify potential thioarsenic anions in C&D debris landfill leachate. Sodium thioarsenate compo unds were synthesized and their identities were confirmed successfully by molecular mass spectrometry. A gradient elution method based on ion chromatography with conductivity detection was developed to separate thioarsenate anions. The method was evaluated in terms of common validation parameters including precision, linearity and range Using this ion chromatographic method, simulated landfill leachate was tested and thioarsenate anions were found. The spike of arsenite in leachate confirmed the transform ation from arsenite to thioarsenates. In the second part of the experiment, simulated C&D debris landfills were constructed in the laboratory to investigate how the percentage of drywall disposal affects sulfide levels and consequently arsenic leaching an d speciation. L eachate analysis results showed that the percentage of drywall in the waste and sulfide levels in the leachate were positively correlated A rsenic leaching, however, didn t have a simple linear relationship with drywall percentages At both low and high levels of sulfide, when sulfide level s w ere in a certain range of around 1000 g/L of sulfide. It is speculated that the formation of arsenic sulfide minerals and thioarsenic anionic species could account for at least part of the observation. The relative percentage of thioarsenic components was shown to be positively dependent on the sulfide concentration. Thioarsenate anions are more sensitive to air exposure, especially tetrathioarsenate. When performing total arsenic analysis, the p reservation method of the C&D debris
135 landfill leachate may affect the r ecovery of arsenic. Conventional acid preservation method may cause significant loss of arsenic and large var iation in analy tical results. The third part of the experiment was designed to evaluate the adsorption characteristics of thioarsenates. Hematite coated quartz sand was made in the laboratory. The adsorption isotherms of monothioarsenate, arsenite, and ar senate at different pH values were compared. S imilar adsorption pattern s w ere found for m onothioarsenate and arsenate, which both differ ed from arsenite. The adsorption capacities of both arsenate and monothioarsenate showed a decreasing trend with increas ing pH. Arsenite adsorption on hematite coated sand in sulfide solution s at pH 7 and pH 10 showed slow increase s initially and lower adsorption capacities, compared to the adsorption in non sulfide solution s At pH 5, however, a significantly large adsorpt ion capacity and a sharp increase w ere seen with increasing arsenic concentration. The slow increase of arsenite adsorption was probably due to the formation of iron sulfide minerals. The large adsorption capacity at pH 5 was suggested to be due to the pre cipitation of arsenic sulfide minerals. In the last experiment arsenic and iron mobility in soils underneath unlined C&D debris landfills was explored by flush ing leachate from simulated landfills through hematite coated sand columns. I ron mobilization i n either C&D debris landfills or in soils underneath could be affected by sulfide levels. H igh levels of sulfide can either inhibit or promote mobilization of iron in landfill conditions, depending on the relative abundance s between sulfide and iron oxide minerals I ron oxide minerals, such as hematite, can be transformed to iron sulfide minerals under high levels of sulfide, as evidenced by XPS analysis results. T he formation of iron sulfide minerals did not seem to promote arsenic
136 adsorption on coated san d; instead, it seemed to inhibit the retention of arsenic. The dissolution of iron oxide s may be another reason for low arsenic removal efficiency observed in the sand column flushed with higher level of sulfide leachate. T here was not sufficient evidence to demonstrate the formation of arsenic co precipitates with iron or sulfur. Conclusions As described in detail in previous chapters, this dissertation developed and used an ion chromatographic method to identify and quantify thioarsenate anions in the lea chate of C&D debris landfills. As the major factor contributing to sulfide production in C&D debris landfills, the effect of drywall disposal on the leaching and speciation of arsenic was studied. The adsorption of arsenic/thioarsenic on iron oxide mineral s in sulfid ic and non sulfid ic solutions was investigated, as it pertains to the mobility of arsenic in soils underneath unlined C&D debris landfills. Specifically, the following conclusions may be drawn: Molecular identities of four thioarsenate anions, m ono di tri and tetra thioarsenate, were confirmed using molecular mass spectrometry. A gradient elution ion chromatographic method using a conductivity detector and sodium hydroxide as the eluent was developed. The method was shown to be cap able of s eparat ing the four thioarsenate anions from common inorganic anions, such as chloride, nitrate, sulfide, sulfate and phosphate, with ideal resolution within a Based on the method validation results, the developed method had fairly good pre cision with about 5.0% relative standard deviation (RSD) at all the concentration levels tested. The linear dynamic ranges for di and tri thioarsenate were estimated to be 0.2 to 5.0 mg As/L and 0.1 to 1.8 mg As/L, respectively. Within the concentration r ange tested, no deviation was observed. The linear range for mono thioarsenate was about 0.1 to 2.0 mg As/L, according to the obtained curve. Significant peak tailing occurred at higher arsenic concentrations.
