An experimental study on metal precipitation driven by f luid m ixing:im p lications for genesisof carbonate-hosted lead-zinc ore deposits

Yan Zhang·Runsheng Han·Xing Ding·Junjie He·Yurong W ang

Abstract A type of carbonate-hosted lead—zinc(Pb—Zn)ore deposits,known as M ississippi Valley Type(MVT)deposits,constitutesan importantcategory of lead—zinc ore deposits.Previous studies proposed a f luid-m ixing model to account formetal precipitationmechanism of the MVT ore deposits,inwhich f luidswithmetal-chloride complexes happen to mix w ith f luids w ith reduced sulfur,producing metal sulf ide deposition.In this hypothesis,however,the detailed chem ical kinetic process ofm ixing reactions,and especially the controlling factorson themetal precipitation are not yet clearly stated.In this paper,a series ofm ixing experiments under ambient temperature and pressure conditions were conducted to simulate the f luid mixing process,by titrating themetal-chloride solutions,doping with State Key Laboratory of Isotope Geochemistry,Guangzhou Institute of Geochemistry,Chinese Academy of Sciences,Guangzhou 510640,China Key Laboratory of M ineralogy and Metallogenic,Guangzhou Institute of Geochemistry,Chinese Academy of Sciences,Guangzhou 510640,China University of Chinese Academy of Sciences,Beijing 100049,China

Keywords Metal precipitation·Fluid m ixing·Sulfur species·MVT lead—zinc ore deposits·Carbonate-hosted lead—zinc deposits.

1 Introduction

Fluidm ixing is considered to play an important role in the formation of varioushydrothermalore deposits,such as the Carlin-type gold deposits in the Jerritt Canyon,Nevada(Cooke 1996;Hofstra et al.1988),Creede polymetallic vein deposits,Colorado(Henley 1984;Henley etal.1984;Plum lee 1994),M ississippi Valley-type lead—zinc ore deposits(Anderson 1975;Leach etal.2005;Plum lee etal.1994),epithermal deposits in the Pacif ic Rim(Jiang et al.2004),Olympic Dam-type Cu—U—Au deposits(Fang and Li 2014;Hayneset al.1995),andporphyrycopper—molybdenum—gold deposits(Fan et al.2001;Li and Liu 2002).An important reason is that f luid m ixing can effectively facilitate the precipitation andm ineralization of ore-form ingmetalelements(Leach etal.2005;Reed 2006;Reed and Spycher 1985).Therefore,studies on f luid m ixing process can considerably promote better understanding of the genesisof largeand gianthydrothermalore deposits.

Fluidm ixing is themost favorablegenesismodel for the formation of M ississippi Valley Type(MVT)Pb—Zn ore deposits(Anderson 1975;Beales1975;Bealesand Jackson 1966),besidessulfatereductionmodel(Anderson 1973,1991;Barton 1967),and reduced sulfur model(Anderson 1973,1975)(Table 1).A pioneering study from Beales and Jackson(1966)on the Pine Point Pb—Zn ore deposit,Canada,proposed the f irst f luid mixingmodel.In the model,some kind of chloride complex-bearing f luid originated from a distal basin migrated into the metallogenic area,and m ixed w ith local H2S-bearing f luid,resulting in the precipitation of metals and deposition of sulf ides.Corbella et al.(2004)demonstrated that f luid m ixing is an eff icientway to produce the MVT ore deposit and associated carbonate dissolution.In recent years,this model received increasing support from studieson ore f ield structures(Brown 1970;Ohle 1985),isotopes(Bottrell et al.2001),homogenization temperature,and salinity of f luid inclusions(Kesler et al.1997;Leach et al.1993),halogen and inert gases(Grandia et al.2003;Kendrick et al.2002),and numerical simulations(Anderson and Garven 1987;Appold and Garven 2000;Corbella et al.2004;Plum lee et al.1994;Reed 2006;Reed and Spycher 1985).A general understanding is that during the f luid mixing process Pb and/or Zn ions chem ically react w ith reduced sulfur species in the f luids,producing sulf ide precipitation.In a hydrothermal system,however,reduced sulfur could occur in the form of H2S,S2-,S3-,or HS-(Manning 2011;Pokrovskiand Dubrovinsky 2011;Tossell 2012).Which form of reduced sulfur dom inates themetal precipitation of the MVT Pb—Zn ore deposits and is involved dynamic process during themetal precipitation is still unclear.

