ENVIRONMENTAL PROCESSES CONTROLLING THE FATE AND TRANSPORT OF ARISTOLOCHIC ACID IN AGRICULTURAL SOIL AND COPPER IN CONTAMINATED LAKE SEDIMENT By Chaiyanun Tangtong A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Environmental Engineering – Doctor of Philosophy 2014 ABSTRACT ENVIRONMENTAL PROCESSES CONTROLLING THE FATE AND TRANSPORT OF ARISTOLOCHIC ACID IN AGRICULTURAL SOIL AND COPPER IN CONTAMINATED LAKE SEDIMENT By Chaiyanun Tangtong Fate and transport of toxic chemicals are important processes that describe how chemicals move and transform in the environment. Environmental processes such as adsorption, solubility, complexation, dissolution/precipitation, oxidation/reduction, plant uptake and biodegradation play major roles in controlling the fate and transport of chemicals. Understanding these processes is essential to assess the potential of human health risks, the exposure pathway or even the methods for prevention and remediation of these risks. Environmental problems can be caused by organic or inorganic chemicals and they have different fate and transport behavior in the environment. In this study, the fate and transport of Aristolochic acids (AAs) in soil and copper in sediment were used as examples to show the different behavior of organic and inorganic chemicals that induced problems in environment. AAs were believed to be causal agents that induced Balkan Endemic Nephropathy (BEN) by food contamination. They are active chemicals in Aristolochia species plants which are extensively found in the endemic villages. This study examined the essential environmental partitioning processes that control fate and transport of AAs. The results showed that the octanolwater partitioning coefficient (Kow) of AAs decreased when pH increased, which indicated the different hydrophobicity between neutral and anion forms. This trend was similar to the soil water partitioning coefficient (Kd). Solubility (Sw) increased when pH increased. These suggested that AAs will be highly mobile in an alkaline environment. The Kow and Sw were increased when a calcium ion presented in solution. Even if AAs had a high sorption capacity to the soils, they had a high tendency to be desorbed too. The soil adsorption and desorption experiment indicated the cation bridging mechanism may play a major role in soil processes. Root exudates are not the main pathway that release AAs to the soil, but their seed decomposition can release a large amount of AAs, which can be degraded by microorganisms. The plant uptake showed AAs had high accumulation in the roots but less translocation to shoots. All evidence suggested the food contamination hypothesis is possible. Torch Lake, Houghton County, Michigan, which was impacted by copper mining waste, showed a persistent high level of copper in the top sediment. This copper can never be remediated by natural processes. I hypothesized that copper was released from mining waste by microbial mediate reactions and was sequestrated by organic matter and bacteria in the post sediment. To test this hypothesis, the fate and transport of copper were studied with the Phreeqc computer modelwhere the biogeochemical processes (e.g. complexation, precipitation, adsorption, reductive dissolution and biodegradation) were applied to the diffusion process. The results showed that oxidation/reduction conditions highly impact copper fate and transport in sediment. The TEAP process causes the sediment to enter the reducing condition and dissolve the copper waste to pore water. The bacteria in the top sediment adsorbed a large amount of copper but this sorption turned to more stable minerals as time proceeded. Due to the removal of copper by precipitation process, the copper which contributed the solid phases in the top sediment were expected from overlay water. This study showed that this modeling approach is an effective tool to describe the persistent high concentration of copper inTorch Lake sediment. ACKNOWLEDGEMENTS First of all, I would like to express my especially thank to my Advisors, Dr. Thomas C. Voice and Dr. David T. Long for their guidance and support over my study in the U.S. They always give me all aspect of learning. Without them, my dissertation could not be completed. I also would like to express special thank for my other committee members, Dr. Phanikumar S. Mantha and Dr. Terence L. Marsh, for their supports and valuable comments and suggestions for my dissertation. I would like to express my special thank to Dr. Alvin J.M. Smucker for his providing the soil samples for this research and Dr. Alison Cupples to gave me valuable advices about microbiology . Second, I would like to thank to Yan Lang Pan to help me setting up all analytical instrument and Lori who always help me to order the chemicals. Thank for Lisveth for starting up the experiment. Thanks for Lulu, student in my research group, my colleagues Indu, Fernanda, Jean and many others. I could not have fun life at MSU without them. Thank for Han and Tuan from Dr. Phani lab for helping me to install computer program. I extend my thanks to Thai friends, Thorn, Mink and Michie, Ben, Toh, Oah and many others. I greatly value their friendship. I would to thank the owner of Thai restaurants, Thai princess and Taste of Thai who always give me the delicious food during my study here. I would like to thank for Royal Thai government for giving me the financial support throughout my study in the U.S. Without their funding, I could not have valuable experience of Ph.D. study at Michigan State University. iv I wish to extend my love and thank to my Fiancee’ Tik for her support, patience and love and her family here, Ningnong and Roger and Jim and Jeff who alway support me with their true loves. Finally, I would like to express my heart-felt gratitude for my family in Thailand, my Dad, my Mom, my Sister and her daughter, and my Brother. All their loves are the source of my strength. My Ph.D. degree would not be possible without them. v TABLE OF CONTENTS LIST OF TABLES xiii LIST OF FIGURES xvi SECTION I Fate and Transport of Aristolochic Acids as Responsible Agent of Balkan Endemic Nephropathy (BEN) and Exposure Pathway 1 Chapter 1 Introduction 1.1 The problem concern : Aristolochic acids and Health risks 1.2 Aristolochic acids and Balkan endamic nephropathy (BEN) 1.3 Environmental characteristic of BEN area study 1.4 Aristolochic acids structure and their physical-chemical properties 1.5 Hypothesis and objective REFERENCES 1 1 3 5 6 7 9 Chapter 2 Octanol water partitioning coefficient 2.1 Introduction : Octonol-water and its environmental significance 2.2 Literature reviews 2.2.1 Kow of neutral organic compounds 2.2.2 Kow of ionized organic compounds and their dependent 2.3 Materials and methods 2.3.1 Kow by shake flask method 2.3.1.1 Buffer aqueous phase 2.3.1.2 Preparation of pre-saturation of aqueous and octanol phase 2.3.1.3 Analytical method and validation 2.3.1.4 Procedure 2.3.1.5 pH and ionic strength effect 2.3.2 Kow by RP-HPLC method 2.4 Result and discussion 2.4.1 Ionization of Aristolochic acid 2.4.2 Calibration curve and validation 2.4.3 Log Kow by shake flask method 2.4.4 pH and ionic strength effect 2.4.5 Log Kow by HPLC method 2.5 Conclusion REFERENCES 13 13 15 15 16 22 23 23 24 24 24 26 26 28 28 28 30 31 34 36 37 vi Chapter 3 Aqueous solubility 3.1 Introduction 3.2 Literature reviews 3.2.1 Solubility of organic acids and pH effect 3.2.2 The ionic strength, ions presence and complexation formation effect 3.3 Materials and Methods 3.3.1 The kinetic test and calcium ion effect 3.3.2 The pH effect 3.3.3 The type of cation and ionic strength effect 3.4 Result and discussion 3.4.1 Kinetic study and calcium ion effect 3.4.2 The pH effect 3.4.3 Cation types and ionic strength effect 3.5 Conclusion REFERENCES 42 42 43 43 46 48 48 49 49 50 50 53 54 56 57 Chapter 4 Soil-water partitioning coefficient 4.1 Introduction 4.2 Literature reviews 4.2.1 The adsorption components in soil 4.2.1.1 Organic matter 4.2.1.2 Clays 4.2.1.3 Aluminum /Iron oxides 4.2.2 Adsorption mechanisms 4.2.2.1 Hydrophobic interaction 4.2.2.2 H-bonding 4.2.2.3 Ion exchange 4.2.2.4 Ligand exchange 4.2.2.5 Cation bridging 4.2.3 Adsorption Isotherm 4.2.4 Soil sorption of non-ionized and ionized organic compounds 4.2.4.1 Non-ionized organic compounds 4.2.4.2 Ionized organic compounds 4.2.4.3 Effect of pH 4.2.4.4 Effect of ionic strength 4.2.4.5 Effect of cations 4.2.4.6 Effect of anions 4.2.5 Desorption 4.2.6 Desorption isotherm 4.2.7 Desorption Hysteresis index 4.2.8 Factors effect to desorption hysteresis 4.2.8.1 Effect of organic matter content 4.2.8.2 Initial concentration 4.2.8.3 Contacting time 4.2.8.4 Effect of pH 59 59 60 60 60 61 61 62 63 63 63 64 64 64 66 66 66 68 68 69 69 70 70 72 73 73 74 74 74 vii 4.2.8.5 Effect of cations 4.2.8.6 Effect of anions 4.3 Material and method 4.3.1 Soil samples and their properties 4.3.2 Methods for determination of CEC and AEC in soils 4.3.3 Analytical method validation 4.3.3.1 Accuracy and precision 4.3.3.2 Linearity 4.3.3.3 Limit of detection(LOD) and limit of quantification (LOQ) 4.3.4 Adsorption experiment 4.3.4.1 Preliminary and kinetic study 4.3.4.2 Adsorption Isotherm 4.3.4.3 Effect of pH 4.3.4.4 Effect of calcium ion 4.3.4.5 Effect cations/anions 4.3.5 Desorption Experiment 4.3.5.1 Desorption kinetic 4.3.5.2 Desorption isotherm 4.3.5.3 Effect of pH 4.3.5.4 Effect of cations /anions 4.4 Result and discussion 4.4.1 AAs analytical method validation and matrix effect 4.4.2 Adsorption experiment 4.4.2.1 Adsorption kinetic 4.4.2.2 Adsorption Isotherm 4.4.2.3 Effect of soil properties 4.4.2.4 Effect of pH 4.4.2.5 Effect of calcium ion 4.4.2.6 Effect of Cations/ Anions 4.4.3 Desorption experiment 4.4.3.1 Desorption kinetic 4.4.3.2 Desorption isotherm 4.4.3.3 Effect of initial concentration 4.4.3.4 Effect of pH 4.4.3.5 Effect of cations/anions 4.5 Conclusion REFERENCES Chapter 5 Plant uptake 5.1 Introduction 5.2 Literature review 5.2.1 Root uptake mechanism and pathway 5.2.2 Translocation, accumulation and metabolism in plant 5.2.3 Plant uptake descriptor 5.2.4 Plant uptake of non-ionic compounds viii 75 76 76 76 77 78 79 79 79 80 80 82 82 82 83 83 83 84 84 85 85 85 88 88 90 94 95 97 99 102 102 104 107 109 113 118 120 127 127 128 128 130 132 134 5.2.5 Plant uptake of ionized compounds 5.3 Materials and methods 5.3.1 Calibration curve 5.3.2 Analytical method validation 5.3.2.1 Accuracy and precision 5.3.2.2 Linearity 5.3.2.3 Limit of detection(LOD) and limit of quantification (LOQ) 5.3.3 Nutrient solution preparation 5.3.4 Hydroponic culture experiment 5.3.4.1 Extraction method and recovery test 5.3.4.2 Degradation of Aristolochic acids in nutrient solution 5.3.5 Sand culture experiment 5.3.5.1 Sand preparation 5.3.5.2 The uptake experiment 5.3.5.3 Determination of AAs degradation in sand 5.4 Result and discussion 5.4.1 Analytical method validation 5.4.1.1 The calibration curve in plant matrix solution 5.4.1.2 Accuracy, precision and limit of detection 5.4.2 Recovery of plant extraction method 5.4.3 Hydroponic culture experiment 5.4.3.1 The stability of AAs in Knop’s nutrient solution 5.4.3.2 Kinetic uptake experiment 5.4.3.3 AAs accumulation in plant part 5.4.4 Sand culture experiment 5.4.4.1 The AAs available in treating sand 5.4.4.2 AAs accumulation in plant part 5.5 Conclusion REFERENCES Root exudates of Aristolochia plants and leaching and decomposition of Aristolochia Clematitis seeds 6.1 Introduction 6.2 Literature reviews 6.2.1 Plant root exudates 6.2.2 Root exudates collect method 6.2.3 Plant litter leaching and decomposition 6.2.4 Cellulose degradation in activated sludge medium 6.2.5 Cellulose degradation in anaerobic environment 6.3 Material and method 6.3.1 Root exudates from hydroponic culture 6.3.2 Root exudates from sands culture 6.3.3 Leaching from decomposition of Aristolochia Clematitis seeds 6.3.3.1 The decomposition in aerobic condition by activated sludge and soil 136 139 140 141 141 142 142 142 143 146 148 148 148 149 150 151 151 151 153 154 155 155 156 158 162 162 163 165 166 Chapter 6 ix 171 171 173 173 175 176 177 178 179 179 181 182 182 6.3.3.2 The decomposition in anaerobic condition by activated sludge and soil 6.4 Results and discussion 6.4.1 Root exudates of Aristolochia plant from hydroponic and sand culture 6.4.2 Leaching from the decomposition of Aristolochia Clematitis seeds 6.4.2.1 The decomposition in aerobic condition by activated sludge and soil 6.4.2.2 The decomposition in anaerobic condition by activated sludge and soil 6.5 Conclusion REFERENCES SECTION II Biogeochemical Modeling of Fate and Transport of Copper from Mining Waste in Torch Lake Sediment 184 185 185 187 187 190 191 192 196 Chapter 7 Introduction 7.1 The problem of concern and Torch Lake history 7.2 Torch Lake study site 7.3 The copper state in Torch Lake sediment 7.4 State of recovery 7.5 The redox state 7.6 The evidence of microbial copper sequestration 7.7 Torch Lake Sediment composition and mineralogy 7.8 Hypothesis and Approach REFERENCES 196 196 197 198 201 202 202 205 206 208 Chapter 8 Literature reviews 8.1 Copper speciation, control mechanism and transport in natural water 8.1.1 Complexation 8.1.2 Precipitation/dissolution 8.1.3 Adsorption 8.1.3.1 Sorption to metal hydrous oxides 8.1.3.2 Sorption to clays 8.1.3.3 Soption to organic matter 8.1.3.4 Sorption to microorganisms 8.2 Factors control copper mobility in sediment 8.2.1 pH 8.2.2 Redox state 8.3 Microbial mediated reaction 8.4 Releasing of heavy metal by reductive dissolution of iron oxide REFERENCES 211 211 211 214 216 217 217 218 219 219 220 220 222 224 226 x Chapter 9 Materials and methods 9.1 Copper speciation and saturation indices modeling 9.1.1 Surface water and top sediment porewater modeling 9.1.2 Mine tailing porewater modeling 9.2 Adsorption modeling 9.2.1 Adsorption to Hydrous Ferric Oxide (HFO) modeling 9.2.2 Adsorption to organic matter modeling 9.2.3 Adsorption /ion exchange on clay modeling 9.2.4 Adsorption to bacteria surface modeling 9.2.5 Contribution of all sorbent modeling 9.3 Oxidation of dissolve organic compounds by microorganism or TEAP modeling 9.4 Reductive dissolution of iron oxides modeling 9.5 Diffusion transport modeling 9.6 Diffusion transport with reductive dissolution of iron oxides and TEAP modeling 9.7 Full combined processes modeling APPENDICES Appendix A.1 Copper speciation in Torch lake surface water modeling Appendix A.2 Copper speciation in top sediment pore water modeling Appendix A.3 Chemical composition in mine tailing pore water modeling Appendix B.1 Adsorption to Hydrous Ferric Oxide (HFO) modeling Appendix B.2 Adsorption to organic matter modeling Appendix B.3 Adsorption / ion exchange on clay modeling Appendix B.4 Adsorption to Bacteria surface modeling Appendix B.5 Contribution of all sorbent modeling Appendix C.1 Oxidation of dissolved organic compounds by microorganism or TEAP modeling Appendix D.1 Reductive dissolution modeling Appendix E.1 Diffusion transport with reductive dissolution and TEAP modeling Appendix F.1 Full combined processes modeling REFERENCES 231 231 232 232 233 234 235 239 239 241 Chapter 10 Result and discussion 10.1 Copper speciation and saturation indices modeling 10.1.1 Surface water and top sediment porewater modeling 10.1.2 Mine tailing porewater modeling 10.2 Adsorption Modeling 10.2.1 Adsorption to Hydrous Ferric Oxide (HFO) modeling 10.2.2 Adsorption to organic matter modeling 10.2.3 Adsorption/ion exchange on clay modeling 10.2.4 Adsorption to bacteria surface modeling 10.2.5 Contribution from various sorbents 280 280 280 282 283 283 284 285 285 286 xi 241 244 245 246 248 250 251 252 253 254 256 258 259 261 263 265 268 273 278 10.3 Oxidation of dissolve organic compounds by microorganism or TEAP modeling 10.4 Reductive dissolution modeling 10.5 Diffusion transport with reductive dissolution and TEAP modeling 10.6 Full combined processes modeling 10.6.1 The dissolved redox species and dissolved copper concentration 10.6.2 Copper solid phase concentration REFERENCES Chapter 11 Conclusion 287 289 291 294 295 298 302 304 xii LIST OF TABLES Table 1.1 Aristolochia species, parts and their used (International Agency for Research on Cancer Staff., 2002) 2 Table 1.2 Chemical structure of Aristolochic acid I and II 6 Table 2.1 The molecular structure and literature values of pKa and log Kow of reference compounds 27 Table 2.2 The log Kow of AA I and II by shake flask method at three different concentrations 30 Table 2.3 The log Kow of reference compounds and their retention time from RP-HPLC system 35 Table 3.1 The pH of buffer solution and mixing ratio between H3PO4 0.01M and K2HPO4 0.01M 49 Table 4.1 Physical, chemical properties and composition of test soil samples 77 Table 4.2 % recovery (accuracy) and % RSD (intra-day and inter-day precision) of AAs analytical method 86 Table 4.3 Linear Ranges, correlation coefficient, quantification and detection limit of Aristolochic Acid I and II 87 Table 4.4 Linear and Freundlich isotherm parameters for sorption of AAI to four soils at natural soil pH 92 Table 4.5 Desorption coefficient (kdes) and 1/ndes obtained from fitting the model and hysteresis index (HI= ) for sorption of AA I and II in all tested soils 107 xiii Table 5.1 The literature values of RCF and TSCF of some non-ionized and ionized chemicals 139 Table 5.2 % recovery (accuracy) and % RSD (intra-day and inter-day precision) of analysis ALs in leave matrix 153 Table 6.1 The organic compounds identified in root exudates (modified from Yoshitomi, 2001) 173 Table 6.2 The AAs analysis in root exudates from three Aristolochia plants over 12 day periods 186 Table 8.1 Stability constant of complex formation of copper (Nriagu, 1979; Stiff, 1971) 213 Table 8.2 Solubility of copper minerals (Elder & Horne, 1978; Nriagu, 1979; Parkhurst & Appelo, 2013) 215 Table 8.3 The biodegradation reaction of sequential terminal electron acceptor 224 Table 8.4 Secondary redox reactions 224 Table 9.1 Modeling input of surface water and top sediment pore water 233 Table 9.2 Acidity constant of HFO and Equilibrium constants (Log K) of copper surface complexation on HFO 234 Table 9.3 The surface characterization parameters of HFO used in the model 235 Table 9.4 Proton dissociation constants and surface complexation constants of copper binding to humic acid in Tipping and Hurley’s database, WHAM (Windermere Humic Acid Model of Tipping and Hurley) modeling 236 Table 9.5 Copper ion exchange reaction and its constant 239 xiv Table 9.6 Deprotonation and stability constants of the copper and anionic functional groups on B. Subtilis cell surface 240 Table 9.7 The concentration of each functional groups per unit weight of bacteria B. Subtilis 241 Table 9.8 The biodegradation reaction of sequential terminal electron acceptor 242 Table 9.9 The kinetic rate law of sequential biodegradation reaction 243 Table 9.10 Kinetic Parameter constants of microbial mediation reaction using in the model 243 Table 9.11 Some physical transport properties of Torch Lake sediment in transport modeling 246 Table 9.12 Secondary redox reactions 247 Table 9.13 The kinetic rate law of secondary oxidation reaction 247 Table 9.14 Kinetic parameter constant of secondary oxidation reaction 248 Table 10.1 Oxidation state and speciation of copper in surface water and top sediment porewater 281 Table 10.2 Saturation index of copper minerals in surface water and top sediment porewater 282 Table 10.3 Some properties and copper composition from mine tailing porewater modeling 283 xv LIST OF FIGURES Figure 2.1 The extent of ionization as function of pH where AH and A- is neutral and ionized form of organic acids and B and BH+ is neutral and ionized form of organic bases(Kah & Brown, 2006) 18 Figure 2.2 The apparent 20 as a function of pH (modified from Kah & Brown, 2008) Figure 2.3 Log Kow of organic acids as a function of log ionic strength (Log ) 22 Figure 2.4 Octanol and aqueous phase separation in shake flask experiment(For the interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this dissertation) 25 Figure 2.5 Aqueous speciation of Aristolochic acid I and II as function of pH 28 Figure 2.6 The calibration curve of AA I and II in aqueous (a) and octanol (b) solution 29 Figure 2.7 Experimental and the fitted model curve versus pH after equilibrium: AA I (a) and AA II (b) 31 Figure 2.8 Experimental and fitted model at different ionic strength in aqueous phase: AA I (a) and AA II (b) 33 Figure 2.9 of anion of AA I and II at high pH as a function of logarithm of ionic strength (Log µ) 34 Figure 2.10 The linear correlation between log Kow and log RT of reference compounds and the extrapolation of AA I and II 35 Figure 3.1 Pentachlorophenol (PCP) solubility as a function of pH (modified from Wightman & Fein, 1999) 45 xvi Figure 3.2 Concentration of AA I (a) and AA II (b) dissolved in pure water and CaCl2 0.01M as a function of time 51 Figure 3.3 Concentration AA I and II dissolved in pure water (a) and CaCl2 solution (b) as a function of time 52 Figure 3.4 The experimental solubility of AA I (a) and AA II (b) as a function of pH at 7 day of equilibration. Solid line represents the fitted model curve of Eq.3.7 53 Figure 3.5 The solubility of AA I (a), AA II (b) and the pH (c) at the different type of cation and their concentration 55 Figure 4.1 The various types of sorption isotherm of Linear (a), Freundlich (b) and Langmuir (c) (modified from Schwarzenbach et al., 2003) 66 Figure 4.2 The two-compartment desorption isotherm and the splitting linear and exponential compartment model (modified from Barriuso, Baer, & Calvet, 1992) 72 Figure 4.3 AAs equilibrated with soil solution in mechanical shaker 81 Figure 4.4 Separation aqueous phase from soil after centrifuge 81 Figure 4.5 The calibration curve of AA I and II in soil matrix solution 86 Figure 4.6 Matrix effect of analysis AA I (a) and AA II (b) in soil matrix and pure methanol 88 Figure 4.7 Sorption kinetic of AA I (a) and II (b) on four soils 89 Figure 4.8 Sorption isotherm of AA I (a) and AA II (b) in all four soils at native soil pH equilibrium 91 xvii Figure 4.9 Adsorption isotherm of AA I in all four soils at adjusted pH equilibrium 93 Figure 4.10 Soil sorption coefficient (Kd) as a function of % organic carbon (a) and cation exchange capacity (b) of soil 94 Figure 4.11 Soil sorption coefficient as a function of pH for Wooster soils (a) and Hoytville soils (b) 96 Figure 4.12 The sorption coefficient (Kd) and pH (second axis) as a function log CaCl2 concentration on WT CT soil (a), WT NF soil (b), HY CT soil(c) and HY NF soil (d) 98 Figure 4.13 Adsorption isotherm of AA I (a) and II (b) on Hotyville NF soil in different cations background solution 99 Figure 4.14 Adsorption isotherm of AA I (a) and II (b) on Hotyville NF soil in different anion background solution 101 Figure 4.15 % Desorption of AAI (a) and AA II (b) as function of time for all tested soils 103 Figure 4.16 Sorption and desorption isotherm of AA I and II in WT CT soil (a and b), WT NF soil (c and d), HY CT soil (e and f) and HY NF soil (g and h) (filled symbols-adsorption point open symbols-desorption point 105 Figure 4.17 The hysteresis index of AA I (a) and II (b) as a function of initial concentration for all tested soils 108 Figure 4.18 AAs sorption –desorption isotherm of AA I and II on Hoytville NF soil at adjusted pH 6.0 (a) and (b), original pH 6.56 (c) and (d), adjusted pH8.36 (e) and (f). The pH shown in figure was from an average of pH of adsorption process 110 Figure 4.19 Hysteresis index calculated form sortion- desorption isotherm of all soil as a function with pH of AA I (a) and AA II (b) 112 xviii Figure 4.20 The sorption/desorption isotherm of AA I on Hoytville NF soil in different cation background solution (a) SrCl2 (b) CaCl2 (c) MgCl2 (d) KCl (e) NaCl (filled symbols-adsorption point, open symbols-desorption point) 114 Figure 4.21 The hysteresis index of desorption of AA I and II on HYNF soil at initial concentration 0.2-0.9 µg/ml with different cation background solution 115 Figure 4.22 The sorption/desorption isotherm of AA I on Hoytville NF soil in different anion background solution : KNO3(a), K2SO4 (b), KCl (c) (filled symbolsadsorption point, open symbols-desorption point) 117 Figure 4.23 The hysteresis index of desorption of AA I and II on HYNF soil at initial concentration 0.2-0.9 µg/ml with different anion background solution 118 Figure 5.1 The diffusion pathway of symplastic and apoplastic (Dettenmaier, 2008) 129 Figure 5.2 The relationship between root concentration factor (RCF) and the octanol – water partition coefficient (Log Kow) of root uptake of omethylcarbamoyloximes (open circle) and substituted phenylurea (cross mark) by barley plants from nutrient solution (Briggs et al., 1982) 134 Figure 5.3 The relationship between the translocation stream concentration factor and octanol-water partition coefficient of the root uptake of 0methylcarbamoyloximes (open circle) and substituted phenylurea (cross mark) by Barley plants from nutrient solution (Briggs et al., 1982) 135 Figure 5.4 The relationship between stem concentration factor (SCF) and and octanolwater partition coefficient (Log Kow) of o-methylcarbamoyloximes in the stem bases (close circle) and central stem(open circle) sections, and substituted phenylnreas in the stem bases (closed triangle) and central stem (open triangle) sections taken up by barley plants from nutrient solution (Briggs et al., 1983) 136 Figure 5.5 The root concentration factor of 2,4 dichloro-phenoxyacetic acids(open circle) , and 3,5 dichloro-phenoxyacetic acids (closed circle) uptake by barley as a function of pH of nutrient solution (Briggs et al., 1987) 137 xix Figure 5.6 The accumulation of weak acids between xylem and phloem by ion trapping effect (Hellstrom, 2004) 138 Figure 5.7 The derivatization of Aristolochic acids to Aristolactam by using zinc powder in acid solution (Chan et al., 2007) 141 Figure 5.8 Cucumber seedlings germinated on wetted towel in box 144 Figure 5.9 AAs uptake by hydroponic culture experimental setup 145 Figure 5.10 Plant parts separation after harvesting 145 Figure 5.11 Plant parts maceration by pestle and mortar 146 Figure 5.12 Plant extracted by sonication in ultrasonic bath 147 Figure 5.13 Supernatant separation from plant tissue after centrifuge 147 Figure 5.14 Plant uptake by sand culture experimental setup 149 Figure 5.15 HPLC-FLD chromatogram of derivative Aristolactum I and II in leaves (a), stem (b) and root (c) matrices solution spiked with AA I and II 151 Figure 5.16 The calibration curve of AL I (a) and AL II (b) in leaf, stem, root matrix and methanol 152 Figure 5.17 The % Recovery from spiked AA I (a) and II (b) standard solution to leaves and stem matrices 154 Figure 5.18 The stability of AAs in Knop’s solution as a function of time 156 Figure 5.19 The AA I and II remaining in solution as a function of exposure time 157 xx Figure 5.20 The AA I and II concentration in shoot as a function of exposure time 157 Figure 5.21 The percentage of AA I (a) and AA II (b) in different parts in cucumber plant parts and residual in media 158 Figure 5.22 AA I (a) and AA II (b) distribution in cucumber plant parts in hydroponic culture 160 Figure 5.23 AAs concentration found in stem and root tissues (a) and stem and leaf concentration factor (b) (SCF and RCF) of cucumber plant uptake AAs from spiked nutrient solution 161 Figure 5.24 AA I and II concentration available in sand as a function of incubation time 162 Figure 5.25 AA I (a) and AA II (b) distribution in cucumber plant parts in sand culture 163 Figure 5.26 AAs concentration found in stem and root tissues (a) and stem and leaf concentration factor (b) (SCF and RCF) of cucumber plants grown in Aristolochia Clematitis seed treating sand 164 Figure 6.1 The experimental setup to collect root exudates in hydroponic culture 180 Figure 6.2 The condensation of collected root exudates by evaporation unit 181 Figure 6.3 The experimental setup to collect root exudates in sand culture 182 Figure 6.4 The prepared bottles for decomposition of A.Clematitis seeds in aerobic condition 183 Figure 6.5 The set of bottles prepared in glove box under N2 atmosphere 184 Figure 6.6 AA I (a) and II (b) leaching from the decomposition of A. Clematis seeds in aerobic condition from bottle set A, B, C, D and E as a function of time : A=no microorganism + sterile, B=activated sludge + sterile, C=activated sludge + non sterile, D =soil + sterile, E=soil + non sterile 188 xxi Figure 6.7 The growing of microorganism from activated sludge in bottle A (a) , bottle B (b) and botle C (c) in aerobic decomposition experiments 189 Figure 6.8 The chromatogram of AA I and II leaching from A.Clematitis (bottle set A) analyzed by HPLC-DAD 189 Figure 6.9 AA I (a) and II (b) leaching from the decomposition of A. Clematis seeds in anaerobic condition from bottle set A, B ,C, D and E as a function of time : A=no microorganism + sterile, B=activated sludge + sterile, C=activated sludge + non sterile, D =soil + sterile, E=soil + non sterile 190 Figure 7.1 Map showing location of Torch Lake, Keweenaw Peninsula, Michigan (Fett, 2003) 198 Figure 7.2 Torch Lake sediment core (McDonald & Urban) 200 Figure 7.3 Bulk density (open circles) and solid phase copper concentration (solid circles) profiles of Torch lake sediment at different water depth (a)10 m (b) 13 m and (c) 20 m (modified from C. P. McDonald et al., 2010) 200 Figure 7.4 The location of Gratiot Lake, Torch Lake and Portage Lake (Fett, 2003) 201 Figure 7.5 Normalized concentration of Fe, Mn and Cu in Torch Lake sediment (Fett, 2003) 202 Figure 7.6 Scanning electron microscopy of Ralstonia isolated from Torch lake grown on copper free agar (A) and (C); grown on copper-supplemented agar (B) and (D) (modified from Konstantinidis et al., 2003) 203 Figure 7.7 TEM and SEM of R. pickettii strain 12 J grown in absence copper (TEM: A, C; SEM: E) and presence copper (TEM: B, D; SEM: F) (modified from Yang et al., 2010) 204 Figure 7.8 Model of copper upward diffusion from mine tailing sediment and sequestration by organic matter and bacteria at the surface sediment 206 xxii Figure 8.1 Calculated copper speciation in fresh water when absence (a) and presence (b) of organic chelation NTA (modified from Elder & Horne, 1978) 214 Figure 8.2 The stability diagram for ternary system Cu(II)-H2O-CO2(g) (Nriagu, 1979) 215 Figure 8.3 Sorption of copper by clay minerals as a function of pH (modified from Nriagu, 1979) 218 Figure 8.4 The stability relation between copper compound in system of Cu+H2O+O2+ S +CO2 (PCO2 = 10-3.5 atm) , total dissolved sulfur species 10-1 M at 25oc and 1 atm (Nriagu, 1979) 221 Figure 8.5 The equilibrium aqueous concentration of copper species as a function of pE (Nriagu, 1979) 222 Figure 8.6 Biodegradation with sequential terminal electron acceptors 223 Figure 9.1 The 1-D column for diffusion process modeling between surface water and mine tailing porewater 246 Figure 9.2 The diagram of all processes combined as full model 249 Figure 10.1 The simulation result of adsorption of Cu+ (a) and Cu2+(b) on strong and weak surface sites of hydrous ferric oxide as a function of pH 284 Figure 10.2 The simulation result of adsorption of Cu2+ on organic matter as a function of pH 284 Figure 10.3 The simulation result of adsorption/ion exchange of Cu2+ on clay 285 Figure 10.4 The simulation result of adsorption of Cu2+ onto bacteria surface 286 Figure 10.5 The simulation result of the distribution of copper adsorption on hydrous ferric oxide (HFO), organic matter, clay, bacteria surface and total sorption of all sites at different dissolved copper concentration 287 xxiii Figure 10.6 Concentration of redox species as a function of time result from the sequential biodegradation or TEAP modeling 288 Figure 10.7 Concentration of some redox species, dissolved copper, Tenorite, pE and pH as a function of time in reductive dissolution modeling 290 Figure 10.8 The simulation results of pE, pH, some redox species, dissolved copper and Tenorite concentration profile from diffusion transport with reductive dissolution and TEAP modeling 293 Figure 10.9 The simulation results of dissolved redox species concentration as a function of depth and time in the full combined processes modeling 296 Figure 10.10 The simulation results of copper solid phases concentration as a function of depth and time in the full combined processes modeling 300 xxiv SECTION I Fate and Transport of Aristolochic Acids as Responsible Agent of Balkan Endemic Nephropathy (BEN) and Exposure Pathway Chapter 1 Introduction 1.1 The problem concern : Aristolochic acids and health risks Aristolochic acids (AAs) are chemicals that naturally found in plants of Aristolochia and Asarum species. Many types of aristolochic acids (AAs) and aristololactums (ALs) are found, but among of them, Aristolochic acid I and II are major components (Chan, Lee, Liu, & Cai, 2007; Zhang et al., 2006). Aristololactam I and II (AL I and II) are also found as metabolites of Aristolochic I and II in animals (Krumbiegel, Hallensleben, Mennicke, Rittmann, & Roth, 1987) In the past, AAs were widely used as traditional herbal medicine as well as botanicalcontaining dietary supplement. Table 1.1 shows some parts of Aristolochia plants which made of the medicines and their used. 1 Table 1.1 Aristolochia species, parts and their used (International Agency for Research on Cancer Staff., 2002). Aristolochia species Part used Used Aristololchia fangchi Root Oedema, antipyretic and analgesis remedies Aristololchia manshuriensis Stem Anti-flammatory and diuretic for acute infection of the urinary system and as emmenagogue and galactagogue for amenorrhoea Aristololchia contorta Fruit, Herb Haemorrhoids, cough and asthma, epigastric pain, arthralgia and oedema Aristololchia debelis Fruit, Herb and Root Same as A. contorta and dizziness, headache, abdominal pain, and snake and insect bites Aristolochic acids containing medicines were used until they were identified as potential carcinogen in rodent in 1980’s. The rats treated with Aristolochic acids developed a tumor in forestomach, lung, lymphoid organs and uterus as well as benign and malignant tumor in kidney and urinary tract (Mengs, 1988). High dose of Aristolochid acids treated rats were died from acute renal failure (Mengs, 1987). The evidence which support Aristolochic acids as nephrotoxic and carcinogenic in human had been reported in early’s 1990. At least 100 cases of rapidly developed renal disease in young women who took the drugs which were made from Aristolochia species during a slimming regimen were reported in Belgium (Vanherweghem et al., 1993). In 1999, the first 2 cases of specific nephropathy related to ingestion of Chinese herbal medicines had been reported in U.K. 2 However, by different sources from the cases in Belgium, the patients had taken the eczama herbal preparations. The analysis of these preparations by High Performance Liquid Chromatrography and Mass Spectrometry (HPLC-MS) found Aristolochic acid I and II in them (Lord, Tagore, Cook, Gower, & Pusey, 1999). After the original report in Belgium, similar cases were reported worldwide such as United Kingdom, France, Spain, Germany, United States, China, Japan and Taiwan in context of using AAs containing medicines (Debelle, Vanherweghem, & Nortier, 2008). Because of founding renal failure disease which outbreak in patients taking pills containing Chinese herb, this renal disorder was termed as Chinese herb nephropathy (CHN). The AAs were confirmed as a cause of CHN when the doctor detected AA-DNA adducts in the kidney tissue of CHN patients which showing evidence of exposure to AAs (Arlt, Pfohl-Leszkowicz, Cosyns, & Schmeiser, 2001). Moreover, rats which were treated with sliming pill same as CHN patients had developed a renal interstitial fibrosis and showed similar pattern of AA-DNA adduct found in CHN patients. The specific AA-DNA adducts were also found in the kidney tissue samples from urothelial carcinoma patients (Nortier et al., 2000). These evidence supported hypothesis that AAs play a major role in CHN patients. Consequently, the term Chinese herb nephropathy (CHN) had been replaced by Aristolochic acid nephropathy (AAN). 