order to recover gold from the ammonium thiosulfate solution, ion exchange ... In
this study, the results also demonstrated the advantage of gold recovery from ...
Materials Transactions, Vol. 44, No. 10 (2003) pp. 2099 to 2107 # The Mining and Materials Processing Institute of Japan
Gold Recovery from Nickel Catalyzed Ammonium Thiosulfate Solution by Strongly Basic Anion Exchange Resin Harunobu Arima1; * , Toyohisa Fujita2 and Wan-Tai Yen1 1 2
Department of Mining Engineering, Queen’s University, Kingston, Ontario K7L 3N6, Canada RACE, Department of Geosystem Engineering, The University of Tokyo, Tokyo 113-8656, Japan
Copper catalyzed ammonium thiosulfate leaching process has been proposed as an alternative gold extraction process to cyanidation. In order to recover gold from the ammonium thiosulfate solution, ion exchange process has been proposed as a gold recovery method using strongly basic anion exchange resin. However, the co-adsorption of copper along with gold causes difficulty in separating gold and copper at the gold elution stage. Our previous study has demonstrated that nickel catalyzed ammonium thiosulfate solution for gold extraction has the advantage of reducing thiosulfate consumption. In this study, the results also demonstrated the advantage of gold recovery from the nickel catalyzed ammonium thiosulfate solution by strongly basic anion exchange resin. The optimal gold loading conditions on a 1 g/dm3 strongly base anion exchange resin (wet base value) are investigated in several ion concentrations and 95 kg-Au/t-resin has been obtained. The alternative gold eluant was investigated as the gold loaded resin cannot be eluted by conventional hydrochloric acid. Results showed that the elution efficiency was in the order of OH < Cl NO3 < Br I < ClO4 . The maximum gold recovery by using 2.5 mol/dm3 ClO4 was around 98% with the stripped resin assayed as 0.2 kg/t Au. The feasibility of resin recycling has demonstrated that there was no deterioration in gold adsorption and desorption for four cycles. (Received March 17, 2003; Accepted July 18, 2003) Keywords: gold, thiosulfate, copper, nickel, perchlorate, ion exchange resin, adsorption, elution
1.
Introduction
While cyanide, a highly toxic chemical, has been used for gold leaching since the late 19th century, thiosulfate has been suggested as the best cyanide substitute.1–9) Thiosulfate (S2 O3 2 ) is the least toxic and cheapest chemical which can stabilize gold in aqueous solution. Up to 12 g thiosulfate can be orally ingested with no adverse effects.10) The current market price for ammonium thiosulfate is US$0.30/kg.11) In addition to thiosulfate as a major reagent, both ammonia and cupric or nickel ion are required in the gold leaching process. Ammonia is required to stabilize the catalyst and modify the pH. The toxicity of ammonia weighted exposure average limit is 35 mg/dm3 .10) However, there are some problems in commercializing the thiosulfate process in the gold milling industry. One of major reasons is the uncertainty in the gold recovery from the leach solution. In cyanidation, the Merill Crowe process (Cementation) is one of the classical gold recovery methods. The process is widely applied in current mill production.12–17) In the 1970s, the CIP/CIL (Carbon in Pulp/Carbon in Leach) process was developed as another gold recovery technique.18) The process utilizes activated carbon to recover gold from a gold pregnant solution. The affinity of aurous cyanide complexes to activated carbon was known in early 1910s.19,20) Activated carbon may also adsorb gold chloride, bromide and thiourea complexes.21–25) The RIP/RIL (Resin in Pulp/Resin in Leach) process has been also proposed but is not widely applied on an industrial scale.26,27) Among these three conventional gold recovery processes, CIP/CIL is considered to be least applicable in the ammonium thiosulfate leaching process because the aurous thiosulfate complex does not have affinity for carbon.28–30) *Graduate