In this paper,we conduct a series of f luid m ixing experiments at ambient temperature and pressure conditions,through titrating Pb and/or Zn-bearing chloride solutions w ith sulfur-bearing solution.We also used the experimental results,combined w ith the thermodynamic calculations,to depict the detailed chemical kinetic process of mixing reactions,the geochemical pathways on metal precipitation associated to the carbonate-hosted Pb—Zn deposits,and the main controlling factors on the metal precipitation.

2 Experimentalmethods

2.1 Startingmaterials

Solution preparation was operated at the hydrothermal laboratoryof the high-temperature andhigh-pressure experimental platform in the Guangzhou Institute of Geochem istry(GIG),Chinese Academy of Sciences.Three kindsof Pb and/or Zn chloride solutions,aswellasa NaHS solution,weremade up to be the analoguesof initialmetal chloride-containing and sulfur-bearing f luids,respectively.Lead and/or zinc chloride solutions prepared for the experiments are as follows:(1)0.01mol/L ZnCl2and 1mol/L NaCl;(2)0.0005mol/L PbCl2and 1mol/L NaCl;and(3)0.01 mol/L ZnCl2,0.0005mol/L PbCl2,0.002 mol/LCaCl2,0.002 mol/LMgCl2,and 1mol/L NaCl.Lead and zinc concentrations inall three initial solutions are approximately 100 ppm and 650 ppm,respectively.These valuesarequite similar to the Lead and zinc concentrations in ore-forming f luids of the MVT Pb—Zn ore deposits(Carpenter et al.1974;Czamanske et al.1963;Pinckney and Haffty 1970;Stoffelletal.2008;Yardley 2005).Initial NaHSsolutionsare variousat the concentrationsof0.0002,0.002,0.001,0.01,0.02,and 0.1 mol/L.All reagents usedduring the preparation process of initial solutions are of analytical purity and the water used was deionized.

Table 1 Representativemodels for them igration and deposition of sulf ides in the MVT deposits

To facilitate the observation of sulf ide precipitates produced during f luid m ixing process,a dolom ite,the most commonhost rocks for carbonate-hostedPb—Znore deposits(Gerdemann and Myers 1972;Gregg 1985;Han etal.2007;Lyle1977;Rowan 1987;Sass-Gustkiewicz and Dzulyn′ski1998;Zhang etal.2009),wasalso prepared.The dolomite sample was fresh and f ine-grained.It was collected from the Lower Carboniferous Baizuo Formation,located in the Huize M ining Area of Yunnan Province,China.The dolom ite consisted ofmore than 80%dolom ite andm inor calcite,w ith a greatquantity of Ca,Mg,Fe,Mn and other trace elements(Fraser et al.1989;Jazi et al.2017).The sample was f irstly crushed and then powdered in an agatemortar to 40mesh in dimension.

2.2 Experimental strategy and procedure

Given that MVT Pb—Zn ore deposits are the product of regional or sub-continental scale hydrothermal processes which have nothing to do w ith magmas(Leach et al.1993,2001,2005;Stoffell et al.2008;Sverjensky 1986).They form at very low pressure and temperature conditions,for instance,mainly at 50—200°Cand w ithin 100 MPa(Banks and Russell 1992;Grandia et al.2003;Gratz and M isra 1987;Leach et al.1996,2004,2005;Marieand Kesler2000;Roedder1977;Samson and Russell 1987;Savard etal.2000).At these conditions,stabilitiesof Pb and Zn chloride complexes are seldom affected by the pressure and temperature(Fig.1).Because of this,we designed a series of mixing experiments under ambient temperature and pressure conditions to simulate the f luid mixingprocess,bytitratingmetal-chloride solutions,dopingw ith orw ithoutdolom ite,and using NaHS solution.Onthe basis of these experimental results,we can extrapolate chem ical kinetic processes to the conditions related to the MVT Pb—Zn ore deposits.

The experiments were designed in two groups,f luidmixing experiments and f luid—rock reaction experiments.Allof the experimentswere performed at the hydrothermal laboratory high-temperature and high-pressure experimental platform in the GIG.

(1)Fluid-mixing experiments were conducted through a method of titrating Pb and/or Zn chloride solutions w ith sulfur-bearing solutions,including three sets of parallel experiments as follows:

1.M ixing of solutions containing Zn and NaHS 10 m L of the prepared Zn chloride solution was placed in a 50 m L plastic vial.Hydrochloric acid or sodium hydroxide was added to adjust solution acidity to pH=2.09,2.70,3.76,5.00,and 5.80.The solutionsw ith differentpH values were titrated using 0.002mol/L NaHS solution.

