A Refined Set of Universal Force Field Parameters for Some Metal Nodes in Metal-Organic Frameworks

Yutao Li,Xin Jin,Elias Moubarak,Berend Smit

DOI: https://doi.org/10.26434/chemrxiv-2024-btm9c

2024-08-23

Abstract:Metal-organic frameworks (MOFs) exhibit promise as porous materials for carbon capture due to their design versatility and large pore sizes. The generic force field (e.g. UFF and Dreiding) uses one set of Lennard-Jones parameters for each element, while MOFs have a much richer local chemical environment than those used to fit the UFF. When MOFs contain hard-Lewis acid metals, UFF systematically overestimates $CO2$ uptakes within MOFs. To address this, we developed a workflow to affordably and efficiently generate reliable force fields to predict \ce{CO2} adoption isotherms of MOFs containing metals from groups IIA (e.g. Mg, Ca, Sr, Ba) and IIIA (e.g., Al, Ga, In), connected to various carboxylate ligands. This method uses experimental isotherms as input. The optimal parameters are obtained by minimizing the loss function of the experimental and simulated isotherms, in which we use the Multistate Bennett Acceptance Ratio (MBAR) theory can be used to derive the functionality relationship of loss functions in terms of force field parameters.

Chemistry

What problem does this paper attempt to address?

This paper attempts to solve the following problems:

1. **The applicability of general force fields in metal - organic frameworks (MOFs)**
The paper points out that general force fields (such as UFF and Dreiding) have limitations when describing MOFs. These force fields use only one set of Lennard - Jones parameters for each element, while MOFs have more complex local chemical environments. For example, when MOFs contain hard Lewis - acid metals (such as Al, Mg, etc.), UFF will systematically overestimate the adsorption amount of CO₂.

2. **Improve force - field parameters to enhance prediction accuracy**
In order to improve the prediction accuracy of CO₂ adsorption properties in MOFs, the author developed a workflow for generating reliable force - field parameters. This method uses experimental adsorption isotherms as input and optimizes the force - field parameters by minimizing the loss function between the experimental and simulated isotherms.

3. **The transferability of force - field parameters**
The author not only focuses on the force - field optimization of specific MOFs but also studies whether the optimized force - field can be transferred to other similar MOFs. For example, whether the force - field optimized for MOFs containing Al³⁺ is applicable to MOFs of other Group III metals (such as In³⁺, Ga³⁺) or Group IIA metals (such as Mg²⁺, Ca²⁺).

---

### Specific solutions

1. **Analyze the deficiencies of the UFF force field**
- An in - depth analysis was carried out on the overestimation of the CO₂ adsorption isotherm of Al - MOFs by UFF, and it was found that similar problems may exist in MOFs with other metal nodes.
- It was proposed that the force - field parameters need to be refitted, especially those parts related to metal nodes (such as Al, In, Mg, etc.) and their coordination environments.

2. **Develop a workflow for optimizing the force field**
- Use Grand Canonical Monte Carlo (GCMC) simulations to generate adsorption isotherms.
- Use Multistate Bennett Acceptance Ratio (MBAR) theory to derive the differentiable relationship between the adsorption amount and force - field parameters.
- Optimize the force - field parameters by the gradient - descent method to make them consistent with the experimental isotherms in the low - pressure range.
- Verify the performance of the optimized force - field in binding energy and complete isotherms.

3. **Optimize specific force - field parameters**
- For Al - MOFs, the Lennard - Jones parameters (ϵO and ϵC) of O and C were optimized so that the new force - field can accurately reproduce the experimental data.
- For other metals (such as In³⁺, Ga³⁺, Mg²⁺, etc.), the transferability of the optimized force - field was verified.

4. **Consider the ion - polarization effect**
- Analyzed the polarization effect caused by the unconsidered metal - oxygen coordination in UFF, and proposed to reduce the ϵ value of oxygen to reflect the actual polarization situation.

---

### Key formulas

1. Lennard - Jones potential formula:
$$
u_{\text{non - bonding}}(r_{ij}) = 4 \epsilon_{ij} \left[ \left( \frac{\sigma_{ij}}{r_{ij}} \right)^{12} - \left( \frac{\sigma_{ij}}{r_{ij}} \right)^6 \right] + \frac{q_i q_j}{4 \pi \epsilon_0 r_{ij}}
$$

2. Definition of the loss function:
$$
\text{Error} = \frac{1}{N} \sum_{i = 1}^N |n_{\text{exp},i} - n_{\text{sim},i}|
$$
where $n_{\text{exp},i}$ and $n_{\text{sim},i}$ represent the experimental and simulated gas adsorption amounts respectively.

3. The core formula of MBAR theory:
$$
\langle A \rangle_2 = \frac{\langle A \exp[-\be