Oxidation of Sulfide Minerals Present in Duluth Complex Rock: A

---

Chapter 36

Oxidation of Sulfide Minerals Present in Duluth Complex Rock A Laboratory Study Kim A. Lapakko and David A. Antonson Division of Minerals, Minnesota Department of Natural Resources, 1525 Third Avenue East, Hibbing, MN 55746-1461

-1

-1

The average rate of sulfate release(mol(grock) s )from sixteen 75g Duluth Complex rock samples (0.053 < d < 0.149 mm) was described by[d(SO 2-)/dt] = (5.97 x 10 ) S (n = 32, r = 0.801), where S is the solid-phase sulfur content in percent. The sulfate-release rate also increased as drainage pH decreased below 4. Drainage pH decreased with increased sulfur content and experimental duration. After 150 weeks the minimum drainage pH from samples containing 0.18-0.40% S was 6.1, while that from samples containing 0.41-0.71% S ranged from 4.8 to 5.3. Minimum drainage pH values from samples containing 1.12-1.64% S ranged from 4.3 to 4.9 (69 weeks), while those from samples containing 2.06 and 3.12% S were 4.3 and 3.5, respectively (78 weeks). -13

4

0.984

2

ave

The dissolution of rocks and their component minerals has become a topic of interest for mine-waste management. Such dissolution determines the quality of drainage generated by pit walls, waste rock, and tailings. Of particular concern is the generation of acidic drainage by abandoned mine wastes. By predicting mine-waste drainage quality prior to the inception of mining, plans for mineral-resource development and mine-waste management can be developed to minimize adverse environmental impacts. The Duluth Complex in northeastern Minnesota is a large copper and nickel resource (1), and contains elevated levels of platinum group elements (2). The Minnesota Department of Natural Resources (MDNR), Division of Minerals was aware of the possibility of mineral resource development in this formation. The Division was also cognizant that predicting the quality of drainage generated over period of decades and centuries by abandoned mine wastes is a relatively new and complex science. Consequently a program was developed to examine the quality of drainage generated by potential Duluth Complex mining wastes.

0097-6156/94/0550-0593$06.00/0 © 1994 American Chemical Society

594

ENVIRONMENTAL GEOCHEMISTRY OF SULFIDE OXIDATION

The major water-quality concern regarding mine waste is the generation of acidic drainage. The extent of acid release is dependent on the balance of acidproducing and acid-consuming mineral-dissolution reactions. Acid is produced as a result of the oxidation of iron-sulfide minerals present in mine waste. Some or all of the acid produced may be neutralized by dissolution of minerals present in the host rock. Pyrrhotite (Fe^S) is the predominant iron sulfide in the Duluth Complex, occurring in both monoclinic and hexagonal forms (3), and the presence of troilite (FeS) has also been reported (4). Other sulfides present, in decreasing order of abundance, are chalcopyrite, cubanite, and pentlandite, with minor amounts of bornite, sphalerite, and pyrite (5). The complete oxidation of one mole of iron sulfide (either FeS or Fe^S) releases two moles of acid and one mole of sulfate, as indicated by reaction 1. Fe^Sis) + [(5-3x)/2]H 0 + [(9-3x)/4]0 (g) 2

2

-> +

2

(l-x)Fe(OH) (s) + 2H (aq) + S0 (aq) 3

4

(1)

The acid production is the net result of the oxidation of ferrous iron and the subsequent precipitation of ferric iron and the oxidation of elemental sulfur present. The ferric hydroxide may react further to form lepidocrocite, for example (6). The most effective minerals for neutralizing acid are calcium carbonate and magnesium carbonate. The effectiveness of such neutralization was observed with calcite occurring naturally in the Duluth Complex (7) and with limestone added as a mitigative measure (8). However, the carbonate-mineral content of the Duluth Complex is typically very low. The dominant host-rock minerals in the Duluth Complex are plagioclase, olivine, and pyroxenes. Dissolution of these minerals neutralizes acid as indicated by reactions 2 and 3 (9, 10). +

2+

CaAl Si 0 (s) + 2H (aq) + H 0 -> Ca (aq) + Al Si 0 (OH) (s)

(2)

2+

(3)

2

2

8

2

2

Mg Si0 (s) + 4iT(aq) -> 2Mg (aq) + H Si0 (aq) 2

4

4

4

2

5

4

The acid neutralization by these minerals will be rapid during the initial phase of mineral dissolution, when hydrogen ions are rapidly and reversibly exchanged with alkali ions on the mineral surface (11, 12). The rate of dissolution and consequent acid neutralization decreases, eventually becoming linear with respect to time (12, 13). The objective of this program is to correlate laboratory drainage quality with rock composition and to compare this correlation with field results. An earlier dissolution experiment was conducted on 10 drill-core samples and a test-shaft sample from the Duluth Complex. The quality of drainage generated over a 17-week period was related to the rock chemistry, mineralogy, and surface area, and compared with field data (7).