1.2 Aristolochic acids and Balkan endamic nephropathy (BEN) Balkan endemic nephropathy or BEN is a chronic fatal kidney disease that had first reported around 60 years ago. It was believed as environmentally induced disease which affecting to people living in the rural area of Balkan countries included Bosnia-Herzegovina, Bulgaria, Croatia, Romania and Serbia(Hranjec et al., 2005). Even extensive of researches for etiology of BEN, the causes are still unclear. Over than 20 years, many hypothesis had been set 3 for the BEN causes such as trace element deficiency, mycotoxin or ochratoxin A which is also nephrotoxic chemical, plant toxin (aristolochic acids) and polycyclic aromatic hydrocarbon (PAHs) from coal deposit which may be leached and contaminated to groundwater (Castegnaro et al., 2006; USGS, 2001; Voice et al., 2006). Among these hypotheses, Aristolochic acids are most possible. The pathological and clinical features of BEN are very similar to those associated with AAN except for AAN is a rapid decline of renal function in 6 month to 2 year where BEN is slower progression and longer accumulation process from 10 -20 years to end-stage renal failure or urothelial cancer (Broe, 2014). Recently, Grollman et al. had identified the AA-derived DNA adduct dA-AL (Aristolactum) and dG-AL in the renal tissue of patients in BEN area, but not found in patients with other chronic renal diseases living in non-BEN area which was the first time proving AAs exposure by BEN patients (Grollman et al., 2007). Moreover, after sequencing the DNA from tumour tissue patients who resided in endemic area for at least 15 year, they found significant mutation pattern which clearly showed the link between urothelial tumour, p53 mutation and AAs exposure (Arlt et al., 2007). So, the most recent researches were more intent to AAs. The long term consumption of dietary contaminated with AAs was hypothesized as cause of BEN (Grollman & Jelakovic, 2007) . Because the Aristolochic Clematits plants were found to extensively grow as a weed in wheat field in BEN areas, scientists believed that Aristochia clematitis seeds were mixed with wheat grain during the harvest and the contaminated flour was used to prepare for home-made bread. However, the life cycle of the weed and the grain were too different where Aristolochia plant fruits matured in fall but wheat was harvested in the mid summer and the Aristolochia plant fruits have a large pulp form which easily remove from the 4 wheat grain, so the contamination was hard to occur and this hypothesis still has some limitation (Pitt, 2011; Voice et al., 2006). The long term consumption of dietary contaminated with AAs hypothesis had been investigated from the BEN patients and healthy residents from endemic villages (Hranjec et al., 2005). The majority of subjects reported the extensive growing Aristolochia Clematitis in the wheat field more than 30 years ago and the BEN patients had expose significantly more frequently than the controls. These results can imply that the patients from endemic village consume the bread contaminated with A.Clematitis seed. However, the changing of lifestyle and farm handling such as applying the herbicide to field had reduced the dietary exposure to aristolochic acids and expect as a cause of decreasing of incidence rate of BEN. 1.3 Environmental characteristic of BEN area study Due to the relation with environmental agents of BEN, the knowing about environmental characteristics between BEN and non-BEN areas is necessary. The environmental differences between endemic and non-endemic village had been studied. These studies showed that there were many sources of nitrate on the soil and high level of nitrate had been found in groundwater. However, there was no statistically significant for concentration of different nitrogen species between BEN and non-BEN village. The water samples that supplied to both village were fail to show high level of organic compounds in BEN areas (Voice et al., 2006). The soil geochemistry between BEN and non-BEN areas had been tested. The results showed that concentration of metals and other elements were usual. However, there were some notes that some elements e.g. Li, Ca, Mg in soil and well water from both villages were higher than world average concentration. K, Ca, Pb, Cd, Mo, Hg and As in soil were greater in endemic villages whereas 5 Mg, Ba, Mn, Fe, Al, Co and Se were lower in endemic village (Long et al., 2001). The different of soil chemistry between BEN and non-BEN village may response for the occurrence of BEN in Balkan areas. 1.4 Aristolochic acids structure and their physical-chemical properties Aristolochic acids molecules contain the mixing of hydrophobic and polar structure including nitrophenanthrene, dioxolo and carboxylic acid as shown in Table 1.2 (Pfau, Schmeiser, & Wiessler, 1990). AAs are yellow solid, slightly dissolve in water but completely dissolve in alcohols, acetic acid, acetone, chloroform and diethyl ether (International Agency for Research on Cancer Staff., 2002). AA I and II are different by methoxy group in AA I which cause it less polar than AA II. Table 1.2 Chemical structure of Aristolochic acid I an II Compound Structure Aristolochic acid I Formula C17H11NO7 (8-methoxy-6-nitro-phenanthro (3,4-d)- 1,3-dioxolo-5-carboxylic acid) Aristolochic acid II C16H9NO6 (6-nitro-phenanthro (3,4-d)1,3-dioxolo-5-carboxylic acid) 6 Aristolochic acids are ionizable organic compounds which can dissociate when dissolved in water. They have two different forms, neutral or ionized forms where the neutral molecules can deprotonated by losing one proton (H+) from carboxylic (COOH) group and become as negative charged ions. The extent of ionization will depend on their pKa and pH of solution. Fu et al., 2011 presented the dissociation constant (pKa) of AA I and II at 3.3±0.1 and 3.2±0.2, respectively. No aqueous solubility had been reported. 1.5 Hypotheses and objectives Because the linkage between consumption of AA-contaminated bread and BEN is not yet confirmed, other exposure pathways should be considered. In this research, a new exposure pathway was investigated. I hypothesize that AAs can be released by Aristolochia species plants via root exudates or leaching from the decomposition of dead plant tissues. Once AAs are released to soil, they can move efficiently with water and may be taken up by crop plants, which may be consumed by BEN patients. To study fate and transport of chemicals in the soil, there are many essential parameters to be determined. The octanol-water partition coefficient shows the hydrophobicity of chemicals and is also related to other environmental partitioning processes. Solubility is a parameter that indicates the ability of chemicals to dissolve in water. Soil sorption and desorption are the main processes in controlling the mobility in soil and availability in other compartments. The root exudates and leaching from decomposition of plant tissues determine the sources of chemicals that are transferred from plants to soils. The uptake of chemicals by crop plants shows the exposure potential to humans and food chain contamination. 7 In this study, I examined the fate and transport properties of AAs including the octanolwater partition coefficient, solubility, and the soil-water partition coefficient. I also determined AAs’ release from Aristolochia species plants via root exudation and leaching from their seeds decomposition. The plants’ ability to uptake AAs was also determined. 8 REFERENCES 9 REFERENCES Arlt, V. M., Stiborova, M., vom Brocke, J., Simoes, M. L., Lord, G. M., Nortier, J. L., . . . Schmeiser, H. H. (2007). Aristolochic acid mutagenesis: molecular clues to the aetiology of Balkan endemic nephropathy-associated urothelial cancer. Carcinogenesis, 28(11), 2253-2261. doi: 10.1093/carcin/bgm082 Broe, M. D. E. (2014). Balkan endemic nephropathy. from http://www.uptodate.com/contents/balkan-endemic-nephropathy Castegnaro, M., Canadas, D., Vrabcheva, T., Petkova-Bocharova, T., Chernozemsky, I. N., & Pfohl-Leszkowicz, A. (2006). Balkan endemic nephropathy: Role of ochratoxins A through biomarkers. Molecular Nutrition & Food Research, 50(6), 519-529. doi: 10.1002/mnfr.200500182 Chan, W., Lee, K.-C., Liu, N., & Cai, Z. (2007). A sensitivity enhanced high-performance liquid chromatography fluorescence method for the detection of nephrotoxic and carcinogenic aristolochic acid in herbal medicines. Journal of Chromatography A, 1164(1-2). doi: 10.1016/j.chroma.2007.06.055 Debelle, F. D., Vanherweghem, J. L., & Nortier, J. L. (2008). Aristolochic acid nephropathy: A worldwide problem. Kidney International, 74(2), 158-169. doi: 10.1038/ki.2008.129 DuarteDavidson, R., & Jones, K. C. (1996). Screening the environmental fate of organic contaminants in sewage sludge applied to agricultural soils .2. The potential for transfers to plants and grazing animals. Science of the Total Environment, 185(1-3), 59-70. doi: 10.1016/0048-9697(96)05042-5 Dudka, S., & Miller, W. P. (1999). Accumulation of potentially toxic elements in plants and their transfer to human food chain. Journal of Environmental Science and Health Part BPesticides Food Contaminants and Agricultural Wastes, 34(4), 681-708. doi: 10.1080/03601239909373221 Fu, X. F., Liu, Y., Li, W., Bai, Y., Liao, Y. P., & Liu, H. W. (2011). Determination of dissociation constants of aristolochic acid I and II by capillary electrophoresis with carboxymethyl chitosan-coated capillary. Talanta, 85(1), 813-815. doi: 10.1016/j.talanta.2011.03.088 Gao, Y. Z., & Zhu, L. Z. (2004). Plant uptake, accumulation and translocation of phenanthrene and pyrene in soils. Chemosphere, 55(9), 1169-1178. doi: 10.1016/j.chemosphere.2004.01.037 Grollman, A. P., & Jelakovic, B. (2007). Role of environmental toxins in endemic (Balkan) nephropathy. Journal of the American Society of Nephrology, 18(11), 2817-2823. doi: 10.1681/asn.2007050537 10 Grollman, A. P., Shibutani, S., Moriya, M., Miller, F., Wu, L., Moll, U., . . . Jelakovic, B. (2007). Aristolochic acid and the etiology of endemic (Balkan) nephropathy. Proceedings of the National Academy of Sciences of the United States of America, 104(29), 12129-12134. doi: 10.1073/pnas.0701248104 Hranjec, T., Kovac, A., Kos, J., Mao, W. Y., Chen, J. J., Grollman, A. P., & Jelakovic, B. (2005). Endemic nephropathy: the case for chronic poisoning by Aristolochia. Croatian Medical Journal, 46(1), 116-125. International Agency for Research on Cancer Staff. (2002). Some Traditional Herbal Medicines, Some Mycotoxins, Naphthalene and Styrene Iarc Monographs (pp. 590 p.). Retrieved from http://ezproxy.msu.edu:2047/login?url=http://site.ebrary.com/lib/michstate/Top?id=1025 2491 Krumbiegel, G., Hallensleben, J., Mennicke, W. H., Rittmann, N., & Roth, H. J. (1987). STUDIES ON THE METABOLISM OF ARISTOLOCHIC ACID-I AND ACID-II. Xenobiotica, 17(8), 981-991. Long, D. T., Icopini, G., Ganev, V., Petropoulos, E., Havezov, I., Voice, T., . . . Stein, A. (2001). Geochemistry of Bulgarian soils in villages affected and not affected by Balkan Endemic Nephropathy: A pilot study. International Journal of Occupational Medicine and Environmental Health, 14(2), 193-196. Lord, G. M., Tagore, R., Cook, T., Gower, P., & Pusey, C. D. (1999). Nephropathy caused by Chinese herbs in the Uh. Lancet, 354(9177), 481-482. doi: 10.1016/s01406736(99)03380-2 Mengs, U. (1987). ACUTE TOXICITY OF ARISTOLOCHIC ACID IN RODENTS. Archives of Toxicology, 59(5), 328-331. doi: 10.1007/bf00295084 Mengs, U. (1988). TUMOR-INDUCTION IN MICE FOLLOWING EXPOSURE TO ARISTOLOCHIC ACID. Archives of Toxicology, 61(6), 504-505. doi: 10.1007/bf00293699 Ney, R. E. (1995). Fate and transport of organic chemicals in the environment : a practical guide (2nd ed.). Rockville, Md.: Government Institutes. Nortier, J. L., Martinez, M. M., Schmeiser, H. H., Arlt, V. M., Bieler, C. A., Petein, M., . . . Vanherweghem, J. L. (2000). Urothelial carcinoma associated with the use of a Chinese herb (Aristolochia fangchi). New England Journal of Medicine, 342(23), 1686-1692. doi: 10.1056/nejm200006083422301 Pavlovic', N. M., Maksimovic', V., Maksimovic , J. D., Orem, W. H., Tatu, C. A., Lerch, H. E., . . . Paunescu, V. (2012). Possible health impacts of naturally occurring uptake of 11 aristolochic acids by maize and cucumber roots: links to the etiology of endemic (Balkan) nephropathy. Environmetal Geochemical Health. Pfau, W., Schmeiser, H. H., & Wiessler, M. (1990). ARISTOLOCHIC ACID BINDS COVALENTLY TO THE EXOCYCLIC AMINO GROUP OF PURINE NUCLEOTIDES IN DNA. Carcinogenesis, 11(2), 313-319. doi: 10.1093/carcin/11.2.313 Pitt, J. I. (2011). Chinese herbal medicines, aristolochic acid and Balkan endemic nephropathy. Occupational and Environmental Medicine, 68(4), 237-237. doi: 10.1136/oem.2010.061663 USGS. (2001). Health effect of toxic organic compounds from coal-The case of Balkan Endemic Nephropathy. Vanherweghem, J. L., Depierreux, M., Tielemans, C., Abramowicz, D., Dratwa, M., Jadoul, M., . . . Vanhaelen, M. (1993). RAPIDLY PROGRESSIVE INTERSTITIAL RENAL FIBROSIS IN YOUNG-WOMEN - ASSOCIATION WITH SLIMMING REGIMEN INCLUDING CHINESE HERBS. Lancet, 341(8842), 387-391. Voice, T. C., Long, D. T., Radovanovic, Z., Atkins, J. L., McElmurry, S. P., Niagolova, N. D., . . . Ganev, V. S. (2006). Critical evaluation of environmental exposure agents suspected in the etiology of Balkan endemic nephropathy. International Journal of Occupational and Environmental Health, 12(4), 369-376. Zhang, C. Y., Wang, X., Shang, M. Y., Yu, J., Xu, Y. Q., Li, Z. G., . . . Namba, T. (2006). Simultaneous determination of five aristolochic acids and two aristololactams in Aristolochia plants by high-performance liquid chromatography. Biomedical Chromatography, 20(4), 309-318. doi: 10.1002/bmc.565 12 Chapter 2 Octanol water partitioning coefficient 2.1 Introduction: Octonol-water and its environmental significance The octanol water partition coefficient (Kow or P) is indicator of hydrophobicity of chemicals. It shows how chemicals favor to partition in water phase or organic liquid phase. Kow widely used in term of prediction of chemical fate and transport in environment compartment because it describes the partitioning between the water and environmental organic medias such as soil, sediment and organic matter (Ney, 1995). Octanol water partition coefficient also relate to other partition coefficients such as soil-water, lipid-water partition coefficient or bioconcentration factor (BCF) which called Linear Free Energy Relationship (LFER). The LFER principle was widely used in environmental field. One is Quantitative Structural Activity Relationships or QSARs which is a regression model relating the physical-chemical properties of chemical to their molecule structure(Yuying, Guanghui, Ying, Zhuang, & Cheng, 2009). The example of application of QSAR is using Kow to predict soil sorption of chemical (Karickhoff, Brown, & Scott, 1979; R. P. Schwarzenbach & Westall, 1981). The high log Kow means the greater potential for chemicals to be sorped in soil, low mobility, low potential to degrade by physical, chemical or biological which lead to persistence in the environment(Ney, 1995). Kow is also used to predict the bioaccumulation potential of terrestrial and aquatic biota which taken the chemicals from contaminated soil or sediment, these can represent in term of bioconcentration factor (BCF). The high log Kow chemicals have high potential for food chain biomagnifications too (Brandt, Becker, & Porta, 2002; Fisk, Norstrom, Cymbalisty, & Muir, 1998). It related to the 13 toxicity parameters e.g. LD50 where higher Kow means higher toxicity (lower LD50) (Veith, Call, & Brooke, 1983). The Kow also is related to ability of chemicals to be uptake by plant root where lipophilic organic chemicals possess a greater tendency to partition into plant root lipids than hydrophilic chemicals and to distribution in various plant parts (Briggs, Bromilow, Evans, & Williams, 1983; E. M. Dettenmaier, Doucette, & Bugbee, 2009). Due to AAs are toxic chemicals that causing development of nephropathy and urothelial cancer in human, the physical - chemical properties to predict their fate and transport e.g. octanol-water partition coefficient is very important. However, there are very few literatures reported their log Kow. Most of them are from calculation from various programs that calculated by fragment method. However, for complex structure like AAs, it may have large error due to lack of fragmental value and inaccuracy of database (Han, Qiao, Zhang, Lian, & Ge, 2012). Because of the difficulty in measurement AAs at low concentration and high cost of high quality AAs standard chemicals, the experimental literature values were rarely reported. Moreover, because AAs can ionize when dissolve in water into two different forms, the log Kow of neutral molecules may be different from ionized form for many order of magnitude that make measurement Kow more difficult in term of different method using for each form (R.P. Schwarzenbach, Gschwend, & Imboden, 2003). The first experimental log Kow of AA I and II were reported by Han et al., 2012 using RP-HPLC method which tried to relate the log Kow of AAs to log Kow of some neutral and ionized compounds. They reported log Kow of AA I and II in neutral form at 4.45 and 3.99 respectively. These numbers indicated that AAs in neutral form are highly hydrophobic(Han et al., 2012). Some programs provided the calculated log Kow online, for example, log Kow of AA I and II from ACD/Labs are 3.41 and 3.50, KOWWIN are 4.19 and 4.11 and ALOGPs are 2.69 14 and 2.68 (VCCLAB, 2011) . The different between these Kow number means the calculation methods cannot predict them accurately because of highly complex molecule structure and lack of fragmental values. In this study, the octanol water partition coefficient of Aristolochic acid I and II will be determined by conventional shake flask method and indirectly RP-HPLC method. The shake flask method measured by direct contacting between octanol and water which contained the chemicals. The concentration ratio in both phases will be determine at equilibrium and calculated as Kow, whereas the RP-HPLC method will correlate between retention behavior on C-18 column and the hydrophobicity of reference compounds. This study try to accurately determine the Kow of AAs and their dependence on environment factors which is necessary for prediction the fate and transport when AAs were released to environment. 2.2 Literature reviews 2.2.1 Kow of neutral organic compounds Kow is a parameter that describes chemical’s hydrophobic properties. Octanol is used to represent the lipid-water partition because their polarity is similar to biotic lipid. The Kow is defined by the concentration ratio of compounds distributed in octanol phase to the aqueous phase, when two phases are equilibrium in each other as shown in Eq.2.1. This coefficient indicates the phase that chemicals are favored to partition in. (2.1) The concentration in this equation is not solubility of solutes in octanol or water, but it is the concentration in water-saturated octanol and octanol-saturated water. Since Kow of chemicals 15 have varied in many orders of magnitude, so they always present as base 10 logarithm (Log Kow or Log P). The compounds with high log Kow (more than 3) are determined as hydrophobic whereas the chemical with small log Kow (1 or less) will be determined as hydrophilic. Generally, most hydrophobic compounds are nonpolar and nonionic (Word, 2002). The Kow of nonpolar compounds relates to degree of sorption to organic carbon of natural sorbents which dominated by hydrophobic interaction. Many studies showed the organic carbon-water partition coefficient (Koc) can be related to Kow by Eq. 2.2 where a and b are constant for given compounds (Karickhoff et al., 1979; B. E. Nowosielski & Fein, 1998) (2.2) However, this relation may be not fitted for the polar and especially ionizable organic compounds, because variety interactions e.g. ionic exchange reaction rather than hydrophobic interaction had involved to the sorption and the other parameters such as pH and ionic strength become primary order importance(Jafvert, Westall, Grieder, & Schwarzenbach, 1990). 2.2.2 Kow of ionized organic compounds and their dependent The ionized compounds such as weakly acids will be partially ionized in water which give anion form (A-) and hydronium ion (HA) as express in Eq. 2.3. The extent of ionization depend on dissociation constant (Ka) and pH. (2.3) From dissociation equation, we can rearrange Eq. 2.3 in term of the ratio of concentration in neutral and anionic form as a function of pKa and pH as expressed in Eq. 2.4 – 2.7 (Kah & Brown, 2006). 16 (2.4) (2.5) (2.6) (2.7) Besides, the proportion species presented in solution can be shown by extent of ionization (αia) which define as the fraction of neutral species that present in the solution (R.P. Schwarzenbach et al., 2003) (2.8) (2.9) This equation can be plotted between fraction of neutral and ionized form as a fuction of pH as shown in Figure 2.1. This figure shows that, at pH<>pKa, the ionized form becomes dominant instead, and at pH equal to pKa , the fraction of neutral and ionized form will be equal. 17 Figure 2.1 The extent of ionization as function of pH where AH and A- is neutral and ionized form of organic acids and B and BH+ is neutral and ionized form of organic bases(Kah & Brown, 2006) Because neutral and ionic species perform different polarities, their hydrophobicity are also different. The ionic species is more soluble in water and would be expected lower Kow. From octanol water partition coefficient definition, we can expand the definition to include the ionic and neutral species in each phase. This partition coefficient is known as “apparent coefficient” ( ) which accounts for all contribution species in solution(Westall, Leuenberger, & Schwarzenbach, 1985). (2.10) Where [HAoc] and [HAw] are concentration of neutral species which equilibrium in octanol and aqueous phase respectively and [Ao-] and [Aw-] are concentration of ionic species which equilibrium in octanol and aqueous phase. The is rearranged to express in term of pKa and pH for more practical using as shown in Eq. 2.11 (Word, 2002) 18 (2.11) (2.12) Introducing, the partition coefficient of neutral species, coefficient of ionized species, ,and from Eq. 2.7 , ,and partition ,the will become as (2.13) This equation is very useful to predict the apparent Kow of chemicals at given pKa and pH. It also can predict the octanol-water coefficient of neutral form and ionized form by fitting the experimental data with this equation. From Eq.2.13 the relationship between log and the pH can be represented as sigmoidal curve as shown in Figure 2.2 (solid line). However, some chemicals are very low dissociated in non-aqueous phase, so we can neglect the ionized species in octanol phase and the can be express as Eq. 2.14 (Bogdan E. Nowosielski, 1998). This shows as dash line in Figure 2.2 where log will be indefinitely decrease when pH increase (Kah & Brown, 2008). (2.14) From this figure, the Kow of each form of ionized chemicals may be completely different in many orders of magnitude. 19 Figure 2.2 The apparent as a function of pH (modified from Kah & Brown, 2008) The type of sigmoidal curve (solid line) had been observed in many ionizable pharmaceutical compounds (Berthod, Carda-Broch, & Garcia-Alvarez-Coque, 1999) , trichlorophenol, pentachlorophenol and their derivative (Jafvert et al., 1990; B. E. Nowosielski & Fein, 1998) whereas the continually decreasing of Kow with pH (dash line) was found in dichlorprop (Riise & Salbu, 1992) and pentachlorophenol (Kaiser & Valdmanis, 1982). The difference of log Kow between neutral and ionic form (log Kow ) of some chemicals has been reported, for example, 2.9 for chlorophenol (B. E. Nowosielski & Fein, 1998) and 2.2 for Dichlorprop (Riise & Salbu, 1992), which mean the neutral species are 100-1,000 times more hydrophobic than the ionized species. To measure Kow of neutral species, the pH at pKa - 2 was recommended. However, to measure the Kow- of ionic species, the measured pH will depend on the difference of hydrophobicity of neutral and ionic form, which closed to pKa + log is very small , + 2. If will become large and the pH that determine log Kow- will never reach (Kah & Brown, 2008). 20 The apparent octanol water partition coefficient is also effect by ionic strength and type of counter ions that present in aqueous phase. The ionized organic compounds may form the complex with these counter ions to neutralize their charged and partition in octanol or water as ion-pair which make the partition coefficient more complicated. The apparent calculation that includes ion pairs can be expressed in Eq.2.15 (B. E. Nowosielski & Fein, 1998) (2.15) Where is ion-pair between monovalent inorganic cation and ionized organic species in octanol phase and aqueous phase, respectively. Many studies showed that Kow of some organic acids was a function of ionic strength. They found that, at pH large higher than pKa where the anion form dominated, the Kow will linearly increase with increasing ionic strength on the log scale as shown in Figure 2.3 (Jafvert et al., 1990; B. E. Nowosielski & Fein, 1998). Moreover, Kow for aqueous solution that contained divalent cations e.g. Ca2+ and Mg2+ were more than Kow of solution contained monovalent cations e.g. Li+, Na+, K+. This indicated the organic acid anions can form the ion pair with cations in aqueous solution. More concentration of cations will give more chance to form the ion pairs and cross over from aqueous phase to octanol phase. This process will raise the Kow (Jafvert et al., 1990; Lee, Rao, Nkedikizza, & Delfino, 1990; B. E. Nowosielski & Fein, 1998; Word, 2002). 21 Log Kow Log  Figure 2.3 Log Kow of organic acids as a function of log ionic strength (Log ) (modified from B. E. Nowosielski & Fein, 1998) 2.3 Materials and methods In this study, Kow of Aristolochic acid I & II were measured by direct shake flask method, and indirect RP-HPLC method. Shake flask method is traditional and widely used to determine Kow because it has good repeatability and reliable for the compounds with log Kow between 0 - 4. Although the shake flask method give a reliable result but there are some deficiency especially when dealing with highly hydrophobic compounds. These chemicals will strongly partition into octanol phase and consequently have very low concentration in aqueous phase, so large amount of aqueous phase sample are required for analysis. Moreover, the agitation in shake flask method will produce large amount of droplet of octanol in the aqueous. These droplets may remain after phase separation which induce the overestimation of concentration in aqueous phase, so lower the actual Kow (Brooke, Dobbs, & Williams, 1986). This method is not suitable for highly hydrophobic chemical (log Kow > 4-5) (Debruijn, Busser, Seinen, & Hermens, 1989). 