*Shannon Shaw, M.Sc. is a Geochemist with Robertson Geoconsultants Inc. in Vancouver

Introduction

Acid Rock Drainage (ARD), as described in the Introduction Page, is the product formed by the atmospheric oxidation (i.e. by water, oxygen and carbon dioxide) of the relatively common iron sulfide minerals pyrite (FeS2) and pyrrhotite (FeS) in the presence of (catalyzed by) bacteria (Thiobacillus ferrooxidans), and any other products generated as a consequence of these oxidation reactions (i.e. heavy metals solubilized by acidic solutions). To understand this more fully the basic chemistry must be examined.

Basic Chemistry of ARD Generation

The following chemical reactions describe the oxidation of pyrite (FeS2) to the products that constitute the contaminants generically termed Acid Rock Drainage (ARD), although similar equations may be written for the oxidation of pyrrhotite (FeS).

FeS2 + 7Fe2(SO4)3 + 8H2O = 15FeSO4 + 8H2SO4 (1)
FeS2 + Fe2(SO4)3 = 3FeSO4 + 2S (2)
4FeSO4 + O2 + 2H2SO4 bacteria = 2Fe2(SO4)3 + 2H2O (3)
2S + 3O2 + 2H2O bacteria = 2H2SO4 (4)
4FeS2 + 15O2 + 2H2O = 2Fe2(SO4)3 + 2H2SO4 (5)
S0 + 3Fe2(SO4)3 + 4H2O = 6FeSO4 + 4H2SO4 (6)

Neutralization of the acidic metal-rich solutions that can be generated by the above chemical reactions may also occur as a result of dissolution of neutralizing minerals (most importantly carbonates) that come in contact with the acidic solutions. These chemical reactions must also be examined to fully understand the processes occurring and to be able to predict the chemistry of solutions resulting from the combination of oxidation and neutralization processes. Examples of these reactions include:

Calcite dissolution by sulfuric acid
CaCO3 + H2SO4 -> CaSO4 + H2O + CO2 (7)
Dolomite dissolution by sulfuric acid
CaMg(CO3)2 + 2H2SO4 -> CaSO4 + MgSO4 + 2H2O + CO2 (8)

Reactions of other gangue minerals may also contribute to the neutralization potential of the ‘system’ under specific pH conditions (Ritchie, 1994), for example:

Muscovite dissolution
KAl2[AlSi3O10](OH)2(s) + H+ + 3/2H2O -> K+ + 3/2Al2Si2O5(OH)4(S) (9)
Biotite dissolution
KMg1.5Fe1.5AlSi3O10(OH)2(s) + 7H+ + 1/2H2O -> K+ + 1.5 Mg2+ + 1.5 Fe2+ + H4SiO04 + 1/2Al2Si2O5(OH)4(s) (10)
Albite dissolution
NaAlSi3O8(s) + H+ + 9/2H2O -> Na+ + 2H4SiO04 + 1/2Al2Si2O5(OH)4(s) (11)
Anorthite dissolution
CaAl2Si2O8(s) + 2H+ + H2O -> Ca2+ + Al2Si2O5(OH)4(s) (12)
K-feldspar dissolution
KAlSi3O8(s) + H+ + 9/2H2O -> K+ + 2H4SiO04 + 1/2Al2Si2O5(OH)4(s) (13)
Iron oxy-hydroxide dissolution
Fe(OH)3(s) + 3H+ -> Fe3+ + 3H2O (14)

When studying these reactions and the resulting ‘systems’ it should be clear that mineralogy is the key factor. However, in many instances where ARD ‘systems’ are studied in order to assess or predict water quality, it is the mineralogy that is largely ignored. The primary excuse for excluding or doing only minimal mineralogical characterization is that detailed mineralogy is often time consuming, expensive and specialized. In recent years, more and more people in the industry have acknowledged that mineralogical characterization is necessary, however there remains a lack of detailed work and/or interpretation of it.

Methods and Techniques

Petrographic or mineralogic examination of samples in ARD predictive work is usually conducted by transmitted and reflected light microscopy, and by various X-ray diffraction (XRD) techniques. Although electron probe microanalysis (EPMA), scanning electron microscopy (SEM) and other more specialized techniques are employed, their use is generally confined to sulphide minerals where compositional abnormalities affect ARD testwork interpretation. Such techniques are particularly useful in the determination of the chemical composition of sulphide oxidation products such as rims, inclusions and amorphous (non-crystalline) species.

Transmitted light microscopy utilizes thin (30 m) sections of samples and reflected light microscopy utilizes polished mounted samples. Current practice combines the two sample preparation procedures to produce polished thin sections, so that the two techniques can be used on the same sample. McCrone et al. (1979) describe in detail the techniques of thin and polished section microscopy. Samples may be prepared from whole rock in the form of drill core, or from fragmented material such as humidity cell feed and residue samples, or from tailings.

