Meta & GeorgiaTech Researchers Release a New Dataset and Associated AI Models to Help Accelerate Research on Direct Air Capture to Combat Climate Change

The global community faces a challenge in tackling the impact of rising carbon dioxide (CO2) levels on climate change. To address this, innovative technologies are being developed. Direct Air Capture (DAC) is a very important approach. DAC involves capturing CO2 directly from the atmosphere, and its implementation is crucial in the fight against climate change. However, the high costs associated with DAC have hindered its widespread adoption.

An important aspect of DAC is its reliance on sorbent materials, and among the various options, Metal-Organic Frameworks (MOFs) have gained attention. MOFs offer advantages such as modularity, flexibility, and tunability. In contrast to conventional absorbent materials that require a lot of energy to be restored, Metal-Organic Frameworks (MOFs) offer a more energy-efficient alternative by allowing regeneration at lower temperatures. This makes MOFs a promising and environmentally friendly choice for various applications.

But, identifying suitable sorbents for DAC is a complex task due to the vast chemical space to explore and the need to understand material behaviour under different humidity and temperature conditions. Humidity, in particular, poses a significant challenge, as it can affect adsorption and lead to sorbent degradation over time. 

In response to this challenge, the OpenDAC project has emerged as a collaborative research effort between Fundamental AI Research (FAIR) at Meta and Georgia Tech. The primary goal of OpenDAC is to significantly reduce the cost of DAC by identifying novel sorbents — materials capable of efficiently pulling CO2 from the air. Discovering such sorbents is key to making DAC economically viable and scalable.

The researchers performed extensive research, resulting in the creation of the OpenDAC 2023 (ODAC23) dataset. This dataset is a compilation of over 38 million density functional theory (DFT) calculations on more than 8,800 MOF materials, encompassing adsorbed CO2 and H2O. ODAC23 is the largest dataset of MOF adsorption calculations at the DFT level, offering valuable insights into the properties and structural relaxation of MOFs.

Also, OpenDAC released the ODAC23 dataset to the broader research community and the emerging DAC industry. The aim is to foster collaboration and provide a foundational resource for developing machine learning (ML) models. 

Researchers can identify MOFs easily by approximating DFT-level calculations using cutting-edge machine-learning models trained on the ODAC23 dataset.

In conclusion, the OpenDAC project represents a significant advancement in improving Direct Air Capture’s (DAC) affordability and accessibility. By leveraging Metal-Organic Frameworks (MOF) strengths and employing cutting-edge computational methods, OpenDAC is well-positioned to drive progress in carbon capture technology. The ODAC23 dataset, now open to the public, marks a contribution to the collective effort to combat climate change, offering a wealth of information beyond DAC applications.

Check out the Paper and Project. All credit for this research goes to the researchers of this project. Also, don’t forget to join our 32k+ ML SubReddit, 40k+ Facebook Community, Discord Channel, and Email Newsletter, where we share the latest AI research news, cool AI projects, and more.

If you like our work, you will love our newsletter..

We are also on Telegram and WhatsApp.