36. LAPAKKO & ANTONSON

Oxidation of Sulfide Minerals

595

The experiment presently in progress is examining the dissolution of 16 Duluth Complex samples collected from blast holes at the Dunka mine. The quality of drainage generated for periods of 69 to 150 weeks, considerably longer than the earlier experiment, is described in this paper. The data presented focus on the variation of pH and sulfide-oxidation rate as a function of solid-phase sulfur content. Methods Materials. The samples were collected from blast holes at the Dunka mine near Babbitt in northeastern Minnesota. Particle size was reduced either by hand, using a bucking maul, or mechanically with a pulverizor. The samples were sieved and particles with diameters from 0.053 to 0.149 mm (-100/+270 mesh) were retained for experimental use. Samples with sulfur contents of 1.12, 1.16, 1.40, 1.44, and 1.64 percent were wet-sieved while the remaining samples were dry-sieved. The maximum diameter was selected to ensure liberation of the sulfide minerals present

(2). The sulfur contents of the 16 samples, as determined with a LECO furnace, ranged from 0.18 to 3.12 percent. Metal content of the solids was determined using acid digestion (14) and subsequent analysis with a Perkin Elmer 603 atomic absorption spectrophotometer. The trace-metal content of the samples was fairly uniform, with trace-metal content decreasing in the order Cu > Ni > Zn > Co (Table I). Procedures. Samples (75 g), run in duplicate, were placed into the upper segment, or reactor, of a two-stage filter unit. The solids were placed on a glass-fiber filter which rested on a perforated plastic plate near the bottom of the reactor. To each reactor, 200 mL of distilled-deionized water was added, allowed to remain in contact with the solids for four to seven minutes, and then filtered through a 0.45-um filter on top of the lower stage of the filter unit. This rinsing was repeated three times at the inception of the experiment to remove oxidation products which accumulated between the time of sample crushing and the beginning of the experiment. The solids were subsequently rinsed weekly with a single 200-mL volume. Reaction Conditions. Between rinses the solids were retained in the reactors to oxidize. The reactors were stored in individual cubicles which formed a rectangular matrix within a topless housing with a perforated base. A thermostatically controlled heating pad was placed beneath the housing to control temperature. The housing was stored in a small room equipped with an automatic humidifier and dehumidifier, to maintain a stable range of humidity. Temperature and relative humidity were monitored two to three times a week, and the average weekly values were determined. Variations in temperature and humidity did occur, due largely to seasonal variations in these parameters (Figure 1). The experiment began on 14 February 1989 with reactors 1-24. Reactors 1-20 are yet in progress and data from the first 150 weeks are presented. Reactors 21-24 were terminated after 78 weeks, and reactors 29-38 were started at week 81 (4

596

ENVIRONMENTAL GEOCHEMISTRY OF SULFIDE OXIDATION

li ; ι ι ; ° ι »0 LV \ ο li Q\

\\

%

\

ι

1

ι

ι

ι

Week 21 ο Week 78 Week 150 ° Δ

ο

0

0

\A \io\ \\

\

Δ

0

\k

\

-

y

\

\ \

V\

a\

\

\ ο

-

\

\

V) \ • V \ D \\ 0\ A

\

Δ Δ

\ 0

Δ

\

\

4.5

\ °

0.0

i 0.5

I 1.0

ι 1.5

^^

\

Ο

As

^ s

Δ

1 1 \ 1 2.0 2.5 3.0

3.5

Percent Sulfur in Solid Phase Figure 1. Temperature and relative humidity versus time.

36.

