Uptake of Pb by hydrozincite, Zn 5(CO 3) 2(OH) 6—Implications for remediation

Please download to get full document.

View again

of 7
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.
Categories
Published
Uptake of Pb by hydrozincite, Zn 5(CO 3) 2(OH) 6—Implications for remediation
   Journal of Hazardous Materials 177 (2010) 1138–1144 Contents lists available at ScienceDirect  JournalofHazardousMaterials  journal homepage: www.elsevier.com/locate/jhazmat Short communication Uptake of Pb by hydrozincite, Zn 5 (CO 3 ) 2 (OH) 6 —Implications for remediation Pierfranco Lattanzi a , ∗ , Carlo Meneghini b , c , Giovanni De Giudici a , Francesca Podda a a Dipartimento di Scienze della Terra, Università di Cagliari, via Trentino 51, I-09127 Cagliari, Italy b Dipartimento di Fisica, Università Roma Tre, Italy c CNR-TASC, c/o GILDA, ESRF, Grenoble, France a r t i c l e i n f o  Article history: Received 6 July 2009Received in revised form13 November 2009Accepted 4 December 2009 Available online 16 December 2009 Keywords: EnvironmentHeavy metalsBioremediationSardiniaX-ray absorption a b s t r a c t Hydrozincite,Zn 5 (CO 3 ) 2 (OH) 6 ,periodicallyprecipitatesfromheavymetalcontaminatedwatersoftheRioNaracaulistream,Sardinia,inassociationwithabiologicalphotosyntheticcommunity.Theprecipitationremoves not only zinc from the waters, but also other toxic “heavy metals”, such as Cd, Cu, Pb. Thephenomenon is therefore of potential interest for “soft” remediation of contaminated waters. PreviouscationexchangeexperimentssuggestedthatbindingofPbtohydrozinciteisfairlystrong.Thissuggestionis in agreement with new release tests in deionized water and X-ray absorption spectroscopy (XAS)spectra collected at the Pb L  III  edge for natural hydrozincites from Naracauli, and synthetic Pb-dopedhydrozincites. The results suggest that, up to bulk concentration of 1.5wt.% Pb, uptake of this metaloccursintwodistinctways:(1)asasubstitutionforZninthetetrahedralsiteofthehydrozincitestructure,possibly via formation of a surface mononuclear tridentate inner sphere complex; (2) as an ill-defined,presumably amorphous, phase with a local atomic structure similar to cerussite. These data support theconcept that Pb binding to hydrozincite is strong enough to make this mineral a potential sink for themetal. © 2009 Elsevier B.V. All rights reserved. 1. Introduction Hydrozincite, Zn 5 (CO 3 ) 2 (OH) 6 , is a relatively common super-gene mineral in zinc deposits (e.g. [1]); it was also found in zinc-rich carbonate soils (e.g. [2]). Some years ago, it was observed as a primary mineral directly precipitated from the heavy metalcontaminated waters of Rio Naracauli, Sardinia, in associationwith a biological photosynthetic community, composed of an alga( Chlorella  sp.) and of a cyanobacterium ( Scytonema  sp.; [3,4]). The precipitationremovesnotonlyzincfromthewaters,butalsoothermetals,suchasCd,Cu,Ni,Pb,causingthe“heavymetal”contentsof the Rio Naracauli waters to fall below the recommended limits fordrinking waters. The phenomenon emphasizes the potential envi-ronmentalimportanceofhydrozincite,andmayhaveapplicationin“soft”bioremediationofcontaminatedwaters.Toreachthisend,anallimportantpointisunderstandingthewayinwhichmetalsotherthan zinc are bound to hydrozincite, since this will be the ultimatecontrol on the long-term effectiveness of their removal. Previousstudies [3,4], including conventional X-ray diffraction, SEM/EDS observations, and bulk cation exchange tests, could not provide adefinite answer to