Student, Queen’s University. Present address: Akita Zinc Co., Ltd., Akita 011-8555, Japan
The order of affinity of complexed gold on activated carbon is: SCN > SC(NH2 )2 > CN S2 O3 2 . Wan and Brierley (1997) reported that ammonium thiosulfate was applicable for gold extraction from a carbonaceous ore, which causes preg-robbing problems in cyanidation. The process explains a poor affinity of aurous thiosulfate complex toward carbonaceous materials.20) Marchbank et al. (1996) proposed the addition of cyanide to adsorb gold from aurous thiocyanate.31) However, this process is not regarded as a wholly non-cyanide process. Thus, the CIP/CIL process is not feasible to recover gold from a thiosulfate leach solution. Cementation and the RIP/RIL processes are considered as possible methods for gold recovery from an ammonium thiosulfate solution. Cementation process was presented in our previous study.32,33) The following reviews the various studies on the gold recovery from both ammonium thiosulfate and cyanide solutions by the RIP/RIL processes. Although the RIP/RIL (Resin in Pulp/Resin in Leach) process is not commonly used in the gold mining industry, these processes may recover not only gold cyanide but also gold chloride, thiourea and thiosulfate complexes.34–36) Ryan (1987) has considered the RIP process as an alternative to the CIP process.37) Several investigations have proposed the strongly basic anion exchange resin as an alternative gold recovery method from a copper catalyzed ammonium thiosulfate solution.38–43) The first problem of the RIP process was the co-adsorption of copper along with gold and the difficulty of copper and gold separation. The second problem was to elute the gold from the loaded resin. Some eluants, such as thiocyanate, nitrate and polythionate, have been suggested but are still not viable.38–43) Both thiocyanate and polythionate caused sulfur precipitation at the resin regeneration stage by using conventional regenerant of hydrochloric acid,27,44) and the precipitate is required to be isolated from the resin and the eluant in the resin recycling stage. Again, thiocyanate may be
H. Arima, T. Fujita and W.-T. Yen
disproportionate and emit highly toxic fumes of cyanide, when heated or in contact with acid or acid fumes.10) Concerning nitrate as an eluant, a maximum concentration of effluent at 200 300 mg/dm3 Au was produced from a 3 kg/t of gold loaded resin by using 2.0 mol/dm3 NO3 eluant,43) while the Zadra process in cyanidation obtained 1000 mg/dm3 Au from a 4 5 kg/t Au loaded carbon.45) Elution efficiency with nitrate was about 70%.43) To prevent the co-adsorption of catalyst (copper) along with gold onto strongly basic anion exchange resin, nickel catalyzed ammonium thiosulfate solution was utilized in this study. The catalytic action of nickel in ammonium thiosulfate leaching process was found and demonstrated in our previous study.1) Firstly, the stability of copper and nickel with respect to the RIP/RIL process is discussed using the Eh-pH diagram. The effect of various ionic species (nickel, copper, ammonia and thiosulfate) on gold recovery is then experimentally investigated on a 100 mg/dm3 Au solution. Investigating the best aurous thiosulfate eluant from the loaded resin is the secondary objective of this study. Various eluants, such as NaOH, NaCl, NaBr, NaNO3 , KI and NaClO4 , were utilized to evaluate the gold elution efficiency from a 5 kg/t Au loaded resin. Perchlorate (ClO4 ), which has not been reported so far, was found to demonstrate excellent gold elution performance as compared to the other eluants. The optimum elution condition by perchlorate was then investigated. 2.
