Quantifying carbon fixation in trace minerals from processed kimberlite: A comparative study of quantitative methods using X-ray powder diffraction data with applications to the Diavik Diamond Mine, Northwest Territories, Canada

Abstract

The capacity of mine waste to trap CO₂ is, in some cases, much larger than the greenhouse gas production of a mining operation. In mine tailings, the presence of secondary carbonate minerals that trap CO₂ can therefore represent substantial fixation of this greenhouse gas. The abilities of three methods of quantitative phase analysis to measure trace nesquehonite (MgCO₃·3H₂O) in samples of processed kimberlite have been assessed: the method of reference intensity ratios (RIR), the internal standard method, and the Rietveld method with X-ray powder diffraction data. Tests on synthetic mixtures made to resemble processed kimberlite indicate that both the RIR and Rietveld methods can be used accurately to quantify nesquehonite to a lower limit of approximately 0.5 wt.% for conditions used in the laboratory. Below this value, estimates can be made to a limit of approximately 0.1 wt.% using a calibration curve according to the internal standard method. The RIR method becomes increasingly unreliable with decreasing abundance of nesquehonite, primarily as a result of an unpredictable decline in preferred orientation of crystallites. For Rietveld refinements, structureless pattern fitting was used to account for planar disorder in lizardite by considering it as an amorphous phase. Rietveld refinement of data collected from specimens that were serrated to minimize preferred orientation of crystallites gives rise to systematic overestimates of refined abundances for lizardite and underestimates for other phases. The resulting pattern of misestimates may be mistaken for the effect of amorphous and/or nanocrystalline material in samples. This effect is mitigated by collecting data from non-serrated specimens, which typically give relative errors on refined abundances for major and minor phases in the range of 5–20%. However, relative error can increase rapidly for abundances less than 5 wt.%. Nonetheless, absolute errors are sufficiently small that estimates can be made for the amount of CO₂ stored in secondary nesquehonite using the RIR method or the Rietveld method for abundances ⩾0.5 wt.% and a calibration curve for abundances <0.5 wt.%. The extent to which C is being mineralized in an active mine setting at the Diavik Diamond Mine, Northwest Territories, Canada, has been investigated. Rietveld refinement results and calibrated abundances for trace nesquehonite are used to estimate the amount of CO₂ trapped in Diavik tailings. Results of quantitative phase analysis are also used to calculate neutralization potentials for the kimberlite mine tailings and to estimate the contribution made by secondary nesquehonite.

Introduction

Emission of anthropogenic greenhouse gases (e.g., CO₂, CH₄, and N₂O) has been implicated as a cause of current warming of the Earth’s climate. Carbon dioxide is by far the most significant of these greenhouse gases, representing 77% of total anthropogenic emissions in 2004 (IPCC, 2007). By 2005, the global atmospheric concentration of CO₂ had increased to 379 ppm from a pre-1750 (i.e., pre-industrial) value of 280 ± 20 ppm. Approximately 2/3 of this increase is attributed to combustion of fossil fuels and ⅓ to changes in land-use patterns during the past 260 a (IPCC, 2007). It has been suggested that strategies for decarbonizing energy sources, increasing energy efficiency, and trapping and storing CO₂ must be developed and implemented in order to stabilize concentrations of atmospheric CO₂ and curtail the most damaging effects of anthropogenic climate change (e.g., Hoffert et al., 2002, Lackner, 2003, Pacala and Socolow, 2004, Broecker, 2007, IPCC, 2007).

