Evolved-Gas Analysis - American Chemical Society

Mining, Hydrology, Sedimentology, and Reclamation; Lexington, KY, 1985; pp. 139-145. 5. Hammack, R.W. In Proceedings of the Seventh Annual West Virgin...
0 downloads 0 Views 1MB Size
Chapter 28

Evolved-Gas Analysis

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 27, 2017 | http://pubs.acs.org Publication Date: December 20, 1993 | doi: 10.1021/bk-1994-0550.ch028

A Method for Determining Pyrite, Marcasite, and Alkaline-Earth Carbonates Richard W. Hammack Pittsburgh Research Center, U.S. Bureau of Mines, P.O. Box 18070, Cochrans Mill Road, Pittsburgh, PA 15236

An evolved-gas analysis technique has been developed for the simultaneous determination of pyrite and alkaline-earth carbonates in geologic materials. The technique can also be used to estimate oxidation rates for coal pyrite and therefore may improve the prediction of acid discharges resulting from mining activity. A programmable tube furnace was used to heat crushed-rock samples in a 10% oxygen atmosphere. The evolution of sulfur dioxide and carbon dioxide was monitored with respect to time and temperature using a quadrupole mass spectrometer. Sulfur-dioxide peaks attributable to the oxidation of sedimentary pyrite occurred between 380°C and 440°C; sulfur-dioxide peaks attributable to the oxidation of hydrothermal pyrite were present between 475°C and 520°C. Two CO peaks resulting from the decomposition of calcium carbonate (calcite) occurred at 380-370°C and at 550-650°C. Dolomite (CaMg(CO ) ) decomposed to yield CO at 780-900°C. Calibration curves were prepared by plotting SO - and CO -evolved gas peak areas versus the concentration of pyrite and calcium carbonate, respectively. The detection limit for pyrite was found to be about 3% pyrite, with a working range that extended beyond 30% pyrite. The detection limit for calcite was less than 0.5% calcite, with a working range to about 14% calcite. Evolved-gas analysis cannot distinguish between hydrothermal pyrite and sulfur bound to aromatic organic compounds. This limitation may prevent the use of this technique in the analysis of anthracite coal and related strata. 2

3 2

2

2

2

Acidic mine drainage (AMD) is one of the most persistent and serious sources of industrial pollution in the United States. Premining prediction of acidic mine drainage alerts mine operators to potential sources of acid discharge and allows them to plan mining operations and reclamation to minimize water-quality degradation. In geographical areas or coal seams that have historically been sources of AMD, This chapter not subject to U.S. copyright Published 1994 American Chemical Society Alpers and Blowes; Environmental Geochemistry of Sulfide Oxidation ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 27, 2017 | http://pubs.acs.org Publication Date: December 20, 1993 | doi: 10.1021/bk-1994-0550.ch028

432

ENVIRONMENTAL GEOCHEMISTRY OF SULFIDE OXIDATION

mine operators are required by law (1) to identify possible hydrologie consequences prior to opening a new mine. Concern in the United States regarding the possibility of acidic drainage has led to the requirement that an assessment of acid-discharge potential accompany each new mine-permit application. Overburden analysis to determine acid-discharge potential includes chemical tests to quantify the acidic or alkaline weathering products from each stratum overlying or directly underlying the coalbed to be mined. There are many proposed overburden analysis techniques ranging from direct chemical determinations to simulated weathering methods (2). The acidftase account, a direct chemical technique, is the most widely used because of its simplicity and low cost. The method is based on measuring the total-sulfur content of each lithologie unit and converting that value to an acid potential based on the stoichiometry of complete pyrite oxidation. The neutralization potential is determined for each lithology by its ability to neutralize strong acid. The two values, acid potential and neutralization potential, are both represented as calcium-carbonate equivalents and the net excess or deficiency of neutralizes is calculated. The acid/base account was originally developed as a quick method of identifying acid or alkaline weathering material for revegetation planning. Use of the acid/base account to predict drainage quality from heterogeneous mine spoils without considering the various other contributory factors (mine type, extent of reclamation, climate, and hydrology) is a serious overextension of the original intent of the method. Erickson (2) indicated that the acid/base account, as typically applied to overburden analysis, has Utile ability to predict drainage quality. The tendency of the acid/base account to overestimate both the acid potential and the neutralization potential may be partly responsible for its poor predictive capability. The acid potential is based on the total-sulfur content even though all sulfur compounds do not contribute acidity. For example, the sulfur in gypsum does not react to form acidity. The overestimation of the neutralization potential results from the fact that, although all carbonate forms are soluble to some degree in strong acid, not all forms are readily available in the mine environment. In addition, the dissolution of manganese and iron carbonates will provide no net neutralization if the mine-waste drainage is ultimately exposed to an aerobic environment, resulting in the oxidation of manganese and iron. This study was undertaken to develop an analytical method that would improve the predictive capability of the acid/base account. The first step was to improve the selectivity of analyses typically employed by the acid/base account. An ideal analytical procedure would be one that quickly and directly determined the sulfur species contributing acidity, and the carbonate species contributing alkalinity under field conditions. A previous study (3) indicated that the technique conventionally used for total-sulfur determinations (combustion-furnace ignition with infrared S0 detection) can be made more selective by operating at lower temperatures. For example, at 500°C, only S0 from the combustion of pyritic and organic sulfur is detected because sulfate sulfur is stable at this temperature. In noncarbonaceous samples, where the organic-sulfur content is negligible, the low temperature technique results in the direct determination of pyritic sulfur. Later 2