Thermodynamic Discussion on Ion Exchange Process
Eh/V
In our previous thermodynamic study of copper catalyzed system, at the most favourable reagent composition; [0.3 mol/dm3 NH4 OH, 0.05 mol/dm3 S2 O3 2 and 0.0001 mol/dm3 Cu], the Eh-pH diagram in Fig. 1 showed that anionic cuprous thiosulfate complex [Cu(S2 O3 )2 ]3 is more stable than the cationic cuprous ammine complex [Cu(NH3 )2 ]þ thermodynamically.32) This supports the fact
Eh/V
2100
Fig. 2 Eh-pH diagram for the Ni-NH3 -S2 O3 2 -H2 O system. Condition : 0.5 mol/dm3 NH4 OH, 0.0001 mol/dm3 Ni2þ and 0.05 mol/dm3 S2 O3 2 .1)
that copper ion is co-adsorbed on anion exchange resin.38–43) The co-adsorption of copper along with gold is the major problem for selective elution of gold and copper from the loaded resin. In terms of the thermodynamic evaluation of the most favourable reagent composition of nickel catalyzed system; [0.5 mol/dm3 NH4 OH, 0.05 mol/dm3 S2 O3 2 and 0.0001 mol/dm3 Ni], the Eh-pH diagram in Fig. 2 showed that [Ni(NH3 )6 ]2þ is more stable than [Ni(S2 O3 )2 ]2 .1) This suggests that the nickel is not co-adsorbed along with gold in anion exchange resin thermodynamically. It is obvious that gold elution from the anion exchange resin can be simpler in the nickel catalyzed thiosulfate solution system than in the copper catalyzed thiosulfate system. In the previous study of the copper catalyzed ammonium thiosulfate leaching system, it was confirmed that gold could be recovered efficiently at high thiosulfate and low copper and ammonia concentrations by zinc, copper and aluminium powders.33) This fact indicates that the aurous ion is stable in the form of a thiosulfate complex. The affinity of the gold complex toward anion exchange resin rather than cation exchange resin has been reported by many researchers.38–43) The pH of gold leach solution should be around 9.5 to stabilize the catalyst as ammine complex. Thus, strongly basic anion exchange resin would be selected to recover anionic aurous thiosulfate complex in ammonium thiosulfate solution since weakly basic anion exchange resin reduces the pH value below 9.0.46,47) The most common structure of strongly basic anion exchange resin (polystyrene base) is shown as the following:48,49) ð1Þ
Fig. 1 Eh-pH equilibrium diagram for the Cu-NH3 -S2 O3 2 -H2 O system at reagent combination of 0.3 mol/dm3 NH4 OH, 0.0001 mol/dm3 Cu2þ and 0.05 mol/dm3 S2 O3 2 .32) The gray lines in the stable region of [Cu(S2 O3 )2 ]3 indicate the stable region of copper species in the absence of thiosulfate (0 mol/dm3 S2 O3 2 ).
The positively charged ammoniated site of aromatic nuclei, as shown in chemical formula (1), has a function to adsorb anionic species. Generally strongly basic anion exchange resin should be conditioned to chloride form. Equation (2) shows that aurous thiosulfate complex is exchanged with chloride ion of anion resin:
Gold Recovery from Nickel Catalyzed Ammonium Thiosulfate Solution by Strongly Basic Anion Exchange Resin
4.
3[CH2 N(CH3 )3 ]þ Cl þ [Au(S2 O3 )2 ]3 ! [CH2 N(CH3 )3 ]3 þ [Au(S2 O3 )2 ]3 þ 3Cl 3.
ð2Þ
Selection of Gold Eluant
It has been established that the aurous thiosulfate on a resin surface cannot be eluted by conventional hydrochloric acid due to the strong bonding.38–43) Furthermore, thiosulfate is not stable in acidic solution as shown in eqs. (3)–(5):50) 5S2 O3 2 þ 6Hþ ! 2S þ 2S4 O6 2 þ 3H2 O
ð3Þ
S2 O3 2 þ Hþ ! S þ HSO3
ð4Þ
3S2 O3 2 þ 2Hþ ! 4S þ 2SO4 2 þ H2 O
ð5Þ
Thus, the preferable pH for gold elution is higher than 7.0. Table 1 shows the level of relative selective coefficient for various anionic species, which are potential eluants for gold thiosulfate complex. The coefficients in Table 1 indicate that the order of the anionic ion affinity towards anion exchange resin is: OH < Cl < Br < NO3 < I < ClO4
ð6Þ
O’Malley (2001) has reported the gold elution by nitrate salts from aurous thiosulfate loaded strongly basic anion exchange resin.42) It would be more effective to elute aurous thiosulfate by iodide and perchlorate because these anions have higher affinity than nitrate toward anion exchange resin. The choice of perchlorate as an eluant has the following advantages:44,51–53) . Perchlorate is stable, and does not form a complex with metallic species. . Perchlorate is a well known electrolyte in the study of electrochemistry. This suggests that the gold effluent may be subjected to electrowinning successfully. The relative selective coefficient in Table 1 is evaluated using the following ion exchange reaction (7): [RCH2 N]þ Cl þ A ! [RCH2 N]þ A þ Cl
