You are on page 1of 8

Canadian

Metallurgical

Quarterly,

0008--4433/90 $3.00 + .00
Canadian Institute of Mining and Metallurgy
Pergamon Press pIc

Vol. 29, No.1, pp. 13-20, 1990

Printed in Great Britain

ACID PRESSURE OXIDATION OF ARSENOPYRITE:
PART II, REACTION KINETICS
V. G. PAPANGELAKIS
Department

and G. P. DEMOPOULOS

of Mining and Metallurgical Engineering, McGill University,
Quebec H3A 2A7, Canada

(Received 9 February 1989; and in revised/arm

28 September

Montreal,

1989)

Abstract-The
kinetics of pressure oxidation of narrow-sized arsenopyrite fractions (triclinic and monoclinic) was experimentally
determined in the temperature
range 120-180°C and pressure range of 2-20
atm 02' A surface reaction control shrinking core model was found adequately to describe the kinetic
results. The partial formation of molten elemental sulphur and the massive precipitation of scorodite was
found not to interfere with the progress of the heterogeneous
oxidation process. The dependence of the
apparent rate constant on temperature (EA = 66-72 kJ mol-I), oxygen pressure (first order), and particle
size (ks vs 11ro linear) were in agreement with the surface reaction control model. The true rate determining
step was postulated to be electrochemical in nature involving the first electron transfer during reduction
of the surface-adsorbed
oxygen. Finally the following intrinsic rate equation was developed to represent
the pressure oxidation of arsenopyrite in a batch reactor:
dN
--=49.527exp
Sdt

(-8672)
-T

Po molmm-.Icm-2
2

arsenate which precipitates

INTRODUCTION

3

Fe + (aq) +H3As04(aq)

Acid pressure oxidation is currently considered as the most
effective process for the treatment of refractory gold concentrates [1]. Arsenopyrite and pyrite are the most common
host minerals of refractory gold. From a process optimization
standpoint it is important to know the behaviour of each of
these minerals during acid pressure oxidation. Although the
aqueous oxidation of pyrite by molecular oxygen has been
investigated in the past by various researchers [2, 3] the same
is not true for arsenopyrite. As a part of an on-going research
program focussing on the chemistry, kinetics and modelling of
the pressure oxidation process the system FeAsS-H2S04-02
has been studied. The subject-matter
of the present paper, is
the kinetics of pressure oxidation of arsenopyrite. The reaction
chemistry of the system is described in Part I [4]. Other papers
report on the mathematical
modelling of the batch [5] and
continuous oxidation [6] of arsenopyrite.
Pressure oxidation of arsenopyrite has been found [4] to be
stoichiometric and to follow two parallel competing paths:
4FeAsS(s) + 1302(aq) +6H20(l)

--+

4FeAsS(s) + 702(aq) +8H+ (aq) +2H20(l)

--+

(1)

(2)

The SOforming reaction occurs to a significant degree (approx.
20% yield) only at temperatures below 160°C and in the presence of high sulphuric acid concentrations.
Arsenic reports in
solution directly as H3AsOiaq)
while ferrous ion is further
oxidized to yield ferric ion
2

2Fe + (aq) +102(aq)
Finally,

+2H+ (aq)

ferric ion combines

--+

3

2Fe + (aq) +H20(I).

with H3As04(aq)

FeAs04

2H20(s) + 3H+ (aq).

(4)

For comparison reasons the kinetics of pyrite oxidation is
briefly reviewed here in view of their common occurrence in
refractory gold concentrates. The pressure oxidation of pyrite
has been found [7-10] to be chemically controlled exhibiting
predominantly
first order dependence [7, 8] at the lower O2
pressure region ( < 20 atm) and half order dependence at higher
pressures [9] and temperatures [10]. Activation energies in the
order of 50 to 55 kJ mol- I have been reported for temperatures
below 160°C [7-9]. H2S04 was found by Bailey and Peters [9]
to have a beneficial effect on pyrite leaching kinetics while
McKay and Halpern [7], and Gerlach and coworkers [8] failed
to detect a similar effect in their respective studies. Apart from
the principal research papers quoted above, others have also
reported on the kinetics of pyrite oxidation [11-13]. Their
results however are not directly comparable to the present study
since temperatures
below the melting point of SO had been
employed or poorly identified pyrite specimens had been used.
The experimental kinetic results have been interpreted by early
investigators on the grounds of an oxygen chemisorption mechanism followed by a slow chemical reaction [7, 8, 10, 11]. However, Bailey and Peters [9] in a radical departure from previous
investigators have convincingly demonstrated
the mechanism
of pressure oxidation of pyrite to be electrochemical involving
coupled anodic (pyrite oxidation)
and cathodic
(oxygen
reduction) reactions.

