Ind. Eng. Chem. Res. 1988,27, 571-576
57 1
Kinetic Study of Oxidation of Pyrite Slurry by Ferric Chloride Koei Kawakami,*tJunko Sato,t Koichiro Kusunoki,+Katsuki Kusakabe,*and Shigeharu Morookat Department of Chemical Engineering and Department of Applied Chemistry, Kyushu University, Fukuoka 812, Japan
The aqueous oxidation of pulverized pyrite by ferric ion was investigated in a concentrated, acidic chloride solution. The ratio of the concentration of sulfate ion to elemental sulfur produced during the reaction was approximately 1.3, regardless of the leaching conditions such as temperature and acid concentration. The oxidation rate increased with decreasing acid concentration and increasing initial ferric chloride concentration and was retarded by an addition of ferrous ion. It was also found that the oxidation rate was not linearly related to the pyrite loading and the reciprocal of the particle diameter. A kinetic model was proposed assuming that fine mineral powder and irregular pits on the outer surface of the crushed pyrite are responsible for the high initial reaction rates. The kinetic expression derived in the Langmuir-Hinshelwood form gave good agreement with the experimental results. by the same form as eq 3. They reported that the oxidation rate decreased with decreasing pH and was proportional to pyrite loading and surface area. King and Perlmutter (1977) measured the rate of pyrite oxidation in concentrated (0.1 and 1.0 km01.m-~)ferric chloride solutions using a stirred slurry reactor. They found possible occurrence of the sulfur side reaction 1from the simultaneous measurement of total ferric and ferrous ion concentrations. The effects of temperature, pyrite loading, and particle size were investigated. The experimental data were empirically correlated by the following two-parameter kinetic model:
The oxidation of pyrite in a ferric iron salt solution is of great importance in hydrometallurgical processing for sulfide ores. The treatment of pulverized coal with an aqueous ferric sulfate or chloride solution as a leaching agent has also been proposed for removal of inorganic sulfur (pyrite) from coal (Meyers, 1977). The oxidation of pyrite by ferric ion in the absence of oxygen is reported to produce sulfate ion as well as elemental sulfur according to the following reactions (Hamersma et al., 1973; King and Perlmutter, 1977; Lowson, 1982): FeS2
-
+ 2Fe3+
FeS2 + 14Fe3++ 8H20
-
3Fe2+ + 2s
(1)
15Fe2++ 2SOd2-+ 16H+ (2)
The chemical equilibrium in an acidic solution of iron salts is very complicated, and there exists a number of complexes between iron ions and anions depending on ionic strength of the solution (Rabinowitch and Stockmayer, 1942; Helgeson, 1969; Strahm et al., 1979). From survey of the literature, it seems that, under the condition of dilute solution of ferric salt, sulfate ion is the only sulfur product (Garrels and Thompson, 1960; Mathews and Robins, 1972; McKibben and Barnes, 1986). Meanwhile, at higher concentrations of ferric ion, e.g., more than 0.1 k m ~ l - m -the ~ , pyrite appears to be oxidized to produce simultaneously both sulfur and sulfate according to reactions 1and 2 (Hamenma et al., 1973; King and Perlmutter, 1977). However, there is still disagreement in the literature as to kinetic expressions derived. Garrels and Thompson (1960) studied the pyrite oxidation in dilute ferric sulfate solutions (10-5-10-3 k~nol-m-~) and found the stoichiometry to be consistent with eq 2. It was also indicated that the rate was controlled by competitive adsorption of ferrous and ferric ions on the pyrite surface and was proportional to the fraction of pyrite surface occupied by ferric ions, leading to the rate expression k[Fe(III)] r= (3) [Fe(III)] + [Fe(II)] Mathews and Robins (1972) studied the pyrite oxidation with dilute ferric sulfate solution (-0.01 k m ~ l - m - using ~) a gas-lift percolator apparatus. The overall stoichiometry of eq 2 was confirmed, and the rate equation was expressed 'Department of Chemical Engineering. 1Department of Applied Chemistry. 0888-5885/88/2627-0571$01.50/0