ð7Þ
Where, R is hydrocarbon chain; A is anion listed in Table 1, and Cl is the reference ion. Thus, the relative selective coefficient for A relative to Cl can be expressed by the following eq. (8):
A KCl ¼
½RCH 2 N þ A ½Cl ½RCH 2 N þ Cl ½A
ð8Þ
Table 1 Relative selectivity coefficient of strongly basic anion exchange resins.45) Anion
OH
Cl
Br
NO3
I
ClO4
Coefficient
0.1
1
3
3.8
9
10
Table 2 Type of ion Exchanger Strongly basic anion exchange resin
2101
Experimental
4.1 Gold adsorption test The reagent grade of ammonium thiosulfate, ammonia, copper sulfate, nickel sulfate and gold sodium thiosulfate salt (Na3 [Au(S2 O3 )2 ]) were used to prepare the aurous ammonium thiosulfate solution. All reagents were supplied from Aldrich Chemical and Alfa Aesar. Table 2 shows the functional group and particle size of strongly basic anion exchange resin (DOWEX 21K) used in this study. The resin was conditioned with a 5% HCl solution prior to adsorption test. This polystyrene based resin is one of the typical ion-exchangers proposed for gold recovery.38–43) In the gold adsorption test, the interference of various ionic species (0:0005 0:005 mol/dm3 : Cu2þ , Ni2þ , 0:5 3 mol/ dm3 : NH4 OH, 0:05 0:3 mol/dm3 : S2 O3 2 ) to recover 100 mg/dm3 Au was investigated. This was done by adding a selected amount of resin (wet base value) in a 100 cm3 solution, which was agitated in a 250 cm3 flask at 200 rpm. The standard reagent composition was adjusted at 0.0005 mol/dm3 Ni2þ , Cu2þ , 0.5 mol/dm3 NH4 OH and 0.05 mol/dm3 S2 O3 2 , based on the results obtained in the previous gold leaching and cementation studies. The barren solution was collected to analyse for gold or other metal values by an atomic absorption spectrophotometer (Perkin Elmer 3300). 4.2 Gold elution test Various eluants were prepared for gold elution test by dissolving reagent grade sodium or potassium form salts, such as NaOH, NaCl, NaBr, NaNO3 , KI and NaClO4 (supplied by BDH, Fisher Scientific Company and Malinckrodt), in distilled water. The pH of all eluants was 7.0, while that of the hydroxide solution was 11.5. The initial gold loading capacity was set at 5 kg-Au/t-resin, which is the common value in the RIP process.41,43) A 2 g of 5 kg/t Au loaded strongly basic anion exchange resin was prepared and transferred to a 10 cm3 burette for gold elution. The unit of flow rate for eluants was defined as BV/H (bed volume per hour) in this experiment. One bed volume for a 2 g resin is 2.5 cm3 . The higher gold loaded resin used for elution was 1.3 cm3 for a 1.0 g of 10 kg/t Au and 0.6 cm3 for a 0.5 g of 20 kg/t Au. The average eluted pregnant solution by the Zadra process is varied in the range between 140 400 mg/dm3 Au (maximum gold concentration in elution profile is 1000 mg/dm3 ) leaving the stripped carbon assay as 0.15 kg/t Au. The final stripped carbon assay depends on the gold processing plant.43) The objective of this study was to produce an average 300 mg/dm3 Au pregnant solution with a stripped resin assayed at less than 0.1 kg/t Au. A 30 cm3 eluant was passed through a 10 cm3 burette packed with 2.5 cm3 of 5 kg/t Au loaded resin by a vertical top to
Ion exchange resin used in this study.46)
Functionally group
Size
Trade name
Quarterly Amine
0:8 1:2 mm
DOWEX 21K
Producer The dow chemical company
2102
H. Arima, T. Fujita and W.-T. Yen
bottom flow operation at ambient temperature. The initial flow rate was set at 12 BV/H for the fundamental investigation to choose the best eluant. Then, the optimal flow rate was investigated on the best eluant. At the end of the elution, a 2 bed volume of distilled water was used to wash out all remaining eluant and aurous thiosulfate complex in the column. The eluted resin was regenerated by a 10 vol% HCl solution and recycled to the gold adsorption stage. During the elution stage, consecutive 5 cm3 of effluent was collected for gold assay by an atomic absorption spectrophotometer. The gold value in the stripped resin was evaluated by the fire assay method.54,55)
Gold recovery /%
80
60
40
1 g/dm 3 3 g/dm 3 5 g/dm 3 10 g/dm3 20 g/dm3
20
Results and Discussion
0 0
100
-
Gold loading capacity / kg . t -1
80
60
40
20
0 0
2
4
6
8
Time, t /h
Fig. 3 Gold adsorption profile onto the 1 g/dm3 of strongly basic anion exchange resin. Solution composition: 100 mg/dm3 Au, 0.05 mol/dm3 (NH4 )2 S2 O3 , 0.5 mol/dm3 NH4 OH, 0.0005 mol/dm3 NiSO4 ; Agitation speed: 200 rpm.
