Reduction of silver solubility by humic acid and thiol ligands during acanthite (β-Ag 2S) dissolution

Reduction of silver solubility by humic acid and thiol ligands during acanthite (β-Ag 2S) dissolution

Please download to get full document.

View again

of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.

Health & Fitness

Publish on:

Views: 6 | Pages: 9

Extension: PDF | Download: 0

  Reduction of silver solubility by humic acid and thiolligands during acanthite ( b -Ag 2 S) dissolution Astrid R. Jacobson a, ) , Carmen E. Martı ´nez a,1 , Matteo Spagnuolo b ,Murray B. McBride a , Philippe Baveye a a Department of Crop and Soil Science, Cornell University, 1002 Bradfield Hall, Ithaca, NY 14853, USA b Dipartimento di Biologia e Chimica Agroforestale ed Ambientale, Universita ` degli Studi di Bari, via Amendola 165/a, 70126 Bari, Italy Received 28 May 2004; accepted 19 October 2004 The presence of humic acids and thiol ligands reduced Ag dissolution from acanthite particleswith mixed-oxidation surface layers. Abstract Precipitation of highly insoluble metal sulfide minerals like acanthite ( b -Ag 2 S) or red cinnabar (HgS) is in principle an effectivemeans to reduce metal availability and toxicity in contaminated soils. Unfortunately, experiments have shown that red cinnabar maybe solubilized in the presence of dissolved organic matter or thiol ligands. To determine whether the same applies to acanthite,a laboratory synthesized  b -Ag 2 S mineral was incubated for up to 3 weeks in the presence of KNO 3 , dissolved humic acids, cysteine,methionine and thiosulfate. XPS analysis identified Ag 2 O (52%), Ag 2 SO 4  (8%) and Ag 2 S (40%) on the particle surfaces. Ag wasreleased into solution in the presence of KNO 3  and methionine, presumably from mixed-oxidation surface layers. Contrary to earlierresults with cinnabar, however, humic acids reduced Ag concentrations in solution by about 75%, and cysteine and thiosulfate, eachcontaining a free –SH functional group, almost completely suppressed Ag release into solution.   2004 Elsevier Ltd. All rights reserved. Keywords:  Acanthite; Cysteine; Humic acids; Methionine; Silver sulfide; Thiols; Thiosulfate 1. Introduction In reducing environments, metals react with dissolvedsulfides to form stable metal sulfide species that tend tobe insoluble. Among the most insoluble metal sulfidesthat form are  a -HgS (red cinnabar, log  K  0red Z  52.03)and  b -Ag 2 S (acanthite, log  K  0 Z  48.97) (Lindsay,1979). As a result, high concentrations of metals areoften found in sulfide-rich coastal sediments (Shaferet al., 1998; Sundelin and Eriksson, 2001). Variousremediation schemes try to put this process to good use,e.g., by developing wetlands to treat acid-mine drainageor by using sulfide-rich sediments to sequester metalsfrom the water column or soil solution at metal-contaminated sites (Peltier et al., 2003).Studies with acanthite have found that it formsquickly under natural conditions, either by the pre-cipitation of Ag and sulfide, or through the substitutionof Ag for Fe in iron sulfide (Adams and Kramer, 1998). A number of NMR investigations described in a reviewby Bell and Kramer (1999) suggest that substitutionreactions by thiolate groups on silver are rapid at ) Corresponding author. Tel.: C 1 607 255 3156; fax: C 1 607 2558615. E-mail address: (A.R. Jacobson). 