J. De Vos

High-Throughput Screening of Covalent Organic Frameworks for Carbon Capture Using Machine Learning

J. De Vos, S. Ravichandran, S. Borgmans, L. Vanduyfhuys, P. Van der Voort, S.M.J. Rogge, V. Van Speybroeck

Chemistry of Materials

36, 9, 4315-4330

2024

A1

Abstract

Postcombustion carbon capture provides a high-potential pathway to reduce anthropogenic CO₂ emissions in the short term. In this respect, nanoporous materials, such as covalent organic frameworks (COFs), offer a promising platform as adsorbent beds. However, due to the modular nature of COFs, an almost unlimited number of structures can possibly be synthesized. To efficiently identify promising materials and reveal performance trends within the COF material space, we present a computational high-throughput screening of 268,687 COFs for their ability to efficiently and selectively separate CO₂ from the flue gas of power plants using a pressure swing adsorption process. Furthermore, we demonstrate that our screening can be significantly accelerated using machine learning to identify a set of promising materials. These are subsequently characterized with high accuracy, taking into account the effects of competitive adsorption and coadsorption. Our screening reveals that imide, (keto)enamine, and (acyl)hydrazone COFs are particularly interesting for carbon capture. Additionally, the best-performing materials are 3D COFs possessing strong CO₂ adsorption sites between aromatic rings at opposite sides of pores with a diameter of 1.0 nm. In 2D COFs, a significant influence of the framework chemistry, such as imide linkages or fluoro groups, is observed. Our design rules can guide experimental researchers to construct high-performing COFs for CO₂ capture.

Gold Open Access

DOI

http://dx.doi.org/10.1021/acs.chemmater.3c03230

Engineering of Phenylpyridine- and Bipyridine-Based Covalent Organic Frameworks for Photocatalytic Tandem Aerobic Oxidation/Povarov Cyclization

M. Debruyne, S. Borgmans, S. Radhakrishnan, E. Breynaert, H. Vrielinck, K. Leus, A. Laemont, J. De Vos, K. S. Rawat, S. Vanlommel, H. Rijckaert, H. Salemi, J. Everaert, F. Vanden Bussche, D. Poelman, R. Morent, N. De Geyter, P. Van der Voort, V. Van Speybroeck, C.V. Stevens

ACS Applied Materials & Interfaces

15, 29, 35092–35106

2023

A1

Abstract

Covalent organic frameworks (COFs) are emerging as a new class of photoactive organic semiconductors, which possess crystalline ordered structures and high surface areas. COFs can be tailor-made toward specific (photocatalytic) applications, and the size and position of their band gaps can be tuned by the choice of building blocks and linkages. However, many types of building blocks are still unexplored as photocatalytic moieties and the scope of reactions photocatalyzed by COFs remains quite limited. In this work, we report the synthesis and application of two bipyridine- or phenylpyridine-based COFs: TpBpyCOF and TpPpyCOF. Due to their good photocatalytic properties, both materials were applied as metal-free photocatalysts for the tandem aerobic oxidation/Povarov cyclization and α-oxidation of N-aryl glycine derivatives, with the bipyridine-based TpBpyCOF exhibiting the highest activity. By expanding the range of reactions that can be photocatalyzed by COFs, this work paves the way toward the more widespread application of COFs as metal-free heterogeneous photocatalysts as a convenient alternative for commonly used homogeneous (metal-based) photocatalysts.

Open Access version available at UGent repository

DOI

http://dx.doi.org/

ReDD-COFFEE: A ready-to-use database of covalent organic framework structures and accurate force fields to enable high-throughput screenings

J. De Vos, S. Borgmans, P. Van der Voort, S.M.J. Rogge, V. Van Speybroeck

J. Mater. Chem. A

11, 14, 7468-7487

2023

A1

Abstract

Covalent organic frameworks (COFs) are a versatile class of building block materials with outstanding properties thanks to their strong covalent bonds and low density. Given the sheer number of hypothetical COFs envisioned via reticular synthesis, only a fraction of all COFs have been synthesized so far. Computational high-throughput screenings offer a valuable alternative to speed-up such materials discovery. Yet, such screenings vitally depend on the availability of diverse databases and accurate interatomic potentials to efficiently predict each hypothetical COF’s macroscopic behavior, which is currently lacking. Therefore, we herein present ReDD-COFFEE, the Ready-to-use and Diverse Database of Covalent Organic Frameworks with Force field based Energy Evaluation, containing 268 687 COFs and accompanying ab initio derived force fields that are shown to outperform generic ones. Our structure assembly approach results in a huge amount of computer-ready structures with a high diversity in terms of geometric properties, linker cores, and linkage types. Furthermore, the textural properties of the database are analyzed and the most promising COFs for vehicular methane storage are identified. By making the database freely accessible, we hope it may also inspire others to further explore the potential of these intriguing functional materials.