this question. Cation exchange tests [4] pointed outthatmetalslikePb,Ni,andCuarenotlooselyadsorbedontothe ∗ Corresponding author. Tel.: +39 070 6757722; fax: +39 070 282236. E-mail address:  lattanzp@unica.it (P. Lattanzi). hydrozincite surface, but are bounded in a stronger manner. It washypothesized that Ni and Cu, having ionic radii very similar to Zn,could be incorporated in the hydrozincite structure as substitutingions, in agreement with other evidence from literature [5,6]. The behaviourofalargeionlikePbismoreproblematic;moreover,thismetalisofenvironmentalconcernbecauseofitstoxicity.Thereforeweperformednewbulkreleasetests,andanX-rayabsorptionspec-troscopy (XAS) study at the Pb L  III  edge on natural and syntheticPb-doped hydrozincite. XAS is a powerful chemical selective tech-nique able to probe the local atomic structure around a specificabsorbing ion in the sample. This is the first study that providesspecific evidence on the nature of Pb binding to hydrozincite. Theresults have consequences for the potential use of bioprecipitatedhydrozincite for remediation of contaminated waters. 2. Experimental  2.1. Sample collection/preparation Samples for this study (Table 1) include two natural (samples N19 and N20) and three synthetic, Pb-doped, hydrozincites (sam-ples S3, S5, and S9). The reference samples were Pb-doped calcite(sample4-pbc;synthesizedaccordingto[7]),andnaturalcerussite (PbCO 3 ;sample22-cer).Naturalsamples(N19,N20)werecollectedat the Rio Naracauli in spring 1997 and spring 2001. Their con-ventional XRD patterns are consistent with hydrozincite (PDF N ◦ 0304-3894/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jhazmat.2009.12.026  P. Lattanzi et al. / Journal of Hazardous Materials 177 (2010) 1138–1144 1139  Table 1 Samples analyzed at European Synchrotron Radiation Facility (ESRF).Sample Description Pb (ppm)S3 Pb–hydrozincite, hot synthesis a 14,800S5 Pb–hydrozincite, cold synthesis a 4070S9 Pb–hydrozincite, hot synthesis a 1950N19 Natural hydrozincite b 3700N20 Natural hydrozincite c 65004-pbc Pb–calcite, cold synthesis a 53222-cer Natural cerussite (PbCO 3 ) – a See text for synthesis conditions. b Collected June 26, 2001; srcinal label ING4; other trace metals include750mg/kg Cd, and 100mg/kg Cu. c Collected May 7, 1997; srcinal label ING0N; other trace metals include540mg/kg Cd, and 260mg/kg Cu. 19-1458),andsuggestalowcrystallinity(typicalforthesehydroz-incites; see [8]). This low crystallinity corresponds to quite large surface areas (in the order of tens of m 2 g − 1 : [8]). Ageing effects were not detected (i.e., the XRD patterns are essentially simi-lar a few days and several years after collection in the field). Noother phase was recognized in the XRD patterns (including a fewtests with a synchrotron source, BM08 beamline, ESRF, Grenoble,France), nor by scanning electron microscope (SEM) observation.Thedetectionlimitofacrystallinephaseinsynchrotron-basedXRDpatterns is estimated in the order of 0.1% mass; for SEM observa-tions, it is more difficult to give a reliable detection limit.Synthetic hydrozincites were prepared in two ways: hot syn-thesis (samples S3 and S9) and cold synthesis (sample S5). Thehot synthesis route is a modification of the method describedby [9]. Hydrozincite was precipitated by mixing, at boiling point, equal amounts of a 32mM (NH 4 ) 2 CO 3  solution and a80mM Zn(NO 3 ) 2 · 6H 2 O solution; the molar ratio Zn/CO 3  in themixed solution was thus the stoichiometric ratio for hydroz-incite. Thermodynamic