Evidence of irreversible CO₂ intercalation in montmorillonite

Abstract

Mitigation of the global climate change via sequestration of anthropogenic carbon dioxide (CO₂) in geologic formations requires assessment of the reservoir storage capacity and cap rock seal integrity. The typical cap rock is shale or mudstone rich in clay minerals that may significantly affect the effectiveness of the CO₂ trapping. Specific objectives of this study were to conduct experimental investigation into the processes associated with CO₂ and H₂O trapped in swelling clay, namely, Wyoming and Texas montmorillonite powder. Combined (same-sample) multi-technique data – manometric sorption isotherm hysteresis, diffuse reflectance infrared spectroscopy ‘trapped CO₂’ fingerprints, irreversible X-ray diffraction patterns for the clay interlayer in intermediate hydration state, and HF acid digestion resulting in formation of non-extractable F:CO₂ adducts – corroborate a hypothesis that carbon dioxide molecules can be irreversibly trapped via anomalous extreme confinement in the galleries associated with montmorillonite interlayer, which may result in formation of carbonates in the longer term. Validation on Arizona montmorillonite lumps substantiated the evidence that such processes may occur in natural clay deposits but possibly on a different scale and at a different rate.

Introduction

Sequestration of greenhouse gases (GHGs) in geologic formations is a viable approach to mitigation of the global climate change. Options for geologic storage of anthropogenic carbon dioxide (CO₂) vary from saline aquifers and depleted oil and gas fields to unmineable coal seams (DOE/NETL, 2010). Important consideration is potential changes in caprock seal integrity, owing to interaction with CO₂. The typical caprock is shale or mudstone rich in clay minerals that may significantly affect the effectiveness of CO₂ trapping. Swelling clay minerals, such as smectite with aluminosilicate structure controlled by low-charge layers, can accommodate water and, potentially, carbon dioxide molecules in the interlayer region (Romanov et al., 2009a, Knudson and McAtee, 1973).

Dioctahedral smectites have a 2:1 lamellar structure with an octahedral sheet of repeating AlO₆ units sandwiched between two tetrahedral silicate (SiO₄) sheets. Isomorphic metal-ion substitutions in either layer (e.g., Al³⁺ replaced by Mg²⁺ or Fe²⁺ in the octahedral sheet or Si⁴⁺ replaced by Al³⁺ in the tetrahedral sheet) result in fixed-charge imbalance that is compensated by metal cations (e.g., Na⁺ or Ca²⁺) residing at or near the clay surfaces. The lamellae, thus held together by electrostatic forces, can be expanded by penetration of polar (or polarizable) molecules into the interlayer (Romanov et al., 2009a, Ferrage et al., 2005). Heterogeneity of smectites can lead to coexistence of layers with substantially different degrees of hydration, within the same crystallite (Ferrage et al., 2005).