22 The RP-HPLC method is based on the principle the partitioning of chemical on stationary phase in column and mobile phase in HPLC column which is proportion to their octanol-water patition coefficient. The interaction of solute to hydrophobic stationary phase and hydrophilic mobile phase is similar to the interaction to octanol and aqueous phase. The low Kow chemical will come first whereas the high Kow will come last. This method gives good reproducibility but the accuracy will depend on the correctness of their reported Kow of reference compounds. The set of these compounds have to select from the chemicals which have similar structure to the test chemical (Poole & Poole, 2003). Their Kow also should cover the range of Kow of the test chemical to minimize the error from extrapolation. Although, this method is convenient and cheap, it has limited for determining the Kow of some complex chemicals because of lack of the reliable literature log Kow of reference compounds. 2.3.1 Kow by shake flask method In general, the standard AAs were contacted with pre-saturated octanol and water in certain volume ratio and shake until reach equilibrium. After phase separation, the solute concentration were directly measured in both phase and Kow can be calculated from concentration ratio between octanol to aqueous phase (Qiao, Xia, & Ma, 2008) 2.3.1.1 Buffer aqueous phase Due to level of ionization depends on pH, so buffered solution was necessary to control the pH and form of AAs. The aqueous phase were prepared by mixing 10mM H3PO4 (pH=2) and 5 mM K2HPO4 (pH=9) and adjusted to desired pH by adding a small amount of 1 M HCl or NaOH . 23 2.3.1.2 Preparation of pre-saturation of aqueous and octanol phase Before mixing, the two solvents, water and octanol had been pre-saturated with each other to reduce the changing of solvent volume. The octanol-saturated aqueous phase was prepared by adding 100 ml octanol to 100 ml of adjusted pH- buffered water. The mixture was shaken for 1 day in mechanical shaker and left standing for at least 1 day to separate the both phases. The saturated octanol phase and saturated aqueous phase were removed into individual bottles and these solutions were ready to use for shake flask experiment. 2.3.1.3 Analytical method and validation The reverse-phase HPLC, Pelkin Elmer, equipped with Supelco Discovery C-18 column (25 cm x 4.6 mm, 5 µm) and diode array detector (DAD) setting wavelength at 255 nm was used to measure AA I and II in octanol and aqueous phase. The mobile phase was pre-mixed solution between methanol and ultrapure water (70:30 of methanol: water) and adjusted the pH to 2.3 by adding aliquot of 85% orthophosphoric acid. The flow rate of mobile phase was constant at 1ml/min. The injection volume was 30 µl by autosampler. The calibration curve of AA I & II in buffer solution and octanol were prepared separately. To prepare calibration curve, the series concentration of AAs standard solution were prepared separately in water and octanol and analyzed with HPLC-DAD. The linearity range and limit of detection (LOD) of both phases had been investigated. The LOD were determined by three time of signal to noise ratio (S/N). 2.3.1.4 Procedure Octanol-water partition coefficient was determined by shake flask method following OECD 107(OECD, 1995). In 25 ml glass centrifuge tubes, the 10 ml of octanol-saturated water was overlaid by 5 ml of water- saturated octanol which contain AAs in different concentration 24 (the volumetric ratio of octanol to water was 1:2). Then, tubes were placed in mechanical shaker at 180 rpm and room temperature for 24 hr to reach equilibrium. To separate the phase, tubes were put in centrifuge machine and rotate at high 5,000 rpm for 15 min and allow to settle for 24 hr. Figure 2.4 shows the phase separation between octanol and aqueous phase. Then the sample in both phase were taken. The 100 µl octanol sample was taken from the top layer by micropipette. The aqueous phase sample was taken by syringe with removable needle to minimize risk of including traces of octanol. Then, after removal the octanol phase, the pH of aqueous phase was measured by glass electrode pH meter, Orion 550A. Figure 2.4 Octanol and aqueous phase separation in shake flask experiment (For the interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this dissertation) The concentration of AA I and II were measured in both phase samples with RP-HPLC. The aqueous phase sample was directly measured whereas the octanol phase sample will be diluted 10 fold with methanol before measurement. The Kow were calculated from AA I and II concentration ratio between octanol phase and aqueous phase. The mass balance calculation 25 indicated mass loss by adsorption to cap and tube was less than 10%. The experiment was done at room temperature. 2.3.1.5 pH and ionic strength effect The experiment was set up similar to the procedure above except the buffered aqueous solution was adjusted pH from 2 to11 by adding aliquot of 1M HCl or NaOH until it reach desired pH. The AAs stock solution at fixed concentration was added into octanol and then two solutions were mixed together. At equilibrium after taking the sample of both phase, the pH of aqueous phase were measured again as final pH. The obtained log Kow were plotted with final pH and the non-linear regression according to Eq. 2.13 was fitted to experimental data by using template created in Excel spreadsheet. To determine effect of ionic strength, KCl was used as adjusting salt which was added to the pH adjusted buffer aqueous solution to vary the ionic strength from 0.01 - 1.0 M. The plot between log Kow and pH in different ionic strength solution were compared. 2.3.2 Kow by RP-HPLC method The Kow were also determined by RP-HPLC method which followed to OECD 117(OECD, 1989). This experiment will be conducted by Pelkin Elmer Reverse phase HPLC model with binary pump and vacuum degasser, autosampler and Supelco Discovery C-18 column (25 cm x 4.6 mm, 5 µm pore diameter). The six reference compounds were chosen from weakly ionizable monocarboxylic acids which have structure related to AAs and had the reported Log Kow. The set of benzoic and naphthalene carboxylic acid and their derivatives were chosen as the reference standard because their structures are quite similar to AAs. Their structure, pKa and log Kow were shown in Table 2.1. These values were based on literature values tested by 26 shake flask method. These reference solutions at 0.1 mg/ml were prepared in methanol and inject to HPLC column. Their retention times (RT) were determined with Diode Array Detector at wavelength 255 nm. Because these reference chemicals are ionizable, their retention times will be determined under phosphoric acid buffered mobile phase at pH 3 to suppress the ionization. The regression equation will be obtained from plotting between their Log Kow (molecular form) and logarithm of retention time (log RT). The log Kow of Aristolochic acid I and II were calculated from regression equation. Table 2.1 The molecular structure and literature values of pKa and log Kow of reference compounds Chemical Structure MW pKa Log Kow benzoic acid 122.12 4.21a 1.87a 3-nitrobenzoic acid 167.12 3.46a 1.83a 1-naphthoic acid (1- naphthalene carboxylic acid 172.18 3.69b 3.1b 1-naphthaleneacetic acid 186.2 4.24c 2.24d 1-naphthoxyacetic acid (1-Naphthyloxyacetic acid) 202.2 3.18e 2.60d a from Ming, Han, Qi, Sheng, & Lian, 2009; b from Han et al., 2012; c from Dippy, Hughes, & Laxton, 1954; d from Burgos & Pisutpaisal, 2006 ; e from "1-Naphthoxyacetic acid," 27 2.4 Result and discussion 2.4.1 Ionization of Aristolochic acid From the apparent Kow equation (Eq. 2.13) and pKa of AA I and II of 3.3 and 3.2(X. F. Fu et al., 2011) , we can calculate the speciation of neutral and ionized form correspond to pH as shown in Figure 2.5. It shows that, at pH < 2, most of AAs will be in neutral form but they will be most in ionized form at pH > 6. In natural environment where the pH is about 6-8, anions will be major form and contribute the behavior in the environment. The neutral and anion form behavior are expected to be different in term of water solubility, soil adsorption and mobility in environment. b) AA II 1.2 1.2 1 1 0.8 0.8 % of species % of species a) AA I 0.6 AAI 0.4 0.6 0.4 AAII AAI0.2 0.2 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 pH AAII- 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 pH Figure 2.5 Aqueous speciation of Aristolochic acid I and II as function of pH 2.4.2 Calibration curve and validation From Figure 2.6, the calibration curve of AA I and II were created from both aqueous and octanol phase. The figures showed linear line over concentration range 0.045 – 5.5µg/ml of AAs in both aqueous solution and octanol solution with r2 over 0.999. The limit of detection (LOD) 28 of AA I and II were 0.01 and 0.014µg/ml in aqueous solution and 0.008 and 0.009 µg/ml in octanol solution, respectively. This indicated HPLC with DAD detector had a good sensitivity and accuracy in analysis AAs in aqueous and octanol solution. a) Aqueous solution 800000 AA I : y = 134124x - 6732.2 R² = 0.9994 700000 AA II: y = 115339x - 6662.2 R² = 0.9995 600000 peak area 500000 400000 300000 200000 100000 AA I AA II 0 0 1 2 3 4 5 6 concentration (µg/ml) b) Octanol solution peak area 800000 700000 AA I : y = 148895x - 6283.1 R² = 0.9997 600000 AA II: y = 127504x - 9334.3 R² = 0.9995 500000 400000 300000 200000 AA I AA II 100000 0 0 1 2 3 4 5 6 concentration (µg/ml) Figure 2.6 The calibration curve of AA I and II in aqueous (a) and octanol (b) solution 29 2.4.3 Log Kow by shake flask method The Kow of AAs was determined by shake flask method at different three initial concentrations and native pH of 5-6. The results are shown in Table 2.2. The measured log Kow of AA I and II were quite constant at different initial concentration. The average log Kow of AA I and II were 1.65 and 1.23, respectively which indicate AAs were likely to partition in water. These Kow are lower than Kow of normal PAHs structure compounds such as naphthalene and phenanthrene which expected from weakly acid property of AAs. Log Kow of AA I was more than AA II. This expects to cause by one more alkyl branch on the AA I molecule structure that bring AA I more hydrophobic. Table 2.2 The log Kow of AA I and II by shake flask method at three different concentrations AAs Spiked amount (µg) pH Log Kow AAI AAII 80 6.60 1.64 1.24 120 5.66 1.65 1.22 160 5.88 1.66 1.22 Average 6.05 1.65 1.23 30 2.4.4 pH and ionic strength effect Figure 2.7 shows the apparent octanol water coefficient (log function of pH of aqueous solution. At the pH < 3, the log However, when the pH increased, the log ) of AA I & II as a was high and independent to pH. linearly decreased for 4 orders of magnitude until after pH 9, it was independent to pH again. The data were fitted well to the Eq.2.13 with correlation coefficient (r2)> 0.99. The figures indicated log Kow of AA I and II were 3.85 and 3.63 for neutral form and -0.55 and -0.85 for anion form, respectively. The log Kow of neutral form and anion form were 4 log unit or10,000 times difference. The neutral form of AAs is very hydrophobic and has high affinity to octanol phase whereas the ionized form is highly hydrophilic and has high affinity to water. This indicated that a pH is very important factor that control AAs behavior in environment. (a) AA I (b) AA II 4 4 Experimental 3.5 model 3 Model 3 2.5 2.5 log Kow" log Kow" Experimental 3.5 2 1.5 1 2 1.5 1 0.5 0.5 0 0 -0.5 -0.5 -1 -1 0 2 4 Figure 2.7 Experimental (a) and AA II (b) 6 pH 8 10 12 0 2 4 6 pH 8 10 and the fitted model curve versus pH after equilibrium: AA I 31 12 From fitted parameter with Eq. 2.13, Kow of AA I and II at various pH can be calculated by Eq. 2.16 and 2.17. (2.16) (2.17) Figure 2.8 shows effect of ionic strength on log Kow” over range of pH. It shows that ionic strength had no effect when the pH was low or most molecules were in neutral form. However, ionic strength had obvious effect at the high pH or most molecules were in anion form where the increasing of ionic strength caused increasing of log Kow. This evidence indicated that the complexation between cations (in this case K+) and AA- is possible and negative charge of AAs was turned to be neutral and caused AAs transfer to octanol phase. This finding was according to the ion pair transfer mechanism to maintain the electroneutrality of the solution (Word, 2002). The plotting of log Kow of anion form (at pH>10) with logarithm of ionic strength in solution shows the linear relationship as shown in Figure 2.9. This finding agreed with previous studies which indicated the Kow of organic anions linearly increased with increasing ionic strength on the log scale (Jafvert et al., 1990; B. E. Nowosielski & Fein, 1998). The results suggest that cations in soil solution can alter mobility of AAs in environment especially in alkaline condition.The regression equations of AA I and II had similar slope 0.6 and the equations can be used to predict log Kow at other ionic strengths. 32 (a) AA I 4 AA II, I=0.01 M 3.5 AA II , I=0.1M 3 AA II, I=0.5 M log Kow" 2.5 AA II, I= 1 M 2 1.5 1 0.5 0 -0.5 -1 0 1 2 3 4 5 6 pH 7 8 9 10 11 12 (b) AA II 4 AA I , I =0.01M AA I, I=0.1M AA I, I=0.5 M AA I, I = 1 M 3.5 3 log Kow" 2.5 2 1.5 1 0.5 0 -0.5 -1 0 Figure 2.8 Experimental AA I (a) and AA II (b) 1 2 3 4 5 6 pH 7 8 9 10 11 12 and fitted model at different ionic strength in aqueous phase: 33 0.8 0.6 AA I: y = 0.6308x + 0.6773 R² = 0.9944 0.4 Log Kow" 0.2 AA II : y = 0.6085x + 0.3201 R² = 0.9939 0 -0.2 -0.4 -0.6 -0.8 AA I AA II -1 -2.5 -2 -1.5 -1 -0.5 0 Logµ Figure 2.9 strength (Log µ) of anion of AA I and II at high pH as a function of logarithm of ionic 2.4.5 Log Kow by HPLC method To measure log Kow by HPLC method, the reference compounds and AA I and II were injected to C-18 column with isocratic mobile phase (methanol: water is 70:30) at the pH 3. The retention time (RT) of their peaks and log Kow are shown in Table 2.3. The RT found to increase with the molecular weight and log Kow of compounds which is general trend for hydrophobic chemicals. Figure 2.10 shows the relationship between logarithm of retention time and log Kow of reference compounds and AA I and II. The RT of AA I and II were over RT from reference compounds because of the larger size of molecules. The reference compounds that had molecular weight more than AA I and II and had reported log Kow by shake flask method from literatures were not found. The linear regression gave equation: log Kow = 7.14logRT-2.61. The log Kow of AA I and II calucalated from this regression equation were 3.71 and 3.27, respectively. The low correlation coefficient (r2 = 0.69) was expected from the unclear degree of ionization of reference compounds from literatures which resulted in uncertainty of obtained Kow value. This 34 method quite sensitive to the correctness of Kow of reference compounds (Chamberlain, Evans, & Bromilow, 1996; Paschke, Neitzel, Walther, & Schuurmann, 2004). Table 2.3 The log Kow of reference compounds and their retention time from RP-HPLC system Chemical MW pKa Log Kow RT benzoic acid 122.12 4.21 1.87 4.11 3-nitrobenzoic acid 167.12 3.46 1.83 4.33 1-naphthoic acid 172.18 3.69 3.1 5.4 1-naphthaleneacetic acid 186.2 4.24 2.24 5.18 1-naphthoxyacetic acid 202.2 3.18 2.60 5.7 AA I 341.27 3.3 - 7.66 AA II 311.25 3.2 - 6.64 (1- naphthalene carboxylic acid) (1-naphthyloxyacetic acid) 4.5 4 y = 7.14x - 2.61 R² = 0.69 3.5 Log Kow 3 AA I AA II 2.5 2 1.5 1 0.5 0 0.50 0.55 0.60 0.65 0.70 0.75 0.80 Log RT 0.85 0.90 0.95 1.00 Figure 2.10 The linear correlation between log Kow and log RT of reference compounds and the extrapolation of AA I and II 35 2.5 Conclusion This study had measured octanol-water partition coefficient of AA I and II by direct shake flask method and indirect HPLC method. The shake flask method gives reliable Kow if concentrations in both phases are over the detection limit of analytical method. At natural pH, the Kow of AA I and II were 1.65 and 1.23 which indicated AAs were quite hydrophilic. The Kow obtained from HPLC method was questionable because of lacking the accuracy reported Kow of reference compounds. Kow of AAs was strongly effect by the pH where the Kow decreased when pH increased. The AAs in acid solution were very hydrophobic (log Kow > 3) but they were very hydrophilic in basic solution (log Kow < 0). At high pH where most molecules were in anion form, the Kow increased when ionic strength increased. This expected from the complex formation between anion of AAs and counter cation. This study shows that the prediction fate and transport of AAs in environment need to account for the pH and ionic strength of surrounding media. 36 REFERENCES 37 REFERENCES 1-Naphthoxyacetic acid. Retrieved March 10, 2013, from http://www.chemexper.net/specification_d/chemicals/supplier/cas/1Naphthoxyacetic%20acid.asp Berthod, A., Carda-Broch, S., & Garcia-Alvarez-Coque, M. C. (1999). Hydrophobicity of ionizable compounds. A theoretical study and measurements of diuretic octanol-water partition coefficients by countercurrent chromatography. Analytical Chemistry, 71(4), 879-888. doi: 10.1021/ac9810563 Brandt, C. A., Becker, J. M., & Porta, A. (2002). Distribution of polycyclic aromatic hydrocarbons in soils and terrestrial biota after a spill of crude oil in Trecate, Italy. Environmental Toxicology and Chemistry, 21(8), 1638-1643. doi: 10.1897/15515028(2002)021<1638:dopahi>2.0.co;2 Briggs, G. G., Bromilow, R. H., Evans, A. A., & Williams, M. (1983). RELATIONSHIPS BETWEEN LIPOPHILICITY AND THE DISTRIBUTION OF NON-IONIZED CHEMICALS IN BARLEY SHOOTS FOLLOWING UPTAKE BY THE ROOTS. Pesticide Science, 14(5), 492-500. doi: 10.1002/ps.2780140506 Brooke, D. N., Dobbs, A. J., & Williams, N. (1986). OCTANOL - WATER PARTITION COEFFICIENTS(P) - MEASUREMENT, ESTIMATION, AND INTERPRETATION, PARTICULARLY FOR CHEMICALS WITH P-GREATER THAN-10(5). Ecotoxicology and Environmental Safety, 11(3), 251-260. doi: 10.1016/01476513(86)90099-0 Burgos, W. D., & Pisutpaisal, N. (2006). Sorption of naphthoic acids and quinoline compounds to estuarine sediment. Journal of Contaminant Hydrology, 84(3-4), 107-126. doi: 10.1016/j.jconhyd.2005.12.008 Chamberlain, K., Evans, A. A., & Bromilow, R. H. (1996). 1-octanol/water partition coefficient (K-ow) and pK(a) for ionisable pesticides measured by a pH-metric method. Pesticide Science, 47(3), 265-271. doi: 10.1002/(sici)1096-9063(199607)47:3<265::aidps416>3.0.co;2-f Debruijn, J., Busser, F., Seinen, W., & Hermens, J. (1989). DETERMINATION OF OCTANOL WATER PARTITION-COEFFICIENTS FOR HYDROPHOBIC ORGANICCHEMICALS WITH THE SLOW-STIRRING METHOD. Environmental Toxicology and Chemistry, 8(6), 499-512. doi: 10.1897/1552-8618(1989)8[499:dowpcf]2.0.co;2 Dettenmaier, E. M., Doucette, W. J., & Bugbee, B. (2009). Chemical Hydrophobicity and Uptake by Plant Roots. Environmental Science & Technology, 43(2), 324-329. doi: 10.1021/es801751x 38 Dippy, J. F. J., Hughes, S. R. C., & Laxton, J. W. (1954). CHEMICAL CONSTITUTION AND THE DISSOCIATION CONSTANTS OF MONOCARBOXYLIC ACIDS .14. MONOMETHYLCYCLOHEXANECARBOXYLIC ACIDS. Journal of the Chemical Society(DEC), 4102-4106. doi: 10.1039/jr9540004102 Fisk, A. T., Norstrom, R. J., Cymbalisty, C. D., & Muir, D. C. G. (1998). Dietary accumulation and depuration of hydrophobic organochlorines: Bioaccumulation parameters and their relationship with the octanol/water partition coefficient. Environmental Toxicology and Chemistry, 17(5), 951-961. doi: 10.1897/1551-5028(1998)017<0951:daadoh>2.3.co;2 Fu, X. F., Liu, Y., Li, W., Bai, Y., Liao, Y. P., & Liu, H. W. (2011). Determination of dissociation constants of aristolochic acid I and II by capillary electrophoresis with carboxymethyl chitosan-coated capillary. Talanta, 85(1), 813-815. doi: 10.1016/j.talanta.2011.03.088 Han, S. Y., Qiao, J. Q., Zhang, Y. Y., Lian, H. Z., & Ge, X. (2012). Determination of noctanol/water partition coefficients of weak ionizable solutes by RP-HPLC with neutral model compounds. Talanta, 97, 355-361. doi: 10.1016/j.talanta.2012.04.045 Jafvert, C. T., Westall, J. C., Grieder, E., & Schwarzenbach, R. P. (1990). DISTRIBUTION OF HYDROPHOBIC IONOGENIC ORGANIC-COMPOUNDS BETWEEN OCTANOL AND WATER - ORGANIC-ACIDS. Environmental Science & Technology, 24(12), 1795-1803. doi: 10.1021/es00082a002 Kah, M., & Brown, C. D. (2006). Adsorption of ionisable pesticides in soils. Reviews of Environmental Contamination and Toxicology, Vol 188, 188. doi: 10.1007/978-0-38732964-2_5 Kah, M., & Brown, C. D. (2008). LogD: Lipophilicity for ionisable compounds. Chemosphere, 72(10), 1401-1408. doi: 10.1016/j.chemosphere.2008.04.074 Kaiser, K. L. E., & Valdmanis, I. (1982). APPARENT OCTANOL WATER PARTITIONCOEFFICIENTS OF PENTACHLOROPHENOL AS A FUNCTION OF PH. Canadian Journal of Chemistry-Revue Canadienne De Chimie, 60(16), 2104-2106. doi: 10.1139/v82-300 Karickhoff, S. W., Brown, D. S., & Scott, T. A. (1979). SORPTION OF HYDROPHOBIC POLLUTANTS ON NATURAL SEDIMENTS. Water Research, 13(3), 241-248. doi: 10.1016/0043-1354(79)90201-x Lee, L. S., Rao, P. S. C., Nkedikizza, P., & Delfino, J. J. (1990). INFLUENCE OF SOLVENT AND SORBENT CHARACTERISTICS ON DISTRIBUTION OF PENTACHLOROPHENOL IN OCTANOL-WATER AND SOIL-WATER SYSTEMS. Environmental Science & Technology, 24(5), 654-661. doi: 10.1021/es00075a006 39 Ming, X., Han, S. Y., Qi, Z. C., Sheng, D., & Lian, H. Z. (2009). Chromatographic retention prediction and octanol-water partition coefficient determination of monobasic weak acidic compounds in ion-suppression reversed-phase liquid chromatography using acids as ion-suppressors. Talanta, 79(3), 752-761. doi: 10.1016/j.talanta.2009.04.069 Ney, R. E. (1995). Fate and transport of organic chemicals in the environment : a practical guide (2nd ed.). Rockville, Md.: Government Institutes. Nowosielski, B. E. (1998). Experimental study of octanol-water partition coefficients for 2,4,6trichlorophenol and pentachlorophenol: Derivation of an empirical model of chlorophenol partitioning behaviour. (M.Sc. MQ44235), McGill University (Canada), Canada. Retrieved from http://ezproxy.msu.edu/login?url=http://search.proquest.com/docview/304504053?accoun tid=12598 ProQuest Dissertations & Theses (PQDT); ProQuest Dissertations & Theses A&I database. Nowosielski, B. E., & Fein, J. B. (1998). Experimental study of octanol-water partition coefficients for 2,4,6-trichlorophenol and pentachlorophenol: Derivation of an empirical model of chlorophenol partitioning behaviour. Applied Geochemistry, 13(7), 893-904. doi: 10.1016/s0883-2927(98)00015-8 OECD. (1989). OECD Guideline for the testing of chemicals Partition Coefficient (noctanol/water), High Performance Liquid Chromatography (HPLC) Method. OECD. (1995). OECD Guideline for the testing of chemicals Partition Coefficient (noctanol/water): Shake Flask Method. Paschke, A., Neitzel, P. L., Walther, W., & Schuurmann, G. (2004). Octanol/water partition coefficient of selected herbicides: Determination using shake-flask method and reversedphase high-performance liquid chromatography. Journal of Chemical and Engineering Data, 49(6), 1639-1642. doi: 10.1021/je049947x Poole, S. K., & Poole, C. F. (2003). Separation methods for estimating octanol-water partition coefficients. Journal of Chromatography B-Analytical Technologies in the Biomedical and Life Sciences, 797(1-2), 3-19. doi: 10.1016/j.jchromb.2003.08.032 Qiao, Y., Xia, S., & Ma, P. (2008). Octanol/water partition coefficient of substituted benzene derivatives containing halogens and carboxyls: Determination using the shake-flask method and estimation using the fragment method. Journal of Chemical and Engineering Data, 53(1), 280-282. doi: 10.1021/je700381u Riise, G., & Salbu, B. (1992). MOBILITY OF DICHLORPROP IN THE SOIL-WATER SYSTEM AS A FUNCTION OF DIFFERENT ENVIRONMENTAL-FACTORS .1. A BATCH EXPERIMENT. Science of the Total Environment, 123, 399-409. doi: 10.1016/0048-9697(92)90163-m 40 Schwarzenbach, R. P., Gschwend, P. M., & Imboden, D. M. (2003). Environmental Organic Chemistry (2nd Ed.). New Jersey: John Wiley & Sons, Inc. Schwarzenbach, R. P., & Westall, J. (1981). TRANSPORT OF NON-POLAR ORGANICCOMPOUNDS FROM SURFACE-WATER TO GROUNDWATER - LABORATORY SORPTION STUDIES. Environmental Science & Technology, 15(11), 1360-1367. doi: 10.1021/es00093a009 VCCLAB. (2011). Virtual Computational Chemistry Laboratory. 2012, from http://www.vcclab.org/lab/alogps/ Veith, G. D., Call, D. J., & Brooke, L. T. (1983). STRUCTURE TOXICITY RELATIONSHIPS FOR THE FATHEAD MINNOW, PIMEPHALES-PROMELAS - NARCOTIC INDUSTRIAL-CHEMICALS. Canadian Journal of Fisheries and Aquatic Sciences, 40(6), 743-748. Westall, J. C., Leuenberger, C., & Schwarzenbach, R. P. (1985). INFLUENCE OF PH AND IONIC-STRENGTH ON THE AQUEOUS NONAQUEOUS DISTRIBUTION OF CHLORINATED PHENOLS. Environmental Science & Technology, 19(2), 193-198. doi: 10.1021/es00132a014 Word, R. C. (2002). Partition of ionized and neutral halogenated phenols between octanol and water, lipid membranes and water, and sarcoplasmic reticulum and water. (Doctor of Philosophy), Portland State University. Yuying, D., Guanghui, D., Ying, C., Zhuang, W., & Cheng, S. (2009). Determination and estimation of partitioning properties for substituted phosphates and thiophosphates. Environmental Monitoring and Assessment, 152(1-4), 443-450. doi: 10.1007/s10661008-0328-0 41 Chapter 3 Aqueous solubility 3.1 Introduction Aqueous solubility is an important parameter that controls many fate and transport processes such as soil adsorption, complexation, bioaccumulation and plant uptake. It determines amount of chemicals that can be available in environment (P.G. Wightman, 1997). It predicts how much of chemicals can be leach into groundwater or runoff with surface water. It is also related to the aqueous phase partitioning properties such as octanol-water partition coefficient (Kow) or soil-water partition coefficient (Kd) (Arcand, Hawari, & Guiot, 1995). Due to its importance to environmental processes, the knowledge about the solubility of organic pollutants is necessary (M. T. C. Figueroa, 1989)(M. T. C. Figueroa, 1989). The definition of aqueous solubility is amount of chemicals (solids, liquids and gases) that equilibrated as pure phase in the water. It is the saturation concentration where chemicals can be dissolved in pure water at given temperature. For non-polar hydrophobic compounds, solubility mainly effect by temperature and dissolved inorganic salt. The increasing of temperature will increase the solubility of solid due to less energy required to melt the solid at high temperature. In general, the presence of inorganic ions such as Na, K, Mg, Ca, Cl, HCO3, SO4 will decrease solubility by increasing activity coefficient of compounds which referred as “salting out” effect. The presence of highly water soluble organic compounds e.g. methanol ethanol can increase the solubility of organic solid by “cosolvent” phenomenon (R.P. Schwarzenbach et al., 2003). 42 Unlike the non-polar hydrophobic chemicals, the solubility of ionized organic chemicals was largely effect by pH because of different forms of solute molecules, other than temperature. The ionic strength and the presence of other ions also had been reported to alter the solubility by affecting their activity coefficient (M. T. C. Figueroa, 1989). The ion-pair formation between organic and inorganic were reported which raise the solubility of ionized organic compounds. Moreover, the complexation may alter to sorption properties of both metal and organic acids by changing their electrostatic charge which affect to mobility of these chemicals in environment (Daughney & Fein, 1997). Due to the solubility of AAs in aqueous phase can govern many processes, so the knowledge about aqueous solubility behavior of AAs is essential. Since environmental factors such as pH, ion presence, ionic strength affected the solubility, so these dependence parameters were also included in this study. 3.2 Literature reviews 3.2.1 Solubility of organic acids and pH effect Organic acids when dissolved in water can be dissociated as ionized molecules and can exist as 2 forms, neutral and ionized. The aqueous solubility among these 2 forms was reported to be largely different. Generally, the neutral molecules are relative hydrophobic and not likely to dissolve in water and their solubility are independent to pH. In the other hand, the charged molecules are very polar and interact with water very well and then high solubility (Arcand et al., 1995; P. G. Wightman & Fein, 1999). 43 To develop the model to estimate solubility of organic acid, we assume the organic acid solid phase is dissolved in water until equilibrium as Eq.3.1. This dissolved organic acids (HA) can be dissociated when pH>pKa as anion species and proton as Eq. 3.2. (3.1) (3.2) The total solubility of organic acids is the summation of molality of unionized and ionized species which express as Eq. 3.3 (P.G. Wightman, 1997). (3.3) From dissociation constant (Ka) equation of the acid which is expressed in Eq 3.4, it can be rearrange to present in term of molarity of anion in Eq. 3.5 by assuming acitivity coefficient of neutral species equal to 1. (3.4) (3.5) From and , and assuming in dilute solution, the molarity of anion can be written as Eq.3.6 (3.6) Put this term in Eq. 3.3, and finally the total solubility will be expressed as Eq. 3.7. (3.7) 44 This equation is very useful to estimate the solubility of organic acids at any pH when and pKa are known (Arcand et al., 1995; P. G. Wightman & Fein, 1999). For example, Wightman & Fein, 1999 showed this model were fitted well with pentachlorophenol (PCP) solubility data as shown in Figure 3.1. Log molality -2 -3 -4 -5 pH 2 6 4 8 Figure 3.1 Pentachlorophenol (PCP) solubility as a function of pH (modified from Wightman & Fein, 1999) This figure shows that the solubility increase when the pH increase. At pH<4, the solubility is constant which indicates the solubility of undissociated form is not effect by pH. At pH > 5, the log solubility increase with pH and close to linear relationship. However, this equation is best applied in pure water or dilute electrolyte solution, it could not be generalized to concentrated electrolyte solution. Wightman, 1999 showed the solubility of PCP was enhanced and deviated out from the linear line at high pH and high concentration of salts. This non-linear increasing was explained by metal-PCP complex formation. 45 3.2.2 The ionic strength, ions presence and complexation formation effect In non-dilute electrolyte solution, the activity coefficient of ionized compounds ( ) in Eq.3.5 is not equal to 1 and the solubility will be effected by ionic strength by changing their activity coefficient. From Debye-Huckle equation, the activity coefficient were related to ionic strength as shown in Eq.3.8 (M. T. C. Figueroa, 1989). (3.8) Hence, from combining between Eq. (5) and (7) , the solubility can be related to ionic strength as shown Eq. 3.9. This equation shows the solubility will increase with increasing ionic strength. (3.9) The presence of different ions also has influence to solubility of organic acids. Some studies show the organic acids have high tendency to form the complex with metal that dissolved in the solution (Daughney & Fein, 1997; P. G. Wightman & Fein, 1999). The metal – organic acid complexes were known to exist in aqueous solution that contains the anion and metal in high concentration. These complexes can enhance the solubility of organic acids. The potentiometric studies which measure activities of free metal ions e.g. Cd2+, Cu2+ in solution that contain organic acids showed the activity of free metals decreased when the concentration of organic acids increased which was expected from complexation betweeen metals and organic anion (Daughney & Fein, 1997). 