2" ‘1 I) ’ 2.- 31;} . ../ g h #3.}. S h: . .. m..%.....um~.... . rs... -flfi....fimn.u .. Gm-.. . tfiisizialF i . .15... .4 a . .. . .. .e m... .h ,. . .. f... 35%»... l. .5... an... . I . - HT. .. wakfirm. ....fl.......:._ 3.. 4 .. .... . ...l ‘1 .4! J- .. .ufiw...fimw.$m% S :i , Ea. 5.5... .1.» r 3.. 1|. . . . t . . i . . . . .35. .gmmmr... wavy..- F... 7& $5.»... .. 1.. 94%.. m.%.. 12...?) .tm....!. “by In .5 L b! "It... . .21.“... «m. #2 ha... . m... .mv Um. .. .n .. unique. 4 . 1.. u .9 ”1.3.95.3 ;.. .. 4;..de 1.... .. t. n... 4...... .. .13.... . .2. . . . . . . .. $5.3. _ 'n‘r. '. 4; J. J . : u." .. $1.11. It. .L1kl .1 :l . .. H. .p Isizm. . in Mariam—a.“ A ‘ ~‘V 2%.... $5.». . Q... 5%.... .. 1...... Lawrm ..-r _ .. tnfifflwe 1 3.3.1.3.? .hsr‘at 4.11 < 3.3.. . ~ 21...... $3.. , finch... “Aidié A... . 1 1......“ .. 41.-....“m. .... 1...... .1. .3. . 1n. . . .. . .. ,m...n......wvum.u....u...é.u£u. a. . ......n.u....m.uuww...r.i.v....... .. ., -nraur ..>...fu......1 ..... .m.....vaem...um~n..?fflk...u.fiwu .9. I .. . Lu» .0 :. V I . ' . Y '1 I V . ‘ t u . . 5331...».”gynamakflw. m...» . . . .. 5.... .1... Tuba... .- 3......”er .3... x g? 3.: j: "if f.-. nNIH‘h...-ltt . . . r... k... in 5. k: 2. .3.-- . A. .. a... .31.... 0.. . .. . .....di..i. .. .13-.. 21...... 113.1. 5...? I Mu. - .w... .3 .. -.. lvtt . L. E .h..wn....un?.n. i... - . .5 in»... ....nu...m.mw,t..1§.. . .5... . an. 5 MM... 4...... H. a... 5...... . - e. flung-Ft... fig .. . .... 3.... u...........3nm..b.. .zuflfimhawwm .. . . . . .. . .P.fiwW..w. i... t .ufivorfl huh 0.0. . 2. .. yuhwflfiflhflbg- fifihgflvwfinaflkhvm. ..... . 4...»... ukfmnSafimnikufifia auntknum... .55.... . . . ufiwnmlaufidhfiunfiummuwwmuvuarr1hfi .5... .ivfiwina...fls..fl .11.? ...L...f»l...,..w...n..wmnu.. hunt. .. War. flml.u.urflu.r.......ri.. . Win... . n r. . ..r..n.&.u.s...¢.n..u....., .31... .- 1 . n. ”.34.”...59rfiflhgaba. .1. , .2-.. 1.... . - . - .... ...u..u.mb...u+n.fifl...»..":2.u}nm.nr.flfi . .fimmfififlufim . .. a... --...r.....-i. . -..-szs(u..\$..lt .. .3». "and...” 3.-.“... lets-.. . -I... i. I. L I '7!”- fl. . . . awn-«m. o.» 7..."... r .51"? 1 1 . . -WN .1 ii... 5.... ... 3..-... u . u. u... .- - - . -- J... :33" ‘1 if t {:1 "Pi g. ; ‘ a 4‘ l .5 33*. . . h .1: 1 2 a .. ; 4.3.12.9! 91.1151: viii... . xls..|..i1.:....nux .».'..I”.rhn!?vn if! .9 5...... ..LL.§.I. r5... -3 -Awlno... LL ll . N... 1? Ir)...“ '5 [unwrahn .ufll‘. . r "Him: . {I . .u. .3... Ix- . .-......... .134“ s. 1.; «(Lufuu : 30...... u... f!...u1i{..,..v...r. 9.41. .1. .. .ilvl,~.1h\ul.ll{tltnh«1 ’2. TI‘ .. 1 (ll {51"} .5 l§l. .I 'I.‘ .i . K . it .3725}...on31 . 3.311.101 . . 2.. . . :1 .12).! . I! .lHfl...-...ls1-lt.: 0...... .4;l . . p: .35... . - 1.. .. 1-.. 4 . E . 1.11 . (:4; J11! \W-.|.u . I [‘15 . It‘s" IQ ( \I nJWI o. - ... u . . .. {at}... A!) II . . . .. s‘l\l . it .17“... .\l.! I I . I . . u..." Nets..!v..nx 1.3:... (.L....c1..l1..l| lite...’ . {4h 13.. n .l\.v:vn|u .rbl. .1LL....?..\ giitluvu 30.1011 . .. 4......umuiwadr :Vkr...o.fi.u.+.q..h U1 2...... . . H.4HNIJ-.AH..D.N.H .hfituu. .1..I......l...l.‘..4k it. 61.1“. all“ .. .. .H 0.5..- .1? ....r.r. u... 9...»... . 'ml.‘.¢ t I. 11;! v . .. I t... .r... it- HI. 0.... . A. .2. 1:11? .L 0 . (‘CIiiI’lnl . 1).. .25! 6.01. an)!!! . ll 1....ru“D|P)\. ,lilplw . I f-..\ .0’. {n.‘-l|'.1v r A.’\.«.. 3|...bb.v .3. .1: .9! I‘ll)? .0... .0. .. pl- 3 l: . 1|...5thl'l It). {thinlif .llvovu . . a. 21‘- I . . .. . . . . ... . . . -.\..¢. . .1. lut"'..|v-. . . l . . .ILI ... .lv‘ (.~ I! b l . It: I V II. 4 xi; . .. 1!.1... .- .. . .» va Pint} .1 . Lowleflw. 1...... n?....r..... rub. r........r.u£... .«n .. ... . la . . . H. . . THESlS LIB BIERAR 111111111111111111111111111111111111111 31293 01571 16 11111 This is to certify that the dissertation entitled COPPER ADSORPTION/DESORPTION IN PHOSPHATE- AND SLUDGE- TREATED OXISOLS: KINETICS AND EFFECTS OF AGING AND pH presented by Luiz Roberto G. Guilherme has been accepted towards fulfillment of the requirements for Ph.D. Crop and Soil Sciences/ Environmental Toxicology degree in J6me WW Major professor Date 7' '797 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 LIBRARY Michigan State Unlverslty PLACE II RETURN BOX to remove We checkout {tom your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE MSU Ie An Affirmative AdIoNEqnl Oppomnlty lnetltulon WWI COPPER ADSORPTION/DESORPTION IN PHOSPHATE- AND SLUDGE-TREATED OXISOLS: KINETICS AND EFFECTS OF AGING AND pH By Luiz Roberto Guimaraes Guilherme A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Crop and Soil Sciences Institute for Environmental Toxicology 1997 ABSTRACT COPPER ADSORPTION/DESORPTION IN PHOSPHATE- AND SLUDGE-TREATED OXISOLS: KINETICS AND AGING AND pH EFFECTS By Luiz Roberto Guimaraes Guilherme Copper mobility in the enviromnent typically is controlled by adsorption/desorption reactions. This research studied the kinetics and the effects of aging and pH upon Cu adsorption/desorption in phosphate- and sludge-treated A and B horizons of two Brazilian Oxisols. Phosphate fertilization is nearly ubiquitous in Oxisols. Sewage sludge application to agricultural land is increasing in Brazil. The time-dependence of Cu adsorption/desorption revealed that initial adsorption rate constants were up to 45% greater in P-treated, and up to 70% greater in sludge-treated than control soils. In P-treated soils, the fraction of Cu desorbed (Cum/Cum) was half that in control soils, and smaller in B than A horizons. Sludge treatment affected Cum/Cum somewhat less than did P treatment. Little Cu was desorbed fi'om P-treated soils after 15 min, whereas desorption occurred up to l d for sludge-treated, and up to 18 d for control soils. Increasing adsorption reaction time (aging 1h to 54 d) increased Cu sorption distribution coefficients (Km and Km); Km was 2 to 8 times greater for 54-d than l-h aging, whereas Kdeg increased 2- to 4-fold over the same aging period. Aging caused up to an 80% decrease in Cu desorption. Samples aged 3 d or less exhibited rapid desorption followed by readsorption. Km, and K,“ values and aging effects on K“, and Km followed the trends P- treated > sludge-treated > control, 50 > 150 W Cu, and A > B horizon. Increasing pH from 4.5 to 6.5 increased adsorption up to 3-fold and decreased Cum/Cum, to < 0.01. At a given pH, more Cu was adsorbed on A than B horizons, and on pretreated than control samples. Copper desorption from aged soils can be overestimated more than 2-fold by measuring desorption after a 24- or 72-h adsorption reaction, as is typical in laboratory experiments. The large increase in initial Cu adsorption and the decrease in Cum/Cum, for P- or sludge-treated soils is noteworthy as this will affect Cu mobility and availability in Oxisols. To all of those who have contributed to my education, especially to my parents Teresa and Roberto. iv ACKNOWLEDGEMENTS I would like to express my gratitude to CNPq (National Council for Scientific and Technological Development, Brazil) and to Universidade Federal de Lavras (UF LA), whose financial support made possible the completion of my degree. I am truly indebted to my major Professor, Dr. Sharon Anderson, for her guidance, support and friendship. I also thank Dr. Boyd Ellis, Dr. Lee Jacobs, and Dr. David Long, for their enlightening suggestions and assistance. I would like to acknowledge Michigan State University for providing me with the opportunity to participate in the multidisciplinary doctoral program in Environmental Toxicology offered by the Department of Crop and Soil Sciences and the Institute for Environmental Toxicology. The important support from my friends at the Environmental Geochemistry and the Environmental Soil Chemistry laboratories, particularly Ralph DiCosty and Jon Kolak, is gratefully appreciated. I thank my colleagues from the Department of Soil Science at UFLA, Lavras, MG, Brazil, for their encouragement. I also thank my friends fi'om the Brazilian Community Association at Michigan State University, especially my fellow students at the Department of Crop and Soil Sciences, Carlos Paglis, Carlos Silva, Djail Santos, José Cora, José Lima, Moacir Dias Jr., and Ricardo Balardin, for providing a home-like atmosphere during this long period we stayed far from our homeland. Special thanks goes to my wife, Cristina, and my daughter, Fernanda, for their moral support and understanding. TABLE OF CONTENTS LIST OF TABLES ........................................................................................................... viii LIST OF FIGURES ............................................................................................................. ix Chapter 1 GENERAL INTRODUCTION ............................................................................................ 1 Rationale ........................................................................................................................... 1 Origin of Cu in Oxisols ..................................................................................................... 1 Reactions of Cu in Oxisols ................................................................................................ 2 Kinetics of Soil Reactions ................................................................................................. 6 List of References .............................................................................................................. 8 Chapter 2 COPPER ABSORPTION/DESORPTION KINETICS IN PHOSPHATE- AND SLUDGE-TREATED OXISOLS ....................................................................................... 10 Abstract ........................................................................................................................... 10 Introduction ..................................................................................................................... l 1 Materials And Methods ................................................................................................... 14 Soil Material ................................................................................................................ 14 Adsorption/Desorption Kinetics .................................................................................. 19 Initial Rate Constants ................................................................................................... 20 Results ............................................................................................................................. 21 Soil Chemical Properties ............................................................................................. 21 Adsorption Kinetics ..................................................................................................... 22 Desorption Kinetics ..................................................................................................... 25 Discussion ....................................................................................................................... 28 List of References ............................................................................................................ 32 Chapter 3 AGING EFFECTS ON COPPER ABSORPTION AND DESORPTION IN PHOSPHATE- OR SLUDGE-TREATED OXISOLS. ...................................................... 37 Abstract ........................................................................................................................... 37 Introduction ..................................................................................................................... 38 Materials And Methods ................................................................................................... 40 Soil Material ................................................................................................................ 40 Soil Pretreatment .......................................................................................................... 42 Cu Sorption .................................................................................................................. 42 Results ............................................................................................................................. 43 Adsorption Distribution Coefficient ............................................................................ 43 Fraction of Copper Desorbcd ....................................................................................... 44 vi Desorption Distribution Coefficient ............................................................................ 48 Discussion ....................................................................................................................... 55 List of References ............................................................................................................ 58 Chapter 4 COPPER ADSORPTION/DESORPTION IN PHOSPHATE- OR SLUDGE-TREATED OXISOLS AS AFFECTED BY pH ................................................................................... 62 Abstract ........................................................................................................................... 62 Introduction ..................................................................................................................... 63 Materials And Methods ................................................................................................... 66 Soil Material ................................................................................................................ 66 Cu Sorption .................................................................................................................. 66 Results ............................................................................................................................. 68 Copper Adsorption Capacity ....................................................................................... 68 Copper Adsorption ....................................................................................................... 70 Copper Desorption ....................................................................................................... 74 Distribution Coefficients .............................................................................................. 76 Discussion ....................................................................................................................... 78 List of References ............................................................................................................ 84 Appendix Al - Sludge characterization .............................................................................. 89 Appendix A2 - Chemical properties of control (Ctrl) and P- or sludge-treated A- and B- horizon samples of two Brazilian Oxisols.. ............................................................................................................................................ 90 Appendix A3 - Titration curves for the determination of the Pszse of control, P-treated, and sludge-treated A- and B-horizon samples of two Brazilian Oxisols. ............................................................................................................................................ 91 Appendix A4 - Specific surface area, total porosity, and pore distribution (micro and mesoporosity) of control (Ctrl) and P- or sludge-treated A- and B-horizon samples of two Brazilian Oxisols ............................................................................................................................................ 92 vii LIST OF TABLES Table 1.1. Specific surface area (SSA) and crystallographic estimates of maximum values of qH/S and qOH/S for selected soil minerals presented in Oxisols ........................................ 4 Table 1.2. Adsorption equilibria on Fe and Al oxide (and silicate) surfaces ....................... 5 Table 2.1. Selected properties of A and B horizons of two Oxisols from Brazil. .............. 15 Table 2.2. Chemical properties of control (Ctrl) and P- or sludge-treated soil samples. 17 Table 3.1. Selected properties of A and B horizons of two Oxisols from Brazil. .............. 41 Table 4.1. Copper sorption capacity (Cu¢ max.) at pH 4.5, 5.5, and 6.5 of control and P- or sludge-treated soil samples. ................................................................................................ 69 Table 4.2. Copper speciation at pH 4.5, 5.5, and 6.5 (pCO2 = 3.5) in 5 mM Ca(NO,)2 from different initial total solution-phase Cu concentration, [Cu-r] as simulated by MINTEQA2 (Allison et al., 1990). ......................................................................................................... 8] viii LIST OF FIGURES Figure 1.1. Oxisol distribution in Brazil indicated by dark shading (EMBRAPA, 1981) (Brazil’s area = 8,511,965 km2) ........................................................................................... 3 Figure 1.2. Transport processes in solid-liquid soil reactions - Nonactivated processes: 1. Transport in the soil solution, 2. Transport across a liquid film at the solid-liquid interface, 3. Transport in a liquid filled micropore; Activated processes: 4. Difussion of a sorbate at the surface of the solid, 5. Diffusion of a sorbate occluded in a micropore, 6. Diffusion in the bulk of the solid (from Aharoni and Sparks, 1991). ....................................................... 6 Figure 2.1. Initial rate constants for Cu adsorption (km) in control and P- or sludge-treated A and B horizon samples of two Oxisols (Gme = Dark-Red Latosol; Kth = Yellow- Red Latosol). ...................................................................................................................... 23 Figure 2.2. Time-dependence of Cu adsorption in control and P- or sludge-treated A and B horizon samples of two Oxisols (Gme = Dark-Red Latosol; Kth = Yellow-Red Latosol). Left axis is CWCuW; right axis is adsorbed Cu ........................................ 24 Figure 2.3. Copper desorbed during initial 15-min desorption reaction for control and P- or sludge-treated A and B horizon samples of two Oxisols (Gme = Dark-Red Latosol; Kth = Yellow-Red Latosol). ............................................................................................. 26 Figure 2.4. Time-dependence of fraction of Cu desorbed (Cum/Cum1 ad) in control and P- or sludge-treated A and B horizon sample of two Oxisols (Gme = Dark-Red Latosol; Kth = Yellow-Red Latosol). ............................................................................................. 