PETROGRAPHY AND APPLICATION OF THE RIETVELD METHOD TO THE QUANTITATIVE ANALYSIS OF PHASES OF NATURAL CLINKER GENERATED BY COAL SPONTANEOUS COMBUSTION

Fine-grained and mainly reddish color, compact and slightly breccious and vesicular pyrometamorphic rocks (natural clinker) are associated to the spontaneous combustion of coal seams of the Cerrejón Formation exploited by Carbones del Cerrejón Limited in La Guajira Peninsula (Caribbean Region of Colombia). These rocks constitute remaining inorganic materials derived from claystones, mudstones and sandstones originally associated with the coal and are essentially a complex mixture of various amorphous and crystalline inorganic constituents. In this paper, a petrographic characterization of natural clinker, aswell as the application of the X-ray diffraction (Rietveld method) by mean of quantitative analysis of its mineral phases were carried out. The RIQAS program was used for the refinement of X ray powder diffraction profiles, analyzing the importance of using the correct isostructural models for each of the existing phases, which were obtained from the Inorganic Crystal Structure Database (ICSD). The results obtained in this investigation show that the Rietveld method can be used as a powerful tool in the quantitative analysis of phases in polycrystalline samples, which has been a traditional problem in geology.

José Antonio Henao M.1, Angélica María Carreño P.2, John Alexander Quintero D.3, Sol Angel Candela H.4, Carlos Alberto Ríos R.5, Miguel A. Ramos G.1 y Jesús Andelfo Pinilla A.1

1 Grupo de Investigación en Química Estructural (GIQUE), Escuela de Química - Universidad Industrial de Santander

2 ECOPETROL-Instituto Colombiano del Petróleo, Piedecuesta, Colombia

3 Carbones del Cerrejón Limited, Bogotá, Colombia

4 ECOPETROL S.A., Colombia

5Escuela de Geología, Universidad Industrial de Santander, Bucaramanga, Colombia Corresponding author: C. A. Ríos, E-mail: carios@uis.edu.co Tel/Fax: 57 7 6343457

Manuscript received: 28/01/2010 Accepted for publication: 14/05/2010

---

ABSTRACT

Key words: Rietveld method, Cerrejón, X-ray powder diffraction, natural clinker.

---

RESUMEN

Rocas pirometamórficas (clinker natural) de grano fino y coloración principalmente rojiza, compactas, y ligeramente brechadas a vesiculares están asociadas a la combustión espontánea de mantos de carbón de la Formación Cerrejón explotados por Carbones del Cerrejón Limited en la península de La Guajira (Región Caribe de Colombia). Estas rocas constituyen materiales inorgánicos remanentes derivados de arcillolitas, limolitas y areniscas originalmente asociadas con carbón y son esencialmente una mezcla compleja de diferentes constituyentes inorgánicos amorfos y cristalinos. En el presente trabajo, se llevó a cabo la caracterización petrográfica del clinker natural, así como la aplicación de la difracción de rayos-X de polvo (método de Rietveld) para el análisis cuantitativo de las fases minerales presentes de las rocas en estudio. El programa RIQAS fue utilizado para el refinamiento de perfiles de difracción de rayos-X de polvo, analizando la importancia de utilizar los modelos isoestructurales adecuados para cada una de las fases presentes, los cuales se obtuvieron de la base de datos Inorganic Crystal Structure Database (ICSD). Los resultados obtenidos en la presente investigación muestran que el método de Rietveld puede usarse como una poderosa herramienta en el análisis cuantitativo de fases en muestras policristalinas, lo cual ha sido un problema tradicional en el campo de las geociencias.

Palabras clave: método de Rietveld, Cerrejón, difracción de rayos X de polvo, clinker natural.

