ARTICLE IN PRESS

WAT E R R E S E A R C H

42 (2008) 661 668

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/watres

Coprecipitation of arsenate with iron(III) in aqueous sulfate media: Effect of time, lime as base and co-ions on arsenic retention
Yongfeng Jiaa, George P. Demopoulosb,
a b

Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110016, China Department of Mining, Metals and Materials Engineering, McGill University, Montreal, QC, Canada H3A 2B2

art i cle info

Article history: Received 30 May 2007 Received in revised form 20 August 2007 Accepted 21 August 2007 Available online 25 August 2007 Keywords: Coprecipitation Arsenate Iron

ab st rac t
The removal and immobilization of arsenic from industrial mineral-processing efuents typically involves lime neutralization and coprecipitation of arsenate with ferric iron. Despite the wide practice and environmental importance of this technique, no laboratory study has focused on the roles of lime as base and third ions like Ca2+, Ni2+ and SO2 on the 4 kinetics of arsenic retention by the coprecipitates. In this work, coprecipitation was performed at 22 1C by fast (10 min) neutralization of industrially relevant concentrated arsenateiron(III) (Fe/As 2, 4) acidic sulfate solutions to different pHs (4, 6, 8) in batch reactors, and the concentration of arsenic was monitored up to 1 year. The tests showed that maximum removal of arsenic was achieved upon neutralization to the target pH. Arsenic was found to be released back into solution from the precipitates upon continuing mild agitation at constant pH. Near-equilibrium was attained at different times depending on the applied pH: 10 days at pH 4, 6 months at pH 6 and 9 months at pH 8. An aging treatment at pH 4 signicantly enhanced arsenic retention (arsenic release was reduced by at least 50%) after the system was nally stabilized at pH 8. The retention of arsenic at pH 8 was multifold improved (by a factor 25) when lime was used instead of NaOH. Similarly, the retention of arsenic was enhanced by the presence of calcium and nickel ions in the starting solution. Finally, evidence of Ca(II)Fe(III)As(V) association was found, but not sulfate incorporation at pH 8. & 2007 Elsevier Ltd. All rights reserved.

1.

Introduction

Arsenic present in mineral-processing solutions and efuents is a major industrial contaminant. It is derived during extraction of metals by oxidation and acid dissolution of the arsenic-containing minerals such as arsenopyrite (FeAsS) and members of tetrahedritetennantite series ((Cu,Fe)12Sb4S13 (Cu,Fe)12As4S13) in many base-metal and gold ores or Ni CoFe arsenides and sulfarsenides in many uranium ores (Jambor and Dutrizac, 1995). Recent reviews have summarized the methods in industrial practice employed to remove
Corresponding author. Tel.: +1 514 398 2046; fax: +1 514 398 4492.

arsenic liberated during various mineral-processing and metallurgical operations (Harris and Krause, 1993; Twidwell et al., 1999; Harris, 2000; Riveros et al., 2001). Among the available technologies, lime neutralization accompanied by coprecipitation of arsenic with ferric iron is the industrial method of choice, when it comes to the removal of arsenic from acidic mineral-processing efuents (Harris, 2003; Twidwell et al., 2005). The bulk of arsenic in mineralprocessing solutions is generally present as arsenate, i.e. As(V). Arsenate is efciently removed from the solution to sub-ppm level, when excess ferric iron (Fe(III)/As(V) molar

E-mail addresses: yongfeng.jia@iae.ac.cn (Y. Jia), george.demopoulos@mcgill.ca (G.P. Demopoulos). 0043-1354/$ - see front matter & 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.watres.2007.08.017