2.Mixing of solutions containing Pb and NaHS 10 m L of the prepared Pb chloride solution was placed in a 50 m L plastic vial.Hydrochloric acid or sodium hydroxide was added to adjust solution acidity to pH=1.38,2.38,2.42,3.00,3.20,4.00,5.00,and 6.00.The solutions w ith differentpHvaluesweretitratedusing 0.0002mol/L NaHS solution.

3.Mixing ofsolutions containing Pb—Zn and NaHS 10m L of the prepared Pb—Zn chloride solution wasplaced in a 50m L plastic vial.Hydrochloric acid or sodium hydroxide was added to adjust solution acidity to pH=1.66,2.40,3.32,and3.90.The solutions w ith different pH values were titrated using 0.002 mol/L NaHS solution.

During slow titration of the NaHS solutions into the Pb—Zn chloridesolutions,continuousshaking of theplastic vial was necessary to prevent the precipitates from local oversaturation of NaHS and to speed up chem ical reactions.Meanwhile,we kept close watch on the precipitation processes from the solution m ixtures.Once the precipitates were observed,an excess amount of NaHS solution was rapidly added into the m ixtures.During the procedure above,the pH value was measured at different steps.Finally,the solid precipitates and residue solutions in the plastic vialwere separated from each other by f iltration for XRD and ICPMS analysis,respectively.

(2)Fluid-rock reaction experiments were carried out through titrating NaHS solution into the Pb—Zn chloride solutions doping w ith 2 g of 40-mesh f inegrained dolomite powder collected from the Huize M ining Area,China.The purpose of these experimentswas to observe the role ofwall rock onmetal precipitation during the f luid m ixing process.Considering that the pH value of the solutions in these experiments could be varied due to the disequilibrated f luid—rock reaction,it therefore cannot truly ref lect the chem ical kinetics of the f luid m ixing process.So,the pH values of the solutions were no longermeasured.After the titration,the solid phases were separated from the solutions,by f iltration and dry for electronm icroprobe analysis.

2.3 Analyticalm ethods

pH measurements of solutionswere performed in the GIG hydrothermal laboratory by a FiveEasyTMpH metermanufactured by Mettler-Toledo International,Inc.Calibrationsusingapreparedbuffer solution(GGJ-119).Calibration was carried out before each pH measurement.XRD and EPMA analyses were f inished in the test center of the South China University of Technology.The equipment used for XRD analysis is the D8ADVANCE diffractometermanufactured by Bruker Corporation,Germany.The analytical conditionsare as follows:wavelength of incident ray is 0.15418 nm;tube pressure and f low were 40 kVand 40mA,respectively;scanning range,step length,and speed are 5°—90°,0.02°,and 19.2 s/step,respectively;and the slit was DS 0.5°RS 8mm(corresponding to a LynxExe array detector).For the EPMA analysis,the EPMA-1600 manufactured by Shimadzu Corporation,Japan was employed,together with the Genesis energy spectrometermanufactured by EDAX Inc.,the United States.The experimental conditions are as follows:acceleration voltage was 2.0 kV,resolution of secondary electronswas 6 nm,X-ray detection angle was 52.5°,and energy resolution is 120 eV.

3 Experimental results

3.1 Titration curve for f luid m ixing experiments

In the f luidmixing experiments,w ith the addition of NaHS solutions into various initial pH metal chloride solutions,the pH values of all them ixed solutions increased slow ly.Once the titration amountof NaHS solution was enough to deposit all the Pb and/or Zn ions,the pH values clearly declined(Table A.1,Fig.2).For example,in experiment Hh-3,0.002mol/L NaHS solution was titrated into the Zn chloride solution w ith an initial pH value of 3.76.W ith the increaseof titration volumesof NaHSsolution from 3.50 to 14.00m L,the pH value ofmixed solutions changed from 3.88 to 4.05.When the titration volume of NaHS solution reached 60.00 m L,however,the pH value ofm ixed solutions declined to 2.53.This tendency occurs in notonly for single Pb or Zn chloride solutions,butalso complex Pb—Zn chloride solutions aswell(Fig.2).

