Multi-scale engineering toolbox for systematic assessment of porous materials in the context of adsorption and membrane separations

Lead Research Organisation: University of Edinburgh
Department Name: Sch of Engineering

Abstract

We will integrate structure characterization, molecular simulation and process modelling methods into a single computational toolbox and apply this toolbox to explore the scope and accuracy of multi-scale approaches in the assessment of performance of porous materials in adsorption and membrane separation processes. Separation processes consume about 10-15% of global energy, while high energy cost of carbon capture still presents a major hurdle in the implementation of this technology. Recent discovery of new families of porous materials opens unprecedented opportunities to advance energy efficient adsorption and membrane separations; however the large number of new materials demands a transition from traditional trial-and-error process design to rational selection of materials based on computational screening. In this project, we develop computational tools required for this transition, test them against bench scale experiments, and explore their robustness in screening materials for realistic process configurations. In the latter case, we use portable oxygen concentration technologies as a source of extensive reference data to test computational predictions. At the same time, we use this case as an opportunity to apply multi-scale approaches to explore further optimization of portable oxygen concentrators (POC) to make these medical devices even lighter with longer battery life.

Planned Impact

In this project we will develop a multi-scale software toolbox, aimed at the systematized and streamlined design of energy efficient adsorption and membrane separation processes, capitalizing on the recent developments in material engineering, molecular simulations and process optimization.

1) This toolbox will probe whether accurate ranking of porous materials for adsorption and membrane separations can be obtained within a fully self-contained computational framework.
2) It will open opportunities for design of adsorption and membrane separation processes based on simultaneous material and process optimization, shifting from the traditional trial-and-error paradigm.
3) Within the toolbox novel combinations of adsorption and membrane separations can be explored.
4) It will be applied to optimize Portable Oxygen Concentration (POC) devices for health, sports and military applications.

No simulation suite of this type exists at the moment. Currently available software, commercial or public, specializes in either the molecular simulation level of description (Accelrys, Materials Design) or process design and optimization (AdSim from Aspen Tech, gPROMS from PSE). Industrial companies working in the field of adsorption and membrane technologies are identified as the immediate beneficiaries of the project. The proposed toolbox alleviates the need for the companies to develop their own software, which would be time consuming and unaffordable for smaller companies. This will have a direct impact on the competitiveness of the UK in large scale industrial separations and other technologies. Specifically, research in carbon capture (within a larger theme of energy efficiency) and materials science have been a UK strength, however process optimization through material engineering and technology commercialization have been lagging behind in the UK compared to other countries such as the USA, where the "Materials Genome" project has become a frontline multi-agency initiative, championed by the White House. Having a technology developed based on PIMs would be particularly symbolic and impactful for the UK competitiveness in this area, as these materials have been discovered in the UK.

On the academic side, the most immediate beneficiaries of this project will be eight academic groups within the School of Engineering (UoE), including those of applicants, working in the fields of adsorption, membranes, molecular simulation and process optimization. These groups will receive at their disposal a powerful tool which will expand the type of problems and analysis they can tackle, leading to new research outcomes, ideas and collaborations. In addition to the group of Prof. Kaskel at TUD, several other academic groups expressed their interest in using the software, including Profs. Russell Morris (St. Andrews), Andy Copper (Liverpool), Neil McKeown (Edinburgh) and six of the current users of the Poreblazer software. The groups mentioned above specialise in material synthesis and characterization; for these groups it is important to explore properties of newly discovered materials in real applications and the proposed toolbox will allow them to do that, without developing their own codes.

Another group of beneficiaries of this project will be companies working on the development of the POC devices, medical and health services, companies and services associated with sports medicine and health, various sectors within the defence industry, and patients with different types of health conditions, such as the Chronic Obstructive Pulmonary Disease. They will benefit from this project if it produces new, optimized processes for portable oxygen concentration, based on novel materials.
 
Description Carbon dioxide capture from industrial streams, such as power plant flue gas, is an important strategy in control of green-house gas emissions and in climate change mitigation. Pressure-swing adsorption (PSA) technologies have been considered as an energy efficient alternative to the conventional amine-based absorption processes. At the heart of the PSA process is adsorbate material. It has been proposed that promising materials for carbon capture must exist among tens of thousands of already synthesized and potentially millions of hypothetical MOFs and ZIFs. Computational screening strategies have been employed to identify these materials, with more recent studies based on realistic process simulations. It is now recognized that the optimal performance of a material, and hence its ranking, is not a function of its intrinsic adsorptive properties only, but also depends on other variables such as how the material is structured within the adsorption column and characteristics of the cycle. In this project we combined molecular simulations with advanced process modelling to explore performance of materials in real dynamic PSA cycles. Overall, our findings can be summarized as follows:

1 Equilibrium adsorption characteristics are not on their own sufficient to predict the behaviour of the material in a real process simulation. This behaviour also depends on other factors such as pellet size, configuration of the size etc which can be optimized for each individual material specifically.
2 We discovered several characteristics and metrics of the optimal process configurations that are material-independent and can be used to expedite the computational screening methods in a search for the most optimal materials for carbon capture and other applications.

These key findings have been presented in several articles associated with the project. The most recent article has been published in a prestigious journal of Energy and Environmental Science.