1 Present address: Department of Crop and Soil Sciences,Pennsylvania State University, 116 ASI Building, University Park,PA 16802, USA.0269-7491/$ - see front matter    2004 Elsevier Ltd. All rights reserved.doi:10.1016/j.envpol.2004.10.017Environmental Pollution 135 (2005) 1–  ambient temperatures resulting in Ag(I) being readilyexchanged between thiolate aggregates and explainingwhy in the presence of HS  , soluble silver thiolates areso easily converted to Ag 2 S and the free mercaptan. Thereverse process, in which silver moves from inorganic toorganic complexes is thought to be limited because of the extraordinary insolubility and stability of Ag 2 S (Belland Kramer, 1999). Therefore, although Ag(I) may alsobe photoreduced to Ag 0 or precipitate as a halide salt,Ag 2 S is thought to be the ultimate sink of Ag(I), insulfur-rich environments such as reduced soils andsediments (Bell and Kramer, 1999).Unfortunately, the perception of sulfides, like cinna-bar and acanthite, as the ultimate sink for soil- orsediment-borne metals may not be appropriate in thepresence of specific organic ligands or ligands in humicacids. Ravichandran et al. (1998) observed that disso-lution of the highly insoluble mercury sulfide, cinnabar,was enhanced in the presence of organic matter isolatedfrom the Florida everglades and the sulfur-containingligands cysteine and mercaptoacetic acid. Since acan-thite is more soluble than cinnabar (Lindsay, 1979), andthe Ag(I) ion is softer than the Hg(II) ion, it isconceivable that acanthite dissolution will also beenhanced by ligands associated with dissolved organicmatter or sulfur-containing ligands. In the case of silver,such remobilization has been demonstrated by Adamsand Kramer (1998) who observed the re-solubilizationof Ag, by the addition of 3-mercaptopropanoic acid,from FeS particles to which it had previously beensorbed. Appreciable concentrations of these reducedsulfur species have been shown to exist even in oxicwaters. Thus, Ag(I) complexation with such speciescould represent a mechanism for its mobilization fromsediments to overlying waters resulting in a low, butconstant turnover of Ag at the sediment/water interface(Bell and Kramer, 1999; Benoit and Rozan, 1997;Cowan et al., 1985; Rozan et al., 2000).Formation of complexes by Ag C with variousorganic ligands may affect not just the stability of acanthite, but is also believed to reduce the bioavail-ability and/or toxicity of Ag C to a variety of organisms.However, although Ag C complexes readily with manyfunctional groups and ligands present, such as amines,carboxylates, and especially thiols (almost exclusivelywhen they are present), work by Fortin and Campbell(2001) shows that some silver complexes may also beable to cross algal membranes. Similarly, Bell andKramer (1999) suggest that when Ag–thiolate aggre-gates or complexes come into contact with a cell orcell-receptor containing free mercaptans, silver couldtransfer from the aqueous-phase to the cellular mercap-tan. Whereas complexation may not necessarily protectaquatic organisms from the uptake of Ag, the accumu-lation of complexed Ag does not necessarily result ina toxic effect.In this general context, the purpose of the presentstudy was to investigate the effect of sulfur-containingligands on the dissolution of acanthite. The sulfurligands we investigated included cysteine, methionine,thiosulfate, and a humic acid isolated from an oak-forest soil. Cysteine is found in aquatic environmentsand has a free thiol group (–SH) that is known tocomplex Ag C strongly. Its formation constant based on1:1 complex stoichiometry is p K  f  Z 11.9 (Bell andKramer, 1999). The sulfur in methionine is bound to 2methyl groups and thus is not able to complex Ag C asstrongly (log  K  25  C,0.1 Z 3.17; Sillen and Martell, 1964).The humic acids contain 2.9% S most likely ina combination of free and bound thiols and moreoxidized S functional groups. Thiosulfate is an inorganiccompound containing a free thiol group that formsstrong hydrophilic complexes with silver (log  K  Z 8.82;Sillen and Martell, 1964). Not only is thiosulfate used bythe photographic industry and therefore present in itswaste (Smith and Carson, 1977), but Ag–thiosulfatecomplexes have the potential to cross green algalmembranes via an anion transporter (Fortin andCampbell, 2001). 