137 Thioarsenate anions were detected in leachate from laboratory simulated landfill leachate. The increase in peak intensity of thioarsenate after an a rsenite spike in landfill leachate confirmed they were arsenic species. Sulfide levels affect arsenic leaching in a counter intuitive way. Suppressed arse nic concentrations were seen in leachate samples with sulfide concentration s around 1000 g/L. Higher arsenic concentrations were seen for either lower or higher sulfide levels. The special shape relationship between sulfide and arsenic concentrations imply that arsenic leaching in C&D debris landfills is controlled by both insoluble arsenic sulfide precipitation and formation of soluble thioarsenic species. The percentage of drywall disposal in C&D debris landfills is positively cor related with sulfide concentration in leachate. Accordingly, higher concentrations of thioarsenates were found in leachate with higher levels of sulfide. Ferrous iron concentrations in C&D debris landfill leachate are inversely correlated with sulfide concentrations, possibly regulated by the equilibrium of iron sulfide precipitation. Thioarsenate anions, particularly tetrathioarsenate, were shown to be unstable when exposed to air. Storage in a refrigerator with minimum air exposure proved to be effective in preventing thioar senic species transformation. The selection of a preservation method is critical in total arsenic analysis for C&D debris landfill leachate with high levels of sulfide. The c ommonly used acid preservation method was shown to cause the loss of arsenic and l arger variations. The hydrogen peroxide pretreatment before acid preservation improved arsenic recovery and analysis precision. The adsorption behavior of monothioarsenate on hematite coated sand was found to be more similar to that of arsenate than that o f arsenite, with increasing adsorption capacity at lower pH. Arsenite, however, was shown to have the highest adsorption capacity at pH 7, compared to those at pH 5 and pH 10. At pH 7, Arsenite had a highe r adsorption capacity on hematite coated sand than either monothioarsenate or arsenate. At pH 7 and pH 10, arsenite adsorbs less, particularly at low arsenic concentrations, on hematite coated sand in sulfide solution than in a solution without sulfide. The formation of iron sulfide minerals was suggested as a possible contributing factor to the low adsorption. The much higher adsorption capacity of arsenite in sulfide than in non sulfide solutions at pH 5 was very likely due to the formation of arsenic sulfide precipitates. Arsenic removal efficiency was 1 8% in a hematite coated sand column when simulated C&D debris landfill leachate with a high sulfide concentration was used
138 as influent. The efficiency was 31% when leachate with low sulfide concentratio n was used, even though higher initial arsenic concent ration was seen in the latter leachate influent. X ray photoelectron spectroscopy results showed that hematite was transformed to iron sulfide minerals in the sand column using leachate with high sulfide levels. The formation of iron sulfide possibly accou nted for the low arsenic removal efficiency when sulfide levels were high in leachate. Significant increases in ferrous iron concentration were seen in effluents from both hematite coated sand columns, but more drastically for the column with high sulfide leachate as the influent. Iron reductive dissolution was also suggested to be one reason for the lower arsenic removal efficiency seen in the column using leachate with high sulfide concentrations. F uture Work The research described in this dissertation ha s provided some preliminary understanding of thioarsenic species, their formation, stability, adsorption, and detection, in a C&D debris landfill environment. However, the work conducted was still far from being extensive and elaborative. The following top ics may deserve further in depth research efforts. Method Development : The ion chromatographic method developed in this dissertation separates thioarsenate anions from other inorganic arsenic anions including arsenate and arsenite. However, the method is n ot able to identify commonly seen organic arsenic species, such as monomethylarsonic acid (MMA) and dimethylarsonic acid (DMA). Usually, methods based on high performance liquid chromatography (HPLC) (Komorowicz and Baralkiewicz, 2011) and capillary electrophoresis (CE) (Kannamkumarath et al., 2002) have been widely employed in detecting both inorganic and organic arsenic species. Unfortunately, the application of those HPLC or CE based methods in testing thioarsenate anions has not been reported. Work needs to be done