Zheng et al. (1986) studied the kinetics of the pyrite oxidation in aqueous ferric sulfate using a packed-bed reactor. For the calculation of the oxidation rate, the stoichiometry according to eq 2 was assumed. The experimental data under a constant acid concentration were found to best fit the Hougen-Watson dual-site adsorption model for total ferric and ferrous ions, with the rate expression given by
r=
k1[Fe(III)]1/2- k2[Fe(II)]1/2 1
+ K3[Fe(III)]1/2+ K4[Fe(II)]1/2
(5)
Those authors also developed a reaction model and rate expression in terms of the concentrations of free ferric ions, ferrous ions, and sulfate ion by considering a number of chemical equilibria which exist among many ions in solution. Recently, McKibben and Barnes (1986) studied the oxidation of pyrite with millimolar levels of ferric ion in acidic solutions and determined the rate law, indicating that the rate was proportional to the square root of free ferric ion concentration. From their SEM study, it was concluded that the reactive surface area was substantially different than the total surface area, and the oxidation occurred preferentially on sites such as grain edges and corners, defects, solid and fluid inclusion pits, cleavages, and fractures. In this paper, the kinetics of pyrite oxidation with a concentrated solution of ferric chloride was extensively studied as a function of operating variables such as temperature, acid concentration, pyrite loading, particle size, and initial concentration of ferric and ferrous chloride. For
0 1988 American Chemical Society
572 Ind. Eng. Chem. Res., Vol. 27, No. 4, 1988 363K
0-0-
333K
--o---+-
Conc. of HCI, 0.1 k m ~ l , m - ~ Init. Conc. of Fe(lll1, 0.5 kmol.ni3 Slurry Conc.. 1 2 kg.m3 Particle Size. - 53 F m
?
0
0
12
36
24
40
Temp., 333 K Init. Conc. of Fe(1U). 0.5km0l.m-~ Slurry Conc.. 1 2 k g . d 53 p m Fbrticle Size,
60
0
-
12
24
60 I
O f 0
36
40
60
40
60
Time [ h l
Time [ h l
1
I 12
24 Time
36
40
60
[ h l
Figure 1. Effect of temperature on formation of (a) Fe(I1) ion and (b) sulfate ion and elemental sulfur.
a practical design, a kinetic model was developed to explain a wide range of experimental results.
Experimental Section The pyrite sample used was from Yanahara, Japan, and was obtained from Nippon Chikagaku-sha Co. in the powder form under 500 pm. Chemical analysis of the pyrite indicated 46.3% Fe and 52.5% S on a weight basis. All the experiments were carried out in a well-stirred batch reactor (75 mm i.d. and 115 mm high), which was made of glass and equipped with four baffles. Agitation was provided by impellers with four flat paddles at a stirring rate of 800 rpm. The pyrite slurry was prepared by using 0.01-1.0 k m ~ l - mHC1 - ~ as a diluent medium. The reaction was initiated by adding a concentrated ferric chloride solution of 50 cm3 in a pyrite slurry of 250 cm3. The final concentration of ferric chloride was 0.05-1.0 km~i.m-~. The nitrogen gas was continuously bubbled into the pyrite slurry at a low flow rate to exclude atmospheric oxygen and was vented through a water-cooled condenser. Slurry samples were withdrawn at appropriate time intervals. The concentration of sulfate ion in the solution was measured by ion chromatography (TOY0 SODA IC-Anion PW column). The concentration of Fe(I1) ion was measured by absorption spectroscopy using o-phenanthroline. The residual pyrite was contacted with toluene to extract the elemental sulfur remaining on the pyrite surface, and
0
12
24
36 Time [ h l
Figure 2. Effect of acid concentration on formation of (a) Fe(I1) ion and (b) sulfate ion and elemental sulfur.
the sulfur content was determined by the combustion method (JIS M8813).