---

Introduction

Spontaneous combustion of coal has been reported in different coalfields of the world (USA: Heffern etal., 1983, 1993; foit etal., 1987; Cosca etal., 1989; Heffern & Coates, 1997, Lyman & Volkmer, 2001; India: Prakash et al.,1997; Romania: Radan & Radan, 1998; Australia: Ellyett & Fleming, 1974, New Zealand: Lindqvist et al., 1985, China: Zhang, 1998; de Boer et al., 2001). Fires associated with coal mines, which are preserved in the rock record as reduced volume burnt coal beds and pyrometamorphic rocks (natural clinker), represent global catastrophes that may be initiated for various reasons and are responsible for coronary and respiratory diseases, human death and destruction of ecosystems, contaminating air, water and soil (Stracher, 2007).

In Colombia, Candela & Quintero (2004) conducted a geological mapping of natural clinker at El Cerrejón mining operation, located in the Guajira Peninsula, northeastern Colombia. Carbones del Cerrejón Limited is dedicated to the exploration, production, transportation and shipment ofhigh quality thermal coal), which includes monitoring and structural control at depth. According to Quintero et al. (2009), the "in situ" spontaneous combustion of coal, with the consequent development of natural clinker, is a phenomenon that hinders the development of coal mining, not only because it creates areas of geotechnical instability, given the characteristics acquired by the adjacent rock when subjected to high temperatures reached during combustion of coal but also because it represents a reduction in tonnage of workable coal.

The term "natural clinker" refers to a series of pyrometamorphic rocks generated as a result of thermal transformation of sedimentary protoliths from the spontaneous combustion of coal beds (Figure 1). Although many coalfields contain such rocks, there is little information regarding quantitative data about them, and almost nothing about their forming conditions. The mineralogy of natural clinker depends on physical and chemical variables, such as chemical composition of the sedimentary protolith, temperature, degree of fusion and oxidation state, among others, and how one or more of these factors affects the stability of mineral phases may vary on a microscopic scale (Cosca et al., 1989). However, some field and petrographic observations combined with analytical data can help to deduce the relative importance of these factors. Moreover, the natural clinker should be studied not only in relation to its origin but also to its potential use as an indicator of coal resources. Therefore, future studies should consider benefits obtained by using mine tailings as natural clinker based on knowledge of their petrographic, structural, physical and chemical properties.

In general, natural clinker is a mineralogical and chemically complex mixture in which the amount of each mineral phase present depends on the original composition of the protolith from which it was formed by pyrometamorphism, the mixture characteristics, the burning conditions, and the thermal history of the material, among others (Jones et al., 1984; Cosca etal., 1989, Coates & Heffern, 1999; Heffern et al., 1993). Although a geological study was not carried out on the protolith from which natural clinker has been generated, we can deduct some mineralogical changes suffered in particular clay strata after burning by spontaneous combustion of coal beds. It can be inferred that the increase in temperature caused by spontaneous combustion evidently produced the dehydration and vitrification of clay minerals, the dissolution of carbonates and the formation of aluminosilicates. Therefore, mineralogical and textural changes observed probably occurred on a system with disequilibrium on a small scale, with the development of high-temperature metamorphic reactions (pyrometamorphism).

Petrographic studies on glass-bearing burned rocks reveal the occurrence solid solution series as magnetite-hercynite (FeAl2O4), ulvöspinel (Fe2TiO4), hematiteilmenite (FeTiO3), y magnesiumferrite along the spinel (MgAl2O4)-magnesiumferrite (MgFe2O4) join of Sp80Mf20 and along the magnesioferrita-hematita join of Mf60Hm40(Cosca etal., 1989). Typical high temperature non magnetic minerals found in pyrometamorphic rocks associated with coal combustion appear to include several Fe and Al rich clinopyroxene, melilita, cristobalite, tridymite, mullite, cordierite and fayalite solid solutions, as well as glass (Foit et al., 1987; Cosca etal., 1989; Râdan & Râdan, 1998). Glass compositions have been used by Cosca et al. (1989) to estimate the minimum temperatures of formation of natural clinker, which vary between 1020 and 1400 oC. These authors suggest that different degrees of melting, mixing, crystallization, volatilization and liquid immiscibility in conditions of disequilibrium may be involved in the formation of such pyrometamorphic rocks. However, it is important to establish the type and amount of these magnetic phases in natural clinker, which affect its properties and, therefore, is necessary to identify and quantify them.