27 Figure 3.1. Distribution coefficients (Cu adsorbed/Cu supernatant) for Cu adsorption in control and P- and sludge-treated A and B horizon samples of two Oxisols reacted with 150 and 50 M Cu for 1 h, 3 d, 18 d, and 54 (1 (Gme = Dark-Red Latosol; Kth = Yellow-Red Latosol). ......................................................................................................... 45 Figure 3.2. Time-dependence of fraction of Cu desorbed fiom control and P- or sludge- treated A and B horizon samples of two Oxisols reacted with 150 M Cu for 1h, 3 d, 18 d, and 54 (1 (Gme = Dark-Red Latosol; Kth = Yellow-Red Latosol). ............................... 46 Figure 3.3. Time-dependence of fraction of Cu desorbed from control and P- or sludge- treated A and B horizon samples of two Oxisols reacted with 50 M Cu for 1 h, 3 d, 18 d, and 54 (1 (Gme = Dark-Red Latosol; Kth = Yellow-Red Latosol). ............................... 47 ix Figure 3.4. Distribution coefficients of Cu (Cu remaining adsorbed/Cu supernatant) after 1 h of Cu desorption fi'om control and P- or sludge-treated A and B horizon samples of two Oxisols reacted with 150 and 50 W Cu for 1 h, 3 d, 18 d, and 54 (1 (Gme = Dark-Red Latosol; Kth = Yellow-Red Latosol). ............................................................................... 49 Figure 3.5. Ratio of Km , h/Km in control and P- or sludge-treated A and B horizon samples of two Oxisols reacted with 150 and 50 M Cu for l h, 3 d, 18 d, and 54 d ........ 51 Figure 3.6. Sorbed Cu concentration as a function of solution-phase Cu concentration in control Oxisols samples reacted with 150 and 50 M Cu for l h, 3 d, 18 d, and 54 d ....... 52 Figure 3.7. Sorbed Cu concentration as a function of solution-phase Cu concentration in P-treated Oxisols reacted with 150 and 50 M Cu for l h, 3 d, 18 d, and 54 d. ................ 53 Figure 3.8. Sorbed Cu concentration as a function of solution-phase Cu concentration in sludge-treated Oxisols reacted with 150 and 50 M Cu for l h, 3 d, 18 d, and 54 d. ........ 54 Figure 4.1. Copper adsorption isotherms in control and P- or sludge-treated A and B horizon samples of two Oxisols reacted with Cu at pH 4.5, 5.5, and 6.5 (Gme = Dark- Red Latosol; Kth = Yellow-Red Latosol). ....................................................................... 71 Figure 4.2. Fraction of Cu adsorbed as a function of pH in control and P- or sludge-treated A and B horizon samples of two Oxisols reacted with 5, 50, and 150 1.1M Cu (1:100 soil:solution) (Gme = Dark-Red Latosol; Kth = Yellow-Red Latosol) ......................... 72 Figure 4.3. Fraction of Cu desorbed in 5 mM Ca(NO,)2 (pH 5.5) from control and P- or sludge-treated A and B horizon samples of two Oxisols reacted with 5, 50, and 150 M Cu (1:100 soil:solution) at pH 4.5, 5.5, and 6.5 (Gme = Dark-Red Latosol; Kth = Yellow-Red Latosol). ................... , ...................................................................................... 75 Figure 4.4. Adsorption and desorption distribution coefficients of Cu in control and P- or sludge-treated A and B horizon samples of two Oxisols reacted with 150 [TM Cu (1:100 soil:solution) at pH 4.5, 5.5, and 6.5 (Gme = Dark-Red Latosol; Kth = Yellow-Red Latosol) ........ . ....................................................................................................................... 7 7 Chapter 1 GENERAL INTRODUCTION Rationale Phosphate fertilization is nearly ubiquitous in Oxisols used for agriculture. High concentrations of Cu in some P fertilizers have been of concern in Brazil. Application of sewage sludge to agricultural land is also becoming an increasing practice in Brazil. Thus, there is a need for better understanding of the effects of such agricultural inputs on Cu sorption and transport in Oxisols. This research is proposed to determine the effects of phosphate and sludge treatment on Cu adsorption/desorption in A and B horizons of two Brazilian Oxisols with similar Fe oxides content but different kaolinitezgibbsite and goethite:hematite ratios. The kinetics of Cu adsorption/desorption, as well as the effects of aging and pH, on Cu adsorption and desorption were measured at initial Cu concentrations from 5 to 150 M. Origin of Cu in Oxisols Average Cu concentrations in soils range from 20 to 30 mg kg‘1 (Baker, 1990), though Oxisols can contain much greater Cu concentrations (Baker, 1990; Kabata-Pendias and Pendias, 1992), mainly if the parent material originally has a high Cu concentration (Curi and Franzrneier, 1987). A total Cu content as high as 210 mg kg'I was reported for mafic-derived Oxisols from Brazil (Curi, 1983). Total Cu concentration in Brazilian soils has shown a close association with total Fe and clay contents (Pérez et al., 1995). Heavy metal additions to soils normally are regulated based on cation exchange capacity (CBC) and soil pH, but Mattiazzo and Gloria (1995) have suggested that clay content and concentrations of Fe and Al oxides seem to be better criteria for Brazilian soils. Anthropogenic inputs of Cu to soils are very diverse and include fungicides, fertilizers, lime, animal manures, sewage sludges, and atmospheric deposition (Baker, 1990). On a local scale, inputs fi'om anthropogenic sources can greatly exceed natural Cu contents (Tiller and Merry, 1981). Copper concentrations (dry weight basis) in some agricultural sources of Cu are: fungicides = 12000 to 50000 mg kg"; sewage sludges = 50 to 17000 mg kg"; P fertilizers = l to 300 mg kg"; farmyard manure = 2 to 172 mg kg"; lime = 2 to 125 mg kg"; N fertilizers = <1 to 15 mg kg"(Alloway, 1990; Baker, 1990; Kabata-Pendias and Pendias, 1992). Annual additions of Cu from P fertilizers in Brazil are estimated at 213 t per year (Malavolta, 1994), which adds on average 2.25 pg Cu kg soil" year" to Brazilian agricultural soils. Annual additions from lime and phosphogypsum represent 302 t Cu per year, and contributions from sewage sludge can be even greater. Copper concentrations of 1455 mg kg" have been reported for sewage sludges in Brazil (Berton et al., 1989). Applications of only 10 t ha" y" of this sludge could double, in three years, the concentration of Cu in soils having an initial concentration of 20 mg Cu kg" soil. Reactions of Cu in Oxisols Reactions controlling availability of Cu in soil solutions include adsorption/desorption, precipitation/dissolution, and soluble complex formation. At Cu concentrations typically found in soils, Cu is removed from solution by adsorption rather than precipitation (Ellis and Knezek, 1972; James and Barrow, 1981). This is especially true for most Oxisols, since the low solution concentration as well as the low pH do not allow precipitation to occur easily. Oxisols, which are highly weathered, acid soils with low to medium organic matter content and variable clay content, represent about 60 % of the Brazilian territory (Figure 1.1). The mineralogy of the clay fraction is characterized by the predominance of Fe and A1 oxides and kaolinite (Oliveira et al., 1992), all of which have low CBC. Goethite (a-FeOOII) and hematite (a-Fe203) are the most common Fe oxides, whereas gibbsite (y-Al(OH)3) is the main Al oxide present in these soils. These oxides have points of zero charge (PZC) in the range 7.5 - 9.0, which means that they will be positively charged at the natural soil pH of most Oxisols. Organic matter decreases the PZC to near 4.0, so Oxisol A horizons are negatively charged. 70" 62' 54° 46° 38° Figure 1.1. Oxisol distribution in Brazil indicated by dark shading (EMBRAPA, 1981) (Brazil’s area = 8,511,965 km2) The relative importance of a particular mineral for cation and anion sorption depends on the surface area and on the surface charge density (sites per unit area), both of which depend on crystallinity as well as on crystal structure. Specific surface area and crystallographic estimates of maximum moles of complexed proton charge per unit area (qH/S) and of complexed hydroxyl charge per unit area (qOH/S) on proton-selective surface functional groups of minerals in the clay fraction of Oxisols are presented in Table 1.1. Table 1.1. Specific surface area (SSA) and crystallographic estimates of maximum values of qH/S and qOH/S for selected soil minerals presented in Oxisols Mineral SSA“) qH/Sm qOH/Sa) ng. _um01.m‘2_ Goethite 30-120 (avg.= 104) 4.4 6.7 Gibbsite 20 2.8 5.6 Kaolinite 10-150 (avg.= 52) 0.35 1.0 Hematite 10-150 (avg.= 64) ‘0 Source: Resende et al. (1988) ‘2’ Source: Sposito (1984) Copper concentrations in natural soil solutions are controlled by adsorption on oxides and organic materials to a far greater extent than by adsorption on clay minerals, the influence of which may be negligible in some soils (McLaren et al., 1981). In addition, Cu desorption from organic matter and Fe and Al oxides is very small (McLaren et al., 1983), because activation energies are large for desorption of inner-sphere complexes (McBride, 1989). Adsorption equilibria involving Fe and Al oxide (and silicate) surfaces can be illustrated by reactions of Table 1.2. One important effect of Cu2+ adsorption (and also proton transfer, and ligand adsorption), as shown by Table 1.2, is the change in surface charge upon adsorption. Specifically adsorbed cations like Cu” increase the pH of the point of zero charge (pHPZC), also known as the isoeletric point, but lower the pH of the point of zero net proton charge (pHPmC) (Stumm, 1992). The pHPZC increases upon Cu2+ adsorption because the fixed surface charge becomes zero only at higher pH values (higher OH' solution concentration). Conversely, the pHpZNPC is shifted toward lower pH because protons are released as a consequence of Cu2+ adsorption. Table 1.2. Adsorption equilibria on Fe and Al oxide (and silicate) surfaces Acid base equilibria S-OH + H“ c s-OH,+(1) s-on + OH‘ <:> so + up (2) Metal (Cufi) binding S-OH + Cu2+ o S-OCu+ + H+ (3) 2 son + Cu“ <2> (S-O),Cu + 2 H“ (4) S-OH + Cu2+ + H20 <=> S-OCuOH + 2 H+ (5) Ligand exchange (L = ligand, e. g. H ,P0; or HPOX') S-OH + L‘ a S-L + OH‘ (6) 2 S-OH + L’ a Sz-L+ + 2 OH' (7) S-OH + L2“ 4: S-L' + OH' (8) 2 S-OH + L2' <:> Sz-L + 2 OH’ (9) Ternary surface complex formation (metal = Cu“) S-OH + L'+ Cu“ <=> S-L-Cu2+ + OH' (I 1) S-OH + Lz' + Cu” c:> S-L-le + OH' (11) 2 S-OH + Lz' + Cu2+ <=> 2 S-L-Cu + 2 OH' (12) S = Fe or Al (or Si). Modified from Stumm (1992) Specifically adsorbed anions such as H,PO,' or HPO,2 ’, cause the opposite effects on pHPZC and pHm,C of variable-charge mineral surfaces. Some agricultural inputs such as lime, fertilizers, manures, and sewage sludge, which are potential sources of Cu to soils, can modify Cu sorption behavior in soils by changing the net surface charge or generating mobile organic and inorganic colloids, which can enhance Cu transport throughout the soil profile (McCarthy and Zachara, 1989). Kinetics of Soil Reactions Most soil reactions are described as heterogeneous solid-liquid reactions that take place by a multistep mechanism comprising transport processes as well as chemical reactions (Aharoni and Sparks, 1991) (Figure 1.2). Rates of ion exchange processes are especially likely to be transport controlled in the case of readily exchangeable ions, whereas specifically adsorbed ions (6. g. Cu2+ , H2P04', HPOK') may participate in reactions that are surface controlled (Sposito, 1994), although effects of transport and chemical processes are often experimentally inseparable (Sparks, 1989). 1 2 —> -|-_>1 ————> b» —> ——> 1 1 1 LIquld Fllm Figure 1.2. Transport processes in solid-liquid soil reactions - Nonactivated processes: 1. Transport in the soil solution, 2. Transport across a liquid film at the solid-liquid interface, 3. Transport in a liquid filled micropore; Activated processes: 4. Difussion of a sorbate at the surface of the solid, 5. Diffusion of a sorbate occluded in a micropore, 6. Diffusion in the bulk of the solid (from Aharoni and Sparks, 1991). The difficulty of separating the effects of transport phenomena and chemical reaction, and the heterogeneous character of soil ion exchangers restrict the use of simple kinetic models such as first- or second-order rate equations in studies concerning soil reactions kinetics (Sparks, 1989). However, attempts have been made to treat adsorption as a simple reaction in which the surface and the sorbing solute are reactants and the sorbed solute a product (Aharoni and Sparks, 1991 ). The objectives of this research were to: a) study the time-dependence of Cu adsorption/desorption in A and B horizon samples of two Oxisols as affected by phosphate (P) or sludge pretreatment (Chapter 2); b) detemiine the effect of aging upon Cu desorption from untreated and P- or sludge-treated A and B-horizon samples of two Oxisols (Chapter 3); c) determine the effect of pH upon Cu adsorption/desorption in untreated and P- or sludge-treated A and B-horizon samples of two Oxisols (Chapter 4). LIST OF REFERENCES Aharoni, C., and D.L. Sparks. 1991. Kinetics of soil chemical reactions - a theoretical treatment. p. 1-18. In D.L. Sparks and D.L. Suarez (ed.) Rates of soil chemical processes. SSSA Spec. Publ. 27. SSSA, Madison. Alloway, B.J. 1990. The origin of heavy metals in soils. p. 29-39. In B.J. Alloway (ed.) Heavy metals in soils. New York, John Wiley and Sons. Baker, DE. 1990. Copper. p. 151-176. In B.J. Alloway (ed.) Heavy metals in soils. John Wiley and Sons, New York. Berton, R.S., O.A. Camargo, and J .M.A.S. Valadares. 1989. Absorcao de nutrientes pelo milho em resposta a adicao de lodo de esgoto a cinco solos paulistas. R. bras. Ci. Solo 13:187-192. Curi, N. 1983. Lithosequense and toposequence of Oxisols from Goias and Minas Gerais States, Brazil. Ph.D. thesis. Purdue Univ., W. Lafayette, IN (Diss. Abst. Int. 44: 1674-8). Curi, N., and DP. Franzmeier. 1987. Effect of parent rocks on chemical and mineralogical properties of some Oxisols in Brazil. Soil Sci. Soc. Am. J. 51:153-158. Ellis, BC. and B.D. Knezek. 1972. Adsorption reactions of micronutrients in soils. In J .J . Mortvedt et al. (ed.) Micronutrients in Agriculture. p. 59-78. SSSA, Madison. EMBRAPA. Servico Nacional de Levantamento e Conservacao de Solos. 1981. Mapa de solos do Brasil. Rio de J aneiro, Ministério da Agricultura. James, R.O., and NJ. Barrow. 1981. Copper reactions with inorganic components of soils including uptake by oxide and silicate minerals. p. 47-68. In J.F. Loneragan, A.D. Robson, and RD. Graham. (ed.) Copper in soils and plants. Academic Press, Sydney. Kabata-Pendias, A. and H. Pendias. 1992. Trace elements in soils and plants. 2 ed. CRC Press, Boca Raton. Malavolta, E. 1994. Metais pesados nos fertilizantes - mitos, mistificacbes e fatos. CENA/U SP, Piracicaba. 153p. McBride, MB. 1989. Reactions controlling heavy metal solubility in soils. Adv. Soil Sci. 10:1-56. McCarthy, J.F., and J .M. Zachara. 1989. Subsurface transport of contaminants. Environ. Sci. Technol. 23:496-502. McLaren, R.G., J .G. Williams, and RS. Swift. 1983. Some observations on the desorption and distribution behavior of copper with soil components. J. Soil Sci. 34:325-331. McLaren, R.G., R.S. Swift, and J .G. Williams. 1981. The adsorption of copper by soil materials at low equilibrium solution concentrations. J. Soil Sci. 32:247-256. Mattiazzo, M.E., and NA. Gloria. 1995. Parametros para adicao de residues contendo metais. I: Estudos com solucoes. p. 2315-2317. In: Resumos expandidos. Congresso Brasileiro de Ciéncia do Solo 25th, Vicosa. 23-29 July 1995. SBCS/UFV, Vicosa, Brazil. Oliveira, J .B., P.K.T. Jacomine, and MN. Camargo. 1992. Classes gerais de solos do Brasil: guia auxiliar para seu reconhecimento. FUNEP, Jaboticabal. 210p. Pérez, D.V., M.F.C. Saldanha, and NA. Meneguelli. 1995. Avaliacao dos teores totais de alguns elementos micronutrientes e metais pesados em alguns solos. p. 214-216. In: Resumos expandidos. Congresso Brasileiro de Ciéncia do Solo 25th, Vicosa. 23-29 July 1995. SBCS/UFV, Vicosa, Brazil. Resende, M., N. Curi, and DP. Santana. 1988. Pedologia e fertilidade do solo: interacoes e aplicacOes. MEC/ESAL/POTAF OS, Brasilia. 81p. Sparks, D.L. 1989. Kinetics of soil chemical process. Academic Press, New York. 210p. Sposito, G. 1984. The surface chemistry of soils. Oxford University Press, New York. 234p Sposito, G. 1994. Chemical equilibria and kinetics in soils. Oxford University Press, New York. 268p. Stumm, W. 1992. Chemistry of the solid-water interface. John Wiley and Sons, New York. 428 p. Tiller, K.G., and RH. Merry. 1981. Copper pollution of agricultural soils. p. 119-137. In J .F . Loneragan, A.D. Robson, and RD. Graham. (ed.) Copper in soils and plants. Academic Press, Sydney. Chapter 2 COPPER ABSORPTION/DESORPTION KINETICS IN PHOSPHATE- AND SLUDGE-TREATED OXISOLS Abstract Phosphate fertilization is nearly ubiquitous in Oxisols used for agriculture, and the application of sewage sludge to agricultural land is increasing in Brazil and elsewhere. This study measured the effects of phosphate and sludge pretreatment on Cu sorption kinetics and Cu availability in A and B horizon samples of two Oxisols at pH 5.5. Initial rate constants for Cu adsorption were up to 45% greater in P-trcated than control soils, and up to 70% greater in sludge-treated than control soils. The large increase in the initial sorption rate for P- and sludge-treated soils may be caused by an increase in the number of sites with low activation energy, or by a decrease in the activation energy for sorption at either surface-organo or surface-phosphate groups compared with surface-OH groups. Phosphate treatment caused a marked decrease in the initial fractional Cu desorption. In P- treated soils, the fraction of Cu desorbed was about half that in control soils, and smaller in B than A horizons. Sludge treatment affected fractional desorption in a similar way, but the effects were not as remarkable as those for P treatment. Little Cu was desorbed from P-treated soils alter the initial 15-min reaction time, whereas Cu desorption increased with increasing reaction times up to 1 d for sludge-treated soils and up to 18 d for control soils. The large increase in initial Cu adsorption and the decrease in fractional desorption for P- or sludge-treated soils is noteworthy because of its implications for the mobility and availibility of Cu in Oxisols. 