LAPAKKO & ANTONSON

Oxidation of Sulfide Minerals

597

September 1990). The last set of reactors is yet in progress and pH data for the first 78 weeks and sulfate data for 69 weeks are presented. Analysis. The volume offilteredeffluent rinse water, or drainage, was determined by weighing the drainage. Samples were then filtered and analyzed on site to determine pH, alkalinity, acidity, and specific conductance. Samples were also taken for subsequent determination of metal and sulfate concentrations. Samples taken for metal analyses were acidified with 0.2 mL AR Select nitric acid (Mallinckrodt) per 50 mL sample. An Orion SA 72 pH meter, with a Ross combination pH electrode (8165), was used for pH analysis and a Myron L conductivity meter was used to determine specific conductance. Alkalinity and acidity were analyzed using standard titration techniques (15). Sulfate was analyzed using an HF Scientific DRT-100 nephelometer for the barium sulfate turbidimetric method (15). Metals were analyzed with a Perkin Elmer 603 atomic absorption spectrophotometer. Calculation of Sulfate-Release Rates. The mass of sulfate released was calculated as the product of the observed sulfate concentration and the volume of drainage. Although all drainage samples were not analyzed for sulfate, specific conductance was analyzed weekly. Missing sulfate values were estimated based on regression analyses of the measured sulfate concentrations vs specific conductance observed for each of the sixteen solids (40 < η < 123). The r values for 13 of the solids ranged from 0.56 to 0.95, with a median value of 0.84. The r values for the solids containing 0.18,0.22, and 0.51 percent sulfur were 0.45,0.22, and 0.41, respectively. For most of the reactors the sulfate release in the initial weeks was inconsistent with that observed over the experiment as a whole. This initial period was ignored in the calculation of all sulfate-release rates. The average release rate was calculated as the total sulfate release during this modified period of record divided by the number of weeks in the modified period. To determine the variation in release rates during the experiment, cumulative sulfate release was plotted as a function of time for each reactor. Periods of linear release were selected based on visual examination of the graphs produced, and the release rate for each period was determined by linear regression. 2

2

Results and Discussion The agreement between duplicate samples for all three rates was generally quite good, with the difference from the mean value for the two reactors typically less than ten percent. Using the chemical data from Table I, it was estimated that 0.9 grams of pyrrhotite were present in the sample containing 0.71 percent sulfur. Assuming all sulfate release from this sample was due to pyrrhotite oxidation yields an oxidation rate of about 2.4 χ 10" mol s" g" pyrrhotite. Assuming the relationship between the pyrrhotite particle size and specific surface area is the same as that between quartz sand particle size and specific surface area (16) allows conversion of this pyrrhotite oxidation rate to 8.0 χ 10* mol m" s". Assuming equal oxidation rates for all metal-sulfide minerals present, and that release of one mole of sulfate indicates the oxidation of one mole of metal-sulfide mineral, yields a metal-sulfide oxidation rate of 4.9 χ 10" mol m" s". This rate is reasonably consistent with the 11

1

1

10

10

2

1

2

1

598

ENVIRONMENTAL GEOCHEMISTRY OF SULFIDE OXIDATION

Table I. Chemical Composition of Dunka Blast Samples Subjected to Dissolution Experiments

s

a

b

wt%

Cu wt%

Ni wt%

Co wt%

Zn wt%

S w/ FeS" wt%

0.18 0.22 0.40 0.41 0.51 0.57 0.58 0.71 1.12 1.16" 1.16" 1.40" 1.40" 1.44 1.63 1.64 2.06 3.12

0.190 0.180 0.072 0.105 0.179 0.179 0.208 0.190 0.124 0.163 0.183 0.290 0.318 0.174 0.181 0.333 0.187 0.201

0.055 0.062 0.032 0.047 0.038 0.041 0.055 0.059 0.042 0.063 0.066 0.087 0.097 0.052 0.059 0.084 0.058 0.050

0.011 0.010 0.005 0.009 0.009 0.011 0.018 0.018 0.009 0.012 0.011 0.013 0.015 0.011 0.015 0.013 0.009 0.007

0.018 0.019 0.032 0.021 0.027 0.027 0.025 0.021 0.041 0.022 0.022 0.024 0.024 0.026 0.026 0.031 0.037 0.035

0.039 0.081 0.328 0.316 0.381 0.438 0.423 0.562 1.010 1.026 1.015 1.187 1.167 1.305 1.486 1.404 1.911 2.970

Calculated as the sulfur not bound by Cu, Ni, Co, or Zn assuming all trace metals were bound by sulfur in a 1:1 molar ratio of sulfur to metal. Duplicate analyses for metals.