139 to check the suitability of those methods in the analysis of a variety of arsenic species including thioarsenics, or new method s should be developed. Surface Complex ation Modeling : Extensive studies have been conducted in modeling the surface complexation processes between arsenite/arsenate in solution and the surface of iron oxide or clay minerals. In contrast, the surface complexation model between thioarsenic speci es and iron oxide/clay minerals is rarely seen. The same holds true between arsenic or thioarsenic species and iron sulfide minerals. The findings from such studies would help in understand ing macroscopic interactions between arsenic and soils in the envir onment and aid in hydrogeochemical modeling. Toxicity of Thioarsenic : Compared to the already established data of arsenite and arsenate toxicity, very little research ha s been done to investigate the toxicity of thioarsenic species. Based on Microtox acute toxicity test which us ed various concentration ratios of sulfide and arsenite, it was suggested that the formation of thioarsenic species reduce s the toxicity (Rader et al., 2004) In another study (Planer Friedrich et al., 2008) thioarsenates showed varying toxicities, with monothioarsenate being less toxic than arsenate or arsenite, while trithioarsenate ha d a simi lar toxic ity However, lower toxicity was observed in arsenite solutions with a higher ratio of sulfide to arsenite. In terms of organic thioarsenicals produced during metabolism, it was found that monomethylmonothioarsonic acid (MMMTA) was a more toxic ar senic metabolite than non thiolated monomethylarsonous acid (MMA) (Naranmandura et al., 2010) It is believed that the formation of an arsenic sulfur bond accounts for the toxicity in the body; however, more research focusing on how thioarsenic species bio trans forms in mammal ian body systems is needed
140 Puzzle of Thioarsenate or Thioarsenite : The existence and the exact form s of thioarsenic species still seem mysterious. The i nitial speculation was that arsenic exists as thioarsenite or polymeric thioarsenic spec ies (Eary, 1992) due to the hig h concentration of sulfide and highly reducing condition s However, experimental data gave two totally opposite answers. Chromatographic based methods found that thioarsenates are formed by the reaction between arsenite and sulfide (Stauder et al., 2005; Wallschlager and Stadey, 2007) and a mechanism explaining the transformation between arsenite and thioarsenate was proposed (Stauder et al., 2 005) On the other hand, spectroscopic methods on bulk solutions suggested the sole existence of thioarsenites rather than thioarsenates (Beak et al., 2008; Bostick et al., 2005) A recent paper (Helz and Tossell, 2008) using theoretical thermodynamic calculations proposed that both thioarsenates and thioarsen ites are possible and they can occur simutaneously, depending on the exact conditions such as pH and total sulfide concentration More research is needed to solve this dilemma.
141 APPENDIX A ADDITIONAL MATERIAL FOR THIOARSENATE IDENTIFICATION An Appendix is formatting requirements are relaxed so that some documents may be presented in their original format. However, the margin requirements remain in effect throughout the entire document regardless of Figure A 1. Ion chromatographic separation and mass spectrometric analysis of anions
142 Figure A 2. A diagram of ASRS working principle The diagram shows that anions (chloride in this case) are separated with sepa ration column and converted into neutral species (HCl in this case) after the suppressor. The ASRS eliminates adverse salt effect on mass spectrometer caused by some common cations such as sodium or potassium in samples.
143 Figure A 3. Synthesized thioa rsenic compound: monothioarsenate Figure A 4. Ion chromatogram of synthesized thioarsenic compound (top) Monothioarsenate; (bottom) arsenate The comparison of the two chromatograms show the separation between arsenate and monothioarsenate anions. Th e small peak at ~20 minute is dithioarsenate anion.
144 Calibration of mass error in mass spectrometric data: An error (mass accuracy) is calculated based on the observed and the theoretical masses. When a known compound is used to calibrate the instrument, t he mass accuracy indicates the deviation of the instrument response from the known calculated monoisotopic mass. Usually the mass accuracy is expressed in parts per million (ppm) as shown in the equation below: (A 1) For identification purposes, a mass accuracy of less than 10 ppm can be usually considered as a good confirm ation of the formula. Table A 1 Determination of injection volume Original Arsenate Concentration (mg/L) Original Arsenic Concentration (mg/L) Fraction Volume (mL) Fraction Arsenic Concentration Injection Volume (mL) Average Injection Volume 100.00 53.93 25.00 159.6888 0.074024 73.78 158.0484 0.073263 159.7212 0.074039 Table A 2 Concentrations and corresponding peak areas of thioarsenates Replicate Monothioarsenate Dithioarsenate Trithioarsenate Tetrathioarsenate Concentrat ion (mg As/L) Area Concentration (mg As/L) Area Concentration (mg As/L) Area Concentration (mg As/L) Area 1 2.0914 92312.8 6.1379 610859.9 3.401 9 846781.8 2.103 9 251143.4 2 2.3507 97317.6 6.1179 696594.4 3.3729 963436.4 1.7203 287186.3 3 2.1135 96249.5 5.800 8 736451.4 3.2137 1011921 1.9029 304522.4 Average 2.1852 95293.3 6.018 9 681301.9 3.3295 940713.2 1.9090 280950.7 Standard deviation 0.1438 0.1892 0.1013 0.1919
145 APPENDIX B ADDITIONAL MATERIAL FOR SIMULATED LANDFILL EXPERIMENT Figure B 1 Eh pH diagram of As S O system. The diagram was obtained using program. Shaded green box shows the range within which C&D debris landfill ram, orpiment or realgar are the predominant species under common C&D debris landfill conditions.