4H3As04(aq)

+4Fe2+ (aq) +4S°(l)

+2H20(l)
--+

4H3As04(aq)

+4Fe2+ (aq) +SO~- (aq)

as crystalline scorodite

EXPERIMENT

(3)

AL

High-grade arsenopyrite mineral specimens of both mono-

to form ferric

clinic and triclinic structure were used in the present study. The
13

Scanning electron micrographs of the two arsenopyrite specimens: (a) monoclinic. Quartz was the only impurity phase identified (120/0in monoclinic and 1. acid concentration.) . 1013 kPa O2.21 solution. the following variables were then studied: stirring speed. From a number of tests. aliquots of leach liquor were withdrawn from the pressure reactor and analyzed for As(tot) and occasionally for Fe(tot) by AA spectrophotometry. PART II RESULTS As it was established in Part I of this work [4].0 0 1lJ 130 C 10 atm O2 0.PAPANGELAKIS and G.1 % solids. (Experimental conditions: Boac. More details on the experimental procedure have been given previously in Part I [4]. la). All the results reported here were conducted at a constant 0. Periodically. namely -147 ± 104 /lm. The massive mineral specimens after crushing. no· significant effect on the oxidation rate was found. the stirring speed of 730 rev. a slurry density of 0.30/0in triclinic). As was chosen as the main element to monitor the oxidation kinetics of FeAsS as a precaution against possible interference in the iron determination from corrosion of the sampling valve made of stainless steel.2 g of arsenopyrite was charged into the autoclave along with 220 ml of sulphuric acid solution. Results from these experiments expressed in terms of "fraction extracted". possible O2 mass transfer problems were effectively prevented thus securing the unambiguous measurement of the intrinsic reaction kinetics. throughout this series of experiments the slurry density was kept constant at 0.14 V. while the other variables were kept constant at 130°C. Lower speeds than 530. Utah (monoclinic variety) and Mexico (triclinic variety)./min were· not applied since this was found to be the minimum speed to give adequate solids suspension. rev. In that way the oxidation kinetics of FeAsS could be studied by monitoring simply the dissolved arsenic and/or iron concentrations and thus greatly facilitating the overall investigation. grinding and wet sieving were split into three narrow-sized fractions. and particle size. ~ ~ 930 rpm 730 o 530 c 0 30 ~ ft 60 TIME (min) 120 Fig. In Fig. correction factors were applied to account for the volume and mass losses due to sampling.1 % solids was found to give precipitation-free conditions.2 g FeAsS/0. Stirring speeds in the reactor were varied from 530 to 930 rev. 1 typical ground mineral particle morphologies are depicted. (b) triclinic. Thus. Over this range of stirring speeds. as it was determined by carrying out a series of agitation tests in glass liner that allowed visual observation. However. 1. oxygen pressure. No detectable sulphide/ arsenide impurity phases were present in either specimen. by using a low slurry density. t\ ~ 10 Fig./min. are shown in Fig. -74+ 53 . where the amount of arsenopyrite mineral leached was varied. 2.5 N H2S04 concentration. Temperature and pressure were kept constant throughout each experiment. Having thus established the slurry density for the kinetic investigations. 10 atm O2 and 0. temperature. All pressure oxidation tests were conducted in a 300 ml Parr mini autoclave with the internal parts made of titanium. The effect of stirring speed on arsenopyrite conversion. Typically for each experiment 0.5 N H2SO4 -200+270 mesh 0 I- u « 0:: I- x 1lJ g 0./min was chosen for all subsequent tests. It was decided thus at the early stage of the kinetic investigation to establish a slurry density that would not be associated with precipitation.5 z g 0 I- u « 0:: ll. For the calculation of conversion or equivalently "fraction extracted" (R). that is 0. Some quartz-totally liberatedparticles are seen to be present in the monoclinic specimen (Fig.urn. P. DEMOPOULOS: ACID PRESSURE OXIDATION OF ARSENOPYRITE: mineral specimens originated from Gold Hill. The acid generated by the oxi1. it is important to clarify that the "fraction extracted" used to illustrate the results in this paper is based principally on As measurement.44 + 37 /lm. G. In several instances "fraction extracted" based on Fe measurement was determined as well for comparison p~rposes. Based on these data. Furthermore.during pressure oxidation of arsenopyrite the highly insoluble scorodite compound forms. -74+ 53 /lm and . Before the effects of the other variables are presented.5 N H2S04. 2.

