Chmical Engineering Science, Vol. 47, No. 8, pp. 1933-1941, 1992. ooo9-2509/92 ss.00 + 0.

00
Printed in Great Britain. 0 1992 Prrsaroon Press Ltd

MATHEMATICAL MODELLING OF METAL LEACHING IN

NONPOROUS MINERALS

GRETCHEN LAPIDUS
Departamento de Ingenieriade Procesos e Hidraulica, Universidad Autbnoma Metropolitana-Iztapalapa,
Apartado Postal 55-532, Mexico 13, DF 09340, Mexico

(First received22 January 1991; accepted in revisedform 12 December 1991)

Abstract-A model based on the shrinking-core principle is presented for three cases of oxidative metal
leaching from nonporous mineral particles. Several limitations inherent in the assumptions of the tradi-
tional shrinking-core model, such as those of oxidation agent in excess and first-orderirreversiblekinetics,
are overcome. In each case, an unconventional behavior observed in experimental data is explained
through numerical leaching simulations.

INTRODUCI’LON (Habashi, 1967). The cyanidation reaction is so fast

In the literature, very few attempts have been made to with respect to the time scale involved in the leaching
model the oxidative leaching kinetics of metals from process (24 h) that it is considered to be in equilibrium
nonporous mineral particles. Even in these cases, the at the reaction boundary. In the last case, the interest-
traditional shrinking-core model (SCM) with nth- ing aspects of the former case are further complicated
order irreversible kinetics and the pseudo-steady state by complexing phenomena coupled with diffusion.
(PSS) assumption is employed (Warren et al., 1985; The most important simplification used in all of the
Rath et al., 1988; Fuerstenau et al., 1986; Lapidus and cases is that the models only take into account the
Mosqueira, 1988). These assumptions only apply to principal metal to be leached. The effects that other
a very limited set of conditions. Specifically, the PSS impurities have on consumption of the oxidating
assumption implies that the concentration of the oxi- agent, or on the created porosity in the mineral, are
dation agent remains constant in time. In most indus- the subject of another study which will be published in
trial applications, near-stoichiometric concentrations the future.
of the oxidation agent are employed, a situation which
causes a reduction in its level as the reaction proceeds. THEORY
Taylor et al. (1983) presented an extension of the SCM The generalized shrinking-core theory assumes the
which does not assume a PSS. However, the model is following (Yagi and Kunii, 1955):
still quite limited due to the fact that the concentra-
tion of the oxidation agent is considered constant (in
(1) The system is isothermal.
excess) and the kinetics are assumed to be first-order
(2) The desired metal to be leached has an initially
irreversible; consequently, the effect of product accu- uniform distribution throughout the mineral
mulation in the bulk solution on the overall kinetics particle, and the other species present are con-
cannot be accounted for. sidered to be inert or, at least, do not react with
The models presented in this paper, while still based the same oxidant.
on the shrinking-core concept (diffusion and reaction
(3) The mineral particle is spherical and conserves
in series and constant particle radius), incorporate its size and shape during the leaching process.
variable bulk concentrations, irreversible or fast re-
(4) The reaction boundary is considered to be well-
versible reactions with one or more oxidation agents defined and the external mass transfer, the dif-
and the use of more complicated kinetic expressions. fusion and the reaction resistances act in series.
All of these considerations are necessary as a first step
in the direction of building an industrial-reactor simu- The diffusion of any species in the inert layer may
lator. Only then will it be possible to employ laborat- be described by Fick’s law in spherical coordinates,
ory results on monosized particulate systems in the i.e.
scaleup for multisized particulate industrial reactors.
The cases that are addressed in this paper are rep-
R, < R < R, (1)
resentative and were selected in view of certain char-
acteristics of the oxidation reaction. The first case,
where
sphalerite leaching with ferric chloride, is by far the
simplest, because the kinetics are basically a function cl = concentration of species i
of the oxidation agent (Warren et al., 1986). In case II, R = radial position in the inert layer
cyanidation of silver has been treated in view of the Di = effective diffusivity of species i in the inert
interesting observed behavior with respect to pH layer.
1933
1934 GRETCHEF v LAPIDUS