146 APPENDIX C ADDITIONAL MATERIAL FOR ADSORPTION EXPERIMENT Figure C 1. XRD spectrum of hematite. The number of each peak represents the crystal planes in h ematite. The position and intensity pattern of the eight peaks match the structure of hematite. Figure C 2. Adsorption kinetics of arsenate on coated sand at pH5. Error bars represent the standard deviations of triplicate results.
147 Figure C 3. Col or change with increasing arsenic concentration in arsenite adsorption experiment in sulfide solution The color changes from orange red (nominal initial arsenic concentration 37.5 mg/L) to black (nominal initial arsenic concentration 0.75 mg/L) with decre asing arsenite concentration. The color changes indicate the solid phase transformation of hematite occurs.
148 APPENDIX D ADDITIONAL MATERIAL FOR HEMATITE COATED SAND COLUMNS EXPERIMENT Figure D 1 Column experiment setup Leachate is taken directly from the bottom of lysimeter and pumped into sand columns (Experiment column and Control column) using up flow mode. The leachate container is under nitrogen atmosphere during the time of operation.
149 Figure D 2. Appearance of coated sand columns after opera tion. LS CS: hematite coated sand column with low sulfide influent (from DW1 leachate); HS CS: hematite coated sand column with high sulfide influent (from DW12 leachate). The comparison of the color between the bottom parts of the 2 columns shows solid ph ase iron oxide transformation in HS CS column.
150 Figure D 3. Cleaned and dried sand from operated columns for XPS analysis. LS CS: sand from hematite coated sand column with low sulfide influent (from DW1 leachate); HS CS: sand from hematite coated sand column with high sulfide influent (from DW12 leachate).
151 LIST OF REFERENCES 2001. FR Federal Register, pp. 6975 7066. 2003a. Code of Federal Regulations, CFR 261.4(b)(9). 2003b. FR Federal Register, pp. 17366 17372. Agency, U.S.E.P., 2001. Preliminary evaluation of the non dietary hazard and exposure to children from contact with chromated copper arsenate (CCA) treated wood playground structures and CCA contaminated soil. U.S. Environmental Protection Agency (EPA), Washington D.C. Ant elo, J., Avena, M., Fiol, S., Lopez, R., Arce, F., 2005. Effects of pH and ionic strength on the adsorption of phosphate and arsenate at the goethite water interface. Journal of Colloid and Interface Science 285, 476 486. Beak, D.G., Wilkin, R.T., Ford, R. G., Kelly, S.D., 2008. Examination of Arsenic Speciation in Sulfidic Solutions Using X ray Absorption Spectroscopy. Environ. Sci. Technol. 42, 1643 1650. Benner, S.G., Hansel, C.M., Wielinga, B.W., Barber, T.M., Fendorf, S., 2002. Reductive Dissolution and Biomineralization of Iron Hydroxide under Dynamic Flow Conditions. Environmental Science & Technology 36, 1705 1711. Bostick, B.C., Fendorf, S., 2003. Arsenite sorption on troilite (FeS) and pyrite (FeS 2 ). Geochimica et Cosmochimica Acta 67, 909 921. Bost ick, B.C., Fendorf, S., Brown, G.E., 2005. In situ analysis of thioarsenite complexes in neutral to alkaline arsenic sulphide solutions. Mineralogical Magazine 69, 781 795. Britton, H.T.S., Jackson, P., 1934. Physicochemical Studies of Complex Formation In volving Weak Acids. Journal of Chemical Society, 1048 1055. Brookins, D.G., 1988. Eh, pH Diagrams for Geochemistry. Springer Verlag, Berlin. Brooks, W.E., 2008. 2007 Minerals Yearbook: Arsenic, http://minerals.er.usgs.gov/minerals/pubs/commodity/arsenic/myb1 2007 arsen.pdf United States Geological Survey. Catalano, J.G., Park, C., Fenter, P., Zhang, Z., 2008. Simultaneous inner and outer sphere arsenate adsorpt ion on corundum and hematite. Geochimica et Cosmochimica Acta 72, 1986 2004.