6.-.. .------. -74+53 !lm. 0. 5. high slurry densities .. The effect of acid concentration on arsenopyrite conversion. -74+53 !lm.. 1013 kPa 2.) D ° are employed which favour the precipitation of scorodite. The effect of partial oxygen pressure on arsenopyrite conversion. • __ • __ ~ 10 ~ e . 5.. 20 atm /. (Experimental conditions: 1013 kPa 02./.5 N H2S04._ ~ .) ° D Triclinic 10 30 60 TIME(min) Fig.. The effect of particle size on arsenopyrite conversion.---. z o · A/ 10 15 l- ~~ L~~v- N -200+270 mesh 0:: :---- z 9 I- 0.5 U X 0. G. The effect of partial oxygen pressure was determined for the range 202..0.) o to r---.0r-r--~"----. PAPANGELAKIS o and G.(Experimental conditions: 130 C.5 • • D dation of FeAsS itself was negligible in comparison to the 0. 6.-------. 0 0. All the results presented so far were established by using the monoclinic arsenopyrite mineral specimen. SEM examination showed only limited spread of molten sulphur over regions of semi-leached particles along with colonies of scorodite polycrystallites [4]. W w ACID PRESSURE OXIDATION OF ARSENOPYRITE: 10 30 60 TIM E (min) 5 2 120 Fig. 7. W I- U <{ 0:: l- x wO.4 kPa (20 atm) at 130°C. ~ ~ ~ 8~. 1013 kPa O2.5 130°C o 130°C 10 atm 730 rpm 200)(270 mesh PART II to . (Experimental conditions: 130 C.. r--------. 3 illustrate.5 N acid level since a very small mineral quantity was used. P.5 N H2S04. A number of tests were also carried out with the triclinic arsenopyrite variety./:_e.. (Experimental conditions: 1013 kPa 02. At the same time the surface of oxidized arsen- opyrite was found to be irregular exhibiting sites of high and low reactivity.. In industrial pressure oxidation practice.--e e_e~ :::::::. -74+ 53 !lm.25 010 • 60 TI ME (min) . 1013 kPa 2. It is readily apparent that finer material oxidized at a faster rate. 2 demonstrate the degree of reproducibility of the arsenopyrite pressure oxidation tests. 4 the test run at 130°C was duplicated.------.. 0.6 kPa (2 atm) to 2026. 7. The latter precipitate together with the produced elemental sulphur could hinder the progress of the oxidation reaction by covering the surfaces of the partially leached particles. All temperatures employed were higher than the melting point of elemental sulphur (119°C) which forms in the present system.-- 30 H2S04 x o~: ~ ~.50N H2S04 A 0..~ /. Lower acidity levels were investigated as well and were found to result in somewhat reduced reaction rates as the results of Fig. The obtained results are illustrated in Fig.. wO.. The obtained results are given in Fig. temperatures above 150°C are required to result in complete oxidation of arsenopyrite within the 2 h reaction time employed.---r-~-"'-----r-----. ---.. As it can be seen.. the effect of particle size is shown in Fig.) 10 30 60 120 TIME (min) Fig.~. DEMOPOULOS: I- U <{ 0:: l- W I- <{ U <{ 0:: lJ.---r--r--..5 60 TIME 120 (min) Fig.) .5 N H2S04. The effect of temperature on triclinic arsenopyrite conversion.-. It is clearly shown that an increase in oxygen pressure favours the overall oxidation kinetics. o~. 120 Fig. These duplicated results along with the results of Fig. It is evident that the leaching rates are strongly temperature dependent.. 1.--.~ /.V. 4. Finally... (Experimental conditions: 130 C.---r--~___r'--r---r------r-----__.5 N H2S04.. In Fig. Figure 4 illustrates the effect of temperature on arsenopyrite oxidation at 10 atm O2 pressure. The effectof temperature on arsenopyriteconversion. -74+53 !lm. 3. 0.