36. LAPAKKO & ANTONSON

599

Oxidation of Sulfide Minerals 10

2

1

overall metal-sulfide rate for Duluth Complex rock of 2.6 χ 10" mol m* s' previously reported for batch reactor experiments (17). Few rates are reported for the oxidation of pyrrhotite. The pyrrhotite-oxidation rate determined in the present study is roughly an order of magnitude lower than the 1 χ 10" mol m s oxidation rate at pH 6 reported by Nicholson and Scharer (18) for pyrrhotite particles of similar size, and similar to that previously reported for pyrite (18). The similarity with reported rates for pyrite oxidation with those for the mixed sulfides in the present study suggests that the rate-controlling step may be similar. Examination of this hypothesis was beyond the scope of the present study. Previous work on the oxidation of sulfide minerals present in the Duluth Complex was consistent with a surface-reaction mechanism in which the rate limiting step was the rearrangement of molecules and/or electrons at sites at which oxygen was adsorbed to sulfide minerals (17). The studies on pyrrhotite and pyrite previously mentioned (18) were conducted with relatively pure minerals, and the rates in the present study are possibly influenced by the complex mineral assemblage present. For example, Koch (19) reported that the presence of copper in a sulfide such as chalcopyrite stabilizes the ferric iron present. It is also possible that copper released from chalcopyrite may have participated in an exchange reaction with iron present in pyrrhotite and, consequently, inhibited its oxidation. Similar reactions have been reported in which metals in solution exchange with sulfide-bound metals of higher solubility (20). In contrast, if galvanic interactions of chalcopyrite and pyrrhotite contributed significantly to pyrrhotite oxidation (21), lower than normal oxidation rates might be expected for these samples. 8

2

1

Dependence of Sulfate-Release Rate on Sulfur Content A first order of dependence on the sulfide-mineral surface area has been reported for the oxidation of pyrrhotite (18) and mackinawite (6) by oxygen. A similar dependence has been reported for the oxidation of the mixed sulfides present in the Duluth Complex (7, 17). The sulfur content of the solids in the present study was assumed to be proportional to the sulfide-mineral surface area available for reaction because (1) the sulfide minerals were liberated in the size fraction selected for the experiment and (2) the specific surface area of the sulfide minerals was assumed to be constant among sulfide minerals of the same particle size. The weekly release rates calculated were approximatelyfirstorder with respect to solid-phase sulfur content, as indicated by equations 4, 5, and 6. 2

[d(S0 -)/dt] 4

4

4

13

0984

ave

= (5.97 χ Iff ) S

2

[d(S0 )/dt]

0948

= (4.13 χ Iff ) S

2

[d(S0 )/dt]

13

min

13

= (8.74 χ Iff ) S

max

2

2

(4)

2

(5)

, η = 32, r = 0.833

, η = 32, r = 0.801

0954

2

, η = 32, r = 0.669

(6) 1

1

where d(S0 ")/dt = rate of sulfate release in mol (g rock)" s" and S = solid-phase sulfur content in percent 4

600

ENVIRONMENTAL GEOCHEMISTRY OF SULFIDE OXIDATION

These equations were plotted on a single graph along with the observed average sulfate-release rates (Figure 2). Since the dependence on sulfur content was close to first order, linear regression analysis was conducted on the data and yielded the following equations. 2

[d(S0 -)/dt] 4

min

2

13

14

2

(7)

12

13

2

(8)

= 4.96 χ 10 S - 5.39 χ ΙΟ" , η = 32, r = 0.833

[d(S0 -)/dt] = 1.17 χ 10 S - 4.73 χ ΙΟ , η = 32, r = 0.722 4

ave

2

[d(S0 )/dt] 4

12

max

12

2

= 2.41 χ 10 S - 1.26 χ ΙΟ , η = 32, r = 0.661

(9)

