Professional Documents
Culture Documents
www.elsevier.com/locate/jcis
Abstract
Studies on the Merrill–Crowe process as applied to silver recovery have shown that one half of the used zinc powder is wasted in water
reduction at high cyanide concentrations, while the other half reduces silver ions from the cyanide solution. However, the cementation
mechanisms as an electrochemical process taking place on the zinc surface do not explain the split of the electric current resulting from
the anodic dissolution of zinc into two equal values. This study demonstrates that the mechanism for silver precipitation at high and low
cyanide concentrations differs considerably. At low cyanide concentrations cementation is essentially an electrochemically-controlled process
following a shrinking-core behavior. At high cyanide concentrations, the process seems not to be electrochemically controlled. The areas for
zinc dissolution and silver deposition are not connected by an electrical-conducting medium and reduction of silver–cyano complex ions
takes place by hydrogen adsorbed onto silver growing outward from the cementing zinc particles. The results are based on scanning electron
microscopy of solids recovered from cementations in stirred reactors and in situ observations by optical microscopy of the cementation
process on the edge of thin zinc disks in cyanide solutions.
2004 Elsevier Inc. All rights reserved.
the rates for both reactions are the same, instead of (3), the
overall silver cementation reaction should be
Zn0 + Ag(CN)−
2 + H2 O + 2CN
−
− 1
→ Ag0 + Zn(CN)2−
4 + OH + H2 . (6)
2
For this reaction ε = 1, since one mole of zinc precipi-
tates one of silver. This ε value and Reaction (6) have been
verified not only in controlled laboratory experiments at
high cyanide concentrations [1,12], but also in plant oper-
ations [7,9,11].
According to the traditional view of silver cementation
Fig. 1. Schematic representation of the electrochemical behavior of silver as an electrochemical process, the decrease in ε from its
cementation onto zinc in the Merril–Crowe process. maximum value of 2 to 1 seems to be a subject easy to be
understood. It would appear that the cause is a gradual in-
crease in water reduction as cyanide concentration increases,
trons from the zinc oxidation sites up to the silver reduc- taking place simultaneously in the electrochemical process.
tion sites. A schematic picture of this process, as shown in Despite this apparent simplicity the issue is much more com-
Fig. 1, is included in most scientific reports dealing with the plex, since it is necessary to point out that silver and water
Merril–Crowe method to illustrate the cementation process. reduction rates should be of the same order for Reaction (6)
Accordingly, electrochemical techniques have been exten- to be valid. Taking into account that this is true for conditions
sively used to delineate the mechanisms of precious metal as varied as those in plant operations, there is no reason for
cementation in order to optimize plant operations [2–6,8]. both rates to have similar values. In fact, both rates may be
In the Merril–Crowe process as applied to silver recovery completely different, as shown by the studies on the cemen-
from cyanide solutions, the zinc oxidation and silver reduc- tation of precious metals from cyanide solutions using Evans
tion reactions are diagrams by Parga [6]. This researcher did not detect hydro-
gen evolution in the cementation of silver, which occurred
Zn0 + 4CN− → Zn(CN)2− −
4 + 2e , (1)
in gold cementation as reported elsewhere [2,3,5,6]. It can
Ag(CN)− − −
2 + e → Ag + 2CN ,
0 (2) be stated that equality between both rates is not the rule,
but the exception, as deduced from electrochemical stud-
respectively. If side reactions are absent the final stoichiom- ies [2,3,5,6]. A congruent explanation of the origin of ε = 1
etry is is really absent in the literature.
This work aims at thoroughly understanding the mecha-
Zn0 + 2Ag(CN)− 0 2−
2 → 2Ag + Zn(CN)4 . (3) nism of the transition between Reactions (3) and (6) through
The cementation efficiency for Reaction (3) is ε = 2, the the- two different kinds of experiments. The first one was devised
oretical maximum. This value has been found to be true only to model the Merril–Crowe process through cementation
at low free cyanide concentrations, however [1]. As pointed tests in stirred reactors. The second was aimed at explicitly
out above, ε is always closer to 1 than 2 in plant operation. analyzing the validity of an electrochemical behavior for sil-
Besides impurities cementation, the decrease in ε has been ver precipitation onto zinc at high cyanide concentrations.