1. Automated analysis and benchmarking of GCMC simulation programs in application to gas adsorption
RJ Gowers, AH Farmahini, D Friedrich, L Sarkisov Molecular Simulation 44 (4), 309-321

2. From crystal to adsorption column: challenges in multiscale computational screening of materials for adsorption separation processes
AH Farmahini, S Krishnamurthy, D Friedrich, S Brandani, L Sarkisov Industrial & Engineering Chemistry Research 57 (45), 15491-15511

3. Exploring New Sources of Efficiency in Process-Driven Materials Screening for Post-combustion Carbon Capture
AH Farmahini, D Friedrich, S Brandani, L Sarkisov Energy & Environmental Science

There are several other articles currently in the pipeline (including a comprehensive review) which will make the impact of the project even stronger.
Exploitation Route At this stage the most significant impact of the project is a set of results published in the most recent article "Exploring New Sources of Efficiency in Process-Driven Materials Screening for Post-combustion Carbon Capture", AH Farmahini, D Friedrich, S Brandani, L Sarkisov Energy & Environmental Science. There, we identify some general correlations that should expedite the screening algorithms for adsorbent materials in the context of carbon capture and other application. This, in principle, should facilitate the implementation of adsorption-based carbon capture technologies.
Sectors Chemicals,Energy,Environment
 
Description The current thread of research on carbon capture using computational approaches, materials screening and new algorithms has been drawing a lot of scientific and general public attention. The results of this work have been presented at many conferences (AIChE 2018, Fundamentals of Adsorption). Moreover, the PI on the project, Prof. Sarkisov, has been invited to give a lecture on this topic to a broad audience at the Global Innovation Forum in Yerevan, Armenia in 2019 (https://fast.foundation/globalinnovationforum/2019/home). Overall, the impact of this engagement is a higher awareness of the general public on the current research trends in fighting global climate change and carbon emissions.
First Year Of Impact 2020
Sector Chemicals,Education,Other
Impact Types Societal,Policy & public services
 
Title Case studies for computational performance of the Monte Carlo codes 
Description As a first step of the project we develop a database of case studies to probe the computational performance of five publicly available Monte Carlo codes. Specifically, we concentrated on the case of carbon dioxide adsorption in a metal organic framework, IRMOF-1. Five codes have been obtained (RASPA, MuSiC, Towhee, Cassandra, DL_MONTE) and their performance and accuracy compared in application system above. Firstly it was important to establish that all codes are consistent with each other and under what conditions. Secondly, it was important to compare their performance. Currently we are working on associated publication and the database will be released to the general public along with the publicationm 
Type Of Material Computer model/algorithm 
Provided To Others? No  
Impact Material screening for adsorption applications is becoming a very important area of research particularly in application to carbon capture and to methane storage. This requires in general computationally efficient and accurate tools, specifically Monte Carlo methods traditionally used for adsorption problems. In molecular dynamics community, the scale of the community and the interest in large biological system, has been driving the development of benchmarks and case studies which would test the performance and accuracy of the md codes. This is not the case for Monte Carlo community, where systematic comparison of the codes in terms of accuracy and performance is lacking. This database of case studies is the first example of benchmarking the Monte Carlo codes. It will allow the researchers to identify the best code for their applications, to build their own case studies and simulations set up, and it will help to validate the newly developed codes, by providing accurate reference data. 
 
Description A new collaboration with IBM established 
Organisation IBM Research - Brazil
Country Brazil 
Sector Private 
PI Contribution Based on the progress in the initial project funded by EPSRC, we have established a collaboration with IBM in the area of accelerated material discovery using multiscale workflows. To this partnership, my group and in particular PhD student Mr Conor Cleeton bring expertise in molecular and process simulations, where as partners at IBM bring on board expertise in Machine Learning to accelerate the workflows. In particular, this collaboration has been exploring efficient methods to obtain models of partial charges in Metal-Organic Frameworks (MOFs), generation of large-set of reference adsorption data for carbon capture in MOFs, novel approaches to material informatics and metrics for adsorption applications, and uncertainty propagation in multiscale workflows. This has resulted in the Charge-dependent, Reproducible, Accessible, Forcefield-dependent, and Temperature-dependent Exploratory Database (CRAFTED) of adsorption isotherms, recently published online. This dataset contains the simulation of CO2 and N2 adsorption isotherms on 726 metal-organic frameworks taken from the CoRE MOF 2014 database (total of 34,848 adsorption isotherms at different conditions). In the context of the current proposal, the collaborators from IBM will help the project to generate the necessary adsorption data for the case studies and provide their advice and expertise on the development of advanced features of porous materials to be used in conjunction with the ML methods for the prediction of adsorption data.
Collaborator Contribution The IBM partners, in particular the Industrial Technology and Science group at IBM, led by Dr. Mathias Steiner, has made significant contribution to the research by providing the computational facilities, running large scale simulations on their infrastructure and offering expert help with the ML models.
Impact The multidisciplinary collaboration involves molecular simulation, machine learning, material science, process simulation and computer science communities. So far the collaboration resulted into two conference contribution and one article under review. 1. Articles under review CRAFTED -- An exploratory database of simulated adsorption isotherms of metal-organic frameworks Felipe Lopes Oliveira, Conor Cleeton, Rodrigo Neumann Barros Ferreira, Binquan Luan, Amir H. Farmahini, Lev Sarkisov, Mathias Steiner https://arxiv.org/abs/2210.09456 2. Conference contributions The role of partial charges in data-driven analytics of gas adsorption simulations B Luan, R Ferreira, R Giro, T Elengikal, A Gupta, F De Oliveira, S Liu, ... Bulletin of the American Physical Society Automated AI-driven screening of nanoporous materials for enhanced carbon dioxide adsorption B O'Conchuir, RNB Ferreira, F Martelli, B Luan, R Giro, T Elengikal, ... Materials Research Society (MRS) Fall Meeting
Start Year 2020