The determination of mineral phases and chemical composition of natural clinker is vital in resolving problems concerning their genesis, correlation of coal beds, characterization of the protolith and assessment of potential environmental impacts of coal combustion. Natural clinker is one of the contributing materials that could affect the environment, especially water chemistry, as important variables like the original composition of the protolith from which this geomaterial comes, burning conditions that generated it, and its particle size and mineralogy, influence the distribution and mobility of trace elements. Therefore, in response to environmental problems that occur as a result of the disposal of solid waste from coal mining, within which natural clinker lies, it is important to undertake studies to establish the environmental impacts of using this byproduct of coal in the field of materials science, based on its variability in quality and environmental impacts and human health.

The bulk chemical composition similarity of natural clinker with the volcanic materials from which the natural zeolites are originated by post-magmatic hydrothermal activity has motivated attempts of making zeolite from this geomaterial by Ríos and co-workers (Ríos & Williams, 2008; Ríos et al., 2008a, 2008b, 2008c; Sandoval et al.2009), justifying the development of future investigations in the field of the synthesis of new materials with potential industrial applications. According to them, zeolitic materials, such as Na-phillipsite, hydroxysodalite, hydroxycancrinite, K-chabazite, zeolite K-F, faujasite and zeolite Linde Type A can be synthesized by alkaline activation of natural clinker.

Recently, Sandoval et al. (2009) synthesized almost pure analcime with a high degree of crystallinity from natural clinker, optimizing the experimental conditions used by the pioneering work of Ríos & Williams (2008). Although its potential application might consume only a small part of the natural clinker produced by coal combustion, end products could reach a much higher added value than that currently presenting this geomaterial in the coal industry of Colombia.

The quantitative analysis using the Rietveld method has gained popularity in recent years as a rapid and accurate quantitative technique (O'Connor & Raven, 1988, Taylor & Matulis, 1991; Bish & Post, 1993; Weidler et al., 1998; Raudsepp et al., 1999; Gualtieri, 2000; de la Torre et al., 2001). Quantitative analysis by the Rietveld method (Rietveld, 1969; Young, 1993) has huge advantages over traditional methods using integrated intensities ofa small set of reflections in limited angular intervals, since the overlap of the lines of a pattern of powder diffraction, especially in mineralogically complex samples such as the natural clinker, make it virtually impossible to carry out a rigorous analysis by conventional methods. The analytical procedure using the Rietveld method is based on the relationship between the refined scale factor for each phase and its weight fraction in the multiphase mixture (Hill & Howard, 1987; Bish & Howard, 1988). This method makes efficient use of all information contained in the diffraction pattern and converts it into an effective analysis procedure without the use of standards, because it can be done on the diffraction data collected without the addition of an internal standard and without the need for calibration curves.

The only requirement to consider is that there must be a reliable isostructural model of each of phases to be quantified. On the other hand, if the process of quantitative analysis by the Rietveld method is successful on high quality diffraction data, the method can be used to quantify crystalline phases whose weight percentage are below 1.0%, and still can be used to determine the amount of amorphous material present in the multiphase mixture. In the last case, it requires adding a known amount of an internal standard to the sample and thus properly quantifies the material (Gualtieri, 2000).

The purpose of this study was to conduct a petrographic characterization of the natural clinker produced by spontaneous combustion of coal beds Cerrejón Formation and to apply the Rietveld method to the quantitative phase analysis of this geological material. Therefore, the implementation of the proposed methodology could be very important to quality control of geological materials for industrial use.

Materials and methods

Petrographic analysis of different samples of natural clinker were carried out to establish their textural and structural features as well as their mineralogy, using the following analytical techniques: transmitted light microscopy (Nikon triocular microscope, model Labophot2-POL) of the School of Geology and image analyzer Leica of the School of Metallurgical Engineering and Materials Science at the Universidad Industrial de Santander, and scanning electron microscope ZEISS EVO50 of the Research Center of Applied Sciences at the University of Wolverhampton (analytical conditions: EHT = 20.00 kV, beam current 100 uA, Signal A = SE1, WD = 8.0mm).

