Abstract:Direct air capture (DAC) of carbon dioxide is a critical technology for mitigating climate change, but current materials face limitations in efficiency and scalability. We discover novel DAC materials using a combined machine learning (ML) and high-throughput atomistic modeling approach. Our ML model accurately predicts high-quality, density functional theory-computed CO$_{2}$ binding enthalpies for a wide range of nitrogen-bearing moieties. Leveraging this model, we rapidly screen over 1.6 million binding sites from a comprehensive database of theoretically feasible molecules to identify materials with superior CO$_{2}$ binding properties. Additionally, we assess the synthesizability and experimental feasibility of these structures using established ML metrics, discovering nearly 2,500 novel materials suitable for integration into DAC devices. Altogether, our high-fidelity database and ML framework represent a significant advancement in the rational development of scalable, cost-effective carbon dioxide capture technologies, offering a promising pathway to meet key targets in the global initiative to combat climate change.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the efficiency and scalability of materials in Direct Air Capture (DAC) technology. Specifically, the current materials used for DAC have limitations in terms of the ability to adsorb carbon dioxide (CO₂) and stability, and these limitations impede the large - scale application and economic feasibility of DAC technology. The goal of the paper is to discover new amine - functionalized materials by combining machine learning (ML) and high - throughput atomic modeling methods, in order to improve the CO₂ adsorption performance of DAC materials and ensure that these materials are experimentally feasible and synthesizable. This helps to achieve the goal of a CO₂ capture cost per ton below $100, which is a key objective of the U.S. Department of Energy's Carbon Negative Program. ### Main contributions of the paper 1. **Developed a high - precision machine - learning model**: This model can accurately predict the CO₂ binding enthalpy calculated by Density Functional Theory (DFT) and is applicable to a wide range of nitrogen - containing groups. 2. **Large - scale screening of potential materials**: Using this model, the researchers quickly screened more than 1.6 million binding sites and identified materials with excellent CO₂ binding properties from a comprehensive database containing theoretically feasible molecules. 3. **Evaluated the synthetic and experimental feasibility of materials**: By using established machine - learning metrics, the researchers discovered nearly 2,500 new materials that are suitable for integration into DAC devices. 4. **Provided a high - fidelity database and machine - learning framework**: These tools and methods represent a significant advance in the rational design of scalable, cost - effective carbon dioxide capture technologies and provide strong support for achieving global climate change mitigation goals. ### Key problems solved - **Increasing CO₂ adsorption capacity**: The current state - of - the - art amine - based solid adsorbents can usually only reach an adsorption capacity of 2 millimoles of CO₂ per gram, and it decreases significantly after several adsorption/regeneration cycles. The paper aims to discover materials with higher adsorption capacity and stability. - **Reducing capture costs**: By optimizing material properties, reduce the cost of CO₂ capture per ton to make it economically viable. - **Accelerating the discovery of new materials**: Use computational modeling and machine - learning methods to significantly accelerate the discovery speed of new materials, thereby shortening the time from laboratory to practical application. ### Methods and techniques - **High - throughput DFT calculations**: Calculated the CO₂ binding enthalpy of a large number of nitrogen - containing molecules to generate a high - quality training data set. - **Machine - learning model**: Trained a machine - learning model based on DFT calculation results to quickly predict the CO₂ binding performance of a large number of candidate molecules. - **Synthetic feasibility assessment**: Used indicators such as SAscore and GDBscore to assess the synthetic feasibility of candidate molecules to ensure that the discovered materials have practical application value. Through these methods, the paper not only provides new material options for the development of direct air capture technology, but also provides valuable templates and tools for the discovery of future functional materials.

Design of Amine-Functionalized Materials for Direct Air Capture Using Integrated High-Throughput Calculations and Machine Learning

Direct air capture multiscale modelling: From capture material optimization to process simulations

From Lab to Fab: Development and Deployment of Direct Air Capture of CO2

Accelerating CO₂ Direct Air Capture Screening for Metal-Organic Frameworks with a Transferable Machine Learning Force Field

Amine-Functionalized Meso-Macroporous Polymers for Efficient CO 2 Capture from Ambient Air

Making Direct Air Capture Affordable; Technology, Market and Regulatory Approaches

Development of sorbent materials for direct air capture of CO2

Machine Guided Discovery of Novel Carbon Capture Solvents

Thermodynamic Study on Two Adsorption Working Cycles for Direct Air Capture

Research on Current and Future Aspects of Direct Air Capture of CO2

Understandings on Design and Application for Direct Air Capture: from Advanced Sorbents to Thermal Cycles

Direct Air Capture Technology and Its Application

Direct Air Capture (DAC) for Achieving Net-Zero CO2 Emissions: Advances, Applications, and Challenges

Research needs targeting direct air capture of carbon dioxide: Material & process performance characteristics under realistic environmental conditions

Bi-disperse Adsorption Model of the Performance of Amine Adsorbents for Direct Air Capture

Indoor CO2 direct air capture and utilization: Key strategies towards carbon neutrality

Unveiling the Porosity Effect of Superbase Ionic Liquid-Modified Carbon Sorbents in CO 2 Capture from Air

Recent progress on direct air capture of carbon dioxide

Advances in solid adsorbent materials for direct air capture of CO2

Reactive direct air capture of CO 2 to C–C coupled products using multifunctional materials

Synergetic Enhancement of CO2 Direct Air Capture with Monoethanolamine-Impregnated MIL-101(Cr) MOFs