ARTICLE IN PRESS
662
WA T E R R E S E A R C H

42 (2008) 661 668

ratio X3) is present (Harris, 2000). The generated tailings, consisting of Fe(III)As(V) coprecipitated and gypsum solids (produced as a result of the sulfate-based nature of these solutions and the use of lime), are disposed off in specially designed waste storage ponds (Langmuir et al., 1999). Hence, our interest in understanding the fundamental processes and behavior of arsenic retention in this environmentally important Fe(III)As(V)Ca(II)S(VI) system is the focus of this work. There have been extensive studies on the coprecipitation of arsenate with iron(III) regarding the process, the solubility and stability of the arsenic-bearing coprecipitate, the speciation of arsenic and the parameters inuencing these properties (Harris and Monette, 1988; Robins et al., 1988; Krause and Ettel, 1989; Emett and Khoe, 1993; Waychunas et al., 1993; Langmuir et al. 1999, 2006; Nishimura and Umetsu, 2000; Moldovan et al., 2003; Jia et al., 2003; Richmond et al., 2004; Twidwell et al., 2005; Moldovan and Hendry, 2005). The solubility of the arsenate-bearing precipitates was found to decrease with increasing Fe/As molar ratio (Robins et al., 1988; Krause and Ettel, 1989; Nishimura and Umetsu, 2000). The optimal pH for the removal of arsenate lies in a mildly acidic region (i.e. 3.55.5) (Robins et al., 1988; Krause and Ettel, 1989; Nishimura and Umetsu, 2000) after which the concentration of arsenic in solution increases with increasing pH. However, in hydrometallurgical industry, the acidic arsenic-bearing efuents are generally neutralized to a mildly alkaline region (e.g. pH 79) (Langmuir et al., 1999; Moldovan et al., 2003; Moldovan and Hendry, 2005) in order to remove co-existent metal cations (e.g. Co2+, Ni2+, Cu2+, etc.) as metal hydroxide precipitates. The presence of cations in solution, such as Ca2+, Cd2+, Cu2+, Mg2+, Pb2+, Zn2+, has been found to increase the pH range from 47 to 49 for effective arsenic removal (Harris and Monette, 1988; Emett and Khoe, 1993). The main view regarding arsenate speciation in coprecipitates is that arsenate is adsorbed via surface complex formation on ferrihydrite (Robins et al., 1988; Waychunas et al., 1993; Moldovan et al., 2003; Richmond et al., 2004; Moldovan and Hendry, 2005). The overwhelming mineralogical evidence in support of this view is based on coprecipitation studies involving either low arsenic concentrations (typically o1 mM As (Waychunas et al., 1993)) or elevated pH 47 (e.g. pH 8 (Waychunas et al., 1993) or pH 9.8 (Moldovan et al., 2003)). Recent coprecipitation studies involving high As concentrations and the acidic pH range have suggested the formation of ferric arsenate (i.e. Fe/As$1 compound) in the coprecipitates as well (Krause and Ettel, 1989; Langmuir et al., 1999, 2006; Jia et al., 2003). Most of the previous laboratory coprecipitation studies have used NaOH as base for the neutralization of acidic As(V) Fe(III) solutions. In industry, however, lime (CaO or Ca(OH)2) is used as base instead to neutralize acidic arsenateiron(III) solutions or efuents. Hence the porewaters in most industrially generated mineral-processing tailings are gypsumsaturated because mineral-processing efuents are generally sulfate-based due to the use of sulfuric acid as leaching agent or addition of ferric sulfate as coprecipitating agent. The nal solid product is a mixture of gypsum and arsenateiron(III) precipitate (Riveros et al., 2001). It has been speculated that gypsum in the ironarsenate residues may increase their long-term stability (Harris and Krause, 1993). Calcium

arsenate forms only if the Fe/As molar ratio in the solution is low (Riveros et al., 2001). The role of lime (or Ca2+) in arsenic retention during and after the coprecipitation of arsenic(V) with ferric iron from sulfate media has not been studied previously, nor the role of other major cations like Ni2+ present in high concentration in industrial hydrometallurgical solutions (Langmuir et al., 1999; Moldovan and Hendry, 2005). The objectives of this work are (i) to study the kinetics of arsenic retention following the coprecipitation of arsenate with iron(III) at different pHs (4, 6, 8) and (ii) to investigate the effect of lime as opposed to sodium hydroxide as base as well as calcium and nickel ions in solution on the stability of the coprecipitated arsenate. The study emphasizes high arsenic concentrations and ferric sulfate solutions relevant to the treatment of industrial hydrometallurgical process efuents as they are some uranium mill rafnates ranging from 6.67 to 86.67 mM As (Langmuir et al., 1999; Moldovan and Hendry, 2005).

2.

Materials and methods

The acidic arsenateiron(III) sulfate solutions for the various coprecipitation tests were prepared by mixing separately prepared arsenate and ferric sulfate stock solutions. The arsenate stock solution (0.13 M) was prepared by dissolving Na2HAsO4 7H2O in de-ionized water. The pH of the arsenate solution, initially at $8.6, was adjusted with HNO3 to $1.5 before mixing with the iron(III) solution. The iron(III) solution was prepared by dissolving Fe2(SO4)3 5H2O in de-ionized water. Depending on the desired Fe/As molar ratio (2 or 4), 140 or 70 mL of arsenate stock solution was added to the ferric sulfate solution to yield ultimately upon neutralizationsee the next paragraph500 mL slurry with $0.075 M iron concentration. The starting arsenateiron(III) solution was adjusted initially with HNO3 to pH $1.5, at which no precipitation occurred. Coprecipitation took place subsequently by raising the pH of the solution to 4, 6 or 8 via fast addition (in $10 min) of either 1 N NaOH (from burette) or 4% slaked lime (using a pipette), while the mixture was mechanically stirred vigorously. Approximately 105115 mL of 1 N NaOH or slaked lime was consumed to neutralize the acidic solutions to pH 48. The slaked lime was prepared by stirring CaO powder in deionized water for 5 h. The total volume of the neutralized solution (following the attainment of the target pH) was 500 mL with the ferric ion concentration of 0.075 M and sulfate concentration of 0.113 M for all tests. The effect of calcium ions on the retention of arsenic was assessed for the Fe/As 4 coprecipitation system at pH 8. Calcium was added ($300 mL) in the form of clear saturated CaSO4 solution (prepared by gypsum dissolution) to acidic arsenateiron(III) solution and the pH of the mixture was raised to 8 using 1 N NaOH. Once more the volume of the neutralized slurry was 500 mL. A precipitation test in the absence of arsenate was also conducted to compare the fate of calcium ions for the systems with and without arsenate. No gypsum formed on the addition of the CaSO4 solution (equilibrium $4 mM Ca). The effect of nickel ions on the