10 11 Introduction Adsorption/desorption reactions are among the most important controls on the mobility of Cu in the environment (James and Barrow, 1981). Copper concentrations in natural soil solutions and sediment-water systems typically are controlled by adsorption on surface hydroxyl groups of metal oxides and organic matter (McLaren et al., 1981; Piccolo and Stevenson, 1982; Wang and Chen, 1997). Soil organic matter can provide high concentrations of sites for metal sorption (Logan, 1990; Stumm, 1992; Harter and Naidu, 1995; Spark et al., 1997a; Temminghoff et al., 1997). The increase in heavy metal adsorption in sandy soils amended with sludge has been attributed to an increase in the number of sites for metal adsorption from the solid organic matter added (Petruzzelli et al., 1994). Among divalent first-row transition metals, Cu has the greatest affinity for organic matter (Stevenson and Arkadani, 1972). The high degree of selectivity shown by organic matter for Cu is caused by the formation of inner-sphere complexes, often referred to as chemisorption or specific adsorption. Iron and aluminum oxides are also recognized to have a high affinity for Cu (J enne, 1968; Forbes et al., 1976; Schwertrnarm and Taylor, 1977; McBride, 1982; Spark et al., 1995a). Electron spin resonance studies have shown that Al-OH groups sorb Cu mainly in inner-sphere complexes (McBride, 1982; Harsh and Doner, 1984). For kaolinite, ion exchange (outer-sphere complexes) may be important at low pH and low ionic strength, whereas an increase in both ionic strength and pH favor chemisorption at amphoteric surface hydroxyls (Schindler et al., 1987; Spark et al., 1995b). Copper desorption from organic matter and Fe and Al oxides is limited (McLaren 12 et al., 1983) because activation energies are large for desorption of inner-sphere complexes (McBride, 1989). Oxyanions such as phosphate, sulfate, and organic acids can cause either an increase or a decrease in metal sorption by soils, metal oxides, and layersilicates (McBride, 1994; Guilhenne et al., 1995; Murphy and Zachara, 1995; Ali and Dzombak, 1996a; 1996b; Spark et al., 1997c). When an anion and a metal are added to a soil simultaneously, the effect of the anion on metal sorption depends on a) the net surface charge of the soil; b) the affinity of the soil for the metal, anion, and metal-anion complexes; and c) the tendency of the metal and anion to form soluble complexes. The latter depends in part upon the anion:meta1 charge ratio. A large excess of the anion generally suppresses metal adsorption, while charge parity (or less) with the metal generally favors adsorption by ternary complex formation (McBride, 1994). Increased Cu sorption by goethite in the presence of organic acids at low pH has been attributed to a) a decrease in positive surface charge in the low pH region due to sorption of an organic acid anion, resulting in a more favorable electrostatic environment for Cu sorption or b) the sorption of Cu-organic acid complexes, i.e., the formation of “Cu-organic acid-mineral” ternary surface complexes (Ali and Dzombak, 1996a). If an anion is added to a soil or mineral before a metal is added (for example, phosphate fertilization followed later by Cu fimgicide application), the anion will affect metal sorption principally by converting M’+-OH groups into M3+-anion surface functional groups, possibly altering the net surface charge, and promoting the formation of M”- anion-Cu surface ternary complexes (M = A1 or Fe). Beneficial effects of phosphate for reducing injury from Cu fungicide applications have been reported (Chaney and l3 Giordano, 1977) and may be due to Cu sorption by M3*-phosphate surface fimctional groups. Copper sorption by humate-treated minerals has also been attributed to M”- humate-Cu ternary surface complexes (Spark et al., 1997c). Pretreatment with oxyanions may also affect trace metal sorption kinetics. The time-dependence of trace-metal sorption on goethite indicates that the adsorption reaction initially involves adsorption on external surface sites, with subsequent diffusion to internal sorption sites (Briiemmer et al., 1988). Benjamin and Leckie (1981) reported a rapid initial (1 h) adsorption of Cd followed by a much slower second step, possibly related to solid-state diffusion in amorphous Fe oxyhydroxide. Diffusional processes can contribute to desorption hysteresis and reaction irreversibility (Padmanabham, 1983a, 1983b; Barrow, 1985). Most soil reactions can be described as heterogeneous solid-liquid reactions that take place by a multistep mechanism comprising transport processes as well as chemical reactions (Aharoni and Sparks, 1991). In a practical sense, the effects of transport and chemical processes are often experimentally inseparable (Sparks, 1989). The difficulty of separating the effects of transport phenomena and chemical reaction kinetics, along with the heterogeneous character of surface flmctional groups in soils, often restrict the use of simple kinetic models in soils (Sparks, 1989). However, attempts have been made to treat adsorption as a simple reaction in which the surface and the sorbing solute are reactants and the sorbed solute a product (Aringhieri et al., 1985; Aharoni and Sparks, 1991). In one study, Cu desorption data for a Cu-contaminated soil were fit very well by a first-order kinetics equation in which sorbed Cu was considered the reactant (J opony and Young, 1987). Aringhieri et a1. (1985) found that Cu adsorption kinetics for an organic soil could 14 be described by a model wherein the reaction is first-order with respect to both Cu and substrate concentration (second-order overall) and exhibits a dependence on internal diffusion. Long-term sorption reaction rates are probably mass-transfer limited for metal oxides (Van der Zee and Van Riemsdijk, 1991; Brilemmer et al., 1988) and for highly aggregated Oxisols (Nkedi-Kizza et al., 1982), although batch-shake methods may eliminate much of the mass-transfer control normally found in aggregated soils (Lima, 1995) The objective of this work was to determine the effect of phosphate or sludge pretreatment on Cu adsorption/desorption kinetics and Cu availability in Oxisols that differ in mineralogy and organic carbon (OC) concentration. Batch-shake methods were used so that mass transfer limitations due to differences in aggregation would be minimized or eliminated and the resulting initial reaction rates would reflect differences in the surface chemistry of control, P-treated, and sludge-treated soils. Materials and Methods Soil Material Samples of A and B horizons floor two uncultivated Oxisols were collected near S. Joao Del Rei in the Campos das Vertentes region of Minas Gerais, Brazil (latitude 21° 20’ S; longitude 44° 30’ W). Both soils were vegetated with semi-deciduous tropical cerrado (tortuous trees and shrubs scattered above grass and herbaceous plants) and were underlain by mica schists of the Andrelandia group. The first soil, a Dark-Red Latosol (very fine, allitic, isotherrnic Typic Hapludox), formed on steeply inclined strata and was 15 very well drained. The second soil, a Yellow-Red Latosol (very fine, allitic, isotherrnic Typic Hapludox), developed in nearly horizontal strata and was relatively poorly drained. Both soils have pH near 4.5 in the A horizon and 5.5 in the B horizon. Although both soils have similar clay contents (about 700 g kg") and total Fe oxide contents (165 g kg"), differences in drainage have caused the Yellow-Red Latosol to have greater kaolinitezgibbsite and goethite:hematite ratios than the Dark-Red Latosol (Table 2.1). Table 2.1. Selected properties of A and B horizons of two Oxisols fiom Brazil. Horizon Clay OC’f Fedf Fe,t Kt5 Gb§ Gtsz1 SSA” ———g kg" soil g kg" clay cmz g" W A 691 25.3:t0.0 99 1.88 350 510 5.0 59i3 B 753 9.9:I:0.1 114 0.65 350 510 4.1 61i3 KellmL-RedLaLcmLIKLth A 711 23.6:t0.3 101 1.20 480 375 10.2 53:1:4 B 721 8.0:lz0.1 1 14 0.59 480 400 8.3 58i3 1 OC is organic carbon measured by the Walkley-Black method. * Fd and Fee, respectively, are Fe extracted by dithionite-citrate-bicarbonate (Mehra and Jackson, 1960) and ammonium oxalate (pH 3.1) in the dark (Schwertrnann, 1964). § Kt and Gb, respectively, are kaolinite and gibbsite measured by differential thermal analysis in dithionite-treated clay samples (but expressed on a total-clay basis). ‘ Gt:Hm is the goethite:hematite ratio in the clay fraction, measured by X-ray diffraction of NaOH-treated samples (Kampf and Schwertrnann, 1982a; 1982b). " Specific surface area determined by N2 adsorption isotherms (BET). The A and B horizons of each soil differ in OC content as well as in the relative proportions of amorphous and crystalline Fe (estimated as the ratio of oxalate-extractable to dithionite-extractable Fe). Soil samples were gently crushed to break large aggregates l6 and then were sieved to obtain the <2-mm fi’action, which was used in all of the experiments described below. Relevant soil properties are summarized in Table 2.1; additional details concerning soil characterization may be found in Lima and Anderson (1997) Phosphate and Sludge Pretreatment To assess the effect of phosphate pretreatment (hereafter termed P treatment) on Cu sorption kinetics, samples of each soil material were reacted with 10.75 mM Ca(H2PO.,)2'H20 (21.5 mmol P L") at a soil:solution ratio of 2:3 to give a P addition rate of 1 g P kg" soil. The samples were shaken for 48 h on a reciprocating shaker (120 cycles min") and centrifuged for 10 min at 9000 rpm. Excess solution was removed, and the P concentration was measured colorimetrically. The samples were centrifuge-washed once with 5 mM Ca(N03)2 (2:1 solutionzsoil) to remove entrained and readily desorbed P. The P concentration of the supernatant wash solution was measured. Adsorbed P was near 30 mmol kg" for all samples. Sludge pretreatment consisted of reacting each soil material with air-dried, <0.5-mm lime-stabilized sludge at a rate of 5 g kg". Samples treated with sludge were incubated for 30 (1 (until constant pH) at 37 °C at a moisture content of 60% of saturation. The dry sludge had an OC content of 276 i 8 mg kg"; other sludge properties are reported in Appendix A1. The sludge-treated soils contained about 2 g kg" more OC (Appendix A2) than the control soils. Phosphate- and sludge-treated soils were air dried and crushed to pass a 2-mm sieve. The pHPZSE, and DTPA-extractable Cu measured in subsamples of control and P- or sludge-treated soils are reported in Table 2.2. 17 fine 5:85, e5 are 5.5 .3 293285 .6” A2 xmceoqnzv 3E: Nd H mm 358.805. ”£33 :80.“ a. and .«o @052: 05 moi: 3.5308 “cube :8 Ben mo «Eon 05 a 03:3 mas new. 3% 91.5. 3.1.3. Show ~33 new new as m Ea fl. No man no“; 33% 33: o? men one < 3 «at. 8.8 one Lime 33S Same Se a: 2e m $8 on: 3% 33% a. 39% 53.3 m? 8.4 a. < 333% wwx BEE emx 381 seem a Bo omega .2 Eu Seam .2 Bo com £896 coaseom so “so assuage/ES maize .8383 :8 powwobowpgm co -m use 9va 3980 me 8538a Hogono INN 033. 18 pH Adjustment Prior to measuring Cu sorption capacity or Cu sorption kinetics, suspensions of the twelve soil samples (control and P- or sludge-treated samples from A and B horizons of two Oxisols) were adjusted to pH 5.5 using the following procedure. For each set of experimental conditions (described below), triplicate 2.5-g subsamples of each soil material were suspended in 167 mL of 5 mM Ca(N03)2 (pH 5.5) and stirred continuously with a TPFE-coated stirring bar. The pH was measured and readjusted to 5.5 with either saturated Ca(OI-I)2 or 7 mM I-INO,. All samples were shaken for 24 h on a reciprocating shaker (120 cycles min"), suspension pH was measured again, and acid or base was added as needed to readjust to pH 5.5. Samples were shaken again for 24 h and pH was readjusted as necessary before the samples were shaken again for 24 h. During the third 24-h shaking period, suspension pH changed less than 0.05 pH units, so pH was considered to be stable after 72 h. This 72-h shaking time during pH adjustment effectively disaggregates all samples (Lima, 1995). Consequently, the sorption kinetics experiments described below are unaffected by the original degree of aggregation of the soils and are not confounded by initial differences in aggregation between P- or sludge- treated and control samples. The pH-adjusted soil suspensions were used in the Cu sorption capacity and kinetics experiments described below and were prepared fresh immediately before each experiment. Copper Sorption Capacity Copper sorption capacity at pH 5.5 was measured by repeated reaction with 500 W Cu. This “repeated reaction” approach is preferable to a single reaction at a high Cu 19 concentration because Cu concentrations >500 uM could cause Cu precipitation at pH 5.5 (Allison et al., 1990). After suspension pH was adjusted as described above, appropriate amounts of Cu(NO3)2 in Ca(N03)2 (pH 5.5; I=15 mM) were added to each soil suspension to give initial total Cu concentrations of 500 M and a solutionzsoil ratio of 100: 1. Triplicate soil suspensions were shaken with 500 M Cu(NO3)2 in 4.5 mM Ca(NO3)2 (pH 5.5; 1:15 mM) for 72 h, then centrifuged. The supernatant solutions were decanted and saved for Cu analysis, and the mass of entrained solution was recorded. Each soil paste was reacted again with fresh 500 M Cu in Ca(NO3)2, and centrifuged as described above. This Cu reaction-centrifugation-decantation-Cu analysis procedure was repeated until negligible additional Cu was sorbed. Ten such sorption cycles were required, and the supernatant pH was readjusted to 5.5 after every two cycles. The Cu sorption capacities of control and P- or sludge-treated soils are reported in Table 2.2. Adsorption/Desorption Kinetics Alter pH adjustment, appropriate amounts of Cu(NO,)2 in Ca(NO;,)2 (pH 5.5; 1:15 mM) were added to each soil suspension to give initial total Cu concentrations of 50 and 150 M and a solution:soil ratio of 100:1. Samples were shaken at 120 cycles min" on a reciprocating shaker for 10 min and then were centrifuged for 10 min at 9000 rpm. A 1.5- mL aliquot of supernatant solution was withdrawn from each bottle and saved for Cu analysis by flame atomic absorption spectroscopy. The soils were resuspended and shaken as described above. The centrifuging and subsampling procedure was repeated later for 20 total shaking times of 1 h, 6 h, 24 h, 3 d, 6 d, and 18 d. After 18 d, the supernatant solution was decanted carefully to minimize the amount of entrained solution. The soil pastes that remained in the bottles after centrifuging and decanting the adsorption solution were weighed to determine the mass of entrained solution. Copper desorption from these soil pastes was measured by adding 250 mL of 5 mM Ca(NO,)2 to each soil paste. Samples were shaken, centrifuged, and subsarnpled at the same time intervals used for adsorption kinetics. Total Cu desorbed by 5 mM Ca(NO,)2 at each reaction time was corrected for the amount of Cu entrained at the start of the desorption reaction. Initial Rate Constants Rate constants for Cu adsorption (km) were calculated by considering adsorption to be first-order with respect to both total solution-phase Cu, [CuJ and the concentration of adsorption sites, [S]. Using the initial rate method and making the assumption that the desorption reaction makes a negligible contribution to the overall reaction rate during the initial lS-min reaction time, km, may be calculated as -A[CuJ/At = km: [(311.10 [810 [1] where A[Cu,] is the change in [CuJ during the initial 15-min reaction, [Cu,]0 is the initial solution-phase Cu concentration, and [S]0 is the Cu adsorption capacity. Because the speciation of solution-phase and sorbed Cu is unknown, all concentrations are expressed as mM total Cu, so Eq. [1] is an apparent rate expression. We used 15-min as the time interval for the initial reaction time because the reaction stops when soil is separated from solution, and 15 min represents a 10-min shaking time plus half the centrifugation time. 21 These rate constants are precise in that each sample was reacted for exactly the same amount of time prior to centrifugation. However, because of differential settling during centrifugation and the impossibility of determining exactly when the “average” soil particle was no longer in contact with the bulk solution, the initial rate constants may not be appropriate for comparison with other studies that have used other methods and time scales. Nevertheless, these rate constants are suitable for determining the effects of P and sludge treatment and soil composition on kw. Desorption rate constants are not calculated here because the net desorption during the initial 15-min desorption was affected by Cu readsorption, and extensive irreversibility precluded using the relation KW = kws/kdes, as was previously suggested by Aringhieri et al. (1985). Results Soil Chemical Properties Pretreatment with P caused Cu adsorption capacity to increase by as much as 30% and pHPZSE to decrease by 0.25 to 0.9 pH units compared with the control soils (Table 2.2). Phosphate treatment had a greater effect on pszSE in the Kth than Gme soil. Sludge pretreatment, on~ the other hand, had no effect on pHPZSE except in the B horizon of the Kth soil and did not have a statistically significant effect on Cu adsorption capacity (Table 2.2), even though the OC contents of sludge-treated soils were about 2 g kg" greater than for control soils (Appendix A2). Both P and sludge treatment caused increases in DTPA-extractable Cu (Table 2.2), although DTPA-extractable Cu was at least 22 1000 times less than the Cu sorption capacity and thus should have little effect on the Cu adsorption experiments described below. Adsorption Kinetics Rate constants for Cu adsorption (km) measured during the initial lS-min reaction were up to 45% greater for P-treated than control soils and up to 70% greater for sludge- treated than control soils (Figure 2.1). Adsorption rate constants were greater for 50 M than 150 nM Cu for any given sample. The time-dependence of Cu adsorption is plotted both in terms of the fiaction adsorbed (i.e., Cum/Cum“) and the adsorbed Cu concentration (mmol kg") on the left and right axes, respectively, of Figure 2.2. For 50 M Cu (Figures 2.2a-c), Cu adsorption by sludge-treated samples approached steady-state after 6 h, whereas P-treated samples required about 24 h, and control samples did not reach a plateau within the 18-d reaction time. For 150 M Cu, Cu adsorption again was fastest in sludge-treated soils-and slowest in control soils, although Cu adsorption did not reach steady state within the 18-d reaction time for any of the soils (Figures 2.2d-f). Pretreatment with P and sludge had much less effect on Cu adsorption after 18 d (i.e., Cumm) than on either the concentration of Cu adsorbed during the initial 15-min reaction or on the time-dependence of Cu adsorption. For soils reacted with 50 W Cu, sludge treatment caused up to a 70% increase in initial (1 5-min) Cu adsorption but only a 10% increase in Cum,“ Phosphate treatment caused a smaller increase in initial Cu adsorption (up to 60%) but a 30% increase in Cum,” For soils reacted with 150 M Cu, sludge treatment caused a 40 to 75% increase in initial Cu adsorption but only a 25% increase in Cummd; P treatment caused a 30 to 65% increase in initial Cu adsorption and a 10 to 45% increase in Cummd. 