During the titration,precipitates could not be observed until the titration amount of NaHS solution was large enough(Table A.1,Fig.2).Usually the lower pH values of the initialmetal-bearing solutions had,themore NaHS solutionwasneededtoproduceinitial precipitates(Table A.1,Fig.3).For example,in the experimentof Hh-3,0.002mol/L NaHS solution was titrated into the Zn chloridesolutionw ith an initialpH valueof3.76.While the titration volume of NaHS solution reached 1.50m L,no obviousprecipitatescould beobserved,and thepH valueof mixed solutions became to be a little higher,i.e.3.80.Till the NaHS solution was added up to 3.50m L,precipitates were indeed observed for the f irst time.By contrast,in the experimentof Hh-4,itneeds at least6.00m L 0.002mol/L NaHS solution to produce obvious precipitates w ith an initial pH value of 2.70.

Comparing w ith Zn chloride solutions,Pb or Pb—Zn chloride solutionsw ith similar initial pH value needmuch less 0.002mol/L NaHS solution to produce obvious precipitates,as shown in Fig.3.For example,3.50m L 0.002mol/L NaHS solution was titrated into the Zn chloride solution w ith initial pH value of 3.76 and thus can produce obvious precipitates.However,as for Pb chloride solution w ith initial pH value of 4.00 and Pb—Zn chloride solution w ith initial pH value of 3.90,the titration volumes of NaHS solution needed were 0.80m L and 0.20m L,respectively.Considering that the initial pH values of metal-bearing chloride solutions were similar and the concentration of titrated NaHS solution was also constant,such a difference most likely results from initial metal concentrationdifferenceanddiscrepant geochemical behavior between Pb and Zn.

Fig.2 Titration curves for NaHS solution into various initial pHmetal-bearing solutions.a Titration curves for NaHS solution into Zn-bearing solutionsw ith various initial pH;b titration curves for NaHS solution into Pb-bearing solutionsw ith various initial pH;c titration curves for NaHS solution into Pb and Zn-bearing solutionsw ith various initial pH.In these experiments,w ith the addition of NaHS solutions into metal chloride solutions,the pH valuesof all themixed solutions increaseup slow ly;After the titrated amountof NaHSsolution isenough tomakeall the Pb and/or Zn ions precipitate,the pH values decline notably

3.2 Precipitates produced by f luid-rock reaction experiments

Experiments on titrating NaHS solution into the Pb—Zn chloride solutions doping w ith 2 g of 40-mesh f ine-grained dolomitepowderdemonstrate thata largeamountof sulf ide precipitates can be formed.As shown in Fig.4,large numbers of nanoscale precipitateswere found on the surface of the dolom ites.EDS spectrum analysis on these precipitates indicates that the main components of these precipitates are Pb,Zn,and S(Table 1,Fig.4).Furthermore,XRD analysis of these solid precipitates show that they mainly consisted of galena and/or sphalerite(Fig.5).

4 Discussion

4.1 Species and distributions of sulfur in the hydrotherm al f luids related to the MVT Pb-Zn ore deposits

In the f luid m ixing model on the formation of carbonatehosted MVT Pb—Zn ore deposits,it is usually thought that reduced sulfur species in the f luids dominate the metal precipitation(Anderson 1975;Beales and Jackson 1966;Corbella etal.2004;Giordano 2002;Giordano and Barnes 1981;Leachet al.2005,2006;Reed2006).Ina hydrothermalsystem,reduced sulfur can occur in the forms of H2S,S2-,S3-,or HS-(Manning 2011;Pokrovski and Dubrovinsky 2011;Tossell 2012).The free radical S3-is dom inantwhen temperatures are above 250°C,especially at350°C and 0.5 GPa(Pokrovskiand Dubrovinsky 2011;Tossell 2012).Considering that the MVT Pb—Zn ore deposits mainly form at temperatures of no more than 200°C and pressures of no more than 0.1 GPa(Basuki2002;Conliffe et al.2013;Han et al.2016;Leach et al.2005;W ilkinson 2001),S3-radical cannot act as a dom inant sulfur species in the Pb—Zn metallogenic hydrothermal f luids.Therefore,during the f luid m ixing processes related to theMVTPb—Zn oredeposits,H2S,S2-,and HSare the potential dominant species.Based on sulfur's chem ical equilibriums in the hydrothermal system and thermodynamic data(Lin etal.1985),calculated pH-log fO2phase diagrams of sulfur at 298 K and 523 K show that H2S and S2-are stable at acidic and strongly basic pH conditions,respectively,while the stability f ield of HSlies between them(Fig.6).Note that higher temperature compresses the stability f ield of HS-,thus the boundary between HS-and H2S consequently changes from the pH values of about 7.0—7.7.It suggests that at high temperatures H2S can be fairly stable up to neutral pH conditions.Thus,itprovides the theoreticalbasis for extrapolating our experimental resultsatambientconditions to thatof higher temperatures and pressures(Table 2).