The ratio of the maximum rate to the minimum rate was between 1.3 and 2.8 for 75 percent of the reactors. The sulfate release from the sample containing 1.12 percent sulfur was fairly constant, and the ratio was below the typical range (Table II). The variation in sulfate release from sample containing 3.12 percent sulfur was highest, with maximum to minimum rate ratios of 5.8 and 6.8 for the duplicate samples. The maximum to minimum rate ratios were generally higher for the samples with the 150 week period of record than for samples with shorter records, as indicated by median ratios of 2.35 and 1.65, respectively. Examination of Figure 2 indicates that the sulfate-release rates for solids containing 0.40, 1.12, and 3.12 percent sulfur were higher than suggested by the relationship between rate and sulfur content for the samples as a whole. The rates for the 3.12 percent sample were most deviant from the remaining data, and the deviation can be attributed to increased sulfide oxidation at low pH. The accelerated rates observed for the 0.40 and 1.12 percent S samples may be due to greater available sulfide-surface areas and/or chemical or mineralogical differences for these samples. First, the sulfide-surface area of these samples may be higher due to subtle differences in sulfide grain size and/or "roughness" of the mineral surfaces. Second, there may be sulfide minerals present in these two samples which oxidize more rapidly than the sulfides typically present in the samples examined. Third, these two samples have the lowest copper content of the samples examined. This difference, and/or its implication of a difference in chalcopyrite content, may influence the sulfide-oxidation rate for these samples. As previously discussed, it is possible that under certain conditions copper has an inhibitive effect on iron-sulfide oxidation. Dependence of Sulfate-Release Rate on pH. The dependence of the rate of sulfidemineral oxidation by oxygen on pH is reported to be slight. Nicholson and Scharer (18) found little variation in the abiotic pyrrhotite oxidation rate over the pH range of 2 to 6, reporting "maximum differences in oxidation rates were about a factor of 2 or within 50% of the mean rate at a specified temperature." Nelson (6) reported that the rate of mackinawite oxidation by oxygen at pH 6.5 was five times that at pH 9.0. For the mixture of sulfides present in Duluth Complex rock the sulfide-oxidation rate was reported as proportional to [H ] over the pH range of 5 to 8 (17). Research on pyrite oxidation, however, indicates that as "pH decreases to 4.5, ferric iron becomes more soluble and begins to act as an oxidizing agent" (22). As pH further decreases, bacterial oxidation of ferrous iron becomes the rate limiting +

02

36.

LAPAKKO & ANTONSON

601

Oxidation of Sulfide Minerals

0

1

2

3

4

Percent Sulfur in Solid Phase

Figure 2. Average sulfate-release rate as a function of solid phase sulfur content. The lines are regressions for minimum, average, and maximum rates (equations 4-6): [d(S0 -)/dt] = (4.13 χ 10 ) S ; [d(S0 ")/dt] = (5.97 χ 10 ) S ; [d(S0 -)/dt] = (8.74 χ 10 ) S * . 2

4

13

4

0948

min

2

max

2

4

13

0

954

13

ave

0984

602

ENVIRONMENTAL GEOCHEMISTRY OF SULFIDE OXIDATION

Table II. Summary of Sulfate Release Rates and Minimum Drainage pH Sulfur %

Min pH s.u.