ARTICLE IN PRESS
WAT E R R E S E A R C H

42 (2008) 661 668

663

retention of arsenic was assessed by adding NiSO4 6H2O to the starting arsenateiron(III) solution to yield the molar ratio of Ni/As 1. This experiment was performed in conjunction with the Fe/As 4 coprecipitation system at pH 8. For all coprecipitation tests, following neutralization, the produced slurry ($500 mL) was transferred to capped conical asks subjected to mild magnetic stirring and the aqueous arsenic concentration was monitored for up to 1 year. During this period, the pH was maintained constant at the target value with regular (on an hourly basis initially, followed by daily and weekly basis) addition of 0.1 N or 1 N NaOH. In total (over the 1-year duration of each test) 510 mL of NaOH solution was used to stabilize the pH, which represents less than 2% dilution; hence, the dilution effect was not considered in the concentration data. At various time intervals 2-mL aliquots of slurry were taken for analysis using a 10-mL syringe and ltered using 0.2-mm syringe lters. The coprecipitate collected in the syringe lter was washed with 30 mL de-ionized water of the same pH as coprecipitation and digested with 2 N HCl. The concentration of arsenic, iron, sulfur, calcium and nickel was determined by ICP-AES analysis. The powder XRD patterns were obtained on a Philips PW1710 diffractometer equipped with a copper target (CuKa1 radiation, l 1.54060 A), a crystal graphite monochromator and a scintillation detector. The equipment was run at 40 kV and 20 mA by step-scanning from 101 to 1001 2y with increments of 0.11 2y and a counting time of 0.3 s at each step.

100 10 1 pH 8 0.1 0.01 1E-3 1E-4 10 1 0.1 pH 6 0.01 1E-3 1E-4 0 50 100 150 200 250 300 350 400 Time (day) pH 4 Fe/As = 2 pH 8 neutralization to target pH in 10 min Fe/As = 4

3.

Results and discussion

Fig. 1 Concentration of arsenic in solution as a function of time during and after neutralization of As(V)Fe(III) solution to different pH by NaOH.

The fates of arsenate and (where appropriate) sulfate, calcium and nickel in terms of their concentrations in solution or solid as a function of time were monitored and reported in this study. The concentration of iron in solution was less than 0.009 mM for all coprecipitation tests in this work; hence, its fate is not reported. The Fe/As molar ratio in all coprecipitates was close to the nominal ratio of the starting solution (i.e. Fe/As 2 or 4). First the kinetics of arsenic retention is considered when NaOH is used as base, followed by investigation of the roles of lime, calcium and nickel ions and the behavior of sulfate.

3.1.

Neutralization by NaOH

Fig. 1 shows the arsenic concentration in solution as a function of time during and after neutralization of the arsenateiron(III) solutions to pH 4, 6 and 8, respectively, by NaOH. The removal of arsenate from solutions was fast. For the Fe/As 2 system, the arsenic concentration decreased from initially 37.33 mM to o0.0003, 0.002 and 0.02 mM after neutralization to pH 4, 6 and 8, respectively in 10 min. Surprisingly, arsenic was released back into solution after neutralization to target pH (i.e. 4, 6, 8), taking considerable time to reach steady concentration values. This is contrary to conventional understanding that arsenic removal increases with reaction time. Equilibrium was attained at different times depending on the pH of the slurry. Thus, maximum arsenic release occurred in the rst 10 days at pH 4, with its

concentration reaching $0.002 mM and remaining nearly unchanged thereafter. More arsenic was released after neutralization to higher pH. The arsenic concentration of the pH 6 coprecipitation system rose from the lowest point of 0.0016 mM to 0.024 mM in 6 months and remained stable thereafter. For the pH 8 system, the arsenic concentration increased to $1.6 mM in 5 months, after which only an additional $0.07 mM of arsenic was released in the following 4 months. For the Fe/As 4 coprecipitation systems, similarly, arsenic was removed rapidly (within 10 min) from initially 18.67 mM at pH 1.5 to o0.0003 mM after neutralization to pH 6 and 8. The release of arsenic at pH 8 was 100 times greater than that at pH 6. After neutralization to pH 8, the arsenic concentration tended to stabilize at $0.41 mM in 9 months. In some previous coprecipitation studies, the concentration of arsenic was measured after the mixture was stirred for limited time, e.g. 1 h (Nishimura and Umetsu, 2000), 2 h (Richmond et al., 2004), 39 h (Moldovan and Hendry, 2005), 2496 h or 67 days (Robins et al., 1988). This is reasonable for the study of arsenic removal characteristics by coprecipitation. But if treated as arsenic solubility, the reported data are apparently underestimated because after neutralization arsenic concentration rises over time as per the ndings of the present work. The observed release of arsenic after neutralization is most likely related to the speciation of arsenic in the coprecipitated solids. Some preliminary XRD analysis and a comprehensive EXAFS study showed that poorly crystalline ferric arsenate