Fig.3 Relationship between the initial pH values of metal-bearing solutions and the smallest volumes of NaHS solution needed to produceobservable precipitates.The lower the initialpH valuesof the metal-bearing solutions are,themore the titration volumes of NaHS solution to produce initial precipitates need

Given that NaHS is a kind of saltw ith strongly alkaline cation and weakly acidic anion(Oxtoby et al.2012;Petrucci and Harwood 1977),in an aqueous solution,its weakly acidic anion is easily hydrolyzed(Oxtoby et al.2012;Petrucci and Harwood 1977),as shown in the Eq.(1),resulting in the formation of a kind ofweak basic solution:

Meanwhile,the weakly acidic anion can also dissociate to form a kind of weak acidic solution.The dissociation reaction equation is as follows:

Directions of the Eq.s(1)and(2)are dependent on the acidity or basicity of hydrothermal f luids(Oxtoby et al.2012;Petrucci and Harwood 1977).The acidic environment forwards the Eq.(1),while the basic one is favorable for the Eq.(2).Previousstudieshave demonstrated that the MVT Pb—Zn ore depositswere usually formed in theacidic or neutral hydrothermal f luids(Banks et al.2002;Emsbo 2000;Grandia et al.2003;Leach et al.2005),suggesting the Eq.(1)as a dominant reaction to control the proportions of H2S and HS-during the formation of the MVT Pb—Zn ore deposits.In our f luid m ixing experiments,the low initialpH valuesofmetal-containing chloridesolutions are in favor of the progressof the Eq.(1).This can explain the f irst increase of the pH valuesofm ixed solutions in all of the f luid m ixing experiments(Fig.2),and is also in agreement that H2S is stable in the acidic—neutral f luids(Fig.6).

Temperature and pH can inf luence notonly the stability f ields of H2S and HS-in the hydrothermal f luids,butalso the distribution proportions of them.Under the acidic conditions and at room temperature,the proportion of H2S on reduced sulfur is close to 100%(Fig.7).This is the main reason why most studies supposed that H2S dom inates the precipitation of Pb and Zn during f luidsm ixing for the genesismodel of MVT Pb—Zn ore deposits(Anderson 1975;Beales and Jackson 1966;Corbella et al.2004;Leach etal.2005,2006;Reed 2006).However,our calculations show that although higher temperatures can make the stability f ield of HS-compressed into the basic domain(Fig.6),it can also promote the distribution of a proportion of HS-in the acidic hydrothermal f luids.As shown in Fig.7,at473 K and pH of～5.5 conditions the proportion of HS-has increased up to 10%,and while the pH isup to 6 the proportion of HS-ismore than 20%.This may not be negligible for the f luid mixing model for the genesis of the MVT Pb—Zn ore deposits.

4.2 Geochem ical pathways on metal precipitation during the f luid m ixing processes associated to the carbonate-hosted Pb-Zn deposits

In the f luid m ixing model related to the carbonate-hosted Pb—Zn deposits,metal precipitation and deposition is usually attributed to the mixing of a sulfur-def icient but metal chloride-rich f luid w ith another reduced sulfur-rich(i.e.H2Sand/or HS-)at the site of host carbonate(Anderson 1975;Beales 1975;Beales and Jackson 1966).There are,therefore,two kinds of main pathways to account for the m ixing(Fig.8),through(1)topographically-driven,reduced S-rich f luid continually f low ing into the site of host carbonate whichhas beenlargely percolated by metal chloride f luid,such as the Huize Pb—Znore deposits,southwesternChina(Zhanget al.2014b),or(2)Pb—Zn chloride ore f luid m igrating into the carbonate sequences f illed w ith local reduced S-richfluid,for instance,the Northern Arkansas MVT Pb—Zn deposits,North American(W ilkinson et al.2009),and the Pine Point Pb—Zn ore deposit,Canada(Beales and Jackson 1966).During the f luid mixing,geochemical

