Machine Learning Adsorbent Screening via Molecular Interaction Indices

Solution Overview

Problem

Current methods for predicting the adsorptive properties of adsorbates to adsorbents are inefficient and time-consuming, particularly when using quantum chemical calculations, which require significant computational resources and time.

Innovation Solution

A learning device and method that utilize machine learning to generate a learned model by associating descriptors of molecular structures of adsorbates with interaction indices of intermolecular bonds between adsorbates and adsorbents, allowing for the prediction of adsorptive properties without directly calculating adsorption energy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If quantum chemical calculation is used to predict adsorptive properties, then prediction accuracy is improved, but calculation time and computational resources increase significantly

Engineering Contradiction:

Improveprediction accuracyVSAvoidcalculation time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies preliminary action by pre-calculating and storing interaction indices for various molecular structures in a training dataset before actual prediction is needed. The machine learning model is trained in advance on this pre-prepared data, so that during actual use, predictions can be made rapidly without performing time-consuming quantum chemical calculations each time.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent creates a simplified copy or surrogate model of the complex quantum chemical calculation system. Instead of performing full quantum chemical calculations, a machine learning model is trained to replicate the prediction functionality using pre-computed interaction indices from a training set, providing a faster approximation that maintains reasonable accuracy.

Inventive Principle:

Principle #26Copying

2Measurement precision

If quantum chemical calculation is used to predict adsorptive properties, then prediction accuracy is improved, but computational resources required increase significantly

Engineering Contradiction:

Improveprediction accuracyVSAvoidcomputational resources

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent applies preliminary action by pre-calculating and storing interaction indices for various molecular structures in a training dataset before actual prediction is needed. The machine learning model is trained in advance on this pre-prepared data, so that during actual use, predictions can be made rapidly without performing time-consuming quantum chemical calculations each time.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent creates a simplified copy or surrogate model of the complex quantum chemical calculation system. Instead of performing full quantum chemical calculations, a machine learning model is trained to replicate the prediction functionality using pre-computed interaction indices from a training set, providing a faster approximation that maintains reasonable accuracy.

Inventive Principle:

Principle #26Copying

3Productivity

If machine learning model is used to predict adsorptive properties, then calculation speed is improved, but model training complexity increases

Engineering Contradiction:

Improvecalculation speedVSAvoidmodel training complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent applies parameter changes by transforming the complex quantum chemical calculation problem into a machine learning prediction problem with specific parameters: using interaction indices as features and adsorption energy as the target variable. This parameter transformation allows the use of efficient machine learning algorithms that can achieve fast predictions while managing training complexity through proper feature selection.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20250054583A1Learning device, learning method, screening device, and screening method

Publication Date: 2025.02.13 ENEOS HLDG INC

AI summary

A learning device includes a processor; and a memory storing program instructions that cause the processor to generate a learned model by performing machine learning using a training dataset in which a descriptor of a molecular structure of an adsorbate to be adsorbed to an adsorbent is associated with an interaction index of one or more intermolecular bonds of interest between the adsorbate and the adsorbent.