2

Alpers and Blowes; Environmental Geochemistry of Sulfide Oxidation ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

28.

HAMMACK

Evolved-Gas Analysis

433

studies (4,5) indicated that samples with low S0 initiation temperatures were more reactive and generated more acid in laboratory weathering tests. The evolved-gas analysis (EGA) technique used in this study is capable of simultaneously quantifying pyrite and various carbonate minerals in a single run. The EGA method monitors the evolution of S0 and C 0 produced by the following reactions: 2

2

2FeS(s) + 11/2 0 (g) - Fe 0 (s) + 4S0 (g)

(1)

Me C0 (s) - Me O(s) + C0 (g)

(2)

2

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 27, 2017 | http://pubs.acs.org Publication Date: December 20, 1993 | doi: 10.1021/bk-1994-0550.ch028

2

2

x

3

2

x

3

2

2

where Me = monovalent or divalent cation. The type of gas and temperature of evolution is characteristic of the original mineral or compound. Therefore, minerals can be tentatively identified by the temperatures at which certain gases are evolved. The amount of a particular mineral present in a sample is proportional to the partial pressure of the gas evolved. EGA may improve the predictive capability of the acid/base account by: 1.

determining pyritic sulfur more quickly than conventional ASTM sulfur forms speciation; and

2.

providing neutralization potentials based only on carbonates species that are known to generate alkalinity under oxidizing field conditions.

Methods An instrument designed specifically for EGA was assembled for this study (Figure 1). The design was adapted from an evolved-gas instrument constructed by LaCount and others (6). Major components include an electronic mass-flow controller/gas blender, a programmable tube furnace, a quadrupole mass spectrometer, a programmable analog-to-digital (A/D) converter, and a microcomputer. Three hundred mg of -60 mesh sample was diluted with 3 g of tungsten oxide to aid uniform heating. The sample was then placed in a 2.54 cm diameter by 50 cm long quartz tube and secured with either glass wool or quartz wool, depending upon the maximum temperature of the run. A 32 mm (1/8 in), Type Κ thermocouple was inserted into the sample and the whole tube assembly was placed in the furnace. A 10.0% oxygen/90.0% nitrogen gas mixture was introduced into the tube furnace where it flowed through the crushed sample material at a flow rate of 100 mL/min. The tube furnace used for this study was capable of performing two heating ramps with selectable heating rates, dwell temperatures, and dwelltimes.The sample was heated at a rate of 6°C/min from about 70°C until 380°C was attained. The heating rate was then decreased to 3°C/min until the run was terminated.