152 Chakraborty, S., Wolthers, M., Chatterjee, D., Charlet, L., 2007. Adsorption of arsenite and arsenate onto muscovite and biotite mica. Journal of Colloid and Interface Science 309 392 401. Clark, C., Jambeck, J., Townsend, T., 2006. A Review of Construction and Demolition Debris Regulations in the United States. Critical Reviews in Environmental Science and Technology 36, 141 186. Clarke, M.B., Helz, G.R., 2000. Metal Thiometalate Transport of Biologically Active Trace Elements in Sulfidic Environments. 1. Experimental Evidence for CopperThioarsenite Complexing. Environ. Sci. Technol. 34, 1477 1482. Cornell, R.M., Schwertmann, U., 2003. The Iron Oxides: Structure, Properties, React ions, Occurrences and Uses, 2 nd Ed. ed. Wiley VCH, Weinheim, Germany. Cummings, D.E., Caccavo, F., Fendorf, S., Rosenzweig, R.F., 1999. Arsenic Mobilization by the Dissimilatory Fe(III) reducing Bacterium Shewanella Alga BrY. Environ. Sci. Technol. 33, 723 729. deLemos, J.L., Bostick, B.C., Renshaw, C.E., Sturup, S., Feng, X., 2006. Landfill Stimulated Iron Reduction and Arsenic Release at the Coakley Superfund Site (NH). Environ. Sci. Technol. 40, 67 73. Dixit, S., Hering, J.G., 2003. Comparison of Arsenic (V) and Arsenic(III) Sorption onto Iron Oxide Minerals: Implications for Arsenic Mobility. Environ. Sci. Technol. 37, 4182 4189. Dubey, B., 2005. Comparison of Environmental Impacts of Wood Treated with Chromated Copper Arsenate (CCA) and Three As Free Pre servatives. University of Florida, Gainesville, FL. Eary, L.E., 1992. The solubility of amorphous As 2 S 3 from 25 to 90 0 C. Geochimica et Cosmochimica Acta 56, 2267 2280. EPA, 1998. Characterization of Building Related Construction and Demolition Debris in t he United States, in: Agency, U.S.E.P. (Ed.). U.S. Environmental Protection Agency (EPA), Office of Solid Waste and Emergency Response, Washington DC. Flis, I.E., Mishchenko, K.P., Tumanova T.A., 1959. Dissociation of Arsenic Acid. Russian Journal of Ino rganic Chemistry 4, 120 124. Freundlich, H., 1926. Colloid and Capillary Chemistry. Methuen, London, England. Gammonsa, C.H., Frandsen, A.K., 2001. Fate and transport of metals in H 2 S rich waters at a treatment wetland. Geochem. Trans. 1.