2. Materials and methods 2.1. Acanthite preparation Silver sulfide was prepared by mixing equal volumesof 3.0 M silver nitrate (AgNO 3 ) and 1.7 M sodiumsulfide (Na 2 S  9H 2 O) while stirring vigorously. Theresulting suspension was incubated in the dark for 3days. The mix was then centrifuged at 30090 !  g  for10 min and the supernatant decanted. The Ag 2 S pre-cipitate was then washed by filling the centrifuge tubeswith deionized distilled-water (16.7 M U cm at 25   C;dd-water), shaking overnight (in the dark) on a trans-lational shaker, centrifuging, and decanting the super-natant liquid. The washing procedure was repeated threetimes. Finally, the solid was freeze-dried, pulverized inan agate mortar and pestle, and stored under N 2(g)  in thefreezer. The surface area of the precipitate was de-termined by BET to be 3.9 m 2 g  1 , suggesting a relativelycrystalline solid. The Ag 2 S was further characterized byX-ray diffraction (XRD) and identified as acanthite toa detection limit of 1%. The periphery of the solid(5 nm) was investigated further using X-ray photoelec-tron spectroscopy (XPS). The XPS measurements wereperformed on a Leybold LHS10 spectrometer equippedwith an unmonochromatized Al K a source and a SPECSmulti-channel electron analyzer. Survey spectra wereobtained in the fixed retarding ratio mode (retardingratio Z 30); high-resolution spectra were recorded in thefixed analyzer transmission mode, with a pass energy of 30 eV. Calibration of the binding energy scale was 2  A.R. Jacobson et al. / Environmental Pollution 135 (2005) 1–9  performed by using the alkyl component of the C 1sphotoelectron peak (binding energy Z 284.8 eV) as in-ternal reference. 2.2. Humic acid preparation and characterization The humic acids (HA) used in this research wereselected not because of any specific property, but simplybecause they were characterized previously in otherstudies carried out in our laboratory. The HA wereextracted from a dark brown (10YR2/3), calcareous soilfrom Foresta Umbra, Italy, covered with  Quercus cerris .The soil was extracted with 0.5 M NaOH–0.1 MNa 4 P 2 O 7  under a N 2  atmosphere for 24 h with vigorousshaking (100 g of soil per L of extracting solution) toproduce the first humic acid extract. The HA suspensionwas centrifuged at 6000 !  g  for 20 min and the sedi-ments extracted again, overnight. The supernatantsolutions from the two HA extractions were pooled,and then precipitated overnight by acidification to pH 2with 6 M HCl (Schnitzer, 1978; Schnitzer and Khan,1972). After centrifugation at 8000 !  g  for 20 min,pellets were re-suspended with 0.1 M KOH undernitrogen, re-precipitated with HCl, and centrifuged asdescribed above. The resulting solid was then treatedwith 0.1 M HCl–0.1 M HF to reduce the ash content(Piccolo, 1988), dialyzed against deionized water withhourly changes, and freeze-dried. The elemental com-position was determined with an EA 1108 CHNSAnalyzer (Fisons Instruments, Lucino di Rodano,Italy). The degree of polymerization and aromaticitywas evaluated by the ratio of the absorbance at 464 and665 nm (E 4 /E 6 ) (Chen et al., 1977). The HA was also analyzed by fluorescence and FTIR spectroscopy tocharacterize its functional groups (Stevenson, 1982). Elemental analysis, water content, ash content and E 4 /E 6 ratio are shown in Table 1. 