46 Wightman, 1997 found that the solubility of PCP at high salt and high pH condition were largely increased which cannot be described by only ionic strength effect but also with the presence of metal-PCP complex. To incorporate the complexation to the model, Daughney & Fein, 1997 and Wightman & Fein, 1999 introduce the 1:1 complexation model between monovalent metals (M+) with anion of organic acids (A-) where equilibrium reaction at fixed temperature and pressure was shown in Eq. 3.10 (3.10) Where the stability constant (K) express as: (3.11) (3.12) The metal-organic complex activity was accounted in mass balance equation which shows in Eq.3.13. The model that included metal-organic complex found to fit with experimental data very well. (3.13) The degree of increasing solubility by the complexation found to depend on the strength of stability constant between organic acids and metal cations. The strength of stability constant determine by the polarizing effect of cations which increase with increasing the charge (Z) and decreasing of radius (r) of cations or combine as ratio Z/r. So, the stability constants of monovalent cations are expect to be less than divalent or trivalent cations (Orlov & Belkina, 2011) . Mantoura et al. 1978 had determined the stability constant of the complexes of humic 47 material and metals. They found that the order of stability constant follow to the Irving-William series e.g. Mg2+silicate>sulfate>> nitrate > chloride (Hyun, Lee, & Rao, 2003). The studies showed that phosphate and sulfate had inhibited the adsorption of organic anion due to the competition to anion exchange sites. This effect will be much greater for soils that have high content of Fe/Al oxides. There were some studies showed that phosphate can form specific adsorption via ligand exchange with clay minerals or Fe/Al oxides in soil (Hyun et al., 2003; Regitano, Bischoff, Lee, Reichert, & Turco, 1997). These indicate that the high volume of organic chemicals could be release back in to the solution if there are applying phosphate to the soil. 69 4.2.5 Desorption Desorption is the important process that control the fate and transport of chemicals in environment. It determines the amount of adsorbed chemical released back into water which increase risk to environment such as contaminating in surface or ground water. The sorbed chemicals will not be available for degradation by microorganism or uptake by plant before the desorption (Ren, Wang, & Zhou, 2011). The desorption determines whether the sorption process reversible or irreversible. If total desorption is more than 75% of amount adsorbed, the adsorption is considered as reversible. The sorption and desorption isotherm of most organic chemicals, are significantly deviated. Mostly, the desorption is delayed or hindered when compare to adsorption process. Desorption may be described by 2 steps, the fast release step and slow diffuse out step. The fast desorption step represents the molecule adsorbed on non-porous mineral surface sites whereas the slow desorption step represents the molecules adsorbed in micropores in organic matter matrices which are more difficult to release (Ren et al., 2011; G. S. Yuan & Xing, 2001) 4.2.6 Desorption isotherm The desorption isotherm can be determined from several desorption cycles of same sample. The plotting between remaining concentration of chemicals in solid and concentration released in liquid phase at equilibrium of each cycle will be desorption isotherm. Mostly, the desorption isotherm can be described by Freundlich model as expressed in Eq 4.9. (Chefetz, Bilkis, & Polubesova, 2004; Piwowarczyk & Holden, 2012) : (4.9) 70 Where is desorption coefficient, liquid and solid phase respectively and and are concentration of chemicals in is desorption Freundlich exponent which indicates the non-linearity of isotherm. Values of and can be obtained from fitting the model to experimental data. Some studies introduced two-compartment desorption isotherm to represents two different desorption rates, “fast and slow”, which can be described by Eq. 4.10. This model cooperates the linear term and inverse exponential term which correspond to two different sorption sites. The linear compartment represents the weak adsorption and the easily releasing of sorbate to solution whereas the exponential compartment represents the strong retentive force which requires longer time to desorb. This model can describe the desorption isotherm better than Freudlich model especially for extreme point of isotherm (Barriuso, Baer, & Calvet, 1992). (4.10) Where Ce and Cs are equilibrium concentration of chemical in solution and adsorbed in the solid phase, respectively. Cs is concentration of chemical adsorbed in exponential compartment. Kf1 and Kf2 are linear and exponential parameters. The example of twocompartment desorption isotherm and the split up of linear and exponential component are shown in Figure 4.2. 71 Adsorbed concentration Desorption isotherm Linear component Exponential component Solution concentration Figure 4.2 The two-compartment desorption isotherm and the splitting linear and exponential compartment model (modified from Barriuso, Baer, & Calvet, 1992.) 4.2.7 Hysteresis index The desorption deviation is frequently described by hysteresis index (HI). The hysteresis implied the rules that the adsorbed chemical has different limit degree of reversibility which depend on physico-chemical properties of soils and chemicals. Hysteresis coefficient can be calculated from the ratio of Freundlich exponent of desorption isotherm adsorption isotherm to the as shown in Eq. 4.11. (Chefetz et al., 2004; Cox, Koskinen, & Yen, 1997; Piwowarczyk & Holden, 2012) : (4.11) If the Freundlich coefficient of desorption and sorption are equal, HI is 100 and the hysteresis will not be observed. The lower HI value indicates more desorption hysteresis or more difficulty of the sorbed molecules to desorb from the soil matrices The hysteresis occurs by many reasons which may include the effect of experimental procedure such as loss of solute by 72 volatilization, sorption to tube or biodegradation or strongly binding to the sorbent (G. S. Yuan & Xing, 2001). The degree of hysteresis is depended on physico-chemical properties of chemicals, soil and environmental condition. 4.2.8 Factors effect to desorption hysteresis 4.2.8.1 Effect of organic matter content The desorption hysteresis can be impacted by organic matter content in soil. Some studies showed the hysteresis increase with increase organic matter. The resistance to desorption of hydrophobic compounds is believed from the strongly bound of chemicals to soil organic matter (G. M. Fu, Kan, & Tomson, 1994; Jenks, Roeth, Martin, & McCallister, 1998). This hysteresis from organic carbon is also controlled by type and location of organic matter. The aged and more condensed of soil organic matter shows more apparent sorption – desorption hysteresis (Liang, Dang, Liu, & Huang, 2005). The organic matter of aged soil will have reduced and condensed structure like glassy state while the young organic matter are mainly contained oxygen functional group and easily hydrated in aqueous solution like the rubbery state where the sorbed molecules can freely diffuse in and out. The rubbery phase is responsible for linear and simultaneously sorption/desorption of solute whereas the glassy phase is responsible for non-linear and slow sorption by “hole filling” and slow desorption which leading to hysteresis (Chefetz et al., 2004; Lesan & Bhandari, 2003) However, some studies showed the result in opposite direction where small organic matter soils had more desorption hysteresis than high organic matter soils. Even, the larger fraction of molecules adsorbed in high organic matter soil, they were able to desorb rapidly than low organic matter soil. This can be explained that organic matter provides preferential and 73 easily accessible sorption sites, so it retard the chemical to adsorb on mineral surface sites which may form stronger adsorption (Lesan & Bhandari, 2003). 4.2.8.2 Initial concentration The initial concentration of sorbate is also affected to hysteresis. High concentration of sorbate found to increase desorption hysteresis because high concentration gradient will drive molecules into deeper sites in organic matter matrices or microporous region. The strong bond such as H-bond may be formed, resulting in decrease fraction of sorbate to be desorbed, and then the hysteresis increases (Chefetz et al., 2004; Ren et al., 2011). 4.2.8.3 Contacting time The hysteresis also depends on contacting time where slowing desorption was observed when increasing the sorption time. This is because the compounds can diffuse into deeper sites and form the stronger bound to sorption sites which increases hysteresis effect. The sorption is considered to reversible for small contact time. Some studies showed that, the chemicals desorbed from slow sorption in aged soil were 2 or 3 orders of magnitude less than freshly spiked soil. Moreover, the hysteresis also was found to relate with Freudlich coefficient(Kf) which means the hysteresis will increase when greater amount of sorbate had been adsorbed. (J. P. Gao, Maguhn, Spitzauer, & Kettrup, 1998; Lesan & Bhandari, 2003). 4.2.8.4 Effect of pH Desorption of ionized organic compounds is strongly effect by pH. At low pH where the sorbate in molecular form, the desorption and adsorption isotherm are similar or no hysteresis occur. However, when pH increased, most of organic acids are in anion form and charged 74 surface sites become more negative, so the cation bridging may be occurred. As a result, the desorption isotherm at high pH found to be more deviate from adsorption isotherm. These patterns also can be observed by hysteresis coefficient which reduce when pH increase. The more difficulty of the desorption of ionized form is due to the formation of more stronger binding energy of ion exchange or cation-bridging than weaker hydrophobic interaction (Caceres et al., 2010; Ren et al., 2011). 4.2.8.5 Effect of cations Desorption of ionized organic compounds also impacted by presence of cations in the soil solution. For example, Chen, Wang, & Pei, 2014 showed that Ag+, Zn2+, Al3+ had promoted the desorption of Trichlorophenol (TCP) from ash. They implied that these metals replaced the previous adsorbed TCP and the hysteresis decreased when increased valence of cation where TCP was easiest to desorp in Al3+ solution. Weber, 1982 showed that adsorbed pesticides on organic matters/clays were desorbed in paraquat (2+) solution much better than only water. The probably desorption mechanism was cation exchange where paraquat (2+) replaced the adsorbed pesticides. The desorption reaction enhanced by cation exchange reaction was introduced by Wu et al.,2013 and Weber, 1982 which expressed in Eq. 4.12 where M is exchangeable cations and C is ionizable organic compounds. They found that desorption increased with increasing charge of exchange cations e.g. Al3+>Ca2+>Na+. C-Clay/OM + Mn+  M-Clay/OM + nC+ 75 (4.12) However, the presence of cations can also inhibit the desorption if they can form strongly bond between soil and organic compounds. Moreale & Vanbladel, 1979 showed that the desorpion hysteresis of amines adsorbed on saturated cations clays increased with increasing polarizing power of cations e.g.Fe3+> Mg2+>Ca2+>K+>Na+. This was explained by the increasing stabtilty of amines - cations surface complex on clay. The sorption in Fe solution was almost irreversible due to the very strong Fe-amines complex. 4.2.8.6 Effect of anions Hyun et al., 2003 studied the effect of presence of anions to desorption hysteresis of organic compounds that adsorp to soil by anion exchange mechanism. They found that there were no or little hysteresis occur when desorption organic chemicals in Cl- or PO43- solution which indicated the Cl- or PO43- can replace organic chemicals very well. The desorption increase with the effectiveness of replacibilty of anions e.g. the PO43- can displace organic anions better than Cl-. 4.3 Material and method To study the adsorption – desorption process of AAs in the soil, the batch equilibrium method (OECD 106) was followed for experimental set up as describe below(OECD, 2000). 4.3.1 Soil samples and their properties Four types of soils, Wooster CT, Wooster NF, Hoytville CT and Hoytville NF, provided by Department of plant, soil and microbial sciences, Michigan State University had been used in this study. The soil samples were collected from long-term tillage treatment research site at the Ohio Agricultural Research and Development center at depth 0- 5 cm (E. J. Park & Smucker, 76 2005). The CT and NF mean samples taken from conventional tilled (CT) site and site adjoining to native forest (NF), respectively. The Wooster is fine, loamy, mesic Typic Fragiudalf soil and Hoytville is fine, illitic, mesic Mollic Epiaqualf soil. Some physical and chemical properties of theses soils are shown in Table 4.1. Soils were gently ground by mortar and pestle, 2 mm sieved and stored in dry and room temperature prior use. Table 4.1 Physical, chemical properties and composition of test soil samples Soil pHa %OC Composition Soil texture % clay %silt %sand Bulk density CECb AECb (g/cm3) (cmol/kg) (cmol/kg) Wooster CT 5.78 0.9 14.8 64.8 20.4 Silt loam 1.74 7.10 -0.36 Wooster NF 3.91 2.8 11.7 64.3 24.0 Silt loam 1.38 4.00 -0.26 Hoytville CT 5.88 2.3 35.9 45.9 18.2 Silty clay 1.84 10.95 -0.62 1.60 12.48 -0.94 loam Hoytville NF 5.98 7.6 36.7 47.0 16.3 Silty clay loam a Soil pH was determined in CaCl2 0.01 M at soil: solution ratio= 1:5 (Al-Busaidi, Cookson, & Yamamoto, 2005) b CEC and AEC measure at native soil pH range 5-7 by method described below 4.3.2 Methods for determination of CEC and AEC in soils The cation ion exchange capacity and anion exchange capacity were determined by point at zero net charge determination at natural soil pH (Zelazny, He, & Vanwormhoudt, 1996). The 1 g of soil samples were added to 25 ml glass centrifuge tube and measure weight of tube plus soil sample. Soils were saturated with 20 ml of 1 M KCl and shake for 1 hr with mechanical shaker at 150 rpm. Then, centrifuge the soil solution at 7,500 rpm for 30 min and discard the supernatant. Wash these soils with 20 ml 0.01M KCl for ½ hr and centrifuge to remove the 77 supernatant , repeat this washing steps for 3 times. The final supernatants were collected to determine pH, K+ and Cl- in retain solution (C1). Weigh the tube to calculate the retain volume of final wash solution in the soil (V1). The soil pellet were saturated with 10ml 0.5 M NaNO3 solution where the Na+ and NO3- will replace K+ and Cl- in the soil, respectively. Shake the tube for 1 hr, centrifuge at 7,500 rpm for 30 min and collect the replacing solution in volumetric flask. This step was repeated for 3 times and supernatant were bring together and measured the volume (C2). Analyze K+ and Cl- of this displacing solution by Atomic Absorption Spectrophotometer (AAS) or Ion Chromatography(C2). The CEC and AEC in centimole per kg were calculated according to the equation 4.13 and 4.14: (4.13) (4.14) 4.3.3 Analytical method validation The blank soil matrix solution was obtained by equilibrate most adsorpbability soil (Hoytville NF) 0.2 g in 20 ml 0.01 M CaCl2 for overnight and then centrifuge at 7,500 rpm to obtain the supernatant. The certain amount of standard mixture of Aristolochic Acid I and II were added to blank matrix solution within concentration range 0.05 – 1 µg/ml for set up calibration solutions and were analyzed by HPLC- DAD. The accuracy, precision, reproducibility, detection limit and recovery had been investigated. 78 4.3.3.1 Accuracy and precision The accuracy of the measurement method was determined by spiking standard solution of AAI and AAII in soil blank matrix (0.045 , 0.45 and 0.9 µg/ml for AAI and 0.055 , 0.55 and 1.1 µg/ml for AAII) and analyzed for 6 times (n=6). % recovery was calculated from Eq. 4.15 and reported as accuracy at each concentration. (4.15) The precision of AAI and AAII were tested from three standard solutions (0.045 , 0.45 and 0.9 µg/ml for AAI and 0.055 , 0.55 and 1.1 µg/ml of AAII) for 6 time in one day (intraday, n=6) and twice a day over 3 consecutive days (interday, n=6 ). % relative standard deviation (%RSD) of peak area of each concentration were calculated and reported as precision at each concentration (Kuo et al., 2010; J. B. Yuan et al., 2008). 4.3.3.2 Linearity Linearity of the method was tested by six concentration of standard solution of AAI and AAII ranging from 0.045 -0.9 and 0.055-1.1 µg/ml, respectively. Each concentration was analyzed for three times. The average peak areas were plot with concentration. The linear regression equation and correlation coefficient (r2) was obtained from graph. 4.3.3.3 Limit of detection(LOD) and limit of quantification (LOQ) The limit of detection (LOD) and quatification (LOQ) of AAI and AAII were tested by analyzing minimum concentration (0.045 and 0.055 ug/ml of AA I and II respectively) for 7 times and calculated for standard deviation (SD). The LOD was equal to three times of SD (signal to noise ratio = 3) and LOQ was equal to ten times of SD (signal to noise ratio =10). 79 4.3.4 Adsorption experiment 4.3.4.1 Preliminary and kinetic study First, the adsorption kinetic of all soils was determined. The 0.1 g of each soil was added to 7 centrifuge tubes with Teflon-liner screw cab. Soil solutions were pre-equilibrated with 9 ml of 0.01 M CaCl2 solution for overnight (12hr) and then 1 ml of stock AAs solution was spiked to make final volume to 10 ml. The mixtures were shaken with mechanical shaker at 150 rpm as shown in Figure 4.3. Time of mixing of each tube was0.5, 1, 2, 5 , 24, 48, 72 hr . After that, soil solutions were centrifuged at 7,500 rpm for 30 min to separate the aqueous phase from soil, as shown in Figure 4.4. The AAs in aqueous phase were determined by RP-HPLC. The percentage of adsorption and concentration in aqueous phase were calculated and plotted versus time. Two control samples (spiked AAs but no soil) had been included to check stability and adsorption of AAs on the tube and cap. One blank sample of each soil (no spiked AAs but had soil) was prepared to serve as background and to detect other interference compounds. The control and blank samples also were taken at same time interval. From control sample analysis, it showed AA I and II were not adsorbed on centrifuge tube and cap, the recovery of AA I and II in four soils are over 90%. The blank samples showed no AA I and II contained in all original soils. 80 Figure 4.3 AAs equilibrated with soil solution in mechanical shaker Figure 4.4 Separation aqueous phase from soil after centrifuge 81 4.3.4.2 Adsorption Isotherm The soil sorption isotherm of four soils was determined. The method is generally similar to the kinetic study. Different volume of AAs stock solution was added to make seven different initial concentrations ranging from 0.05 -2 g/ml whereas amount of soil was keep constant (soil 0.1 g in solution 10 ml). Soil solutions were equilibrated for 24 hr. The experiments were done at room temperature. pH of adsorption were measured in aqueous phase after take the samples. The AAs concentration in aqueous phase (Ce) was measured by HPLC while the amount of AAs adsorbed to soil (Cs) was calculate from mass balance. 4.3.4.3 Effect of pH To investigate the difference adsorption of neutral and anion molecules, pH of soil solutions were varied. The batch experiments were used same as before but during soil preequilibrating period (first 12 hr), the pH of soil solutions was adjusted in the range of 2 - 11 by adding small amount of 0.1 M HCl or 0.1 M NaOH until the change of target pH was low as possible. After that AAs stock solution was spiked to soil solution. Equilibrate for 24 hr and centrifuge to obtain supernatant. After taken the aqueous samples, the pH of remaining soil solutions were measured as equilibrium pH. 4.3.4.4 Effect of calcium ion To investigate the effect of calcium ion, the CaCl2 at different concentration (0.0005-0.1 M) were used as background electrolyte. No pH adjustment was applied to avoid changing of ionic strength. The sorption was test at one initial concentration (0.4 µg/ml of AA I and 0.5 µg/ml of AA II) for all four soils. The concentration in aqueous phase (Ce) and conconcentration 82 adsorbed in soil (Cs) were determined and calculated for the soil sorption coefficient (Kd) as in Eq. 4.2. These Kd were plotted versus log CaCl2 concentration for each soils. 4.3.4.5 Effect of cations/anions To test the contribution from cation /anion exchange capacity, monovalent cation salt (KCl, NaCl) and divalent cation salt (MgCl2,CaCl2,SrCl2 ) were use as background solution to test cation effect whereas KCl, KSO4 and KNO3 were used as background solution to test anion effect. The Hoytville NF which is highest organic matter and clay content soil was used. The concentration of each background solutions was adjusted to have same ionic strength. The sorption isotherms of each cation/anion type were compared. The measured final equilibrium pH showed that each salt solution gave similar pH, so the obtaioned isotherms were not effect by the different pH. 4.3.5 Desorption Experiment 4.3.5.1 Desorption kinetic The desorption experiment was done by decant and refill technique immediately after sorption equilibrium. The desorption kinetic study was conducted to determine the contact time to reach desorption equilibrium of each soil. First, sets of soil samples were prepared same as adsorption experiment. After mixing for 24 hr, soil solutions were centrifuged at 7,500 rpm for 30 min and 5 ml of supernatant were removed. The removed aqueous phase were replaced by fresh CaCl2 0.01 M in same volume. This soil solutions were shaken again and the aliquot of samples were taken after desorption at interval of time 0.5, 1, 2, 5, 24, 48, 72 hr. Concentration of AAs desorbed in aqueous phase was measured and % desorption was plotted versus time to determine time to reach desorption equilibrium. 83 4.3.5.2 Desorption isotherm Desorption isotherm were determined by sequential decant and refill technique. The desorption kinetic study showed that 24 hr shaking period for each step was enough to reach desorption equilibrium. First, the sets of soils solution were prepared same as adsorption isotherm experiment. After equilibrium of adsorption step, the three different initial concentration of AAs were chosen to study desorption isotherm. After centrifuge, 5 ml of supernatant were removed and replacing with 5 ml of fresh 0.01 M CaCl2 solution and shaking for another 24 hr. This step was repeated for three times in a roll. The desorbed AAs in aqueous phase (Ce) and concentration remaining in soil (Cs) of each cycle were determined. The desorption isotherms were plotted between the obtained Ce and Cs of each desorption cycle. These isotherms were fitted with Freundlich model. The adsorption –desorption hysteresis were quantified by hysteresis index (HI) calculated from Eq. 4.10. 4.3.5.3 Effect of pH The adsorption- desorption were tested at 3 different pH, one of original soil pH and other two adjusted pH. The pH of soil solution were adjusted by spiking very small amount of 1M HCl or NaOH to solution in pre-equilibration in adsorption step. The desorption steps were conducted as three times in a roll same as mention before but no pH adjustment was applied in desorption step. The pH of aqueous phase after taking the sample of each cycle was measured as pH of desorption. The desorption isotherms of each pH were constructed. The hysteresis indexs were calculated and plotted with the pH of adsorption. 84 4.3.5.4 Effect of cations/anions After the cation/anion effect adsorption experiment, the soil solutions were continued to study desorption step where KCl, NaCl, MgCl2, CaCl2, SrCl2 (for cation experiment) and KCl, KSO4 and KNO3 (for anion experiment) were used as fresh replacement solution. The desorption isotherm of each cation/anion were created. The experiment show that the pH of desorption of each cation/anion solution did not deviate much from pH of adsorption, so the desorption isotherms should not be effect by pH. The hysteresis index were calculated and plotted with cation or anion types. 4.4 Result and discussion 4.4.1 AAs analytical method validation and matrix effect The calibration curve of six concentrations of AA I and II ranging from 0.045 -0.9 and 0.055-1.1 µg/ml, were prepared in soil matrix solution and were analyzed by HPLC-DAD. The area of each peak and their concentrations are plotted and showed in Figure 4.5. The linear regression equation of AA I is y=113,345x-5,306.5 and AA II is y = 115,184x – 6,386.6. Both of them show good linearity correlation (r2>0.99). 85 Peak Area 140000 120000 AA I: y = 113345x - 5306.5 R² = 0.9952 100000 AA II: y = 115184x - 6386.6 R² = 0.9948 80000 60000 40000 AAII 20000 AAI 0 0 0.2 0.4 0.6 0.8 concentration (µg/ml) 1 1.2 Figure 4.5 The calibration curve of AA I and II in soil matrix solution The accuracy and precision of this method were reported as % recovery and % relative standard deviation (%RSD) as shown in Table 4.2. The result showed that the precision of intraday and interday are similar. However, the analysis at low concentration gave less accuracy and precision than analysis at high concentration because of closing to the detection limit. Table 4.2 % recovery (accuracy) and % RSD (intra-day and inter-day precision) of AAs analytical method Chemical Concentration AA I AA II %recovery %RSD %RSD (µg/ml) (mean, n=6) (intra-day, n=6)a (inter-day, n=6)b 0.045 147.59 22.95 23.89 0.45 88.97 4.39 8.24 0.9 99.28 3.88 3.25 0.055 169.21 28.82 22.50 0.55 88.36 6.48 6.08 1.1 98.06 3.04 3.79 a The samples were analyzed 6 times in 1 day b The samples were analyzed 6 times over 3 three consecutive days 86 The linearity range and correlation coefficient are presented in Table 4.3. The result showed that the linear range of analysis with HPLC-DAD can extend over two orders of magnitude of concentration. Limit of detection and quantification were defined as the minimum concentration which gave signal to noise ratio (S/N) equal three and ten, respectively. The values were also shown in Table 4.3. The obtained LOD of AA I and II from this study were 0.013 and 0.010 µg/ml, respectively. These detection limits of analysis were according to literature at 0.0109 µg/ml for AA I and 0.0148 ug/ml for AA II (Zhang et al., 2006). To determine soil matrix effect, the standard solution which prepared in methanol was analyzed same as standard solution prepared in soil matrix solution. The calibration curve of both standards were compared together. Figure 4.6 show that both calibration curve of soil matrix and pure methanol were close together which indicated no interference from soil matrix to analysis by HPLC-DAD. Table 4.3 Linear ranges, correlation coefficient, quantification and detection limit of Aristolochic Acid I and II Chemical Linear range r2 (µg/ml) LOD LOQ (µg/ml) (µg/ml) Aristolochic Acid I 0.045-0.9 0.9952 0.013 0.042 Aristolochic Acid II 0.055-1.1 0.9948 0.010 0.034 87 a) AA I 200000 soil matrix 180000 pure methanol 160000 Peak area 140000 120000 100000 80000 60000 40000 20000 0 0 200000 0.5 1 1.5 concentration (µg/ml) 2 b) AA II soil matrix 180000 pure methanol 160000 Peak area 140000 120000 100000 80000 60000 40000 20000 0 0 0.5 1 1.5 concentration (µg/ml) 2 2.5 Figure 4.6 Matrix effect of analysis AA I (a) and AA II (b) in soil matrix and pure methanol 4.4.2 Adsorption experiment 4.4.2.1 Adsorption kinetic The preliminary kinetic studies showed the sorption of AAs to all soils were rapid process where more than 90% of sorption capacity was reached for less than 24 hr of mixing as shown in Figure 4.7. This is according to a simple two-site sorption model , where the sorption 88 take place rapidly on surface of sorbent called “ apparent equilibrium” and the slow process when molecules diffuse into micropore of organic matter (J. P. DiVincenzo & Sparks, 2001; Piwowarczyk & Holden, 2012). However, the slow sorption process was not included in this study. a) AA I 100.00 WT CT 90.00 WT NF HY CT HY NF 80.00 % Adsorption 70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 0 12 24 36 48 Time (hrs) 60 72 84 b) AA II 100.00 WT CT 90.00 WT NF HY CT HY NF 80.00 % Adsorption 70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 0 12 24 36 48 Time (hrs) Figure 4.7 Sorption kinetic of AA I (a) and II (b) on four soils 89 60 72 84 4.4.2.2 Adsorption Isotherm The concentration of AAs remained in aqueous phase (Ce) and AAs sorbed in soil (Cs) were plotted as adsorption isotherm shown in Figure 4.8. The isotherms were fitted well with linear and freundlich isotherm with r2>0.9 as shown in Table 4.4 and the Kd, Kf and 1/n were obtained by fitting the data to the model. The linear isotherm indicated that sorption concentration of AAs were lower the sorption capacity of the soils. The different Kd indicated different sorption capacity of each soils and may relate to soil composition (R.P. Schwarzenbach et al., 2003; W. C. Yang, Mang, Zhang, Zhu, & Chen, 2009). Even AAs were in anion form at this native pH, the log Koc calculated from Eq.4.3 were high in the range of 3.3-3.5. This number contradicted with log Kow of AAs at natural pH which was very low. This indicates that other sorption mechanisms may involve other than simple hydrophobic interaction. From the point of view of AAs structure, we can assume that the phenanthrene group is responsible for hydrophobic interaction with organic matter whereas the carboxylic acid and nitro group are attribute to other specific interaction such as the H-bonding or ion exchange (Senesi, 1992; Westall et al., 1999). Moreover, the Ca2+ presenting in solution may enhance the adsorption by adsorb to negative charged sites in soil and make them less negative, thereby reduce electrostatic repulsion or even forming the complex with AAs by cation bridging mechanism which are general patterns for organic acids (Jafvert et al., 1990). 