Thermal Studies of magnesium silicates from the Great Serpentinite Belt in New South Wales for CO2 sequestration by mineral carbonation in Australia

This present contribution addresses the most important impediment for a large scale implementation of sequestering CO₂ by mineralisation, the energy cost of dehydroxylation of serpentinite ores. The current work, based on a practical heat activation strategy, provides reasonable energy and cost estimates of the thermal treatment of serpentine. A practical heat activation strategy comprises prograde heating to yield a material containing 20 % OHres and employs heat integration to recover ~80 % of the sensible heat from the dehydroxylated mineral. This study also reconfirms that crystallinity primarily controls magnesium availability. Once dehydroxylated, overheating must be avoided to prevent the material from recrystallising. Significant loss of crystallinity commences by ~ 50 % OHres, peaks at about 20 % OHres, and terminates beyond 5 % OHres. Besides limiting crystallinity, entrapment of liberated water and the formation of hematite must be minimised or altogether eliminated. The fluidisation of the reactor bed and the use of CO₂ as purge gas combined with prior demagnetisation circumvent these limiting factors. Furthermore, the rapid thermal treatment of 75 µm particles allows for optimal conditions of temperature and rate of dehydroxylation and presents a practical means to increase the throughput and minimise reactor size. Examination of the prior studies on the dehydroxylation of serpentine minerals identified the knowledge gaps that require addressing. This review surveys the key concepts, including the chemical and physical transformations involving the proposed mechanisms, thermal stability, reaction kinetics, the formation of intermediates and products, associated heat requirements, factors that influence the reaction, as well as associated enhancements in both dissolution and carbonation. The review finds the activation processes must avoid recrystallisation of disordered phases to forsterite and enstatite, and minimise the partial pressure of water vapour, which engenders reverse reaction. In particular, this current study determined the suitability of the serpentinite resource from the Great Serpentinite Belt in New South Wales for thermal activation using thermal, X-ray and spectroscopic techniques. Experimental investigations designed to evaluate the thermal behaviour and structural transformations of serpentine minerals include thermogravimetry-derivative thermogravimetry, mass spectrometry, X-ray diffraction and magic angle spinning nuclear magnetic resonance. We focussed on antigorite as the model serpentine mineral for these studies, however, we include the other serpentine minerals, lizardite and chrysotile as well as partially serpentinised samples in subsequent energy measurements by thermogravimetry-differential scanning calorimetry. We first analysed the changes in antigorite’s derivative thermogravimetric curves (DTG) and deduced factors affecting the mass loss profiles to serve as guide in both fundamental studies and in the design of a scale-up dehydroxylation reactor. The imposed heating rate, type of purge gas, type of comminution and sample mass all influence the dehydroxylation curve. However, the results show no influence of material of construction of the heating vessel and flow rate of the purge gas. We report an important effect of oxidation of Fe²⁺ under air purge gas that occurs prior to dehydroxylation and leads to formation of hematite skins on serpentinite particles, slowing down subsequent mass transfer and increasing the processing temperature. Design considerations of a practical scale up reactor for activating serpentinite ores for storing CO₂ by carbonation ought to comprise rapid heating, proper size reduction, prior demagnetisation and fluidisation of the powdered bed. We then demonstrate the optimisation of heat treatment of antigorite, to provide a benchmark of an extreme case of activation among serpentine minerals. Antigorite was investigated non-isothermally via thermogravimetry-mass spectrometry and in-situ X-ray powder diffraction, its thermal reaction sequence elucidated and reaction kinetics subsequently modelled. Here we describe a predictive framework expedient to the thermal processing of serpentinites for the mineralisation of CO₂. We successfully modelled an optimised thermal dehydroxylation strategy for antigorite using a three-dimensional phase boundary reaction model (R3), with activation energy Ea of 160 kJ mol⁻¹ and a frequency factor A of 5.7 ± 4.1 × 10⁵ s⁻¹ (5.7 × 10⁵ s⁻¹ for dynamic and 1.6 × 10⁵ s⁻¹ for static stage). This strategy translates to a fast and efficient thermal processing in an optimally-sized calcining vessel. Furthermore, the results imply that activation of the more common serpentine minerals lizardite and chrysotile would be significantly less energy intensive as their dehydroxylation proceeds at lower temperatures than that of antigorite. We also investigated the local chemical environment of thermally treated antigorite to elucidate the structural transformations that take place during its dehydroxylation. A combination of ²⁹Si and ²⁵Mg magic angle spinning nuclear magnetic resonance (MAS NMR) spectroscopy, phase compositional analysis and morphological observations served to characterise these transformations. The results indicate that, whilst the octahedral sheet remained relatively immobile, the tetrahedral sheet demonstrated a range of movements. This progressive break-up of the silicate sheet leads to various SiO₄ tetrahedral polymerisation and the formation of two intermediate phases, dehydroxylate 1 and dehydroxylate 2. These metastable phases possess similar Mg octahedral environment, but differ in the local configuration of SiO₄ tetrahedra. The results suggest that crystallinity plays a significant role in antigorite dehydroxylation wherein partially amorphised regions dehydroxylate preferentially. For applications to mineralisation of CO₂, once dehydroxylated, the formation of new crystalline phases must be minimised in order to maximise the loss of crystallinity. The combined chemical, morphological and physical evidence implies that significant loss of crystallinity commences at about 50 % OHres until about 5 % OHres. Finally, we present realistic cost estimates based on a practical heat activation strategy for serpentinites. By comparing serpentinites from the Great Serpentinite Belt to those found in Coolac Serpentinite Belt, we found the former more suitable for heat activation under a practical heat activation strategy. Based on the research findings, the thermal activation of serpentine minerals from the Great Serpentinite Belt (GSB) presents a practical option to New South Wales for the mineralisation of CO₂. Heat activation of the GSB serpentinites is best performed via prograde heating to about 680 °C to yield an active material with 20 % OHres and requires an energy input of at least 541 MJ (tSerpentinite)⁻¹. The operational cost of thermally activating GSB serpentinites is about AU$ 1.25 per tonne of available active serpentine and with a corresponding CO₂ penalty of about 7 %.