calculations (by the PHREEQC code [10];for hydrozincite, the thermodynamic data reported by [11] wereused) suggest that at these synthesis conditions supersaturationwithrespecttohydrozincite,expressedasSI=logIAP/ K  ,wasabout5. Pb was added in the form of Pb (NO 3 ) 2 ; the initial concentra-tionsinthemixedsolutionswere0.39mMforsampleS3(Table1),and 0.039mM for sample S9. In the first case, the solution wassupersaturated with respect to PbCO 3  (cerussite), Pb 3 (CO 3 ) 2 (OH) 2 (hydrocerussite), and Pb(OH) 2  (SI=1.5, 3.6 and 1.9, respectively);inthesecondcase,thesolutionwasclosetoequilibriumwiththesephases (SI=0.6, 0.1, 0.1). None of these phases was ever observedin the XRD patterns (including those obtained with a synchrotronsource)neitherofprecipitates,norbySEMobservation.AdditionaltestsathigherPbconcentrationsshowedthatthestrongestdiffrac-tion peaks of cerussite appear in the (conventional) XRD patternwhen Pb concentration exceeds 3.86mM. Hydrocerussite did notform even for initial concentration as high as 80mM, unless pHwas raised beyond 10. As for lead hydroxide, its occurrence in thepresence of carbonate is unlikely (cf. [12]).The cold synthesis method is derived from that used by [7] f or the synthesis of Pb-doped calcite, which is in turn adapted from[13]. One liter of a 50mM Zn(NO 3 ) 2 · 6H 2 O  − 280mM NH 4 Cl solu-tion was placed into a 1.5l closed reactor. The upper part of thereactor was in communication with a vial containing (NH 4 ) 2 CO 3 as a fine powder. The slow decomposition of this salt releases NH 3 and CO 2  to the solution. Hydrozincite began to precipitate at thewater–vapour interface after about 24–48h. When precipitationwaswelladvanced,PbwasintroducedseparatelyasPbCl 2  solution,to give an initial concentration of 0.048mM. In such an inherentlyinhomogeneous, dynamic system it is difficult to perform mean-ingful equilibrium modelling. However, as soon as hydrozincitebegan to precipitate, the solution was sampled as close as possi-ble to the forming solid, and analyzed for zinc and alkalinity, tocalculate the carbonate concentration. The pH was also measured.From these data, the SI of hydrozincite was calculated to be about − 1(i.e.,thebulksolutionisnominallyundersaturatedwithrespectto this phase). Probably, saturation conditions are reached only atthe water–vapour interface.Precipitates were filtered either on Whatman N ◦ 54 disks (hotsynthesis), or on Nuclepore 0.4  m membranes (cold synthesis),washed with distilled water, and dried at room temperature. XRDpatterns in all cases showed no phase other than hydrozincite.Synthetic hydrozincites also show a low degree of crystallinity.However,theaveragecrystallitesize(asdeterminedbyTEMobser-vationandfromthewidthofXRDpeaksviatheScherrer’sequation)is slightly larger than natural hydrozincite from Naracauli [8]. A portion(about0.1g)oftheprecipitatesandofnaturalsampleswasdissolvedin2mLofHNO 3  10vol.%(CarloErbaSuprapur),andMilli-Q water was added up to 10mL; the solutions were analyzed forPb by ICP-OES. The calculated Pb contents in the solid phases arereported in Table 1.  