The preparation and assembly of natural clinker samples for X-ray diffraction (XRD) was carried out at the Laboratory of Structural Chemistry of the School of Chemistry at the Universidad Industrial de Santander. Samples were homogenized in an agate mortar, and then mounted on an aluminum specimen holder using the technique of lateral filling (McMurdie, 1986). The process of disintegration and separation of clays (about 2µm) involved the separation of minerals by particle size using the Stokes' Law. Chemical and thermal treatments were conducted as follows: saturation with K+ and Mg2+ ions, calcinations at 350 and 550 °C and solvation with ethylene glycol. XRD patterns ofthe natural clinker were recorded with a RIGAKU D/MAX IIIB diffractometer operating in Bragg-Brentano geometry with Cu-a1 radiation (k = 1.5406 Å) at 40 kV and 20 mA and graphite monochromation. The scan parameters were step size 0.02°, dwell time 2 s and 20 range 2-70°. Phase identification was performed through the process of comparison search (Search/Match) using the crystallographic database Powder Diffraction File (PDF-2) from International Centre for Diffraction Data (ICDD). Full-pattern Rietveld refinement using RIQAS versión 4.0 (MDI, Rietveld Quantitative Análisis and Whole Pattern Fitting, 2006) was performed to quantify the amounts of phases in the natural clinker using a 20% by weight of aluminum oxide as internal standard.

Results and discussion Outcropping clinker

One of the most remarkable features in the areas of Cerrejón coal mining complex is the occurrence of natural clinker as an indicator of ancient fires produced by spontaneous combustion of coal seams (Candela & Quintero, 2004; Quintero et al., 2009). Natural clinker includes a variety of thermally affected rocks, which are easily recognizable on the field thanks to their mainly red color in a variety of intensities. Figure 1 illustrates a typical natural clinker outcrop, showing the reddening intensity due to the rock's alteration degree. However, it can also display several tones of black, gray, green and cream or even white, easily contrasting with the grayish hue of coal. It is a fine-grained particle size, compact and highly fractured rock, mainly reddish brown to red in colour due to the presence of hematite. These rocks represent a complex mixture of various amorphous and crystalline inorganic constituents and contain remaining inorganic materials derived from shales, sandy mudstones and grey sandstones originally associated with coal. In general, it is constituted by rocks of mainly clay texture but it could also be sandy. Natural clinker also retains much of the original rock characteristics as relict sedimentary structures and even fossil plant remains within layered sandstone and mudstone survive. The occurrence and distribution of natural clinker is described in more detail by Quintero et al. (2009). Evidence supports the fact that claystones, siltstones and sandstones adjacent to coal beds were metamorphosed by the spontaneous combustion of coal, recording continuous mineralogical and textural changes from relatively unaltered rocks to molten material (vesicular slag) during pyrometamorphism.

Petrography

Natural clinker includes a variety of pyrometamorphic rocks, which vary in their degree of thermal alteration, according to their texture and structure, original composition and position on the burning area. These rocks are characterized by fine-grained particle size and usually red color in different shades, although black, gray, green, cream or white varieties also occur. It is relatively hard and compact but disintegrates at the touch, and it can show a sugary and slight brecciate vesicular appearance. General macroscopic characteristics of natural clinker varieties of the Cerrejón Formation are shown in Figure 2. Natural clinker usually shows cooked sandy and clay textures (Figures 2a and 2b), the last of them similar to a common building brick, which is characterized by its reddish-orange color, shiny fracture rather than opaque similar to jasper and intense oxidation, which has been accompanied by a hardening of the rock through the transformation of clay minerals in hematite. This variety sometimes exhibits very specific characteristics, taking into account that relics of the original sedimentary structure are preserved, which are still recognized by the millimeter scale "tigerstripe" pattern of its color (Figure 2c), with orange-red, locally purple, areas and light gray areas, the latter as relics of the original claystone with scalloped and parallel edges to the stratification surface. Orange-red areas present red edges, due to the formation of hematite, which tends to permeate the rock, producing relatively hard cement. Light gray areas are darker toward the contact with red-orange areas, due to the presence of black magnetite. Figure 2d illustrates an example of a natural clinker similar to a vesicular slag with a characteristic spongy texture. The distinguishing characteristics of this variety are shown in Figure 3. A glassy material of gray to grayish brown color occurs filling vesicles or along fractures that are arranged perpendicular or diagonal and, in some cases, parallel to the stratification surface (Figure 3, top). A microphotography captured in an image analyzer (Figure 3, bottom) illustrates the vesicular slag and sponge appearance of this variety. Under transmitted light microscopy, it is completely isotropic, except for tiny particles of quartz. Rogers (1918) interprets this variety as the result of hot gases (350°C) which escaped from the zone of spontaneous combustion of coal beds and affected the adjacent rock to the molten state, which then hardens into a vesicular glass.