23 [Cu] Initial = 150 ”M E Control 1:1 Phosphate -1 -1 ka d8 (L mmol h ) O 6 r///////. Sludge 4 2 0 A B A B Gme Kth Figure 2.1. Initial rate constants for Cu adsorption (km) in control and P- or sludge-treated A and B horizon samples of two Oxisols (Gme = Dark-Red Latosol; Kth = Yellow- Red Latosol). 24 so Bases a as am: ”35355 a as «3 .2823 Bases». n 5Q £823 Basso u Swag £8me 23 .«o 8383 canto: m 98 < “6335-035? .6 -m 98 35:8 5 5:88pm :0 .«o oueopooqopefifi .~.N Bowman Ev 2.5... 3039.033 coo? 00.. o_. _. rd cor or F v.0 cor or v _..o Qm iiiiiiiiiirliiIiiiriftiu Nd 3 :1 o9 ".3330. m. no BTW wed a 3m woo m. can mod w 0 . . m . V Mme. 1. w so m. n 0 wofi - .. ed M O n we? .U ) w 3.3 m . 1 "3.; p m cm - .. 2 om .0. m .m . L m 4, zone: o- m hon- m. 11:88:17 m L 4 g n m.m 1. g W p m 5.135 2282a .228 25 Desorption Kinetics Concentrations of Cu desorbed during the initial 15-min desorption reaction are shown in Figure 2.3. For 50 M Cu, P and sludge treatment caused initial Cu desorption to decrease by about 50% (Figure 2.3a), even though the adsorbed Cu concentration was greater for P-treated and sludge-treated soils than for control soils (Figures 2.2a-c). For 150 M Cu, P treatment caused a 4 to 16% decrease in initial Cu desorption, and sludge treatment a 16 to 30% decrease in initial Cu desorption compared to control soils (Figure 2.3b). For both Cu concentrations, adsorbed Cu was greater and desorbed Cu less in Gme than Kth soils. For 50 W Cu, desorbed Cu was less in A than B horizons, although sludge and P treatment minimized the differences in initial Cu desorption between A and B horizons. These effects of soil type and pretreatment on fi'actional desorption can be seen more clearly in Figure 2.4. For soils initially reacted with 50 M Cu, only 2 to 9% of sorbed Cu was desorbed during the 18-d desorption reaction (Figures 2.4a-c), whereas soils reacted with 150 W Cu released 8 to 18% of adsorbed Cu (Figures 2.4d-f). For both Cu concentrations, the fraction of Cu initially desorbed from P-treated and sludge- treated soils was about half that desorbed from control soils. Fractional Cu desorption generally was smaller in A than B horizons. and in the Gme than Kth soil, although P- treatment greatly diminished the differences among soils, and both P and sludge treatment had greater effects on fractional desorption than did soil properties. Both P and sludge treatment had notable effects on the time dependence of Cu desorption (Figure 2.4). At both Cu concentrations, Cu desorption from control soils increased steadily throughout the 18-d reaction time (Figures 2.4a and d), whereas Cu desorption from P-treated soils was nearly complete after the initial 15-min desorption reaction (Figures 2.4b and e). Net Cu desorption from sludge-treated soils increased 26 during the first 24 h, after which some initially desorbed Cu was readsorbed by the soils (Figures 2.4c and f). 0.3 '3’ 0.2 E” g 0.1 E 5,3 0.0 13° 1.6 6 .D '5 1.2 In G) D 0.8 8 0.4 0.0 [Cu] Initial: 50 “M [CU] [mug]: 150 11M m Control (b) [:1 Phosphate m Sludge A A B Gme Kth Figure 2.3. Copper desorbed during initial 15-min desorption reaction for control and P- or sludge-treated A and B horizon samples of two Oxisols (Gme = Dark-Red Latosol; Kth = Yellow-Red Latosol). 27 .9808..— pom§o=o> u “OHM m38:5 come—3Q n EEDOV £8me 95 Co 2953 £55: m 28 < woueobowpsfi .8 -m can 75:8 5 333301.38 39—836 :0 mo :03an .3 005253-08? .vd DBMS E as: 5:988 coo? Dow or _. _..o cow 9. _. _..o cow or _. rd wood :1 o9 u 3.5.8. ("1 ""°n0/’°"no) pemoseo narrows 4 acute... m o . 141 coutoz < 101 Q Show. E a. m E. emu-gm cannamoca .ohcoo 28 Discussion The decrease in pHPZSE and the increase in Cu adsorption capacity caused by P pretreatment are consistent with prior studies wherein phosphate adsorption in inner- sphere complexes on metal oxides and Oxisols causes a decrease in PHch- The very small or negligible effects of sludge pretreatment on pHPZSE and on Cu adsorption capacity, however, are unexpected, particularly since the sludge-treated soils contained 2 g kg" more OC than did control soils. Other researchers have shown that humic acid adsorption on metal oxides and kaolinite causes pHPZC to decrease (Kretzschmar et al., 1997; Spark et al., 1997b). Further, an increase in Cu adsorption has been reported for goethite in the presence of organic acids (Ali and Dzombak, 1996a) and humic acid (Spark et al., 1997c) and attributed to M3+-organo surface functional groups. Increased Cu adsorption by sandy soils after sludge treatment, on the other hand, has been attributed to adsorption by sites on solid organic matter that is not adsorbed by soil minerals (Petruzzelli et al., 1994). The negligible effect of sludge pretreatment on PHste in the present study may simply reflect the fact that pHste is not a true zero point of charge and cannot be assumed equal to pHPZC in a system as complex as a soil amended with lime-stabilized sludge. Prior to pHPZSE measurements (but not before Cu adsorption experiments) the sludge-treated soils were washed with dilute acid to remove CaCO3 because PHsts for unwashed sludge-treated soils were greater than those for control soils. Thus, the pHpZSE values for the sludge- treated soils cannot be interpreted as true zero points of charge. Both P and sludge treatment had much greater effects on km and on initially adsorbed Cu than on either Cu adsorption capacity or Cum,” Further, sludge had a greater effect than P on kw, whereas P had a greater effect than sludge on Cu adsorption capacity and Pszsra- Thus, P and sludge treatment must affect Cu adsorption kinetics by increasing the proportion of sites with low activation energies for Cu adsorption, not simply by increasing the total number of Cu adsorption sites. It is not likely that Cu 29 adsorption kinetics were affected by solution-phase complex formation between Cu and either dissolved phosphate or organic matter, because the solution-phase P and organic C concentrations during the batch adsorption reaction were too low to promote significant complex formation (unpublished data, 1996). The slow Cu sorption reaction that occurs between 15 min and 18 (1 may be caused by adsorption on sites with greater activation energies or by slow diffusion of Cu to internal adsorption sites in metal oxides. Initial Cu adsorption was greater, and slow reaction less important, in P- and sludge-treated than control soils. This suggests that the concentration of readily accessible, highly reactive sites was greater in the P- and sludge- treated soils, even though the Cu adsorption capacities differed little between treatments. Phosphate and sludge treatment convert M3+-OH groups into M3i-phosphate and M”- organo fiinctional groups, thereby modifying the reactivity of the site toward Cu but not increasing the number of surface M3+ sites, which exert overall control on the Cu adsorption capacity. Fast metal adsorption on Fe oxide has been attributed to adsorption on doubly coordinated (bridging) OH groups, and slower adsorption to terminal OH groups (Grossl and Sparks, 1995). In contrast, oxyanions more readily form surface complexes at terminal OH groups, particularly protonated OH groups (Sposito, 1984; Hiemstra et al., 1989). Thus, P or organic matter adsorption on positively charged (or neutral) terminal OH groups would decrease the activation energy for Cu adsorption on that site by decreasing the electrostatic repulsion (increasing the electrostatic attraction) between Cu and the site while having little effect on the Cu adsorption capacity. This modification of the native M3*-OH groups also explains why P and sludge treatment tended to diminish any differences between A and B horizons of Gme and Kth soils that were pronounced in the control soil. The fact that km, was greater for 50 pM than 150 W Cu indicates either that it was inappropriate to make the assumption that the reverse (desorption) reaction made a negligible contribution to the reaction rate during the initial lS-min reaction, or else that 30 the initial reaction was influenced by transport processes such as diffusion. Aringhieri et al. (1985) found that Cd adsorption was influenced by diffusion at high concentrations and short reaction times. In addition, Cu adsorption kinetics may depend on the concentration of CuOH“ or some other species, not [Cu,], which is used to calculate km in Eq. [1]. If the proportionality between [CuOIF] and [CuJ differs greatly between 50 and 150 M, then formulating the rate expression in terms of [CuJ is inappropriate. However, because the identity of the rate-controlling Cu species is unknown, Eq. [1] is as suitable as any other formulation of the rate expression. The marked decrease in Cu fractional desorption observed in this study is typical for trace metal adsorption/desorption by organic matter and oxides (McLaren et al., 1983) and soils (Sparks, 1985). The smaller fractional desorption caused by P and sludge treatment, especially in B horizons, is also consistent with the explanation that these treatments change surface M3+-OH groups into M3*-phosphate and M3+-organo surface fimctional groups that have greater affinity for Cu than do the normal M3i-OH groups. The smaller fi'actional desorption for samples reacted with 50 M Cu than 150 M Cu is also consistent with the hypothesis that there are a limited number of sites with very great affinity for Cu. The smaller fractional desorption in A than B horizons is consistent with the generally observed decrease in metal desorption as OC content increases (Sparks, 1985). In addition, A horizons had 2.5 to 3 times more oxalate-extractable Fe (F e0) than did B horizons, and other researchers have reported that isotopically exchangeable Cd and Zn decreases with increasing Feo (F uj ii and Corey, 1986). The smaller fractional desorption for the Gme than Kth soil may be attributed to the formation of stronger bonds between Cu and gibbsite compared with Cu-kaolinite bonds or to the greater Feo concentrations in the Gme than Kth soils. In smnmary, phosphate and sludge pretreatrnents have much greater effects on Cu sorption kinetics and fractional desorption than on net Cu adsorption after 18 d or on Cu adsorption capacity. In a transient system such as a soil with alternating leaching and 31 drying events, the initial rates of adsorption and desorption may have greater impacts on Cu mobility and availability than would the Cu adsorption capacity. Thus, the 45 to 70% increase in initial Cu adsorption in P and sludge-treated soils, combined with the roughly 50% decrease in initial Cu desorption, suggest that sludge and P treatment will greatly decrease the mobility of Cu in Oxisols. 32 LIST OF REFERENCES Aharoni, C., and D.L. Sparks. 1991. Kinetics of soil chemical reactions - a theoretical treatment. In D.L. Sparks and D.L. Suarez (ed.) Rates of Soil Chemical Processes. p. 1-18. SSSA Spec. Publ. 27. SSSA, Madison. Ali, M.A., and DA. Dzombak. 1996a. Effects of simple organic acids on sorption of Cu“ and Ca2+ on goethite. Geochim. Cosmochim. Acta 60:291-304. Ali, M.A., and DA. Dzombak. 1996b. Interactions of copper, organic acids, and sulfate in goethite suspensions. Geochim. Cosmochim. Acta 60:5045-5053. Allison, J .D., D.S. Brown, and K.J. Novo-Gradac. 1990. MINTEQA2/PRODEFA2, a geochemical assessment model for environmental systems: Version 3.00 user’s manual. EPA-600/3-91-021. USEPA, Athens. Aringhieri, R., P. Carrai, and G. Petruzzelli. 1985. Kinetics of Cu2+ and Cd2+ adsorption by an Italian soil. Soil Sci. 139:196-204. Baker, DE, and MC. Amacher. 1982. Nickel, copper, zinc, and cadmium. In AL. Page (ed.) Methods of soil analysis, Part 2. 2nd ed. Agronomy 9:323-346. Barrow, NJ. 1985. Reactions of anions and cations with variable-charge soils. Adv. Agronomy 38:183-230. Benjamin, M.M., and J .O. Leckie. 1981. Multiple-site adsorption of Cd, Cu, Zn, and Pb on amorphous iron oxyhydroxide. J. Colloid Interface Sci. 79:209-221. Brfiemmer, G.W., J. Gerth, and K.G. Tiller. 1988. Reaction kinetics of adsorption and desorption of nickel, zinc, cadmium by goethite. 1. Adsorption and diffusion of metals. J. Soil Sci. 39:37-52. Chaney, KL, and RM. Giordano. 1977. Microelements as related to plant deficiencies and toxicities. In L.F. Elliot and F.J. Stevenson (ed.) Soils for Management of Organic Wastes and Waste Waters. p. 234-279. ASA, Madison. Forbes, E.A., A.M. Posner, and J .P. Quirk. 1976. The specific adsorption of divalent Cd, Co, Cu, Pb and Zn on goethite. J. Soil Sci. 27:154-166. 33 Fuj ii, R. and RB. Corey. 1986. Estimation of isotopically exchangeable cadmium and zinc in soils. Soil Sci. Soc. Am. J. 50:306-308. Grossl, RR. and D.L. Sparks. 1995. Evaluation of contaminant ion adsorption/desorption on goethite using pressure-jump relaxation techniques. Geoderrna, 67: 87-101. Guilherrne, L.R.G., J .M. Lima, and SJ. Anderson. 1995. Efeito do fosforo na adsorcao de cobre em horizontes A e B de latossolos do Estado de Minas Gerais. p. 316-318. In Resumos expandidos. Congresso Brasileiro de Ciéncia do Solo 25‘“, Vicosa. 23-29 July 1995. SBCS/UFV, Vicosa, Brazil. Harsh, J .B., and HE. Doner. 1984. Specific adsorption of copper on an hydroxy- aluminum-montrnorillonite complex. Soil Sci. Soc. Am. J. 48:1034-1039. Harter, RD, and R. Naidu. 1995. Role of metal-organic complexation in metal sorption by soils. Adv. Agronomy 55: 219-263. Hiemstra, T., J .C.M. Wit, and W.H. Van Riemsdijk. 1989. Multisite proton adsorption modeling at the solid/solution interface of (hydr)oxides: a new approach. 11. Application to various important (hydr)oxides. J Colloid Interface Sci. 133: 105-1 17. James, R.O., and NJ. Barrow. 1981. Copper reactions with inorganic components of soils including uptake by oxide and silicate minerals. In J.F. Loneragan, A.D. Robson, and RD. Graham (ed.) Copper in Soils and Plants. p. 47-68. Academic Press, Sydney. J enne, EA. 1968. Controls on Mn, Fe, Co, Ni, Cu, and Zn concentrations in soils and water; the significant role of hydrous Mn and Fe oxides. In Trace Inorganics in Water. Advances in Chemistry Series, 73. Am. Chem. Soc., p.:337-387. J opony, M., and D. Young. 1987 . A constant potential titration method for studying the kinetics of Cu” desorption fi'om soil and clay minerals. J. Soil Sci. 38:219-228. Kttmpf, N., and U. Schwertrnann. 1982a. Quantitative determination of goethite and hematite in kaolinitic soils by x-ray diffraction. Clay Miner. 17:359-363. Kampf, N., and U. Schwertrnann. 1982b. The 5-M-NaOH concentration treatment for iron oxides in soils. Clays and Clay Miner. 30:401-408. Kretzschmar, R., D. Hesterberg, and H. Sticher. 1997. Effects of adsorbed humic acid on the surface charge and flocculation of kaolinite. Soil Sci. Soc. Am. J. 61: 101-108. Lima, J .M. 1995. Relation between phosphate sorption and aggregation in Oxisols from Brazil. East Lansing: Michigan State University, 87p. (Ph.D. Dissertation). 34 Lima, J .M., and S]. Anderson. 1997. Effect of aggregation and aggregate size on extractable Fe and Al in two Brazilian Typic Hapludoxs. Soil Sci. Soc. Am. J. 61:965-970. Logan, T.J. 1990. Chemical degradation of soils. Adv. Soil Sci. 11:187-221. McBride, MB. 1982. Cu2+ adsorption characteristics of aluminum hydroxides and oxyhydroxides. Clays Clay Miner. 30:21-28. McBride, MB. 1989. Reactions controlling heavy metal solubility in soils. Adv. Soil Sci. 10: 1-56. McBride, MB. 1994. Environmental Chemistry of Soils. Oxford University Press, New York. 406p. McLaren, R.G., R.S. Swift, and J .G. Williams. 1981. The adsorption of copper by soil materials at low equilibrium solution concentrations. J. Soil Sci. 32:247-256. McLaren, R.G., J.G. Williams, and RS. Swift. 1983. Some observations on the desorption and distribution behavior of copper with soil components. J. Soil Sci. 34:325-331. Mehra, GP, and ML. Jackson. 1960. Iron oxide removal from soils and clays by a dithionite-citrate system buffered with sodium bicarbonate. Clays Clay Miner. 7:317-327. Murphy, E.M., and J .M. Zachara. 1995. The role of humic substances on the distribution of organic and inorganic contaminants in groundwater. Geoderma 67:103-124. Nkedi-Kizza, P., P.S.C. Rao, R.E. Jessup, and J.M. Davidson. 1982. Ion exchange and diffusive mass transfer during miscible displacement through an aggregated Oxisol. Soil Sci. Soc. Am. J. 46: 471-476. Padrnanabharn, M. 1983a. Adsorption-desorption behavior of copper (II) at the goethite- solution interface. Aust. J. Soil Res. 21: 309-320. Padrnanabham, M. 1983b. Comparative study of the adsorption-desorption behavior of copper (H), zinc (II), cobalt (II) and lead (11) at the goethite-solution interface. Aust. J. Soil Res. 21: 515-525. Petruzzelli, G., L. Lubrano, B.M. Petronio, M.C. Gennaro, A. Vanni, and A. Liberatori. 1994. Soil sorption of heavy metals as influenced by sewage sludge addition. J. Environ. Sci. Health A29 (1):31:50. 35 Piccolo, A., and F .J. Stevenson. 1982. Infrared spectra of Cu”, Pb”, and Ca2+ complexes of soil humic substances. Geoderma, 27:195-208. Raij, B. van, and M. Peech, M. 1972. Electrochemical properties of some Oxisols and Alfisols in the tropics. Soil Sci. Soc. Am. Proc. 36:587-593. Schindler, P.W., P. Liechti., and J .C. Westall. 1987. Adsorption of copper, cadmium and lead fi'om aqueous solution at the kaolinite/water interface. Netherlands J. of Agric. Sci. 35:219-230. Schwertrnann, U. 1964. Differenzienmg der eisenoxide des bondes durch extraktion mit arnmonium-oxalat-losung. Z. Pflanzenemahrung. 105:194-202. Schwertrnann, U., and RM. Taylor. 1977. Iron oxides. In J .B. Dixon, and SB. Weed (ed.) Minerals in Soil Environments. p. 145-180. SSSA, Madison. Spark, K.M., J .D. Wells, and BB. Johnson. 1995a. Characterizing heavy-metal adsorption on oxides and oxyhydroxides. Europ. J. Soil Sci.46:621-631. Spark, K.M., J .D. Wells, and BB. Johnson. 1995b. Characterizing trace metal adsorption on kaolinite. Europ. J. Soil Sci.46:633-640. Spark, K.M., J .D. Wells, and BB. Johnson. 1997a. The interaction of a humic acid with heavy metals. Aust. J. Soil Res. 35289-101. Spark, K.M., J .D. Wells, and BB. Johnson. 1997b. Characteristics of sorption of humic acid by soil minerals. Aust. J. Soil Res. 35:103-112. Spark, K.M., J.D. Wells, and BB. Johnson. 1997c. Sorption of heavy metals by mineral- humic acid substrates. Aust. J. Soil Res. 35:113-122. Sparks, D.L. 1985. Kinetics of ionic reactions in clay minerals and soils. Adv. Agronomy 38 :23 1-266. Sparks, D.L. 1989. Kinetics of Soil Chemical Processes. Academic Press, New York. 210p. Sposito, G. 1984. The Surface Chemistry of Soils. Oxford University Press, New York. 234p Stevenson, F.J., and MS. Arkadani. 