If complexing reactions occur in the solution, these The reaction rate expression (RRE) may assume any
are considered to be very fast compared to the leach- irreversible or reversible form as long as the diffusion
ing reaction, and eq. (1) may be used to describe the of all species included in the RRE are described. If,
diffusion of the total chemical species through the instead, local equilibrium is chosen as the boundary
inert layer (Lapidus, 1991). condition, eq. (6) would be replaced by the thermo-
The initial conditions correspond to the total con- dynamic equilibrium constant, and the fluxes of the
centration of species i added to the system and the different components at the reaction boundary would
equivalence of the two radii: be related through the stoichiometric coefficients of
the oxidation reaction.
for t = 0,
The core-shrinkage rate may be described by a ma-
ci = cp (2) terial balance with respect to the flux of the desired
metal, species j:
R, = R,. (3)
dR,
-=--
Di dci
For t > 0 in the bulk solution, a material balance
dt biPM aR R.
must be made for the consumption or production of
each reacting species. This balance is related to the where
percentage of solids in the system in the following
pm = molar density of the desired metal in the
manner:
mineral.
cf = 60 +b. E;pI %. x The conversion of the desired metal is related to the
‘MWlOO-%%.
position of the shrinking core by the following expres-
= c;O + b*aX (4) sion (Levenspiel, 1972):
3
where

cf = concentration of species i in the bulk

solution To solve
x+.
(> P

the problems presented in this study, the

cf* = initial concentration of species i in the equations were nondimensionalized to fix the moving
bulk solution boundary using the same variables as presented by
bi = stoichiometric coefficient of species i, in- Taylor et al. (1983):
cluding its sign, in the oxidation reaction
with respect to that of the desired metal
product
F = weight fraction of the desired metal in the Once this was performed, the differential equations
mineral and their respective boundary conditions were dis-
P, = leaching solution density cretized using orthogonal collocation (with three col-
MW = molecular weight of the desired metal location points) and the implicit Euler method in the
%, = percentage of solids in the system radial and temporal coordinates, respectively. The
X = fractional conversion of the desired entire system of equations, including the initial condi-
metal. tions, was solved using the Newton-Raphson method,
When there are external mass transfer limitations, starting at a very small time and integrating up to the
e.g. in the case of a fast reaction rate combined with desired time or conversion.
the lack of particle suspension in the reactor, the
difference between the concentrations of species i in Case I: sphaierite leaching with ferric chloride
the bulk solution and at the particle surface may be The oxidation reaction for this system is the follow-
appreciable. In this case, a specific correlation for the ing:
mass transfer coefficient, hi, is necessary and the fol-
ZnS + 2Fe3+ + 2Fe*+ + Zn*+ + So.
lowing equation is used: at R, (particle surface),
The leaching kinetics have been found mostly to be
Dgi = h,(cf - cf). reaction controlled. However, in the high-temper-
'3R Al,
ature region (above about 7O”C, depending on the size
Under normal laboratory operating conditions, an fraction) diffusion control becomes significant. Al-
effort is made to eliminate the external mass transfer though there is disagreement among authors about
limitations using rigorous agitation. When this is pos- the order of the reaction (Warren et al., 1986; Rath et
sible, cf = cf because hi is large. al., 1989; Palacios et al., 1989), using the SCM under
At the reaction boundary, the flux of the oxidation PSS conditions, 2CrJpzn < 0.1, and a large excess of
agent or agents is equal to its rate of reaction as ferric chloride, a power-law expression using any n fits
follows: at R, (reaction boundary), all of the experimental data quite well. However, when
the concentration of ferric ion is not constant, as in
aCi
= reaction rate expression. (6) the case of an initial near-stoichiometric amount of
D%i J& the oxidant, or when, in the mixed kinetic region, the
 Mathematical modelling of metal leaching in nonporous minerals 1935
simple expression for reaction control from the SCM, Case II: silver leaching with sodium cyanide and air
1 -(l -X) 113, does not give a straight line. In this system, rotating disc studies have deter-
In Fig. 1, data obtained by Palacios et al. (1989) for mined that the kinetics are very fast compared to the
ferric chloride leaching of a sphalerite concentrate (at mass transport phenomena (Habashi, 1967). How-
80°C and pH = 1) are plotted as the SCM reaction ever, it has been found in plant practice and laborat-
control function vs time for four different ferric ion ory studies that there is a notable decrease in the silver
concentrations, which correspond to 700, 200, 100 extraction rate at high values of pH ( > 12) which
and 50% of the stoichiometric amount required. For cannot be solely explained by differences in diffusion
the highest concentration, 700%, the data fit rates of the reacting species. This has been attributed
a straight line. However, as the concentration de- to several factors, such as cyanide decomposition or
creases, this is not the case and determination of the reduced oxygen solubility at higher pH values, which
kinetic rate constant from the slope of the line is not are unconvincing because the same tendency is not
possible. For this reason, a more complex model is found in gold extraction (Habashi, 1967).
needed to describe the leaching kinetics. The principal difference between the gold and silver
Since the reaction is essentially irreversible, only the systems is that the former has an equilibrium constant
diffusion of ferric ion need be described. For this for the oxidation reaction eight orders of magnitude
system the following equations were used: larger than the latter. Since the hydroxyl ion is one of
the reaction products, it may be expected that local
equilibrium at the reaction boundary is one of the
R, -K R < R, (9) causes of the rate decrease at higher initial hydroxide
concentrations.
cFc = CR and R, = R,, for f = 0. (10, 11) To demonstrate this effect, a simulation was per-
formed using the oxidation reaction proposed by
Since the agitation was sufficiently high in these ex-
Bodlanger (1896):
periments to reduce the external mass transfer resist-
. . 2Ag0 + 4CN- + O2 + 2H20 ti ZAg(CN);
ante to a mmlmum, c$, x &. The boundary condi-
tion at the particle surface reduces to a material bal-
+ 20H- + H202.
ance of the ferric ion that remains in the bulk solution:
at R,, The diffusion behaviour of each of the species in the
0
above reaction, except water due to its presence in
CFe = CFe - aaxz,, for t > 0. (12) excess, may be described by the corresponding forms
At the reaction boundary, the ferric ion flux equals of eq. (1). Nondimensionalization is performed using
its reaction rate, i.e. at R,, the concentration and the effective diffusivity of dis-
solved oxygen as reference values.
D+e The boundary conditions at the particle surface are
kc”,.. (13)
Fc aR similar to those employed in case T, material balances
=