50
100
150
200
Time, t /min
Fig. 4 Gold recovery profile from the nickel catalyzed ammonium thiosulfate solution at the different resin concentrations. Solution composition: 100 mg/dm3 Au, 0.05 mol/dm3 (NH4 )2 S2 O3 , 0.5 mol/dm3 NH4 OH, 0.0005 mol/dm3 NiSO4 ; Agitation speed: 200 rpm.
0.006 100 0.005 90 Gold recovery / %
5.1 Gold adsorption 5.1.1 Kinetics and gold loading capacity Figure 3 shows the kinetics of gold adsorption on the resin from a 100 cm3 solution containing 100 mg/dm3 Au, 0.5 mol/dm3 NH4 OH, 0.05 mol/dm3 (NH4 )2 S2 O3 and 0.0005 mol/dm3 NiSO4 at pH 9.5. Gold loading capacity reached a plateau at 95 kg-Au/t-resin after 3 hours. Figure 4 shows the kinetics of gold recovery with various resin concentrations in the same gold solution. At 1 g/dm3 of resin, 65% of gold recovery was attained in one hour. 100% gold recovery was obtained at resin concentrations of 5, 10 and 20 g/dm3 at retention times of 90, 60 and 30 minutes respectively. The gold loading capacity of 5 kg-Au/t-resin was achieved in 30 minutes at the 20 g/dm3 resin concentration. There was no nickel adsorption on the resin. 5.1.2 Effect of various ionic species on gold recovery The effect of various species, such as nickel, copper, thiosulfate and ammonia, on gold adsorption was studied at a resin concentration of 20 g/dm3 for one hour contact time. Figure 5 shows the effect of the initial nickel concentration on gold recovery and the final nickel concentration. In the range of 0:0005 0:005 mol/dm3 of nickel concentration, there was no effect of nickel on the gold recovery. The final
0.004
0.003 80 0.002 70 0.001 60 0
0.002 Initial Ni concentration,
0.004
Final Ni cocentration / M/mol . dm -3
5.
100
0.000 0.006
M /mol . dm -3
Fig. 5 Effect of initial nickel concentration on gold recovery and final nickel concentration. Fixed value in the solution: 100 mg/dm3 Au, 0.05 mol/dm3 (NH4 )2 S2 O3 , 0.5 mol/dm3 NH4 OH; Resin concentration: 20 g/dm3 ; Retention time: 60 minutes; Agitation speed: 200 rpm.