Results and Discussion Figure 1 shows the effect of temperature on the time course of the concentrations of oxidation products, i.e., Fe(I1) ion, sulfate ion, and elemental sulfur left on the pyrite surface. The oxidation rate increased significantly with increasing temperature. The apparent activation energy was about 95 kJ-mol-' for each product formation. This value is consistent with an activation energy of 92 kJ-mol-' reported by Mathews and Robins (1972). The effect of concentration of HC1 on the oxidation is shown in Figure 2. The formation rate of each oxidation product decreased with increasing acid concentration. Such an observation is dso reported in the literature (Mathews and Robins, 1972; McKibben and Barnes, 1986). A concentrated ferric chloride solution contains various complexes, such as FeC12+and FeC12+,other than free ferric ion, Fe3+, and the equilibrium among these ions depends on the total ferric and chloride ions concentrations (Strahm et al., 1979). Therefore, the decreased oxidation rate with increasing HCl concentration may be closely associated with increasing complex ions bearing less strong oxidizing power. Figure 3 shows the effect of pyrite loading on the formation of Fe(I1) ion. The abscissa of the figure is given
Ind. Eng. Chem. Res., Vol. 27, No. 4, 1988 573
I 0 3
I
Temp, 348K Conc. of HCI, 0 1 kmol m-3 Slurry Conc, 10 kg.m-3 Particle Slze. 3 7 - 53 i m
(a)
Keys are shown be!ow
-
-
Temp.. 348K Conc. of HCI, 0.1 km01.m.~ lnit Corc.of Fe(lll), 0.5 kmo1.ni3 Particle Size, - 53 pm
0
1200
2400
3600
+!
4800
6000
7200
[rmn kg m-31
Figure 3. Effect of slurry concentration on formation of Fe(I1) ion.
0
0
60
120
180
240
300
360
2LO
300
360
240
300
360
Time [minl
30
-
n
'E
20
I n i t . Conc. of Fe(l0). 0 2 kmo1.m') Slurry Conc., 6 67 kg.m-3
1-
I
v)
Fhrticle Size
"0.. C
P
-5 IO p C
I
u"
100 150 200 t I d, [ min,pm-'l Figure 4. Effect of particle size on formation of Fe(I1) ion. 0
50
by the product of pyrite loading, wp, and reaction time. If the formation rate of Fe(II) ion is proportional to the pyrite loading, the time-course data should be correlated by a single curve. However, the result shows that this is not the case, reflecting a more complicated nature of the reaction. The effect of particle size is shown in Figure 4. In this case, the time course of the formation of Fe(I1) ion is plotted versus the product of the reciprocal of median particle diameter for each sieved fraction, l / d , and reaction time. This result also shows a loss of Pinearlity between reaction rate and l/dp, Le., external surface area of the initial pyrite. Figure 5 shows the time course of the concentrations of Fe(I1) ion and sulfate ion at various initial concentrations of Fe(II1) and Fe(I1) ions. The formation rates increased with increasing initial concentration of Fe(II1) ion from 0.05 to 0.5 k m ~ l e m and - ~ leveled off at the highest initial This may also be due to concentration of 1.0 km~l.m-~. an increase in the proportion of less reactive Fe3+-C1complexes with increasing concentration of ferric chloride. The formation of Fe(I1) and sulfate ions was appreciably inhibited by the addition of Fe(I1) ion. Typical SEM photomicrographs are illustrated in Figure 6. Figure 6a shows the pyrite surface before oxidation. Fine mineral powders adhere to the surface, and irregular pitting is also visible on the surface. Parts b and c of Figure 6 show the pyrite surface after 7 days leaching at 333 and 363 K, respectively. Most of the fine powders disappear, and the oxidation extends over almost the whole surface of the pyrite, leaving a very rough and ragged, texture.
0 0
60
120
180 Time [ m ~ n l
-'E E 20 0
"0' v) L
0
C
z
-
g10 E
0
0
0
60
120
180 Time
[minl
Figure 5. Effect of initial concentration of Fe(II1) and Fe(I1) ions on formation of (a) Fe(I1) ion and (b and c) sulfate ion.
Overall Stoichiometry
As the oxidation of pyrite in concentrated ferric chloride solutions brings about simultaneous formation of elemental sulfur and sulfate ion, it is necessary to determine an overall stoichiometry. Let aBbe the molar fraction of total pyrite that reads via eq 1; then the summation of eq 1and 2 gives FeS2 + (14 - 12a,)Fe3+ + 8(1 - a,)H20 (15 - 12a,)Fe2++ 2asS + 2(1 - a,)SO:-
-+
16(1- a,)H+ (6)
This equation for the overall reaction results in the fol-
574 Ind. Eng. Chem. Res., Vol. 27, No. 4,1988 0.5
/
Conc. of HCI. 0.1 km01.m'~ Init. Conc. of Fe(II1). 0.5 km01.m~~ Slurry Conc.. 12 kg.m-3 Particle Size, - 53pm P
0.4
,
E I
-
" 20