153 Ghosh, A.K., Bhat tacharyya, P., 2004. Arsenate sorption by reduced and reoxidised rice soils under the influence of organic matter amendments. Environmental Geology 45, 1010 1016. Gupta, V.K., Saini, V.K., Jain, N., 2005. Adsorption of As(III) from aqueous solutions by iro n oxide coated sand. Journal of Colloid and Interface Science 288, 55 60. Harvey, C.F., Swartz, C., Badruzzamann, A.B.M., Keon Blute, N., Yu, W., Ashraf, A.M., Jay, J., Beckie, R., Niedan, V., BRABANDER, D., Oates, P.M., Ashfaque, K.N., Islam, S., Hemond, H.F., Ahmed, M.F., 2002. Arsenic Mobility and Groundwater Extraction in Bangladesh. Science 298. Helz, G.R., Tossell, J.A., 2008. Thermodynamic model for arsenic speciation in sulfidic waters: A novel use of ab initio computations. Geochimica et Cosmochimi ca Acta 72, 4457 4468. Helz, G.R., Tossell, J.A., Charnock, J.M., Pattrick, R.A.D., Vaughan, D.J., David Garner, C., 1995. Oligomerization in As (III) sulfide solutions: Theoretical constraints and spectroscopic evidence. Geochimica et Cosmochimica Acta 59 4591 4604. Herbel, M., Fendorf, S., 2006. Biogeochemical processes controlling the speciation and transport of arsenic within iron coated sands. Chemical Geology 228, 16 32. Hering, J.G., Kneebone, P.E., 2001. Biogeochemical Controls on Arsenic Occurrenc e and Mobility in Water Supplies, in: Frankenberger, W. (Ed.), Environmental Chemistry of Arsenic. Marcel Dekker, New York, pp. pp. 155 181. Hopenhayn, C., 2006. Arsenic in Drinking Water: Impact on Human Health. Elements 2, 103 107. Hounslow, A.W., 1980. Ground Water Geochemistry: Arsenic in Landfills. Ground Water 18, 331 333. Jambeck, J., 2004. The Disposal of CCA treated Wood in Simulated Landfills: Potential Impacts. University of Florida, Gainesville, FL. Jambeck, J., Weitz, K., Solo Gabriele, H., Tow nsend, T., Thorneloe, S., 2007. CCA Treated wood disposed in landfills and life cycle trade offs with waste to energy and MSW landfill disposal. Waste Management 27, S21 S28. Jang, Y. C., 2000. A Study of Construction and Demolition Waste Leachate from Lab oratory Landfill Simulators. University of Florida, Gainesville, FL.
154 Jang, Y. C., Townsend, T.G., 2003. Effect of Waste Depth on Leachate Quality from Laboratory Construction and Demolition Debris Landfills. Environmental Engineering Science 20, 183 196. J iang, W., Zhang, S., Shan, X. q., Feng, M., Zhu, Y. G., McLaren, R.G., 2005. Adsorption of arsenate on soils. Part 2: Modeling the relationship between adsorption capacity and soil physiochemical properties using 16 Chinese soils. Environmental Pollution 1 38, 285 289. Kannamkumarath, S.S., Wrobel, K., Wrobel, K., B'Hymer, C., Caruso, J.A., 2002. Capillary electrophoresis inductively coupled plasma mass spectrometry: an attractive complementary technique for elemental speciation analysis. Journal of Chromato graphy A 975, 245 266. Keimowitz, A.R., Simpson, H.J., Stute, M., Datta, S., Chillrud, S.N., Ross, J., Tsang, M., 2005. Naturally occurring arsenic: Mobilization at a landfill in Maine and implications for remediation. Applied Geochemistry 20, 1985 2002. K han, B.I., Jambeck, J., Solo Gabriele, H.M., Townsend, T.G., Cai, Y., 2006a. Release of Arsenic to the Environment from CCA Treated Wood. 2. Leaching and Speciation during Disposal. Environ. Sci. Technol. 40, 994 999. Khan, B.I., Solo Gabriele, H.M., Towns end, T.G., Cai, Y., 2006b. Release of Arsenic to the Environment from CCA Treated Wood. 1. Leaching and Speciation during Service. Environ. Sci. Technol. 40, 988 993. Kocar, B.D., Herbel, M.J., Tufano, K.J., Fendorf, S., 2006. Contrasting Effects of Dissimilatory Iron(III) and Arsenic(V) Reduction on Arsenic Retention and Transport. Environ. Sci. Technol. 40, 6715 6721. Komorowicz, I., Baralkiewicz, D., 2011. Arsenic an d its speciation in water samples by high performance liquid chromatography inductively coupled plasma mass spectrometry -Last decade review. Talanta 84, 247 261. Lal, R., Shukla, M.K., 2004. Principles of Soil Physics. Marcel Dekker, Inc., New York. Lan gmuir, I., 1918. The Adsorption of Gases on Plane Surfaces of Glass, Mica, and Platinum. Journal of the American Chemical Society 40, 1361 1403. Langner, H.W., Inskeep, W.P., 2000. Microbial Reduction of Arsenate in the Presence of Ferrihydrite. Environ. S ci. Technol. 34, 3131 3136.