1

Rate, mol S0 s" g rock' χ 10 4

1

Min

a

Avg

13

Rate, mol S0 s mol S" χ 10

1

4

1

9

Max

Min

Avg

Max

Time weeks

0.18 0.18 0.22 0.22 0.40 0.40

6.20 6.25 6.40 6.33 6.16 6.10

1.23 1.17 1.10 1.12 2.38 1.98

1.67 1.72 1.43 1.45 4.16 3.72

2.29 3.26 2.29 2.77 7.39 6.84

2.20 2.08 1.61 1.64 1.91 1.59

2.99 3.06 2.09 2.12 3.34 2.99

4.08 5.81 3.34 4.05 5.94 5.50

150 150 150 150 150 150

0.41 0.41 0.51 0.51 0.57 0.57 0.58 0.58 0.71 0.71

5.30 5.20 5.08 4.98 5.25 5.35 4.85 4.82 4.98 5.05

1.36 1.36 1.61 1.69 2.33 2.40 2.05 1.83 2.20 1.63

2.38 2.07 2.60 2.66 3.08 3.15 2.60 2.46 2.82 2.82

4.25 3.10 4.11 4.09 3.54 3.67 3.65 4.03 3.94 4.88

1.07 1.07 1.01 1.07 1.31 1.35 1.13 1.01 1.00 0.74

1.86 1.62 1.64 1.68 1.74 1.77 1.44 1.37 1.27 1.27

3.33 2.43 2.59 2.58 2.00 2.07 2.02 2.23 1.78 2.21

150 150 150 150 150 150 150 150 150 150

1.12 1.12 1.16 1.16 1.40 1.40 1.44 1.44 1.63 1.63 1.63 1.63 1.64 1.64

4.62 4.75 4.69 4.60 4.78 4.92 4.59 4.41 4.95 4.75 4.37 3.90 4.42 4.32

9.46 9.66 3.96 4.05 4.53 4.84 3.85 3.87 4.58 6.80 3.98 6.71 7.59 9.50

9.77 9.90 5.13 5.17 6.09 6.53 4.86 4.97 6.05 6.80 5.59 10.10 9.26 11.48

10.10 10.05 6.73 6.62 7.88 7.50 5.61 5.63 6.62 6.80 6.62 18.22 9.86 19.12

2.71 2.77 1.10 1.12 1.04 1.11 0.86 0.86 0.90 1.34 0.79 1.32 1.49 1.86

2.80 2.84 1.42 1.43 1.40 1.50 1.09 1.11 1.19 1.34 1.10 1.99 1.81 2.25

2.90 2.88 1.87 1.83 1.81 1.72 1.25 1.26 1.31 1.34 1.31 3.59 1.93 3.75

69 69 69 69 69 69 69 69 69 69 150 150 69 69

2.06 2.06 3.12 3.12

4.20 4.30 3.70 3.50

10.54 10.93 15.88 15.91

14.23 28.23 12.96 21.63 38.24 91.48 52.65 104.32

1.64 1.71 1.64 1.64

2.22 2.02 3.94 5.42

4.40 3.37 9.42 10.75

78 78 78 78

a

a

Sixty-nine week data included for 1.63% S sample for comparison with samples of similar sulfur content.

36. LAPAKKO & ANTONSON

Oxidation of Sulfide Minerals

603

step in the oxidation of pyrite by ferric iron (23), which is the only significant oxidant in this pH range (22, 23, 24). Consequently, under these low-pH conditions the reaction is independent of sulfide-mineral surface area (24). The large variation observed for the rate of sulfate release from the sample containing 3.12 percent sulfur suggests that the rate was affected by the low-pH conditions created by rapid iron-sulfide oxidation. The pH of drainage from this sample decreased steadily from around 6.0 to about 3.5. As pH decreased below 4.0, the sulfate-release rate increased rapidly. The rates observed in the pH range of 3.5 to 4.05 were roughly six to seven times those observed in the pH range of 5.35 to 6.1. This acceleration was probably due to increased biologically mediated ferric-iron oxidation of the sulfide minerals as pH decreased below 4.0. Abiotic oxidation may have also increased, although the extent of increase was most likely slight relative to the increase in biological oxidation (22). Because the decreased pH conditions which accelerated sulfide oxidation were produced as a result of elevated iron-sulfide oxidation, the variation in sulfate release with sulfur content is influenced by a form of autocatalysis. Dependence of Sulfate-Release Rate on Temperature and Relative Humidity. Analysis of the data indicates, qualitatively, that variations in sulfide-oxidation rates were related to changes in temperature and relative humidity in the reaction environment, as well as changes in pH within the reactor. Temperature and/or relative humidity may have influenced the rate of sulfide-mineral oxidation in a manner similar to that reported for pyrite oxidation (26). Inspection of the rates reported by Nicholson and Scharer (18) indicates that the pyrrhotite oxidation rate at pH 6 roughly quadrupled with a temperature increase of 10°C. Nelson (6) reported that the rate of mackinawite oxidation was approximately doubled by a temperature increase of 8 - 10°C. Recent studies indicate that the effect of relative humidity on pyrite oxidation is dependent on the type of pyrite oxidized (27). The observed temperature and relative humidity during periods of maximum and minimum sulfate release qualitatively indicate that sulfide-oxidation rates were affected by temperature and, perhaps, relative humidity. The qualitative assessment of these effects considered 23 reactors for which the ratio of maximum to minimum sulfate-release rate exceeded 1.5. It excluded the sample containing 3.12 percent sulfur. For 20 of the 23 cases, maximum rates occurred during times of elevated temperature, specifically weeks 6 to 28, 65 to 81, and 117 to 134. The relative humidity was above average in the middle period and near average in the others. Likewise, for 20 of the 23 cases, minimum rates occurred during periods of low temperature: weeks 36 to 58, 86 to 109, and 136 to 150. The relative humidity during the first two periods was near average and below average during the third (Figure 1). Dependence of Drainage pH on Sulfur Content. Assuming iron-sulfide oxidation was the major contributor of sulfate, the rate of acid production also increased linearly with solid-phase sulfur content. The observed drainage pH data support this hypothesis. The pH of drainage from the 16 samples was observed generally to decrease as solid-phase sulfur content and time of dissolution increased (Figure 3,

