Data-driven Screening and Discovery of Metal- organic Frameworks as C2 Adsorbents from Over 900 Experimental Isotherms

Heather Kulik,Matthew Rivera,Gianmarco Terrones,Tae Hoon Lee,Zachary Smith

DOI: https://doi.org/10.26434/chemrxiv-2024-lfm84

2024-08-08

Abstract:The separation of ethylene from ethane accounts for almost 100 million tons of CO2 emissions annually and 0.3% of global primary energy usage. Replacing current cryogenic distillation units with adsorption separation units, especially for the minor component of ethane, would enable significant efficiency gains. Metal-organic frameworks (MOFs) are well-suited for adsorption separation due to their high surface areas and tunable chemical properties; exploring all possible MOFs is a daunting experimental challenge, motivating in silico screening with machine learning models. We present a database of 948 experimentally measured pure-component C2 isotherms from 192 MOFs gathered from the literature and use it to train machine learning models to predict MOF ethane and ethylene uptake across a range of temperature and pressure conditions. The models have high accuracy in interpolative tasks (mean absolute error ~0.05 mmol/g) when trained on only 20% of available data. Performance on unseen structures was also reasonably accurate with MAEs ~0.7 mmol/g. We apply the models to screen the CoRE ASR 2019 database and identify the most promising candidates. Several MOFs containing lanthanide metals were predicted to have high ethane selectivity, suggesting that this class of MOFs may be worth further investigation. Feature importance analysis suggests that MOF SBU design and optimizing the process conditions at which the sorbent will operate are critical for enabling ethane-selective separation. We synthesize a MOF predicted to exhibit high ethane selectivity and experimentally validate qualitative agreement of model predictions, highlighting the utility of both the dataset and model in discovering unexplored C2 adsorbents.

Chemistry

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to screen and discover metal - organic frameworks (MOFs) suitable for separating ethylene (C2H4) and ethane (C2H6) through data - driven methods. Specifically, the paper aims to:

1. **Reduce energy consumption and carbon emissions**: The current low - temperature distillation technology used to separate ethylene and ethane generates nearly 100 million tons of carbon dioxide emissions annually and accounts for 0.3% of global primary energy use. Using adsorption separation technology to replace the existing low - temperature distillation units can significantly improve efficiency and reduce energy consumption.

2. **Explore new MOF materials**: MOFs are very suitable for adsorption separation because of their high specific surface area and tunable chemical properties. However, experimentally exploring all possible MOFs is a huge challenge. Therefore, researchers use machine - learning models to predict the adsorption performance of MOFs for ethane and ethylene, thereby accelerating the screening of new MOF materials.

3. **Establish a high - quality data set**: The authors collected 948 experimentally measured pure - component C2 isotherms from 192 MOFs and used these data to train machine - learning models to predict the MOF adsorption amounts under different temperature and pressure conditions.

4. **Develop an efficient prediction model**: By training machine - learning models, the authors were able to accurately predict the ethane and ethylene adsorption amounts of MOFs in interpolation tasks (with an average absolute error of approximately 0.05 mmol/g) and also showed reasonable accuracy on unseen structures (MAE of approximately 0.7 mmol/g).

5. **Identify promising MOF candidate materials**: These models were applied to screen the CoRE ASR 2019 database to identify the most promising MOF candidate materials. The study found that MOFs containing lanthanide metals showed high ethane selectivity, indicating that this type of MOFs is worthy of further study.

6. **Verify model predictions**: A MOF predicted to have high ethane selectivity was synthesized, and the qualitative consistency of the model predictions was verified experimentally, demonstrating the practicality of the data set and the model in discovering new C2 adsorbents.

Through these efforts, this research provides an important theoretical basis and experimental evidence for the development of more efficient and environmentally - friendly ethylene/ethane separation technologies.