nickel concentration was not changed. Figure 6 shows the effect of initial copper concentration on gold recovery and final copper concentration. In the range of 0:0005 0:005 mol/dm3 of copper, there was no effect of copper on the gold recovery by strongly basic anion exchange resin. However, the adsorption of copper increased with the increase of copper concentration. The total adsorbed copper in the resin at 0.0005 mol/dm3 , 0.001 mol/dm3 , 0.003 mol/ dm3 and 0.005 mol/dm3 of initial copper concentration was 0.0004 mol/dm3 , 0.0008 mol/dm3 , 0.0023 mol/dm3 and 0.0035 mol/dm3 respectively, which indicated 70 80% of copper adsorption. This fact demonstrates the co-adsorption of the copper ion along with gold on anion exchange resin as reported by many investigators.38–43) Moreover, the result indicates that co-adsorbed copper requires a selective elution of gold and copper from loaded resin, while the nickel ion was not adsorbed on the resin as shown in Fig. 5. Figure 7 shows the effect of thiosulfate concentration on
Gold Recovery from Nickel Catalyzed Ammonium Thiosulfate Solution by Strongly Basic Anion Exchange Resin 0.0008
0.0020
0.0005
70
60 0
0.001
0.002
0.003
0.004
0.005
Initial Cu concentration, M/mol . dm
0.0000 0.006
0.0006 90 Gold recovery /%
Gold recovery /%
0.0010 80
Final Cu concentration, M/mol . dm -3
0.0015 90
0.0004 80
0.0002
70
60
0.0000 0
1
OHCBrNO3IClO-
80
80
0.0002
0.0000 0
0.1
0.2
0.3
Thiosulfate concentration, M/mol . dm-3
60
Gold recovery /%
0.0004
Final Ni c concentration, M/mol . dm-3
0.0006 90
-3
100
100
60
3
Fig. 8 Effect of initial ammonia concentration on gold recovery and final nickel concentration. Fixed value in the solution: 100 mg/dm3 Au, 0.05 mol/dm3 (NH4 )2 S2 O3 , 0.5 mol/dm3 NH4 OH, 0.0005 mol/dm3 NiSO4 ; Resin concentration: 20 g/dm3 ; Retention time: 60 minutes; Agitation speed: 200 rpm.
0.0008
70
2
Ammonia concentration, M/mol . dm
-3
Fig. 6 Effect of initial copper concentration on gold recovery and final copper concentration. Fixed value in the solution: 100 mg/dm3 Au, 0.05 mol/dm3 (NH4 )2 S2 O3 , 0.5 mol/dm3 NH4 OH, 0.0005 mol/dm3 NiSO4 ; Resin concentration: 20 g/dm3 ; Retention time: 60 minutes; Agitation speed: 200 rpm.
Final Ni concentration/ Mmol . dm -3
100
100
Gold recovery / %
2103
40
20
0 0
20
40
60
Time, t /min
Fig. 7 Effect of initial thiosulfate concentration on gold recovery and final nickel concentration. Fixed value in the solution: 100 mg/dm3 Au, 0.05 mol/dm3 (NH4 )2 S2 O3 , 0.5 mol/dm3 NH4 OH, 0.0005 mol/dm3 NiSO4 ; Resin concentration: 20 g/dm3 ; Retention time: 60 minutes; Agitation speed: 200 rpm.
gold recovery and final nickel concentration. In the range of 0:05 0:3 mol/dm3 of thiosulfate, strongly basic anion exchange resin could recover 100% of gold. There was no nickel adsorption at thiosulfate concentration below 0.2 mol/ dm3 . Thus, the preferable thiosulfate concentration should be less than 0.2 mol/dm3 at a combination of 0.5 mol/dm3 NH4 OH and 0.0005 mol/dm3 NiSO4 , so as to prevent the coadsorption of nickel along with gold. This co-adsorption of nickel at higher thiosulfate concentration can be attributed to the increase of nickel thiosulfate stability. The formation of the nickel thiosulfate complex is confirmed thermodynamically as shown in Fig. 2. Figure 8 shows the effect of ammonia on gold recovery and the final nickel concentration. Despite the increase of ammonia concentration from 0.5 mol/dm3 to 3 mol/dm3 , gold was well stable in the anionic form of the thiosulfate
Fig. 9 Aurous thiosulfate ([Au(S2 O3 )2 ]3 ) elution profile from 5 kg/t of gold loaded strongly basic anion exchange resins by various 1 mol/dm3 eluants. Resin weight and volume: 2 g, 2.5 cm3 ; Flow rate: 12 bed volume per hour.
complex. 100% of gold was recovered by the anion exchange resin, and nickel adsorption was not detectable. These results demonstrate that the gold content in the ammonium thiosulfate solution can be recovered in a wide range of ammonia concentration without nickel adsorption. 5.2 Gold Elution 5.2.1 Investigation of eluant Figure 9 shows the gold elution by various eluants at 1 mol/dm3 . The result shows that both the OH and Cl ions could not elute gold from resin, while nitrate and bromide could elute 10% and 20% of gold respectively in one hour. The superior eluants were iodide and perchlorate, which eluted 62% and 72% of gold respectively in one hour. The feasibility of gold elution essentially agreed with the relative selectivity coefficient as shown in Table 1.