604

ENVIRONMENTAL GEOCHEMISTRY OF SULFIDE OXIDATION

36. LAPAKKO & ANTONSON

605

Oxidation of Sulfide Minerals

Table Π). The decrease in drainage pH over time can be attributed to the depletion of the more reactive neutralizing components. The samples were divided into four classes based on sulfur content and the minimum pH of drainage from the samples (Table II). This classification was complicated by the variable periods of record for the samples and was based partly on the "best guess" extrapolation of drainage pH over time. As the experiment progresses, and as other samples are examined, the classification is likely to be revised. The terms describing the drainage pH ranges (circumneutral to strongly acidic) were used to simplify data presentation. They were selected as descriptions relative to the experimental data generated, and not with respect to mine waste drainage in general. The gross rates of sulfate release (not normalized for sulfur content) associated with the various groups are also presented. The rates were not normalized in order to provide a measure of the relative magnitudes of acid production for the four classes. Samples containing 0.18 to 0.40 percent sulfur generated minimum drainage pH values of 6.10 to 6.40 over the 150 week period of record. The pH of drainage from these solids meets the typical water-quality standard of pH 6.0, thus these samples present the lowest potential for generating acidic drainage. The rates of sulfide oxidation were the lowest observed, averaging approximately 1.5 χ 10" mol (g rock)" s", and were consistent with the general trends observed for the solids. Apparently the dissolution of silicate minerals present in these samples (for example reactions 2, 3) neutralized the acid resulting from iron-sulfide oxidation. Samples containing 0.41 to 0.71 percent sulfur produced "mildly acidic" drainage, with minimum drainage pH values of 4.82 to 5.25 over a period of 150 weeks (Figure 3). The average sulfide-oxidation rate for these samples was in the range of 2.2 χ 10" to 3.3 χ ΙΟ" mol (g rock)" s*, with maximum rates roughly 50 percent higher. Although mitigative measures might be required for such drainages to meet pH standards, the extent of mitigation required would be relatively low. Samples containing 1.12 to 1.64 percent sulfur produced "moderately acidic" drainage, with minimum drainage pH values of 4.32 to 4.95 over a period of 69 weeks. Although the minimum drainage pH range for these samples intersects that for the Group 2 samples, the period of record for the Group 3 samples is considerably shorter. Based on the observed temporal variation, the pH of drainage from these samples would be expected to decrease as dissolution continues (Figure 3). Indeed, the period of record for reactors containing the 1.63 percent sulfur sample is 150 weeks, and the minimum drainage pH values for this period are 4.37 and 3.90. It is likely that other samples in this group will exhibit similar decreases in drainage pH over time. The average sulfide-oxidation rates from these samples ranged from about 4.4 χ 10" to 11 χ 10" mol (g rock)" s", indicating a higher rate of acid production than samples from the first two groups. The oxidation rate for the 1.12 percent sulfur sample was anomalously high. Unlike most samples, the rate of sulfate release from this sample was also relatively constant This suggests that the sulfide-mineral composition of this sample may be slightly different from that of the samples in general. Mineralogical analyses are planned to elucidate these phenomena. The samples containing 2.06 and 3.12 percent sulfur produced "strongly acidic" drainage, with minimum pH values of 4.20 and 3.50, respectively, over a 13

1

1

13

13

13

13

1

1

1

1

606

ENVIRONMENTAL GEOCHEMISTRY OF SULFIDE OXIDATION

period of 78 weeks (Figure 3). The average sulfide-oxidation rates for these samples were roughly 13 χ 10" and 46 χ 10" mol (g rock)" s*. With the observed pH variation, accelerated sulfide oxidation would be expected at the lower pH conditions observed for the higher sulfur content. Based on samples examined to date, Duluth Complex mine wastes with similar sulfur contents would require the most rigorous mitigation to meet water-quality standards for pH. 13

13

1

1

Summary The release of sulfate from Duluth Complex rock samples was observed to be first order with respect to solid-phase sulfur content. The majority of the sulfate released was assumed to be due to the oxidation of iron sulfide, mainly pyrrhotite, which is the dominant sulfide in the Complex. An observed decrease in drainage pH with sulfur content was consistent with this hypothesis. Based on the observed drainage quality, the samples were divided into four ranges of sulfur content. Distinct ranges for minimum drainage pH and sulfideoxidation rate were observed for each group. Continued testing of the present samples, dissolution tests on additional well-characterized Duluth Complex samples, and field verification will be necessary to evaluate the validity of the results presented. Trace metal release will also be examined. Additional data analysis is planned to quantify the effects of pH, temperature, and relative humidity on the observed sulfide-oxidation rate. Drainage samples will be analyzed for calcium and magnesium to obtain insight into dissolution reactions which contribute to acid neutralization in various pH ranges. Furthermore, the rock samples examined will be subjected to more extensive physical, chemical, and mineralogical analysis. The rates of sulfide oxidation and acid neutralization will be compared with other reported values, and factors which potentially control dissolution will be addressed. Acknowledgments Bill Conger of LTV Steel Mining Company was responsible for particle size reduction and sample sieving. Kate Willis conducted the majority of the laboratory experiments and Jean Matthew analyzed samples for sulfate. Anne Jagunich and Jason Perala were responsible for data input and Jim Porter managed data output. The U. S. Bureau of Mines provided funding to support data compilation and analysis as part of a cooperative project on mine waste dissolution.