= 131\T m·-k' Pb~. number of moles of oxygen at time t r nidius of unreacted core at time t k' apparent intrinsic rate constant (= Khll k) P o~ partial pressure of O2 From the stoichiometry The concentration of dissolved oxygen is considered to be proportional to its partial pressure according to Henry's law [16] i. I POry. where R is the "fraction extracted" at time t and ks. with eqs. arsenopyrite by the following equation fraction extracted. The pressure instead of the dissolved oxygen concentration is used here which implies that Henry's constant is incorporated into the intrinsic rate constant.16 V. On the other hand. Second order polynomial regression was performed to fit the experimental data of Fig. The shrinking core "inodel equations can be applied to the present heterogeneous reaction system provided that some basic assumptions are met. l-1R-(1-R)2/3 for product centration in the bulk of the solution remains constant. The rate of arsenopyrite oxidation expressed as moles of oxygen disappearing per unit time. II and (14) o A plot of In dRjdt at t = 0 versus In P o~ would result in a straight line with slope a. the obtained experimental data are analysed in this section with the aid of shrinking core models which have been previously established [14. (b) the product layer is porous. IS]. lVO-1V R=-- where b the arsenopyrite stoichiometric coefficient. G.~ dt (7) for mixed control.e. per particle is dN layer diffusion control and = 4nr2k' P/ o~ . P. (c) the con- I of reaction dl\To~ 13dlV dt 4 dt ( 12) where N number of moles of FeAsS present at time t. Disregarding diffusion control through the external boundary layer (see Fig. 2). mixed control). equation 14 is obtained dR dt 13 and combining 16 . These constants are given in equations 8. the mixed-control model equation 7 is applicable provided the reaction is of first order and the pseudosteady state assumption holds [15. M molecular weight of FeAsS p density of FeAsS ro initial particle radius k intrinsic rate constant Dc effective diffusivity of 02(aq) through the product layer [02] bulk concentration of dissolved oxygen a reaction order in terms of dissolved oxygen concentration.Ls) [4] assumption (d) is believed to be met in view of the very dilute slurry employed. 5. In addition. DEMOPOULOS: ACID PRESSURE OXIDATION OF ARSENOPYRITE: DISCUSSION Shrinking core models The object of the present kinetic study was primarily to develop a rate equation useful for reactor design and process modelling. Moreover a mechanistic interpretation of the kinetic results is presented. By differentiating equation 12. Initial reaction order Before the shrinking core model equations are applied to the experimental kinetic data the order of the initial reaction rate with respect to the partial pressure of oxygen will be established. (d) the telnperature is uniform throughout the heterogeneous reaction zone. 9 and 10. (11) where (9) k PART II I\T 0 R. In the present system the product layer may consist of molten SO and scorodite particles adhering on the unreacted mineral surfaces.e. . kD and km are apparent rate constants. is given ( 13) where No initial number of moles of arsenopyrite. PAPANGELAKIS and G. The pressure was kept constant. [02] = Kt. To this end. 17]. In the case of spherical (or more precisely equiaxed) particles the following shrinking core models are available for single particles and constant solution concentration [IS] 1-(1-R)1/3 = kst (5) = kDt (6) for surface reaction control. By taking the logarithms of the initial slopes (SJ of the regression polynomials at t = 0 and plotting them against the logarithms of the respective oxygen pressures the graph of Fig. These are: (a) the particle retains its equiaxed shape. bkM[02Y ks=---pro (8) = bkM[02] m pro· (10) Moreover No. The derivatives of R are taken at t = 0 since no elemental sulphur has yet been formed to interfere with the oxidation reaction. then the pressure oxidation of arsenopyrite is expected to be controlled by either the surface reaction or the diffusion process through the product layer or combination of the two steps (i.-c. Finally despite the high exothermic nature of the oxidation reaction I (~H~9S = 1430 kJ molj. equal to ~ since reaction 1 dominates in the present system.