2104
H. Arima, T. Fujita and W.-T. Yen 500
1000 OHClBrNO3IClO4-
300
800 Au Concentration/ kg . m -3
Au Concentration/ kg . m -3
400
200
100
0 0
2
4
6
8
10
12
14
1.0 mol/dm3 1.5 mol/dm3 2.0 mol/dm3 2.5 mol/dm3
600
400
200
0
16
0
2
Bed Volume Number
4
6
8
10
12
Bed Volume Numnber
Fig. 10 Total elution curve of the aurous thiosulfate ([Au(S2 O3 )2 ]3 ) elution test from 5 kg/t Au loaded strongly basic anion exchange resins by various 1 mol/dm3 eluants. Resin weight and volume: 2.0 g, 2.5 cm3 ; Flow Rate: 12 bed volume per hour.
Figure 10 shows the relationship between eluant volume and gold concentration of effluent. The highest gold concentration was about 450 mg/dm3 in the perchlorate solution. The peak of gold concentration was shifted to the right side with a proportional decrease of gold elution performance. In conclusion, perchlorate was chosen as the aurous thiosulfate eluant for a strongly basic anion exchange resin. 5.2.2 Investigation of optimal perchlorate concentration on gold elution Figure 11 shows the effect of perchlorate concentration on the gold elution kinetics from a 2.5 ml of 5 kg/t Au resin. The solution of 1.0 mol/dm3 , 1.5 mol/dm3 and 2.0 mol/dm3 ClO4 eluted 72%, 82% and 92% of gold, respectively in 60 minutes. The maximum gold recovery of 98% was obtained from a 2.5 mol/dm3 ClO4 in 60 minutes. Even when the perchlorate concentration was increased to 3.0 mol/ dm3 , gold recovery remained unchanged at 98%. 100
Fig. 12 Effect of perchlorate (ClO4 ) concentration on total elution curve of the aurous thiosulfate ([Au(S2 O3 )2 ]3 ) elution test from 5 kg/t Au loaded strongly basic anion exchange resins. Resin weight and volume: 2.0 g, 2.5 cm3 ; Flow Rate: 12 bed volume per hour.
Figure 12 shows the relationship of eluant volume at various concentrations and gold concentration of effluent. The maximum gold concentrations of 450 mg/dm3 , 500 mg/ dm3 , 750 mg/dm3 and 850 mg/dm3 were obtained from 1.0 mol/dm3 , 1.5 mol/dm3 , 2.0 mol/dm3 and 2.5 mol/dm3 ClO4 eluants, respectively. Overall, 2.5 mol/dm3 ClO4 was the optimal eluant concentration to obtain 98% gold recovery in 12 BV/H in one hour. The stripped resin was assayed as 0.24 kg/t Au, which is exceedingly high to be accepted for the industrial application. Further unproved optimization tests will be described in the next section. 5.2.3 Investigation of optimal flow rate on gold elution In order to achieve a lower gold value in the stripped resin, the test was conducted at a slower flow rate. Figure 13 shows the effect of the eluant flow rate on the gold concentration profile by using a 2.5 mol/dm3 ClO4 solution to elute gold from a 5 kg/t Au resin. The decrease of the flow rate from 12 BV/H to 1.5 BV/H increased the gold concentration of the eluate. The maximum gold concentrations eluted at 12 BV/
80
1400
60
40 1.0mol/dm3 1.5mol/dm3 2.0mol/dm3 2.5mol/dm3
20
0 0
20
40
60
80
Time, t /min
Au Concentration / kg . m -3
Gold recovery /%
12BV/H 1200
6BV/H
1000
3BV/H 1.5BV/H
800 600 400 200 0 0
2
4
6
8
10
12
Bed Volume Number
Fig. 11 Effect of perchlorate (ClO4 ) concentration on aurous thiosulfate ([Au(S2 O3 )2 ]3 ) elution profile from 5 kg/t of gold loaded strongly basic anion exchange resins. Resin weight and volume: 2.0 g, 2.5 cm3 ; Flow rate: 12 bed volume per hour.