研究顯示：精神科護士一年內遭受工作場所暴力發生率為57.0%～99.0%[2,12-13]。本次調查顯示：我院精神科護士1年內遭受工作場所暴力發生率為75.3%。精神科護士遭受工作場所暴力高的原因為：①精神病病人大多存在思維、行為異常，受幻覺、妄想影響，隨時可能出現沖動傷人行為，而護士與病人接觸機會最多，遭受暴力危害最直接[14]；②在護士職業道德影響下，護士對病人暴力行為常采取忍讓態度，處于較為被動地位，不能有效控制暴力；③精神科保護性約束或隔離措施的使用與暴力發生密切相關[15]，護士在執行保護性約束和隔離措施時，經常遭受病人暴力。

reactions causing metal precipitation are considerably different.

Fig.4 EPMA spectra and photomicrograph of precipitates from f luid—rock reaction experiments.a Sample of Hh-2p;b sample of Hh-6p;c sample of Hh-10p.The results indicate that themain components of these precipitates are Pb,Zn,and S

Fig.5 XRD analysis spectra of precipitates from f luid—rock reaction experiments,demonstrating the grow th of sulf ide on the surface of dolom ite.a Sample of Hh-3 show ing the precipitates as sphalerite;b sample of Hh-7 show ing the precipitates as galena;c sample of Hh-10 show ing the precipitates as sphalerite and galena;d standard XRD spectrum of sphalerite and galena,data from http://rruff.info.M ineral abbreviation:Ga-galena,Sp-sphalerite

Fig.6 pH-log f(O2)phase diagram of S in the hydrothermal system at various temperatures.The complete calculationmethod is presented in Online Appendix A.H2S dominates for pH below 7,HS-between 7 and 13,S2-above 12 at T=298 K.The pH range for H2S stability was offset towards alkalinity by 1.0 unitsw ith the temperature changing from 298 to 473 K

Table 2 The results of EDS spectrum analysis on precipitates

Fig.7 Sulfur species fraction versus solution pH at T=298 K and 473 K,respectively,the curve of T=473 K modif ied from Reed(2006).Concentrationsof H2S,HS-,S2-as functionsof pH(Bjerrum plot)calculated from Online Appendix C.When T=298 K,at pH≤4,the sulfur in the solution mainly exists as H2S;at pH≥7,the sulfur in the solution ismainly presented as HS-;4＜pH＜7,H2S and HS-coexists.As the temperature increases,the dom inant f ield of H2Smoves to the right about one pH unit

Given thatmetal chloride complexes aremore stable in an acidic f luid and at high temperature(And et al.2003;Reed 2006;Seward 1984;Seward and Barnes 1997;Tagirov et al.2007a,b;Tagirov and Seward 2010),in the f irstmain pathway(Fig.8a),metal-chloride-rich ore f luids with low pH values percolate into the site of host carbonate,easily resulting in instabilities ofmetal chloride complexes due to Pb and Zn hydrolysis according to:

Dehydration of the hydrolyzed hydroxide products can take the form:

Equation(3)supplies a large amount of hydrogen ions to dissolve the host carbonate,resulting in massive alteration and even karsting(Pirajno 1992)and formation of secondary carbonate m inerals(M isra 2000).Because of exhaustion of hydrogen ions,the ore f luidsbecome closely neutral(Robb 2005);this causes Eq.(1)in the reverse direction and thus forming more HS-in the f luids,when reduced sulfur-rich f luids come to mix in the site of host carbonate.Considering thatsulf ideusually has the smallest solubility product constant(Reed 2006)and the solubility of galena and sphalerite in a NaCl system under various conditions is very low(less than 1 ppm)(Barrett and Anderson 1982,1988;Daskalakis and Helz 1993;Hayashi et al.1990;Hennig 1971;Melent'Yev et al.1969),those hydroxides and oxides of Pb and Zn in f luids or on the surface of carbonate can easily transform into sulf ides,according to:

The transformation from hydroxides and/or oxides to sulf ides could plausibly interpret the formation of sulf ide zonation in the carbonate-hosted Pb—Zn deposits,such as the Huize,Zhaotong and some other Pb—Zn ore deposits,southwestern China(Zhang et al.2014a,b).