8 was obtained.) 60 120 TIME (min) Fig. it appears that the minor liquid sulphur formation (10-20% yield [4]) especially at the lower temperature range investigated. which represents the results of Fig.---'-_. in which precipitation had taken place.----JI. In other words. gave excellent agreement with the experimental data. The linear plot obtained is given in Fig.0 2. liquid sulphur does not form a protective impervious layer around the arsenopyrite particles. where liquid sulphur has been found essentially to stop the reaction below the 1000/0oxidation level [9]. 0. 1013 kPa O2.p.(1. The linearity of the "1. Plots of the shrinkingcore--ehemicalreactioncontrolmodelat different temperatures. of concern to the present reaction system is also the possible effect of precipitation of ferric arsenate could have on the progress of the oxidation reaction.1 0:0.--..0 --=::.1 % solids) experiment is plotted for the purpose of comparison. This is in contrast to the pyrite pressure oxidation system. The totally different behaviour of the two minerals can be explained by comparing their respective Pilling-Bedwarth 17°~/'60/ · .L. . ----.----L 10 60 30 TIME for arsenopyrite. does not interfere with the reaction kinetics of arsenopyrite oxidation.5 N H2S04. it can be concluded that ferric arsenate precipitation does not interfere with the overall oxidation kinetics of arsenopyrite. /. in the case of FeAsS (z < 1) the produced sulphur does not suffice to form a similar protective layer thus allowing the reaction to proceed to completion without any hindrance. data from a precipitation-free (0... Application of the shrinking core-reaction control model to the experiments represented by Figs 2 and 5 published in Part I [4]. Plot of the shrinking core-ehemical 240 reaction control model for differentslurry densities. This observation supports the fundamental assumption made in the early stages of the present investigation to monitor the kinetics of the oxidation reaction under precipitation-free conditions. The resulting straight line (correlation coef. the z values are 20 / o PART II z = 2 x 16.&. 9. 119°C). 0.. (Data from Fig. 9.6 n ~0. 11) was constructed for both the monoclinic (Fig. It can be clearly seen that z is greater than unity in the case of FeS2 (assuming that all sulphidic sulphur oxidizes to SO)and this can result in a sulphur layer enveloping completely the reacting pyrite particle. defined as the volume of insoluble product formed per unit volume of solid reactant.36 for pyrite and -1 z = 16. -74+53 }lm..--. P. . For sulphur. The other two model equations 6 and 7 failed to yield satisfactory correlation.---. . 0. Since the ks values are directly proportional to the intrinsic rate constant (eqn 8) and the data represent experimental runs of the same particle size 00.. 7) arsenopyrite oxidation test results.0 lnP02 Fig./ .~ /..-0. Surface reaction control Analysis of the arsenopyrite pressure oxidation results (Figs 2-7) indicated that all were in very good correlation with equation 5.2 I 130°C • __ % ds YO. Apart from the elemental sulphur. However.-.2 • • // / lwe 0. the value refers to crystalline monoclinic Sf3 (m. PAPANGELAKIS and G.62 26. From these findings..(1. From the slopes of the "1.5 -to 1200C • 2.R) 1/3 vs time" plots the ks values were derived and the Arrhenius plot (Fig.06 -2 1. G.5 cm molpeAss (j) c n = l06:!: 0. Determination of the reaction order with respect to oxygen pressure by considering initial kinetics. In the same Figure.) .R) 1/3 vs time" plots suggests that the reaction controlling step of the pressure oxidation of arsenopyrite is the chemical reaction taking place at the surface of the mineral..997) has a slope of 1 indicating that the reaction is of first order.3 cm3 g-atomsol 24 3 1-1 cm mo FeS2 = 1. DEMOPOULOS: 2 10 ACID PRESSURE OXIDATION OF ARSENOPYRITE: 17 ratios [18].5 -'-- --' 120 (min) Fig.. 4./.V.z. -----.4 140·C . / /. Thus./ / .. 10.--.3 cm3 g-atomso1 ./ _ . 4) and triclinic (Fig. 4.. 3 I = 0. The molar volume values were calculated from Weast [19]. ~ :::::::::::::::--./~.(Experimentalconditions: Boac. At temperatures above 119°C the molar volume would increase but it is not expected to significantly affect the z values. 8..4 0: • or- 0.0 0.J. Typically the agreement obtained is illustrated in Fig. 10./. Thus. __ ..5 3.