Alpers and Blowes; Environmental Geochemistry of Sulfide Oxidation ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

ENVIRONMENTAL GEOCHEMISTRY OF SULFIDE OXIDATION

434

Evolved gases were detected with a quadrupole mass spectrometer. The capillary inlet to the mass spectrometer was placed in the gas stream, immediately downstream from the sample. This placement minimized lag time between gas evolution and detection. The mass spectrometer could simultaneously monitor the ion current at 12 user-selected mass to charge ratios (M/e). For each gas, the ion current was then converted to partial pressure by a calibration factor. Gases typically monitored included: S0 (M/e = 64), C 0 (M/e = 44), COS (M/e = 60), H 0 (M/e = 18), H S (M/e = 34), 0 (M/e = 32), and C ^ (M/e = 30). The partial pressures of all monitored gases were transmitted to a microcomputer at each 1.5°C increase in temperature. Run time, temperature, and gas partial pressures were converted to ASCII files and written to floppy disk. Periodically, these files were transferred to a mainframe computer where graphics, Gaussian peak fitting, and peak integration were performed using library functions. 2

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 27, 2017 | http://pubs.acs.org Publication Date: December 20, 1993 | doi: 10.1021/bk-1994-0550.ch028

2

2

2

2

Results and Discussion Pyrite Determination. Initial efforts were directed toward resolving S0 peaks resulting from the oxidation of pyrite or marcasite from those attributable to the combustion of organic sulfur. Figure 2 is a thermogram of Rasa Coal, from the former Yugoslavia, which contains 8.67% organic sulfur and 0.08% pyritic sulfur. Because the pyritic-sulfur content is two orders of magnitude less than the organicsulfur content in this sample, it provides an EGA thermogram of organic sulfur in coal without the usual pyrite interferences. The thermogram (Figure 2) of the Rasa Coal exhibits coincident C0 , H 0 and S0 peaks at 325°C and 475°C. Aliphatic compounds, due to their higher hydrogen content, would be expected to evolve more water during combustion than aromatic compounds. In Figure 2, the H 0 peak area relative to the C 0 peak area is much greater at 325°C than at 475°C. This difference indicates that the peaks at 325°C and 475°C are likely due to the combustion of aliphatic and aromatic compounds, respectively. Less S0 evolution accompanies C 0 evolution at 325°C (aliphatic hydrocarbons) than at 475°C (aromatic hydrocarbons). LaCount and others (7) have shown that the evolution of S0 at 320°C is due to the oxidation of sulfur in nonaromatic structures, whereas the S0 evolution near 480°C is due to the oxidation of thiophenic and aryl sulfide type structures. Standards of sedimentary and hydrothermal pyrite were tested to determine the temperature range for sulfur dioxide evolution (Figure 3). Hydrothermal pyrite from the Noranda Mine in Québec, Canada evolved S0 between 430°C and 520°C, with the predominant peak at 485°C (Figure 3B). For hydrothermal pyrite, the S0 peak-evolution temperature decreased as the pyrite content increased. Because hydrothermal pyrite evolves S0 in the same temperature range as organic sulfur, the two species could not be distinguished if present in the same sample. Fortunately, organic sulfur and hydrothermal pyrite are found together only in the most thermally mature types of coal, such as the anthracite. Coal pyrite standards were obtained as the sink fraction from the heavy media (bromoform) separation of pyritic coal. Combustion furnace analysis indicated that these concentrates contained greater than 90% pyrite. Two S0 peaks attributable to 2

2

2

2

2

2

2

2

2 2

2

2

2

2

Alpers and Blowes; Environmental Geochemistry of Sulfide Oxidation ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

28.

HAMMACK

Evolved-G as

Flow controller

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 27, 2017 | http://pubs.acs.org Publication Date: December 20, 1993 | doi: 10.1021/bk-1994-0550.ch028

435

Analysis

X Valve Exit gas

Oxygen Flow controller Π

t7# . XValve



i Capillary

Programmable tube furnace 4-way valve, nitrogen and calibration gases Nitrogen

(ΣΖ3 m Personal computer

Quadrupole mass spectrometer

Figure 1. Schematic of the evolved gas analysis system used in this study.

Alpers and Blowes; Environmental Geochemistry of Sulfide Oxidation ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

436

ENVIRONMENTAL GEOCHEMISTRY OF SULFIDE OXIDATION

t .2

I

1

1

,

Γ

1

A K

1.0 -

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 27, 2017 | http://pubs.acs.org Publication Date: December 20, 1993 | doi: 10.1021/bk-1994-0550.ch028

ft

E

Y

\

1.7 pet 3.3 pet 6.7 pet 10.0 pet

|-

i: :i :: :

.6

•4 h

ε