90 a) AA I 180 Concentration adsorbed in soil (µg/g) WT CT, pH 6.45 y = 228.39x 160 WT NF, pH 4.65 140 HY CT, pH 6.37 120 y = 79.212x HY NF, pH 6.35 100 y = 46.394x 80 60 y = 18.727x 40 20 0 0.00 0.50 1.00 1.50 2.00 Concentration in solution (µg/ml) 2.50 b) AA II 180 Concentration adsorbed in soil (µg/g) WT CT, pH 6.45 160 y = 148.51x WT NF, pH 4.65 140 HY CT, pH 6.37 120 HY NF, pH 6.35 100 y = 67.571x 80 60 y = 23.127x 40 20 0 0.00 y = 10.434x 0.50 1.00 1.50 2.00 Concentration in solution (µg/ml) 2.50 Figure 4.8 Sorption isotherm of AA I (a) and AA II (b) in all four soils at native soil pH equilibrium 91 Table 4.4 Linear and Freundlich isotherm parameters for sorption of AAI to four soils at natural soil pH Soil pH Linear isotherm Kd (ml/g) r2 Freundlich isotherm Log Koc Kf (µg1-1/n g-1)(ml1/n) 1/n r2 Wooster CT 6.45 18.73 0.93 3.32 18.23 1.07 0.96 Wooster NF 4.65 46.39 0.99 3.22 46.01 1.04 0.99 Hoytville CT 6.37 79.21 0.99 3.54 79.19 0.98 0.99 6.35 228.39 0.99 3.48 220.07 0.94 0.99 Hoytville NF However, to compare the sorption of organic acid where pH is dominant factor, the similar adjusted pH would be more accurate. The adjusted pH sorption isotherms were shown in Figure 4.9. It was observed that, after adjusted the pH, the sorption of WT NF was lower than WT CT even it had higher organic carbon content. Instead, it was positively correlated to the clay content. This indicated that, for low organic carbon content soil e.g. Wooster soil, the AAs has higher affinity to clay rather than organic matter in soil. 92 a) AA I 180 y = 181.56x WT CT, pH 6.13 Concentration adsorbed in soil (µg/g) 160 WT NF, pH 6.13 140 HY CT, pH 6.25 HY NF, pH 6.29 120 y = 60.529x 100 80 60 y = 17.205x 40 y = 11.123x 20 0 0 0.5 1 1.5 2 2.5 3 Concentration in solution (µg/ml) b) AA II Concentration adsorbed in soil (µg/g) 200 WT CT, pH 6.13 y = 137.93x 180 WT NF, pH 6.13 160 HY CT, pH 6.25 HY NF, pH 6.29 140 y = 60.724x 120 100 80 60 40 y = 9.0107x 20 y = 6.8272x 0 0 0.5 1 1.5 2 2.5 Concentration in solution (µg/ml) 3 3.5 Figure 4.9 Adsorption isotherm of AA I (a) and AA II (b) in all four soils at adjusted pH equilibrium 93 4.4.2.3 Effect of soil properties To determine the effect soil properties, the obtained soil partition coefficient (Kd) of each soil were plotted as a function of %organic carbaon content and cation exchange capacity (CEC) as shown in Figure 4.10 (a) and (b). The results show that Kd had low correlation with % organic carbon but instead, it had better correlation with cation exchange capacity (CEC). These finding indicate that AAs sorption mechanism is not only simple hydrophobic partitioning but rather related to cation exchange mechanism. a) % organic carbon soil sorption coefficient (Kd) (ml/g) 200 180 AA I 160 AA II 140 120 100 80 60 40 20 0 0 2 4 % organic carbon 6 8 Figure 4.10 Soil sorption coefficient (Kd) as a function of % organic carbon (a) and cation exchange capacity (b) of soil 94 Figure 4.10 (cont’d) b) cation exchange capacity soil sorption coefficient (Kd) (ml/g) 200 180 AA I 160 AA II 140 120 100 80 60 40 20 0 0 5 10 15 CEC (cmol/kg) 4.4.2.4 Effect of pH The sorption coefficient (Kd) of all soil were plotted with pH as shown in Figure 4.11. The figures show that AAs adsorption was strongly depended on the pH of soil solution. The shape of Kd with pH were sigmoidal curve for all soils where the Kd found to decrease 20-100 times when the pH increased from 2 to 6 but the effect may differ between soil. These data were fitted well with apparent sorption coefficient ( ) equation considering the contribution from neutral and anion forms of AAs (Eq.4.8). The ionized molecules at high pH were much less adsorbed than neutral molecules at low pH. This may be explained that the neutral molecules were partition in organic matter in soil very well. However, anion molecules which had largely less hydrophobicity were less partition to organic matter. Also, at high pH, the surface of soils had negative charge which will repulse with anion of AAs (Hyun et al., 2003). These evidence may imply that AAs adsorb to the soil by hydrophobic partitioning at low pH but by ion exchange at high pH. 95 a) Wooster soil 3000 WT CT experiment WT NF experiment 2500 soil sorption coefficient (Kd) (ml/g) WT CT model WT NF model 2000 1500 1000 500 0 0 2 4 6 8 10 pH b) Hoytville soil 3000 HY CT experiment soil sorption coefficient (Kd) (ml/g) HY NF experiment 2500 HY CT model HY NF model 2000 1500 1000 500 0 0 2 4 pH 6 8 10 Figure 4.11 Soil sorption coefficient as a function of pH for Wooster soils (a) and Hoytville soils (b) 96 4.4.2.5 Effect of calcium ion The influence of ions presented in solution on adsorption was examined. The single point adsorption at fixed initial concentration was determined in CaCl2 background solution from 0.0005 – 0.1 M. The soil sorption coefficient (Kd) were calculated and plotted with log CaCl2 concentration for all soil as shown in Figure 4.12 (a)-(d). The figures show that Ca2+ enhanced the adsorption of AAs in only Hoytville NF soil which is highest CEC soil but there were no effect observed on other lower CEC soils. This indicated that cation exchange was important mechanism. The Ca2+ enhanced adsorption can be explained by two possible mechanisms; first ,Ca2+ which is highly exchange capacity cation were adsorbed at negative charged sites, so the electrostatic repulsion was reduced or the surface complexation between anion AAs and Ca2+ via cation bridging mechanism was possible (Jafvert, 1990). Second, the complex formation between anion AAs and Ca2+ in solution may turned the molecules to positive charge and they favored to be adsorbed on the cation exchange sites (Westall et al., 1999). 97 b) WT NF soil 300 300 7.5 7.5 Kd 200 7 250 6.5 200 6 150 5.5 100 Kd (ml/g) pH pH 6.5 6 150 5.5 100 5 5 50 50 4.5 4.5 0 0.0001 0 0.0001 4 0.001 0.01 0.1 Log CaCl2 conc.(M) 1 c) HY CT soil 7.5 250 7 1 300 7.5 250 7 Kd 150 6 pH 5.5 100 5 50 4.5 6.5 200 6 150 5.5 100 5 Kd 50 4.5 pH 4 0.001 0.01 0.1 Log CaCl2 Conc.(M) Kd (ml/g) 6.5 200 0 0.0001 4 0.001 0.01 0.1 Log CaCl2 Conc.(M) d) HY NF soil 300 pH Kd (ml/g) 7 pH 0 0.0001 1 4 0.001 0.01 0.1 Log CaCl2 Conc.(M) Figure 4.12 The sorption coefficient (Kd) and pH (second axis) as a function log CaCl2 concentration on WT CT soil (a), WT NF soil (b), HY CT soil(c) and HY NF soil (d) 98 1 pH Kd (ml/g) 250 Kd pH a) WT CT soil 4.4.2.6 Effect of Cations/ Anions To test the effect of cation types on the adsorption of AAs, the salt solution of CaCl2, MgCl2, SrCl2, KCl, NaCl and also deionized water were used as background solution and sorption isotherm of each salt were determined as shown in Figure 4.13. The results show that the presence of divalent cation will enhance the adsorption more than monovalent cation e.g. SrCl2 CaCl2  MgCl2 > KCl  NaCl > DI. This order accords to the power of replaceability of cations where the higher valent cation will have more replacing power e.g. Sr2+ > Ca2+ > Mg2+ > K+ > Na+ > Li+ (Carroll, 1959; R. A. Figueroa et al., 2004). This sorption enhancement by power of replacibility of cation suggested the cation –bridging mechanism because divalent cations can form stronger bond to cation exchange site than monovalent cations (Hyun & Lee, 2005). a) AA I 45.00 40.00 35.00 CaCl2 MgCl2 SrCl2 Cs(µg/g) 30.00 KCl 25.00 20.00 NaCl DI 15.00 10.00 5.00 0.00 0.000 0.050 0.100 0.150 0.200 Ce(µg/ml) 0.250 0.300 0.350 Figure 4.13 Adsorption isotherm of AA I (a) and II (b) on Hotyville NF soil in different cations background solution 99 Figure 4.13 (cont’d) b) AA II 45.00 CaCl2 40.00 MgCl2 35.00 SrCl2 Cs(µg/g) 30.00 KCl 25.00 NaCl 20.00 DI 15.00 10.00 5.00 0.00 0.000 0.100 0.200 0.300 0.400 Ce(µg/ml) 0.500 0.600 0.700 The effect of anions on the adsorption of AAs was also tested and the results showed in Figure 4.14.This figures show that the adsorption of AAs in KCl as background solution were similar to KNO3 and K2SO4. The non-difference in AAs sorption from different anion type even Cl- , SO42-, NO3-, have different affinity to anion exchange capacity sites. This indicated that adsorption of AAs were not effect by anions exchange mechanisms. Also, because of the tested soil at high pH has low AEC, so the contribution from anion exchange was limited. 100 a) AA I 45.00 KCl 40.00 KNO3 35.00 K2SO4 Cs(µg/g) 30.00 DI 25.00 20.00 15.00 10.00 5.00 0.00 0.000 0.050 0.100 0.150 0.200 Ce(µg/ml) 0.250 0.300 b) AA II 45.00 KCl Cs(µg/g) 40.00 KNO3 35.00 K2SO4 30.00 DI 25.00 20.00 15.00 10.00 5.00 0.00 0.000 0.100 0.200 0.300 0.400 Ce(µg/ml) 0.500 0.600 Figure 4.14 Adsorption isotherm of AA I (a) and II (b) on Hotyville NF soil in different anion background solution. 101 4.4.3 Desorption experiment 4.4.3.1 Desorption kinetic Figure 4.15 show % desorption of AA I and II from all soils as function of time. The figures show that the desorption of AAs were quite rapid where nearly 90% of desorbed amount released in first 2 hr. The desorption equilibrium can be reach about 24 hr for Wooster CT and NF soils. However, the desorption for Hoytville CT and NF were slower and were reach to equilibrium about 48 hr. The rapid desorption can be explained that mostly sorbed AAs were attached on the surface sites and not diffused to stronger adsorption sites in the organic matter matrices.The resuls show that the adsorbed AAs on Wooster soil can be desorb more than Hoytville soil which may due to the higher clay and organic matter content of Hoytville soil. This supports the idea that soils which give less adsorb will have greater tendency to desorb (Piwowarczyk & Holden, 2012). . 102 a) AA I 100.00 WT CT 90.00 WT NF HY CT HY NF 80.00 %desorption 70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 0 12 24 36 48 time (hrs) 60 72 84 b) AA II 100.00 WT CT 90.00 WT NF HY CT HY NF 80.00 %desorption 70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 0 12 24 36 48 time (hr) 60 72 84 Figure 4.15 % Desorption of AAI (a) and AA II (b) as function of time for all tested soils 103 4.4.3.2 Desorption isotherm To determine the desorption isotherm, the supernatant was remove and replaced by fresh CaCl2 background solution. This technique will reduce AAs in solution and the adsorbed AAs will be released to reach the equilibrium again. The concentration remaining in solution and sorbed in soil were plotted as desorption isotherm as shown in Figure 4.16. It show that most of desorption isotherms were fitted well with Freundlich equation (r2> 0.9) as shown in Table 4.5. The figures show that the sorption and desorption isotherm of all tested soils were not much deviated. The hysteresis index (HI) which calculate from the ratio of exponent of desorption isotherm to sorption isotherm (ndes/nads) according to Eq. 4.11 were shown in Table 4.5. The HI values of all soils were nearly to100 at low initial concentration which indicated the sorption of AAs to soils is largely reversible. However, at higher concentration, the HI found to decrease where the hysteresis increase. The easy desorption of AAs may cause by the short contact time where AAs cannot get deep inside into organic matter matrices (Lesan & Bhandari, 2003) 104 b) WT CT AA II, original pH 6.39 a) WT CT, AA I, original pH 6.39 80.00 120.00 adsorption desorption 0.7 µg/ml desorption 1.1 µg/ml desorption 1.4 µg/ml 100.00 70.00 60.00 Cs(µg/g) Cs(µg/g) 80.00 adsorption Desorption 0.7 µg/ml desorption 1.2 µg/ml Desorption 1.5 µg/ml 60.00 50.00 40.00 30.00 40.00 20.00 20.00 10.00 0.00 0.000 0.500 1.000 Ce(µg/ml) 1.500 0.00 0.000 2.000 1.000 1.500 2.000 Ce(µg/ml) c) WT NF AA I, original pH 4.57 d) WT NF AA II, original pH 4.57 80.00 120.00 sorption AA I 70.00 desorption 0.7 µg/ml 100.00 60.00 desorption 1.2 µg/ml 80.00 desorption 1.4 µg/ml Cs(µg/g) Cs(µg/g) 0.500 60.00 50.00 desorption 1 µg/ml desorption 1.7 µg/ml desorption 2.1 µg/ml 40.00 30.00 40.00 20.00 20.00 0.00 0.000 10.00 0.500 1.000 1.500 0.00 0.000 2.000 0.500 1.000 1.500 2.000 Ce(µg/ml) Ce(µg/ml) Figure 4.16 Sorption and desorption isotherm of AA I and II in WT CT soil (a and b), WT NF soil (c and d), HY CT soil (e and f) and HY NF soil (g and h) (filled symbols-adsorption point, open symbols-desorption point) 105 Figure 4.16 (cont’d) f) HY CT, AA II, original pH 6.66 e) HY CT, AA I, original pH 6.66 80.00 120.00 Adsorption Adsorption 100.00 70.00 Desorption 1 µg/ml Desorption 1.6 µg/ml Desorption 2 µg/ml Cs(µg/g) Cs(µg/g) 80.00 60.00 40.00 desorption 1.6 µg/ml 50.00 desorption 2 µg/ml 40.00 30.00 20.00 20.00 0.00 0.000 Desorption 1 µg/ml 60.00 10.00 0.500 1.000 1.500 0.00 0.000 2.000 0.500 1.000 1.500 2.000 Ce(µg/ml) Ce(µg/ml) h) HY NF, AA II, original pH 6.56 g) HY NF, AA I, original pH 6.56 120.00 80.00 70.00 100.00 60.00 Cs(ug/g) Cs(µg/g) 80.00 60.00 40.00 30.00 40.00 Adsorption Desorption 1 µg/ml desorption 1.6 µg/ml desorption 2 µg/ml Adsorption 20.00 0.00 0.000 50.00 Desorption 1 µg/ml 20.00 Desorption 1.6 µg/ml 10.00 Desorption 2 µg/ml 0.500 1.000 1.500 2.000 Ce(µg/ml) 106 0.00 0.000 0.500 1.000 Ce(µg/ml) 1.500 2.000 Table 4.5 Desorption coefficient (kdes) and 1/ndes obtained from fitting the model and hysteresis index (HI= ) for sorption of AA I and II in all tested soils Soil AA I Initial AA II r2 Kdes HI concentration WT CT WT NF HY CT HY NF Initial Kdes r2 HI concentration 0.7 g/ml 49.41 1.37 0.97 133 0.7 g/ml 23.26 1.29 0.97 126 1.1 g/ml 39.36 0.6 0.99 58 1.2 g/ml 22.89 0.53 0.99 52 1.4 g/ml 35.84 0.61 0.98 58 1.5 g/ml 20.81 0.75 0.90 73 0.7 g/ml 52.19 1.55 0.85 111 1 g/ml 16.99 0.78 0.94 52 1.2 g/ml 55.06 1.46 0.97 105 1.7 g/ml 26.03 0.87 0.98 58 1.4 g/ml 47.10 0.72 0.92 52 2.1 g/ml 31.88 0.45 0.99 30 1 g/ml 37.34 1.17 0.92 125 1 g/ml 20.28 1.16 0.98 116 1.6 g/ml 38.54 0.53 0.98 57 1.6 g/ml 21.55 0.46 0.99 46 2 g/ml 33.25 0.84 0.94 91 2 g/ml 21.24 0.51 0.98 50 1 g/ml 106.93 0.65 0.99 82 1 g/ml 71.35 0.77 0.99 98 1.6 g/ml 99.21 0.46 0.99 58 1.6 g/ml 61.67 0.64 0.99 81 2 g/ml 102.72 0.58 0.92 74 2 g/ml 64.67 0.60 0.99 76 4.4.3.3 Effect of initial concentration The effect of different initial concentration to desorption hysteresis of each soil are shown in Figure 4.17. The figures show that hysteresis index (HI) had been decrease when concentration increased which indicated the desorption was more difficult at higher concentration. This can be explained that, at low concentration, AAs had low concentration gradient where the few molecules could penetrate to organic matter matrix. However, at high concentration, the AAs had more driving force to go deeper sites in the soil, and they can form 107 the stronger bond such as H-bonding with organic matter which caused them less to desorb (Chefetz et al., 2004). a) AA I 140 WT CT 120 WT NF HY CT Hysteresis index 100 HY NF 80 60 40 20 0 0 0.5 1 1.5 concentration (µg/ml) 2 2.5 1 1.5 concentration (µg/ml) 2 2.5 b) AA II 140 Hysteresis index 120 100 80 60 WT CT 40 WT NF HY CT 20 HY NF 0 0 0.5 Figure 4.17 The hysteresis index of AA I (a) and II (b) as a function of initial concentration for all tested soils 108 4.4.3.4 Effect of pH To investigate the effect of pH on desorption process, the desorption at different pH were tested with all soils. Because similar trend was found, the only isotherms of Hoytville soil are presented here. From Figure 4.18, the sorption and desorption isotherm at low pH are quite similar. However, when the pH increase, the desorption increasingly deviated from adsorption isotherm. Figure 4.19 plots the calculated hysteresis index (HI) of desorption isotherm of all soil at different pH. The figures show the HI values were gradually decreased when the pH increased. These evidence showed the different desorption mechanism between neutral and anion form of AAs. The desorption was readily when AAs were in neutral form but the desorption will be harder when AAs were in anion form at high pH. The anion molecules can form specific interaction e.g. cation bridging mechanism to soil which was quite strong and therefore resist to desorption (Caceres et al., 2010; Ren et al., 2011). 109 a) HY NF, AA I, adjust pH 6.00 b) HY NF, AA II, adjust pH 6.00 120.00 90.00 80.00 100.00 70.00 60.00 Cs(µg/g) Cs(µg/g) 80.00 60.00 Adsorption 40.00 20.00 0.00 0.000 50.00 40.00 Desorption 0.8 µg/ml 30.00 Desorption 1.3 µg/ml 20.00 Desorption 1.6 µg/ml 10.00 0.500 1.000 0.00 0.000 1.500 Adsorption Desorption 1.0 µg/ml desorption 1.6 µg/ml desorption 1.9 µg/ml 0.500 1.000 1.500 Ce(µg/ml) Ce(µg/ml) c) HY NF, AA I, original pH 6.56 d) HY NF, AA II, original pH 6.56 120.00 90.00 80.00 100.00 70.00 60.00 Cs(µg/g) Cs(µg/g) 80.00 60.00 50.00 40.00 Adsorption Adsorption 40.00 30.00 Desorption 1 µg/ml 20.00 desorption 1.6 µg/ml 10.00 desorption 2 µg/ml Desorption 1 µg/ml Desorption 1.6 µg/ml 20.00 Desorption 2 µg/ml 0.00 0.000 0.500 1.000 1.500 Ce(µg/ml) 0.00 0.000 0.500 1.000 1.500 Ce(µg/ml) Figure 4.18 AAs sorption –desorption isotherm of AA I and II on Hoytville NF soil at adjusted pH 6.0 (a) and (b), original pH 6.56 (c) and (d), adjusted pH8.36 (e) and (f). The pH shown in figure was from an average of pH of adsorption process. 110 Figure 4.18 (cont’d) e) HY NF, AA I, adjust pH 8.36 120.00 100.00 f) HY NF, AA II , adjust pH 8.36 Adsorption 90.00 Desorption 0.9 µg/ml 80.00 Adsorption Desorption 1.1 µg/ml 70.00 desorption 1.1 µg/ml 60.00 desorption 1.9 µg/ml Desorption 1.8 µg/ml Cs(µg/g) Cs(µg/g) 80.00 60.00 40.00 Desorption 0.9 µg/ml 50.00 40.00 30.00 20.00 20.00 10.00 0.00 0.000 0.500 1.000 1.500 0.00 0.000 0.500 1.000 Ce(µg/ml) Ce(µg/ml) 111 1.500 a) AA I 140 WT CT 120 WT NF HY CT Hysteresis Index 100 HY NF 80 60 40 20 0 0 2 4 6 8 10 6 8 10 pH b) AA II 140 WT CT 120 WT NF HY CT 100 HY NF HI 80 60 40 20 0 0 2 4 pH Figure 4.19 Hysteresis index calculated form sortion- desorption isotherm of all soil as a function with pH of AA I (a) and AA II (b) 112 4.4.3.5 Effect of cations/anions The effect of cation type to desorption process was tested on Hoytville NF soil. The desorption isotherm of AA I with different cation background solution had been created as shown in Figure 4.20. The hysteresis indexs were calculated and plotted with cations types which showed in Figure 4.21. Divalent cation (Mg2+, Ca2+ and Sr2+) found to have more hysteresis than monovalent cations (Na+ and K+). This indicated the adsorption was largely reversible when AAs dissolved in deionized water or monovalent cations solution, whereas the irreversibility or hysteresis increased when AAs dissolved in divalent cations solution. The reversibiltiy of sorption in monovalent cation solution suggests non-specific cation exchange sorption reaction were dominant whereas higher hysteresis in divalent cation solution indicated stronger specific binding mechanisms e.g. cation bridging were dominant. The hysteresis increase in order of Na2.3.co;2 Ren, W. J., Wang, M. E., & Zhou, Q. X. (2011). Effect of soil pH and organic matter on desorption hysteresis of chlorimuron-ethyl in two typical Chinese soils. Journal of Soils and Sediments, 11(4), 552-561. doi: 10.1007/s11368-011-0337-4 124 Sassman, S. A., & Lee, L. S. (2007). Sorption and degradation in soils of veterinary ionophore antibiotics: monensin and lasalocid. Environmental Toxicology and Chemistry, 26(8), 1614-1621. doi: 10.1897/07-073r.1 Schwarzenbach, R. P., Gschwend, P. M., & Imboden, D. M. (2003). Environmental Organic Chemistry (2nd Ed.). New Jersey: John Wiley & Sons, Inc. Senesi, N. (1992). BINDING MECHANISMS OF PESTICIDES TO SOIL HUMIC SUBSTANCES. Science of the Total Environment, 123, 63-76. doi: 10.1016/00489697(92)90133-d ter Laak, T. L., Gebbink, W. A., & Tolls, J. (2006). The effect of pH and ionic strength on the sorption of sulfachloropyridazine, tylosin, and oxytetracycline to soil. Environmental Toxicology and Chemistry, 25(4), 904-911. doi: 10.1897/05-232r.1 Tolls, J. (2001). Sorption of veterinary pharmaceuticals in soils: A review. Environmental Science & Technology, 35(17), 3397-3406. doi: 10.1021/es0003021 Vonoepen, B., Kordel, W., & Klein, W. (1991). SORPTION OF NONPOLAR AND POLAR COMPOUNDS TO SOILS - PROCESSES, MEASUREMENTS AND EXPERIENCE WITH THE APPLICABILITY OF THE MODIFIED OECD-GUIDELINE-106. Chemosphere, 22(3-4), 285-304. doi: 10.1016/0045-6535(91)90318-8 Westall, J. C., Chen, H., Zhang, W. J., & Brownawell, B. J. (1999). Sorption of linear alkylbenzenesulfonates on sediment materials. Environmental Science & Technology, 33(18), 3110-3118. doi: 10.1021/es9804316 Yang, W. C., Mang, J., Zhang, C. D., Zhu, L. Y., & Chen, W. (2009). Sorption and Resistant Desorption of Atrazine in Typical Chinese Soils. Journal of Environmental Quality, 38(1), 171-179. doi: 10.2134/jeq2007.0674 Yuan, G. S., & Xing, B. S. (2001). Effects of metal cations on sorption and desorption of organic compounds in humic acids. Soil Science, 166(2), 107-115. doi: 10.1097/00010694200102000-00004 Yuan, J. B., Liu, Q., Zhu, W. F., Ding, L., Tang, F., & Yao, S. Z. (2008). Simultaneous analysis of six aristolochic acids and five aristolactams in herbal plants and their preparations by high-performance liquid chromatography-diode array detection-fluorescence detection. Journal of Chromatography A, 1182(1), 85-92. doi: 10.1016/j.chroma.2007.12.076 Zelazny, L. W., He, L., & Vanwormhoudt, A. (1996). Charge analysis of soils and anion exchange (D. L. e. a. Sparks Ed.). Madison, WI: SSSA. 125 Zhang, C. Y., Wang, X., Shang, M. Y., Yu, J., Xu, Y. Q., Li, Z. G., . . . Namba, T. (2006). Simultaneous determination of five aristolochic acids and two aristololactams in Aristolochia plants by high-performance liquid chromatography. Biomedical Chromatography, 20(4), 309-318. doi: 10.1002/bmc.565 126 Chapter 5 Plant uptake 5.1 Introduction Uptake of organic chemicals by plants grown in contaminated area is an important process to access the human health risk by food contamination. Plants can take many of organic chemicals such as chlorinated solvents, pesticides, PAHs, PCBs and pharmaceuticals from soil that contained these chemicals or was irrigated with contaminated wastewater (Aslund, Rutter, Reimer, & Zeeb, 2008; Kipopoulou, Manoli, & Samara, 1999; Shenker, Harush, Ben-Ari, & Chefetz, 2011). These chemicals will be taken up directly from root and translocated to aboveground tissue e.g.stems, leaves and fruits. Some edible vegetables and fruits have been detected these chemicals over the safe exposure limit. For example, the diedrin and endrin which are extremely persistent pesticides had been detected in cucumber fruits that grown in contaminated sites exceeding the concentration set by the food sanitation law (Hashimoto, 2005). Therefore, the plant uptake process is important for accumulation of organic chemicals in environment by increasing concentration of the chemicals in food chain. The pollutant chemicals can be transfered to plant tissues in various extent which depend on concentration in the medias, physicochemical properties of contaminants (hydrophobicity, water solubility), plant species, exposure time, soil properties (pH , organic matter content, cation exchange capacity) and the interaction between contaminants and plant composition (Chiou, Sheng, & Manes, 2001; Su, Zhu, & Liang, 2009). The pollutant organic chemicals can be taken by plant roots in proportion to amount of water transpired via root tips and root hairs. 127 Chemicalsenter into the plants by two different processes, first is passive process which is diffusion across membrane by concentration gradient and second is active process where chemicals incorporate to nutrient uptake and energy consuming processes (Su, Zhu, & Du, 2005; Trapp, 2004). After chemicals get into the root, they will transport via 2 vessel system, xylem and phloem where chemical can move from root to shoot or shoot to root follow concentration gradient (Hellstrom, 2004; Su, Liu, & Liang, 2010) . The study of distribution of chemicals into the plants mostly will be described by root concentration factor (RCF) and transpiration stream concentration factor (TSCF) which is define as concentration accumulated in root tissue or concentration in transpiration stream divide by concentration in external media concentration, respectively (Briggs et al., 1983; Su et al., 2010). The obtained values of RCF and TSCF are useful to predict the contamination in the crops. They evaluate the capacity of chemical that can adsorp on the plant root and transpiration across the root membrane to xylem and translocation to other parts e.g. stems and leaves. To evaluate the AAs potential exposure pathway via food crop contamination, the determination the plant uptake and accumulation capacity is important. This study will provide the basic plant-AAs uptake data. It will addressed the potential of AAs enter to plant through the root and translocate and accumulate in the shoots in hydroponic and sand culture system. 5.2 Literature reviews 5.2.1 Root uptake mechanism and pathway Organic chemicals can move from external solution into the plant by transpiration process with water. The chemical transfer process can divide into passive and active mechanism. For the passive process, organic chemicals will diffuse from root pass through bundle of root 128 cells until reach the xylem column driving by concentration gradient. Typically, there are two pathways for transportation organic chemicals in root cells, the symplastic and apoplastic. In symplastic, the water and chemicals will move across the plasma membrane to each root cell and endodermis until they get into the xylem whereas, for apoplastic, the water and chemicals will move between the root cell membrane until reaching the xylem as shown in Figure 5.1 The transport of compounds may be combination between these two pathways (E. Dettenmaier, 2008; Steudle & Peterson, 1998). Most organic compounds move by passive process and the uptake rate is controlled by diffusion across these lipid membrane and concentration gradient as transport driving energy. Figure 5.1 The diffusion pathway of symplastic and apoplastic (Dettenmaier, 2008) 129 Unlike the neutral organic compounds, ionized compounds are not likely to passive transport across lipid membrane by concentration driving because of high activation energy required. However, they cross the membrane rather by synergistic with specific protein spanning in the membrane in ADP-ATP proton pumping process. The hydrogen ions are pumping in and out cross the membrane during cell respiration which produces concentration and electrical potential gradient. This charge gradient will drive cationic or anionic species to cross the membrane to obtain the electrical potential balance. This is called “active process” (E. Dettenmaier, 2008) Except the hormone-liked chemicals e.g. 2,4-dichlorophenoxyacetic acid, there are no anthropogenic chemicals taken up by active process. 5.2.2 Translocation, accumulation and metabolism in plant Once the compounds pass through the root cell, they diffuse from cell to cell into the xylem where the compounds can move to other part of the plant by xylem flow (E. Dettenmaier, 2008). In xylem, the water and soluble mineral nutrients will be transported from the root to aerial parts e.g. stems and leaves to replace the water lost in transpiration and photosynthesis processes. The water and solute move in xylem by mass flow or pressure potential induced by transpiration process rather than cell diffusion. In Pholem, water and food e.g. sugar and amino acids produced by photosynthesis will be moved to storage organs e.g. roots, seeds fruit, tubers and bulbs. The organic substances will be transport by diffusion gradient and phloem sap moved by positive hydrostatic pressure. The phloem located on the outer side of vascular bundle whereas the xylem located inside the bundle (Diffen; Hellstrom, 2004). 130 The organic chemicals that were taken up by root will translocate throughout the plant tissues via xylem and phloem vessel. The accumulation may be presented in photosynthesis areas or sites with great transpiration, e.g. mature leaf. Most nonionized compound can partition to the stem xylem tissue which is dominant factor that impede the long distance movement of compounds. The transport of organic compounds in stem is similar to reverse phase column chromatography. They may adsorb to the vessel wall or partitioning to lipophilic component in xylem tissues which also related to the hydrophobicity (Kow) of chemicals (McCrady, McFarlane, & Lindstrom, 1987). The more hydrophobic chemicals are likely to retain by the stem base. In xylem, the compounds can also diffuse to adjacent tissue e.g. phloem but degree of exchange depend on characteristic tissue to the chemicals (Trapp & McFarlane, 1995). Many studies show that some plants are also effective in translocation and accumulation organic and inorganic compounds in the leaf (Cui et al., 2014; Lin, Zhu, He, & Tu, 2006) and in fruit(Hulster, Muller, & Marschner, 1994) as well, even if they were much less than root and stem. The accumulation in leaf will be higher if plant has high transpiration rate (Polder, Hulzebos, & Jager, 1995). Although plant can uptake and accumulate the organic pollutants, live cells plants can metabolic degradation the chemicals that foreign to them e.g. pesticides. Plant can metabolize even persistent chemicals such as the insecticide, DDT or fungicide, hexachlorobenzene (Sandermann, Scheel, & Vandertrenck, 1984). Plant metabolism are similar to liver processes ,by transform the xenobiotic compounds to non-toxic chemical and then form conjugate with glucose or amino acid or even cell wall in the plant. The different between plant and liver metabolism is the conjugates will be excreted in animals but they will be stored or compartmentalized in the plant cells (Hellstrom, 2004; Trapp & McFarlane, 1995). For example, atrazine was metabolited 131 in corn seedling to OH-derivative of atrazine. This transformation was readily after parent compounds enter the plant (Raveton, Ravanel, Serre, Nurit, & Tissut, 1997). Palazzo, 1986 reported the founding TNT metabolites in terrestrial plants expose to TNT in hydroponic solution. Although metabolic fate of organic compounds in plant had been intensively studied, it is still difficult to predict and largely differ among plant species. 5.2.3 Plant uptake descriptor The potential uptake of chemical by plant can quantitatively described by bioconcentration factor (BCF) and transpiration stream concentration factor (TSCF). The bioconcentration factor, similar using in fish, use to determine the ability of chemicals accumulated in plant tissue and simply define by ratio of concentration in plant tissue to concentration in the media that plant growing in e.g. nutrient concentration or soil concentration (Briggs, Bromilow, & Evans, 1982; Y. Z. Gao & Zhu, 2004; Oconnor, Kiehl, Eiceman, & Ryan, 1990; Trapp, 2000). The root concentration factor is used to describe the accumulation of chemical in root as Eq. 5.1: (5.1) Likewise the root concentration factor (RCF), Brigg et al. also defined stem concentration factor (SCF) to describe ability of compounds accumulate in stem which is the ratio of concentration in stem to external solution as defined in Eq. 5.2 (5.2) However, some substances can accumulated in leaf with high concentration more than stem if transpiration rate at leaf is high. So, the stem concentration factor may not a good ratio to 132 represent the concentration in plant. Many studies use the whole aboveground plant tissue e.g. stem and leaves to show ability of translocation from root to shoot. The shoot concentration factor defined as ratio of concentration of chemicals in whole shoot to concentration in the medium as shown in Eq. 5.3 (Y. Z. Gao & Zhu, 2004; Polder et al., 1995) (5.3) The ability of compounds to be passively transport from root to shoot with the transpiration stream can be expressed by transpiration stream concentration factor (TSCF) and defined by the ratio of concentration in xylem to the external solution and expressed as Eq.5.4 (Briggs et al., 1982). However, because the concentration of xylem is difficult to measure directly, it is usually determined from mass of chemical accumulated in the shoot compare to known amount of water transpired (e.g. over 24-48 hr) as shown in Eq. 5.5 (Briggs et al., 1982; Trapp, 2000). This approach assumes that the degradation of chemicals in transpiration stream is neglect and transport back to root by phloem is insignificant. (5.4) (5.5) The TSCF of water is 1, the nutrients that actively taken by plant e.g. nitrogen, potassium and phosphorus have TSCF value greater than 1. The TSCF of most organic pollutants are less than 1 indicated that they are passively move from root to shoot with transpiration water (Orita, 2012). 