2 /_ •..-::-~-· 10 6.4 kPa).8 kJ/mol • EA=65. _. 3 were similarly analyzed with the aid of the shrinking core model and the dependence of the apparent rate constant (ks) on H+ concentration was found to be 0.1±5.998 for the tests involving the triclinic mineral variety. First order dependence on pressure has also been reported for pyrite oxidation at temperatures and pressures in the range of 25160°C and 0. This plot confirms that pressure -6 V) oX c.-__ .I (17.0 ~I.995). where 1'0 is the geometric average radius of the particles corresponding to each size fraction.) • triclinic .18 V.2 kcal mol-I) with a correlation coef..9 kcal mol-I) with a correlation coefficient of 0.0. 13) a linear plot with slope essentially 1 (correlation coefficient 0. Thus by plotting the logarithm of ks versus the logarithm of oxygen pressure (Fig. The Arrhenius plot revealed an activation energy of 72.. Fig.5±4.527exp(-8672) --T- . shows stronger temperature dependence for the former mineral. • •.• ./ 10 .0 50 TI ME 20 .-1/ . 11.. 5.03 1. ~0. G. No tendency towards fractional order was detected in the present arsenopyrite oxidation system as it is testified by the plots of Figs 8 and 13.5 1/ T (K-~103) Fig..9 kJ/mol /. 12. . (Data from Fig. The results of Fig. Such a "normalization" plot is shown in Fig.. (Data from Figs 4 and 7.) (-74+ 53 flm). Finally the explicit expression of the intrinsic rate constant for the pressure oxidation of FeAsS (monoclinic) becomes k'=49.8 2.1 kJ mol. Bailey and Peters [9] (15) Po (atm) .5 order dependence has been reported for pyrite [9]. This range of activation energy supports the view that the pressure oxidation of arsenopyrite is controlled by the rate of the chemical reaction at the surface of the particles. Particle size and acid effects Considering equations 5 and 8 and for a given temperature and oxygen pressure.._. However. 8.988) was obtained.3 (correlation coefficient 0. For the estimation of free H+ concentration. 0...-- 5 .-7 -8 05 / / • 20 / n= 0..-/. 5 2 10 Overall oxygen pressure dependence The linear plots of Fig. 14.I). /' ~. plots of 1'0[1. for higher pressures a 0. ..4 • EA=72.2 30 • monoclinic • PART II 23 2A 2.0 3. Icm-2atm-I molmln- oxidation of arsenopyrite is first order with respect to oxygen pressure for the whole pressure range studied (2-20 atm).2-10 atm.. since its second dissociation constant is only 8 x 10-3 at 130°C [21].--g---." 6.- 2 atm 120 (min) Fig. The consistency of the first order dependence on pressure which was determined using initial rates (Fig.Arrhenius plots for monoclinic and triclinic arsenopyrite. PAPANGELAKIS and G.0 1nPo.. are plotted in terms of the chemical reaction control model (eqn 5). As it is seen.. P.5 kJ mol-I (15. DEMOPOULOS: TEMPERATURE (Oe) 160 140 180 ACID PRESSURE OXIDATION OF ARSENOPYRITE: 120 0. correlation of the three size fractions is very good.90! 0. 12. The straight lines obtained indicate that the reaction mechanism does not change within the oxygen pressure range of 2-20 atm (202._.~. Here the results of Fig. The pertinent data are given in Table 1. Determination of the reaction order with respect to oxygen pressure by considering overall kinetics. / . The pre-exponential factor of the intrinsic rate constant was also determined with the aid of the Arrhenius plot (monoclinic specimen).~ /. the activation energy calculated is directly related to the intrinsic kinetics.R) 113] vs time would be expected to give a common straight line for the different particle sizes investigated. Comparison of the activation energy determined for the pressure oxidation of arsenopyrite with those reported for pyrite oxidation (50-55 kJ mol. representing experiments run at various partial oxygen pressures.6-2026.988 for the tests involving the monoclinic arsenopyrite specimen and 66.(1. Plots of the shrinking core-ehemical reaction control model at different oxygen pressures.. H2S04 was taken as being 500/0dissociated. 5. 13. respectively [7. ~. 20].0 2. . 8) was verified with the shrinking core model results of Fig.. 12 add further support to the finding that the oxidation process is chemically controlled..