As concentration (mM)

pH 6

ARTICLE IN PRESS
664
WA T E R R E S E A R C H

42 (2008) 661 668

was present in the coprecipitate formed at acidic pH from concentrated arsenateferric sulfate solutions, as are the ones used in this work (Jia et al., 2003; Chen et al., submitted for publication). Moreover, poorly crystalline ferric arsenate was found to form on the surface of schwertmannite (Carlson et al., 2002) or ferrihydrite at acidic pH (i.e. pH 3, 4) (Jia et al., 2006, 2007). The solubility of both poorly crystalline ferric arsenate and surface-complexed arsenate on ferrihydrite is known to increase with increasing pH (Robins et al., 1988; Krause and Ettel, 1989; Raven et al., 1998; Nishimura and Umetsu, 2000; Jia and Demopoulos, 2005). Therefore, irrespective of arsenic speciation in the coprecipitates, some of the previously removed arsenate at lower pH may be expected to be partially released back into solution after neutralization to target pH as it was observed in this work. Since the coprecipitation systems showed lower arsenic solubility at the pH 46 region, it was thought that a two-stage equilibration process may render higher stability than a single-stage one. To investigate this possibility, the Fe/As 2 coprecipitation system was previously aged (for a period of 6 months) at pH 4 before it was nally adjusted to pH 8. Fig. 2 compares the arsenic release over time of the two-stage neutralization process with the single-stage one. The arsenic concentration increased after the previously pH-4-aged coprecipitation system was raised to pH 8 and reached 0.73 mM in 10 months. This level is signicantly lower than the value of the corresponding 1-stage neutralized system ($1.67 mM As in $10 months). This indicates that arsenic release was markedly suppressed if the coprecipitation system had previously been aged at lower pH. This observation can have signicant implications for the disposal of industrial neutralization tailings. Thus, industrial operations that are known to have adopted staged neutralization processes, e.g. two-stage neutralization process at JEM mill (rst stage at pH 4 and second stage at pH 8) (Rowson private

communication) and three-stage neutralization process at Rabbit Lake uranium mine (rst stage at pH 3.5, second stage at pH 6.5 and third stage at pH 8.5) (Moldovan et al., 2003) according to the ndings of this work should lead to enhanced coprecipitate stability. It is postulated that, due to the formation of poorly crystalline ferric arsenate phase (i.e. Fe/As 1) at acidic pH (Jia et al., 2003), the Fe/As ratio of the remaining arsenateferrihydrite phase in the coprecipitate is much higher than the nominal ratio. This leads to the reduced arsenic concentration for the multi-stage neutralized systems compared with the single-stage neutralization to pH 8.

3.2.

Neutralization by CaO

The effect of lime and calcium ions on the kinetics of arsenic retention by the coprecipitate is considered in this part of the study. Fig. 3 shows the effect of CaO and added calcium ions on the concentration of arsenic in solution after neutralization of As(V)Fe(III) sulfate solutions to pH 8. It is apparent that the CaO-neutralized system equilibrated with much lower arsenic concentration. The lowest point after neutralization of the Fe/As 2 system with CaO was o0.0003 mM compared with 0.02 mM of the NaOH-neutralized system. Similar to the NaOH-neutralized system, arsenic was also released back into solution when CaO was used as base but at much lower level. For example, the arsenic concentration increased from o0.0003 to $0.067 mM after 8 months for the Fe/As 2 system neutralized with lime. This is 25 times lower than that of the corresponding NaOH-neutralized system, indicating that the use of CaO as base suppressed arsenic release.

100 10 1
neutralization to target pH in 10 min

2.0

NaOH NaOH with Ca CaO

2+

0.1
1-stage

As concentration (mM)

0.01 1E-3 1E-4 10

As concentration (mM)

1.5

Fe/As = 4

1.0 2-stage 0.5

NaOH

1 0.1
CaO

0.0 0 50 100 150 200 250 Time (day) 300 350 400

0.01 1E-3 1E-4 0 50 100 150 200 250 300 350 400 Time (day)
Fig. 3 Effect of CaO as base and calcium ions on arsenic release after neutralization of As(V)Fe(III) solutions to pH 8.
Fe/As = 2

Fig. 2 Comparison of single-stage with two-stage neutralization in terms of arsenic release from coprecipitate (Fe/As 2; single-stage: neutralization to pH 8; two-stage: neutralization to pH 4 and aged for 6 months, then raised to pH 8).