2.2. Bulk release tests Previous cation exchange tests [4] were conducted with an 1MMgCl 2  solution (solid to liquid ratio 1g to 40mL; pH ∼ 7, adjustedbyNaOH).Inthisstudy,weconductedbulkbatchreleasetestswithdemineralisedwatertosimulateinteractionwithmeteoricwaters.The tests were performed on natural hydrozincite (sample N20;Table 1), and on the two synthetic samples with the highest andthe lowest bulk Pb contents (S3 and S9, respectively). Approxi-mately0.6gofeachsamplewereplacedintoabeakerwith660mL ofMilli-Qwaterandamagneticstirrertofacilitatethereaction.Thesupernatant solution was then periodically sampled, and analyzedfor Zn and Pb (ICP-MS). In a first long-term ( ∼ 1100h) experiment,we determined the time necessary for the system to level off to astationary state for both metals, i.e. no further change (within theanalytical uncertainty of   ± 10%) was observed in their concentra-tions(Fig.1a).CalculationswiththePHREEQCcodeindicatethatat these conditions the system was close to saturation with hydroz-incite (SI ∼ 0), strongly undersaturated with respect to cerussite(SI ∼− 2) and hydrocerussite (SI ∼− 6), and slightly undersaturatedwith respect to smithsonite (SI ∼− 1). The experiment was thenrepeated for a shorter time (528h), but with more detailed sam-pling.  2.3. XAS experiment  The XAS experiments were carried out at the Pb L  III  edge(13035eV) at the BM08-GILDA (General Italian beamLine forDiffraction and Absorption) beamline [14] at the ESRF (European Synchrotron Radiation Facility), Grenoble (France).XAS spectra were measured in fluorescence geometry, exceptfor cerussite sample. Samples for fluorescence XAS data collectionwere prepared by suspending sample powders on aqueous solu-tion (deionized water) and deposing the powders as thin filmson Millipore membranes. Because the solubility of hydrozincite indeionized water is not negligible, small amounts of Na 2 CO 3  wereadded to the suspensions to obtain an alkaline environment. ThesuspensionhadapHofabout11.5.Afterfiltration,theliquidphasewas analyzed for Zn and Pb by ICP-MS. From the metal content,weestimatethathydrozincitedissolutionduringmanipulationwasless than 0.1wt.%. The films were sealed by Kapton tape and trans-ferred on the XAS measurement chamber. The cerussite sample(22-cer), measured in transmission geometry, was prepared in theform of a pressed pellet by mixing sample powders (about 16mg)and pure cellulose (about 100mg), then pressing at 5kbar for fewminutes.  1140  P. Lattanzi et al. / Journal of Hazardous Materials 177 (2010) 1138–1144 Fig. 1.  (a) Variations with time of Zn and Pb concentrations during batch reactionof hydrozincite with Milli-Q water. The parabolic best fits to experimental data areshown.Theerrorbarscorrespondtotheanalyticaluncertaintyof10%.(b)Variationswith time of the Pb/Zn concentration ratios during batch reaction of hydrozincitewith Milli-Q water, normalized to the Pb content of the solid. The error bars corre-spond to an overall uncertainty of 15%. XAS measurements were performed keeping the sample at theliquidnitrogentemperature(77K)inordertominimizethethermaldisordereffectontheXASstructuralsignal.Fluorescencemeasure-mentswereperformedmeasuringthePbfluorescence( I  f  )byahighpuritysolidGemultidetector(13elements).AthinAlfilterwasusedin order to attenuate the intense Zn fluorescence peak on hydroz-incitesamples.TheincomingX-raybeamintensity I  o wasmeasuredby a N 2 -filled ionization chamber. XAS scans were repeated 2–5times in order to improve the counting statistics. The overall dataacquisition for each fluorescence sample required up to 10h.Transmission XAS spectra on cerussite sample were collectedby measuring the X-ray intensity transmitted through the sample I  t , using a second N 2 -filled ionization chamber.Standardprocedures[15]wereusedfordatanormalizationandto extract the structural EXAFS (extended X-ray absorption finestructure) signal   ( k ); these procedures