The Kinetics of CO2 Mineralization Processes Using Thermally Activated Serpentine

Abstract

A fully functional Carbon Capture and Storage scheme has to ensure a safe and permanent storage of the captured CO2. Ex situ CO2 mineralization would enable this by storing CO2 in a chemically stable carbonate form. Magnesium and calcium silicate rocks found in nature are the main sources of the metal cations needed for the mineralization process. The large abundance of the serpentine group of magnesium silicate minerals makes it an attractive feedstock for the mineralization process. In an aqueous mineralization process, the magnesium ions are leached from serpentine under a CO2 gas atmosphere and then precipitated as magnesium carbonates. The rate limiting step for this process is the leaching of magnesium ions, and high reaction temperatures and CO2 pressures are typically needed in a process that uses natural serpentine. The partial dehydroxylation of serpentine through thermal activation has been shown to produce a more reactive silicate material that can form magnesium carbonates at low temperatures and CO2 pressures. However, the kinetics of this entire process is not known, and the dynamics of the mineralization processes at low temperatures and CO2 pressures is complex and non-trivial. The aim of this thesis is to evaluate the kinetics of the individual steps involved in this mineralization process, and demonstrate that these kinetic models can predict the mineralization performances of partially dehydroxylated lizardite (a polymorph of serpentine). In a follow-up work, the models can be used to determine the optimal operating conditions for a mineralization process using partially dehydroxylated lizardite. A non-steady state kinetic model was first developed that describes the dissolution kinetics of partially dehydroxylated lizardite at far-from-equilibrium operating conditions. Based on two simplified dehydroxylation pathways that represent the extreme cases, namely of a fully homogeneous dehydroxylation and a heterogeneous dehydroxylation, two different particle structures for the dehydroxylated material were considered. Surface complexation mechanisms were used to describe the specific dissolution rates of the different silicate species present in the particles and model parameters were estimated by fitting experimental profiles that were measured in an earlier study. The kinetic model was then updated in order to include the description of dissolution at near-equilibrium solution compositions. Batch dissolution experiments for partially dehydroxylated lizardite were carefully designed to explore a wide range of operating conditions without precipitating any secondary phases, namely amorphous silica or magnesium carbonates. The operational range included solutions whose compositions were very close to solubility of the dissolving silicate species. Expressions for specific dissolution rates were updated in order to describe the experimentally measured near-equilibrium, non-steady state dissolution profiles. With regard to the precipitation of magnesium carbonates, the seeded growth kinetics of synthetic hydromagnesite was investigated at 90°C and under low fugacities of CO2 in a semi-batch reactor. The kinetics of CO2 desorption from the solution was measured independently and was found to affect the precipitation rates of hydromagnesite. A kinetic model together with the population balance equation was developed in order to describe CO2 desorption and hydromagnesite growth processes that occur simultaneously in the reactor. The solubility of synthetic hydromagnesite was experimentally measured and was found to be higher than that of crystalline hydromagnesite that is formed in natural environments. Surface nucleation mechanism was found to control the growth rate of synthetic hydromagnesite at 90°C and at low supersaturations. Finally the passivation effect of precipitated amorphous silica on the dissolution of partially dehydroxylated lizardite was measured and an empirical description for this effect was included in the kinetic model. The precipitation of amorphous silica on the dissolving silicate particles was found to inhibit the Mg release rates, and the inhibitory effect was stronger at 60°C than at 30°C. The presence of NaCl or NaHCO3 did not reduce the passivation behavior of amorphous silica or improve CO2 mineralization performances at these operating conditions. The validity of the various kinetic models that were developed in this thesis was tested through CO2 mineralization experiments. The model employing the combined kinetics of the dissolution of partially dehydroxylated lizardite particles, the precipitation of hydromagnesite, and the CO2 mass transfer rates was able to accurately predict the dynamics of single- and two-step CO2 mineralization processes. --> A fully functional Carbon Capture and Storage scheme has to ensure a safe and permanent storage of the captured CO2. Ex situ CO2 mineralization would enable this by storing CO2 in a chemically stable carbonate form. Magnesium and calcium silicate rocks found in nature are the main sources of the metal cations needed for the mineralization process. The large abundance of the serpentine group of magnesium silicate minerals makes it an attractive feedstock for the mineralization process. In an aqueous mineralization process, the magnesium ions are leached from serpentine under a CO2 gas atmosphere and then precipitated as magnesium carbonates. The rate limiting step for this process is the leaching of magnesium ions, and high reaction temperatures and CO2 pressures are typically needed in a process that uses natural serpentine. The partial dehydroxylation of serpentine through thermal activation has been shown to produce a more reactive silicate material that can form magnesium carbonates at low temperatures and CO2 pressures. However, the kinetics of this entire process is not known, and the dynamics of the mineralization processes at low temperatures and CO2 pressures is complex and non-trivial. The aim of this thesis is to evaluate the kinetics of the individual steps involved in this mineralization process, and demonstrate that these kinetic models can predict the mineralization performances of partially dehydroxylated lizardite (a polymorph of serpentine). In a follow-up work, the models can be used to determine the optimal operating conditions for a mineralization process using partially dehydroxylated lizardite.A non-steady state kinetic model was first developed that describes the dissolution kinetics of partially dehydroxylated lizardite at far-from-equilibrium operating conditions. Based on two simplified dehydroxylation pathways that represent the extreme cases, namely of a fully homogeneous dehydroxylation and a heterogeneous dehydroxylation, two different particle structures for the dehydroxylated material were considered. Surface complexation mechanisms were used to describe the specific dissolution rates of the different silicate species present in the particles and model parameters were estimated by fitting experimental profiles that were measured in an earlier study.The kinetic model was then updated in order to include the description of dissolution at near-equilibrium solution compositions. Batch dissolution experiments for partially dehydroxylated lizardite were carefully designed to explore a wide range of operating conditions without precipitating any secondary phases, namely amorphous silica or magnesium carbonates. The operational range included solutions whose compositions were very close to solubility of the dissolving silicate species. Expressions for specific dissolution rates were updated in order to describe the experimentally measured near-equilibrium, non-steady state dissolution profiles. With regard to the precipitation of magnesium carbonates, the seeded growth kinetics of synthetic hydromagnesite was investigated at 90°C and under low fugacities of CO2 in a semi-batch reactor. The kinetics of CO2 desorption from the solution was measured independently and was found to affect the precipitation rates of hydromagnesite. A kinetic model together with the population balance equation was developed in order to describe CO2 desorption and hydromagnesite growth processes that occur simultaneously in the reactor. The solubility of synthetic hydromagnesite was experimentally measured and was found to be higher than that of crystalline hydromagnesite that is formed in natural environments. Surface nucleation mechanism was found to control the growth rate of synthetic hydromagnesite at 90°C and at low supersaturations.Finally the passivation effect of precipitated amorphous silica on the dissolution of partially dehydroxylated lizardite was measured and an empirical description for this effect was included in the kinetic model. The precipitation of amorphous silica on the dissolving silicate particles was found to inhibit the Mg release rates, and the inhibitory effect was stronger at 60°C than at 30°C. The presence of NaCl or NaHCO3 did not reduce the passivation behavior of amorphous silica or improve CO2 mineralization performances at these operating conditions. The validity of the various kinetic models that were developed in this thesis was tested through CO2 mineralization experiments. The model employing the combined kinetics of the dissolution of partially dehydroxylated lizardite particles, the precipitation of hydromagnesite, and the CO2 mass transfer rates was able to accurately predict the dynamics of single- and two-step CO2 mineralization processes.