133 5.2.4 Plant uptake of non-ionic compounds Uptake and translocation of non-ionized chemicals are largely determined by hydrophobicity of chemicals which can describe by octanol water partition coefficient (Kow). The RCF increases with increasing the hydrophobicity of chemicals as shown in Figure 5.2. Briggs et al., 1982 had setup the equation that showed the relation between RCF and log Kow as shown in Eq. 5.6. (5.6) From the incresing uptake of more hydrophobic compounds and the positive relation between RCF and root lipid content suggest the partitioning of hydrophobic compounds in lipophilic root solid is main mechanism (Briggs et al., 1982; Briggs et al., 1983; Y. Z. Gao & Zhu, 2004). This behavior is supported by the similar sorption coefficient of compounds to macerated root to the coefficient from plant uptake experiment which indicates the partition process is most account for observed RCF of compounds (Briggs et al., 1982; de Carvalho, Bromilow, & Greenwood, 2007b) Figure 5.2 The relationship between root concentration factor (RCF) and the octanol – water partition coefficient (Log Kow) of root uptake of o-methylcarbamoyloximes (open circle) and substituted phenylurea (cross mark) by barley plants from nutrient solution (Briggs et al., 1982) 134 The translocation of non-ionized compounds is unlike to the root uptake. It fitted to bellshape curve with most effective translocation at intermediate of hydrophobicity (log Kow  2). The highly hydrophobic (log Kow >4.5) or highly polar (log Kow<1) compounds have small translocation capacity as shown in Figure 5.3 (Briggs et al., 1982; de Carvalho, Bromilow, & Greenwood, 2007a). The reason of this most efficient log Kow is unclear. The polar compounds have difficulties in diffusion cross the lipid-like membrane of endodermis and the highly hydrophobic compound can cross the endodermis much less than water may be the explanation (Trapp & McFarlane, 1995). This evidence had been mention in that most hydrophobic chemical found to uptake by root but no translocation to the shoot (Hellstrom, 2004). This indicated that there are selective rejections for highly polar and hydrophobic chemical occuring at membrane barrier in the root. Briggs et al., 1982 had shown the Gausian curve that fitted to TSCF data and log Kow of nonionized compounds as shown in Eq.5.7. (5.7) Figure 5.3 The relationship between the translocation stream concentration factor and octanolwater partition coefficient of the root uptake of 0-methylcarbamoyloximes (open circle) and substituted phenylurea (cross mark) by Barley plants from nutrient solution (Briggs et al., 1982) 135 For accumulation in stem, the stem concentration factor (SCF) or amount of chemicals that partition to stem found increase with increasing hydrophobicity. The mainly mechanism is partition to solid phase of stem. However, the maximum calculated SCF showed at log Kow about 4.5, then the SCF will decrease with the same reason of decreasing of TSCF as shown in Figure 5.4 (Briggs et al., 1983). Figure 5.4 The relationship between stem concentration factor (SCF) and and octanol-water partition coefficient (Log Kow) of o-methylcarbamoyloximes in the stem bases (close circle) and central stem(open circle) sections, and substituted phenylnreas in the stem bases (closed triangle) and central stem (open triangle) sections taken up by barley plants from nutrient solution (Briggs et al., 1983) 5.2.5 Plant uptake of ionized compounds The transport of ionizable compounds across the membrane is more complex because pH in the plant compartments is different. The molecules can change their from when partition in different compartment as a result in different membrane permeation property (Briggs, Rigitano, & Bromilow, 1987). The diffusion across the membrane in cell protoplast of neutral molecules is quite rapid whereas it is much slower for the ionized molecules, so chemicals tend to 136 accumulate in high-pH plants compartment e.g. cytoplasm. This process is called “ion trapping” (Trapp & McFarlane, 1995). Because the permeability of membrane to anion is very low, so the transport of anion across the endodermis is not efficient (Trapp & McFarlane, 1995). Orita, 2012 indicated the ionized organic compounds are not taken up well as the non-ionized compounds because they are cross membrane by proton pumping process which required high activation energy from ATP-ADP reaction. The uptake of weak organic acids is found to increase when the pH of nutrient solution decrease where the compounds are in neutral form. This was confirmed from the increasing of RCF of 2,4 D from 0.92 at pH 7 to 33.3 at pH 4 as shown in Figure 5.5 (Briggs et al., 1987; Inoue, Chamberlain, & Bromilow, 1998). Figure 5.5 The root concentration factor of 2,4 dichloro-phenoxyacetic acids(open circle) , and 3,5 dichloro-phenoxyacetic acids (closed circle) uptake by barley as a function of pH of nutrient solution (Briggs et al., 1987) 137 Weak acids are major compounds that mobile and accumulate in phloem by ion trapping. This process can be explained that the neutral molecules can freely cross the lipophilic membrane from xylem to phloem. However, the pH in phloem ( 8) are more basic than xylem ( 5.5), so compounds will be ionized in phloem and these ionized molecules cannot cross the membrane back to xylem , so they are in trapped and accumulated in phloem as shown in Figure 5.6. This process reduce the long distance transport in the plant and accumulation in the leave of weak acids (Hellstrom, 2004; Trapp & McFarlane, 1995) Figure 5.6 The accumulation of weak acids between xylem and phloem by ion trapping effect (Hellstrom, 2004) The RCF, TSCF and SCF of some non-ionized and ionized chemicals had been collected from literatures which show in Table 5.1. However, these values are subject to be different depend on measured methods (hyponic culture or pressure chamber), length of exposure, initial concentration and plant species. 138 Table 5.1 The literature values of RCF and TSCF of some non-ionized and ionized chemicals Chemicals pH Log Kow RCF TSCF (ml/g) (ml/g) References Non-ionized chemicals Toluene - 2.73 - 0.64 Dettenmaier et al.,2009 Benzene - 2.13 - 0.59 Dettenmaier et al.,2009 Aldoxycarb - -0.57 0.65 0.18 Briggs et al. 1982 3-Phenoxybenzaldehyde - 3.12 8.62 0.29 Briggs et al. 1982 4 0.06 2.62 0.12 Briggs et al. 1987 5-6.5 1.53 1.85 0.75 Shone et al. 1974 2,4 Dichloro-phenoxyaceticacid 4 2.81 88.4 3.12 Shone et al. 1974 2,4 Dichloro-phenoxyaceticacid 6.5 2.81 8.07 0.142 Shone et al. 1974 0-Methylcarbamoyloximes Ionized chemicals 4-mesyl POA phenoxyaceticacid Atrazine 5.3 Materials and methods The plant uptake capacity of AAs is determined by root uptake through nutrient solution in hydroponic culture system and sand culture system by cucumber plants. The experiment was started from incubation the cucumber seeds in germination box until plants were big enough before transfer to AAs spiked nutrient solution in hydroponic culture or sand applied with Aristolochia Clematitis seed in sand culture. After period of time, plants were harvest and analyzed the AAs in plant tissues included leaves, stems and roots by HPLC-FLD with precolumn derivatization reaction. The detail of experiment was described below. 139 5.3.1 Calibration curve To prepare the plant matrix solution, 0.5 g fresh wt. of cucumber plant parts (leaves, stems, and roots) were macerated with pestle and mortar and extracted in 20 ml of 70% methanol. Solutions were left overnight ,then sonicated in ultrasonic bath for 15 min. After that the mixtures were shaken in mechanical shaker at 180 rpm for 1 hr, and then centrifuged at 7,500 rpm for 30 min to separate the plants tissues, the supernatants were collected as leaves, stems and roots matrix solution. The calibration solutions were prepared by spiking AAs in various amounts to plant matrix solution. Due to low concentration of AAs was expected in plant tissues, the HPLC-FLD with pre-column derivatizaiton had been used to analyze AAs in the extract solution. To analyze, 1 ml of calibration solution were transferred to 1.5 ml Eppendorf tube, added 10 mg of Zinc powder and 50 µl of concentrated acetic acid with periodic mixing with vortex mixer for 15 min. This procedure will derivertize Aristolochic acid I and II to Aristolactam I and II(ALs) as shown in Figure 5.7. Then the solutions were centrifuge at 13,000 rpm for 5 min and the supernatants were collect to analyze by HPLC-FLD(Chan et al., 2007). The Pelkin Elmer HPLC system equipped with Supelco Discovery C-18 column 25 cm x 4.6 mm, and 5 µm particle diameter was used to analyzed ALs. The fluorescence detector (FLD) was used at excitation and emission wavelength 393 and 455 nm, respectively. The mobile phase was the mixing solution between acetonitrile and 0.1% phosphoric acid water at the gradient of 40% to 80% of acetronitrile in 10 min and held for 3 min before recondition to starting condition. The flow rate was 1 ml/min. The calibration curve obtained from leaves, stems and roots solution were compared. The accuracy, precision, reproducibility, detection limit and recovery also had been determined. 140 Figure 5.7 The derivatization of Aristolochic acids to Aristolactam by using zinc powder in acid solution (Chan et al., 2007) 5.3.2 Analytical method validation To verify the above analytical method for determination of Arisolochic acids in plant matrix solution, the method validation included accuracy, precision, linearity and detection limit were determined. 5.3.2.1 Accuracy and precision The accuracy of the HPLC-FLD method was tested by spiking stock solution of AA I and II to leave, stem or root matrix solution at three concentration (0.0225 , 0.09 and 0.225 µg/ml for AAI and 0.0275 , 0.11and 0.275 µg/ml for AAII), derivatization to Aristolactam I and II and measured for 6 time in each concentration (n=6) . The percent recovery was calculated from Eq.5.8 and reported as accuracy. (5.8) 141 The precision of the method was checked by analyzing the above standard solution for 6 time in one day (intraday, n=6) and twice a day for 3 consecutive days (interday, n=6 ). The percent relative standard deviation (%RSD) of peak area of each concentration were calculated and reported as precision at each concentration 5.3.2.2 Linearity The linearity of method was check by prepared six standard solution of Aristolactum I and II which derivatized from Aristolochic I and II 0.0225 -0.225 and 0.0275-0.275 ug/ml, respectively. Each concentration was analyzed for three times. The average peak areas were plotted with concentration. The linear regression equation and correlation coefficient (r2) were obtained from the graph. 5.3.2.3 Limit of detection(LOD) and limit of quantification (LOQ) The limit of detection and quantification were measured by analyzing the lowest standard concentration of AL I and II for 7 times and calculated the standard deviation (SD). The LOD calculated from 3 times of standard deviation (3 times of signal to noise ratio) and LOQ calculated from 10 of times of standard deviation (10 time of signal to noise ratio). 5.3.3 Nutrient solution preparation The Knop’s solution were used as nutrient solution in experiment which prepared by dissolved Ca(NO3)2 1 g, KH2PO4 0.25 g, KCl 0.125 g, MgSO4 0.25 g and trace of FeSO4 in 1 liter of deionized water (The Gale Group, 2010) . The pH of Knop’s solution was adjusted to 5.9 by adding aliquot of 1M NaOH before use. 142 5.3.4 Hydroponic culture experiment The cucumber seeds were germinated on towel paper saturated with 0.05 mM CaSO4 in germination box as shown in Figure 5.8. Seven day later, after seedling had developed the cotyledons and about 5 cm root length, all seedling were transferred to grow hydroponically in 25 ml glass bottle (2 plant per bottle) which filled with 20 ml of half strength of Knop’s nutrient solution under 10/14 day/night cycle in well ventilated area . The light was supplied by 60 Watt full wavelength lamp. After 14 day of pre-culture, plants in each bottle were transferred to new nutrient solution containing the AA I and II at 1.4 and 3 µg/ml, respectively. The 5 ml of fresh half strength Knop’s solution was added to each tube in every 2 days to avoid nutrient deficiency. The amounts of AAs in initial solution were measured to calculate the mass balance. The bottles were wraped up with aluminum foil to protect from light and prevent the development of algae in nutrient solution. The hydroponic culture experimental set up are shown in Figure 5.9. Duplicate tubes were applied at each treatment. One control bottle was grown with the nutrient solution without spiked AAs. pH of solution was measured before and after treatment. After 20 days of exposure to AAs, the all plants were harvested and separated into different parts e.g. roots, stems and leaves and weighed as fresh weight as shown in Figure 5.10 . Roots were additional washed by soaking in 2.5 mM CaSO4 for overnight to remove adsorbs AAs on the surface, rinsed with deionized water and blotted dry with Kim wipe paper before weighing. The washed solutions were kept to analyze the AAs desorbed from root surface. AAs remaining in nutrient solution was measured by taking 1ml of nutrient solution and centrifuged with Eppendorf tube at 13,000 rpm for 5 min to remove plant particles and analyzed by HPLCDAD similar to soil sorption/octanol-water coefficient experiment. 143 The analyzed concentrations in plant parts and in nutrient solution were calculated as Root concentration factor (RCF) and stem concentration factor (SCF) as shown in Eq 5.1 and 5.2. The concentration in leaves, stems or roots was based on fresh weight basis. The reported concentration did not correct with recovery efficiency. Degradation of AAs in plants was assumed as negligible and concentration left in nutrient solutions was used as concentration in external solution. Figure 5.8 Cucumber seedlings germinated on wetted towel in box 144 Figure 5.9 AAs uptake by hydroponic culture experimental setup Figure 5.10 Plant parts separation after harvesting 145 5.3.4.1 Extraction method and recovery test The combined plants parts were macerated with pestle and mortar as shown in Figure 5.11 and dissolved with 5 ml of 70% methanol. These mixture solutions were let stand overnight. To extract AAs out from the plant tissues, the solutions were sonicated in ultra-sonic bath for 15 min as shown in Figure5.12 and then shaken in mechanical shaker for 1 hr at 180 rpm. After that, the solutions were centrifuged at 7,500 rpm for 20 min to separate the plant tissue and supernatants were collect to analysis by HPLC-FLD as described above as shown in Figure 5.13. This extraction method was selected because of widely using in extraction the AAs from Aristolochia species plants (Sun, Wu, & Jia, 2001; J. B. Yuan et al., 2008; Zhang et al., 2006) Figure 5.11 Plant parts maceration by pestle and mortar 146 Figure 5.12 Plant extracted by sonication in ultrasonic bath Figure 5.13 Supernatant separation from plant tissue after centrifuge 147 The recovery of this extraction method was determined by spiking the aliquot of standard stock AAs solution at three different concentration to 0.5 g fresh weight of leaves, stems and roots which already homogenized by pestle and mortar . The spiked samples were left for 24 hr to allow spiking solution penetrated to plant tissues. After that the samples were extracted by same method as described above. % of recovery of each plant parts and spiked concentration were calculated. 5.3.4.2 Degradation of Aristolochic acids in nutrient solution The degradation of Aristolochic acids in Knop’s solution were tested to ensure that AAs were not degraded during uptake experiment. The several bottles of Knop’s solution containing AAs were prepared in the same manner as uptake experiment. The 1 ml of solution was taken from bottles in day 4, 8, 12, 16, and 20 and analyzed AAs concentration remaining in solution by HPLC-DAD. pH also measured before and after experiment. 5.3.5 Sand culture experiment 5.3.5.1 Sand preparation F-65 Ottawa Silica sand (SiO2 99.77%, Laguna Clay Co. ) was used as a media to determine AAs uptake because it has low ability to adsorb the AAs and inert for chemical reactions. Before apply to experiment, sand was washed with deionized water for 2 times and dried in the fume hood. Then the dried sand was steriled by autoclaving (10 min heat and 10 min dried cycle) to inhibit microbial activity. The ground Aristolochia Clematits seeds were added to the sand as a source of AAs. The AAs leaching from seeds and available concentration in sand were determined as described below. 148 5.3.5.2 The uptake experiment The cucumber seedlings were germinated on towel paper saturated with 0.05 mM CaSO4 in germination box same as in hydroponic experiment. After 7-10 days when the seedlings were long enough, they were transferred to grow in nutrient solution for 2 weeks. When true leaves came out, the seedlings were transferred to 4 oz. straight-side glass jar contained 40 g of treated sand and 0.02 g of ground Aristolochia Clematitis seeds. The 5 ml half strength Knop’s nutrient solution was irrigated the cucumber plants in every two day to replace water evaporator and prevent nutrient deficiency. Three pots were prepared for AAs treatment plants and one control pot was included where no ground Aristolochia Clematitis seeds was applied. The plant uptake by sand culture experiment is shown in Figure 5.14 Figure 5.14 Plant uptake by sand culture experimental setup 149 Plants were harvest after 20 days of growing in sand. They were cut into different parts, weighed and rinsed with deionized water before extraction. The extraction and analysis procedure were similar to hydroponic culture experiments. To determine available concentration of AAs to root, after plants were harvested, sand was irrigated with 5 ml of half strength Knop’s solution, equilibrated for 1 hr and 0.5 ml of clear solution on top were collected and put into eppendorf tube for centrifuge to remove sand particle. The supernatants were collected to analyze AAs with HPLC-DAD. These concentrations were used as concentration in external media to calculate plant bioconcentration factors (RCF and SCF). 5.3.5.3 Determination of AAs degradation in sand AAs degradation in sand was tested by batch incubation technique (Jenks et al., 1998). The 40 g of washed and sterilized Ottawa sand was mixed with 0.02 g of ground Aristolochia clematitis seeds and put into 4 oz. straight-side glass jar. These AAs mixed sand were cover with aluminum foil with small holes on the top for air flow. The jars were incubated for 20 days with measuring AAs concentration in sand in every two days. To determine the concentration of AAs in sand, 5 ml of half strength Knop’s nutrient solution was applied to the jar and equilibrated for 1 hr before 0.5 ml of clear solution on top was taken to eppendorf tube for centrifuge to remove sand particle. The supernatant was analyzed for AAs by HPLC-DAD. The samples were prepared in duplicate. 150 5.4 Result and discussion 5.4.1 Analytical method validation 5.4.1.1 The calibration curve in plant matrix solution The ALs standards solution of various plant matrices (leaves, stems, roots) were prepared and injected to HPLD-FLD system. The chromatograms of each plant matrix were shown in Figure 5.15. The retention time of AL I and II was 13.7 and 11.5 min, respectively. The figures indicate that plant extract compounds e.g. chlorophyll had no interference to ALs HPLC signals. a) Leave matrix AL II b) Stem matrix AL I AL I AL II c) Root matrix AL I AL II Figure 5.15 HPLC-FLD chromatogram of derivative Aristolactum I and II in leaves (a), stem (b) and root (c) matrices solution spiked with AA I and II 151 The peak areas of six standard solutions were measured for three times and the average peak areas were plotted with concentration as calibration curve as shown in Figure 5.16. The figures show that leaves, stem, root extract matrices did not have any enhancement or suppression effect on the ALs signal of HPLC. The calibration was linear over the range of concentration 0.0225 -0.225 for AA I and 0.0275-0.275 ug/ml for AA II, respectively. a) AA I 350,000 Leaf Stem Root Methanol 300,000 Peak area 250,000 200,000 150,000 100,000 50,000 0 0.05 0.1 0.15 Concentration (µg/ml) 0.2 0.25 b) AA II Peak Area 160,000 leaf 140,000 stem 120,000 root Methanol 100,000 80,000 60,000 40,000 20,000 0 0.1 0.2 concentration (µg/ml) 0.3 Figure 5.16 The calibration curve of AL I (a) and AL II (b) in leaf, stem, root matrix and methanol 152 5.4.1.2 Accuracy, precision and limit of detection The accuracy of HPLC-FLD method was determine at low, medium and high concentration in leaves matrix solution and reported as % recovery while the precision reported as % relative standard deviation (%RSD). These data shows in Table 5.2. The results show that % recovery of AL I and II are over than 100% and % RSD less than 4%. The intraday and interday precisions were not much different. However, accuracy and precision at low concentration were less than at high concentration solution. These data indicated the HPLC-FLD with precolumn derivatization had good performance to analyzed AAs in plant matrices. Table 5.2 % recovery (accuracy) and % RSD (intra-day and inter-day precision) of analysis ALs in leave matrix Chemical Concentration (µg/ml) AL I AL II %Recovery %RSD %RSD (mean, n=6) (intra-day, n=6)a (inter-day, n=6)b 0.0225 114.77 3.22 3.78 0.09 98.64 1.24 1.82 0.225 104.56 0.51 1.17 0.0275 122.64 8.08 6.77 0.11 100.16 1.35 1.77 0.275 101.91 0.62 0.75 a The samples were analyzed 6 times in 1 day. b The samples were analyzed 6 times over 3 three consecutive days The limit of detection (LOD) and limit of quantification (LOQ) of ALs were determined from the 3 and 10 times of standard deviation of peak area at lowest calibration concentration. The results show the LOD of AA I and II were 0.002 and 0.0045 µg/ml and LOQ of AA I and II 0.007 and 0.015 µg/ml. The sensitivity of AA II showed slightly less than AA I. The obtained 153 LOD and LOQ were higher than the reported from Chan et al., 2007 one order of magnitude which were 0.00039 µg/ml for AA I and 0.00052 µg/ml for AA II. 5.4.2 Recovery of plant extraction method The recovery of plant extraction method was determined by spiking AAs standard solution to the leave and stem tissues at three different concentration. The extraction solutions were analyzed for ALs peaks and compared to the peaks obtained from standard solution at same concentration. The results show that the average %recovery of AA I and II were 89.24% and 70.38% for leaves and 91.02 and 79.92 for stems, as shown in Figure 5.17. The figures show that %recovery of AA I was more than AA II and % recovery from stems was more than leaves. This finding may indicated the sorption of AAs in leaves may be stronger than in stem. The % recovery from root tissue was not determined because the volume of root tissues was too small to adsorb the solution. The high % recovery indicated the good performance of extraction method. a) AA I 120 Leaves Stem % recovery 100 80 60 40 20 0 1.00 2.00 3.00 Spiked concentration (µg/g plant) Figure 5.17 The % Recovery from spiked AA I (a) and II (b) standard solution to leaves and stem matrices 154 Figure 5.17 (cont’d) b) AA II 120 Leaves Stem % Recovery 100 80 60 40 20 0 1.39 2.78 4.17 Spiked concentration (µg/g plant) 5.4.3 Hydroponic culture experiment 5.4.3.1 The stability of AAs in Knop’s nutrient solution The AAs degradation in Knop’s nutrient solution was tested. From Figure 5.18 , it shows that AAs were stable in a solution for whole experimental period (20 days). This indicated that AAs were not degraded by chemical processes in nutrient solution. However, this experiment did not account for the degradation by microorganisms which may happen when grow the plant. 155 Concentration in nutrien solution (µg/ml) 3.5 3 2.5 2 1.5 1 0.5 AA I AA II 16 20 0 0 4 8 12 time (days) 24 Figure 5.18 The stability of AAs in Knop’s solution as a function of time 5.4.3.2 Kinetic uptake experiment The plant uptake kinetic had been tested by growing the cucumber plants in nutrient solution applied AAs and harvested in every 4 days interval. The AAs remaining in solution and concentration in shoot (stems and leaves) had been determined and showed in Figure 5.19 and 5.20, respectively .The AAs remaining in nutrient solution found to decrease in first 4-8 day and then slightly decrease until the end of exposure period (20 days). The presence of AAs in solution at end of periods indicated AAs were not completely degraded by microoragnisms, so the absence of AAs from solution should reflect to amount of AAs that taken up by plant. However, the concentration found in shoot did not show the relationship with the exposure time. The uptake ability is expected to depend on individual plant and it is difficult to predict plant uptake equilibrium (Briggs et al., 1983). 156 70 AA remain in Nutrient solution (µg) AAI 60 AAII 50 40 30 20 10 0 0 5 10 15 20 25 Days Figure 5.19 The AA I and II remaining in solution as a function of exposure time AA accumulated in shoot (ug/g shoot) 0.90 AA I 0.80 AA II 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 0 5 10 15 20 25 Day Figure 5.20 The AA I and II concentration in shoot as a function of exposure time 157 5.4.3.3 AAs accumulation in plant part The AAs concentration in different parts of cucumber plants are shown in Figure 5.21. The concentration in nutrient solution and root wash solution were also determined. From the mass balance calculation, the average AA I and II in remaining nutrient solution 61 and 59 %, in root wash solution 6 and 3%, in stem tissues 1.8 and 1.4 %, in root tissues 6 and 3%, and mass loss 24 and 32% ,respectively. Very small AAs amount were detected in leaves. The loss portion was expected to be occurred in sample handling during experiment (Zhu, Han, Xiao, & Jin, 2008). The plant metabolism and non-extractable of AAs in plant tissues may also account for this loss (Redshaw, Wootton, & Rowland, 2008). a) AA I 70 plant 1 plant 2 plant 3 60 % mass 50 40 30 20 10 0 Media Root wash Leaves solution Stem Root Loss Figure 5.21 The percentage of AA I (a) and AA II (b) in different parts in cucumber plant parts and residual in media 158 Figure 5.21 (cont’d) b) AA II 70 plant 1 plant 2 plant 3 60 % mass 50 40 30 20 10 0 Media Root wash Leaves solution Stem Root Loss Figure 5.22 shows % distribution of AA I and II that accumulated in different cucumber parts. It shows that about 70% of AAs was accumulated in the roots, 22-30% in stem and 0-2% in leaves. This indicated AAs tended to accumulate in the roots more than the stems and leaves. The low translocation to the shoot was expected from the difficulty to cross the lipid membrane of high polarity anion AAs. The very low concentration in leaves also indicated AAs were not mainly translocated by water in transpiration process to the leave (Shenker et al., 2011; Ucisik, Trapp, & Kusk, 2007). This may be caused by the partitioning of AAs in lipid components in the stem or trapped in the pholem by ion trapping mechanism (Hellstrom, 2004). 159 a) AA I b) AA II 0.00 2.64 22.68 30.69 Leaves Stem Stem 74.68 Leaves 69.31 Root Root Figure 5.22 AA I (a) and AA II (b) distribution in cucumber plant parts in hydroponic culture Figure 5.23 (a) shows concentration of AA I and II found in root and stem tissues and Figure 5.23 (b) shows the calculated root concentration factor (RCF) and stem concentration factor (SCF) of AA I and II. The RCF of AA I and II were 8.48 and 4.37 and the SCF of AA I and II were 0.78 and 0.61, respectively. The obtained RCF values were large when compare to other common organic compounds indicated that AA I and II can be effectively taken by root. RCF and SCF of AA I are more than AA II indicated AA I can be taken up by plants more than AA II which caused by more hydrophobicity of AA I. 160 a) Concentration (µg of AA/g plant) 9 AA I 8 AA II 7 6 5 4 3 2 1 0 stem root b) 9.00 Plant concentration Factor 8.00 7.00 AA I AA II 6.00 5.00 4.00 3.00 2.00 1.00 0.00 SCF RCF Figure 5.23 AAs concentration found in stem and root tissues (a) and stem and leaf concentration factor (b) (SCF and RCF) of cucumber plant uptake AAs from spiked nutrient solution 161 5.4.4 Sand culture experiment 5.4.4.1 The AAs available in treating sand The leaching AAs concentration in the sand from A.Clematitis seeds was tested as a function of incubation time to determine whether AAs degradation had been occurred. The result is shown in Figure 5.24 where AA I concentration was decrease when time proceeded whereas AA II concentration was quite constant. The reduce concentration of AAs may response by the degradation by microorganisms which may be contaminated from the air, even though sand was sterilized (Y. Z. Gao & Zhu, 2004). The irreversible adsorption in the sand may be the cause too. However, this figure can confirm that there wereAAs available in the sand for a whole experiment period (20 days). 7.00 AA I AAs concentration (µg/ml) 6.00 AA II 5.00 4.00 3.00 2.00 1.00 0.00 0 5 10 15 20 25 Days Figure 5.24 AA I and II concentration available in sand as a function of incubation time 162 5.4.4.2 AAs accumulation in plant parts The AAs concentration distriburtion in the plants which exposed to treated sand are shown in Figure 5.25. However, the distribution pattern was not similar to hydroponic culture where the accumulation in the stems was increase or even larger than concentration in the roots in case of AA I. This evidence can be explained that plants were grown better in sand culture because their roots were fixed. As a result, the transpiration of the plants in sand culture was better in hydroponic culture. AA I was quite effective for translocation from root to shoot. No AA I and II found in leaves. The AAs concentration found in plant tissues are shown in Figure 5.26 (a). The concentration of AA I in plants was higher than AA II because A. Clematitis seeds contained AA I ten times more than AA II. From Figure 5.26 (b), the RCF of AA I and II were 6.28 and 4.39 and SCF of AA I and II were 2.54 and 2.72, respectively. The higher than 1 of these numbers indicated the concentration in plants was relative higher than concentration in sand. When compare to hydroponic culture, the SCF from sand culture was increased whereas the RCF was decreased. These indicated AAs have more translocated and accumulated when growing in sand culture than hydroponic culture. b) AA II a) AA I 0.00 0.00 29.72 47.93 Leaves 70.28 52.07 Stem Leaves Stem Root Root Figure 5.25 AA I (a) and AA II (b) distribution in cucumber plant parts in sand culture 163 a) 25 Concentration (µg of AA/g plant) AA I 20 AA II 15 10 5 0 stem 7.00 root b) AA I Bioconcentration Factor 6.00 AA II 5.00 4.00 3.00 2.00 1.00 0.00 SCF RCF Figure 5.26 AAs concentration found in stem and root tissues (a) and stem and leaf concentration factor (b) (SCF and RCF) of cucumber plants grown in Aristolochia Clematitis seed treating sand 164 5.5 Conclusion The kinetic test showed that AAs can be uptake after 4 days but the plant uptake equilibrium was difficult to determine because it depends on uptake ability of individual plant. Most of taken up AAs was found in root and the less far was translocated to the shoot which expected from the low root membrane permeability of anion AAs. RCF from both hydroponic and sand culture were much higher than 1 inidicated AAs has highly potential to accumulate in the root. The translocation of AAs was better in sand culture when plants growed better. The finding contamination of AAs in the plants that growed in AAs contained media suggested the hypothesis of food chain contamination is possible. 