Transmitted light microscopy is used to examine those minerals that transmit light in thin section, and these include most of the gangue or non-metallic minerals that may have neutralizing capability. Reflected light microscopy is used to examine those minerals that do not transmit light in thin section, but reflect light to varying degrees when polished. Such minerals include metallic sulphides that may oxidize to generate acid.

Both types of microscopy are used (often with supplementary techniques such as selective staining or etching) to identify individual mineral grains, to determine mineral grain size and size distribution, and to identify mineral grain spatial interrelationships.

Reaction products of sulphide oxidation (rimming of grains) are readily observed, as are many other characteristics of mineral grains (such as inclusions) not readily seen by other investigative techniques. These capabilities of microscopic examination are extremely useful in ARD studies of both tailings and waste rock.

The ultimate size limitation of these methods is the wavelength of the light used, and the optical arrangement of the microscope, but it is typically about 1 m grain size. Consequently, samples containing significant quantities of clay minerals (< 2 m) present identification problems to the microscopist.

It is usually possible to determine the frequency of occurrence of individual minerals within a sample by the examination of a number of fields of view. Quantitative mineralogical analysis by this method is termed modal analysis.

For some well-defined, unaltered, rock types it has been possible to calculate a modal analysis (or quantitative mineralogical analysis) from elemental oxide analyses of whole rock samples using mathematical techniques. Such calculations, known as CIPW from the initials of their original developers (Cross et al., 1902), can be undertaken by computer programs such as NewPet for DOS (Clarke, D., 1993). However, for the rock types typically associated with significant sulphide mineralization, CIPW calculations have not yet been shown to have general applicability (Lawrence and Sheske, 1997, Paktunc, 1998c) because they were developed for unaltered magmas, and cannot be applied to sedimentary or metamorphic rocks or to altered igneous rocks. This is most unfortunate, since whole rock elemental oxide analysis is generally less expensive than microscopic analysis and does not require the considerable skill of an mineral microscopist. Paktunc (1998b) has developed a computer program called MODAN that may be used to determine modal mineralogy from whole rock elemental oxide analysis and a knowledge of the mineral species present in a sample. Paktunc (1998a) has also discussed the use of MODAN in conjunction of mineral reactivity data to estimate the Neutralizing Potential (NP) of samples containing neutralizing silicates in addition to carbonates.

Low-power stereoscopic microscopy is also a useful tool, particularly in the examination of tailings and other relatively fine, fragmented samples. This technique is valuable in the determination of degree of sulphide mineral liberation and may also be used as an aid in sulphide mineral identification. The identification of mineral grains is well covered by Jones and Fleming (1965). It is usually preferable to examine screen-size fractions when using low-power stereoscopic microscopy, and it is helpful also to remove the bulk of the gangue mineral particles using one or more heavy liquid procedures at a Relative Density of 2.8 to 2.9 (Mills, 1978,1985).

Some workers (Jambor and Blowes, 1998) obtain an X-ray diffractogram for each sample examined by transmitted and reflected light microscopy and compare the mineral identification data obtained from the two procedures for corroboration. This is a highly recommended approach.

Individual mineral grains greater than about 100 m may be positively identified by XRD using the X-ray film method with Debye-Scherrer or Gandolfi cameras, or with modern microdiffractometry instruments (Jambor and Blowes, 1998). These XRD methods all give qualitative data on mineral identity.

Examples

It is the intention of this site to act as an introduction to the mineralogical and petrological basics associated with ARD in an effort to emphasize the importance of mineralogical characterization in ARD studies and prediction. One of the major criticisms of petrological characterization of minerals is that it is largely non-quantitative or semi-quantitative. In an effort to quantify sulfide oxidation, Blowes and Jambor (1990) have developed a Sulfide Alteration Index tailored to the oxidation of mine wastes. Although it is objective it serves to put into relative terms the degree of oxidation occurring or having occurred at one site and give an indication of the progression and remaining oxidation of a particular waste as it is easily adapted to specific site mineralogy. The index (developed initially for the Waite Amulet Tailings, Quebec) is reproduced below.

Numerical Scale

Degree of Alteration of Sulfides

10

Pyrrhotite and pyrite obliterated; only traces of sulfide, typically chalcopyrite are present.

9

Similar to 10, but with a few scattered remnant grains of pyrite.

8-7

First appearance of trace amounts of pyrrhotite (at scale 8); at scale 7 the vestiges of strongly altered pyrrhotite increases in abundance or degree of preservation.