ARTICLE IN PRESS
WAT E R R E S E A R C H

42 (2008) 661 668

665

To clarify if the observed benecial effect was derived from calcium, a test was done with NaOH in which calcium had been added as clear CaSO4 solution (refer to the Fe/As 4 data in Fig. 3). Again, the release of arsenic was strongly suppressed in both cases (i.e. with CaO and with NaOH/ CaSO4). The stronger effect of lime is attributed to the higher concentration of calcium in solution, i.e. $16 mM for the CaOneutralized systems, $4 mM for the NaOH/Ca(II) system. Previous studies have also noted the benecial effect of calcium on the stability of arsenic-bearing coprecipitates (Harris and Monette, 1988; Emett and Khoe, 1993) and arsenateferrihydrite adsorption systems (Wilkie and Hering, 1996; Jia and Demopoulos, 2005; Masue et al., 2007). Fig. 4 compares the calcium content in the solids precipitated from As(V)Fe(III)Ca(II) and Fe(III)Ca(II) solutions neutralized with NaOH to pH 8. There was an appreciable uptake of calcium by the precipitates in both cases. No gypsum crystallized in the two systems due to the low Ca(II) ion concentration involved ($4 mM). This was conrmed by XRD analysis. Calcium ions can be adsorbed to an appreciable extent by ferrihydrite (Dzombak and Morel, 1990; Jambor and Dutrizac, 1998). The calcium content in the solid precipitated from As(V)Fe(III)Ca(II) solution is $1.8 times higher than that in ferrihydrite. This suggests that calcium ions have been involved in direct association with arsenate in the precipitate in addition to adsorption by ferrihydrite. Bulk precipitation of calcium arsenate phases is unlikely to have occurred under the conditions of this work. For the Ca(II)As(V)H2O system, the precipitated phase was Ca3 (AsO4)2 xH2O at $pH 8, with an equilibrium concentration of 8.8 mM As and 8 mM Ca in solution (Nishimura and Robins, 1998). The Ksp of this compound is three orders of magnitude higher than the IAP for the coprecipitation systems in the present study, indicating that the solution was undersaturated with respect to calcium arsenate. However, the formation of calcium arsenate surface precipitate cannot be ruled out. It was reported previously that surface precipitation of

ferric phosphate on goethite (Ler and Stanforth, 2003), ferric arsenate on ferrihydrite (Jia et al., 2006, 2007) and zinc arsenate on goethite (Grafe et al., 2004) occurred under apparently undersaturated conditions. Additionally, it is proposed here that some type of Ca(II)Fe(III)As(V) association occurred in the solid precipitated from As(V)Fe(III) solution in the presence of calcium ions, added either as CaO or as CaSO4 solution that stabilized arsenic. To conrm this hypothesis, the CaO-neutralized Fe/As 2 coprecipitate slurry (obtained after 24 h equilibration) was subjected to an accelerated aging test at pH 7.58.0, 75 1C and 7 weeks to force the conversion of amorphous phases to XRD-recognizable crystalline forms. After washing off with water the crystallized gypsum from the aged precipitate, XRD analysis revealed the presence of crystalline yukonite (Ca2Fe3(AsO4)4(OH) 12H2O) (see Fig. 5). Crystalline yukonite was also observed to form during dissolution of scorodite (FeAsO4 2H2O) at 75 1C, pH 7 in the presence of gypsum (Demopoulos, 2005). The formation of a calcium ironarsenate phase had been hypothesized previously as a possible cause to explain the increased uptake of arsenate by ferrihydrite in the presence of calcium (Harris and Monette, 1988) without presenting mineralogical evidence. Mineralogical evidence of the presence of calciumironarsenate phases (like the yukonite phase detected here) was found though in industrial gold mine tailings (Paktunc et al., 2003).

3.3.

Coprecipitation in the presence of Ni2+

Fig. 6 shows the effect of nickel on arsenic retention during and after neutralization of the Fe/As 4 system to pH 8. NaOH was used as base in order to simplify the system and clarify the role of nickel. Nickel is also compared with calcium in Fig. 6 in terms of their effects on arsenic release after coprecipitation. The removal of arsenic was improved in the presence of nickel, as reected by the lowest point of concentration reached upon neutralization to pH 8. Similar

Time (h) 0 10 Ca concentration in solution (mM) 20 40 60 80 100 120 0.14 0.12 8 0.10 6 0.08 0.06 4 0.04 0.02 2 0 20 40 60 80 100 120 0.00 140 Ca concentration in solid (mol-Ca/mol-Fe)

100

Reference: yukonite (PDF 45-1358) Ca2Fe3(AsO4)4(OH).12H2O

80
Relative intensity

60

40

20

0 0 20 40 60 80 100

Time (h)

Fig. 4 Comparison of aqueous Ca concentration and solid Ca/Fe molar ratio of As(V)Fe(III)-Ca(II) system (Fe/As 4) with Fe(III)Ca(II) system after neutralization to pH 8 with NaOH.

2
Fig. 5 XRD patterns of the product from 7-week accelerated ageing at 70 1C of CaO- neutralized Fe/As 2 coprecipitate at pH 8.