R,

for each species, except for oxygen. For this species,

Since the flux of ferric ion is related through the the concentration in the bulk solution is considered
oxidation reaction to that of zinc ion at R,, the core- constant because of continuous aeration of the pulp.
shrinkage rate may be expressed as follows:
At R,,
dR,_
_--
DFe aCFs
(14)
dt 2PM aR K.’ = c& = constant due to the aeration
co*
of the solution (15)
Solving these equations for n = 0 and l/2, the
best fit (for both orders) to the experimental data is CcN = c& - 2aX (16)
shown in Fig. 2 as fractional conversion vs time CA&ZN)z = clx (17)
(Palacios et al., 1989). As shown in the table below,
the simulation gives consistently increasing or de- cOH = c& + aX (18)
creasing values for the reaction rate constants for
cn202 = +aX (19)
n = 0 and l/2, respectively (Fig. 3).

n=O ?l= l/2

Ferric ion Damkdhler Damlciihler

concentration (M) k x 1O-9 number k x 1O-6 number

0.00625 3.6 6.1 1.9 4.4

0.0125 4.7 9.1 1.7 9.1
0.0250 5.0 1.8 1.3 1.8
0.0970 5.5 0.9 0.6 0.6
1936 GRETCHEN LAPIDUS

0 Fet3 - 0.00625 M
A Fet3 = 0.0125 M
o Fet3 - O-0260 M
-<
* Fet3 = 0.0870 M .w
_A I

TIME. MINUTES

Fig.1. Sphalerite leaching with ferric chloride. Experimental data reported by Palacios er al. (1989) at 80°C
and plotted as the reaction control function of the traditional SCM, 1 - (1 - X)“j vs time in minutes.