6-2

At scale 6 the pyrrhotite grains have broad alteration rims, but the cores of numerous grains are preserved; gradation to scale 2 is marked by the appearance of narrower alteration rims and a predominance of unaltered grains.

1-0

Only a few grains of pyrrhotite are weakly altered along rims and fractures; >95% of the grains have sharp, fresh margins.

A second criticism of petrological characterization of minerals is that it is relatively abstruse to the majority of people studying in this field. The following photomicrographs (1-9) and their corresponding descriptions and interpretations have been included as examples of sulfide oxidation and dissolution of neutralizing minerals from various sites throughout North America.

Photomicrograph 1: Reflected light photomicrograph of altered pyrrhotite grain.
Alteration is proceeding as preferential replacement along the {0001} crystallographic parting plane. The alteration (secondary) mineral phase is predominantly an iron-oxyhydroxide.

Photomicrograph 2: Reflected light photomicrograph of altered pyrrhotite and unaltered pyrite grains.
This photomicrograph demonstrates the different alteration rates between pyrrhotite and pyrite. Alteration of pyrrhotite in this case is occurring both around the grain periphery and along the parting plane.

Photomicrograph 3: Reflected light photomicrograph of altered pyrrhotite and pentlandite grains.
Both the pyrrhotite and pentlandite exhibit alteration to iron-oxyhydroxides at approximately equal rates of oxidation, however, the alteration textures are significantly different. Pyrrhotite shows predominant alteration along the grain's parting plane, whereas pentlandite alteration is seemingly non-preferential, primarily along fractures in the grain.

Photomicrograph 4: Photomicrograph of altering pyrrhotite grains and associated primary minerals 'cemented' by secondary gypsum and iron-oxyhydroxides (a) in reflected light, (b) under crossed Nicols.
This photomicrograph suggests that significant sulfide alteration has occurred producing iron and sulfates that have re-precipitated out of solution as iron-oxyhydroxides (orangey-red) and gypsum (white). These 'cementing' phases likely serve to slow further oxygen penetration and therefore sulfide oxidation.

Photomicrograph 5: Photomicrograph of pyrrhotite pseudomorph (a) in reflected light, (b) under crossed Nicols.
Complete pseudomorphic replacement of pyrrhotite. In reflected light the alteration product looks relatively homogeneous, under crossed Nicols however it is clear that it is a multi-phased and progressive replacement. This grain was examined in further detail as described below.

Many of the mineralogical processes of interest and importance that control the acid generation/metal leaching character of mine waste occur on a small scale (micron level), in particular for tailings. For this reason, the use of the Scanning Electron Microscope (SEM) can greatly aid in the understanding and therefore interpretation of these mineralogical controls. For instance the pyrrhotite pseudomorph shown in Figure 5 above (turned 90 degrees) was examined under the SEM using both back-scattered electron imagery (BSE) and elemental x-ray maps as shown below.

Photomicrograph 6: Back Scattered Electron image of pyrrhotite pseudomorph (field of view=300 microns).
The BSE image allows for detailed examination of the texture of this grain. The pseudomorphic replacement occurred along a preferred orientation, most likely along the {0001} crystallographic parting plane. It also appears that portions of the pseudomorph are more crystalline than others.

Photomicrograph 7: Element x-ray maps of the same pseudomorph (field of view=300 microns).
The x-ray maps allow for non-destructive, semi-quantitative analysis of the chemistry of alteration products.

These mineralogical "tools" , although documenting a snapshot in time, illustrate clearly that alteration of sulfides is a progressive process with complex mineralogical changes occurring throughout and various intermediate phases dominating different stages. The above grain would be classified on the S.A.I. given above as a 9 or 10, depending on the alteration of other sulfides in the sample. If this grain was typical of an entire site, it may suggest that the rate of oxidation, and therefore acid generation, would be slowing. This conclusion could affect the selection of particular control technologies or closure requirements. For example, the selection of a complex tailings cover to prevent oxygen infiltration as a means of controlling sulfide oxidation at a site where the sulfides were typical of the above example may not be appropriate.

Other predictive tools that are commonly used include the acid-base-accounting (ABA) tests and other static and kinetic laboratory procedures. Interpretation of these results is critical and as proven more than once, very site specific. One of the issues most difficult to interpret is the neutralization potential (NP) of the ABA test. Carbonates play a dominant role in acid neutralization, but other minerals, depending on pH conditions, can contribute to NP. Plagioclase feldspar and biotite are two such minerals as seen in the two photomicrographs (8 and 9) presented below.

Photomicrograph 8: Transmitted light photomicrograph of weakly altered plagioclase feldspar.
Weak alteration of the plagioclase feldspar known as saussuritization is seen in this photomicrograph, the grain is also rimmed by fine grained iron-oxyhydroxide precipitating from solution.