ARTICLE IN PRESS
666
WA T E R R E S E A R C H

42 (2008) 661 668

100 10 As concentration (mM) 1 0.1 0.01 1E-3 1E-4 0 50 100 150 200 250 Time (day)
neutralization to target pH in 10 min
As(V)-Fe(III) As(V)-Fe(III)-Ni(II) As(V)-Fe(III)-Ni(II) As(V)-Fe(III)-Ca(II)

100

0.40 0.35
As(V)-Fe(III) As(V)-Fe(III)-Ni(II) As(V)-Fe(III)-Ca(II)

10 Ni concentration (mM)

0.30 0.25 0.20 0.15 Sulfate content in solid (mol-SO4 /mol-Fe) 0.10 0.05 0.00 1 2 3 4 pH 5 6 7
neutralization to pH 8

As concentration

0.1
Ni concentration

0.01 300

8 0

20

40

60

80 100

Time (day)

Fig. 6 Concentration of arsenic and nickel in solution as a function of time during and after neutralization to pH 8 by NaOH (Fe/As 4, nickel added as NiSO4 (Ni/As 1), calcium added as CaSO4(Ca/As 0.5)).

0.20

pH 4 pH 8 pH 8 (previously aged at pH 4)

0.15

to calcium, nickel also suppressed arsenic release, while at the same time it was removed from the solution. The concentration of nickel decreased from initially $18.74 to $0.005 mM after 4 months retention, indicating that all of the added nickel has been incorporated into the coprecipitate. The nickel content in the coprecipitate was Ni/Fe $0.25, which was greater than the calcium content of Ca/Fe $0.13. However, arsenic was released to a similar extent for both cases, suggesting that calcium was advantageous over nickel in stabilizing the arsenic-bearing coprecipitate. In a previous study, the solubility of annabergite (Ni3 (AsO4)2 8H2O) was determined to be $0.10 mM Ni and $0.64 mM As at pH 8 (Yuan et al., 2005). Hence, the present coprecipitation system ($0.03 mM Ni, $0.13 mM As) was undersaturated with respect to annabergite, thus bulk precipitation of annabergite was unlikely to have occurred. However, the formation of nickel arsenate surface precipitate on ferrihydrite remains a possibility since surface precipitation may occur at concentrations below the bulk saturation state (Ler and Stanforth, 2003). Additionally, the enhanced removal of arsenic (and nickel) in the Ni(II)Fe(III)As(V) system may be due to adsorption on coprecipitated nickelferric hydroxide.

0.10

0.05

0.00 0 50 100 Time (day) 150 200

Fig. 7 Sulfate contents in the coprecipitates neutralized by NaOH: (top) as a function of pH during neutralization and as a function of time after neutralization of Fe/As 4 solutions; (bottom) as a function of time after neutralization of Fe/As 2 solutions.

3.4.

Deportment of sulfate

Fig. 7 shows the sulfate contents in the coprecipitated solids as a function of pH and time. For the Fe/As 2 coprecipitation systems (bottom part of the gure), negligible amount of sulfate was incorporated into coprecipitate at pH 8. In comparison, the incorporation of sulfate was signicant at pH 4. Upon neutralization to pH 4, the sulfate content was SO4/Fe$0.2. The sulfate content decreased gradually over time and reached a relatively stable level of SO4/Fe$0.1 after 25 days. However, this was signicantly higher than that of the corresponding adsorption system (SO4/Fe$0.01 for the Fe/As 2, pH 4 adsorption system), indicating a different mechanism than simple adsorption mode of retention.

Nevertheless, upon elevation to pH 8, all incorporated sulfate (previously aged at pH 4) was released back into solution after about 2 weeks. The sulfate content of the Fe/As 4 coprecipitation solids was also monitored as a function of pH and time as illustrated in Fig. 7 (top part of the gure). In the process of neutralization (5 min per pH pointin total 35 min neutralization time) of arsenateiron(III) solution from pH 1.5 to 8, the molar ratio of SO4/Fe decreased sharply with increasing pH. The presence of calcium ions in solution (4 mM) did not help the incorporation of sulfate. In comparison, the presence of nickel enhanced initial uptake of sulfate by the precipitate. In this case, the sulfate content was SO4/Fe$0.25 upon neutralization to pH 8, but decreased gradually over time and approached 0 after 85 days. The results suggest that sulfate can only be retained by the coprecipitate in the acidic region. When the acidic solution is neutralized to pH 8, all previously incorporated sulfate is released from the solid. The reason for the release of sulfate is unknown. This may be due to either the formation of sulfate-containing

ARTICLE IN PRESS
WAT E R R E S E A R C H

42 (2008) 661 668

667

minerals at lower pH that break down with increasing pH, or simply adsorptiondesorption behavior as a function of pH.

R E F E R E N C E S

3.5.

Environmental implications

The use of lime as base as opposed to sodium hydroxide to neutralize acidic mineral-processing solutions for the removal and immobilization of arsenic enhances arsenic stability. The risk of arsenic release from the coprecipitates is reduced if staged neutralization is practiced or if nickel and calcium ions as per the ndings of this work are present. Lime-neutralized coprecipitates probably contain some type of Ca(II)Fe(III)As(V) association, which may convert to calcium ferric arsenate (e.g. yukonite) upon aging.

4.