Fig. 13 Effect of flow rate on the gold elution profile by 2.5 mol/dm3 ClO4 from 5 kg/t Au loaded strongly basic anion exchange resin after passing 12 bed volume of eluant. Resin weight and volume: 2.0 g, 2.5 cm3 .
Gold Recovery from Nickel Catalyzed Ammonium Thiosulfate Solution by Strongly Basic Anion Exchange Resin 100.0
0.30
24BV/H 12BV/H 6BV/H 3BV/H
99.0
0.20
98.5
0.15
98.0
0.10
97.5
0.05
97.0
0.00 0
2
4
6
8
10
12
1200 Au Concentration, kg . m -3
0.25
Gold Value in Stripped Resin /kg . t -1
Gold recovery /%
1500
99.5
2105
900
600 Optimum Bed
300
0 0
14
4
8
12
16
20
24
Bed Volume Number
Flow Rate, BV/H
Fig. 14 Effect of flow rate on final gold recovery and gold assayed result of stripped resin eluted by 2.5 mol/dm3 ClO4 from 5 kg/t Au loaded strongly basic anion exchange resin. Resin weight and volume: 2.0 g, 2.5 cm3 .
3
H, 6 BV/H, 3 BV/H and 1.5 BV/H were about 850 mg/dm , 1100 mg/dm3 , 1230 mg/dm3 and 1250 mg/dm3 , respectively. Figure 14 shows the effect of the eluant flow rate on gold recovery and the gold value of stripped resin. The gold recovery and the associated gold value in the stripped resin eluted at the flow rates of 12 BV/H and 6 BV/H were 97.8% with 0.24 kg-Au/t-resin and 98.4% with 0.15 kg-Au/t-resin respectively. With further decrease of the flow rate to 3 BV/ H and 1.5 BV/H, the gold recovery was 99.5% with 0.05 kgAu/t-resin and 100.0% with 0.00 kg-Au/t-resin respectively. Thus, the optimal flow rate on a 2.5 cm3 of 5 kg/t Au resin was chosen at 3 BV/H, which produced an average 300 mg/ dm3 Au pregnant solution with a maximum concentration of 1250 mg/dm3 Au. 5.2.4 Investigation of optimal flow rate on gold elution for a higher gold loaded resin The optimal flow rate for 1 g (wet) of 10 kg/t Au resin and 0.5 g of 20 kg/t Au resin was also investigated. Figure 15 shows the effect of the eluant flow rate on the gold elution profile from a 1 g of 10 kg-Au/t-resin. The maximum gold concentration of 1400 mg/dm3 was obtained at a flow rate lower than 6 BV/H. This maximum gold concentration was
Fig. 15 Effect of flow rate on the gold elution profile by 2.5 mol/dm3 ClO4 from 10 kg/t Au loaded strongly basic anion exchange resin. Resin weight and volume: 1.0 g, 1.25 cm3 .
about 200 mg/dm3 Au higher than the result of a 5 kg/t Au resin as shown in Fig. 13. As a result no gold value appeared in the effluent at the bed volume of 16 20. The result indicates that an average of 400 mg/dm3 gold pregnant solution can be produced from a 10 kg/t Au loaded resin. The gold value of the stripped resin was less than 0.05 kg-Au/tresin at a flow rate slower than 6 BV/H. In terms of a 0.6 cm3 of 20 kg/t Au loaded resin, the maximum gold concentration of 1700 mg/dm3 was obtained at a flow rate slower than 12 BV/H. An average of 400 mg/ dm3 Au pregnant solution could be produced. The gold stripped resin was assayed at less than 0.05 kg/t Au. 5.2.5 Feasibility of perchlorate elution process Table 3 shows the optimal gold elution conditions and results by using a 2.5 mol/dm3 ClO4 eluant as compared with that of the conventional Zadra Process in cyanidation.43) It is obvious that perchlorate elution is faster and more efficient than the Zadra Process, which required 1 2 BV/H at 363 373 K and 400 500 kPa to give 96 98% gold recovery leaving the stripped carbon assayed at 0.15 kg/t Au. The gold recovery using perchlorate eluant was higher than 99.5% leaving the stripped resin assayed at 99:5%
>99:5%
>99:5%
96 98%
Gold assayed stripped resin/carbon [kg/t]