Compared to the f irstmain pathway for the f luidmixing,the othermain pathway does not involve inmultiplemetal depositions(Fig.8b).During the inf lux of reduced sulfur-rich f luids into the carbonate sequences,the distribution of S species in ore f luids is controlled by Eq.(1)or:

Fig.8 Model diagrams on two kinds of f luidm ixing processes associated to carbonate-hosted MVT Pb—Zn deposits.Me represents the Pb and Zn metal ions

Due to the occurrence of the host carbonate,hydrogen ionsare consumed and consequently,the f luid isneutralor weakly alkaline(Anderson 1997).In thiskind of f luid,both H2S and HS-are the dominant species(Fig.7).Considering that Pb and Zn bisulf ide complexes are stable in low temperature,neutral to alkaline,lowsalinity solutions(Akinf iev and Tagirov 2014;Bourcier and Barnes 1987;Giordano and Barnes1979;Zhong etal.2015)in the form:

metal precipitation cannot happen by the reaction:

Whenmetal-chloride-rich ore f luidsare injected into the site of host carbonate and the neutralized orweakly alkaline S-rich f luids,Pb and Zn chloride complexes are unstable,some of which are transformed into bisulf ide complexes according to the Eq.(10),some react directly w ith H2S in the f luids:

For Eqs.(10)and(12),it is not likely that Pb and Zn ionshydrolyze because of suff icientsulfur occurring in the mixing f luids,which isdifferent from the f irstgeochemical pathway.Meanwhile,because H2S in the f luids are consumed during sulf ide deposition by Eq.(12),it leads to unstable Pb and Zn bisulf ide complexes and thus releases partial HS-to form H2S,till reaching the equilibrium amongthe sulfur species,metal ions,andbisulf ide complexes.

The true scenario on the f luidm ixing associated w ith the carbonate-hosted Pb—Zn deposits ismost likely attributed to successivegeochemicalprocesses involving notonly the f irstmain pathway but the second one.Itprobably depends on the paleogeographic and tectonic features on the deposition site(Anderson 1975;Corbella et al.2004;Leach et al.2005).However,while considering metal precipitation,the f irstpathway canmake themetal deposition reach itsmaximum size,probably form ing a series of larger size and higher-grade ore deposits than the second one.

4.3 Control factors on metal precipitation reaction during the f luid m ixing

It iswellknown thatdecrease in temperatureand change in the properties or compositionof the ore f luidcansignif icantlypromotemetalprecipitationduring hydrothermal processes(Fan et al.2001;Reed 2006;Seward and Barnes 1997).Asmentioned above,the MVTtype Pb—Zn deposits usually form atvery low pressure and temperature conditions(Banks and Russell 1992;Ganino and Arndt 2012;Grandia et al.2003;Leach et al.1996,2005;Marie and Kesler 2000;Savard et al.2000).Instabilities ofmetal complexes and thusmetal precipitation are seldom affected by temperature.Although the processes can change the properties or composition of the ore f luids such as increasing the sulfur concentration,f luid oxidation and m ixing w ith groundwater(Seward and Barnes 1997),they are diff icult to control during our f luid mixing experiments.

Interestingly,when the NaHS solution was titrated into the metal chloride solution the pH value of the m ixed solutionsbecame higher and no precipitateswere observed until the titration volume was large enough(Fig.2).This implies that the reactions between metal ions and sulfur species were indeed inf luenced by the environmental pH and the stability ofmetal complex.Given that HS-reacts directly w ithmetal ionsaccording to the Eq.(11),released hydrogen ions should promote the acidity of the f luids while precipitates are observed.On the contrary,the low pH primarily leads to sulfur species transformation from HS-to H2S according to the Eq.(1),releasing OH-to neutralize the f luids.W ith the neutralization of the f luids,metal chloride complexes are increasingly unstable(Reed 2006),which consequently makes bisulf ide complexation and sulf ide precipitation happen.Because Eq.(12)produce not only the sulf ide precipitations but also hydrogen ions,massivemetal precipitation is usually accompanied by the release of hydrogen ions.This distinctly interprets the decline of the pH values at the f inal stages of our f luid mixing experiments(Fig.2).

Another interesting observation is thatwhen the initial pH value of themetal chloride solution isapproximately 6 drops of NaHS solution titrating,which can immediately produce precipitates(Figs.2,3).It no doubt suggests that metal precipitation is much easier to happen at weakly acidic—neutral conditions.At these conditions,on the one hand,metal chloride complexes were less stable(Reed 2006),some of which were replaced by the bisulf ide complexes follow ing the Eq.(10);on the other hand,most of HS-are transformed into H2S according to the Eq.(1).Therefore,they causemost Pb and Zn ions react directly w ith H2S in the ore f luidsaccording to Eq.(12)to produce precipitates.