Conclusions

Coprecipitation of arsenate with ferric iron was conducted in batch reactors by neutralizing the acidic sulfate solutions to pH 48. The effect of slaked lime as base as well as the presence of nickel and calcium ions on the retention of arsenic by the coprecipitated solids has been investigated by monitoring the concentration of arsenic in the solutions over the period of a year. The major ndings of the study are the following:

(1) The removal of arsenic from solution was very fast during the coprecipitation process. Maximum removal of arsenic was achieved after the system was neutralized to the target pH. However, arsenic was released back into solution from the coprecipitates over time and equilibrium was attained at different equilibration time (from 10 days to 9 months) depending on the pH and Fe/As molar ratio. (2) The use of lime instead of sodium hydroxide as base to neutralize the acidic arsenic-bearing solutions signicantly enhanced the stability of the coprecipitates. Similarly the presence of nickel and calcium ions also strongly inhibited the release of arsenic from the coprecipitates. Evidence was obtained for the association of calciumironarsenate in the coprecipitates, which was converted to crystalline yukonite at 75 1C. (3) Sulfate was incorporated into the coprecipitates at acidic pH, but was released completely when the pH was raised to 8.

Acknowledgments
The authors thank NSERC (Canada) for the support of this work through a Strategic Project Grant. This research was sponsored by Areva Resources, Barrick Gold Corporation, and Hatch Ltd. Yongfeng Jia also thanks the National Natural Science Foundation of China (NSFC, no. 40673079) for their support to this work.

Carlson, L., Bigham, J.M., Schwertmann, U., Kyek, A., Wagner, F., 2002. Scavenging of As from mine drainage by schwertmannite and ferrihydrite: a comparison with synthetic analogues. Environ. Sci. Technol. 36, 17121719. Chen, N., Jiang, D.T., Cutler, J., Kotzer, T., Jia, Y.F., Demopoulos, G.P., Rowson, J.W., Structural characterization of amorphous scorodite, iron(iii)-arsenate co-precipitates and uranium mill neutralized rafnate solids using X-ray absorption ne structure spectroscopy. Geochim. Cosmochim. Acta (submitted). Demopoulos, G.P., 2005. On the preparation and stability of scorodite. In: Reddy, R.G., Ramachandran, V. (Eds.), Arsenic Metallurgy. TMS, Warrendale, PA, pp. 2550. Dzombak, D.A., Morel, F.M.M., 1990. Surface complexation modeling: hydrous ferric oxide. Wiley-Interscience, New York, pp. 173178. Emett, M.T., Khoe, G.H., 1993. Environmental stability of arsenic bearing hydrous iron oxide. In: Warren, G. (Ed.), EPD Congress 1994. TMS, Warrendale, PA, pp. 153166. Grafe, M., Nachtegaal, M., Sparks, D.L., 2004. Formation of metal arsenate precipitates at the goethitewater interface. Environ. Sci. Technol. 38, 65616570. Harris, G.B., Krause, E., 1993. The disposal of arsenic from metallurgical processes: its status regarding ferric arsenate. In: Reddy, R.G., Weizenbach, R.N. (Eds.), Extractive Metallurgy of Copper, Nickel and Cobalt, vol. I. TMS, Warrendale, PA, pp. 12211237. Harris, G.B., Monette, S., 1988. The stability of arsenic-bearing residues. In: Reddy, R.G., Hendrix, J.L., Queneau, P.B. (Eds.), Arsenic Metallurgy: Fundamentals and Applications. TMS, Warrendale, PA, pp. 469498. Harris, G.B., 2000. The removal and stabilization of arsenic from aqueous process solutions: past, present and future. In: Young, C.A. (Ed.), Minor Elements 2000. SME, Littleton, CO, pp. 320. Harris, B., 2003. The removal of arsenic from process solutions: theory and industrial practice. In: Young, C., Alfantazi, A., Anderson, C., James, A., Dreisinger, D., Harris, B. (Eds.), Hydrometallurgy 2003Proceedings of the International Symposium Honoring Professor Ian M. Ritchie, vol. 2. TMS, Warrendale, PA, pp. 18891902. Jambor, J.L., Dutrizac, J.E., 1995. Solid solutions in the annabergiteerythritehornesite synthetic system. Can. Mineral. 33, 10631071. Jambor, J.L., Dutrizac, J.E., 1998. Occurrence and constitution of natural and synthetic ferrihydrite, a widespread iron oxyhydroxide. Chem. Rev. 98, 25492585. Jia, Y.F., Demopoulos, G.P., 2005. Adsorption of arsenate onto ferrihydrite from aqueous solution: inuence of media (sulfate vs. nitrate), added gypsum, and pH alteration. Environ. Sci. Technol. 39, 95239527. Jia, Y.F., Demopoulos, G.P., Chen, N., Cutler, J.N., Jiang, D.-T., 2003. Preparation, characterization and solubilities of adsorbed and co-precipitated iron(iii)arsenate solids. In: Young, C., Alfantazi, A., Anderson, C., James, A., Dreisinger, D., Harris, B. (Eds.), Hydrometallurgy 2003Proceedings of the International Symposium Honoring Professor Ian M. Ritchie, vol. 2. TMS, Warrendale, PA, pp. 19231935. Jia, Y.F., Xu, L., Fang, Z., Demopoulos, G.P., 2006. Observation of surface precipitation of arsenate on ferrihydrite. Environ. Sci. Technol. 40, 32483253. Jia, Y.F., Xu, L., Wang, X., Demopoulos, G.P., 2007. Infrared spectroscopic and X-ray diffraction characterization of the nature of adsorbed arsenate on ferrihydrite. Geochim. Cosmochim. Acta 71, 16431654. Krause, E., Ettel, V.A., 1989. Solubilities and stabilities of ferric arsenates compounds. Hydrometallurgy 22, 311337.