= 0.097 M
0 Fe+3 = 0.025 M

- 0.00326 M

0 30 60 90 120 150 180 210 240 270 300 330 3Qo

TIME, MINUTES

Fig. 2. Sphalerite leaching with ferric chloride. Model simulations compared with the experimental data of
Palacios et al. (1989) plotted as fractional conversion, X, vs time in minutes.

where At R,.
X = fractional conversion or extraction of silver. K, = h,CNdz (cOW)2(~HzO+)
= 2.98 x lo8 (20)
At the reaction boundary, local equilibrium for the b~)(‘kN)~

oxidation reaction is assumed and relative fluxes of all

of the reacting species are dictated by the stoichiomet- acAs(CNh aco2
D As(CN)r = -2Do,-
ric coefficients. Equilibrium is only considered at this aR R, aR R,
(21)

boundary, akd nowhere else in the diffusion region,

because at this point the concentrations of the react-
ants are the lowest and those of the products the
highest.
= -to,q R,
(22)
 Mathematical modelling of metal leaching in nonporous minerals 1937

tration dissolved in water at the same temperature

(23) and 1 atm.

h-I202 Do, =2.4x lo-’

= 2h.02 y . (24) DcN = 1.7 x 1O-5
Rc
D AS(,JN)2
= 0.3 x lo- ’
The equation for the core-shrinkage rate is
Doa = 2.7 x 1O-5
dRc
-=- DAS,CN)~acA,s(cNh D H2o2 = 2.3 x 1O-s.
dt 2p, aR R,’
The results of the simulation are plotted in Fig. 4
The values of the free diffusivity of each species which shows conversion of silver vs dimensionless
(listed below) were taken from the literature (Habashi, time for different initial pH values at 25°C. The pulp
1967; Schmid and Curley, 1975; Geaukoplis, 1972). consisted of 50% total solids of a mineral containing
The equilibrium constant for the oxidation reaction 200 g/ton of metallic silver. The times are dimension-
was calculated with the redox half-potentials at 25”C, less, Do,tfR& since experimental data for monosized
changing the reference state for oxygen to its concen- particles are not available in the literature.

on-0

o.al 0.01 0-W 0.03 0.04 0.05 0.08 0.07 0.08 o.cm 0.10
Fe+3 ooncerttrotion. M

Fig. 3. Sphaleriteleaching with ferricchloride. Estimated rate constants, k. for assumed reaction orders of
0 and I/t as a function of ferric ion concentration in mol/l.

I mu
________-___e,
______------
0-Q zesi___

0.6

0.7

0.6

X 0.5

0.4

0.3

0.2

0.1

0.0’
0.0 0.5 I .o 3.5 4.0
DMEr&&-: TIblE5WRg

Fig. 4. Silver leaching with cyanide. Model simulationsshowing the et%ct of the initial solution pH on the
extraction of silver from a 200 g/ton mineral as fractional conversion, X, vs dimensionless time, Dt/Rf .
1938 GRETCHEN LAPIDUS

The numerical results show that metallic silver it should be expected at pH values below about 2.5,
leaching is reduced for values of pH both above and where hydroxide complexing of the ferric ion is insig-
below 12 at low effective concentrations of sodium nificant, that the silver extraction would be inde-
cyanide (0.20 kg/ton mineral, which corresponds to pendent of the acid concentration in solution. How-
approximately 10% more cyanide than the stoichio- ever, Tellez et al. (1986) observed a pronounced effect
metric amount necessary to leach all of the silver). At vriithargentite similar to that observed for gold. This
pH 10, this behavior is principally due to the lower behavior has been attributed to a secondary oxidation
concentration of free cyanide in solution, while at pH reaction by an oxidized species of thiourea, for-
14 it is due to the high hydroxide concentration. In mamidine disulfide, whose reduction is favored by
both of these cases, equilibrium is reached in the bulk increased acid concentration (Groenwald, 1975).
solution. In all of the simulations at this concentration This and other observations, such as the unconven-
of cyanide, evidence exists from the concentration tional dependence of the extraction rate on the ferric
profiles that there is a change in the rate-controlling ion concentration, are evidence that the complexing
step from oxygen diffusion at small times to product phenomena occurring in the solution should necessar-
diffusion at large times (Fig. 5). ily be taken into account when modelling the leaching
A low effective cyanide concentration, similar to the kinetics (Hiskey, 1981).
one presented here, presents itself for the case of In the present system, local equilibrium is con-
highly oxidized minerals which contain large amounts sidered at the reaction boundary because the
of cyanicide metals. This behavior is not observed oxidative reaction rate for eq. (26) has been proved to
with gold probably due to the higher equilibrium be very fast (Chen et al., 1980). Additionally, it is
constant of the oxidation reaction, K, = 8.4 x lOi assumed that, in the entire solution phase, the follow-
(Lapidus, 1988). ing spontaneous equilibria exist:

