Machine Learning Enables Highly Accurate Predictions of Photophysical Properties of Organic Fluorescent Materials: Emission Wavelengths and Quantum Yields

Ju, Cheng‐Wei; Bai, Hanzhi; Li, Bo; Liu, Rizhang

doi:10.26434/chemrxiv.12111060.v3

Cited by 18 publications

(40 citation statements)

References 74 publications

(86 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We identified two transcription errors by Ju et al in the CGSD solvent descriptors (available from https://figshare.com/articles/dataset/ChemFluor/12110619/3) used to train their GBRT models. 16 • In row 22 of Solvent Descriptors.xlsx, the E T (30) value for 1-methyl-2-pyrrolidinone should be 42.2 rather than 48 according to entry no. 284 of Table 2 of Reichardt's work.…”

Section: Author Contributionsmentioning

confidence: 99%

“…Entries Dye Solvent Absorption Emission Other ChemDataExtractor 14 8,467 SMILES Name λ max , max --ChemFluor 16 4,386 SMILES Name Among the previous studies on predicting absorption peak wavelengths or excitation energies, the work of Ju et al, 16 Kang et al, 25 and Joung et al 26 is particularly noteworthy because of the size of their training datasets and the accuracies this enabled them to achieve. Although these recent works achieved impressive accuracies, their reported performance may be more representative of how they would perform in substituent-selection applications as opposed to de novo design tasks with unseen chemistries.…”

Section: Datasetmentioning

confidence: 99%

“…a single dye family) because of the limited availability of large UV/Vis datasets. This data sparsity has been addressed recently with the publication of several experimental datasets, 1,[11][12][13][14][15][16][17][18] described in Table 1. There are also several large computed datasets of excitation energies available (Table 2).…”

Section: Introductionmentioning

confidence: 99%

“…"Full" refers to the full absorption/emission spectrum as xy-coordinate pairs, λ max is the peak wavelength, max is the peak molar attenuation coefficient (also called the molar extinction coefficient or molar absorptivity), σ is the peak FWHM (bandwidth), Φ is the quantum yield, and τ is the fluorescence lifetime. A subset of the data in the ChemFluor 16 set was extracted from the Fluorophores 12 set. The number of entries for the UV/Vis+ dataset includes the count of the dye entries only, and the entries for NIST do not include ions.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Multi-fidelity prediction of molecular optical peaks with deep learning

Greenman

Green

Gómez‐Bombarelli

2021

Preprint

View full text Add to dashboard Cite

Optical properties are central to molecular design for many applications, including solar cells and biomedical imaging. A variety of ab initio and statistical methods have been developed for their prediction, each with a trade-off between accuracy, generality, and cost. Existing theoretical methods such as time-dependent density functional theory (TD-DFT) are generalizable across chemical space because of their robust physics-based foundations but still exhibit random and systematic errors with respect to experiment despite their high computational cost. Statistical methods can achieve high accuracy at a lower cost, but data sparsity and unoptimized molecule and solvent representations often limit their ability to generalize. Here, we utilize directed message passing neural networks (D-MPNNs) to represent both dye molecules and solvents for predictions of molecular absorption peaks in solution. Additionally, we demonstrate a multi-fidelity approach based on an auxiliary model trained on over 28,000 TD-DFT calculations that further improves accuracy and generalizability, as shown through rigorous splitting strategies. Combining several openly-available experimental datasets, we benchmark these methods against a state-of-the-art regression tree algorithm and compare the D-MPNN solvent representation to several alternatives. Finally, we explore the interpretability of the learned representations using dimensionality reduction and evaluate the use of ensemble variance as an estimator of the epistemic uncertainty in our predictions of molecular peak absorption in solution. The prediction methods proposed herein can be integrated with active learning, generative modeling, and experimental workflows to enable the more rapid design of molecules with targeted optical properties.

show abstract

Section: Author Contributionsmentioning

confidence: 99%

Section: Datasetmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Multi-fidelity prediction of molecular optical peaks with deep learning