ARTICLE IN PRESS
668
WA T E R R E S E A R C H

42 (2008) 661 668

Langmuir, D., Mahoney, J., MacDonald, A., Rowson, J., 1999. Predicting arsenic concentrations in the porewaters of buried uranium mill tailings. Geochim. Cosmochim. Acta 63, 33793394. Langmuir, D., Mahoney, J., Rowson, J., 2006. Solubility products of amorphous ferric arsenate and crystalline scorodite (FeAsO4 2H2O) and their application to arsenic behavior in buried mine tailings. Geochim. Cosmochim. Acta 70, 29422956. Ler, A., Stanforth, R., 2003. Evidence of surface precipitation of phosphate on goethite. Environ. Sci. Technol. 37, 26942700. Masue, Y., Loeppert, R., Kramer, T., 2007. Arsenate and arsenite adsorption and desorption behavior on coprecipitated aluminum: iron hydroxides. Environ. Sci. Technol. 41, 837842. Moldovan, B.J., Hendry, M.J., 2005. Characterizing and quantifying controls on arsenic solubility over a pH range of 111 in a uranium mill-scale experiment. Environ. Sci. Technol. 39, 49134920. Moldovan, B.J., Jiang, D.T., Hendry, M.J., 2003. Mineralogical characterization of arsenic in uranium mine tailings precipitated from iron-rich hydrometallurgical solutions. Environ. Sci. Technol. 37, 873879. Nishimura, T., Robins, R.G., 1998. An re-evaluation of the solubility and stability regions of calcium arsenites and calcium arsenates in aqueous solution at 25 1C. Miner. Process. Extr. Metall Rev. 18, 283308. Nishimura, T., Umetsu, Y., 2000. Chemistry on elimination of arsenic, antimony and selenium from aqueous solution with iron(III) species. In: Young, C.A. (Ed.), Minor Elements 2000. SME, Littleton, CO, pp. 105112. Paktunc, D., Foster, A., Laamme, G., 2003. Speciation and characterization of arsenic in Ketza River mine tailings using X-ray absorption spectroscopy. Environ. Sci. Technol. 37, 20672074.

Raven, K.P., Jain, A., Loeppert, R.H., 1998. Arsenite and arsenate adsorption on ferrihydrite: kinetics, equilibrium, and adsorption envelopes. Environ. Sci. Technol. 32, 344349. Richmond, W.R., Loan, M., Morton, J., Parkinson, G.M., 2004. Arsenic removal from aqueous solution via ferrihydrite crystallization control. Environ. Sci. Technol. 38, 23682372. Riveros, P.A., Dutrizac, J.E., Spencer, P., 2001. Arsenic disposal practices in the metallurgical industry. Can. Metall. Q. 40, 395420. Robins, R.G., Huang, J.C.Y., Nishimura, T., Khoe, G.H., 1988. The adsorption of arsenate ion by ferric hydroxide. In: Reddy, R.G., Hendrix, J.L., Queneau, P.B. (Eds.), Arsenic Metallurgy: Fundamentals and Applications. TMS, Warrendale, PA, pp. 99112. Rowson, J. Private communication. AREVA Resources Inc., Saskatoon, Sask., Canada. Twidwell, L.G., McCloskey, J., Miranda, P., Gale, ., 1999. Technologies and potential technologies for removing as from process and wastewater. In: Gaballah, I., et al. (Eds.), REWAS99. The Metallurgical Society of AIME, Warrendale, PA, pp. 17151726. Twidwell, L.G., Robins, R.G., Hohn, J.W., 2005. The removal of arsenic from aqueous solution by coprecipitation with iron (III). In: Reddy, R.G., Ramachandran, V. (Eds.), Arsenic Metallurgy. TMS, Warrendale, PA, pp. 324. Waychunas, G.A., Rea, C.B.A., Fuller, C., Davis, J.A., 1993. Surface chemistry of ferrihydrite: Part 1. EXAFS studies of the geometry of coprecipitated and adsorbed arsenate. Geochim. Cosmochim. Acta 57, 22512269. Wilkie, J.A., Hering, J.G., 1996. Adsorption of arsenic onto hydrous ferric oxide: effects of adsorbate/adsorbent ratios and cooccurring solutes. Colloids Surf. A 107, 97110. Yuan, T.C., Jia, Y.F., Demopoulos, G.P., 2005. Synthesis and solubility of crystalline annabergite (Ni3(AsO4)2 8H2O). Can. Metall. Q. 44, 449456.