Literature Cited 1. 2. 3. 4.

Kingston, G. Α.; Carillo, F. V.; Gray, J.J.; Mallory, P. U.S. Bur. Mines IC 8469, 1970, 57 pp. Anonymous. Skillings' Mining Review, 1992, 81, 5. Foose, M.; Weiblen, P. In Geology and Metallogenv of Copper Deposits, Freidrich, G. H., Ed., Springer-Verlag: Berlin, 1986, pp. 8-24. Boucher, M . L. U.S. Bur. Mines RI 8084, 1975, 55 pp.

36. LAPAKKO & ANTONSON Oxidation of Sulfide Minerals

5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

607

Stevenson, R. J.; Kreisman, P. J.; Sather, N . P. In Minnesota Regional Copper-Nickel Study. Chap. 1. MN Environ. Quality Board: St. Paul, MN, 1979; Vol. 3, 70 pp. Nelson, M . Ph.D. Thesis, Stanford University, Palo Alto, CA, 1978. Lapakko, K. A. In Proc. 1988 Mine Drainage and Surface Mine Reclamation Conference. U.S. Bur. Mines IC9183, 1, 1988, pp. 180-190. Lapakko, Κ. Α.; Antonson, D. A. In Proc. Second International Conference on the Abatement of Acidic Drainage, MEND (Mine Environment Neutral Drainage), Montreal, QC, 1991, pp. 343-358. Busenberg, E. Geochim. Cosmochim. Acta 1978, 42, 1679-1686. Hem, J. D. U.S. Geol. Survey Water-Supply Paper 1473, 1970, 363 pp. Holdren, G. R.; Berner, R. A. Geochim. Cosmochim. Acta 1979, 43, 11611171. Busenberg, E.; Clemency, C. Geochim. Cosmochim. Acta 1976, 40, 41-49. White, A. F.; Classen, H. C. In Chemical Modeling in Aqueous Systems, Jenne, Ε. Α., Ed., ACS Symposium Series No. 93, American Chemical Society: Washington, DC, 1979, pp. 447-473. U.S. Bur. Mines. U.S. Bur. Mines RI 8480, 1980; 14 pp. American Public Health Association; American Water Works Association; Water Environment Federation. Standard Methods for the Examination of Water and Wastewater, 18th edition, American Public Health Association: Washington DC, 1992. Parks, G. A. In Mineral-Water Interface Geochemistry, Rev. Mineral. 1990, 23, 133-175. Lapakko, K. A. M.S. Thesis. University of Minnesota, Minneapolis, MN, 1980. Nicholson, R. V.; Scharer, J. M. In this volume. Koch, D. F. In Modem Aspects of Electrochemistry, Bockris, J. O.; Conway, Β. E., Eds.; Plenum Press: New York, New York, 1975, Vol. 10, pp. 211-237. Gaudin, A. M.; Fuerstenau, D. W.; Mao, G. W. Trans. AIME, 1959, 214, 430-432. Ahonen, L.; Hiltunen, P.; Tuovinen, Ο. H. In Fundamental and Applied Biohydrometallurgy; Lawrence, R. W.; Branion, R. M . R.; Ebner, H. G. Eds., Elsevier: Amsterdam, 1986; pp. 13-22. Nordstrom, D. K. Acid Sulfate Weathering, Soil Sci. Soc. Am.: Madison, WI. 1982, pp. 37-56. Singer, P. C.; Stumm, W. Science 1970, 167, 1121-1123. Kleinmann, R. L. P.; Crerar, D. Α.; Pacelli, R. R. Mining Eng. 1981, 300305. Stumm, W.; Morgan, J. J. Aquatic Chemistry, Wiley: New York, 1981, 780 pp. Morth, A. H.; Smith, Ε. E. In preprints from ACS Division of Fuel Chemistry Symposium on Fossil Fuels and Environmental Pollution; American Chemical Society: Washington, DC, 1966, Vol. 10 (1), pp. 83-92. Borek, S. L. In this volume.

RECEIVED October 20,

1993