155 Lee, M. K.S., James A.; Wilkin, Richard T.; Mohammad, Shahnewaz. 2005. Geochemical modeling of arsenic speciation and mobilization: Implications for bioremediation. in: O'Day, P.A. (Ed.), Advances in arsenic research : integ ration of experimental and observational studies and implications for mitigation American Chemical Society : Distributed by Oxford University Press, Washington, DC Lovley, D.R., 1991. Dissimilatory Fe(III) and Mn(IV) Reduction. Microbiology Reviews 55, 25 9 287. Lovley, D.R., Holmes, D.E., Nevin, K.P., Poole, 2004. Dissimilatory Fe(III) and Mn(IV) Reduction, Advances in Microbial Physiology. Academic Press, pp. 219 286. Luengo, C., Brigante, M., Avena, M., 2007. Adsorption kinetics of phosphate and arsenate on goethite. A comparative study. Journal of Colloid and Interface Science 311, 354 360. Manning, B.A., Fendorf, S.E., Goldberg, S., 1998. Surface Structures and Stability of Arsenic(III) on Goethite: Spectroscopic Evidence for Inner Sphere Complexes. Env ironmental Science & Technology 32, 2383 2388. Manning, B.A., Goldberg, S., 1997a. Adsorption and Stability of Arsenic(III) at the Clay Mineral Water Interface. Environ. Sci. Technol. 31, 2005 2011. Manning, B.A., Goldberg, S., 1997b. Arsenic(III) and Arse nic(V) Adsorption on Three California Soils. Soil Science 162, 886 895. Masue, Y., Loeppert, R.H., Kramer, T.A., 2007. Arsenate and Arsenite Adsorption and Desorption Behavior on Coprecipitated Aluminum:Iron Hydroxides. Environmental Science & Technology 4 1, 837 842. Moldovan, B.J., Hendry, M.J., 2005. Characterizing and Quantifying Controls on Arsenic Solubility over a pH Range of 1 11 in a Uranium Mill Scale Experiment. Environ. Sci. Technol. 39, 4913 4920. Naranmandura, H., Bu, N., Suzuki, K.T., Lou, Y., Ogra, Y., 2010. Distribution and speciation of arsenic after intravenous administration of monomethylmonothioarsonic acid in rats. Chemosphere 81, 206 213. Neal, A.L., Techkarnjanruk, S., Dohnalkova, A., McCready, D., Peyton, B.M., Geesey, G.G., 2001. Iro n sulfides and sulfur species produced at hematite surfaces in the presence of sulfate reducing bacteria. Geochimica et Cosmochimica Acta 65, 223 235. Nickson, R.T., McArthur, J.M., Ravenscroft, P., Burgess, W.G., Ahmed, K.M., 2000. Mechanism of Arsenic Re lease to Groundwater, Bangladesh and West Bengal. Appl. Geochem. 15, 403 413.
156 Nordstrom, D.K., Archer, D.G., 2003. Arsenic thermodynamic data and environmental geochemistry., in: Welch, A.H., Stollenwerk, K.G. (Eds.), Arsenic in ground water. Kluwer Academ ic Publications. and iron on dissolved arsenic concentrations in the shallow subsurface under changing redox conditions. PNAS 101, 13703 13708. Pierce, M.L., Moore, C.B., 19 80. Adsorption of arsenite on amorphous iron hydroxide from dilute aqueous solution. Environmental Science & Technology 14, 214 216. Pierce, M.L., Moore, C.B., 1982. Adsorption of arsenite and arsenate on amorphous iron hydroxide. Water Research 16, 1247 1 253. Planer Friedrich, B., Franke, D., Merkel, B., Wallschlger, D., 2008. Acute toxicity of thioarsenates to Vibrio fischeri. ENVIRONMENTAL TOXICOLOGY AND CHEMISTRY 27, 2027 2035. Ponthieu, M., Pinel Raffaitin, P., Le Hecho, I., Mazeas, L., Amouroux, D., Donard, O.F.X., Potin Gautier, M., 2007. Speciation analysis of arsenic in landfill leachate. Water Research 41, 3177 3185. Rader, K.J., Dombrowski, P.M., Farley, K.J., Mahony, J.D., Di Toro, D.M., 2004. Effect of thioarsenite formation on arsenic(III) tox icity. ENVIRONMENTAL TOXICOLOGY AND CHEMISTRY 23, 1649 1654. Raven, K.P., Jain, A., Loeppert, R.H., 1998. Arsenite and Arsenate Adsorption on Ferrihydrite: Kinetics, Equilibrium, and Adsorption Envelopes. Environmental Science & Technology 32, 344 349. Red man, A.D., Macalady, D.L., Ahmann, D., 2002. Natural Organic Matter Affects Arsenic Speciation and Sorption onto Hematite. Environmental Science & Technology 36, 2889 2896. Reese, J.R.G., 2002. Commodity Summaries 2002: Arsenic, http://minerals.er.usgs.gov/minerals/pubs/commodity/arsenic/160302.pdf United States Geological Survey. Rhue, D., Townsend, T., Wang, Y., Dubey, B., 2008. Soils Underneath Florida Landfills and T heir Role in the Occurence and Fate of Iron and Arsenic in Groundwater. Hinkley Center for Soils and Hazardous Waste Management, Gainesville, FL. Rochette, E.A., Bostick, B.C., Li, G., Fendorf, S., 2000. Kinetics of Arsenate Reduction by Dissolved Sulfide. Environ. Sci. Technol. 34, 4714 4720.