include pre-edge back-ground removal, spline modeling of bare atomic background   o ,and edge step normalization. Polynomial splines of third degreewere used to model the post-edge bare atomic background   o between the edge (13035eV) and the end of the spectra. The pho-toelectron wavevector: k = ¯ h − 1   2 m e ( E  − E  0 )was calculated choosing the edge energy  E  o  at the first inflectionpoint (first derivative maximum) of    ( E  ) and refined during thedata fitting.The EXAFS data refinement was performed using the FITEXAcode[16]whichusestheMINUITroutinesofCERNFortranlibraries[17] f or accurate least square refinement and statistical error anal-ysis. Data refinement was performed on the raw  k  ( k ) spectra (i.e.without Fourier filtering), applying the standard EXAFS formula[15].Modelatomicclustersobtainedfromcrystallographicmodels[18] were used to calculate theoretical amplitude and phase func-tions with FEFF8.2 code [19,20]. These clusters were the starting points for the refinements. Coordination numbers (CN), bond dis-tances ( R ), Debye–Waller factors (   2 ) and the experimental edgeenergyshift(  E  )wereletfreetovaryintherefinements.Errorsonparameters were calculated using the MINOS subroutine from theMINUITpackage[17],whichalsotakesintoaccountthecorrelation between parameters. The  R 2 factor was used to evaluate the bestfitquality;thisfactorrepresentsthemeasureoftheabsolutemisfitbetween experimental data and theory [21], and is defined as R 2 =  i (  ( k i ) −  th ( k i )) 2  i (  ( k i )) 2 where   ( k i ) and   th ( k i ) are respectively the experimental data andthe theoretical values at the point  k i . 3. Results  3.1. Bulk release tests Inthelong-termpreliminarytest,theexperimentaldataforbothPb and Zn can be fitted, within experimental error, to two nearlyparallelparabolictrends(Fig.1a),consistentwithbothmetalsbeing released from hydrozincite at approximately the same rate (thus,conceivably, with similar mechanisms). During the shorter tests,thedissolutionrateofthevarioussampleswassomewhatvariable,even for different aliquots of the same bulk sample; moreover, theamount of Pb released to solution was dependent on the bulk con-tentofthestartingsolid.Therefore,webelievethatthebestwaytocompare the ability of each sample to retain Pb is to normalize theamount of released Pb to the initial concentration in the solid, andto the amount of released Zn (which is a measure of the bulk dis-solution rate). Under this approach, Fig. 1b reports, as functions of  time,thePb/Znratiosintheleachsolutionsincontactwithhydroz-incite, normalized to the initial Pb concentration in the coexistingbulk solid. For all samples, the ratios tend to decrease with time,although for S9 the trend is less pronounced. In all three cases, thePb/Znmolarratioinsolutionislowerthaninthecoexistinghydroz-incite (Table 2), i.e. Pb is even less mobile than Zn. This finding is consistent with metal ratios in waters coexisting with the naturalsample (Pb/Zn molar ratio ∼ 6.5 × 10 − 4 ; [4]).  