Dissolution of Thermally Activated Serpentine at Flue Gas Conditions – Near Equilibrium Dynamics

The rate limiting step for a CO₂ mineralization process is in ensuring a continuous supply of free Mg²⁺/Ca²⁺ cations at high concentrations for carbonate precipitation. The metal cations are usually obtained by dissolving silicate rocks. The great abundance of the serpentine group makes it an attractive feedstock for mineral carbonation. However, it suffers from slow dissolution kinetics when compared to several other less abundant natural minerals. Thermal activation through dehydroxylation has shown to accelerate the dissolution kinetics for serpentine [1]. However, heat activation is an energy intensive process and adds substantially to the costs. The use of direct flue-gas mineralization in aqueous medium avoids the costs associated with the capture of CO₂.

Dissolving activated serpentine offers several advantages over a process that uses pH-tuning additives to dissolve natural serpentine. These include the simplicity of the aqueous chemistry, the non-necessity for additive recovery and the dissolution at lean operating conditions. Earlier, we had studied the far-from-equilibrium, non-steady state dissolution kinetics for 75% dehydroxylated lizardite particles under temperatures (T) and partial pressures of CO₂ (pCO₂) not exceeding 120°C and 2 bar, respectively. The dissolution was performed in a liquid and gas flow-through set-up operated at high flow rates to keep the Mg²⁺ concentrations low. Up to 80% of the magnesium could be extracted within 100 min [2]. We had also developed the first kinetic model that describes the dissolution of dehydroxylated lizardite particles at far-from-equilibrium conditions [3]. In a real CO₂mineralization process, the dissolution of these particles will be subject to influences arising from thermodynamic constraints and secondary precipitation phenomena. Nevertheless, the far-from-equilibrium kinetic study experimentally demonstrated the rapid dissolution kinetics for dehydroxylated lizardite. It also allowed for the study of the intrinsic dissolution kinetics, independent of thermodynamic effects. This is important since when one develops an optimal dissolution strategy, solutions aimed at tackling kinetic and thermodynamic limitations will not be identical.

Having understood the far-from-equilibrium kinetics, we have followed it up by studying the dissolution dynamics under environments representative of a real process condition, i.e. near-equilibrium conditions. A parametric study on the non-steady state dissolution of dehydroxylated lizardite particles was performed over temperature ranges 30°C ≤ T ≤ 90°C and CO₂ partial pressures of 0.025 ≤ pCO₂ ≤ 1.0 bar in batch operation inside a 100 mL Teflon reactor. A conductivity probe (InLab®731, Mettler Toledo) was calibrated that enabled continuous online measurement of the concentration of Mg²⁺ ions during dissolution. Experiments were performed for durations up to 3 hours. Aqueous samples were also taken and analyzed for dissolved silica concentrations. Different amounts of activated serpentine were dissolved in order to achieve high values for the activity ratio term a_Mg2+/(a_H+)². The upper limit values for this activity ratio term were determined by the solubility of the kinetically relevant magnesium carbonate at the operating conditions applied. This was done to avoid precipitation of the magnesium carbonate during the dissolution step. Two different particle size fractions were also tested. A total of 135 non-steady state experiments were performed in this study.

The far-from-equilibrium model that we had developed [3] is being updated in order to accommodate the equilibrium features that affect the dissolution dynamics. These include:

1)      A higher operating pH (5.5 – 9): Specific dissolution rate expression for forsterite (one of the silicate species that is present in activated serpentine) changes with pH due to a change in the dissolution mechanism [4].

2)      Reduction in the thermodynamic driving forces for dissolution as each silicate species approaches its solubility limit.

3)      Diffusion limitations arising from precipitation of secondary phases, namely amorphous silica and quartz on the surface of the dissolving particles [5].

Each of these three phenomena could have significant influence on the dissolution dynamics. The experiments were done in a systematic manner so that the effects of each phenomenon can be observed independent of other phenomena. This allows for accurate descriptions of each phenomenon in the model. The presentation will describe the experimental procedure we have used, the rationale behind our strategy of choosing the experimental conditions and simulations from the updated model to describe the near equilibrium dynamics.

References

[1] O’Connor et al., “Aqueous mineral carbonation: mineral availability, pretreatment, reaction parametrics and process studies”, Albany Research Center DOE/ARC-TR-04-002 (2005).

[2] Werner et al., “Dissolution of dehydroxylated lizardite at flue gas conditions: I. Experimental study”, Chem.Eng. J. (2014), 241, 301-313.

[3] Hariharan et al., “Dissolution of dehydroxylated lizardite at flue gas conditions: II. Kinetic modeling”,

Chem.Eng. J. (2014), 241, 314-326

[4] Pokrovsky et al., “Kinetics and mechanism of forsterite dissolution at 25°C and pH from 1 to 12”, Geochim. Cosmochim. Ac. (2000), 64, 3313–3325

[5] Daval et al., “Influence of amorphous silica layer formation on the dissolution rate of olivine at 90° C and elevated pCO₂”, Chem. Geol. (2011), 284.1, 193-209