Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “molecular design”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 records

RLMolLM: Reinforcement Learning-Enhanced Language Model Framework for Inverse Molecular Design

Inverse molecular design faces significant challenges due to vast chemical space and complex property requirements. While language models show promise for molecular generation, they struggle with validity, multi-property optimization, and structural constraints. This work presents RLMolLM, a reinforcement learning framework combining Proximal Policy Optimization (PPO) with genetic algorithms to address these limitations. Our approach optimizes multiple user-specified properties including quantitative estimates of drug-likeness (QED), synthetic accessibility (SA), and ADMET (absorption, distribution, metabolism, excretion, and toxicity) endpoints without requiring complete model retraining, while maintaining capability for scaffold-constrained generation where specific substructures must be preserved. We outperform state-of-the-art methods for molecular optimization, achieving best QED scores across GDB13, Moses, and Zinc datasets with up to 31% improvement over previous methods while maintaining excellent validity, uniqueness, and novelty metrics. For simultaneous multi-property optimization, our framework achieves substantial improvements in ADMET properties including 4.5-fold reduction in hERG toxicity and enhanced Caco-2 permeability compared to Moses dataset. Under structural constraints, the framework significantly improves molecular validity while preserving scaffolds and effectively optimizing properties. In conclusion, this versatile solution advances pharmaceutical and materials molecular design through effective integration of reinforcement learning and genetic algorithms with multi-property optimization and scaffold preservation.

Genetic algorithms↗

New molecular design for blue BODIPYs

Dihydro analogues of BODIPYs exhibit spectral features ( Φ f ∼ 0.4–0.9) resembling aminocoumarins and suggest applications for broad-band photosensitization or where large Stokes shifts are desired.

Wu, Zhiyuan↗

A molecular design principle of lyotropic liquid-crystalline conjugated polymers with directed alignment capability for plastic electronics

Conjugated polymers with a one-dimensional p-orbital overlap exhibit optoelectronic anisotropy. Their unique anisotropic properties can be fully realized in device applications only when the conjugated chains are aligned. Here, we report a molecular design principle of conjugated polymers to achieve concentration-regulated chain planarization, self-assembly, liquid-crystal-like good mobility and non-interdigitated side chains. As a consequence of these intra- and intermolecular attributes, chain alignment along an applied flow field occurs. This liquid-crystalline conjugated polymer was realized by incorporating intramolecular sulphur–fluorine interactions and bulky side chains linked to a tetrahedral carbon having a large form factor. By optimizing the polymer concentration and the flow field, we could achieve a high dichroic ratio of 16.67 in emission from conducting conjugated polymer films. Two-dimensional grazing-incidence X-ray diffraction was performed to analyse a well-defined conjugated polymer alignment. Thin-film transistors built on highly aligned conjugated polymer films showed more than three orders of magnitude faster carrier mobility along the conjugated polymer alignment direction than the perpendicular direction.

36 MATERIALS SCIENCE↗

Probing the molecular design of hyper-branched aryl polyesters towards lubricant applications

We report novel polymeric materials that may be used as viscosity index improvers (VII) for lubricant applications. Our efforts included probing the comb-burst hyper-branched aryl polyester architecture for beneficial viscosity and friction behavior when utilized as an additive in a group I oil. The monomer was designed as to undergo polymerization via polycondensation within the architectural construct (AB2), typical of hyperbranched polymers. The monomer design was comprised of aliphatic arms (12 or 16 methylenes) to provide the necessary lipophilicity to achieve solubility in a non-polar medium. Once polymerized, via catalyst and heat, the surface alcohols were functionalized with fatty acids (lauric and palmitic). Controlling the aliphatic nature of the internal arms and peripheral end-groups provided four unique flexible polymer designs. Changing the reaction time and concentration provided opportunities to investigate the influence of molecular weight and branching density on oil-solubility, viscosity, and friction. Oil-solubility was found to decrease with fewer internal carbons, but the number of internal carbons appears to have little influence on the bulk solution viscosity. At concentrations of 2 wt % in a group I base oil, these polymer additives demonstrated an improved viscosity index and reduced friction coefficient, validating the basic approach.

Robinson, Joshua W.↗

Molecular Design Principle of All-organic Dyes for Dye-Sensitized Solar Cells

All-organic dyes have shown promising potential as an effective sensitizer in dye-sensitized solar cells (DSSCs). The design concept of all-organic dyes to improve light-to-electric-energy conversion is discussed based on the absorption, electron injection, dye regeneration, and recombination. How the electron-donor–acceptor-type framework can provide better light harvesting through bandgap-tuning and why proper arrangement of acceptor/anchoring groups within a conjugated dye frame is important in suppressing improper charge recombination in DSSCs are discussed. Separating the electron acceptor from the anchoring unit in the donor–acceptor-type organic dye would be a promising strategy to reduce recombination and improve photocurrent generation.

14 SOLAR ENERGY↗

Amphiphilic chlorins and bacteriochlorins in micellar environments. Molecular design, de novo synthesis, and photophysical properties

The incorporation of amphiphilic tetrapyrrole macrocycles in organized media is of great value for a variety of fundamental photochemical studies, yet work to date has chiefly employed porphyrins rather than chlorins or bacteriochlorins. The latter absorb strongly in the red or near-infrared spectral region, respectively. Here, eight amphiphilic macrocycles (six chlorins and two bacteriochlorins) have been designed, synthesized and characterized; the compounds differ in long wavelength absorption (610–745 nm) and peripheral substituents (type of auxochrome, hydrophobic/hydrophilic groups). A methyl pyridinium or benzoic acid substituent at the 15-position provides a polar “tail” whereas a hydrophobic group distal thereto (in the chlorins) provides a lipophilic “head” for the spontaneous incorporation in organized media. The eight (bacterio)chlorins are characterized by static and time-resolved absorption and fluorescence spectroscopy in N,N-dimethylformamide (DMF) and three micellar environments (TX-100, CTAB, and SDS) as well as ultrafast transient absorption studies in DMF.