125 0. = -4Fk 8[H+] Ie 2 Similarly for the anodic oxidation be shown that exp -acFI1¢e fYlT (21) . 8 fraction of arsenopyrite surface sites covered by dissolved oxygen (surface coverage by intermediates is neglected).5 0. Arsenopyrite is also a semiconducting mineral (rest potential = 0. Among the several reaction paths proposed previously to account for the electro catalytic reduction of O2 [28. is considered here. The behaviour of arsenopyrite as an electrocatalyst for oxygen reduction has been shown to be similar to that of other sulphide minerals such as pyrite [27].. the following equation applies (since reactions 16 and 17 do not carry the same number of electrons) 4ia = -13ie which yields 02(aq) + 2FeAsS(s) ~ 2FeAsS(0)ads (18) (23) 2.31xl0-3 ks The first electron transfer step (reaction 19) is taken as the rate determining step. '0 •••-147+104 0-74+53 0-44 + 37 and G.40 x 10-3 1. f!lt.5 V [23]) has been interpreted as an electrochemically controlled process [24]. Similarly the pressure leaching of chalcopyrite (rest potential 0.94 x 10-3 2.) also found that increased H 2S0 4 concentrations eficial effect on pyrite pressure leaching. of FeAsS (reaction 16) it can (22) where ie' ia cathodic and anodic current densities k2 and ka forward rate constants for cathodic (19) and anodic (16) reactions._ 3.. 14. gas constant and temperature respectively. At the mixed potentiall1¢m. the last equation becomes (25) Equation 25 suggests the reaction rate to have half-order depen- .1 0. 29] the following is postulated as representing the present system 1. Normalization plot of the shrinking core model. The surface of arsenopyrite is envisaged to consist of anodic and cathodic sites on which the following two principal reactions* are believed to occur FeAsS(s) +8H20(l) ~ Fe2+ (aq) +H3As04(aq) + SO~. l1¢a cathodic and anodic potentials F. (24) a 2(FeAsS(OH)ads+H+ +e~ ~FeAsS+H20) (20) Assuming aa * Only the sulphate-forming anodic reaction. Second electron-transfer Ia - 13Fka 22 gives (k 28[H+])aa/aa+ac k . aa transfer coefficients for the forward cathodic and anodic reactions respectively l1¢e. which dominates at high temperatures. G. and T Faraday constant.(aq) + 13H+ (aq) + 13e- 02(aq) +4H+ (aq) +4e- ~ 2H20(l). The dependence of rate constant ks on H+ concentration at 130aC ~m [H2SO4] Electrochemical PART II + H+ +e. (16) (17) The cathodic reduction of dissolved molecular oxygen at the mineral surface according to reaction 17 is considered the key process affecting the overall reaction kinetics.58 [25]) varying from p-type to n-type [26].. P.25 1. 6. DEMOPOULOS: ACID PRESSURE OXIDATION OF ARSENOPYRITE: (In ks = 0. PAPANGELAKIS E :::L. 30]. .25 0. the possible influence of the Fe 3+ /Fe2+ couple on the overall mechanism via the alteration of the mixed potential on the mineral surface is neglected. First electron-transfer 2(FeAsS(0)ads 19 Table 1.V.. On the other hand in high pressure leaching systems involving pyrite [9] and chalcopyrite [24] a similar effect with the present study has been found.~ FeAsS(OH)ads) (19) Applying equation 23 to equation . This is in agreement with the positive effect of H+ concentration on the oxidation kinetics (see Table 1). l1¢m = l1¢e = l1¢a.62 V [22]) has been well established to be electrochemically controlled [9]. = ac. Alternative mechanisms call for negative or no effect at all as it has been shown in the case of low pressure ( < 1 atm) pyrite systems [20. have a ben- mechanism The pressure oxidation of pyrite (rest potential 0. The rate of reaction 19 with the aid of the Butler-Volmer equation (high field approximation) [29] can be expressed as follows . (Data from Fig.3 In [H+]) 120 60 30 TIME (min) Fig. It is therefore a reasonable assumption that its oxidation mechanism is of electrochemical nature in a similar fashion with that of pyrite. [H+] concentration of H+ ae..05 0. Chemisorption of O2 (N) [H+] (M) min-1 0.