Greenman

Green

Gómez‐Bombarelli

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…For our database we needed molecules that possess rich spectroscopic and photochemical properties but are notorious in theoretical studies due to SIE. We selected 3,926 such species from existing databases, including 1,941 solar cell materials from the Harvard Clean Energy Project (CEP), 101,102 904 pharmaceutically significant compounds from the DeepChem database, 103 431 fluorescence species from the ChemFluor database, 104 337 organic photovoltaic (OPV) molecules from the Harvard Organic Photovoltaic Dataset (HOPV15), 105 84 organic light-emitting diode (OLED) materials studied by Aspuru-Guzik and coworkers, 106 and 229 oligomers added by us in the present work. These compounds were randomly distributed into a training set of 1,970 and a test set of 1,956, and their structures were provided as simplified molecular-input line-entry system (SMILES) strings in the Supporting Information (SI).…”

mentioning

confidence: 99%

Stacked Ensemble Machine Learning for Range-Separation Parameters

French

Geva

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

High-throughput virtual materials and drug discovery based on density functional theory has achieved tremendous success in recent decades, but its power on organic semiconducting molecules suffered catastrophically from the self-interaction error until the optimally tuned range-separated hybrid (OT-RSH) exchange-correlation functionals were developed. The accurate but expensive �first-principles OT-RSH transitions from a short-range (semi-)local functional to a long-range Hartree-Fock exchange at a distance characterized by the inverse of a molecule-specific, non-empirically-determined range-separation parameter (ω). In the present study, we proposed a promising stacked ensemble machine learning (SEML) model that provides an accelerated alternative of OT-RSH based on system-dependent structural and electronic configurations. We trained ML-ωPBE, the first functional in our series, using a database of 1,970 organic semiconducting molecules with sufficient structural diversity, and assessed its accuracy and efficiency using another 1,956 molecules. Compared with the �first-principles OT-ωPBE, our ML-ωPBE reached a mean absolute error of 0:00504a_0^{-1} for the optimal value of ω, reduced the computational cost for the test set by 2.66 orders of magnitude, and achieved comparable predictive powers in various optical properties.

show abstract

Learning Matter: Materials Design with Machine Learning and Atomistic Simulations

et al. 2022

View full text Add to dashboard Cite

Metrics & MoreArticle Recommendations CONSPECTUS: Designing new materials is vital for addressing pressing societal challenges in health, energy, and sustainability. The combination of physicochemical laws and empirical trial and error has long guided material design, but this approach is limited by the cost of experiments and the difficulty of deriving complex guiding principles. The space of hypothetical materials to be considered is incredibly large, and only a small fraction of possible compounds can ever be tested experimentally. The computational techniques of atomistic simulation and machine learning (ML) offer an avenue to rapidly invent new materials and navigate this enormous space. Together, they can be used to infer complex design principles and identify high-quality candidates more rapidly than trial-and-error experimentation. In this Account, we review our group's recent contributions to simulation and ML for materials design. We begin by discussing the numerical representation of materials for use in ML. Representations can be produced through deterministic algorithms, learnable encodings, or physics-based methods and lead to vector, graph, and matrix outputs. We describe how these different approaches offer distinct material-and application-specific advantages. We provide demonstrations from our own work on small-molecule drugs, macromolecules, dyes, electrolytes, and zeolites. In several cases, we show how the appropriate representation led to guiding principles that facilitated experimental materials design. Next, we highlight the development of ML methods for enhancing atomistic simulation. These advances help to improve simulation accuracy and expand the time and length scales that can be explored. They include differentiable atomistic simulations in which ensemble-averaged quantities are differentiated with respect to system parameters, and novel autoregressive methods for enhanced sampling of challenging physical distributions. Other developments include learnable coarse-grained models, which can accelerate molecular dynamics while minimizing the loss of all-atom information, and ML interatomic potentials, which can be trained on maximally informative quantum chemistry data through active learning and adversarial uncertainty attacks. Next, we show how these combined computational advances have enabled high-throughput virtual screening. This has led to the discovery of low-cost organic structure-directing agents for zeolite synthesis, polymer electrolytes, and efficient photoswitches for targeted medicine. We conclude by discussing the limitations of ML and simulation. These include the large data requirements and limited chemical transferability of the former and the speed−accuracy trade-offs of the latter. We predict that advancements in quantum chemistry will further accelerate simulations, while the incorporation of physical principles will improve the reliability of ML.

show abstract

Machine Learning Enables Highly Accurate Predictions of Photophysical Properties of Organic Fluorescent Materials: Emission Wavelengths and Quantum Yields

Cited by 18 publications

References 74 publications

Multi-fidelity prediction of molecular optical peaks with deep learning

Multi-fidelity prediction of molecular optical peaks with deep learning

Stacked Ensemble Machine Learning for Range-Separation Parameters

Learning Matter: Materials Design with Machine Learning and Atomistic Simulations

Contact Info

Product

Resources

About