2.3. Solution preparation One set of ligand solutions was prepared to containequimolar concentrations of total sulfur: 0.5 mMcysteine; 0.5 mM methionine; 0.25 mM sodium thiosul-fate; and 0.5528 g L  1 HA (0.5 mM S). A second set of ligand solutions with varying concentrations was pre-pared to test the idea that tertiary complexes might beforming: 0.05 and 0.1 mM cysteine; 0.05 and 5 mMmethionine; 0.025 and 0.05 mM sodium thiosulfate;and 55.2 and 110.3 mg L  1 Foresta Umbra humic acid(0.01 and 0.05 mM S). All the ligand solutions orsuspensions were prepared in 1 mM potassium nitrate(KNO 3 ) as a background electrolyte and then broughtto pH 6 with dilute potassium hydroxide (KOH). Theeffects of the background electrolyte on acanthitedissolution were tested by using different concentrationsof KNO 3  (0.1, 1 and 10 mM), and different electrolytes(NaNO 3 , NaClO 4 , Ca(NO 3 ) 2 , and dd-H 2 O) at 1 mMconcentrations. The solutions were adjusted to pH 6with dilute solutions of NaOH, KOH or Ca(OH) 2 matching the cation of the base to that of the electrolyte.All the reagents used were analytical reagent or trace-metal grade. 2.4. Dissolution experiments To test the effect of model S ligands on acanthitedissolution with time, 1 L of ligand solution was addedto a 1 L polyethylene bottle containing 2 g acanthite and20 (6-mm) glass beads, and capped under a N 2 (g)atmosphere. The samples were shaken in an orbitalshaker at 175 rpm and 24 G 1   C. At 1 (24 h), 2, 4, 7, 14,and 21 days the solutions were sampled by removinga 100 mL volume from each bottle while vigorouslystirring to keep the solid suspended. The suspensionswere then filtered through 0.2  m m cellulosic microsepfilters and refrigerated. The HA samples had to becentrifuged at 30900 !  g  for 10 min so that the solutioncould be filtered.Samples for all the other studies were set up in 60 mLpolyethylene bottles containing 50 mg acanthite, 10(6-mm) glass beads and 25 mL solution. The solid tosolution ratio (2 gL  1 ) was the same as for the time-study. Based on the results of the time-study, thedissolution of the acanthite samples containing differentconcentrations of S ligands was conducted for 7 days atwhich time the Ag concentration in the solutionsreached steady-state. The samples were then filteredand stored as described above. Since the objective of thestudies with the electrolyte solutions was to see whetheror not acanthite was dissolved and because Ag concen-trations reached near steady-state within 24 h in 1 mMKNO 3 , these studies were conducted for only 24 h. Inaddition, a few samples of acanthite in 1 mM KNO 3 were left open to the atmosphere and a few others weresampled at short time periods of 5, 30, 60 and 240 min todetermine how fast the dissolution occurred. All theexperiments were conducted with duplicate samples. 2.5. Sample analyses Total silver in solution was analyzed by FAAS or byaxially-viewed, inductively coupled argon plasma opti-cal emission spectroscopy (ICP-OES: SPECTRO Ana-lytical Instruments – SPECTRO CIROS CCD ) fitted witha short depth-of-field lens transfer optic (Rutzke, 1999, Table 1Selected characteristics of Foresta Umbra humic acidsElemental analysis Watercontent (%)Ashcontent (%)E 4 /E 6 %C %H %N %S %O50.3 4.8 3.7 2.9 38.2 8.4 1.2 5.73 A.R. Jacobson et al. / Environmental Pollution 135 (2005) 1–9  2000). The detection limits for Ag were 0.2 m M byFAAS and 10nM with the modified ICP-OES. Dis-solved organic carbon (DOC) was determined in theForesta Umbra humic acid samples by persulfateoxidation/CO 2  analysis (OI Analytical Model 1010Total Carbon Analyzer). These values were used todetermine how much of the humic acid was in solutionthroughout the course of the experiment. Total sulfatewas determined in selected samples by the turbidimetricmethod (Clesceri et al., 1998) using a spectrophotometer(Perkin–Elmer Hitachi 200). However, to prevent AgCl 2 precipitation from interfering with the sulfate detection,the buffer solutions were modified to contain Mg(NO 3 ) 2 rather than MgCl 2  H 2 O, and barium acetate rather thanBaCl 2  was used to precipitate the sulfate. Electronparamagnetic resonance (ESR; Bruker EMX EPRspectrometer, X-band; microwave frequency 9.4GHz;power 20mW; temperature 77K with liquid N 2 ) wasused to detect the presence of metallic Ag in the freeze-dried residual solids (200mg samples) collected whenacanthite dissolution reached steady-state. All thespectra were collected at the same gain setting. 