Case III: argentite leaching with ferric sulfate and Fe3+ + SOg-*FeS04+ (27)
thiourea
H+ + SO:--=HSO; (28)
The thiourea and ferric sulfate system has been
considered as an alternative to cyanidation for silver Fe’+ + SO:-Z$F~SO, (29)
extraction from minerals and concentrates. However,
Fe3+ + SOf- + CS(NH2)s~FeS04CS(NH& (30)
due to the complexity of this system, its kinetic behav-
ior cannot be described by simple models which 2Fe3+ + 2CS(NH&e2Fe2+ + 2H+
would allow optimization. For example, the principal
+ (NH,)NHC-SSCNH(NH,) (31)
oxidation reaction may be written as follows:
Ag+ + CS(NH,),=AgCS(NH,): (32)
Ag,S + 2Fe3+ + 6CS(NH,),=2Ag[CS(NH,),];
Ag+ + 2CS(NH,),=Ag[CS(NH,),]: (33)
+ 2Fe2 * + So. (26)
Ag+ + 3CS(NH,),=Ag[CS(NH,),]: (34)
As may be observed, in the above reaction there is no
direct dependence on the solution pH. Consequently, Ag’ + 4CS(NH,),=Ag[CS(NH,),]:. (35)

0.8 -

OXYGEN RT LFRn TIME --.S_ILVERAT LFRn TIME

-----_---_______________
C
0.5 -
CMAX
0.4 -

0.3 -

OXYQN FIT St4XL TIME

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.Q 1.0
DIMENSIONLESS RFIOIUS. R-Rc/Rp-Rc

Fig. 5. Silverleachingwith cyanide. Oxygen- and silver-simulatedconcentrationprofiles at small and large

times plotted as dimensionless concentration, c/c,,,, as a function of dimensionless radius,
R - RJR* - R,.
 Mathematical modelling of metal leaching in nonporous minerals 1939

In order to facilitate manipulation of the equations, for t > 0, at R = R, (bulk solution),

each species is numbered as shown below:
CFTI = cFTLl” (47)
1 = Fe3+ 8 = FeSO,CS(NH&
CSTI = CST1° (48)
2 = sot- 9 = HSO,
D-U = cTTl” (49)
3=H+ 10 = (NH,)NHC-S-S-
CNHfNHz) [AT] = [AT]’ - [PT-J (50)

4 = CS(NH2)2 11 = FeS04 CPU =aX (51)

5 = Fe2+ 12 = Ag+
where
6 = Ag[CS(NH,),],+ 13 = Ag[CS(NH2)J+
X = fractional conversion
7 = FeSOZ 14 = Ag[CS(NH,)J: for t z- 0, at R = R,
15 = Ag[CS(NH,)Jf.
c5cS
Because the concentrations of the different species G! =2 (52)
c1c4
depend on the local chemical environment at each
point in the inert layer, it is more convenient to
flux [F-r] = 0 (53)
calculate the diffusion of the total element or func- flux [ST] = 0 (54)
tional group. In this manner, the complexing reac-
flux [TTI = 0 (55)
tions cancel each other out, and the formation of
a given species from another is automatically taken flux [ATI = - flux [PT-J (56)
into consideration. With this purpose in mind, the
where
following functional groups may be used for this sys-
tem: flux [zT] = c Di(dci/cYR)
i
total iron = [FT] = ci + cg + c7 + cs + cl1 (36)
total sulfate = [ST-J = c2 + c, + cs + cg + c1 1 (37) i = species that contains the element or functional
group.
total thiourea = LTl”l = ca + 2cs + cs
The equation for the core-shrinkage rate in this
+ 2ClO + Cl3 + 3c14 + 4c15 (38) system is
total oxidating agent = [ATI = c1 + c, + cg + 2c,,
pnr $$ = flux[PT-J I&. (57)
(39)