1972. Organic matter reactions involving micronutrients in soils. In J .J . Mortvedt et al. (ed.) Micronutrients in Agriculture. p. 79-114. SSSA, Madison. 36 Stumm, W. 1992. Chemistry of the Solid-Water Interface . John Wiley and Sons, New York. 428 p. Temminghoff, E.J.M., S.E.A.T.M. Van der Zee, and RAM. De Haan. 1997. Copper mobility in a copper-contaminated sandy soil as affected by pH and solid and dissolved organic matter. Environ. Sci. Technol. 31 :1 109-1 1 15. Van der Zee, S.E.A.T.M., and W.H. Van Riemsdijk. 1991. Model for the reaction kinetics of phosphate with oxides and soil. In G.H. Bolt et al. (ed.) Interactions at the Soil Colloid-Soil Solution Interface. p. 205-239. Kluwer Academic Publishers, Dordrecht. Wang, F., and J. Chen. 1997. Modeling sorption of trace metals on natural sediments by surface complexation model. Environ. Sci. Technol. 31:448-453. Chapter 3 AGING EFFECTS ON COPPER ABSORPTION AND DESORPTION IN PHOSPHATE- OR SLUDGE-TREATED OXISOLS. Abstract Copper mobility in soils is influenced by soil properties, solution composition, and the Cu residence time in soil. This study evaluated the effects of adsorption reaction time (aging l h to 54 d) on Cu adsorption and desorption distribution coefficients (Km and K“) and on Cu desorption from phosphate- and sludge-treated Oxisols at pH 5.5 for 50 and 150 pM Cu. Copper desorption for each adsorption aging time was measured at desorption times fiom 0.25 h to 18 (1. Values of K“, were 2 to 8 times greater for 54-d than l-h aging, whereas K,” values increased only 2- to 4-fold over the same aging period. Aging caused Cu desorption to decrease by as much as 81%. Samples that were aged 3 d or less exhibited rapid Cu desorption followed by readsorption, probably because 3 d was insufficient for Cu adsorption to reach steady state. Values of K“, and K,“ and the effects of aging followed the trends P-treated > sludge-treated > control soils, 50 > 150 M Cu, and A > B horizon. This work showed that Cu desorption from aged soils can be overestimated more than two-fold by measuring desorption after a 24- or 72-h adsorption reaction, as is typically done in laboratory experiments. Both P and sludge treatment caused decreased Cu desorption, and such treatments might be used to control the availability and mobility of Cu in Oxisols. 37 38 Introduction The availability and mobility of Cu in the environment is greatly influenced by adsorption/desorption reactions. The importance of organic matter for Cu adsorption in soils and sediments has been discussed by several authors (Stevenson and Arkadani, 1972; McLaren et al., 1981; Petruzzelli et al., 1994; Spark et al., 1997a ; Temminghoff et al. 1997). Numerous studies have reported Cu adsorption characteristics of variable-charge soils (Barrow et al., 1981; James and Barrow, 1981; Barrow; 1985), as well Fe oxides (Jenne, 1968; Forbes et al., 1976; Schwertrnann and Taylor, 1977; Benjamin and Leckie, 1981), Al oxides (McBride, 1982; Harsh and Doner, 1984; Spark et al., 1995a), and kaolinite (Schindler et al., 1987; Spark et al., 1995b). Additional attention has been given to the effects of anions on Cu adsorption in variable-charge soils or pure sorbents (Guilherme et al., 1995; Harter and Naidu, 1995; Murphy and Zachara, 1995; Ali and Dzombak, 1996a; 1996b; Spark eta1., 1997b; Chapter 2, this dissertation). In contrast, relatively fewer studies have addressed Cu desorption and the effects of aging on Cu adsorption and desorption. Copper desorption fiom organic matter and Fe and Al oxides is often limited (McLaren et al., 1983), and aging causes Cu desorption to decrease (Padmanabham, 1983a; Lehmann and Harter, 1984; Schultz et al., 1987; Hogg et al., 1993). Brennan et al. (1980) reported that incubation of soils up to 120 d with Cu caused Cu availability to plants to decrease by up to 70% compared to freshly applied Cu; the observed decreases in Cu availability were generally greatest for soils with the highest OC and Fe and Al oxide contents. Decreased Cu availability due to aging has been attributed to (i) movement of Cu ions fiom low energy sites to higher energy sites, (ii) alteration of site energy under the influence of the Cu ion, (iii) movement of Cu ions to internal sites less accessible to solution phase, or (iv) formation of a separate solid phase (Lehmann and Harter , 1984). Padmanabharn (1983a) suggested that monodentate Cu- surface complexes are readily desorbed from goethite, whereas Cu in bidentate surface 39 complexes is less readily desorbed. Sorption hysteresis is greater for Cu”, Co“, and Zn” than for the much larger P 2" ion, which led Padmanabharn (1983b) to conclude that the smaller ions could be incorporated into the goethite lattice by forming a bridge with two surface Fe atoms on the oxide surface, whereas Pb2+ is too large to be accommodated. Other authors have attributed slow adsorption and desorption of heavy metals to diffusion either into the crystal matrix or into surface sites between aggregates of crystals (Barrow, 1985; 1986b; 1989; Van der Zee and Van Riemsdijk, 1991). Benjamin and Leckie (1981) reported a rapid initial (1 h) adsorption of Cd followed by a much slower second step, possibly related to solid-state diffusion in amorphous Fe hydrous oxide. Aringhieri et al. (1985) also found that Cu sorption kinetics in an organic soil were consistent with a dependence on internal diffusion. The time-dependence of heavy metal sorption on goethite has been attributed to initial adsorption on external surface sites, with subsequent diffusion to internal sorption sites (Brilemmer et al., 1988). Intraparticle diffusion coefficients are reported to be at least one order of magnitude smaller than bulk- solution mass transfer coefficients (Lo and Leckie, 1993; Michard et al., 1996). Intraparticle diffusion is less affected by variables such as pH and particle size, but more affected by pore-size distribution than is external mass transfer (Michard et al., 1996). Differences in mesopore shapes have been used to explain the much slower phosphate desorption from lepidocrocite, with has cylindrical mesopores that are considered less accessible to solution, than from hematite, which has a mixture of cylindrical and slit-shaped mesopores (Madrid and Arambarri, 1985). In highly aggregated Oxisols, long-term sorption rates might be limited by infra-aggregate rather than intraparticle mass-transfer (Nkedi-Kizza et al., 1982). Although batch-shake methods eliminate much intra-aggregate mass-transfer control normally found in aggregated Oxisols (Lima, 1995; Lima and Anderson, 1997), intraparticle diffusion may still be important in batch-shake systems. 40 The objectives of this study were to test the hypothesis that the effect of aging on Cu sorption and desorption differs between P-treated, sludge-treated, and untreated Oxisols, and also depends on soil mineralogy and OC content. A previous experiment has shown that P and sludge treatment in Oxisols cause Cu adsorption kinetics to increase but Cu availability (desorption) to decrease (Chapter 2), but the effect of aging on Cu sorption reversibility in P- and sludge-treated soils has not been studied previously. Materials and Methods Soil Material Samples of A and B horizons from two uncultivated Oxisols, a Dark-Red Latosol and a Yellow-Red Latosol, were used in this study. Both soils have pH near 4.5 in the A horizon and 5.5 in the B horizon. Although both soils are classified as very fine, allitic, isothermic Typic Hapludox, differences in drainage have caused one soil (a Yellow-Red Latosol) to have greater kaolinitezgibbsite and goethite:hematite ratios than the other (a Dark-Red Latosol) (Table 3.1). The A and B horizons of each soil differ in OC content as well as in the relative proportions of amorphous and crystalline Fe, estimated as the ratio of oxalate-extractable to dithionite-extractable Fe. Additional details concerning soil characterization have been described elsewhere (Lima and Anderson, 1997 ). Porosity was calculated from N2 adsorption isotherms (Appendix A4). Total porosity was taken from the experimental isotherms at a partial pressure of 0.99, as suggested by Lippens and de Boer (1964). Micropore (rp < 20 A) volmne was calculated by the method of Horvath and Kawazoe (1983) and accounted for 15 to 18% of the total 41 pore volume in all samples. Mesopores (rp > 20 A) accounted for the remaining pore volume (82 to 85%) and could be classified as slit-shaped pores (de Boer and Lippens, 1964). No macroporosity was detected in any case, as the soil samples showed no steep gradient at partial pressure values near 1.0. Differences in percent micro and mesoporosity between soil samples were not significant (P > 0.05). Table 3.1. Selected properties of A and B horizons of two Oxisols from Brazil. Horizon Clay 0C1 Fed: F e,1 Kt‘ Gb15 Gtsz SSA” Total porosity 2-1 3-1 g kg" soil —-——— g kg" clay m g cm g Dark-KW A 691 25.3:I:0.0 99 1.88 350 510 5.0 59:1:3 0.154i0.019 B 753 9.9i0.l 114 0.65 350 510 4.1 61:1:3 0.137i'0.016 Idiom-WWW A 711 23.6i0.3 101 1.20 480 375 10.2 53:4 01281-0019 B 721 8.0101 114 0.59 480 400 8.3 58i3 0.143i0.024 1 CC is organic carbon measured by the Walkley-Black method. * F d and Feo, respectively, are Fe extracted by dithionite-citrate-bicarbonate (Mehra and Jackson, 1960) and ammonium oxalate (pH 3.1) in the dark (Schwertrnann, 1964). 5 Kt and Gb, respectively, are kaolinite and gibbsite measured by differential thermal analysis in dithionite-treated clay samples (but expressed on a total-clay basis). 1 Gtsz is the goethite:hematite ratio in the clay fiaction, measured by X-ray diffraction of NaOH-treated samples (Kampf and Schwertrnann, 1982a; 1982b). ” Specific surface area determined by N2 adsorption isotherms (BET). 42 Soil Pretreatment Soils were pretreated with phosphate (P) by reacting samples of each soil material with 10.75 mM Ca(H2PO,)2-HZO (21.5 mmol P L") at a soil:solution ratio of 2:3 to give a P addition rate of 1 g P kg" soil, then washing with 5 mM Ca(NO,)2 to remove excess P. The adsorbed P content of all P-treated soils was near 30 mmol kg". For sludge pretreatment, soils were reacted for 30 d with air-dried, <0.5-mm lime-stabilized sludge at a rate of 5 g kg" soil and a moisture content of 60% of saturation. The dry sludge had an OC content of 276 i 8 mg kg"; other sludge properties are reported in Appendix A1. The sludge-treated soils contained about 2 g kg" more OC (Appendix A2) than the control soils. Although P- and sludge-treatrnent caused moderate increases in DTPA-extractable Cu (Chapter 2), the DTPA-extractable Cu concentrations were about three orders of magnitude less than the Cu concentrations adsorbed in the experiments described below. The P- and sludge-treated soils were air-dried and gently crushed to pass a 2-mm sieve. Cu Sorption Copper sorption was measured in batch-shake experiments. Prior to Cu sorption experiments, all soil suspensions were adjusted to pH 5.5 as described in Chapter 2. Copper sorption capacity at pH 5.5 was measured by repeated reaction with 500 M Cu(NO,)2 in a Ca(NO3)2 background electrolyte (I=15 mM) at a soil:solution ratio of 1:100. Suspension pH was readjusted to 5.5 after every two adsorption cycles. Adsorption was complete after nine reactions with 500 W Cu. To assess the effect of aging upon Cu adsorption and desorption, triplicate samples of untreated and P- or sludge-treated soils were reacted at pH 5.5 for l h, 3 d, 18 d, and 54 d with 50 and 150 W Cu(NO3)2 in Ca(NO3)2 (I = 15 mM; 1:100 soil:solution). After the specified adsorption reaction time, the suspensions were centrifuged, the supernatant 43 solutions were decanted, and the centrifuge bottles were weighed to determine the mass of entrained solution. The Cu concentrations in the supernatant solutions were measured by flame atomic absorption spectroscopy. The soil pastes that remained in the centrifuge bottles were then reacted with 5 mM Ca(NO3)2 at pH 5.5. Copper desorption was measured after 0.25 h,‘ 1 h, 6 h, 24 h, 3 d, 6 d, and 18 d. The concentration of Cu desorbed at each successive desorption reaction time was corrected for the amount of copper in the entrained solution. Distribution coefficients for Cu adsorption at each of the four aging times (Km) were calculated with the equation Km: = {Cum/[Curl [1] where {Cum} is the adsorbed Cu concentration (mmol kg"), and [CuT] is the total solution-phase Cu concentration (mmol L") at each aging time. Distribution coefficients for Cu desorption (Km) after each of the four aging times were calculated for the 1-h desorption reaction period with the equation Km = {CMMCUTI [2] where {Cum} is calculated here as the difference between the adsorbed Cu concentration at the end of the aging period and the concentration of Cu desorbed after l-h reaction with 5 mM Ca(NO,),. Results Adsorption Distribution Coefl‘icient The effect of increased adsorption reaction (aging) time on K“, is shown in Figure 3.1. Values of K“, ranged from about 50 in control B horizons at 150 W Cu to 9000 in P- treated A horizons at 50 M Cu. Both the K“, values and the effect of aging on K,d5 followed the general trends P-treated > sludge-treated > control soils, 50 > 150 W Cu, 44 and A > B horizons, though aging effects in sludge-treated soils at 50 [TM Cu were greater for B than A horizons. Although aging effects differed slightly between the Gme and Kth soils, there were no consistent trends, and the effect of aging on Km, depended much less on soil mineralogy and horizon than on soil pretreatment and initial Cu concentration. The overall increases in Kw as aging time increased flour 1 h to 54 d ranged from less than two-fold for control samples reacted with 150 nM Cu to eight-fold for P-treated A-horizon samples reacted with 50 MI Cu (Figure 3.1). As aging time increased from 1 h to 18 (1, Km always increased. However, a further increase in aging time from 18 to 54 (1 caused little or no additional increase in Km: in many of the samples. Only P-treated A- horizon samples exhibited a consistently large increase in Km as aging time increased from 18 to 54 d, whereas Km for P-treated B-horizon samples actually decreased over that time period. The K“, values plotted in Figure 3.1 show that it is not possible to predict a priori whether the largest increases in K“, occur between 1 and 3 d, 3 and 18 d, or between 18 and 54 (1. Fraction of Copper Desorbcd To show the effect of aging on the time-dependence of Cu desorption, Cu desorption is plotted as the fraction of Cu desorbed (i.e., Cu desorbed after the specified desorption time / Cu adsorbed at the end of the specified aging time, hereafter Cum/Cum). The time-dependence and the effects of soil type and pretreatment were the same for both Cu concentrations, even though Cum/Cu“, for 150 M Cu (Figure 3.2) was about twice that for 50 M Cu (Figure 3.3). Copper desorption is least for the samples aged 18 and 54 d, and greatest for samples aged 1 h. The effects of pretreatment, initial Cu concentration, and soil type and horizon on Cum/Cu“, and on the time dependence of Cudfl/Cumls for 45 samples aged 18 d have been discussed in Chapter 2, so only the effects of aging will be emphasized here. :9. ._V 3.: mo. 3 2 3 e .. 9 .d M m 8 MW m 5 ...n m w .r. m .m ] ”I U o C .. m. [ O C 0 u. a. M .0 # m 0 e 5 m 1 fish a w m m .m ] w U . m C .u m [ O C 10007; 1000 — 100 g 1000 - 10000 g :9. ._V ix = Dark-Red Latosol; Kth Figure 3.1. Distribution coefficients (Cu adsorbed/Cu supernatant) for Cu adsorption in control and P- and sludge-treated A and B horizon samples of two Oxisols reacted with 150 and 50 uMCu for 1 h, 3 d,18 d, and 54d(Gme Yellow-Red Latosol). 46 .9893 "3.3:; u 66— €83 Bmsfia u .539 c «m c5 a M: a m .5 .5. so :1 ca 55 382 £8me 93 .3 838.8 canto: m 98 < 3395632.». .8 -m 93 .9580 89a 39.3% :0 mo .5qu .«o oogweogovbgh .N.m gamma 3 2...... 53930: ooow 2:. or r v.0 cow 2. v .3 2:. 2. w .3 2:. S. _‘ rd _ _ . ad . u a I mu a sum m. a on u M a W .0. 3 o a an a a u u a mum m M M w W o 0. n n u a M a. a a c o _..o ) mu m I u u m In I H o M mm a o O o O .flm u n n no u D m I mum m m 0 O D a O O . m. o u u no w 8 . .CQ n: W - .-i I 0.: O n I H H I w W. M a an n u a u I w w u an n m n n Fe m I u o c a C D I m o o o o u a a m m u o o o mummm no man” nova. SW 0 m a o 03. 4 3mm m on a a... a a... _ .2260 _ f . .353: m canto: < canto: m caste: < “0.! Elam» 47 .2803 00M§O=0> .1. 000M 28304 033.qu u 8:09 0 0m 0:0 .0 w" .0 m .n _ 8.“ =0 33 cm 505 00.0000 £805 23 .«o «.2953 comic: m 05 < 0000050320 8 -m 05 .8280 Bob 020800 :0 .3 5080 me 08005900083 .m.m 0§mE E as: 8.5.83 822: 2 F 2. 2: 2 F to 2: 2 P 3 2: 2 P we 3.. Pup-:0 r...- r.... :b fl...- r.... '33.. 3.. run- rip. r3:- .... r33:- rlB... r5:- . .. , J 80 .u. l: .v v ”a: a .fi an mun“ an” momwamu. —mam$mumwuu -.. ...-8.o I I u o . 4. mm mm Um now a mu ” u T W .0 J... a m 1H. 1.....m m mu m u n u m .. L. 1 . a 092.6 ”2.0 a .,..,......,,-H __,. a. ,,,,, i 1..-: n nu" . . u 22m I. I I ¢ . nmmuu umwma o nun ufi .P mam“: 2“. a“: are; u m n n n H M u ”IO—.6 m n 0 225.02... mm; W woe... m Sum“; .32....” e mummmnw mmmummmmWBdm Swim: “gm: .nm mm w w w m 03. . ”.2.0( mm m u no a 3: . n 3 o. . mm _ 30:00 L f . an; Sara: m :38: < c302. m canto: < ~92 Elam 48 The time-dependence of Cum/Cu“, and the effect of aging on the time-dependence of Cu desorption differ greatly between control, P-treated, and sludge-treated soils, as well as between A and B horizons. For samples aged 18 or 54 (1, Cu desorption from all A horizon samples and from P-treated B horizons was complete after 1 h, whereas control and sludge-treated B horizons required 6 to 24 h to reach desorption steady-state. For the 1-h and 3-d adsorption aging times, Cum/Cu“, reached a maximum afier 0.25 to 24 h desorption, with subsequent readsorption at longer desorption times. Readsorption always commenced sooner for samples aged 1 h than for samples aged 3 (1. Increasing the aging time from 1h to 54 d cause a 40 to 80% decline in values of Cum/Cum measured after a l-h desorption reaction. The l-h desorption time was chosen because it is relevant for understanding Cu remobilization in soil during a heavy rain. The effect of aging on Cumm/Cum at 1 h was not consistently greater in P-treated than sludge- treated and control soils, nor in A than B horizons (Figures 3.2 and 3.3). Desorption Distribution Coefi‘icz'ent Values of K,“ calculated for a 1-h desorption time increased as aging increased from 1 to 54 d (Figure 3.4), although aging effects on K“ were only about half those on K“, (Figure 3.1). K,“ values ranged from about 350 for control B-horizon samples aged 1 h with 150 M Cu to 9000 for P-treated A-horizon samples aged 54 d with 50 W Cu. As aging time increased from 1 to 54 d, the increase in K0. ranged from two-fold for control and sludge B horizons at 150 M Cu to four-fold for P-treated A horizons at 50 M Cu. Both the K,“ values and the effect of aging on Km followed the general trend P-treated >sludge > control soil and 50 > 150 M Cu, as previously noted for K“. However, the effects of aging on Kd,s differed less among the different soil pretreatments and initial Cu concentrations (Figure 3.4) than was noted previously for Km (F igure 3.1). 