3. Results and discussion 3.1. Mineral characterization The XRD spectrum collected for the synthesizedAg 2 S mineral shows an exact alignment of the peakswith the reference lines for acanthite, the absence of anyadditional unexplained peaks, and an extremely lowbackground, suggesting that the solid is almost com-pletely crystalline with very little amorphous character(Fig. 1). However, investigation of the surface of themineral to a depth of 5 nm by XPS revealed that the Son the surface of the mineral is approximately83% G 0.5% sulfide and 17% G 0.5% sulfate. Due tooverlap, it was difficult to discriminate the Ag speciesbased on the 3d Ag peaks. Nevertheless, based on thefull width of the peak at half maximum height,corroborating information from the peak oxygen (O)bound to Ag (1s), and the sulfur speciation data, wecalculated that the distribution of Ag on the mineralsurface must be 52% G 1% Ag 2 O, 40% G 1% Ag 2 Sand 8% G 1% Ag 2 SO 4 . Had this distribution been truefor the bulk mineral, the Ag 2 O and Ag 2 SO 4  specieswould have been apparent in the XRD spectrum eitherasaseries ofdiscrete peaksoratthevery leastasahigherbackground. Since this is clearly not the case, the surfaceof the mineral is probably more oxidized than the bulkmineral. This situation occurs naturally in areas withfluctuating redox conditions, e.g., soils near a watertable, or in sediments disturbed by wave or tidal action. 3.2. Visual MINTEQ speciation The dissolution of the synthesized mineral in thebackground electrolyte and under the experimentalconditions used in these studies was modeled for 2different scenarios, using the speciation program  Visual MINTEQ  (Gustafsson, 2003). The calculated speciesincluded Ag C , Ag(HS) 2  , AgHS(aq), AgNO 3 (aq)Ag(OH) 2  , Ag(OH)(aq), AgS  , AgSO 4  , S  2 , HS  , andH 2 S(g). The first set of calculations, assuming a pure b -Ag 2 S solid in 1 mM KNO 3 , predicted a total Agconcentration in solution of 5 ! 10  14 M, which is farbelow the detection limits for Ag. The second set of calculations considered the presence of Ag 2 O, Ag 2 SO 4 ,and Ag 2 S on the surface of the mineral in 1 mM KNO 3 . Fig. 1. X-ray diffraction powder spectrum for the Ag 2 S precipitate (upper) with the  b -acanthite reference lines (lower).4  A.R. Jacobson et al. / Environmental Pollution 135 (2005) 1–9  The amount of each mineral present in the system wasbased on the relative percentages determined by XPS,the depth of penetration of the XPS analysis (5 nm), thesurface area of the solid (3.9 m 2 g  1 ), and the respectivemineral densities. This second scenario, considering anoxidized mineral rind, resulted in a prediction of 2.75 mM total dissolved Ag. 3.3. Acanthite dissolution experiments The dissolution curve (Fig. 2) for acanthite in thebackground electrolyte (1 mM KNO 3  initially at pH 6)is characterized by a fast and sharp rise, reaching steady-state conditions in 24 h. Experiments with differentelectrolytes (NaNO 3 , Ca(NO 3 ) 2 , NaClO 4 ) in concen-trations ranging from 0.1 mM to 10 mM resulted insimilar concentrations of Ag in solution, suggesting thatAg is not being released into solution due to oxidationof the sulfide by KNO 3 , complexation of Ag by NO 3 ,nor by K exchanging for Ag. The Ag concentration(0.125 mM), measured in the background KNO 3  elec-trolyte at steady-state, was much higher than would beexpected for pure acanthite, but was only 5% of whatwould be predicted if one accounts for the fact that theminerals at the surface of the solid