3.2. XAS experiments Fig. 2 reports the near edge region of the normalized spectraof all samples and reference compounds. It is evident from thesecurves that the Pb L  III  edge XANES spectra of all hydrozincite sam-ples (natural and synthetic) are very similar, whereas they areappreciably different from those collected on both Pb-doped cal-cite (sample 4-pbc) and cerussite (sample 22-cer), suggesting alocal environment of Pb strongly different in hydrozincite and inthe reference compounds. Moreover, our Pb L  III  edge XANES of hydrozincitesamplescloselyresemblethespectrumofPb 4 (OH) 44+ as reported in references [23,24]. Fig. 3 reports the Pb EXAFS spec- tra of hydrozincite and cerussite samples (spectra for Pb–calciteis quite noisy, and did not allow further analysis in the extended  Table 2 Molar ratios Pb/Zn in leaching test solutions and in hydrozincite.Sample Pb/Zn ◦ (s) Pb/Zn (aq)N20 3.95 × 10 − 3 7.7 × 10 − 4 S3 9.1 × 10 − 3 2.6 × 10 − 3 S9 0.96 × 10 − 3 2.2 × 10 − 4 Pb/Zn ◦ (s)=molar ratio in hydrozincite (calculated from data in Table 1). Pb/Zn (aq)=molar ratio in the leach solution at the end of the experiment.  P. Lattanzi et al. / Journal of Hazardous Materials 177 (2010) 1138–1144 1141 Fig. 2.  Near edge region of the XAS spectra collected on the measured samples(vertically shifted for clarity). The spectra for hydrozincite samples (S, synthetichydrozincite, N, natural hydrozincite) are different from those of the referencecompounds (4-pbc, Pb–calcite; 22-cer, cerussite). region).TheEXAFSspectraof Fig.3qualitativelyconfirmthestrong analogiesamonghydrozincitesamples,andthedifferencewiththePblocalstructureincerussite.MoredetailsonthePbL  III  EXAFSdatarefinement are shown in Fig. 4. The top panel reports the best fit and the partial contributions in reciprocal ( k ) space; the Fouriertransforms of experimental data, best fit and partial contributionsare shown in the bottom panel. The quantitative results of EXAFSdata refinement are reported in Table 3.TheanalysisofEXAFSdataforthecerussitestandardreproducesthestructuraldatafromliteraturefairlywell[18]:theEXAFSsignal refinementshowsafirstPb–Oshellaround2.63Å,twoPb–Cshellsat about 3.1 and 3.4Å, and a Pb–Pb shell around 4.2Å. Fig. 3.  Experimental EXAFS spectra (triangles) and best fit (full lines) for theanalyzed samples (vertically shifted for clarity). The strong analogies among thehydrozincite samples, and the large difference with the cerussite EXAFS data areevident.  Table 3 Fit parameters for EXAFS analysis.CN a R  (Å)    2 ( × 10 3 Å) b StatisticsS3Pb–O 4.1 (6) 2.36 (2) 17 (8)   E  =2.7Pb–O 2.8 (3) 2.69 (3) 17*  R 2 =1.5Pb–Pb 1.8 (3) 3.9 (1) 16 (6)Pb–Pb 1.7 (3) 4.1 (1) 13 (4)S5Pb–O 3.9 (6) 2.33 (2) 10 (4)   E  =6.7Pb–O 3.2 (4) 2.55 (2) 10*  R 2 =1.3Pb–Zn 0.8 (2) 3.95 (6) 2.6 (2)Pb–Pb 1.4 (4) 4.1 (1) 8.0 (7)S9Pb–O 4.4 (4) 2.30 (2) 9.6 (3)   E  =1.3Pb–O 2.4 (3) 2.65 (2) 9.6*  R 2 =1.1Pb–Zn 1.9 (3) 3.86 (5) 12 (4)N19Pb–O 4.1 (4) 2.34 (2) 14 (6)   E  =2.1Pb–O 2.9 (4) 2.70 (3) 14*  R 2 =0.8Pb–Zn 1.8 (3) 3.94 (4) 5.7 (6)N20Pb–O 4.2 (4) 2.31 (2) 16 (6)   E  =1.9Pb–O 2.7 (4) 2.65 (2) 16*  R 2 =0.9Pb–Zn 2.5 (3) 3.88 (4) 14 (2)22-cerPb–O 9 2.63 (2) 13   E  =3.1Pb–C 3 3.08 (3) 14  R 2 =0.1Pb–C 3 3.44 (5) 6.7Pb–Pb 6 4.18 (4) 7.7Key to symbols–abbreviations: CN=coordination number; R=average interatomicdistance;    2 =Debye–Waller factor;   E  =edge energy shift;  R 2 =misfit factor (seetext). a For cerussite (sample 22-cer), CN was kept fixed to the theoretical values. b The Debye–Waller parameters for the second Pb–O shells (*) were constrainedto the same value as the first shell. The analysis of Pb EXAFS spectra of hydrozincite samples isa more complex task, due to a weaker signal, higher disorder inthe structure, and high correlation among refinement parameters.Aftersomepreliminarytests,weconcludethatthebestinterpreta-tionforthenearestneighbourPb–Oshellisabimodaldistribution:a first Pb–O shell around 2.3Å made of about 4 neighbours, anda second one around 2.6–2.7Å with about 3 neighbours, givingan