Given thataweakly acidic—neutral condition is likely to be themost favorable to form Pb—Zn ore deposits,the host carbonate thus plays a crucial role to adjust the environmental pH.In many MVT Pb—Zn ore deposits the ore bodies occur as cement among the carbonate breccia fragments(Anderson and Garven 1987;Sverjensky 1986).This suggests that the ore f luids m igrate into and completely react w ith the carbonate sequences,consequently resulting in their neutralization.During or after this process,released hydrogen ions dissolved the carbonate and thus lead to later precipitation of calcite and dolomite.Therefore,whatever the pathway for the f luidmixing is the f irstmodel or the second one,the environmental pH was neutralizing or neutralized before themetal precipitation.This key process caused instabilitiesof Pb and Zn chloride complexes and re-distribution of sulfur species,and thus facilitated the hydrolysis of Pb and Zn ions and precipitation of sulf ides,such as galena,sphalerite and so on.

Besides the above,the nature of low solubility product for sulf ide also makes Pb/Zn precipitation happen easily.Previousexperimentshave demonstrated that the solubility of galena and sphalerite under different temperature,NaCl concentration,and pH conditions was less than 1 ppm,even at temperatures of up to 300°C(Ewald and Hladky 1980;Barrettand Anderson 1988).Given that the solubility productsofPbSandZnSat25°CareKsp(PbS)=1.3×10-36and Ksp(ZnS)=1.6×10-24,the salteffectwasm inimaland can be ignored.Therefore,the solubility product of galena and sphalerite in pure water can be used to study the respective precipitation situation of the twometals.

In ourexperiments,since the NaHSsolutionwas titrated slow ly into a large amount of metal-containing solution,the common ion effects caused by the relative excess of metal ions should promote the formation of Pb/Zn precipitates at the f irst stage.However,because the solubility of the sulf ide precipitateswas very small,the coordination effect can be not considered,the acid effect is therefore signif icant during the whole experiments,as described in Sect.4.2.

5 Conclusion

On the basis of a series of m ixing experiments under ambient temperature and pressure conditions on titrating NaHSsolution intometal-chloride solutionsdopingwith or withoutdolomite,we found thatmetal precipitation during the f luid mixing was inf luenced by the stability of metal complex and the environmental pH.Becausemetal chloride complex isstable in an acidic f luidsandmetalbisulf ide complex is in favorofneutral-alkaline environments,sulfur species,the initial pH of ore f luidsand the environmentpH during metal precipitation govern the stabilities of metal complexes.Therefore,the environment pH was a primary factor controlling the Pb and/or Zn metal precipitation during the f luid m ixing associated w ith the carbonatehosted Pb—Zn ore deposits.

Thermodynamic calculations on the pH-logfO2and the pH-x diagram from 25 to 250°C show thatalthough higher temperatures can make the stability f ield of H2S expand from acidic toweakly alkaline domain and correspondingly compress the stability f ield of HS-from weakly acidic to weakly alkaline domain,it also can promote the distribution proportion of HS-in the acidic—neutral hydrothermal f luids.Combined w ith our f luid m ixing experimental results,we think that a weakly acidic—neutral condition is likely to be themost favorable condition to form Pb—Zn ore deposits,in which carbonate plays a crucial role in adjusting the environmental pH.Neutralization of the environmental pH driven by the carbonate-ore f luid interaction promote instabilities of Pb and Zn chloride complexes and re-distribution of sulfur species,and thus facilitates the hydrolysis of Pb and Zn ions and precipitation of sulf ides.During the f luid m ixing reactions H2S,rather than HS-or S2-in the solutions,dom inates the reactions of Pb and/or Zn precipitation.

Acknow ledgem entsWe thank two anonymous reviewers for their constructive comments.This work was supported jointly by the National Key R&D Program of China(No.2016YFC0600408),the National Natural Science Foundation of China(Nos.41572060,41773054,U1133602,41802089),ChinaPostdoctoral Science Foundation(No.2017M 610614),projects of YM Lab(2011)and Innovation Team of Yunnan Province and KMUST(2008 and 2012),and Yunnan and Kunm ing University of Science and Technology Postdoctoral Sustentation Fund.