Nevertheless, faulting locally generates breccia zones that clearly demonstrate that spontaneous combustion existed before the last fault's activity period since it contains mixed fragments of natural clinker, fresh rock and coal (Figure 2e). Quintero et al. (2009) reported that calcareous rocks of the Cogollo Formation in the hanging wall of the Cerrejón thrust fault, and immediately over this tectonic discontinuity, are thermally affected and present reddish, cream and pink to purple colours, preserving already all of the original characteristics including fossil fragments (Figure 2f).

Figure 4 illustrates images of scanning electron microscopy (SEM) of natural clinker particles, which are characterized by their fragmentation and blocky morphology, with blocky, angular fragments with or without vesicles, occasionally exhibit stepped fractures.

X-ray difraction

Figure 5 shows the XRD patterns of natural clinker in the range of 20 from 2-40°, using Cu-a1 radiation, in which sharp reflections are observed corresponding to high intensities of quartz (PDF No. 010-46-1045). Similarly, there are broad bands with low relative intensity which correspond to hematite (PDF No. 010-73-2234), anatase (PDF No. 010-72-7058) and minerals of smectite and mica groups. Taking into account that the peak intensitycorresponding to quartz is larger than those from other mineral phases, it was necessary to carry out the separation of the clay fraction in order to obtain separated XRD patterns (not shown here) for the aluminosilicates. The saturation with Mg2+ and subsequent glycosylation of natural clinker samples (tigerstripe type, NC-1, and highly oxidized type, NC-2) produced displacements of the reflection (001) of the band corresponding to the smectite group from 15.07 to 17.38 Å for sample NC-1 and from 15.17 to 17.52 Å for sample NC-2. Additionally, strong reflections were observed, indicating the presence of montmorillonite and discarding the presence of vermiculite in both samples with predominance of interchangeable ion Mg+2principally in sample NC-2 in which it looks a greater degree of swelling. In the region between 7.14 and 7.20 Å, there is a wide band of very low intensity which disappears by thermal treatment at 550 °C. This behavior is characteristic of kaolinite which collapses at this temperature and is transformed into metakaolin (amorphous phase). Kaolinite was not detected in the qualitative and quantitative analysis because it is below the detection level, but it was identified after the thermal treatment of the clay fraction. Similarly, the presence of illite (interplanar spacings at 9.93-9.97 Å, 4.98-5.00 Å and 3.32-3.34 Å for sample NC-1 and 9.7-10.04 Å, 5.00 Å and 3.32-3.33 Å for sample NC-2) was confirmed, which remains stable after thermal treatments. XRD profiles of the original (untreated) and treated (saturated with ethylene glycol and heated to 550 °C) clay are shown in Figure 6.

Figure 7 shows the observed and calculated XRD patterns and their differences after refinement by the Rietveld method, where it can be observed the good fit of data, indicating that isostructural models chosen for each phase were appropriate. The models used for illite, hematite, quartz and anatase were taken from the Inorganic Database (ICSD, 2007 /1). The montmorillonite model was taken from the literature (Viani et al., 2002), assuming a montmorillonite with some degree of disorder and interchangeable Ca+2 and Na+2 ions.