14 SOLAR ENERGY↗

Adaptive language model training for molecular design

Abstract The vast size of chemical space necessitates computational approaches to automate and accelerate the design of molecular sequences to guide experimental efforts for drug discovery. Genetic algorithms provide a useful framework to incrementally generate molecules by applying mutations to known chemical structures. Recently, masked language models have been applied to automate the mutation process by leveraging large compound libraries to learn commonly occurring chemical sequences (i.e., using tokenization) and predict rearrangements (i.e., using mask prediction). Here, we consider how language models can be adapted to improve molecule generation for different optimization tasks. We use two different generation strategies for comparison, fixed and adaptive. The fixed strategy uses a pre-trained model to generate mutations; the adaptive strategy trains the language model on each new generation of molecules selected for target properties during optimization. Our results show that the adaptive strategy allows the language model to more closely fit the distribution of molecules in the population. Therefore, for enhanced fitness optimization, we suggest the use of the fixed strategy during an initial phase followed by the use of the adaptive strategy. We demonstrate the impact of adaptive training by searching for molecules that optimize both heuristic metrics, drug-likeness and synthesizability, as well as predicted protein binding affinity from a surrogate model. Our results show that the adaptive strategy provides a significant improvement in fitness optimization compared to the fixed pre-trained model, empowering the application of language models to molecular design tasks.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Surfactant-Specific AI-Driven Molecular Design: Integrating Generative Models, Predictive Modeling, and Reinforcement Learning for Tailored Surfactant Synthesis

Molecular design is a critical aspect of various scientific and industrial fields, where the properties of molecules hold significant importance. In this study, a 3-fold methodology design is presented that leverages the power of generative artificial intelligence (AI), predictive modeling, and reinforcement learning to create tailored molecules with desired properties. This model synergistically combines deep learning techniques with Self-Referencing Embedded Strings (SELFIES) molecular representation to build a generative model that generates valid molecules and a graphical neural network model that accurately forecasts molecular properties. The Variational Autoencoder (VAE) coupled with reinforcement learning helps refine molecule generation based on targeted attributes. Data from an experimental study involving surfactants were used to test the framework. A validation of the structural integrity of the molecules generated was conducted, and Tanimoto similarities were used to quantify the similarity and diversity between the original and generated molecular structures. Also, saliency maps for the generated surfactants were produced to identify the features explaining the property values. Lastly, molecular dynamics simulations were used to validate the stability of the generated molecules. The results showed that the proposed framework can effectively produce valid molecules within the set property threshold value.

36 MATERIALS SCIENCE↗

Enhancing generative molecular design via uncertainty-guided fine-tuning of variational autoencoders

In recent years, deep generative models have been successfully applied to various molecular design tasks, particularly in the life and materials sciences. One critical challenge for pre-trained generative molecular design (GMD) models is to fine-tune them to be better suited for downstream design tasks that aim at optimizing specific molecular properties. However, redesigning and training an existing effective generative model from scratch for each new design task are impractical. Furthermore, the black-box nature of typical downstream tasks that involve property prediction makes it nontrivial to optimize the generative model in a task-specific manner. In this work, we propose an uncertainty-guided fine-tuning strategy that can effectively enhance a pre-trained variational autoencoder (VAE) for GMD through performance feedback in an active learning setting. The strategy begins by quantifying the model uncertainty of the generative model using an efficient active subspace-based UQ (uncertainty quantification) scheme. Next, the decoder diversity within the characterized model uncertainty class is explored to expand the viable space of molecular generation. The low-dimensionality of the active subspace makes this exploration tractable using a black-box optimization scheme, which in turn enables us to identify and leverage a diverse set of high-performing models to generate enhanced molecules. Empirical results across six target molecular properties using multiple VAE-based generative models demonstrate that our uncertainty-guided fine-tuning strategy consistently leads to improved models that outperform the original pre-trained models.

97 MATHEMATICS AND COMPUTING↗

Computer-Aided Molecular Design of Bis-phosphine Oxide Lanthanide Extractants

Computer-aided molecular design and high-throughput screening of viable host architectures can significantly reduce the efforts in the design of novel ligands for efficient extraction of rare earth elements. This paper presents a computational approach to the deliberate design of bis-phosphine oxide host architectures that are structurally organized for complexation of trivalent lanthanides. Molecule building software, HostDesigner, was interfaced with molecular mechanics software, PCModel, providing a tool for generating and screening millions of potential R 2 (O)P-link-P(O)R 2 ligand geometries. The molecular mechanics ranking of ligand structures is consistent with both the solution-phase free energies of complexation obtained with density functional theory and the performance of known bis-phosphine oxide extractants. For the case where link is -CH 2 -, evaluation of the ligand geometry provides the first characterization of a steric origin for the ‘anomalous aryl strengthening’ effect. The design approach has identified a number of novel bis-phosphine oxide ligands that are better organized for lanthanide complexation than previously studied examples.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