include 52% Ag 2 O(see Section 3.1).By comparison, all the sulfur-containing solutions weinvestigated depressed the concentration of total dis-solved Ag to various degrees. Methionine had the leasteffect. It was the slowest system to reach a relativelysteady-state (7 days), but the Ag concentration appearsto be slowly approaching that of the control. The humicacids, which contain a variety of C, O and N functionalgroups besides those of S, suppressed Ag solubility byabout 75%. Both the thiosulfate and cysteine, eachcontaining a free –HS group, almost completelyinhibited acanthite dissolution.Given the presence and solubility of the oxidizedsilver species (Ag 2 SO 4  and Ag 2 O) on the mineral rind,the total concentration of soluble Ag is less thanthermodynamic predictions in all the cases. Severalhypotheses, could account for this observation. First,oxidative dissolution of the acanthite may have causedoxidation of the reduced S and concomitant reductionof the Ag C to Ag 0 . Metallic Ag could then bereprecipitating out of solution. In addition, in thesolutions containing the humic acids or model sulfides,the thiols could be complexing to Ag on the surface of the mineral preventing its dissolution, dissolved Agcould be forming ternary complexes, in which it acts asa bridge between the S in the mineral and the dissolvedthiols, or colloidal Ag–S cluster-complexes could beforming that are large enough to be filtered out of solution. These different processes will be exploredfurther in the following sections. 3.4. Evidence of sulfide oxidation and silver reduction3.4.1. Silver reduction and the role of Ag 0 in oxidation All the solutions were initially adjusted to pH 6.Since, to avoid chemical interference, the samples werenot buffered, solution pH was monitored throughout theexperiment. For acanthite in the presence of thiosulfate,cysteine and the humic acids, the pH of the solutionsremained constant (after an initial decrease) and therewas no correlation between the solution pH and theconcentration of total dissolved Ag (Fig. 3). In themethionine and KNO 3  –electrolyte systems, the pHdecreased throughout the course of the experiments(Fig. 3A). However, only in the methionine system thedecrease in pH was correlated with the concentration of total dissolved Ag (Fig. 3B).Explanations for the observed pH changes are limitedin these simple systems. The most probable cause for theincrease in acidity would be the oxidation of S 2  toSO 42  that has been suggested as an acanthite dissolutionmechanism in studies in which Ag-containing sedimentswere aerated (Di Toro et al., 1997; Manolopoulos,1997). This mechanism would be consistent with boththe increased concentration of Ag in solution anddecreases in pH observed in both the methionine-treatedsample and the control (KNO 3 ) for this study. Theinitial decline in the pH of the cysteine– and thiosulfate– acanthite systems from pH 6.0 to approximately pH 3.7could be due to deprotonation reactions in which Agbinding to the thiol group displaces a proton intosolution. The same phenomenon may have occurred toa lesser extent in the humic acid solution in which thepH dropped from an initial value of pH 6 to pH 5 within1 day. In addition, a variety of other functional groups dissolution time, days 0 2 4 6 8 10 12 14 16 18 20 22    A  g  c  o  n  c  e  n   t  r  a   t   i  o  n   i  n  s  o   l  u   t   i  o  n ,      µ    M 020406080140120100160cysteinehumic acidKNO 3 methioninethiosulfate Fig. 2. Acanthite ( b -Ag 2 S) dissolution over time in the presence of model organic ligands and humic acid dissolved in 1 mM KNO 3 . Errorbars represent 1 standard deviation from the mean of duplicatesamples.5 A.R. Jacobson et al. / Environmental Pollution 135 (2005) 1–9
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks