Large scale in silico screening of materials for carbon capture through chemical looping

Abstract

With the advent of large structural databases containing both optimised crystallographic structures and their ground state energies, the goal of rationally designing novel functional materials for a variety of applications can be realised. Through selection of relevant, property-specific parameter(s), screening criteria can then be applied to thousands of potential candidates in silico, efficiently selecting the most promising materials for subsequent experimental testing. Here we describe our work developing screening methodologies for the design of novel materials for carbon capture and storage (CCS) and ionic conductivity. Our holistic approach combines theoretical screening with experimental validation to better understand the trends underpinning the performance of the selected materials. We have made use of the Materials Project database (www.materialsproject.org), which not only contains an extensive variety of calculated structures, but is also constructed in such a way to be amenable to high throughput screening. The first application we consider is CO2 absorption looping for CCS, which requires oxide materials that are able to reversibly absorb CO2 at high temperatures. CaO is the prototypical material for this application, but unfortunately suffers irreversible capacity loss and sintering upon continuous carbonation-regeneration cycles, necessitating materials with improved performance. With large scale screening, we were able to simulate the carbonation equilibria for 640 prospective sorbents and then select a number of candidates based on (i) minimising the energy cost associated with their use and (ii) maximising their theoretical CO2 capture capacity [1]. The accuracy of the screening was validated using structural and thermogravimetric analysis, and the process led to a number of design rules for optimising materials performance, including focussing on ternary oxides and materials containing magnesium and calcium. The second part of this work concerns another CCS technology, chemical looping combustion, which utilises materials that can spontaneously release gaseous O2 in order to efficiently burn fuel in a N2-free environment and create pure CO2 for subsequent capture. This is in contrast to the former process, which separates CO2 post-combustion. Binary metal oxides with multiple oxidation states such as CuO and MnO2 can be reduced under reactor conditions, but materials that reduce at lower temperatures and cycle more stably are desirable in order to minimise energy costs involved with combustion at higher temperatures. Our screening found over 2200 materials that were able to undergo redox reactions under the specific process conditions [2], with further experimental studies revealing a number of promising materials that could be stably cycled between their perovskite and brownmillerite phases. The final part details the use of local structural similarity algorithms to efficiently screen materials for oxygen ionic conductivity. Building on a previous methodology using Voronoi tessellations to ...

Local Structure Order Parameters and Site Fingerprints for Quantification of Coordination Environment and Crystal Structure Similarity

“…The features that are computed with the CE ngerprint are the continuous symmetry measures 101 (CSM) between a given motif and all available ideal coordination environments supported by ChemEnv. 100 In particular, the considered environments are: (1) single neighbor (S:1), dodecahedron with triangular faces (DD:8), (26) dodecahedron with triangular faces-p2345 plane normalized (DDPN:8), (27) hexagonal bipyramid (HB:8), (28) bicapped octahedron (opposed cap faces) (BO_1:8), (29) bicapped octahedron (cap faces with one atom in common) (BO_2:8), (30) bicapped octahedron (cap faces with one edge in common) (BO_3:8), (31) triangular cupola (TC:9), (32) tricapped triangular prism (three square-face caps) (TT_1:9), (33) tricapped triangular prism (two square-face caps and one triangular-face cap) (TT_2:9), (34) tricapped triangular prism (one square-face cap and two triangular-face caps) (TT_3:9), (35) heptagonal dipyramid (HD:9), (36) tridiminished icosahedron (TI:9), (37) square-face monocapped antiprism (SMA:9), (38) square-face capped square prism (SS:9), (39)…”

“…16 As computational resources still continue to grow 18 and to become more omnipresent and accessible, the computational chemistry, physics, and materials science communities have focused their efforts more and more on automation tools for materials database analysis and on employing statistical and machine learning (SML) 19,20 to help expedite materials discoveries and chemical innovations. 21 This includes, for example, predicting properties (e.g., formation energies, crystal structure dimensionalities, phase diagrams, band gaps, elastic moduli, ionic conductivity) of diverse materials from classes and families such as AX binary compounds, 22 M 2 AX ternary phases, 23 delafossite and related layered phases of composition ABX 2 , 24 conventional 25 and double perovskite halides (or elpasolites), 26 zeolites 27,28 and other silicates, 29 and other inorganic materials [30][31][32][33][34][35][36] as well as polymers; 37 indicating possible synthesis approaches by screening and predicting synthesis parameters and reactions of inorganic materials, 38,39 metal-organic frameworks, 40 and organic molecules; 41,42 generating interatomic potentials; [43][44][45][46] and expediting ab initio [47][48][49][50] calculations. [51][52][53][54][55] Another important scientic problem is, in this...…”

A‐ and B‐site Codoped SrFeO3 Oxygen Sorbents for Enhanced Chemical Looping Air Separation

“…However, the oxygen capacity reduces to approximately 0.6 wt % when pressure swing is operated between 1 % and 20 % O 2 . Recently, SrFeO 3 has been identified as a promising oxygen storage material for air separation . The oxygen capacity has been demonstrated to be roughly 2.2 wt % at 550 °C by redox cycling between nitrogen and air .…”

Sr1-xCaxFe1-yCoyO3-δ as facile and tunable oxygen sorbents for chemical looping air separation

“…Perovskite based oxygen sorbents have the advantage of reversibly absorb and desorb oxygen at significantly lower temperatures (e.g. ≤ 600 • C), owing to their structural flexibility to accommodate significant amount of oxygen vacancies [30][31][32][33][34][35][36]. For example, La 0.1 Sr 0.9 Co 0.9 Fe 0.1 O 3-δ has a wide range of oxygen vacancy depending on temperature and oxygen partial pressure [37,38].…”