157 Scheidegger, A.M., Borkovec, M., Sticher, H., 1993. Coating of Silica Sand with Goethite: Preparation and Analytical Identification. Geoderma 58, 43 65. Schwedt, G., Rieckhoff, M., 1996. Separation of thio and oxothi oarsenates by capillary zone electrophoresis and ion chromatography. Journal of Chromatography A 736, 341 350. Schwertmann, U., Cornell, R.M., 2000. Iron Oxides in the Laboratory: Preparation and Characterization, 2 nd Ed. ed. Wiley VCH. Stauder, S., Raue, B., Sacher, F., 2005. Thioarsenates in Sulfidic Waters. Environ. Sci. Technol. 39, 5933 5939. Stollenwerk, K.G., Colman., J.A., 2004. Natural Remediation of Arsenic Contaminated Ground Water Associated with Landfill Leachate. U.S. Geological Survey Tossel l, J.A., 2001. Computing the Properties of the Copper Thioarsenite Complex, CuAsS(SH)(OH). Inorg. Chem. 40, 6487 6492. Townsend, T., Dubey, B., Tolaymat, T., Solo Gabriele, H., 2005. Preservative leaching from weathered CCA treated wood. Journal of Environ mental Management 75, 105 113. Townsend, T., Tolaymat, T., Solo Gabriele, H., Dubey, B., Stook, K., Wadanambi, L., 2004. Leaching of CCA treated wood: implications for waste disposal. Journal of Hazardous Materials 114, 75 91. Wallschlager, D., Stadey, C.J ., 2007. Determination of (Oxy)thioarsenates in Sulfidic Waters. Anal. Chem. 79, 3873 3880. Wang, S., Mulligan, C.N., 2008. Speciation and surface structure of inorganic arsenic in solid phases: A review. Environment International 34, 867 879. Wauchope, R. D., McDowell, L.L., 1984. Adsorption of Phosphate, Arsenate, Methanearsonate, and Cacodylate by Lake and Steam Sediments: Comparison with Soils. Journal of Environmental Quality 13, 499 504. Weber, W.J., Yong Chul, J., Townsend, T.G., Laux, S., 2002. Leach ate from Land Disposed Residential Construction Waste. Journal of Environmental Engineering 128, 237. Webster, J.G., 1990. The solubility of As 2 S 3 and speciation of As in dilute and sulphide bearing fluids at 25 and 90 C. Geochim. Cosmochim. Acta 54, 1009 1017. Wilkin, R.T., Wallschla¨ger, D., Ford, R.G., 2003. Speciation of arsenic in sulfidic waters. Geochem. Trans. 4, 1 7.
158 Williams, L.E., Barnett, M.O., Kramer, T.A., Melville, J.G., 2003. Adsorption and Transport of Arsenic(V) in Experimental Subsurface Systems. J. Environ. Qual. 32, 841 850. Wolthers, M., Charlet, L., van Der Weijden, C.H., van der Linde, P.R., Rickard, D., 2005. Arsenic mobility in the ambient sulfidic environment: Sorption of arsenic(V) and arsenic(III) onto disordered mackinawite. Ge ochimica et Cosmochimica Acta 69, 3483 3492. Wood, S.A., Taitb, C.D., Janecky, D.R., 2002. A Raman spectroscopic study of arsenite and thioarsenite species in aqueous solution at 25 0 C. Geochem. Trans 3, 31 39. Zhang, H., Selim, H.M., 2005. Kinetics of Ars enate Adsorption Desorption in Soils. Environ. Sci. Technol. 39, 6101 6108.
159 BIOGRAPHICAL SKETCH Jianye Zhang was from China, where he obtained his bachelor and master degrees in physical chemistry from Nanjing University and Peking University, respecti vely. After obtaining a master degree in organic chemistry from the Department of Chemistry at the University of Florida in 2007, Jianye started his PhD study in the Department of Environmental Engineering Sciences supported by the University Alumni Gradu ate Award.