assigned to the simultaneous and wasteful consumption of
electrons by the following side reactions [9]:
2. Experimental
O2 + 2H2 O + 4e− → 4OH− , (4)
For all experiments, deionized water and analytical grade
1
H2 O + e− → OH− + H2 , (5) AgNO3 , NaCN, and NaOH reagents were used to prepare
2 alkaline silver–cyanide solution at initial pH 10.8. The so-
Reaction (4) is minimized in Merril–Crowe plants by lutions were purged with oxygen-free highly pure N2 to
deaeration of the pregnant solution prior to the zinc powder remove O2 prior to the zinc addition and along all the ce-
addition. Reaction (5) is less manageable since it depends on mentation period.
the free cyanide concentration, which is determined by the In the first set of experiments, zinc powder cementation
leaching conditions. Water reduction was recognized as the tests were carried out in a 2.5 L glass reaction vessel im-
most electron consuming side reaction soon after the imple- mersed in a temperature-controlled water bath. The zinc was
mentation of the Merril–Crowe method [10,11]. Ever since it added when the O2 concentration was lower than 0.2 mg/L.
has been the main explanation for the loss in zinc cementa- After 1 h the cementation products were collected through
tion efficiency. In adequately deaerated pregnant solutions, filtration with 0.22-µm pore size filters and observed and an-
Eqs. (2) and (5) are the main reduction reactions. Thus, if alyzed using a Phillips XL 30 scanning electron microscope
410 G. Viramontes Gamboa et al. / Journal of Colloid and Interface Science 282 (2005) 408–414
Fig. 2. (a) Schematic view of the zinc disk. (b) Two-dimension-like system
with epoxy resin to avoid the precipitation of silver on the zinc disk faces.
Fig. 5. (a) Characteristic texture of initial zinc particles. (b) Texture of the
inner face of the cemented silver shell at the high cyanide concentration
regime.
species on the outer silver–aqueous solution interface. The which is 1.5 × 10−5 cm2 /s [6]. Molecular hydrogen meets
reduction reaction on this interface may be either a simple this requirement as it is smaller in size in comparison with
redox reaction given by that of Ag(CN)− 2.
1 These novel results demonstrate that some electrochem-
Ag(CN)− 2 + 2 H2 → Ag + 2CN + H
0 − + (8) ical methods, such as Evans diagrams, should be used with
a lot of precautions to carry out silver cementation on zinc
or an electrochemical scheme in which hydrogen oxidation studies particularly at high cyanide concentrations.
and silver reduction take place at physically separated sites
as follows
1 4. Conclusions
H2 → H+ + e − , (9a)
2
The chemical routes and morphological patterns present
Ag(CN)− − −
2 + e → Ag + 2CN ,
0 (9b) in the cementation of silver from cyanide solutions with zinc
1 powder depend on the cyanide concentration. At low free
Ag(CN)−
2 + H2 → Ag0 + 2CN− + H+ . (9c) cyanide concentrations an electrochemical process taking
2
place on the zinc particles describes well experimental ob-
Both Reactions (8) and (9c) leads to the formation of servations. For high free cyanide concentrations, like those
a less-alkaline microenvironment. The chemical species in- commonly used in plant operations, zinc oxidation and silver
volved in those reactions are adsorbed at the outer silver– reduction reactions take place on different, non-electrically
aqueous solution interface. Adding Reactions (7c) and either connected, interfaces. Under this last condition, the absence
(8) or (9c) leads to the overall silver cementation stoichiom- of contact between both metals causes the reduction of the
etry for high cyanide concentrations given by Reaction (6). efficiency of zinc to recover silver, exactly by one half of its
Accordingly, the ε = 1 relation originates because the areas theoretical maximum.
for zinc dissolution and silver deposition are not electrically
but electrolytically connected. Maybe the role of hydrogen
microbubbles is to separate the zinc powder particles from Acknowledgments
the silver shell.
The two different mechanisms of silver deposition at This work has been partially supported by the Consejo
low and high cyanide concentration can be accounted for Nacional de Ciencia y Tecnología, México, Grants G29589E
by the stability of Zn(OH)2 and its relationship to the and ER026 Materiales Biomoleculares, and by Centro de
rate-determining steps of the cementation process. At low Investigación y Desarrollo Tecnológico de Servicios Indus-
cyanide concentration and high pH, Zn(OH)2 is highly sta- triales Peñoles. The authors thank Professor Bernardo José
ble and it has been found that it coats the surface of zinc Luis Arauz Lara for the assistance in the use of the equip-
[1,4,6,16]. Under these conditions the rate-determining step ment at the Laboratorio de Fluidos Complejos, Instituto de
is the zinc oxidation [17]. Thus, silver–cyano species diffuse Física, Universidad Autónoma de San Luis Potosí. Thanks
from the bulk solution to the zinc–aqueous solution interface to Dr. Miguel Angel Vidal for use of the mass spectrometer,
where they are cathodically reduced. Then the cementation Instituto de Investigación en Comunicaciones Opticas, from
process follows a shrinking core behavior as observed in the same University, and for helpful discussions.