N. Warren. 60. The reaction order with respect to oxygen partial pressure was found to be 1 by considering both initial and overall kinetics. 26. p. CONCLUSIONS The pressure oxidation kinetics of arsenopyrite has been studied in the temperature range 120-180°C and 2-20 atm O2 pressure. Trans. K. and P 02 the partial oxygen pressure in the reactor. Biegler. McKay and J. Y. W. H. J. Trans. J. 11. New York (1981). Bautista. 333 (1976). 13. J. of l\1in. H. New York (1969). to represent the aqueous pressure oxidation of arsenopyrite in a batch reactor (sulphate-forming reaction path 1). Des. I (1990).) is also gratefully acknowledged. Chemical Reaction Engineering. 205. dN . 29. l\1innich. R. E. Lin. p. B148. PA (1986). R. 9. A Freundlich type isotherm of the form 8 = kjP6. 21. G. T. Int. 25. Mathews and R. (1972). M.65 (1958). 15. thus further supporting the conclusion that the oxidation process is chemically controlled. OH (1985).527 exp (-8672) --T- . W. Intelf Electl'oeil. J\1etal. \Veast. V. Q. Bockris and B. 31. R. \V. Del'. 14. Can. C. G. J.P. Pray. p. Proc. I. O. Derelopment alld Plant Practice (edited by K. Chem. O'M. Bailey and E. D. Proc. Hydrometallurgy: Research. Rand and \VeIch). 1146 (1952). 642. G. Demopoulos. p. D. R. W. E. Acta 24. A. Neither the liquid elemental sulphur that forms to a certain degree nor the precipitation of scorodite which occurs with high slurry densities appear to interfere with the progress of the heterogeneous oxidation process. \Vadsworth and S. 23. R. E.(1. Engng 15. V. J. E. C. Demopoulos. Pawlek. Rev. Rand and R. B. J. P. 32. P02mol mIn-I cm-2 where N is the number of moles of arsenopyrite. Given that no information on the exact form of this isotherm is available for the present high temperature three phase system and that the comparison with experiment ignores the contribution from the minor reaction path 2 and that a number of simplifying assumptions are inherent to the electrochemical kinetic theory no further analysis is attempted. PART II work was supported by an NSERC Operating Grant. Schlitt. L. 27. J. 22. Demopoulos and V. D. H. Bockris. Amsterdam (1975). Freundlich) are available to describe the relationship between 8 and the applied partial oxygen pressure [31. Oxford (1987). 347. H. Schlitt and J. (1982). E. E.3. L. P.3 for [H+] and 0. M. 19 (1974). A. J. 4. B. Osseo-Asare and J. Damianovic. El-Raghy. Pergamon Press. Q. p. Berk and G. The experimentally determined orders were found to be 0. A. Peters. R. J\1etal. Can. N. Trends in Electrochemistry (edited by J. Bockris and B. Bedworth. Plenum Press. \Vadsworth. Gerlach. Acknowledgement-This REFERENCES 1. 415 (1979). 185. Sohn and M. J. 87 (1987). K. O'M. 12. Cleveland. M. p. Engng Chem. which reduces to first order behaviour (eqn 25) may be assum-ed to apply. Modern Aspects of Electrochemistry (edited by J. W. 3rd edn. appl.9 for p02' Different adsorption isothermes (Langmuir. 28. Majima. Handbook of Chemistry alld Physics. 16. Ind. metall. B (1990) in press. G. . 261. 5. 18. Hiskey). Miller). Co. 7. T. Lowson. Electrocllim. P. Trans. No. J. TMS-AIME. TMS-AIME. J. Aletall. 991 and 1251. \Vyslowzil and G. Int. Marshall and E. J. ro. E. CRC Press. McGraw-Hill. B 148. E. W. Jones. 370. 357. J. \Varrendale. O'M. Hiskey and W. However the proposed mechanism offers satisfactory explanation as to the effect of H+ on the overall oxidation kinetics of arsenopyrite and the measured activation energies in the range of 66-72 kJ mol. Pilling and R. 55. p. 719 (1980). P. W. New York (1976). \Varrendale. Aust. J. based on the surface reaction control model. R. M. 149.461 (1982). Peters and H. Conway). Biegler. p. J. Levenspiel. Chemical and Catalytic Reaction Engineering. t is the time. J. D. Biegler and D. 17. de Matos and G. New York (1964). Elsevier. The acid concentration was also found to have a beneficial effect on the oxidation kinetics. G. H. l\1etals 29. Modern Aspects of Electrochemistry (edited by J. V. P. Temkin.S dt = 49. Inst. Chem. S the total surface area of FeAsS particles. No. Smith. 8. Chem. Inst.5.R) 1/3] vs time resulted in one straight line for various particle size fractions. Semiconducting Ore A1inerals p.44. \Voods. Gileadi and B. 66th edn. Rand and R. 15. Hahne and F. Finally an electrochemical reaction mechanism has been advanced involving chemisorption of oxygen on arsenopyrite surface followed by two single-electron transfer steps. New York (1970). PA (1983). 19. p. Reddy. Warren). 20. Poling.US-AlJ\1E6. G. McDonald). DEMOPOULOS: ACID PRESSURE OXIDATION OF ARSENOPYRITE: dency on the hydrogen ion concentration and on the surface concentration of molecular oxygen (8). B-198. metall. 2. Ind. 24. Papangelakis.20 V. G.529 (1923). P. p. T. J. Symp. Tl\1S-AIME. C. Sci. Hydrometallurgical Reactor Design and Kinetics (edited by R. A. p. M. Demopoulos. \Varrendale.G. Littleton. E. 1Uodern Electroc!lemistry. PA (1986). Finally normalization plots of the shrinking core model in the form of ro[1. Processing 20. New York (1977). D. Chem. A. Schweickert and B. 82. Hydrometallurgical Reactor Design and Kinetics (edited by R. E. Er=metallI9. 151 (1975). 29. 30. Aust. 204. Plenum Press. T. 378.4028 (1966). Shuey. V. Plenum Press. T. Warren. \Voodcock. l\1in. King and J. V. Robins. Papangelakis and G. The activation energy was determined to be in the range of 66 (triclinic arsenopyrite) to 72 (monoclinic arsenopyrite) kJ mol-I for the whole temperature range investigated. D. Part I. The first electron transfer is postulated as being the rate determining step. D. 341. Proc. 6. Cornelius and J. Halpern.I further support the postulated activation control. O'M. J. A.49 (1983). 66 (1966). S. Engng Chem. Warren). R. A. K. SMEAIME. Carberry. G. Conway). 32]. H. 19. P. Plenum Press. Berk and G. Imelfacing Technologies ill Sollition A1ining (edited by W. Aust. 3. Bautista. T. SaIter. Lewis. 291. Beattie and G. 7. on Gold 1\1etalllirgy (edited by R. 346 (1956). Papangelakis. Taylor. M. A McGill University scholarship awarded to one ofus (V. Bockris and A. Papangelakis. The oxidation kinetics was found to follow a shrinking core model with the surface chemical reaction. PAPANGELAKIS and G. 111 (1968). Having determined the activation energy and reaction order in terms of oxygen pressure the following intrinsic rate equation was developed. p. Conway. B-201. D. Martins. 7. Can. Electroanalyt. 70. Pill'S. J. metall: Q. Chemical Engineering Kinetics. T. \Vesley and G. B. 301 (1958). Woods. John \Viley. T. as the rate controlling step. Swift. p. Wesley and G. G. G. McGraw-Hill.