total silver = [pT7 = c6 + c,~ + clj + cl4 + ~~5. The system of equations was solved for 80 pm par-
WI ticles of a nonporous mineral with 200 gAg/ton at
25% solids. The results of the leach simulation using
Therefore, eq. (1) would assume the following form
a solution of 0.03 M thiourea and 0.003 M ferric sul-
for the diffusion of total iron:
fate for pH 1, 1.5 and 1.9 appear in Fig. 6 as fractional
m--l conversion vs dimensionless time, D 1t/R:. As may he
Eat Dl$f+D5$f+D,~ observed, the difference in the extraction rates at the
different pH values is appreciable.
aG3 ac11 Another interesting effect may be seen in Fig. 7
+Ds~+DlldR . where the simulated leaching behavior is presented for
>I
the same system at pH 1 and 0.03 M thiourea at
The differential equations for the remaining four func-
different concentrations of ferric sulfate. The leaching
tional groups ([ST-J, [TIJ, [ATI and [pT7) are ana-
rate is actually lower at the highest ferric ion concen-
logous.
tration than that observed at the intermediate concen-
The initial and boundary conditions for the five
trations. The effect is due to complexing of ferric ion,
diffusion equations are as follows:
eq. (30), which leaves less free thiourea in solution. As
for t = 0, can be expected, this tendency disappears at higher
thiourea concentrations.
[FTI = [FTI initial = [FTjO (41)

CSTI = CST1° (42)

CONCLUSION
D-l-l = CTTIO (43)
It has been shown that the models presented in this
CAT7 = CATlo (44) article can be used to simulate the leaching behavior
under a wider range of conditions than that possible
[PT-J=O (45)
with the traditional SCM. In addition, reversible oxi-
R,= R, (W dation reactions and complexing phenomena may be
1940 GRETCHEN LAPIDUS

I _o _______A----
__---
_____------
0.8 -
___---
*__--*

__---
__.-__-- ___---
0.8 -
Rgx pH - 1 .?/.-*---- ____‘_----&.--w - 1.5
0.7 - ,._- */-
d/S /*r
__-_ _/- -H^
0.6 - *’ -0 _---
.- _-
*’ r-
.’ .. ,/--
x ox- ,’ .* *- ----G w - 1-Q
,I ,* ,y
,’ ,’ _A
,,’ /’ ,I
0.4 - ,’
.,’ ,’ .’
0.3 - ;*;fl’,,-
;,‘,/
0.2 -
;/
0.1 $

0.0r
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
DIWNSICN_ESS TIME, Utr’Rp2

Fig. 6. Argentite leaching using 0.003 M ferricsulfate and 0.03 M thiourea. Model simulations at different
values of the solution pH plotted as fractional conversion, X, vs dimensionless time, Da&/R;.

I .o-
0-Q -

- 0.002 M

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.8 1.0
DIMENSIONLESS TIME. Dt/Rp2

Fig. 7. Argentite leaching using 0.03 M thiourea at pH = 1. Model simulations for different ferric ion
concentrations plotted as fractional conversion, X, vs dimensionless time, Da tfRf.

taken into account with the methods employed. Fur- ing its sign, in the oxidation reaction with
thermore, it is possible to extend the same procedure respect to that of the desired metal product
used in the development of these models to other %J concentration of species i
leaching systems where complex kinetics are involved. Ci initial concentration of species i
Di effective diffusivity of species i
Acknowledgements-The author wishes to thank Dr Susana F weight fraction of the desired metal in the
Ortiz of Imperial College of Science, Technology and Medi- mineral
cine for her assistance in the conceptualization of these
P--l total concentration of iron
models, and Drs Jesus Alvarez and Jose de Jesus Alvarez for
their helpful suggestions on numerical computations. hi external mass transfer coefficient
k oxidation reaction rate constant
MW molecular weight of the desired metal
CpT7 total concentration of silver
NOTATION R radius
f3-U total concentration of oxidizing agent RC radius of the shrinking core (oxidation reac-
bi stoichiometric coefficient of species i, includ- tion boundary)
 Mathematical modelling of metal leaching in nonporous minerals 1941