For illite, hematite, quartz and anatase structure factors were calculated using the atomic scattering factors. The background noise was fitted with a polynomial function and the peak profile was modeled using the Lorentzian function of Pearson VII. The cell constants, scale factor and coefficients corresponding to sample displacement and asymmetry were also refined. For the refinement of montmorillonite, the same previously described factors were calculated, except atomic positions, which were left fixed. This strategy avoided overestimating the content of this phase in natural clinker samples. Similarly, during the refinement of the illite some degree of preferential orientation for the reflection (11-1) was observed. This aspect, which is one of the most serious problems in quantitative analysis, was largely corrected using the March function (Dollase, 1986).

Table 1 shows the residual values defined by RIQAS, Rwp y Ri3ragg, for final refinements by least squares together with the percentages of the phases. Rwp values (20.33% and 20.25% for samples NC-1 and NC-2, respectively) indicate that the refinement as a whole has been satisfactory if we consider that natural clinker samples consist of a mixture of minerals which hinders the achievement of lower refining indicators. Low values of RBragg (5.05-10.26% and 3.99-12.28% for samples NC-1 and NC-2, respectively) achieved in each phase indicate that the structural model selected was appropriate. However, for the montmorillonite slightly higher values (12.51% and 14.32%) were observed, indicating that some reflections have been completely unadjusted. Moreover, we can notice that despite performing the correction for the preferred orientation of the reflection (11-1) of illite, the standard deviation of quantification is the highest (±0.4 and ±0.5 for samples NC-1 and NC-2, respectively), but this value is below the average standard deviations for mineral phases (Gualtieri, 2000). In general, very small deviations in the quantitative analysis of the identificated crystalline phases were observed. In accordance, the higher content of amorphous material that could be associated with NC-2 sample was exposed to a higher temperature during spontaneous combustion, which may be related to a possible conversion of quartz to amorphous silica.

TABLE 1

Residual values of Rwp and RBragg as a whole indicate that the refinement has been satisfactory and that the structural models used were appropriate. The presence of amorphous material (2.7% for sample NC-1 and 3.3% for sample NC-2) is not an impediment to identify and quantify the mineralogy of the natural clinker.

A significant amount of amorphous material appears to reduce the crystalline signals. Many of these signals contain a large number of reflections which overlap with those of the other phases. Quantitative analysis of amorphous material also looked at all the factors that could affect the results. It is evident that natural clinker contains amorphous material, which cannot be detected directly by XRD, although it is important to understand the thermal history of natural clinker. To quantify the amorphous material by using the program RIQAS a crystalline standard (free amorphous) was added to the natural clinker in a known percentage. However, the higher content of amorphous material was observed in the natural clinker with glass material.

Conclusions

In the present work, the structural information on natural clinker associated to spontaneous combustion of coal of the Cerrejón Formation was obtained from high-quality powder diffraction patterns. The obtained crystallographic data were determined by Rietveld analysis.

Rietveld analysis of natural clinker reveals that it contains a large number of crystalline phases, many apparently present in small amounts that could be part of a series of solid solutions as a result of the thermal alteration process.

The results obtained in this study indicate that natural clinker is a complex chemical and mineralogical material. Nevertheless, it has been shown that quantitative analysis by the Rietveld method is appropriate to study the mineralogy and crystallochemistry of phases present in this geomaterial producing reliable results.

Acknowledgments

We thank to Carbones del Cerrejón Limited for supporting S. Candela and J. Quintero during fieldwork to carry out the collection of geological material covered by this study. We gratefully acknowledge the Programme Alban, the European Union Programme of High Level Scholarships for Latin America, scholarship No. E05D060429CO, and the Universidad Industrial de Santander for funding the doctoral program of C. Ríos. Special thanks to these institutions for providing research facilities. The authors also thank the anonymous reviewers for critically reading the article, as well as other persons and entities for the support offered.

References

Bish, D., Howard, S., Quantitative phase analysis using the Rietveld Method, J. Appl. Crystallogr. 21, 1988, pp. 86-91.

Bish, D., Post, J., Quantitative mineralogical analysis using the Rietveld full-pattern fitting method, Am. Mineral., 78, 1993, pp. 932-940.