Original debates around the CO₂ intercalation hypothesis stemmed from interpretations of the B.E.T. (Brunauer et al., 1938) surface area measurements on cation-exchanged montmorillonites, related to differences in sample preparation, sensitivities of cryogenic (−78 °C) adsorption apparatus, and theoretical assumptions with regard to organization of the lamellae in the clay matrix being affected by the cation exchange, dialysis, and drying procedures (Thomas et al., 1970). The first report that CO₂ can penetrate the interlamellar space of Cs-smectite (Thomas and Bohor, 1968) was initially challenged, on the grounds of scientific merit, by Aylmore et al. (1970) but was later re-affirmed by Fripiat et al. (1974), based on additional evidence using an infrared (IR) transmission spectrometer (at room temperature and −78 °C) and X-ray goniometer (20 °C to −70 °C to 20 °C) experiments on the same clays that were used by Thomas and Bohor (1968). Despite the very low CO₂ pressure (<1 atm.) used in their work, Fripiat et al. (1974) reported noticeable orientation dependency of the asymmetric stretching (ν₃) band in the 2350 cm⁻¹ region of the absorbed IR spectra of K- and Cs-smectite and significant swelling of Na- and Ca-smectite due to CO₂ condensation during the cooling cycle. Interestingly, the changes in d₀₀₁ spacing were reversible in both Na-exchanged (10.0 Å to 12.3 Å to 10.0 Å) and Ca-exchanged (13.2 Å to 14.2 Å to 13.2 Å) smectites but the changes in d₀₀₂ and d₀₀₃ peak positions appeared to be irreversible with the Ca-smectite. The IR spectra of Na- and Ca-smectites were not reported. Later, Fahmy et al. (1993) observed that the magnitude of saturated swelling of Mg-montmorillonite clay is not monotonically dependent on CO₂ pressure at 45 °C, which they attributed to a competition between the swelling caused by increasing solubility of added water in CO₂-rich phase and the hydrostatic pressure compaction effect but the exact mechanisms were not understood.Molecular-scale phenomena related to CO₂ interaction with swelling clays have recently drawn attention of environmental scientists. In 2008, NETL researchers observed that CO₂ molecules may permanently intercalate between the clay lamellae (Romanov and Soong, 2008) and suggested a sorption mechanism akin to nano-confinement (Romanov et al., 2009a); independently, Busch et al. (2008) conducted experiments on well-characterized shale samples (Muderong Shale, Australia) and reported a relatively high CO₂ sorption capacity compared to coal as well as poor reproducibility in repeat experiments on the same samples, which they attributed to strong CO₂ interaction with clay minerals present in the shale. Romanov et al. (2009b) reported irreversible swelling of Ca-smectite after exposure to CO₂ in prolonged pressure build-up experiments, with preliminary results suggesting a subsequent carbonate formation in the interlayer region. The sorption of volatile fluids in swelling clay was further investigated using ab initio and classical simulations (Cygan et al., 2010, Botan et al., 2010, Cygan et al., 2012), which showed that the hydrated clay system is capable of sorbing linear CO₂ molecules via intercalation. The disposition of CO₂ molecules and density profiles of atoms within interlayer of Na-montmorillonite from molecular dynamics (MD) simulations (Cygan et al., 2012) are consistent with Monte Carlo simulations for lower CO₂ content (Botan et al., 2010). Comparison of the preliminary experimental and theoretical results presented by Romanov et al. (2010a) was followed by advanced complementary studies (Giesting et al., 2012a, Giesting et al., 2012b, Romanov et al., 2010b). It was reported that the modeling results and experimental observations indicate that sufficient presence of water species in the interlayer region is necessary for intercalation of CO₂ (Romanov et al., 2010b). The following X-ray diffraction (XRD) studies (Giesting et al., 2012a, Giesting et al., 2012b, Schaef et al., 2012, Ilton et al., 2012, Hemmen et al., 2012) focused primarily on changes in d₀₀₁ basal spacing during and after the CO₂ treatment. The in situ (Giesting et al., 2012b) measurements replicating the pressure build-up experiments (Romanov et al., 2009b) confirmed the extent and irreversibility of matrix swelling observed ex situ. The reported XRD results indicate that the nature of the interlayer cations, degree of hydration, and the length of exposure to CO₂ can significantly affect the experimental outcome. However, the outcomes of these studies were inconsistent and specific nature of the CO₂-clay interactions governing this process was not fully understood. Among proposed mechanisms was formation of carbonate complexes that can be trapped in the interlayer via hydrogen bonding (Cole et al., 2010) or as part of the interlayer cation hydration shell (Giesting et al., 2012a).

The overall objective of this work was to develop a better understanding of the molecular interactions associated with CO₂ intercalation and the factors affecting CO₂ sorption in clay, especially the permanence of trapping, and to contribute to development of robust models for CO₂ storage in geologic reservoirs. Specific objectives were to conduct experimental investigation into the processes associated with CO₂ and H₂O trapped in swelling clays, namely, Wyoming and Texas montmorillonites, to provide evidence of the sorption mechanism via intercalation in the interlayer and to investigate the role of cation hydration as well as the longer-term changes in carbon dioxide bonding to the clay surfaces.