R, particle radius Lapidus. G. T., 1991, Modelado matemhtico de la lixiviacibn

total concentration of sulfate de argentita con sulfato f&&o y tiourea. Doctoral thesis.
[=-I
Universidad Aut6noma Metropolitana-Iztapalapa,
t time
Mexico.
l3-v total concentration of thiourea Lapidus, G. T. and Mosqueira, M. L., 1988, Effect of product
X fractional conversion or extraction solubility on the leaching kinetics of non-porous minerals.
Hydrometnliurgy u)(l), 49-64.
Greek letters Lev&spiel, O., iG72, Chemical Reaction Engineering, 2nd
Edition, pp. 361-377. Wiley, New York.
tl accumulation of the desired metal in the Palacios. E. G.. Lobe. R. A. and Laoidus. G. T.. 1989.
bulk solution Deter&naci& de1 olden de reaccibn-de 1; lixivia&n d;
& porosity of the inert layer esfalerita con cloruro f&rico, in Avarices en Metalurgia
leaching solution density Extractiva No Ferrosp1988 (Edited by S. Aguayo and
PI
L. A. Alma&Q, pp. 211-228. Universidad de Sonora,
Phf molar density of the desired metal in the
Hermosillo.
mineral Preisler. P. W. and Berger, L., 1947, Oxidation-reduction
%* percentage of solids in the system potentials of thioldithio sysiems: &iourea-formamidine
disulfide. J. Am. them. Sot. 69. 322-25.
Rath, P. C., Paramguru, R. K. and Jena, P. K., 1988, Kinetics
of dissotution of sulphide minerals in ferric chloride solu-
REFERENCES
tion. I. Dissolution of galena, sphalerite and chalcopyrite.
Bodlanger, G., 1896, Die chemie des Cyanidverfahrens. 2. Trans. Instn Min. Metall. (Sect. C: Mineral Process. Extr.
Angew. Chem. 9. 583-587. Metall.) 97, 150-158.
Chen, C. K., Long, T. N. and Wan, C. C., 1980, A study of the Schmid, G. M. and Curley-Fiorino, M. E., 1975, Gold, in
leaching of gold and silver by acidothioureation. nydro- Encyclopedia of Electrochemisrry of the Elements (Edited
metallurgy 5, 207-2 12. by A. J. Bard), Vol. IV, Chap. 3. Marcel Dekker, New
Hiskev._, J. B.._ 1981. Thiourea as lixiviant for aold and silver,
I
York.
in Gold and Silver: Leaching, Recovery and Economics Smith, R. M. and Martell, A. E., 1982, Critical Stability
(Edited bv W. J. Schlitt. W. C. Larson and J. 8. Hiskey), _. Constants. Vol. 5: Organic and Inorganic Complexes.
pp. 83-91: SME, Little&. Plenum &ss, New York.
Fuerstenau, M. C., Chen, C. C., Han, K. N. and Palmer, Taylor, P. M., de Mates, M. and Martins, G. P., 1983,
B. R., 1986, Kinetics of galena dissolution in ferric chloride Modelling of non-catalytic fluid-solid reactions: the quasi-
solutions. Metali. Trans. 17B. 415423. steady state. Met. Trans. 14B. 49-53.
Geankoplis, C. J., 1972, Mass Transport Phenomena, Chap. 3. Tellez, i., Rosas, G., Bedolla, E.- Y. and Tavera, F. J., 1986,
Holt, Rinehart & Winston, New York. Bstudio de la extracci6n de oro y plata de un mineral
Groenewald, T., 1975, Electrochemical studies on gold aurifero: lixiviacihn acida en tiourea-adsorci6n en
electrodes in acid solutions of thiourea containing carbbn-reducci6n carbot&mica, in Proceedings of 8th
gold@ j-thiourea complex ions. J. appl. Electrochem. Encuentro de Investiaacibn Metaltiraica. DD. 121-141.
5( 1). 71-78. Saltillo, MCxico, l&15 September. - Aa
Habashi, F., 1967, Kinetics and mechanism of gold and silver Warren. G. M.. Henein. H. and Jin. Z. M.. 1985. Reaction
dissolution in cyanide solution. State of Montana Bureau mechanism for the ferric chloride leaching of sphalerite.
of Mines and deology, Bulletin 59, April. Metali. Trans. 16B.
Lapidus, G. T., 1988, Modelaci6n matematica de la Yagi, S. and Kunii, D., 1955, Studies on combustion of
fixivir&n de oro y plata con cianuro en particulas min- carbon particles in flames and fluidized beds, in Proceed-
erales, in Avarices en Ingenieria Quimica (Edited by ings of 5th Znternational Symposium on Combustion, pp.
M. Vizcarra et al.), pp. 187-198. AMIDIQ, Mbxico. 231-244. Reinhold, New York.