165 REFERENCES 166 REFERENCES Aslund, M. L. W., Rutter, A., Reimer, K. J., & Zeeb, B. A. (2008). The effects of repeated planting, planting density, and specific transfer pathways on PCB uptake by Cucurbita pepo grown in field conditions. Science of the Total Environment, 405(1-3), 14-25. doi: 10.1016/j.scitotenv.2008.07.066 Briggs, G. G., Bromilow, R. H., & Evans, A. A. (1982). RELATIONSHIPS BETWEEN LIPOPHILICITY AND ROOT UPTAKE AND TRANSLOCATION OF NONIONIZED CHEMICALS BY BARLEY. Pesticide Science, 13(5), 495-504. doi: 10.1002/ps.2780130506 Briggs, G. G., Bromilow, R. H., Evans, A. A., & Williams, M. (1983). RELATIONSHIPS BETWEEN LIPOPHILICITY AND THE DISTRIBUTION OF NON-IONIZED CHEMICALS IN BARLEY SHOOTS FOLLOWING UPTAKE BY THE ROOTS. Pesticide Science, 14(5), 492-500. doi: 10.1002/ps.2780140506 Briggs, G. G., Rigitano, R. L. O., & Bromilow, R. H. (1987). PHYSICOCHEMICAL FACTORS AFFECTING UPTAKE BY ROOTS AND TRANSLOCATION TO SHOOTS OF WEAK ACIDS IN BARLEY. Pesticide Science, 19(2), 101-112. doi: 10.1002/ps.2780190203 Chan, W., Lee, K.-C., Liu, N., & Cai, Z. (2007). A sensitivity enhanced high-performance liquid chromatography fluorescence method for the detection of nephrotoxic and carcinogenic aristolochic acid in herbal medicines. Journal of Chromatography A, 1164(1-2). doi: 10.1016/j.chroma.2007.06.055 Chiou, C. T., Sheng, G. Y., & Manes, M. (2001). A partition-limited model for the plant uptake of organic contaminants from soil and water. Environmental Science & Technology, 35(7), 1437-1444. doi: 10.1021/es0017561 Cui, L. W., Feng, X. B., Lin, C. J., Wang, X. M., Meng, B., Wang, X., & Wang, H. (2014). ACCUMULATION AND TRANSLOCATION OF (198)HG IN FOUR CROP SPECIES. Environmental Toxicology and Chemistry, 33(2), 334-340. doi: 10.1002/etc.2443 de Carvalho, R. F., Bromilow, R. H., & Greenwood, R. (2007a). Uptake and translocation of non-ionised pesticides in the emergent aquatic plant parrot feather Myriophyllum aquaticum. Pest Management Science, 63(8), 798-802. doi: 10.1002/ps.1394 de Carvalho, R. F., Bromilow, R. H., & Greenwood, R. (2007b). Uptake of pesticides from water by curly waterweed Lagarosiphon major and lesser duckweed Lemna minor. Pest Management Science, 63(8), 789-797. doi: 10.1002/ps.1389 167 Dettenmaier, E. (2008). Measuring ang modeling of plant root uptake of organic chemicals. Utah State University. Diffen. Pholem vs Xylem. from http://www.diffen.com/difference/Phloem_vs_Xylem Gao, Y. Z., & Zhu, L. Z. (2004). Plant uptake, accumulation and translocation of phenanthrene and pyrene in soils. Chemosphere, 55(9), 1169-1178. doi: 10.1016/j.chemosphere.2004.01.037 Hashimoto, Y. (2005). Dieldrin residue in the soil and cucumber from agricultural field in Tokyo. Journal of Pesticide Science, 30(4), 397-402. doi: 10.1584/jpestics.30.397 Hellstrom, A. (2004). Uptake of organic pollutants in plants. http://info1.ma.slu.se/IMA/Publikationer/internserie/2004-01.pdf Hulster, A., Muller, J. F., & Marschner, H. (1994). SOIL-PLANT TRANSFER OF POLYCHLORINATED DIBENZO-P-DIOXINS AND DIBENZOFURANS TO VEGETABLES OF THE CUCUMBER FAMILY (CUCURBITACEAE). Environmental Science & Technology, 28(6), 1110-1115. Inoue, J., Chamberlain, K., & Bromilow, R. H. (1998). Physicochemical factors affecting the uptake by roots and translocation to shoots of amine bases in barley. Pesticide Science, 54(1), 8-21. doi: 10.1002/(sici)1096-9063(199809)54:1<8::aid-ps793>3.0.co;2-e Jenks, B. M., Roeth, F. W., Martin, A. R., & McCallister, D. L. (1998). Influence of surface and subsurface soil properties on atrazine sorption and degradation. Weed Science, 46(1), 132-138. Kipopoulou, A. M., Manoli, E., & Samara, C. (1999). Bioconcentration of polycyclic aromatic hydrocarbons in vegetables grown in an industrial area. Environmental Pollution, 106(3), 369-380. doi: 10.1016/s0269-7491(99)00107-4 Lin, D. H., Zhu, L. Z., He, W., & Tu, Y. T. (2006). Tea plant uptake and translocation of polycyclic aromatic hydrocarbons from water and around air. Journal of Agricultural and Food Chemistry, 54(10), 3658-3662. doi: 10.1021/jf052909c McCrady, J. K., McFarlane, C., & Lindstrom, F. T. (1987). THE TRANSPORT AND AFFINITY OF SUBSTITUTED BENZENES IN SOYBEAN STEMS. Journal of Experimental Botany, 38(196), 1875-1890. doi: 10.1093/jxb/38.11.1875 Oconnor, G. A., Kiehl, D., Eiceman, G. A., & Ryan, J. A. (1990). PLANT UPTAKE OF SLUDGE-BORNE PCBS. Journal of Environmental Quality, 19(1), 113-118. Orita, N. (2012). Root uptake of organic contaminants into plants: species differences. Utah State University. 168 Polder, M. D., Hulzebos, E. M., & Jager, D. T. (1995). VALIDATION OF MODELS ON UPTAKE OF ORGANIC-CHEMICALS BY PLANT-ROOTS. Environmental Toxicology and Chemistry, 14(9), 1615-1623. doi: 10.1897/15528618(1995)14[1615:vomouo]2.0.co;2 Raveton, M., Ravanel, P., Serre, A. M., Nurit, F., & Tissut, M. (1997). Kinetics of uptake and metabolism of atrazine in model plant systems. Pesticide Science, 49(2), 157-163. doi: 10.1002/(sici)1096-9063(199702)49:2<157::aid-ps517>3.0.co;2-m Redshaw, C. H., Wootton, V. G., & Rowland, S. J. (2008). Uptake of the pharmaceutical Fluoxetine Hydrochloride from growth medium by Brassicaceae. Phytochemistry, 69(13), 2510-2516. doi: 10.1016/j.phytochem.2008.06.018 Sandermann, H., Scheel, D., & Vandertrenck, T. (1984). USE OF PLANT-CELL CULTURES TO STUDY THE METABOLISM OF ENVIRONMENTAL CHEMICALS. Ecotoxicology and Environmental Safety, 8(2), 167-182. doi: 10.1016/01476513(84)90059-9 Shenker, M., Harush, D., Ben-Ari, J., & Chefetz, B. (2011). Uptake of carbamazepine by cucumber plants - A case study related to irrigation with reclaimed wastewater. Chemosphere, 82(6), 905-910. doi: 10.1016/j.chemosphere.2010.10.052 Steudle, E., & Peterson, C. A. (1998). How does water get through roots? Journal of Experimental Botany, 49(322), 775-788. doi: 10.1093/jexbot/49.322.775 Su, Y. H., Liu, T., & Liang, Y. C. (2010). Transport via xylem of trichloroethylene in wheat, corn, and tomato seedlings. Journal of Hazardous Materials, 182(1-3), 472-476. doi: 10.1016/j.jhazmat.2010.06.055 Su, Y. H., Zhu, Y. G., & Du, X. (2005). Co-uptake of atrazine and mercury by rice seedlings from water. Pesticide Biochemistry and Physiology, 82(3), 226-232. doi: 10.1016/j.pestbp.2005.03.002 Su, Y. H., Zhu, Y. G., & Liang, Y. C. (2009). Interactions of mixed organic contaminants in uptake by rice seedlings. Chemosphere, 74(7), 890-895. doi: 10.1016/j.chemosphere.2008.10.057 Sun, Q., Wu, Y., & Jia, L. (2001). Quantitative Determination of Aristolochic Acid in Asarum heterotropoides Fr. Schmidt var. mandshuricum(Maxim.) Kitag. Asian Journal of Traditional Medicines. The Gale Group, I. (2010). Nutrient solution. from http://encyclopedia2.thefreedictionary.com/Nutrient+Solution 169 Trapp, S. (2000). Modelling uptake into roots and subsequent translocation of neutral and ionisable organic compounds. Pest Management Science, 56(9), 767-778. doi: 10.1002/1526-4998(200009)56:9<767::aid-ps198>3.3.co;2-h Trapp, S. (2004). Plant uptake and transport models for neutral and ionic chemicals. Environmental Science and Pollution Research, 11(1), 33-39. doi: 10.1065/espr2003.08.169 Trapp, S., & McFarlane, J. C. (1995). Plant contamination-Modeling and simulation of organic chemical processes. USA: Lewis. Ucisik, A. S., Trapp, S., & Kusk, K. O. (2007). Uptake, accumulation, phytotoxicity, and removal of 2,4-dichlorophenol in willow trees. Environmental Toxicology and Chemistry, 26(6), 1165-1171. doi: 10.1897/06-353r1.1 Yuan, J. B., Liu, Q., Zhu, W. F., Ding, L., Tang, F., & Yao, S. Z. (2008). Simultaneous analysis of six aristolochic acids and five aristolactams in herbal plants and their preparations by high-performance liquid chromatography-diode array detection-fluorescence detection. Journal of Chromatography A, 1182(1), 85-92. doi: 10.1016/j.chroma.2007.12.076 Zhang, C. Y., Wang, X., Shang, M. Y., Yu, J., Xu, Y. Q., Li, Z. G., . . . Namba, T. (2006). Simultaneous determination of five aristolochic acids and two aristololactams in Aristolochia plants by high-performance liquid chromatography. Biomedical Chromatography, 20(4), 309-318. doi: 10.1002/bmc.565 Zhu, H., Han, J., Xiao, J. Q., & Jin, Y. (2008). Uptake, translocation, and accumulation of manufactured iron oxide nanoparticles by pumpkin plants. Journal of Environmental Monitoring, 10(6), 713-717. doi: 10.1039/b805998e 170 Chapter 6 Root exudates of Aristolochia plants and leaching and decomposition of Aristolochia Clematitis seeds 6.1 Introduction Plant can release chemicals to soil by 2 major pathways, the root exudates from live plants and the leachates from decomposition of dead plants tissues. These plants and their parts can serve as source of chemicals into the soil and also other environment compartment. These released chemicals can impact to ecosystem around the living areas. Some chemicals are toxic to food crops and induce the problems such as crop rotation failure, crop selection in mix cropping, the growth reduction of vegetables or fruits which affect to agricultural economy or even toxic to human when chemicals incorporate to the food chain (U. K. Sahoo, Vanlalhriatpuia, Upadhyaya, & Roy, 2011; Yu & Matsui, 1994) . Root exudates define as substances that release from plant root to surrounding media. It consists of various compounds e.g. sugar, amino acid, and organic acids. They have both benefit or can be toxic to other plants. They found to enhance the nutrient mineral acquisition which required for plant growth (Dakora & Phillips, 2002). In the other hand, they can inhibit the growth or even kill the other plants (Rietveld, Schlesinger, & Kessler, 1983). This effect called “allelopathy” (Hasanuzzaman; Yu & Matsui, 1994). Other than root exudates, the potential sources of chemicals into the soil can be from leaching and decomposition of dead plant parts e.g. roots, stems, leaves, and seeds. When the 171 plant litter falls to the ground, they will be decomposed and release the chemical. Rain can wash the chemicals from plant biomass and carries to the soil. These chemicals can be accumulated in the soil if the degradation is low. Rietveld et al., 1983 reported the phytotoxic chemical “juglone” from black walnut tree that can harm the nearby trees to declined and died within a few years. The contribution of large amount of leachate from walnut biomass and the limited soil microorganism metabolism can build up the juglone to toxic level. Aristolochia clematitis is a weed plant that had the cycle in crop field in BEN area for long time. The previous studies showed that the BEN field were heavily invaded by living of A. Clematitis plant and soil had accumulation of dead plant tissue, so the releasing of AAs from A.Clematitis plant and dead plant tissues to soil would be expected (Hranjec et al., 2005; Pavlovic et al., 2013). There is evidence that the growth of corn fields in Romania which were largely invaded by Aristolochia Clematitis was suppressed or even damaged. This indicated the interaction between Aristolchia Clematitis plant that released chemicals to the soil and the corn plants that uptake these chemicals (Pavlovic et al., 2013). . To our best knowledge, there are no reports that study about the potential sources of AAs such as root exudates or leaching/biodegradation from Aristolochia plants. Therefore, this study was aimed to investigate the role of root exudates from Aristolochia species plant and the leaching and biodegradation of A.Clematitis seeds as potential sources of AAs that release and accumulate in the soil. 172 6.2 Literature reviews 6.2.1 Plant root exudates Organic compounds can release from plant roots to rhizosphere which we known as root exudates. Although plants in different species give their own exudates, these compounds quite similar but may be different in quantity. The identified compounds in exudates have wide range of chemicas which consist of carbohydrates, amino acids, organic acids, sugars, vitamins and enzymes which are listed in Table 6.1. Table 6.1 The organic compounds identified in root exudates (modified from Yoshitomi, 2001) Organic acids Sugars Vitamins Enzymes Acetic Arabinose p-Amino-benzoate Amylase Butyric Deoxyribose Biotin Invertase Citric Galactose Choline Protease Fumaric Glucose Inositol Glycolic Maltose Nicotinic acid Lactic Ribose Pantothenate Malic Sucrose Oxalic Xylose The quality and quantity of root exudation are affected by many factors such as age of the plant, the presence of microorganisms and environmental stress e.g. soil moisture and nutrient stress. The presence of microorganism can increase exudation of some plants significantly whereas may have no effect to other plant. These results suggested the dependent on plant species(Biondini, Klein, & Redente, 1988). The microbial metabolite may contain plant growth regulator such as gibberellins, auxins, cytokinins and also phytoxins such as phenolic acid and hydrogen cyanide. Moreover, the presence of microbe will utilize the organic carbon compound 173 exude at root surface which increase driving concentration gradient for passive diffusion (Yoshitomi, 2001). The soil moisture also have significant effect to amount of exudates where the release amount of carbon increase for soil which have water stress (Martin, 1977) . The nutrient stress can affect the composition of root exudates where plant will modify the exudates to enhance the uptake of scarce nutrients. Root exudates can be released to the soil by either the passively concentration gradient between root cells and soil solution or actively response to some activation such as metal toxicity, nutrient stress and microbial activity. Even though it has only small amount, exudates has significant role in soil nutrient availability because of their chelating properties and stimulation of microbial activity (Phillips, Erlitz, Bier, & Bernhardt, 2008). Root exudates will occur in narrow zone of soil around root which has closely interaction with microorganism. These may have positive and negative effect to the plant. For example the exudates will promote the plant growth by chemicals released from bacteria . The intensity of bacteria in rhizosphere also enhance degradation of pollutants e.g. PAHs in soil (Aprill & Sims, 1990; Reilley, Banks, & Schwab, 1996). The negative effect may include the soil borne pathogen or parasites infection to plants because of colonization around the root. The root exudates of some plants can damage to other neighboring plants, this plant to plant interaction is called “allelopathy”. The chemicals in exudates that have capability in inhibition seed germination and plan growth called “allelochemical” (Yamane, Nishimura, & Mizutani, 1992). The examples of allellochemicals included phenolic acids, terpinoids, flavinoids, polyacetylene and fatty acids which presence in root exudates and also various plant parts. These chemicals can inhibit the photosynthetic and oxygen evaluation process which is essential process for plant growth. They also can modify microbial community dynamic in 174 rhyzosphere (Buehler, 2010). The example weed which has allelopathy effect to the crop is Quack grass. It is important weed that highly reduce yield of corn fields by interfere nitrogen and potassium uptake by maize. The allelochemical “ethylene” which produce by microbial activity in soil was found at rhizomes of Quack grass and response for interruption of mineral uptake (Hasanuzzaman; Inderjit, 1996). A well known allelochemical from the walnut tree “juglone”, 5-hydroxy-1,4-napthoquinone, found to have high accumulation of in the root zone and decreasing with increasing distance from the trees(von Kiparski, Lee, & Gillespie, 2007). 6.2.2 Root exudates collect method In general, the exudates collection methods will comply with these constraint (a) it must capture exudates before microbial assimilation (b) the medium should not affect the root physiology or adsorp the root exudates and (c) the exudates can be distinguish from other soluble organic compounds in the medium(Phillips et al., 2008). A large number of methods had been developed to collect exudates but most of them are complex and also required a lot of laboratory processes (Matsumoto, Okada, & Takahashi, 1979; Prikryl & Vancura, 1980; Wadhwa & Narula, 2012) . Generally, there are two methods to study chemical composition and concentration of root exudates: in situ direct measuring from the soil and culture-based system which is more often used because exudates can be more trapped and separated from the medium. The culture-based system is also divided into two different types: static and dynamic trap solutions. Both types are mainly similar that the roots are submerged in the medium where exudates will be collected for a period of time. The primary difference is the removing and readding the medium solution in dynamic type to maintain diffusion gradient between root cells and medium solution and minimizing the re-uptake of exudates e.g. sugars and amino acids back to roots (Phillips et al., 2008). 175 The solution culture give advantage in simplicity of sample collection and maintenance. However, it does not have solid matrix to hold the plant which may affect to root morphology and exudation rate. The small glass bead and acid-washed sand are commonly use but they also has some limitation due to the sorption of exudates compounds. Another point to consider is the sterility of culture media. The sterile culture system can ensure that the exudates will not be utilized by microorganisms in the media. However, the microbial may stimulate the exudation, so the sterile culture may reduce rate of exudation and affect to composition of exudates (Phillips et al., 2008) 6.2.3 Plant litter leaching and decomposition When plant died and fall down to the soil, the organic substances in the plant can be released to the soil by decomposition process. The decomposition is the combining of physical, chemical and biological processes altering the chemicals on organic substrate in the plant. The plant litters contained several groups of organic compounds which vary with the parts such as leaves, stems, roots and bark but generally consist of water soluble compounds and hard degraded compounds e.g. cellulose and lignin. The major group of soluble compounds include sugar, phenolic acids and some nutrients. The sugars e.g. mono and oligosaccharide are from metabolism of plants. The phenolic compounds found as defensive agent from insects or precursor of lignin. In early stage of decomposition, these soluble compound are readily released from plant by leaching or dissolution with water and may sequestrated by organic matter and clay in the soil. The degradation process of soluble organic substance e.g. simple sugar glucose and fructose is quite rapid and quickly finish with in few month due to easy of run out with water and 176 microbial utilization (J. S. Singh & Gupta, 1977). The leachate of some plants contain allelochemicals e.g. phenolics, flavonoids and terpenoids which inhibit growth of surrounding plants (U. Sahoo, Jeeceelee, Vanlalhriatpuia, Upadhyaya, & Lalremruati, 2010; H. Singh, Batish, & Kohli). After the soluble was leached out and degraded, the carbohydrate e.g. starch and unshield cellulose which is accessible will be next degraded. Finally, the lignin and cellulose will be the last stage of degradation, the degradation is difficult and rate are much more slowly (Kuiters & Sarink, 1986) . Only some bacteria and fungi have the enzyme which can digest the chemical bonds in lignin and cellulose. The decomposition of soluble or non-soluble e.g. cellulose and lignin in plant substrate can be aerobic and anaerobic. In aerobic condition, the decomposition will give CO2 and release from the soil. In anaerobic condition, such as waterlogged in soil , the organic acids e.g. acetic acid will be produce instead of CO2. The long chain of fiber cellulose will be degraded to soluble short chain of glucose units. Lignin will be degraded until formation the stable humus where rate of decomposition is zero. Some microorganism have ability to completely mineralize lignin to CO2 and H2O (Berg & Mcclaugherty, 2014). Rate of decomposition can measure by mass loss or CO2 release which will be effect by litter composition, soil oxic condition and climate (Bragazza, Buttler, Siegenthaler, & Mitchell, 2009) 6.2.4 Cellulose degradation in activated sludge medium Cellulose and lignin are the carbon substrate that hard to degrade by general microorganism due to their insoluble and special enzyme needed. They can be degraded by cellulolytic fungi found in soil or natural water and celluloytic bacteria which can be found in both aerobic and anaerobic culture media such as in wastewater treatment plant. Normally, 177 activated sludge is mainly used for breakdown the soluble organic material. However, there are some studies show that activatd sludge is efficient to degrade these fibers as well (Edberg & Hofsten, 1975). Verachtert et al. show that the degradation of cellulose of filter paper and cotton wool in activated sludge tank can achieve to 80% and 60% w/w with 4-5 weeks incubation time (Verachtert, Ramasamy, Meyers, & Bevers, 1982). This shows that the cellulolysis degradation is active in activated sludge. The most detected cellulolytic microorganism is gram-negative bacteria such as Sporocytophaga myxococcoides which is most efficient aerobically cellulose degrading bacteria (Edberg & Hofsten, 1975). The cellulolytic bacteria found to grow in close contact or adhere with their substrates which result in long contact time. The direct contact between bacteria and fiber is necessary for efficient degradation. This study showed the ratio between cellulose to lignin was change from 1.2 in primary sludge to 0.4 in activated sludge which indicated the lignin utilization is slower (Edberg & Hofsten, 1975; Verachtert et al., 1982). 6.2.5 Cellulose degradation in anaerobic environment Cellulose degradation also was determined in anaerobic condition. The experiment showed that 60% of cellulose can be degraded by bacteria in aerobic treatment whereas the left 50-60% was degraded during anaerobic digestion(Verachtert et al., 1982). The rate of degradation was found to similar to the digesting with activated sludge in aerobic condition which required several days for degradation of cellulose to be occurred. The temperature had the effect to rate of degradation. For example, the 50% loss of cellulose was obtained in 25 days when incubated at 30oc but it will take 40-45 days if temperature was 17oc . The anaerobic 178 cellulolytic bacteria can be found in soil, sediment or waste water sludge. The most effective anaerobic cellulolytic bacter is Clostridium (Edberg & Hofsten, 1975). The anaerobic digestion is complex biodegradation process that involve the series of reactions including hydrolysis, fermentation and methanogenesis to convert organic compounds to methane and carbon dioxide (Siegert & Banks, 2005). Unlike aerobic decomposition, anaerobic decomposition required diverse microorganisms because they have to perform various fermentation and respiration processes which use various electron acceptors e.g. CO2, NO3- , SO42- instead of O2. The key processes are similar in soil, sediment or anaerobic digester where cellulolytic bacteria produce the enzyme to depolymerize cellulose to glucose and some sugars. These sugars will be fermented by cellulolytic bacteria and yield as CO2, H2, and organic acids such as acetate, propionate, butyrate and alcohols. These CO2, H2 and organic acids will not be released to environment but instead immediately consumed by methanogenic bacteria to change to CH4 (Leschine, 1995). 6.3 Material and method 6.3.1 Root exudates from hydroponic culture The Aristolochia plant species: Aristolochia littoralis, Aristolochia tribota and Aristolochia Macrophylla were plant to collect the root exudates. Aristolochia plants were removed from pots, cleaned the roots with deionized water and transfered to glass jar filled with aerated 50%strength Knop’s solution for hydroponic cultivation. These plants were grown under the light in 14/10 hr day/night cycle. To collect root exudates, the Aristolochia plants were thoroughly rinse with deionized water and move to beaker which filled 200 ml of 50% strength Knop’s solution as shown in Figure 6.1. The exudates were expected to release and mixed with 179 nutrient solution. At day of taking sample, 15 ml nutrient solution were collected, centrifuge at 7,500 rpm for 20 min to remove the root debris and then 10 ml of supernatants were taken to increase concentration by condensation method where the supernatant were dried with evaporation unit under air stream as shown in Figure 6.2. The residual in the tube was redissolved with 0.5 ml 70% methanol (20 fold concentrate). The re-dissolved samples were centrifuge in eppendorf tube at 12,000rpm for 3 min to remove precipitated salt before analyze with HPLC-DAD. The samples were taken in every 2 day for consecutive 12 day cultivation. The 15 ml of fresh nutrient solution was refilled to balance the taken volume sample. Figure 6.1 The experimental setup to collect root exudates in hydroponic culture 180 Figure 6.2 The condensation of collected root exudates by evaporation unit 6.3.2 Root exudates from sands culture After collect root exudates from hydroponic culture, Aristolochia plant: Aristolochia littoralis, Aristolochia tribota and Aristolochia Macrophylla were transfered to beaker which contain 150 g pre-treatment sand. The experimental setup for sand culture showed in Figure 6.3. To prepare pre-treatment sand, F-65 Ottawa Silica sand was washed with deionized water for 2 times and dried in the fume hood. Then the dried sand were autoclaved (10 min heat and 10 min dried cycle) to inhibit microbial activity. The 50% knop nutrient solution was used to irrigate in everyday to keep water level cover the sand surface. After 10 day of acclimation, the root exudates was started to collect. To collect the root exudates, 30ml of 50% Knop solution was added to the beaker and equilibrated for 1 hr to dissolve out all soluble compound from root and sand, then 15ml of leachate wascollected from top solution. These solutions were centrifuged to remove sand and any particulate and10 ml of supernatants was collected to increase concentration by condensation method and analyzed for AAs same as hydroponic cultivation. These root exudates was collected in every 2 day. 181 Figure 6.3 The experimental setup to collect root exudates in sand culture 6.3.3 Leaching from decomposition of Aristolochia Clematitis seeds 6.3.3.1 The decomposition in aerobic condition by activated sludge and soil The AAs leaching from decomposition of Aristolochia Clematitis seeds was conducted by batch experiment. Ottawa sand was used as solid media. Activated sludge and soils were used as sources of microorganisms. The sterile and non-sterile samples by autoclaving were used to determine the effect of biodegradation. Five bottles were prepared which consisted of bottle A contained 30 g of Ottawa sand and 30 ml pure water, bottle B and C contained 30 g of Ottawa sand, 25 ml pure water and 5ml of activated sludge, bottle D and E contained 30 g of Ottawa sand, 25 ml pure wate, and 1 gram of garden soil. The bottle A, B and D were sterilized by autoclaving 10 min cycle to inhibit microorganism activity. The bottle C and E were not autoclaved where the biodegradation by microorganism in activated sludge and soil expect to be occurred. The bottles were prepared in duplicated. Before incubation, 0.02 g of ground A. 182 Clematitis seeds was added to each bottle and then covered with sponge plug to prevent microbial contaminated from the air while the O2 can still pass. The prepared bottles are shown in Figure 6.4. The bottles were incubated by shaking in orbital shaker at 180 rpm under room temperature and keep out from the light. To collect the sample, 1 ml of solution were withdrawn from bottles and put into eppendorf tube to centrifuge at 12,000rpm for 3min, then the supernatants were collect to analyze AAs with HPLC-DAD. The samples were taken at interval time of 5 hr, 1 day, 2day, 4day, and every 2 days until 20 days. Figure 6.4 The prepared bottles for decomposition of A.Clematitis seeds in aerobic condition 183 6.3.3.2 The decomposition in anaerobic condition by activated sludge and soil To determine the decomposition of A.Clematitis seeds in anaerobic condition, five set of bottles were prepared similar to aerobic decomposition but the bottles were purged with the N2 gas before leaving in glove box under N2 atmosphere for 5 day to ensure the solution in anaerobic condition as shown in Figure 6.5. Activated sludge and soil were also used as sources of anaerobic microorganisms and were put in glove box for 5 days too. To collect the sample, the needle syringe was used to draw the solution from rubber septum cap for the time interval similar to aerobic experiment. Figure 6.5 The set of bottles prepared in glove box under N2 atmosphere 184 6.4 Results and discussion 6.4.1 Root exudates of Aristolochia plant from hydroponic and sand culture The analysis for AA I and II in root exudates collected from hydroponic and sand culture were shown in Table 6.2. The table shows that no AA I and II were found in all Aristolochia plant species (Aristolochia littoralis, Aristolochia tribota and Aristolochia Macrophylla ) root exudates for the whole experiment period (12 days). This indicated that root exudates of Aristolochia plants is not the main pathway that released AAs to the soil. However, because our culture system was not sterile, the microorganisms in the Aristolochia species plant root may contain from the original soil and degrade AAs during exudates collection periods. 185 Table 6.2 The AAs analysis in root exudates from three Aristolochia plants over 12 day periods Days Aristolochia Littoralis AA I Aristolochia Tribota AA II AA I Aristolochia Macrophylla AA II AA I AA II Hydroponic Sand Hydroponic Sand Hydroponic Sand Hydroponic Sand Hydroponic Sand Hydroponic Sand culure culture culure culture culure culture culure culture culure culture culure culture 2 NF NF NF NF NF NF NF NF NF NF NF NF 4 NF NF NF NF NF NF NF NF NF NF NF NF 6 NF NF NF NF NF NF NF NF NF NF NF NF 8 NF NF NF NF NF NF NF NF NF NF NF NF 10 NF NF NF NF NF NF NF NF NF NF NF NF 12 NF NF NF NF NF NF NF NF NF NF NF NF NF is not found 186 6.4.2 Leaching from the decomposition of Aristolochia Clematitis seeds 6.4.2.1 The decomposition in aerobic condition by activated sludge and soil The AAs concentration released during the Aristolochia Clematitis seeds decomposition with/without sources of microorganisms as function of incubation time are shown in Figure 6.6. The figures show that AA I and II start to release from all bottles in the first few days after apply the A.Clematitis seeds. For bottle A, B and D where solutions were autoclaved to inhibit microbial activity, the AAs continued to release until they reach the maximum concentration after 12 day. The concentration of AA I which leached from A.Clematitis seed was more than AA II about 10 times which according to the previous report about AAs content in Aristolochia herbal plants (Chan et al., 2007). Not similar to bottle A, B and D, bottle C and E where solutions were not sterilized showed the degradation of AA I and II after 8 days and were completely consumed in 12 days for activated sludge bottle and 18 day for soil bottle. This indicated the aerobic microorganism in activated sludge and soil can degrade AA I and II but microorganisms in activated sludge may degrade AAs faster than soil. Figure 6.7 (c) showed the microbe in activated sludge grown by adhering on A.Clematitis seeds which is typical for cellulolytic bacteria. This evidence cannot be seen in bottle B where the microbe had been killed as shown in Figure 6.7(b). The chromatogram of AA I and II in leachate from bottle A analyzed by HPLC-DAD is shown in Figure 6.8. 