Candela, S.A., Quintero, J.A., Cartografía de las zonas de clinker en las áreas de minería de la mina el Cerrejón. Albania, Guajira, Tesis de pregrado en Geología, Universidad Industrial de Santander, Bucaramanga, 2004.

Coates, D., Heffern, E., Origin and geomorphology of clinker in the Powder River Basin, Wyoming and Montana: Wyoming Geological Association 50th Annual Field Conference Guidebook, 1999, pp. 211-229.

Cosca, M., Essene, E., Geissman, J., Simmons, W., Coates, D., Pyrometamorphic rocks associated with naturally burned coal beds, Powder River Basin, Wyoming, Am. Mineral., 74, 1989, pp. 85-100.

De Boer, C., Dekkers, M., Van Hoof, T., Rock-magnetic properties of TRM carrying baked and molten rocks straddling burnt coal seams, Phys. Earth Planet. Inter., 126, 2001, pp. 93-108.

De la Torre, A., Bruque, S., Aranda, M., Rietveld quantitative amorphous content analysis, J. Appl. Crystallogr., 34, 2001, pp. 196-202.

Dollase, W.A., Correction of intensities for preferred orientation in powder diffractometry: Application of the March model, J. Appl. Crystallogr., 19, 1986, p. 267.

Ellyett, C., Fleming, A., Thermal infrared imagery of the burning mountain coal fire, Remote Sens. Environ., 11, 1974, pp. 221-229.

Foit Jr., F., Hooper, R., Rosenberg, P., An unusual pyroxene, melilite, and iron oxide mineral assemblage in a coal-fire buchite from Buffalo, Wyoming, Am. Mineral., 72, 1987, pp. 137-147.

Gualtieri, A., Accuracy of XRPD QPA using the combined Rietveld-RIR method, J. Appl. Crystallogr., 33, 2000, pp. 267-278.

Heffern, E., Coates, D., Naeser, C., Distribution and age of clinker in northern Powder River Basin, Montana (Abstract), Am. Assoc. Pet. Geol. Bull., 67, 1983, p. 1342.

Heffern, E., Coates, D., Whiteman, J., Ellis, M., Geologic map showing distribution of clinker in the Tertiary Fort Union and Wasatch Formations, northern Powder River Basin, Montana. U.S. Geological Survey Coal Investigations Map C-142, scale 1:175,000, 1993.

Heffern, E., Coates, D., Clinker - its occurrence, uses, and effects on coal mining in the Powder River Basin, in Jones, R.W., and Harris, R.E., eds., Proceedings of the 32nd annual forum on the geology of industrial minerals: Wyoming State Geological Survey Public Information Circular No. 38, 1997, pp. 151-165.

Hill, R., Howard, C., Quantitative phase analysis from neutron powder diffraction data using the Rietveld method, J. Appl. Crystallogr. 32, 1987, pp. 841-849.

Inorganic Crystal Structure Database (ICSD). FIZ Karlsruhe Help Desk Hermann-von-Helmholtz-Platz 176344 Eggenstein-Leopodshafen Germany, 2007/I. http://www.fiz-karlsruhe.de/icsd.html

International Centre for Diffraction Data 12 Campus Boulevard Newtown Square, PA 19073-3273 USA.http://www.icdd.com/

Jones, A., Geissman, J., Coates, D., Clinker deposits, powder river basin, Wyoming and Montana: A new source of high-fidelity paleomagnetic data for the quaternary, Geophys. Res. Lett., 11, 1984, pp. 1231-1234.

Lindqvist, J., Hatherton, T., Mumme, T., Magnetic anomalies resulting from baked sediments over burnt coal seams in southern New Zealand. N.Z., J. Geol. Geophys., 28, 1985, pp. 405-412.

Lyman, R., Volkmer, J., Pyrophoricity (spontaneous combustion) of Powder River Basin coals - considerations for coalbed methane development. Coal Report CR01-1, Wyoming State Geological Survey, Laramie, Wyoming, 2001.