overall Pb–O coordination number (CN) around 7. In syntheticsamples (S3, S5, S9), such a bimodal contribution is mandatory toimprove the fitting. Specifically, the  R 2 factor and the reduced   2 aresignificantlyreduced(upto10%)usingabimodalPb–Odistribu-tioninsteadofasingleshell.Onthecontrary,usingabimodalPb–Odistribution in refining the natural samples, the  R 2 factor and thereduced   2 decrease less than 2%. Therefore, the two alternatives(single or double shell) are indistinguishable from a purely statis-tical point of view. However, because of the similar behaviour of Pb in both natural and synthetic samples during bulk release tests,we find more appropriate to adopt the same solution (i.e., bimodaldistribution) for all hydrozincites.The difficulty to distinguish between bimodal or single shellPb–O distribution is mainly due to the anti-phase interference of the two Pb–O signals at around ∼ 2.3 and ∼ 2.6Å. This interferenceis quite evident in the synthetic samples (Fig. 4), and gives rise to a strong correlation among the structural parameters (mainly CNand Debye–Waller factors). In order to reduce this correlation, weconstrained the Debye–Waller factors of Pb–O shells to the samevalue (Table 3). However, the CN estimates are affected by a fairly large (10–20%) uncertainty.BeyondthefirstpeakintheFouriertransform,anextneighbourcoordination shell is clearly recognized in all samples, even withsome differences. Except for samples S3 and S5, the best fit for this  1142  P. Lattanzi et al. / Journal of Hazardous Materials 177 (2010) 1138–1144 Fig. 4.  Top: results of the Pb L  III  edge EXAFS data refinement on natural and synthetic hydrozincite samples. The top curves represent the experimental data (triangles), andthebestfits(fulllines).Belowareshownthepartialcontributionsusedinthefitting.Thelowermostcurves(gray)representtheresidual(experimental–theory).Thereduced  2   values of the best fit are reported:   1   = (1 /N  −  p )  i [  ( k i ) −  th ( k i )] 2 /    2    in which  N  −  p  is the number of independent data points. Bottom: Fourier transform of experimental data, best fits, and partial contributions used in the Pb EXAFS data refinement: the top curves represent the modulus of experimental data (triangles) and bestfits (full lines). Below are reported (shifted for clarity) the moduli (bold lines) and imaginary parts (gray lines) of Fourier transform for the partial contributions used in therefinements. shell is obtained by assuming a Pb–Zn shell at a distance of  ∼ 3.9Å.In sample S3, this shell must be fitted with two contributions,correspondingtointeratomicdistancesof  ∼ 3.9and ∼ 4.1Å,respec-tively. The component corresponding to the longer distance canbe successfully fitted as a Pb–Pb distance; for the shorter distance(veryweakcontribution;Fig.4)someambiguityremainsbetweena Pb–ZnandaPb–Pbinteraction(i.e.,neithersolutionmodifiessignif-icantlythe R 2 factor;cf.Fig.4).SampleS5alsorequiresanadditional Pb–Pb shell around 4.1Å. This contribution is weak, but definitelyrelevant to improve the best fit ( R 2 factor). 4. Discussion Any model of the interaction of Pb with hydrozincite mustaccount for the following facts:1. Pb is strongly bonded to hydrozincite. Previous cation exchangeexperiments [4] indicated that only 0.08% Pb was removed by a 2.5htreatmentwitha1MMgCl 2  solution.Releasetestsindem-ineralisedwatercarriedoutinthisstudyconfirmthereluctanceof Pb to be removed from hydrozincite.
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