49 .......................... wn.u.u.m.....u. .................................. ooooooooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooooooo oooooooooooooo v ............... 3.3. ..................... oooooooooooooooooooooooo Dark-Red Yellow-Red Latosol). Figure 3.4. Distribution coefficients of Cu (Cu remaining adsorbed/Cu supernatant) after 1 h of Cu desorption from control and P- or sludge-treated A and B horizon samples of two Oxisols reacted with 150 and 50 M Cu for 1 h, 3 d, 18 d, and 54 d (Gme Latosol; Kth 50 Except for P- and sludge-treated samples reacted with 50 M Cu, the effects of aging on Km were similar in A and B horizons, and there was no effect of soil mineralogy. In general, Km (Figure 3.4) was greater than Kw (Figure 3.1), except for P- and sludge-treated soils at 50 M Cu, where the ratio chs/Km was near unity (Figure 3.5). The ratio Km/Km decreases with aging time and was greater for control than P- or sludge- treated soils, for 150 M than 50 W Cu, and for B than A horizons. The ratio of Km/K,‘is ranged from 1 to 2 for P- and sludge-treated samples at 50 W Cu to 5 or 6 for control B samples at 150 M. To evaluate whether Km/K“, values greater than unity are evidence for sorption hysteresis or simply a result of isotherm nonlinearity (i.e., greater K values at lesser sorbed Cu concentrations), the sorbed Cu concentrations are plotted as a function of solution-phase Cu concentration (Figures 3.6-3.8). Although the shape of the adsorption and desorption isotherms is not well defined, because only two initial Cu concentrations were used in this study, it generally is apparent that the adsorption and desorption data do not fall on the same smooth curve, and that sorption exhibits hysteresis. Based on the extent to which desorption points lie above the adsorption data points, hysteresis is greatest for control soils and least for P-treated soils, and is much great for B than A horizons. For most samples, hysteresis is most evident for samples aged 1 h and least evident for samples aged 54 d. Strictly speaking, however, hysteresis cannot be evaluated unless adsorption and desorption times are the same, so it is not possible to use l-h desorption data to ascertain aging effects on sorption hysteresis. [CU] initial 5oflM [ Cu] initial 150M" ..... \\\\\ .. \\ s\\\\\ m h l i h mA .. tA w\ \t3 \\ % mm ////////// itrKr ////////////z \\\\\\\\\\\\\ ..... \\\\\\\\\\\\\\\\\\\\ r _ . r . z _ _ . z z . . z . . _ n 6 4 2 o 6 4 2 o 6 4 2 o samples of two Oxisols reacted with 150 and 50 M Cu for 1 h, 3 d, 18 d, and 54 (1. Figure 3.5. Ratio of Km, , h/KIds in control and P- or sludge-treated A and B horizon 52 Cu adsorbed (mmol kg") i T Supernatant [Cut] (pM) Figure 3.6. Sorbed Cu concentration as a function of solution-phase Cu concentration in control Oxisols samples reacted with 150 and 50 pM Cu for l h, 3 d, 18 d, and 54 d. 53 12 ‘GmeAhon'zon v - 9 . ' - 6 ~ mm mm ":;.' . 1h 0 3 ../ I 3d D A 18d A 8 1 r r l r r 1 l r r I '1 541d 1v 1 l 1 iGmeBhon'zon a; 8 __‘ Luw‘ 3:1 6 4‘ / E 2- E. 'U 0 + i e 1 + l + l l .E 12 KthA mm, a 9— '8 :3 6‘ U 3 _ o a : l l0 l + KthBhoflzm . r . 1 . . . . . l ' ' T I 0 5 1O 15 20 Supernatant [Cut] (pM) Figure 3.7. Sorbed Cu concentration as a function of solution-phase Cu concentration in P-treated Oxisols reacted with 150 and 50 nM Cu for 1 h, 3 d, 18 d, and 54 d. 54 12 5 GmeA horizon Cu adsorbed (mmol kg") o.,.,..,,...,.-., 0 5 10 15 20 Supernatant [Cut] (pM) Figure 3.8. Sorbed Cu concentration as a function of solution-phase Cu concentration in sludge-treated Oxisols reacted with 150 and 50 pM Cu for 1 h, 3 d, 18 d, and 54 d. 55 Discussion The effect of P treatment on both K“B and K,“ was greater than the effect of sludge treatment and of soil 0C content (A vs. B horizon) or soil mineralogy (Kth vs. Gme). However, the effects of initial Cu concentration and aging time were at least as great as the effect of P treatment. The decrease in K“, and K“, as initial Cu concentration increases indicates that sorption sites differ in their affinity for Cu, with high-affinity sites occupied at low surface coverage. The pronounced effect of P treatment on Km: and Km suggests that inner-sphere complexes of P with metal oxides increases the affinity of the surface for Cu, and that this effect is enhanced by aging. An increase in K“, and Km caused by sludge treatment might be related to the high affinity of humic substances for Cu (McLaren at al., 1981), and has been reported previously by Petruzzelli et al. (1994). Similarly, the higher Km observed for A than B horizons is likely caused by the higher 0C content of the A horizon compared with the B horizon. The smaller Km, observed for the Kth soil compared with the Gme soil is consistent with the very low value of Cu distribution coefficient reported for kaolinite by McLaren et a1. (1981). The marked effects of aging upon K“, and K,” have not been formally reported in the literature. However, based on the original data of Padmanabham (1983a) (Figs. 1 and 2, pages 313 and 314) one could estimate a small increase (about 5 to 10%) in K“, for Cu in goethite due to aging from 12 to 19 d. The increases in Km and K“, indicate that short- term adsorption experiments underestimate Cu adsorption and overestimate the solution- phase Cu concentration in aged soils. The fact that samples reacted for either 1 h or 3 (1 showed Cu readsorption during the desorption period is additional evidence that adsorption does not reach steady-state in 3 d . Similar results have been reported for P adsorption onto Fe oxides (Cabrera et al., 1981; Madrid and Arambarri, 1985). The 80% decrease in Cu desorption with increasing aging time corroborates the results of Brennan et al. (1980), who found that Cu S6 availability to plants decreased up to 70% with increased aging. In that study, aging caused Cu availability to decrease more in samples with higher 0C content (Bremen et al. (1980). However, in the present study, aging effects on Cum/Cum did not differ between A and B horizons nor between sludge-treated and control soils (Figures 3.2 and 3.3). The increases in K,“ with increasing aging time, together with the decrease in Km/K,Ids for longer aging times indicates that although Cu desorption decreases with aging time, the reaction may approach steady state after 54 d. Decreased solution-phase ion concentrations with increased aging has been attributed to either diffusional processes (Madrid and Arambarri, 1985; Barrow, 1986a; Strauss et al., 1997) or to changes in sorption mechanism (surface speciation) (Padmanabham, 1983a; 1983b; Lehmann and Harter , 1984). The aging effects observed in the present study may be caused by both changes in surface speciation and by diffusion. Mesopores with radii greater than 20 A compose over 80% of the total porosity in these soils, and internal diffusion of hydrated Cu2+ (radius ~ 3.5 A) would not be restricted by pore size. However, micro/mesoporosity and pore shape did not vary significantly (P > 0.05) between soil samples, so the differences in aging effects between control and P- or sludge-treated soils, and between A vs. B horizon cannot be explained solely on the basis of diffusion. Therefore, specific chemical interactions between Cu and surface functional groups, and changes from monodentate to bidentate complexes are likely to be a more important cause for the greater aging effect in P-treated samples compared with either control or sludge-treated samples, and in A horizon compared with B horizon. In addition, aging likely causes changes in the surface-phosphate and surface-sludge interactions, so aging effects are more difficult to explain precisely in P- and sludge-treated soils. In conclusion, this study suggests that aging effects upon Cu desorption and Cu adsorption and desorption distribution coefficients in Oxisols are likely to be caused mainly by time-dependent changes in surface speciation, though by diffusional processes may also play a role. However, the fact that aging has a greater effect on K“, than on Kdes 57 has implications for the use or misuse of distribution coefficients for modeling purposes. Models that ignore this will certainly overestimate Cu availability in Oxisols. Copper desorption from aged Oxisols can be overestimated as much as two-fold if desorption is measured afier 1- or 3-d adsorption reactions, as is typically done in laboratory experiments. Both P and sludge treatment caused decreased Cu desorption and increased the effect of aging on Cu sorption reactions 58 LIST OF REFERENCES Ali, M.A., and DA. Dzombak. 1996a. Effects of simple organic acids on sorption of Cu2+ and Ca2+ on goethite. Geochim. Cosmochim. Acta 60:291-304. Ali, M.A., and DA. Dzombak. 1996b. Interactions of copper, organic acids, and sulfate in goethite suspensions. Geochim. Cosmochim. Acta 60:5045-5053. Aringhieri, R., P. Carrai, and G. Petruzzelli. 1985. Kinetics of Cu“ and Cd“ adsorption by an Italian soil. Soil Sci. 139:196-204. Barrow, NJ. 1985. Reactions of anions and cations with variable-charge soils. Adv. Agronomy 38:183-230. Barrow, NJ. 1986a. Testing a mechanistic model. I. The effects of time and temperature on the reaction of fluoride and molybdate with soil. J. Soil Sci. 37: 267-275. Barrow, NJ. 1986b. Testing a mechanistic model. H. The effects of time and temperature on the reaction of zinc with soil. J. Soil Sci. 37: 277-286. Barrow, NJ. 1989. The reaction of plant nutrients and pollutants with soil. Aust. J. Soil Res. 27:475-492. Barrow, N.J., J .W. Bowden, A.M. Posner, and J .P. Quirk. 1981. Describing the adsorption of copper, zinc and lead on a variable-charge mineral surface. Aust. J. Soil Res. 19:309-321. Benjamin, M.M., and J .O. Leckie. 1981. Multiple-site adsorption of Cd, Cu, Zn, and Pb on amorphous iron oxyhydroxide. J. Colloid Interface Sci. 79:209-221. Brennan, R.F., J .W. Gartrell, and AD. Robson. 1980. Reactions of copper with soil affecting its availability to plants. I. Effect of soil type and time. Aust. J. Soil Res. 18:447-459. Brilemmer, G.W., J. Gerth, and K.G. Tiller. 1988. Reaction kinetics of adsorption and desorption of nickel, zinc, cadmium by goethite. 1. Adsorption and diffusion of metals. J. Soil Sci. 39:37-52. Cabrera, F ., P. Arambarri, L. Madrid, and CG. Toca. 1981. Desorption of phosphate from iron oxides in relation to equilibrium pH and porosity. Geoderma. 26:203-216. 59 de Boer, J .H., and BC. Lippens. 1964. Studies on pore systems in catalysts. II. The shapes of pores in aluminum oxides systems. J. Catalysis. 3:38-43. Forbes, E.A., A.M. Posner, and J .P. Quirk. 1976. The specific adsorption of divalent Cd, Co, Cu, Pb and Zn on goethite. J. Soil Sci. 27:154-166. Guilherrne, L.R.G., J .M. Lima, and S]. Anderson. 1995. Efeito do fosforo na adsorcao de cobre em horizontes A e B de latossolos do Estado de Minas Gerais. p. 316-318. In: Resumos expandidos. Congresso Brasileiro de Ciéncia do Solo 25‘”, Vicosa. 23-29 July 1995. SBCS/UFV, Vicosa, Brazil. Harsh, J .B., and HE. Doner. 1984. Specific adsorption of copper on an hydroxy- alurninum-montmorillonite complex. Soil Sci. Soc. Am. J. 48:1034-1039. Harter, RD, and R. Naidu. 1995. Role of metal-organic complexation in metal sorption by soils. Adv. Agronomy 55: 219-263. Horvath, G., and K. Kawazoe. 1983. Method for the calculation of effective pore size distribution in molecular sieve carbon. J. Chem. Eng. Japan. 470-475. Hogg, D.S., R.G. McLaren, and RS. Swift. 1993. Desorption of copper from some New Zealand soils. Soil Sci. Soc. Am. J. 57:361-366. James, R.O., and NJ. Barrow. 1981. Copper reactions with inorganic components of soils including uptake by oxide and silicate minerals. In: Copper in Soils and Plants. J .F. Loneragan, A.D. Robson, and RD. Graham (eds.), p. 47-68. Academic Press, Sydney. J enne, EA. 1968. Controls on Mn, Fe, Co, Ni, Cu, and Zn concentrations in soils and water; the significant role of hydrous Mn and Fe oxides. In: Trace Inorganics in Water . Advances in Chemistry Series, 73. Am. Chem. Soc., p.:337-387. Kampf, N., and U. Schwertrnann. 1982a. Quantitative determination of goethite and hematite in kaolinitic soils by X-Ray diffraction. Clay Miner. 17:359-363. Kampf, N., and U. Schwertrnann. 1982b. The 5-M-NaOH concentration treatment for iron oxides in soils. Clays and Clay Miner. 30:401-408. Lehmann, R.G., and RD. Harter. 1984. Assessment of copper-soil bond strength by desorption kinetics. Soil Sci. Soc. Am. J. 48:769-772. Lima, J .M. 1995. Relation between phosphate sorption and aggregation in Oxisols from Brazil. East Lansing: Michigan State University, 87p. (Ph.D. Dissertation). Lima, J .M., and 8.]. Anderson. 1997. Effect of aggregation and aggregate size on extractable Fe and Al in two Brazilian Typic Hapludoxs. Soil Sci. Soc. Am. J. 61:965-970. Lippens, BC, and J .H. de Boer. 1964. Studies on pore systems in catalysts. III. Pore-size distribution curves in aluminum oxide systems. J. Catalysis. 3:44-49. Lo, K.S.L., and J .0. Leckie. 1993. Kinetics studies of adsorption-desorption of Cd and Zn onto A1203/solution interfaces. Water Sci. Technol. 28:39-45. Madrid, L., and P. Arambani. 1985. Adsorption of phosphate by two iron oxides in relation to their porosity. J. Soil Sci. 36:523-530. McBride, MB. 1982. Cu2+ adsorption characteristics of aluminum hydroxides and oxyhydroxides. Clays Clay Miner. 30:21-28. McLaren, R.G., R.S. Swift, and J .G. Williams. 1981. The adsorption of copper by soil materials at low equilibrium solution concentrations. J. Soil Sci. 32:247-256. McLaren, R.G., J .G. Williams, and RS. Swifi. 1983. Some observations on the desorption and distribution behavior of copper with soil components. J. Soil Sci. 34:325-331. Mehra, GP, and ML. Jackson. 1960. Iron oxide removal from soils and clays by a dithionite-citrate system buffered with sodium bicarbonate. Clays Clay Miner. 7 :3 1 7-327 . Michard, P., E. Guibal, T. Vincent, and P. Le Cloirec. 1996. Sorption and desorption of many] ions by silica gel: pH, particle size and porosity effects. Microporous Materials 5:309-324. Murphy, E.M., and J .M. Zachara. 1995. The role of humic substances on the distribution of organic and inorganic contaminants in groundwater. Geoderma 67:103-124. Nkedi-Kizza, P., P.S.C. Rao, R.E. Jessup, and J.M. Davidson. 1982. Ion exchange and diffusive mass transfer during miscible displacement through an aggregated Oxisol. Soil Sci. Soc. Am. J. 46: 471-476. Padmanabham, M. 19833. Adsorption-desorption behavior of copper (II) at the goethite- solution interface. Aust. J. Soil Res. 21: 309-320. Padmanabham, M. 1983b. Comparative study of the adsorption-desorption behavior of copper (H), zinc (II), cobalt (II) and lead (H) at the goethite-solution interface. Aust. J. Soil Res. 21: 515-525. 61 Petruzzelli, G., L. Lubrano, B.M. Petronio, M.C. Gennaro, A. Vanni, and A. Liberatori. 1994. Soil sorption of heavy metals as influenced by sewage sludge addition. J. Environ. Sci. Health A29 (l):3l:50. Schindler, P.W., P. Liechti, and J .C. Westall. 1987. Adsorption of copper, cadmium and lead from aqueous solution at the kaolinite/water interface. Netherlands J. Agric. Sci. 35:219-230. Schultz, M.F., M.M. Benjamin, and J .F . Ferguson. 1987. Adsorption and desorption of metals on ferrihydrite: reversibility of the reaction and sorption properties of the regenerated solid. Environ. Sci. Technol. 21:863-869. Schwertrnann, U. 1964. Differenzierung der eisenoxide des bondes durch extraktion mit ammonium-oxalat-ldsung. Z. Pflanzenernahrung. 105:194-202. Schwertrnann, U., and RM. Taylor. 1977. Iron oxides. In: Minerals in Soil Environments , (J .B. Dixon, and SB. Weed, eds.), p. 145-180. SSSA, Madison. Spark, K.M., J .D. Wells, and BB. Johnson. 1995a. Characterizing heavy-metal adsorption on oxides and oxyhydroxides. Europ. J. Soil Sci. 46:621-631. Spark, K.M., J .D. Wells, and BB. Johnson. 1995b. Characterizing trace metal adsorption on kaolinite. Europ. J. Soil Sci. 46:633-640. Spark, K.M., J .D. Wells, and BB. Johnson. 1997a. The interaction of a humic acid with heavy metals. Aust. J. Soil Res. 35:89-10]. Spark, K.M., J .D. Wells, and BB. Johnson. 1997b. Sorption of heavy metals by mineral- humic acid substrates. Aust. J. Soil Res. 35:113-122. Stevenson, F.J., and MS. Arkadani. 1972. Organic matter reactions involving micronutrients in soils. In: Micronutrients in Agriculture. J .J . Mortvedt et al. (eds.), p. 79-114. SSSA, Madison. Strauss, R., G.W. Briimmer, and NJ. Barrow. 1997. Effects of crystallinity of goethite: H. Rates of sorption and desorption of phosphate. European J. Soil Sci. 48: 101-1 14. Temminghoff, EJ.M., S.E.A.T.M. Van der Zee, and F.A.M. De Haan. 1997. Copper mobility in a copper-contaminated sandy soil as affected by pH and solid and dissolved organic matter. Environ. Sci. Technol. 31 :1 109-1 1 15. Van der Zee, S.E.A.T.M., and W.H. Van Riemsdijk. 1991. Model for the reaction kinetics of phosphate with oxides and soil. In: Interactions at the Soil Colloid-Soil Solution Interface G.H. Bolt et al. (eds.), p. 205-239. Kluwer Academic Publishers, Dordrecht. Chapter 4 COPPER ADSORPTION/DESORPTION IN PHOSPHATE- OR SLUDGE- TREATED OXISOLS AS AFFECTED BY pH Abstract Adsorption/desorption reactions of Cu in soils are affected by surface chemistry and solution composition. This study evaluated the effects of pH upon Cu adsorption/desorption by P-treated, sludge-treated, and control A- and B-horizon samples of two Oxisols reacted with 0, 5, 50, and 150 M Cu. Sorption isotherms changed from a L-type at pH 4.5 to a H-type at pH 6.5. Increasing pH from 4.5 to 6.5 generally caused Cu adsorption to increase more in control than pretreated samples, and in B than A horizons. At a given pH, more Cu was adsorbed by (and generally less Cu was desorbed fi'om) P- and sludge-treated than control samples, and by A than B horizons. Copper adsorption increased by as much as 3.2 times as pH increased fi'om 4.5 to 6.5, with the greatest increase in the B horizon of a Yellow-Red Latosol reacted with 150 M Cu and least increase in the A horizon of a Dark-Red Latosol reacted with 5 uM Cu. Increasing preequilibration pH caused the fraction of Cu desorbed to decrease from as much as 0.35 to less than 0.01 (Cu desorbed/Cu adsorbed < 1%) in all A-horizon samples. The fact that a considerable fraction of Cu still remained adsorbed at a preequilibration pH of 4.5 is noteworthy as this may reduce the availability of Cu in Oxisols even at low pH. 62 63 Introduction Copper concentrations in natural soil solutions and sediment-water systems typically are controlled by adsorption/desorption reactions (Ellis and Knezek, 1972; James and Barrow, 1981), particularly with surface hydroxyl groups of metal oxides and organic matter (McLaren et al., 1981; Piccolo and Stevenson, 1982; Wang and Chen, 1997). The retention of Cu by selective bonding processes at variable-charge mineral surfaces and layer silicate particle edges is a pH-dependent process usually termed chemisorption or specific adsorption (formation of inner-sphere complexes) (Forbes etal., 1976; McBride, 1994). The free energy of adsorption comprises both an intrinsic term that describes the chemical interaction between the metal and the surface and a Coulombic term for the electrostatic attraction at the surface, which may vary as a function of pH and surface coverage (Stumm and Morgan, 1996). Changes in pH affect not only the surface charge but also Cu speciation (Ritchie and Jarvis, 1986). Specifically adsorbed anions might complicate the effects of pH, because inner-sphere complexes of anions with variable- charge surfaces cause net surface charge to become less positive (more negative) and pHPZC to decrease (Sposito, 1989; Stumm, 1992; Kretzschmar et al., 1997; Lima and Anderson, 1997; Spark et al., 1997a). Copper adsorption by soils (Hatter, 1983; Barrow, 1985; Basta and Tabatabai, 1992; Carey et al., 1996), Al oxides (McBride, 1982; Hsu, 1989, Spark et al., 1995a), Fe oxides (McKenzie, 1980, Benjamin and Leckie, 1981, Padmanabham, 1983; Wang and Stumm, 1987; Spark et al., 19953) and organic matter (Kabata-Pendias and Pendias, 1992) increases as pH increases. Copper adsorption increases with increasing pH for two reasons 64 (Barrow, 1989). First, OH groups at variable-charge surfaces deprotonate at high pH, which decreases the electrostatic repulsion between Cu and the surface. Second, CuOH” is apparently more strongly sorbed than is Cu2+ (Barrow et al., 1981) . The pK. for solution- phase hydrolysis of Cu2+ to CuOH+ is 7.7 (Lindsay, 1979). Chernisorption of Cu by metal oxides occurs mainly in the pH range 4.5-6.5 (McBride, 1994). Copper sorption by humic acid is greatest between pH 4 to 5, whereas Cu sorption by fulvic acid is greatest at pH 6 to 7 (Kabata-Pendias and Pendias, 1992). For kaolinite, ion exchange (outer-sphere complexes) is reported to be important at low pH and low ionic strength, whereas an increase in both ionic strength and pH favors chemisorption at amphoteric surface hydroxyls (Schindler et al., 1987; Spark et al., 1995b). Metal hydrolysis and formation of metal oxides or hydroxides are all favored at high pH. Because metal precipitation might occur even before adsorption sites are occupied by the trace metal of interest, and even when bulk solution is undersaturated with respect to the metal hydroxide, the distinction between chemisorption and precipitation is not always clear (James and Healy, 1972; McBride, 1994; Sparks, 1995). However, electron spin resonance (ESR) studies (McBride, 1982) have shown that most Cu sorbed on noncrystalline alumina at pH 5 to 6 was chemisorbed on Al-OH groups (which may be the dominant surface functional group in Oxisol B horizons). Yet, the ESR signal for Cu chemisorption on goethite decreased as pH was raised from 4.8 to 7.6, which indicated that Cu likely sorbed as a surface precipitate at the higher pH (McBride, 1982). A surface precipitation model (which considers the formation of a surface phase whose composition varies continuously between that of the original solid and a pure 65 precipitate of the sorbing material) has been used to describe Cu sorption on amorphous Fe(OH)3 at pH 5.1 (Farley et al., 1985). The increase in Cu adsorption with increasing pH corresponds to a decrease in desorption as pH increases (Schultz et al., 1987; McBride, 1994; Temminghoff et al., 1994; Coughlin and Stone, 1995), probably in part because chemisorbed metals are more readily displaced by H+ ions than by other cations (McBride, 1989). However, even when acidic solutions are used for desorption, metal sorption becomes increasingly less reversible as the pH during adsorption increases (Padmanabham, 1983; Coughlin and Stone, 1995). Liming Oxisols to higher pH has been proved to be an efficient way to control Cu toxicity in coffee seedlings (Gimenez et al., 1992). The objective of this study is to test the hypothesis that the effect of pH on Cu adsorption and desorption will differ between P-treated, sludge-treated, and untreated Oxisols and between A and B horizons because the affinity for Cu of surface frmctional groups in each soil material will differ in pH-dependence. A previous study has shown that Cu adsorption at pH 5.5 was greater and the fractional Cu desorption was smaller in P- and sludge-treated than untreated Oxisols, and in A than B horizons (Chapter 2), but the effect of pH on Cu adsorption/desorption in P- and sludge-treated soils has not been studied previously. 66 Materials and Methods Soil Material Samples of A and B horizons from two uncultivated Oxisols, a Dark-Red Latosol and a Yellow-Red Latosol, were used in this study. Both soils have pH near 4.5 in the A horizon and 5.5 in the B horizon. Although both soils are classified as very fine, allitic, isothermic Typic Hapludox, the Yellow-Red Latosol has a greater kaolinitezgibbsite and goethite:hematite ratios than the Dark-Red Latosol. Soil properties are reported in Chapter 2; additional details concerning soil characterization have been described elsewhere (Lima and Anderson, 1997). Soils were pretreated with sludge or P as described in Chapter 2. Properties of the control, P-treated, and sludge-treated soils are reported in Appendix A2. Cu Sorption Replicate 0.45-g subsamples of each soil material were suspended in 30 mL of 5 mM Ca(NO3)2 and the pH was adjusted to 4.5, 5.5, and 6.5 ( I = 15 mM) with either saturated Ca(OH)2 or 7 mM HNO3. This pH range was selected because the natural pH in the A and B horizons of these soils respectively is 4.5 and 5.5; the soils typically are limed to a maximum pH of 6.5 when used for agriculture. All samples then were shaken for 24 h on a reciprocating shaker (120 cycles min"), and acid or base was added as needed to readjust pH to the desired value. Samples were shaken again for 24 h and pH was readjusted as necessary before the samples were shaken again for a third 24-h period. 67 During the third 24-h shaking period, suspension pH changed less than 0.1 pH units, so pH was considered to be stable. Copper adsorption capacity at pH 4.5, 5.5, and 6.5 was measured by repeatedly reacting the pH-adjusted soil suspensions with sufficient Cu(NO3)2 in a Ca(NO3)2 background electrolyte (I = 15 mM) to give an initial Cu concentration of 500 nM Cu. Because Cu adsorption caused a decrease in pH, especially at pH 6.5, suspension pH was readjusted to 4.5, 5.5, or 6.5 with saturated Ca(OH)2 after every two adsorption cycles, though little or no pH adjustment was needed for the pH 4.5 samples. Adsorption was considered complete when the increment in Cu adsorption was < 2% of the total Cu previously adsorbed, which corresponded to 6 reactions with 500 M, for pH 4.5 and 5.5, and 8 reactions with 500 M, for pH 6.5. However, all samples were reacted 9 times with 500 M Cu. This “repeated reaction” approach is preferable to a single reaction at a high Cu concentration because Cu concentrations >500 uM could cause Cu precipitation in solution either at pH 5.5 or 6.5 (Allison et al., 1990). Although Cu precipitation cannot be ruled out at pH 6.5, the fact that sorption reached a clearly defined plateau is evidence that nearly all sorbed Cu was adsorbed, not precipitated. To assess the effect of pH on Cu adsorption and desorption at lower initial concentrations, appropriate amounts of Cu(NO:,)2 in Ca(NO;.,)2 (pH 4.5, 5.5, and 6.5; I = 15 mM) were added to each pH-adjusted soil suspension to give initial total Cu concentrations of 0, 5, 50, and 150 W and a solutionzsoil ratio of 100:1. Triplicate untreated and P- or sludge treated samples were reacted with Cu for 72 h. After adsorption, the suspensions were centrifirged, the supernatant solutions were decanted and saved for Cu analysis, and the centrifuge tubes were weighed to determine the mass of 68 entrained solution. The soil pastes that remained in the centrifuge tubes were then reacted with 5 mM Ca(NO,)2 at pH 5.5. Copper desorption was measured afler 72 h. The concentration of Cu desorbed at each successive desorption reaction time was corrected for the amount of copper in the entrained solution. The 72-h reaction time was chosen because the amount of Cu adsorbed at 72 h was at least 80% of the total adsorbed after 54 d, and desorption was always at least 95% complete after 72 h (Chapter 3). Total solution- phase Cu concentration (supernatant [CuT]) was analyzed by either electrothermal or flame atomic absorption spectroscopy. Distribution coefficients for Cu adsorption (K) at each pH were calculated as Kd = {Cum/[CU] [1] where {Cum} is the adsorbed Cu concentration (mmol kg“), and [Cu] is the final solution- phase Cu concentration (mmol L") after the 72-h adsorption reaction. Distribution coefficients for Cu desorption were calculated according to equation [1], except that for desorption, {Cum} was calculated as the difference between the adsorbed Cu concentration at the end of adsorption and the concentration of Cu desorbed after a 72-h reaction with 5 mM Ca(NO3)2. Results Copper Adsorption Capacity Copper adsorption capacity increased with increasing pH (Table 4.1). As pH increased from 4.5 to 5.5 the Cu adsorption capacity increased 2.7- to 3.0-fold in A- 69 horizon samples, and was 3.3 to 3.9 times greater at 5.5 than 4.5 in B-horizon samples, with P-treated samples exhibiting the greatest increase. As pH increased from 5.5 to 6.5, Cu sorption capacity increased 2.8- to 36-fold in the A horizon, and 3.3 to 3.9 times in the B horizon, with the greatest increase for control samples. Thus, pH always had a greater effect in B than A horizons, but P-treated samples exhibited the greatest increase from pH 4.5 to 5.5, and control samples from pH 5.5 to 6.5. Table 4.1. Copper adsorption capacity (Cum m.) at pH 4.5, 5.5, and 6.5 of control and P- or sludge-treated soil samples. 0112...... PH 4.5 Cu¢.m.PH 5.5 CMMPH 6-5 Hor. Control P Sludge Control P Sludge Control P Sludge mmol kg’I WW 19:0 24:1 22+1 54:4 70:0 60:2 l92+5 193:2 168:1 B 12:0 15:0 14+O 44:0 50:0 47:2 173:6 167:4 160:3 WW 19:0 21:1 20+0 53:3 62 :2 54:1 184:0 180:6 171:12 B 12+0 14:1 12+0 47:3 49:1 43:3 167:3 183:10 160:4 Pretreatment with P caused Cu adsorption capacity to increase by as much as 25% compared with control soils at pH 4.5 and 30% at pH 5.5, but had little effect at pH 6.5. At pH 4.5 and 5.5, the Cu adsorption capacity of sludge-treated soils generally was intermediate between control and P-treated soils, whereas sludge-treated soils have the smallest Cu adsorption capacity at pH 6.5. 70 Copper Adsorption Adsorption isotherms at pH 4.5, 5.5, and 6.5, are shown in Figure 4.1. The shape of the isotherms (Giles et al., 1960) changed from low to high affinity as pH increases. Based on the shape of the adsorption isotherms, a high-affinity type isotherm is observed for all A-horizon samples as well as all P-treated samples at pH 6.5. Adsorption isotherms for untreated and pretreated samples of both the Dark-Red Latosol (hereafter Gme) and the Yellow-Red Latosol (hereafter Kth) were quite similar in shape, except for control and sludge-treated B-horizon samples at pH 6.5. For B-horizon samples, considerable adsorption occurred at pH values below the pHPZSE (Appendix A2). This can be better seen when the fraction of Cu adsorbed (Cum/CW is plotted as a function of pH (Figure 4.2). The higher the initial [CuT], the greater the increase in Gnu/Cum caused by an increase in pH from 4.5 to 6.5. For all initial [CuT], the increase in Cum/Cum upon increasing pH was always greater in the B than the A horizon. There is also a trend for Cum/Cuadded to increase more in untreated than pretreated samples, and in the Kth soil than the Gme soil, as pH increased from 4.5 to 6.5 (Figure 4.2). Specifically, for samples reacted with 5 uM Cu, increasing pH from 4.5 to 6.5 caused CuijuW in control samples of the Gme soil to increase about 1.2 times in the A horizon and 1.4 times in the B horizon. 71 pH 4.5 pH 5.5 pH 6.5 15 —: 0‘ Control 1 GmeA horizon , 1 -D- Sludge P , * I 10 j + Phosphate / ,1! y 1 / _- j ‘. / I 5 j " II I l ! J i O h " V . in I r r 14 . . . I I I ' r r . 15 - Gme Bhonzon '7 j / c: - TD 3 /// O /_/ E 5 : //,/' a. "' E 1 ' ,- I“. I .3 0 H '| % t % w t .e 15 i Kth A horizon _ ° 1 310 1 /DD : « U 5 ; //,,+C)H ./ .i 3 . ’I i . ’ . l. 0 . , '- i 1 1 - l 15 ~ Kth B honzon 1 pm /0 I // _. / 10 : ’0. '1 ..=}// 53 ‘ -*’ ‘ 1 0 "l“rT"'I"FTW"'l I 1"PIYP"~—"l'l l'l 0 25 50 751000 25 50 75 0 5 01520 25 Supernatant [CUT] (pM) Figure 4.1. Copper adsorption isotherms in control and P- or sludge-treated A and B horizon samples of two Oxisols reacted with Cu at pH 4.5, 5.5, and 6.5 (Gme = Dark- Red Latosol; Kth = Yellow-Red Latosol). 72 .2883 00M§e=0> u “OHM ”—883 033.009 n .539 08028.08 2:H 5 =0 21 on . 0:0 .3 .m 505 02000.. £805 23 .«0 003800 acute: m 0:0 < 0000000330 00 -m 0:0 30:00 E 1&0 500:3 0 8 000.5000 :0 we eozufim .Né 0&3 125.... 0.0 3. 0.0 0.... 3. 0.0 0.0 m... _ _ L hill . fl . 111. F 30:38. L - we ., .. . 1.. T v... i r I- - 0... i I a... a... 20:009.; I41 r to 000:5 I m I .2200 e 17‘ (peppe nomemosne no) names»: narrows f: canto: <.. . canto: m acute: < ll EIQG iill .00. 73 Compared with the Gme soil, the Kth soil reacted with 5 uM Cu exhibited a similar trend with respect to soil horizon, but CWCuW values were 0.03 to 0.06 units smaller for the Kth than the Gme soil at pH 4.5 (Figure 4.2). Pretreated samples showed a greater Cum/Cu,“ than control samples at pH 4.5, but adsorption was almost complete at either pH 5.5 or 6.5, so little differences could be seen between untreated and pretreated samples with respect to Cum/Cum at either pH 5.5 or 6.5. For samples reacted with 50 M Cu, increasing pH from 4.5 to 6.5 caused Cum/Cuw in control samples of the Gme soil to increase almost 1.5 times in the A horizon, and 2.2 times in the B horizon. Pretreated B-horizon samples showed a greater CumJCum than control B- horizon samples at pH 5.5, but little or no differences at either pH 4.5 or 6.5. Again the Gme and the Kth soils reacted with 50 M Cu exhibited a similar trend with respect to pretreatment and soil horizon, but Cum/Cu“Med values were 0.06 to 0.09 units smaller for the Kth soil than the Gme soil. For 150 |J.M Cu, increasing pH fiom 4.5 to 6.5 caused Cum/Cum in control samples of the Gme soil to increase about 2.3 times in the A horizon, and 2.8 times in the B horizon; for the Kth soil, Cuw/Cuw in control samples increased almost 2.5 times in the A horizon, and 3.2 times in the B horizon. Pretreatment generally had a greater effect on Cum/Cu“Med for samples reacted with 150 M Cu than for samples reacted with either 5 or 50 M Cu. The pH at a given fraction of adsorbed Cu (with no apparent saturation) increased with increasing [CuT]. This increase is greater in control than either P- or sludge-treated soils, and in B than A horizon. Another way of seen the effects of soil pretreatment is that, despite the [CuT], for a given fraction of adsorbed Cu, the pH tends to be higher for control than for either P- or sludge-treated samples. 74 Copper Desorption Copper desorption decreased remarkably as preequilibration pH increased. For a given pH, there was a trend of decreasing the fraction of Cu desorbed (Cum/Cum) as [CuT] decreased. There is also a trend of Cum/Cu“, being smaller in A than B horizon, and in the Gme soil than in the Kth soil, though little differences regarding both 0C content or mineralogy occurred at pH 6.5 (Figure 4.3). As [CuT] increases, much more Cu tends to desorb at either pH 5.5 or 6.5, so that increasing preequilibration pH from 4.5 to 6.5 caused the shape of the curves to change fi’om exponential, at low [CuT], to almost linear, at high [CuT]. Soil pretreatment generally decreased Cum/Cu“, and its effects were greater in B than A horizons. Except for B-horizon samples reacted with 150 M Cu, for any other sample Cum/Cu“, values at pH 6.5 were always smaller than 0.01, so that differences in Cum/Cu“, caused by CC content, mineralogy or soil pretreatment are difficult to be seen. For samples reacted with 5 uM Cu, Cum/Cu“, at pH 4.5 in control samples were about 2.3 times smaller in the A horizon than in the B horizon of the Gme soil. In the Kth soil, Cum/Cu“, in control samples was about 1.8 times smaller in the A horizon than in the B horizon. Pretreatment showed little effect in the Gme soil, but caused Cum/Cum, at pH 4.5 to decrease as much as 1.6 times in the Kth soil. For 50 11M Cu, Cum/Cum was about 1.4 times smaller in the A than in the B horizon of both soils at pH 4.5, and about 3 times smaller in the A than in the B horizon of both soils at pH 5.5. 75 .2893 03. -32....» u 5.0. 0893 80-0.30 u 8.30. 3 2a .3. .3. ma .0 €25.80...” 8.H c :0 x... 8. 2a .8 .m 5.3 0282 08.5 as. .6 83.5... 0802. 0 Es < 8.80-003... s -0 05 .808 :80 .3 03 $02.63.: w a. 8.08% :0 .0 802E .3. as»... 10.3.... 0.... 0.0 0.0 0.0 0.0 0.0 #1 #1 Y #1 Iclu I. l 1... 0.2320 I4! 1. w a... m ; 30:5 1.0: . . l .9200 o if H + 50.3.. m .90. E00... < . 03.3.. m 5:00 (Demospe homosep no) mecca uonous 03.3.. .< .. . 76 Pretreated B-horizon samples from both soils desorbed up to 3 times less Cu (on a fractional basis) than control B-horizon samples at pH 5.5, and sludge pretreatment affected Cum/Cum, somewhat less than did P treatment. Little or no differences in the Cum/Cum, values between untreated and pretreated samples were observed at either pH 4.5 or 6.5. For samples reacted with 150 W, soil pretreatment also decreased Cum/Cums at pH 5.5 (up to 2 times), but had little effect at either pH 4.5 or 6.5. Differences between soil horizons with respect to Cum/Cum, were smaller for samples reacted with 150 M Cu than for samples reacted with either 5 or 50 M Cu. As for Cum/Cum (Figure 4.2), pretreatment generally had a greater effect on Cum/Cu“, for samples reacted with 150 M Cu than for samples reacted with either 5 or 50 M Cu (Figure 4.3). Distribution Coefi‘icients The effects of pH on the adsorption and desorption distribution coefficient of Cu (K,) in control and P- or sludge-treated samples reacted with 150 M are presented in (Figure 4.4). For adsorption, increasing pH fiom 4.5 to 5.5 caused K, to increase up to 3.4 times in control, up to 3.9 times in sludge-treated, and up to 6.2 times in P-treated samples. For desorption, the same increase in pH caused K, to increase up to 2.2 times in control, up to 2.9 times in sludge-treated, and up to 3.1 times in P-treated samples. For both adsorption and desorption K,, the pH effect was greater in the A than B horizon. Adsorption K, at pH 6.5 were at least one order of magnitude greater in the B horizon and two orders of magnitude greater in the A horizon than adsorption K, at pH 4.5. The same trend regarding soil horizon was observed for desorption K,, but again the effects of pH 77 were smaller for desorption K, than for adsorption K,. Except for the B horizon of the Kth soil, increasing pH up to 6.5 had a similar effect in adsorption and desorption K, for both soils. Desorption Adsorption .79. ._. av. mo. 3 2 . _ 220000 4 3 2 /////V/V//////////// .\\\\\\\\\\\\.\\. ccccccccccccc 00000000 ..o. ooooo %%%n( ma ...... Sludge Phosphate Control Sludge Phosphate 1000 g 100 .wmx.: v. 10000 g Control Figure 4.4. Adsorption and desorption distribution coefficients of Cu in control and P- or sludge-treated A and B horizon samples of two Oxisols reacted with 150 W Cu (1:100 = Dark-Red Latosol; Kth = Yellow-Red soil:solution) at pH 4.5, 5.5, and 6.5 (Gme Latosol) 78 Desorption K, (Figure 4.4, right) was always greater than adsorption K, (Figure 4.4, left). This difference decreased with increasing preequilibration pH, and is smaller for A compared with B horizons. At pH 6.5, little difference was observed between adsorption and desorption K, for A-horizon samples (K, adsorption/K, desorption ~ 1.0). Discussion A change in the shape of the isotherms from low to high affinity (Figure 4.1) suggests a change in site affinity as pH increases, which may be associated with an increase in the Coulombic energy of attraction (Stumm and Morgan, 1996). Isotherrns of the L-type have been associated with chemisorption (McBride, 1994), and are explained by a high affinity of the adsorbent for the adsorbate at low concentration, which then decreases as concentration increases (Sparks, 1995). An H-type isotherm indicates a very strong adsorbate-adsorbent interaction such as inner-sphere complexes (i. e. , chemisorption), and is an extreme case of the L-type isotherm (McBride, 1994; Sparks, 1995). At pH 4.5, there is possibly a limited number of sites with high affinity for Cu, because the Coulombic energy of attraction is very small (solution pH 5 pHpZSE). As pH increases to 5.5, the number of sites with great affinity for Cu increases, but are still limited, since the slope of the isotherm decreases markedly at high concentrations. This increase is greater in the A horizon compared with B horizon, and in P- and sludge-treated samples compared with control samples. Such observations are reported in a previous study, and have been attributed to the fact that P and sludge treatment might convert M”- OH groups into M3”-phosphate and M3*-organo functional groups, thereby modifying the 79 reactivity of the site toward Cu but not increasing the number of surface M3+ sites, which exert overall control on the Cu adsorption capacity (Chapter 2). A further increase in pH from 5.5 to 6.5 results in a remarkable increase in the number of sites with very great affinity for Cu. A greater increase occurred in the A horizon compared with the B horizon , and in P-treated compared with either sludge-treated or control samples. For P-treated samples, this occur despite of the fact that at high pH, the affinity of the surface for the pre-adsorbed phosphate tends to decrease, particularly in the A horizon. The increase in Cu adsorption as pH increases (Figure 4.2) has been reported by several authors for either soils or pure sorbents. A greater adsorption for surface horizons compared with subsurface horizons has also been reported (Carey et al., 1996), and could be attributed to the very strong interaction between Cu and soil organic matter (McBride, 1989). This and the different behavior of control samples compared with either P- or sludge-treated samples (especially at high [CuT]) suggests that the charge of the surface is of great importance for Cu adsorption. Part of the increase in Cu adsorption with increasing pH has been attributed to an increase in the concentration of CuOH+ species as pH increases (Barrow et al., 1981). However, at pH 6.5, most of the Cu species in solution would still be present as Cu2+ (Table 4.2). These observations all questioned the hypothesis that the identity of the adsorbing ion is much more important for adsorption equilibria than is the surface charge (Spark et al., 1995a), at least within the pH range of this study. 80 030.000 >=00_E0§00.E0.0 8.. N0.0.5.0050 .8 0805.005 .05 00.0.5 88.... 2.8 2.. 8 8.0.8.... n.0508; 5 =0 8.. .888 .2. 8 8.8.... 5 . 03.800 3.028053%... 0. «$03080000 000 .000 000000 0000 0.0 .00 0000800.. 0000000 00 :0 00.. 05.0000 .0: 00 00.00% 00 . .600. .800 0000000003 0000800.. 0.0—0800 20000.»... 0.0880 0. 30. 00. 0.03 0000000.. 00080000 00.00 0.0 $0.50 000000000000 DOG 0000000 .000. 00 00 0000—00. .00 003 009.00 00.030 0030005 . 8... S... - 8... a... - 8... - - g. .005 R... 8... - mm... 8... - 8.... 8... - .6005 8.0 8... - 8... S... - 0.... - - $8005 .0... 0.... - 8.... 2... - 8.... 0.... - 3 N0.0.5 R. 8... - 3... 8... - 5.. a... - .005 8.... 8.8 8.8. 8.8 8.8 8.8. 8.8 8.8 8.8. N.5 8 .3 0.0 m... .00 3 m... m... 0.0 m... 8.8..m 5 8.. 8." 05. x... 8 u 05. x... n u 05. .. 800. J0 .0 0000—3 gawk/:2 .3 00.0.0.5... 00 1:0. 00000000000 5 8.8-88.8 .08. 0...... .880... .80 N00.05 2... m 0. .2 n N00... 0.0 08 .00 .0... =0 .0 8.8.8... 885 .0... 8.80 81 Even though surface charge might play an important role in Cu adsorption, it is noteworthy to mention that a considerable fraction of Cu in B-horizon, control samples was adsorbed at pH values below the pHPZSE (Appendix A2) (especially at low [CuT]), which is a typical behavior of chemisorbed cations (Padmanabham, 1983). The shift in the adsorption curves toward a more alkaline region as [CuT] increases (Figure 4.2) has been reported elsewhere (Benjamin and Leckie, 1981; Farley et al., 1985). This arises from the effect of a high degree of occupancy of adsorption sites in lowering the tendency of further Cu chemisorption (McBride, 1994). However, such effects appear to be of less importance in P- and sludge-treated soils, possibly because of a greater affinity of either P- or sludge-treated soils for Cu. This is consistent with the fact that at a given fraction of adsorbed Cu and despite the [CuT], the pH tends to be higher for control than for either P- or sludge-treated samples, a behavior that has been reported also for trace metal adsorption in combined goethite- and silica-humic acid systems (Spark et al., 1997b). The smaller fractional desorption as preequilibration (adsorption) pH increases (Figure 4.3) has been reported elsewhere (Padmanabham, 1983; Coughlin and Stone, 1995) and may arise from the fact that a monodentate complexation reaction should give way to a bidentate reaction at higher pH, which requires a large activation energy for the desorption (McBride, 1994). A small fractional desorption of Cu in either soils or pure sorbents as observed in this study has been reported also by McLaren and Crawford (1974) and McLaren et al. (1983). According to McLaren et al. (1983), this suggests that some reactions involved in the adsorption processes are irreversible, very slowly reversible, or require a high activation energy for desorption, or involve a combination of 82 all three. A high activation energy for desorption has been hypothesized to explain a smaller fractional Cu desorption at pH 5.5 in P- or sludge-treated samples of Oxisols compared with untreated ones (Chapter 2). Furthermore, the fact that a considerable fraction of Cu in Oxisols remains adsorbed even if a short adsorption time (e.g., l h) is followed by a long desorption period (e.g., 54 (1) (Chapter 3) indicates that a somehow irreversible, or very slowly reversible reaction, occurs at the very beginning of Cu adsorption. The large increase in the adsorption K,l upon increasing pH (Figure 4.4) has been reported elsewhere (Basta and Tabatabai, 1992) and is consistent with the hypothesis that site affinity increases as pH increases (isotherms change from L-type to H-type, Figure 4.1). Yet, the great difference between desorption and adsorption Kd at pH 4.5 (markedly in B-horizon samples) suggests that sites with great affinity for Cu exist even at pH values below the pHpZSE , i. e., the “intrinsic” term, as described by Stumm and Morgan (1996), is an important component of the fi'ee energy of adsorption of Cu in Oxisols, and might be responsible for the small fi'actional Cu desorption (Cu adsorbed/Cu desorbed < 0.45) observed even at pH 4.5. The greater increase in adsorption Kd for A compared with B horizon, and for P- and sludge-treated soils compared with control soils (at pH 5.5) is also consistent with the hypothesis that, for the pH range and the soil samples used in this study, surface charge is much more important for adsorption equilibria than is ion speciation. In conclusion, this study suggests that in the pH range 4.5 to 6.5, Cu availability in Oxisols is likely to be controlled by the surface chemistry rather than Cu speciation in solution. Increasing preequilibration pH from 4.5 to 6.5 caused Cu adsorption to go 83 almost to completion and desorption to decrease markedly. However, the fact that a considerable fraction of Cu still remained adsorbed at pH values below the pHp‘ZSE is noteworthy as this will reduce greatly the availability of Cu in Oxisols. 84 LIST OF REFERENCES Allison, J.D., D.S. Brown, and K.J. Novo-Gradac. 1990. MINTEQA2/PRODEFA2, a geochemical assessment model for environmental systems: Version 3.00 user’s manual. EPA-600/3-9l-021. USEPA, Athens. Barrow, NJ. 1985. Reactions of anions and cations with variable-charge soils. Adv. Agronomy 38:183-230. Barrow, NJ. 1989.The reaction of plant nutrients and pollutants with soil. Aust. J. Soil Res. 27:475-492. Barrow, N.J., J .W. Bowden, A.M. Posner, and J .P. Quirk. 1981. Describing the adsorption of copper, zinc and lead on a variable-charge mineral surface. Aust. J. Soil Res. 19:309-321. Basta, NT, and M.A. Tabatabai. 1992. Effect of cropping systems on adsorption of metals by soils: II. Effect of pH. Soil Sci. 153:195-204. Benjamin, M.M., and J .O. Leckie. 1981. Multiple-site adsorption of Cd, Cu, Zn, and Pb on amorphous iron oxyhydroxide. J. Colloid Interface Sci. 79:209-221. Carey, P.L., R.G. McLaren, and J .A. Adams. 1996. Sorption of cupric, dichromate and arsenate ions in some New Zealand soils. Water, Air, and Soil Pollut. 87: 1 89-203. Coughlin, BR, and AT Stone. 1995. Nonreversible adsorption of divalent metal ions . (Mn", Co", Ni", Cu", and Pb") onto goethite: effects of acidification, F eII addition, and picolinic acid addition. Environ. Sci. Technol. 29:2445-2455. Ellis, B.G. and B.D. Knezek. 1972. Adsorption reactions of micronutrients in soils. In J .J . Mortvedt et al. (ed.) Micronutrients in Agriculture. p. 59-78. SSSA, Madison. Farley, K.J., D.A. Dzombak, and F.M.M. Morel 1985. A surface precipitation model for the sorption of cations on metal oxides. J. Colloid Interface Sci. 106:226-242. Forbes, E.A., A.M. Posner, and LP. Quirk. 1976. The specific adsorption of divalent Cd, Co, Cu, Pb and Zn on goethite. J. Soil Sci. 27 :154-166. Giles, C.H., T.H. MacEwan, S.N. Nakhwa, and D. Smith. 1960. Studies in adsorption: Part XI. A system classification of solution adsorption isotherms, and its use in 85 diagnosis of adsorption mechanisms and in measurement of specific surface area of solids. J. Chem. Soc. p. 3:3973-3993. Gimenez, S.M.N., J .C.D. Chaves, M.A. Pavan, and 1.1. Cruces. 1992. Toxicidade de cobre em mudas de cafeeiro. R. bras. Ci. Solo 16:361-366. Harter, RD. 1983. Effect of soil pH on adsorption of lead, copper, zinc, and nickel. Soil Sci. Soc. Am. J. 47:47-51. Hsu, PH. 1989. Aluminum hydroxides and oxyhydroxides. p. 331-378. In Dixon, J.B., and SB. Weed. (ed.) Minerals in soil environments. 2 ed. Soil Science Society of America, Madison. James, R.O., and NJ. Barrow. 1981. Copper reactions with inorganic components of soils including uptake by oxide and silicate minerals. In J .F. Loneragan, A.D. Robson, and RD. Graham (ed.) Copper in Soils and Plants. p. 47-68. Academic Press, Sydney. James, R.O., and T.W. Healy. 1972. Adsorption of hydrolyzable metal ions at the oxide- water interface. III. A thermodynamic model of adsorption. J. Colloid Interface Sci. 40:65-81. - Kabata-Pendias, A. and H. Pendias. 1992. Trace elements in soils and plants. 2 ed. Boca Raton, CRC Press. Kretzschmar, R., D. Hesterberg, and H. Sticher. 1997. Effects of adsorbed humic acid on the surface charge and flocculation of kaolinite. Soil Sci. Soc. Am. J. 61: 101-108. Lima, J .M., and SJ. Anderson. 1997. Effect of aggregation and aggregate size on extractable Fe and A1 in two Brazilian Typic Hapludoxs. Soil Sci. Soc. Am. J. 61 :965-970. Lindsay, W.L. 1979. Chemical equilibria in soils. Wiley-Interscience, New York. McBride, MB. 1982. Cu2+ adsorption characteristics of aluminum hydroxides and oxyhydroxides. Clays Clay Miner. 30:21-28. McBride, MB. 1989. Reactions controlling heavy metal solubility in soils. Adv. Soil Sci. 10: 1-56. McBride, MB. 1994. Environmental chemistry of soils. New York, Oxford University Press, Inc. 86 McKenzie, R. M. 1980. The adsorption of lead and other heavy metals on oxides of manganese and iron. 1980. Aust. J. Soil Res. 18:61-73. McLaren, R.G., and D.V. Crawford. 1974. Studies on soil copper: III. Isotopically exchangeable copper in soils. J. Soil Sci. 25:111-119. McLaren, R.G., R.S. Swift, and J .G. Williams. 1981. The adsorption of copper by soil materials at low equilibrium solution concentrations. J. Soil Sci. 32:247-256. McLaren, R.G., J .G. Williams, and RS. Swift. 1983. Some observations on the desorption and distribution behavior of copper with soil components. J. Soil Sci. 34:325-331. Padmanabham, M. 1983. Adsorptiomdesorption behavior of copper (II) at the goethite- solution interface. Aust. J. Soil Res. 21: 309-320. Piccolo, A., and F.J. Stevenson. 1982. Infrared spectra of Cu”, Pb“, and Ca2+ complexes of soil humic substances. Geoderma, 27: 195-208. Ritchie, G.S.P. and SC. Jarvis. 1986. Effects of inorganic speciation on the interpretation of copper adsorption by soils. J. Soil Sci. 37:205-210. Schindler, P.W., P. Liechti, and J .C. Westall. 1987. Adsorption of copper, cadmium and lead from aqueous solution at the kaolinite/water interface. Netherlands J. of Agric. Sci. 35:219-230. Schultz, M.F., M.M. Benjamin, and J .F. Ferguson. 1987. Adsorption and desorption of metals on ferrihydrite: reversibility of the reaction and sorption properties of the regenerated solid. Environ. Sci. Technol. 21 :863-869. Spark, K.M., J .D. Wells, and BB. Johnson. 1995a Characterizing heavy-metal adsorption on oxides and oxyhydroxides. Europ. J. Soil Sci.46:621-631. Spark, K.M., J .D. Wells, and BB. Johnson. 1995b. Characterizing trace metal adsorption on kaolinite. Europ. J. Soil Sci.46:633-640. Spark, K.M., J .D. Wells, and BB. Johnson. 1997a. Characteristics of sorption of humic acid by soil minerals. Aust. J. Soil Res. 35: 103-1 12. Spark, K.M., J .D. Wells, and BB. Johnson. 1997b. Sorption of heavy metals by mineral- humic acid substrates. Aust. J. Soil Res. 35:113-122. Sparks, D.L. 1995. Environmental soil chemistry. San Diego, Academic Press, Inc. Sposito, G. 1989. Surface reactions in natural aqueous colloidal systems. Chimia 43:169- 176. 87 Stumm, W. 1992. Chemistry of the solid-water interface. John Wiley and Sons, New York. ‘ Stumm, W., and J .J . Morgan. 1996. Aquatic chemistry; chemical equilibria and rates in natural waters. John Wiley and Sons, New York. Temminghoff, E.J.M., S.E.A.T.M. Van der Zee, and F.A.M. De Haan. 1997. Copper mobility in a copper-contaminated sandy soil as affected by pH and solid and dissolved organic matter. Environ. Sci. Technol. 31 :1 109-1 115. Wang, F., and J. Chen. 1997 . Modeling sorption of trace metals on natural sediments by surface complexation model. Environ. Sci. Technol. 31:448-453. Wang, Z., and W. Stumm. 1987 . Heavy metal complexation by surfaces and humic acids: a brief discourse on assessment by acidimetric titration. Netherlands J. of Agric. Sci. 35:231-239. APPENDICES 89 Appendix A1 Sludge characterization’ Constituent g kg" dry sludge A1 5.88i0.70 B 0.03:1:0.01 Ca 153.372t18.57 Cu 0.24zl:0.01 Fe 14.05i1.61 K 190212.88 Mg 8.92:1.18 Mn 02210.02 Mo 0.05;l:0.01 Na 4.62:l:0.59 P 31.77:l:3.96 Zn 0.48i0.07 Organic Ci 275.751828 Total C§ 313.40i2.53 TElemental analysis done by ICP after extraction of oven-dried (550 °C for 8 h), < 0.5-mm sludge with 6 M HNO,. *Organic C measured by the Walkley-Black method. § Total C determined by hi gh-temperature combustion. 9O m3 m2 no 33:. Show «.33 :2 Sign :3 33¢ 33.” m 84. 3.... o? 25.3 33mm 33% $8 afiommm E: 23.3 33.8 < $48333 Se 3m 26 <32 $32 SHE 3: £8 ohm 33.: S“? m “2. 8+ 5. 33% 23.3 9323 7...: $3: on: Saga 33.3 < 3 lemo— _o~=1l lemu— 381ll| like—all omega .2 .5 $35 m :6 032m m :6 ”mesa m as .5 com mafia so 2325£€§a saga oo ..m_8_x0 5535 9.3 .3 838.8 coutonfi 98 ..< 3:35.032.» .8 -m 98 £va .9980 Mo 89303 30386 2 5:59? 91 Appendix A3 Titration curves for the determination of the Pszse of control, P-treated, and sludge- treated A- and B-horizon samples of two Brazilian Oxisols. 60 Dark-Rod LIIOOOI LA W1 o 180.015molLL'. A l=0.15molL" I I Ll III II I ”9‘9‘9l96'b9le9169l«9«9 9959].. 9l Yellow-Rod Law (A hOf_IZ_On H‘IOH'adsorbod, mmol kg" J I l I I I I I I I I I I I I I I I I l I I I I l I I 50 I I I I I I '59 59 59 6:9 b9 o9 69 «9H«9 '59 59 59 b9 b9 e9 o9 «9 «9 '59 59 .9 <99 6:9 o9 o9 «9 «9 7101120" 60 Yellow-Rod Latosol Control 4. Sludge 40- 20— III‘IIII T ,59 ‘9 59199 (D9 99 H99 «9 H,\9 '59 pH —-I— 92 fivw Eng 35w 32 32 Enom 25.9326 .2933. ~ .o 39ng _ .o Euwm NHNO Nflum m ofivw NHmw Nfimw OHS NHE Nfim fl :5ng w .o anHmm — .o c 5.633 .o 36m mfimm mHVm < 34% 3mm 13% 3.13. 32 WE VHHN woodfim: .o 30.on3 .0 «8.3536 35m fine 38 m flow mfivw 3mm 3: mHE E”: 293356 bmodfimmfio 55.3”ch ~30 $60 mung < g .x. _.w fl:8 l|l|_.w NEIIIII omega m :6 0955 m :6 032m m :6 ”mesa m =6 Sm 56989802 gap—30.83 3680a 30,—.