Materials Data Inc (MDI), Version 4.0. Rietveld Quantitative Analysis and Whole Pattern Fitting, RIQAS, 2006.

McMurdie, H., Morris, M.C., Evans, E.H., Paretzkin, B., Wong-Ng, W., Methods of producing standard X-ray diffraction patterns, Powder Diffr., 1, 1986, p. 40.

O'Connor, B., Raven, M., Application of the Rietveld refinement procedure in assaying powder mixtures, Powder Diffr., 3, 1988, pp. 2-6.

Prakash, A., Gupta, R., Saraf, A., A Landsat TM based comparative study of surface and subsurface fires in the Jharia coalfield, India, Int. J. Remote Sens., 18, 1997, pp. 2463-2469.

Quintero, J.A., Candela, S.A., Ríos, C.A., Montes, C., Uribe, C., Spontaneous combustion of the Upper Paleocene Cerrejón Formation coal and generation of clinker in La Guajira Peninsula (Caribbean Region of Colombia), Int. J. Coal Geol., 80, 2009, pp. 196-210

Râdan, S., Râdan, M., Rock magnetism and paleomagnetism of porcelanites/clinkers from the western Dacic Basin (Romania), Geol. Carpath., 49, 1998, pp. 209-211.

Raudsepp, M., Pani, E., Dipple, G., Measuring mineral abundance in skarn. I. Rietveld method using X-ray powder-diffraction data, Can. Mineral., 37, 1999, pp. 1-15.

Rietveld, H., A profile refinement method for nuclear and magnetic structures, J. Appl. Crystal., 2, 1969, pp. 65-71.

Ríos, C.A., Williams, C.D., Synthesis of zeolitic materials from natural clinker: A new alternative for recycling coal combustion by-products, Fuel, 87, 2008, pp. 2482-2492.

Ríos, C.A., Williams, C.D., Roberts, C.L., Removal of heavy metals from acid mine drainage (AMD) using fly ash, natural clinker and synthetic zeolites, J. Hazard. Mater., 156, 2008a, pp. 23-35.

Ríos, C.A., Williams, C.D., Roberts, C.L., Fullen, M.A., Treatment of acid mine drainage (AMD) using natural clinker-based zeolites, XXXI Annual British Zeolite Association Conference, Keele (England), 31 March - 2 April, 2008b.

Ríos, C.A., Williams, C.D., Roberts, C.L., Fullen, M.A., Synthetic faujasite based on coal by-products for the treatment of acid mine drainage (AMD), Proceedings of the 2nd International Conference on Engineering for Waste Valorization, Patras, Greece, 3-5 June, 2008c.

Rogers, G., Baked shale and slag formed by the burning of coal beds, U.S. Geological Survey Professional Paper 108-A, 1918, pp. 1-10.

Sandoval, M.V., Henao, J.A., Ríos, C.A., Williams, C.D., Apperley, D.C., Synthesis and characterization of zeotype ANA framework by hydrothermal reaction of natural clinker, Fuel, 88, 2009, pp. 272-281.

Stracher, G.B., Geology of coal fires: Case studies from around the world. The Geological Society of America. Boulder, Colorado, 2007.

Taylor, J., Matulis, C., Absorption contrast effects in the quantitative XRD analysis of powders by full multiphase profile refinement, J. Appl. Crystallogr., 24, 1991, pp. 14-17.

Viani, A., Gualtieri, A., Artioli, G., The nature of disorder in montmorillonite by simulation of X-ray powder patterns, Am. Miner., 87, 2002, pp. 966-975.

Young, R., The Rietveld Method. International Union Crystallography, Oxford University Press, New York, 1993.

Weidler, P., Luster, J., Schneider, J., Sticher, H., Gehring, U., The Rietveld method applied to the quantitative mineralogical chemical analysis of a ferralitic soil, Eur. J. Soil Sci., 49, 1998, pp. 95-105.

Zhang, X., Coal Fires in Northwest China: Detection, monitoring, and prediction using remote sensing data. Ph.D. Thesis in Geology, Delft University of Technology, The Netherlands, 1998.