Approximately 90% of C on Earth is fixed within carbonate minerals (Sundquist, 1985, Sundquist, 1993) and it is expected that these minerals will be the ultimate sink for most anthropogenic CO₂ on a timescale of 1 Ma (Kump et al., 2000). Storage of CO₂ in carbonate minerals is recognized as a safe and effective method for the sequestration of anthropogenic C (Seifritz, 1990, Lackner et al., 1995, Lackner, 2003). Precipitation of carbonate minerals in situ by dissolution of silicate mine residues is one potential implementation of this process. The development of secondary carbonate minerals has been documented in tailings at several mine sites in Canada: at the Kidd Creek Cu–Zn mine near Timmins, Ontario (Al et al., 2000), the Lower Williams Lake U mine near Elliot Lake, Ontario (Paktunc and Davé, 2002), and in chrysotile mine tailings at Thetford, Québec (Huot et al., 2003), Clinton Creek, Yukon Territory (Wilson et al., 2004), and Cassiar, British Columbia (Wilson et al., 2005). The C bound within secondary carbonate minerals in the tailings at some of these mines may not have had an atmospheric source. However, Wilson et al. (2009) demonstrate that tailings from the historical chrysotile mines at Clinton Creek, Yukon Territory and Cassiar, British Columbia are trapping and storing CO₂ from the atmosphere. Accelerating the uptake of CO₂ into tailings from active mines could reduce or offset the net greenhouse gas emissions of many mining operations.

Bulk geochemical methods for CO₂ abundance cannot distinguish among various carbonate minerals nor can they discern the difference between atmospheric, bedrock, biological, and industrial sources of C within minerals. However, the sources of bound C can be distinguished using radiocarbon and stable isotopes of C and O (Wilson et al., 2009). Also, automated point-counting techniques (e.g., mineral liberation analysis) cannot be used to quantify fine-grained minerals or hydrous minerals that are easily vaporized by an electron beam. As an alternative to point counting, the amount of CO₂ trapped within fine-grained, hydrous carbonate minerals can be estimated from weight-percent abundances determined with quantitative phase analysis using X-ray powder diffraction (XRPD) data.

At the Diavik Diamond Mine, Northwest Territories, Canada, efflorescent films of Ca-, Na- and Mg-carbonate minerals form in the tailings from the fine and coarse Processed Kimberlite Containment facilities (PKC). These minerals precipitate at the surface of kimberlite waste that is beached along the edge of a central pond used for storing process water. Based on the authors’ observations, the most common secondary carbonate mineral, and the one best preserved at depth, is nesquehonite (MgCO₃·3H₂O). Also, geochemical modelling by Rollo and Jamieson (2006) suggests that C mineralization may be occurring in waste kimberlite at the nearby EKATI Diamond Mine.

Processed kimberlite from Diavik contains a variety of minerals that are characterized by one or more of the following: (1) extensive solid solution, (2) structural disorder, and (3) severe preferred orientation. Furthermore, processed kimberlite at Diavik generally contains abundant serpentine and forsterite with minor to trace amounts of many other phases, resulting in complicated XRPD patterns that consist of many overlapped peak profiles. The combination of these factors presents a challenge for quantifying C mineralization with the Rietveld method (Rietveld, 1969). This is chiefly because the Rietveld method requires that the crystal structures and chemistry of the phases being analyzed be known and also gives less reliable results for minerals present at low abundances (e.g., Raudsepp et al., 1999).

Alternatively, Chung’s (1974) method of normalized reference intensity ratios (RIR) and similar methods have been used successfully to quantify trace abundances of minerals in multi-phase mixtures (e.g., Bish and Chipera, 1991, Omotoso et al., 2006). Calibration curves produced according to the internal standard method (Alexander and Klug, 1948) have also been used successfully to measure trace abundances of minerals (e.g., Sanchez and Gunter, 2006). These three methods are typically used in isolation on very different systems of minerals and, as such, it is difficult to recommend the use of one method over another for quantification of low abundances of minerals. Here the ability of Chung’s RIR method, the internal standard method, and the Rietveld method to measure minor to trace amounts of nesquehonite are assessed using weighed mixtures of pure mineral standards, prepared to simulate processed kimberlite. Based on the results of the tests, a procedure for accurate quantification of CO₂ trapping within trace minerals in kimberlite mine tailings is outlined. This procedure is subsequently applied to samples of natural processed kimberlite from the fine and coarse PKC at Diavik. Results of quantitative phase analysis are used to estimate trapping of CO₂ within nesquehonite at Diavik (from stoichiometry of this mineral) and to determine the contribution of nesquehonite to neutralization potential of the tailings according to the method of Jambor et al. (2007).