Gold Open Access

DOI

http://dx.doi.org/10.1039/D3TA00470H

MicMec: Developing the Micromechanical Model to Investigate the Mechanics of Correlated Node Defects in UiO-66

J. Vandewalle, J. De Vos, S.M.J. Rogge

J. Phys. Chem. C

127, 12, 6060-6070

2023

A1

Abstract

New functional materials, such as mixed matrix membranes and metal–organic framework (MOF) monoliths, outperform traditional materials in gas separation and storage applications, among other pressing challenges. However, while most engineered materials nowadays exhibit spatial heterogeneities on different length scales, available simulation techniques to date cannot capture this spatial complexity fully. Herein, we present the MicMec implementation of the micromechanical model we introduced earlier as a systematic coarse-graining approach to routinely access these larger length scales and characterize the mechanical properties of these materials. We thoroughly discuss the key components of our open-source code and validate it both on an analytical system and a case study on correlated reo defects in the UiO-66 MOF. We reveal that the time step that can be reached in micromechanical simulations is 2 to 3 orders of magnitude larger than for atomistic simulations, while still capturing well the macroscopic mechanical properties of the spatially disordered UiO-66 material. It is our hope that the MicMec implementation discussed here may provide a complementary tool to existing atomistic and coarse-grained software and aid the computational design of new materials for pressing applications.

Gold Open Access

DOI

http://dx.doi.org/10.1021/acs.jpcc.3c00451

Exploring the phase stability in interpenetrated diamondoid covalent organic frameworks

S. Borgmans, S.M.J. Rogge, J. De Vos, P. Van der Voort, V. Van Speybroeck

Communications Chemistry

6, 1, 5

2023

A1

Abstract

Soft porous crystals, which are responsive to external stimuli such as temperature, pressure, or gas adsorption, are being extensively investigated for various technological applications. However, while substantial research has been devoted to stimuli-responsive metal-organic frameworks, structural flexibility in 3D covalent organic frameworks (COFs) remains ill-understood, and is almost exclusively found in COFs exhibiting the diamondoid (dia) topology. Herein, we systemically investigate how the structural decoration of these 3D dia COFs—their specific building blocks and degree of interpenetration—as well as external triggers such as temperature and guest adsorption may promote or suppress their phase transformations, as captured by a collection of 2D free energy landscapes. Together, these provide a comprehensive understanding of the necessary conditions to design flexible diamondoid COFs. This study reveals how their flexibility originates from the balance between steric hindrance and dispersive interactions of the structural decoration, thereby providing insight into how new flexible 3D COFs can be designed.

Open Access version available at UGent repository

Gold Open Access

DOI

http://dx.doi.org/10.1038/s42004-022-00808-y

Quantifying the likelihood of structural models through a dynamically enhanced powder X‐ray diffraction protocol

S. Borgmans, S.M.J. Rogge, J. De Vos, C.V. Stevens, P. Van der Voort, V. Van Speybroeck

Angewandte Chemie int. Ed.

60 (16), 8913-8922

2021

A1

Abstract

Structurally characterizing new materials is tremendously challenging, especially when single crystal structures are hardly available which is often the case for covalent organic frameworks. Yet, knowledge of the atomic structure is key to establish structure‐function relations and enable functional material design. Herein a new protocol is proposed to unambiguously predict the structure of poorly crystalline materials through a likelihood ordering based on the X‐ray diffraction (XRD) pattern. Key of the procedure is the broad set of structures generated from a limited number of building blocks and topologies, which is submitted to operando structural characterization. The dynamic averaging in the latter accounts for the operando conditions and inherent temporal character of experimental measurements, yielding unparalleled agreement with experimental powder XRD patterns. The proposed concept can hence unquestionably identify the structure of experimentally synthesized materials, a crucial step to design next generation functional materials.

Gold Open Access