Fig. 4a, which shows the cemented silver growing toward
the center of the zinc particle.
References
At high cyanide concentration, zinc–cyano complex
species are more stable than Zn(OH)2 [6,16]. Under these [1] G. Viramontes Gamboa, M. Medina Noyola, A. López Valdivieso, Hy-
conditions the cementation process has been reported to be drometallurgy (2004), submitted for publication.
controlled by diffusion through the boundary layer [4,6], [2] I. Barin, H. Barth, A. Yaman, Erzmetall. 33 (7/8) (1980) 399–403.
[3] D.W. Kirk, F.R. Foulkes, J. Electrochem. Soc. 131 (4) (1984) 760–769.
which agrees well with the dense-branching morphology of [4] J.D. Miller, R.Y. Wan, J.R. Parga, in: Proc. of Symp. on Precious and
the silver deposit [1] and a first-order rate for the cemen- Rare Metals Technologies, Albuquerque, NM, USA, 1988, pp. 281–
tation Reaction [6]. When cementation starts, silver–cyano 290.
species are first reduced at the zinc–aqueous solution inter- [5] M.J. Nicol, E. Schalch, P. Balestra, H. Hegedus, J.S. Afr. Inst. Min.
face forming the initial silver layer, which is readily elec- Metall. 79 (1979) 191–197.
[6] J.R. Parga, Analysis of the Zinc Cementation Reaction for Recovery
trically charged driving the silver–cyano species adsorbed of Precious Metals from Cyanide Solutions, Ph.D. dissertation, Dept.
at this interface to be discharged. So the diffusion control of Metallurgy and Metallurgical Engineering, University of Utah, Salt
through the boundary layer is established and the reduction Lake City, 1987.
of silver–cyano species will take place at the silver–aqueous [7] J.R. Parga, R.Y. Wan, J.D. Miller, Miner. Metall. Process. SME (1998)
solution interface making the deposit grow outward. For this 170–176.
[8] E.A. Von Hahn, T.R. Ingraham, Trans. TMS-AIME 239 (1967) 1895–
to occur in the absence of an electrical contact between the 1900.
silver layer and the zinc particle, the reducing species needs [9] C.A. Flemming, Hydrometallurgy 30 (1992) 127–162.
to have a diffusion coefficient higher than that of Ag(CN)− 2, [10] A. James, Cyanide Practice, third ed., 1902.
414 G. Viramontes Gamboa et al. / Journal of Colloid and Interface Science 282 (2005) 408–414
[11] J.V.N. Dorr, F.L. Bosqui, Cyanidation of Gold and Silver Ores, [15] K. Osseo-Assare, in: C.A. Young, A.M. Alfontazi, C.G. Ander-
McGraw–Hill, New York, 1950, pp. 184–187. son, D.B. Dreisinger, B. Harris, A. James (Eds.), Hydrometallurgy
[12] S. Vilchis-Carbajal, Modelación matemática de la cinética de ce- 2003—Fifth International Congress in Honor of Professor Ian Ritchie,
mentación de plata con polvo de zinc a partir de soluciones cianuradas, vol. 2; Electrometallurgy and Environmental Hydromatallurgy, Miner-
M.S. thesis, Instituto Politénico Nacional, Mexico, D.F., 1992. als, Metals, and Materials Society, 2003, pp. 1151–1165.
[13] A.H. Webster, J. Halpern, J. Phys. Chem. 61 (1957) 1245–1248. [16] G. Chi, M.C. Fuerstenau, J.O. Marsden, Int. J. Miner. Process. 49
[14] G. Deshenes, G. Ritcey, G.M. Ghali, in: M.C. Jha, S.D. Hill (Eds.), (1997) 171–183.
Precious Metals ’89, Minerals, Metals, and Materials Society, 1988, [17] S. Vilchis Carbajal, I. González, G.T. Lapidus, J. Appl. Elec-
pp. 341–357. trochem. 30 (2000) 217–229.