187 a) AA I 6.00 Concentration (µg/ml) 5.00 4.00 A 3.00 B 2.00 C D 1.00 E 0 2 4 6 8 10 12 Days 14 16 18 20 22 b) AA II 6.00 Concentration (µg/ml) 5.00 A 4.00 B C 3.00 D E 2.00 1.00 0 2 4 6 8 10 12 Days 14 16 18 20 22 Figure 6.6 AA I (a) and II (b) leaching from the decomposition of A. Clematis seeds in aerobic condition from bottle set A, B, C, D and E as a function of time : A=no microorganism + sterile, B=activated sludge + sterile, C=activated sludge + non sterile, D =soil + sterile, E=soil + non sterile 188 b) a) c) Figure 6.7 The growing of microorganism from activated sludge in bottle A (a) , bottle B (b) and botle C (c) in aerobic decomposition experiments Figure 6.8 The chromatogram of AA I and II leaching from A.Clematitis (bottle set A) analyzed by HPLC-DAD 189 6.4.2.2 The decomposition in anaerobic condition by activated sludge and soil The AAs leaching during decomposition under anaerobic and sterile and non-sterile condition are shown in Figure 6.9. Likewise the aerobic decomposition, the bottles which were autoclaved to kill microorganism (bottle A, B and D) show continually release AAs to solution. However, for the bottles which were not autoclave (bottle C and E), the AAs concentration found to decrease with time until they deplete in 5 days which may be faster than aerobic decomposition. The decreasing AAs concentration in bottles that contain soil was expected from the adsorption of AAs into the soil. These results indicated that AAs can be easily degraded even in anaerobic condition too. a) AA I 7.00 Concentration (µg/ml) 6.00 A 5.00 B C 4.00 D E 3.00 2.00 1.00 0 2 4 6 8 10 12 Days 14 16 18 20 22 Figure 6.9 AA I (a) and II (b) leaching from the decomposition of A. Clematis seeds in anaerobic condition from bottle set A, B ,C, D and E as a function of time : A=no microorganism + sterile, B=activated sludge + sterile, C=activated sludge + non sterile, D =soil + sterile, E=soil + non sterile 190 Figure 6.9 (cont’d) b) AA II 7.00 A B 6.00 Concentration (µg/ml) C 5.00 D E 4.00 3.00 2.00 1.00 0 6.5 2 4 6 8 10 12 Days 14 16 18 20 22 Conclusion The root exudates collecting from different Aristolochia plants by both hydroponic and sand culture did not contain AAs which indicates root exudates is not the main pathway that released AAs to soil. Instead, high concentration of AAs was found in leachates from the decomposition of A.Clematitis seeds and this should be the main releasing pathway of AAs to soil if seeds suspend in water. However, the released AAs were found to be degraded by both aerobic and anaerobic microorganisms which may suggest that AAs are not persistent chemicals in environment. 191 REFERENCES 192 REFERENCES Aprill, W., & Sims, R. C. (1990). EVALUATION OF THE USE OF PRAIRIE GRASSES FOR STIMULATING POLYCYCLIC AROMATIC HYDROCARBON TREATMENT IN SOIL. Chemosphere, 20(1-2), 253-265. doi: 10.1016/0045-6535(90)90100-8 Berg, B., & Mcclaugherty, C. (2014). Plant Litter: Decomposition, Humus Formation, Carbon Sequestration (3rd ed.): Springer. Biondini, M., Klein, D. A., & Redente, E. F. (1988). CARBON AND NITROGEN LOSSES THROUGH ROOT EXUDATION BY AGROPYRON-CRISTATUM, AGROPYRONSMITHII AND BOUTELOUA-GRACILIS. Soil Biology & Biochemistry, 20(4), 477482. doi: 10.1016/0038-0717(88)90061-2 Bragazza, L., Buttler, A., Siegenthaler, A., & Mitchell, E. A. D. (2009). Plant Litter Decomposition and Nutrient Release in Peatlands. Carbon Cycling in Northern Peatlands, 184, 99-110. doi: 10.1029/2008gm000815 Buehler, C. P. (2010). Soil modification and potential allelopathy: an investigation into how the invasive CASUARINA EQUISETIFOLIA L. (Australian pine) modify their environment. (Master of Science in Geosciences ), Mississippi State University. Chan, W., Lee, K.-C., Liu, N., & Cai, Z. (2007). A sensitivity enhanced high-performance liquid chromatography fluorescence method for the detection of nephrotoxic and carcinogenic aristolochic acid in herbal medicines. Journal of Chromatography A, 1164(1-2). doi: 10.1016/j.chroma.2007.06.055 Dakora, F. D., & Phillips, D. A. (2002). Root exudates as mediators of mineral acquisition in low-nutrient environments. Plant and Soil, 245(1), 35-47. doi: 10.1023/a:1020809400075 Edberg, N., & Hofsten, B. V. (1975). CELLULOSE DEGRADATION IN WASTEWATER TREATMENT. Journal Water Pollution Control Federation, 47(5), 1012-1020. Hasanuzzaman, M. Allelopathy. http://hasanuzzaman.weebly.com/uploads/9/3/4/0/934025/allelopathy.pdf Hranjec, T., Kovac, A., Kos, J., Mao, W. Y., Chen, J. J., Grollman, A. P., & Jelakovic, B. (2005). Endemic nephropathy: the case for chronic poisoning by Aristolochia. Croatian Medical Journal, 46(1), 116-125. Inderjit. (1996). Plant phenolics in allelopathy. Botanical Review, 62(2), 186-202. doi: 10.1007/bf02857921 Kuiters, A. T., & Sarink, H. M. (1986). LEACHING OF PHENOLIC-COMPOUNDS FROM LEAF AND NEEDLE LITTER OF SEVERAL DECIDUOUS AND CONIFEROUS 193 TREES. Soil Biology & Biochemistry, 18(5), 475-480. doi: 10.1016/00380717(86)90003-9 Leschine, S. B. (1995). CELLULOSE DEGRADATION IN ANAEROBIC ENVIRONMENTS. Annual Review of Microbiology, 49, 399-426. doi: 10.1146/annurev.mi.49.100195.002151 Martin, J. K. (1977). EFFECT OF SOIL-MOISTURE ON RELEASE OF ORGANIC-CARBON FROM WHEAT ROOTS. Soil Biology & Biochemistry, 9(4), 303-304. doi: 10.1016/0038-0717(77)90039-6 Matsumoto, H., Okada, K., & Takahashi, E. (1979). EXCRETION PRODUCTS OF MAIZE ROOTS FROM SEEDLING TO SEED DEVELOPMENT STAGE. Plant and Soil, 53(12), 17-26. doi: 10.1007/bf02181875 Pavlovic, N. M., Maksimovic, V., Maksimovic, J. D., Orem, W. H., Tatu, C. A., Lerch, H. E., . . . Paunescu, V. (2013). Possible health impacts of naturally occurring uptake of aristolochic acids by maize and cucumber roots: links to the etiology of endemic (Balkan) nephropathy. Environmental Geochemistry and Health, 35(2), 215-226. doi: 10.1007/s10653-012-9477-8 Phillips, R. P., Erlitz, Y., Bier, R., & Bernhardt, E. S. (2008). New approach for capturing soluble root exudates in forest soils. Functional Ecology, 22(6), 990-999. doi: 10.1111/j.1365-2435.2008.01495.x Prikryl, Z., & Vancura, V. (1980). ROOT EXUDATES OF PLANTS .6. WHEAT ROOT EXUDATION AS DEPENDENT ON GROWTH, CONCENTRATION GRADIENT OF EXUDATES AND THE PRESENCE OF BACTERIA. Plant and Soil, 57(1), 69-83. doi: 10.1007/bf02139643 Reilley, K. A., Banks, M. K., & Schwab, A. P. (1996). Dissipation of polycyclic aromatic hydrocarbons in the rhizosphere. Journal of Environmental Quality, 25(2), 212-219. Rietveld, W. J., Schlesinger, R. C., & Kessler, K. J. (1983). ALLELOPATHIC EFFECTS OF BLACK WALNUT ON EUROPEAN BLACK ALDER COPLANTED AS A NURSE SPECIES. Journal of Chemical Ecology, 9(8), 1119-1133. doi: 10.1007/bf00982216 Sahoo, U., Jeeceelee, L., Vanlalhriatpuia, K., Upadhyaya, K., & Lalremruati, J. (2010). Allelopathic Effects of Leaf Leachate of Mangifera indica L. on Initial Growth Parameters of Few Homegarden Food Crops. World Journal of Agricultural Sciences, 6(5), 579-588. Sahoo, U. K., Vanlalhriatpuia, K., Upadhyaya, K., & Roy, S. (2011). Effects of leaf extracts of common home garden trees on food crops. Allelopathy Journal, 28(1), 123-134. 194 Siegert, I., & Banks, C. (2005). The effect of volatile fatty acid additions on the anaerobic digestion of cellulose and glucose in batch reactors. Process Biochemistry, 40(11), 34123418. doi: 10.1016/j.procbio.2005.01.025 Singh, H., Batish, D., & Kohli, R. Allelopathic effect of Leucaena leucocepahala on Zea Mays. Journal of Tropical Forest Science, 11(4), 801-808. Singh, J. S., & Gupta, S. R. (1977). PLANT DECOMPOSITION AND SOIL RESPIRATION IN TERRESTRIAL ECOSYSTEMS. Botanical Review, 43(4), 499-528. Verachtert, H., Ramasamy, K., Meyers, M., & Bevers, J. (1982). INVESTIGATIONS ON CELLULOSE BIODEGRADATION IN ACTIVATED-SLUDGE PLANTS. Journal of Applied Bacteriology, 52(2), 185-190. doi: 10.1111/j.1365-2672.1982.tb04839.x von Kiparski, G. R., Lee, L. S., & Gillespie, A. R. (2007). Occurrence and fate of the phytotoxin juglone in alley soils under black walnut trees. Journal of Environmental Quality, 36(3), 709-717. doi: 10.2134/jeq2006.0231 Wadhwa, K., & Narula, N. (2012). A novel technique to collect root exudates from mustard (Brassica juncea). Journal of Basic Microbiology, 52(5), 613-615. doi: 10.1002/jobm.201100407 Yamane, A., Nishimura, H., & Mizutani, J. (1992). ALLELOPATHY OF YELLOW FIELDCRESS (RORIPPA-SYLVESTRIS) - IDENTIFICATION AND CHARACTERIZATION OF PHYTOTOXIC CONSTITUENTS. Journal of Chemical Ecology, 18(5), 683-691. doi: 10.1007/bf00994606 Yoshitomi, K. J. (2001). Influence of root exudates on soil microbial population. University of Cincinnati. Yu, J. Q., & Matsui, Y. (1994). PHYTOTOXIC SUBSTANCES IN ROOT EXUDATES OF CUCUMBER (CUCUMIS-SATIVUS L). Journal of Chemical Ecology, 20(1), 21-31. doi: 10.1007/bf02065988 195 SECTION II Biogeochemical Modeling of Fate and Transport of Copper from Mining Waste in Torch Lake Sediment Chapter 7 Introduction 7.1 The problem of concern and Torch Lake history Lake contaminated with heavy metal from mining waste is one of major environmental problems which deteriorate water resource. It raises the metal concentration in sediment, overlying water and also in aquatic organism to exceed safe level for consumption and decreasing number of fish population as well (Adams, Atchison, & Vetter, 1980; Moore & Sutherland, 1981). Torch Lake, located in Houghton County, MI, had been impacted by copper mining for a hundred years between 1867-1968. Two hundred million tons of copper waste had been deposited on shoreline and the basin which was estimated as 20% of lake volume. These wastes were also discharged to Lake Superior by Kewenaw waterway which connect with Torch Lake and portage Lake (Lopez & Lee, 1977). Moreover, after ceasing of mining activities, there was a large spill of cupric ammonium carbonate in October, 1971 and June, 1972 which increased copper concentration in Torch lake water from 40 µg/ml at the surface to 100 µg/ml at the bottom. The Cu-bearing ore that lay in bottom sediment and shoreline act as a copper reservoir continual providing to lake water. Those reported dissolved Cu concentrations in Torch 196 Lake were exceeded the criteria of US EPA 30 µg/l and quite persist throughout annual cycle. The copper criteria range of lethal dose and teratogenicity to egg of sensitive species fishes and amphibians is 5-10 µg/L. The recommend a maximum allowable value for copper concentration for aquatic life is 12-43 µg/L (USGS, 1997). Even toxic level concentration, substantial fish population were reported in Torch Lake which may result from the sorption of copper onto iron and manganese hydrous oxides resulting in non-toxic form (Lopez & Lee, 1977). Because of significant copper concentration above background level, high incidence of undetermined cause fish tumor and degradation of benthos in bottom sediment. In 1986, Torch Lake was added to US EPA National Priorities List and declared as Area of Concern and Superfunds site for 3 reason: Fish Tumor and Other Deformities, Restritions on Fish and Wildlife Consumption, and Degradation of Benthos (MDEQ, 2007). US EPA attempted to delist Torch Lake from area of concern by started remediation on 1998. The stamp sand and slags deposits were stabilized through covering by the clean soil and vegetative to reduce metal erosion by wind and water from shoreline into the lake. However, for the sediment , US EPA determine that the contamination had no unacceptable threat to human health and select “no action” or natural attenuation as remediation step except long term monitoring and reviewing in every five years (MDEQ, 2007). 7.2 Torch Lake study site Torch Lake is natural deep lake located on eastern side of Keweenaw Peninsula, Houghton County. It has the surface area around 20.5 km2. The average depth is 15 m and the maximum depth is 37 m. It has 9.3 km long and 2.2 km wide at the widest point, near the middle of lake. It contains two distinct basins, north and south. Torch lake water flows to the Portage 197 Lake which located on the south and continually flows to Lake Superior on the west and Kaweenaw Bay on the east via Kaweenaw waterway as shown in Figure 7.1. Torch Lake is oligotrophic (low nutrient content and low algal production) which has the water residence time about 1 year. The watershed was mostly covered by hard woods and limited farming activities (Cusack & Mihelcic, 1999; J.D. Fett, 2003; Jeong, Urban, & Green, 1999; Lopez & Lee, 1977). Figure 7.1 Map showing location of Torch Lake, Keweenaw Peninsula, Michigan (Fett, 2003) 7.3 The copper state in Torch Lake sediment Although the remediation of shoreline had been greatly improved the water quality and clarity of lake and some sites were delisted from Superfund, many studies indicated that total copper concentration in recent deposited sediment had increased and has higher concentration than mine tailing itself (J.D. Fett, 2003; McDonald, Urban, Barkach, & McCauley, 2010). These evidence showed Torch Lake sediment cannot be remediated by natural processes and raise the question how copper concentration in post sediment increased. 198 The Torch lake sediment divide into 2 clearly separation section, the mine tailing sediment which has grayish-purple color and post mining sediment which has black color as shown in Figure 7.2. The mine tailing sediment mainly consists of clay-sized stamp sand particles containing solid copper concentration more than 1,000 mg/L (Cusack & Mihelcic, 1999). It is extremely fine grain and loose with porosities about 80%. This sediment has 8 m height which reflected the high sedimentation rate in that mining period (14-28 mm/yr). This mine tailing was overlaid by 3 – 7 cm natural sediment after stop mining operation which had much lower sediment accumulation rate 5 – 9 mm/yr. It can refer as cap layer which is typical lake bottom sediments. This porosity of the cap layer is greater than 90% (Konstantinidis et al., 2003). The sedimentation flux rate which measured by sediment trap and the sediment accumulation which calculated by radioisotope dating indicated the re-suspension of surface sediment was not significant. The organic matter in post-mining layer was high about 10-18% whereas organic matter in mine tailing was lower about 1.9-2.7%. The solid copper concentration in sediment showed clearly the different between mine tailing and post-mining sediment as shown in Figure 7.3. The post-mining sediment was 1,800 – 2,800 mg/kg whereas underlying mine tailing was half lower 800 – 1,300 mg/kg,(J.D. Fett, 2003; McDonald et al., 2010). The mean copper concentration of particles in overlay water was 790 – 1,090 mg/kg which was nearly to shoreline tailing deposit. These data showed that settling particle contained less copper than post-mine tailing sediment which means the sedimentation of erosion shoreline particle is not the only process that control copper concentration in sediment but it also has other sources that supply the copper to surface sediment. Moreover, the stamp sand on the shoreline had been covered by soil and vegetation which would greatly reduced copper particle input to the 199 lake. Not only the copper, some metals e.g. Zn, and Pb had similar pattern in the sediment (Konstantinidis et al., 2003). Figure 7.2Torch Lake sediment core (McDonald & Urban) Figure 7.3 Bulk density (open circles) and solid phase copper concentration (solid circles) profiles of Torch lake sediment at different water depth (a) 10 m (b) 13 m and (c) 20 m (modified from C. P. McDonald et al., 2010) 200 7.4 State of recovery The Gratiot Lake located on eastern side of Keeweenaw peninsula was used as reference lakes to study of the recovery state of Torch Lake as shown in Figure 7.4. It has the same bedrock geology, surficial geology and had no impact from mining activities. The plot of Cu/Zn ratio of Torch lake sediment showed that they are largely over than ratio from Gratiot Lake but close to ratio of stamp sand which indicate the continual input from shore line stamp sand. However, the plot of other ions such as the ratio of Co/Zn showed the new sediment from local watershed entering to the Torch Lake. This indicate the state of recovery of Torch lake except the copper (J.D. Fett, 2003). Figure 7.4 The location of Gratiot Lake, Torch Lake and Portage Lake (Fett, 2003) 201 7.5 The redox state The redox condition in Torch lake sediment can be presented by the concentration profile of redox sensitive elements such as Fe and Mn as shown in Figure7.5. This figure shows that the oxidizing condition had occur at surface sediment due to peak concentration of Fe and Mn. In deeper sediment, their concentration decrease and remain constant where indicate reducing condition. However, the copper profile differed from Fe and Mn, it was relatively uneffect by changing in redox condition where concentration of copper should be higher in reducing condition at deep sediment (Joel D Fett, 2003). Figure 7.5 Normalized concentration of Fe, Mn and Cu in Torch Lake sediment (Fett, 2003) 7.6 The evidence of microbial copper sequestration From the long history of dumping high concentration copper waste, Torch lake sediment is extremely environment for development and evolution of microbial communities. However, the study of structure of community can identify at least 20 phylotypes in there. These isolated 202 strains are classified in only two genera , Ralstonia and Arthrobacter bacteria, that can aerobically grow in high level of Cu ( >800 ppm) and resistance to other metals, e.g. Zn , Cd and Ni as well (Konstantinidis et al., 2003). The Ralsotia species can turn the colonies to green hue when plating with CuSO4 which indicated that the cells can sequestrate copper. This sequestration can be seen by scanning electron microscopy (SEM) images as shown in Figure 7.6. Cells grown in presence of copper will have extracellular material appeared on outer envelope whereas the smooth surface had been observed for cells grown in absence of copper (Konstantinidis et al., 2003). These blebs were expected to contain copper excreted from the cells. These evidences indicated that microbial sequestration could be responsible of high copper concentration in Torch lake surface sediment. A B C D Figure 7.6 Scanning electron microscopy of Ralstonia isolated from Torch lake grown on copper free agar (A) and (C); grown on copper-supplemented agar (B) and (D) (modified from Konstantinidis et al., 2003) 203 From thin section of transmission electron microscopy (TEM) images (Figure 7.7), intracellular vesicle was observed in non copper contained media cells (Figure 7.7 C) whereas the greater thickness of outer membrane was observed in copper contained media cells (Figure 7.7 D). Figure 7.7 F shows the accumulation of copper at outer envelope. However, the specific location of bound copper was not clear between in outer membrane, periplasm or inner membrane (F. Yang et al., 2010). A C E B F D Figure 7.7 TEM and SEM of R. pickettii strain 12 J grown in absence copper (TEM: A, C ; SEM: E) and presence copper (TEM: B, D ; SEM: F) (modified from Yang et al., 2010) The binding isotherm studies between viable and heat-killed cells of Ralstonia showed that both cells can bind to copper but the viable cells can bind 5-6 times more than heat-killed cells which indicated the copper sequestration was more effective by viable cells (F. Yang et al., 2010). Linear isotherm was described for binding of heat-kill cells whereas the sorption isotherm of viable cells cannot be described by linear, Freundlich or Langmuir model. The study of X-ray absorption spectroscopy showed the valance state of bound copper was Cu(II) and associated 204 with oxygen, nitrogen and carbon ligands. This indicated that the bound copper was not a precipitated (hydr)oxide minerals but associated with organic material instead (F. Yang et al., 2010). The genome sequencing analysis data showed the cell chromosome were enriched with metal resistance and transporter genes e.g. copper binding protein, mercury resistance operon, iron permease, cop ABCD orperon (copper resistance), Czc(cadmium, zinc, cobalt) operon, metal translocating P-type ATPases and heavy metal signal/sensor proteins. This founding of the efflux systems showed cells had effort to transport copper out of their body and those genes revealed the rapid adaption to environment by gene duplication and horizontal transfer. This indicated Ralstonia has highly adaptive capacity to modify their genome to correspond the high concentration of copper and may be the dominant role of their evolution (F. Yang et al., 2010). The capability of Ralstonia copper sequestration may response for the persisting of high copper concentration in Torch Lake surface sediment. 7.7 Torch Lake Sediment composition and mineralogy The analysis of copper in stamp sand deposited along the lakeshores and wetland in Torch Lake area represented the mine tailing in bottom sediment and in post-mining surface sediment showed the different type of minerals. The X-ray diffraction analysis of mining waste showed that Cuprite (Cu2O), Tenorite (CuO), Malachite(Cu2CO3(OH)2), Chalcopyrite(CuFeS2) presented as copper minerals and Calcite (CaCO3), Quartz(SiO4), Hematite (Fe2O3), Orthoclase(KAlSi3O8) and Sanidine ((K,Na)(Si,Al)4O8) presented as the other major minerals. The sequential extraction techniques (SET) showed the carbonates and oxides were main fraction of copper (45-60%) in mining waste whereas the copper carbonates and copper-organic matter were main fractions (61%) in post mining sediment. The changing forms of copper minerals was expected from weathering processes. Some studies proposed that the copper associated with 205 iron/manganese oxides in mine tailing was reduced and released to solution and it was bound to organic matter on the post mining surface sediment (Jeong, 2003; Jeong et al., 1999). 7.8 Hypothesis and Approach From the evidence of raising and persisting of copper in top sediment, I hypothesize that the copper solid phase in top sediment contributed from the dissolution of copper in mining tailing by microbial mediation or TEAP (Terminal Electron Acceptor process) to porewater which diffused upward and was sequestrated by organic matter and bacteria on post surface sediment as shown in Figure 7.8. Surface water Cu sequestration by organic matte/ bacteria Post sediment Mine Tailing Sediment Cu Diffusion Cu dissolved by reductive dissolution Figure 7.8 Model of copper upward diffusion from mine tailing sediment and sequestration by organic matter and bacteria at the surface sediment. To understand the mobility of copper in Torch lake sediment and decide whether natural attenuation can remediate, the biogeochemical modeling was selected as tool to quantify the process that control the fate and transport of copper in the Torch Lake sediment. Because metal behavior in sediment is a result from combining between physical, chemical and biological processes such as adsorption, dissolution/precipitation of minerals, redox reaction, microbial 206 mediated reaction and advection/diffusion, this study combined these processes into the geochemical modeling through mathematical equation. However, the accuracy of the models are highly depend on estimating model parameter obtained from experimental data or literature sources (Steefel & Van Cappellen, 1998). Phreeqc (Parkhurst & Appelo, 2013) is one of widely used geochemical modeling programs that contains equilibrium thermodynamic database of several minerals. It has capabilities to calculate the solubility and speciation of substances. It also has the feature of kinetic control reaction, surface complexation and ion exchange. In addition, Phreeqc also can handle transport simulation by advection or diffusion in porous media. So, in this study, Phreeqc was used to perform a variety of calculation to conduct the copper reactive transport process in Torch Lake sediment. The results from the models were used to describe the existing copper profile in sediment to understanding the cause of persistent and alsothe implication for the future remediation. 207 REFERENCES 208 REFERENCES Adams, T. G., Atchison, G. J., & Vetter, R. J. (1980). THE IMPACT OF AN INDUSTRIALLY CONTAMINATED LAKE ON HEAVY-METAL LEVELS IN ITS EFFLUENT STREAM. Hydrobiologia, 69(1-2), 187-193. doi: 10.1007/bf00016549 Cusack, C. C., & Mihelcic, J. R. (1999). Sediment toxicity from copper in the Torch Lake (MI) Great Lakes Area of Concern. Journal of Great Lakes Research, 25(4), 735-743. Fett, J. D. (2003). Assessing Recovery of Anthropogenically Disturbed Lakes Using Reference Systems and Multi-elemental Techniques. (M.S.), Michigan State University. Jeong, J. (2003). Solid-phase speciation of copper in mine wastes. Bulletin of the Korean Chemical Society, 24(2), 209-218. Jeong, J., Urban, N. R., & Green, S. (1999). Release of copper from mine tailings on the Keweenaw Peninsula. Journal of Great Lakes Research, 25(4), 721-734. Konstantinidis, K. T., Isaacs, N., Fett, J., Simpson, S., Long, D. T., & Marsh, T. L. (2003). Microbial diversity and resistance to copper in metal-contaminated lake sediment. Microbial Ecology, 45(2), 191-202. doi: 10.1007/s00248-002-1035-y Lopez, J. M., & Lee, G. F. (1977). ENVIRONMENTAL CHEMISTRY OF COPPER IN TORCH LAKE, MICHIGAN. Water Air and Soil Pollution, 8(4), 373-385. McDonald, & Urban. Modeling Copper Transport in the sediments of Torch Lake, Houghton County, MI. from http://www.mtcws.mtu.edu/pdf/CWSPoster_1stPlace_McDonald_Cory.pdf McDonald, C. P., Urban, N. R., Barkach, J. H., & McCauley, D. (2010). Copper profiles in the sediments of a mining-impacted lake. Journal of Soils and Sediments, 10(3), 343-348. doi: 10.1007/s11368-009-0171-0 MDEQ. (2007). The Michigan Department of Environmental Quality Biennial Remedial Action Plan Update For Torch Lake Area of Concern. from http://www.epa.gov/glnpo/aoc/trchlke/2007_TorchLkUpdate.pdf Moore, J. W., & Sutherland, D. J. (1981). DISTRIBUTION OF HEAVY-METALS AND RADIONUCLIDES IN SEDIMENTS, WATER, AND FISH IN AN AREA OF GREAT BEAR LAKE CONTAMINATED WITH MINE WASTES. Archives of Environmental Contamination and Toxicology, 10(3), 329-338. Parkhurst, D. L., & Appelo, C. A. J. (2013). Description of Input and Examples for PHREEQC Version 3—A Computer Program for Speciation, Batch-Reaction, One-Dimensional 209 Transport, and Inverse Geochemical Calculations. U.S. Geological Survey, Denver, Colorado. Steefel, C. I., & Van Cappellen, P. (1998). Reactive transport modeling of natural systems Preface. Journal of Hydrology, 209(1-4), 1-7. USGS. (1997). Copper Hazard to Fish, Wildlife and Invertebrates: A Synoptic review. from http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA347472 Yang, F., Pecina, D. A., Kelly, S. D., Kim, S. H., Kemner, K. M., Long, D. T., & Marsh, T. L. (2010). Biosequestration via cooperative binding of copper by Ralstonia pickettii. Environmental Technology, 31(8-9), 1045-1060. doi: 10.1080/09593330.2010.487290 210 Chapter 8 Literature reviews 8.1 Copper speciation, control mechanism and transport in natural water Copper in aquatic environment can present in three phases: the aqueous phase (free ion and soluble complex substances), solid phase (mineral, adsorbed on particulate/colloid) and biological phase (adsorbed to microorganism). However, most of copper present in natural water are in the solid form. Dissolved coppers can present in both free ion state or form the complexes with anions e.g. sulfide, chloride and bicarbonate, phosphate and cyanide which highly reduced copper toxicity to aquatic organism (Flemming & Trevors, 1989). Some anions such as silicate, sulfate and nitrate are considered as non-complex species. The valence state can be I, II, and III but Cu(II) is most common of soluble species in normal oxidation state. Cu(I) is most presented in insoluble phase and Cu(III) are unstable in aqueous media (Nriagu, 1979). There are three main mechanisms that control copper speciation and forms in natural water: complexation, precipitation and adsorption which also control mobility and bioavailability in environment. 8.1.1 Complexation The relative stabilities of Cu(I) and Cu(II) are depended on the complex formation constant and type of ligand presence in water. The Cu(I) is most stable with forming the complexes with some ligands and become as insoluble compounds such as CuI, CuCl, CuBr, Cu2O Cu2S . Copper (II) is strong complexing agent. Most of Cu(II) salts are soluble and these 211 Cu(II) are readily to form complex with other ligands presented in water. The major complexes are hydroxyl species (CuOH+, Cu(OH)20, Cu2(OH)22+ ) and carbonate species (CuCO30, Cu(CO3)22- ) which depend on pH and hardness of solution. The increasing water hardness will highly reduce free Cu(II) (Stiff, 1971). The other important complex species of Cu(II) are sulfide (HS-), phosphate (PO43-), chloride (Cl-) and ammonia (NH3). However, the halide and cyanide are more stable with Cu(I) more than Cu(II). In natural water, Cu(II) can form strong complexes with many organic ligands, typically composed of oxygen ,nitrogen and sulfur doner atom. The stability constant of organic complexes are higher than inorganic complexes in many order of magnitude. The common organic copper complexes found in natural water included cyanide, amino acids, polypeptides and humic substances. The high degree of organic complexation were reported to cover 75-99% of dissolved copper in river water (Stiff, 1971). Some of stability constants of inorganic and organic Cu complexes of Cu(I) and Cu(II) show in Table 8.1. These stability constants are used to calculate copper speciation in natural water. The examples of copper-organic complex were shown in Figure 8.1. The complexation of carbonate and hydroxide are the major forms of copper speciation in fresh water. The monocarbonate complex (CuCO30 ) is major form at neutral pH indicating the strong binding between Cu(II) and carbonate ion. When pH increase, the hydroxide complex (Cu(OH)20) will be dominated and prevented copper from precipitation. However, when presence of organic ligand such as nitrilotriacetic acid (NTA), it can form extremely stable complex with copper and dominant over carbonate and hydroxide complex at low and neutral pH. The complexation is important process that regulate copper concentration in water which affect to precipitation and/or adsorption process. 212 Table 8.1 Stability constant of complex formation of copper (Nriagu, 1979; Stiff, 1971) Complexes Log of stability constants Cu2+ + OH- = CuOH+ 6.1 2Cu2+ + 2OH- = Cu2(OH)22+ 17.7 Cu2+ + CO32- = CuCO30 6.73 Cu2+ + Cl- = CuCl+ 0.5 Cu2+ + F- = CuF+ 1.23 Cu2+ + SO42- = CuSO40 2.3 Cu2+ + 3HS- = Cu(HS)3- 26.5 Cu2+ + H+ + PO43- = Cu(HPO4)0 16.6 Cu2+ + NH3 = Cu(NH3)2+ 5.8 Cu2+ + Glycine = Cu(Glycine) 8.1 Cu2+ + Nitrilotriacetic acid (NTA) = CuNTA 12.7 Cu2+ + 4Citric acid = Cu(Citric acid)4 13.2 Cu2+ + Fluvic acid = Cu(Fluvic acid) 8.69(pH 5) Cu+ + 2Cl- = CuCl2- 5.5 Cu+ + 3Cl- = CuCl32- 5.7 Cu+ + NH3 = Cu(NH3)+ 5.5 Cu+ + 2NH3 = Cu(NH3)2+ 10.3 213 a) ) b) Figure 8.1 Calculated copper speciation in fresh water when absence (a) and presence (b) of organic chelation NTA (modified from Elder & Horne, 1978) 8.1.2 Precipitation/dissolution Next mechanism that control copper in natural water is precipitation which greatly reduced the soluble copper. Copper can react with inorganic ligands as a complexes which have very low solubility and precipitate as solid forms. In surficial natural water pH (6.5 -8) , the major form of precipitates included Cupric hydroxide (Cu(OH)2), Tenorite (CuO), Malachite (Cu2(OH)2(CO3)) and Azurite (Cu3(OH)2(CO3)2) (Sylva, 1976). Some solubility constants of copper minerals are shown in Table 8.2 Figure 8.2 also shows the stability of copper solid forms as a function of pH and pCO2. At the common level of pCO2 in soil and sediment (1