Photomicrograph 9: Transmitted light photomicrograph of altered biotite grain.
Significant alteration has occurred in this biotite grain as evidenced by the loss in colour and pleochroism. Electron probe microanalyses of similar altered biotite grains showed a depletion in K, Fe and Mg indicating alteration to vermiculite. Similar results were reported by
Jambor (1994). This grain, as was seen with the plagioclase in Fig. 8. is also coated with a secondary phase, likely an iron-oxyhydroxide.

The identification of secondary alteration products also plays an important role in the prediction of pore water quality. Geochemical equilibrium speciation models, such as MINTEQA2 (Allison et al., 1991), are frequently used for these predictions. Pore water chemistry is dependent on the solubility constraints of the minerals through which the pore water migrates. Therefore, minerals identified during mineralogical characterization, in particular the secondary alteration products and carbonates, can be used in geochemical models as constraints and controls on the predicted pore water chemistry.

 

The contents of this web page are protected by copyright law. Please contact the authors for permission to re-use the contained information.

 


REFERENCES

Allison, J.D., Brown, D.S., and Novo-Grada, K.J. (1991), MINTEQA2/PRODEFA2, A Geochemical Assessment Model for Environmental Systems, version 3.0, EPA/600/3-91/021, 1991, Office of Research and Development, U.S. Environmental Protection Agency, Athens, GA.

Blowes, D.W. and Jambor, J.L. (1990), The Pore-water Geochemistry and the Mineralogy of the Vadose Zone of Sulfide Tailings, Waite Amulet, Quebec, Canada., Applied Geochemistry, vol. 5, pp. 327-346.

Clarke, D. (1993), NewPet for DOS, Dept. of Earth Sciences, Memorial University of Newfoundland, St. John's, NF.

Cross, W., Iddings, J.P., Pirsson, L.V. and Washington, H.S. (1902), A Quantitative Chemicomineralogical Classification and Nomenclature of Igneous Rocks, Journal of Geology, v10, p555-690.

Jambor, J.L. (1994), Mineralogy of Sulfide-rich Tailings and Their Oxidation Products., in Environmental Geochemistry of Sulfide Mine-wastes, Mineralogical Association of Canada Short Course Vol. 22., pp. 59-102.

Jambor, J.L. and Blowes, D.W. (1998), Theory and Application of Mineralogy in Environmental Studies of Sulfide-Bearing Mine Wastes, in Modern Approaches to Ore and Environmental Mineralogy (Cabri, L.J. and Vaughan, D.J., Eds.), Mineralogical Association of Canada Short Course Vol. 27, p367-402..

Jones, M.P. and Fleming, M.G. (1965), Identification of Mineral Grains, Elsevier, London, 102p.

Kwong, Y.T.J. and Ferguson, K.D. (1997), Mineralogical Changes during NP Determinations and their Implications, Proc. 4th International Conference on Acid Rock Drainage, Vancouver, BC, p435-447.

Lawrence, R.W. and Sheske, M. (1997), A Method to Calculate the Neutralization Potential of Mining Wastes, Environmental Geology, v32, p100-106.

McCrone, W.C., McCrone, L.B. and Delly, J.G. (1979), Polarized Light Microscopy, Ann Arbor Science Publishers, Ann Arbor, Michigan, 251p.

Mills, C. (1985), Specific Gravity Fractionation and Testing with Heavy Liquids, in SME Mineral Processing Handbook, American Institute of Mining, Metallurgical and Petroleum Engineers, New York, 30:44-52.

Mills, C. (1978), Mineralogy and Heavy Liquid Analysis in Gravity Separation, Short Course on Gravity Concentration Technology, University of Nevada, Reno, October, 25p.

Paktunc, A.D. (1998a), Characterization of Mine Wastes for the Prediction of Acid Mine Drainage, in Environmental Impacts of Mining Activities - Emphasis on Mitigation and Remedial Measures (Azcue, J.M., Ed.), Springer-Verlag, Berlin (in print).

Paktunc, A.D. (1998b), MODAN: An Interactive Computer Program for Estimating Mineral Quantities based on Bulk Composition, Computers & Geosciences, v24, n5, p425-431.

Paktunc, A.D. (1998c), Discussion of A Method to Calculate the Neutralization Potential of Mining Wastes, Lawrence, R.W. and Sheske, M., Environmental Geology (in print)

Ritchie, A.I.M. (1994), The Waste-rock Environment, in Environmental Geochemistry of Sulfide Mine-wastes, Mineralogical Association of Canada Shortcourse Handbook (J.L. Jambor and D.W. Blowes, eds.), vol. 22, pp. 133-161.


Return to ARD at Enviromine