Finding the Right Bricks for Molecular Legos: A Data Mining Approach to Organic Semiconductor Design

Künkel, Christian; Schober, Christoph; Margraf, Johannes T.; Reuter, Karsten; Oberhofer, Harald

doi:10.1021/acs.chemmater.8b04436

Cited by 53 publications

(63 citation statements)

References 104 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[48] Schober et al have screened the CSD for monomolecular organic crystals with the objective to identify organic semiconductors with high charge carrier mobility. [31,57,58] For this study, we extracted molecules from the crystals and relaxed them in vacuum with the aforementioned computational parameters. The OE dataset is not yet publicly available.…”

Section: Ii2 Aa: 44 K Amino Acids and Dipeptidesmentioning

confidence: 99%

Chemical diversity in molecular orbital energy predictions with kernel ridge regression

Stuke

Todorović

Rupp

et al. 2019

The Journal of Chemical Physics

Self Cite

111

View full text Add to dashboard Cite

Instant machine learning predictions of molecular properties are desirable for materials design, but the predictive power of the methodology is mainly tested on well-known benchmark datasets. Here, we investigate the performance of machine learning with kernel ridge regression (KRR) for the prediction of molecular orbital energies on three large datasets: the standard QM9 small organic molecules set, amino acid and dipeptide conformers, and organic crystal-forming molecules extracted from the Cambridge Structural Database. We focus on prediction of highest occupied molecular orbital (HOMO) energies, computed at density-functional level of theory. Two different representations that encode molecular structure are compared: the Coulomb matrix (CM) and the many-body tensor representation (MBTR). We find that KRR performance depends significantly on the chemistry of the underlying dataset and that the MBTR is superior to the CM, predicting HOMO energies with a mean absolute error as low as 0.09 eV. To demonstrate the power of our machine learning method, we apply our model to structures of 10k previously unseen molecules. We gain instant energy predictions that allow us to identify interesting molecules for future applications.

show abstract

Section: Ii2 Aa: 44 K Amino Acids and Dipeptidesmentioning

confidence: 99%

Chemical diversity in molecular orbital energy predictions with kernel ridge regression

Stuke

Todorović

Rupp

et al. 2019

The Journal of Chemical Physics

Self Cite

111

View full text Add to dashboard Cite

show abstract

“…[33][34][35] Our study indicated that this machine-learning strategies may provide OFET researchers supporting details to fine-tune the electronic structure and thus the charge transport property of the n-type organic materials. [33][34][35] Our study indicated that this machine-learning strategies may provide OFET researchers supporting details to fine-tune the electronic structure and thus the charge transport property of the n-type organic materials.…”

mentioning

confidence: 76%

Machine Learning for Understanding the Relationship between the Charge Transport Mobility and Electronic Energy Levels for n‐Type Organic Field‐Effect Transistors

Lee

2019

Adv Elect Materials

View full text Add to dashboard Cite

Organic field‐effect transistors (OFETs) have received considerably more attention than inorganic‐based field‐effect transistors for use in next generation of organic circuits. There are a number of variables, for example, the ordering of the OFETs, their energy levels, and the material used for source/drain electrodes, that influence the magnitude of charge transport mobility. Importantly, a suitable energy level match between highest occupied molecular orbital (HOMO) or lowest unoccupied molecular orbital (LUMO) energy level and work function of the electrodes may have a large influence on the measured mobility. An informatics approach, specifically use of machine learning, is proposed for charge transport mobility prediction. Gradient Boosting and Random Forest regression algorithms are used to model previous experimental datasets and HOMO and LUMO energy levels of n‐type materials are optimized using expected machine‐learning methods. The results reveal that Random Forest model benefits the functional analysis of n‐type OFETs in three ways: 1) it provides better understanding of current n‐type organic materials, 2) it may guide the choice of n‐type organic materials and conducting electrodes, and 3) it measures the tradeoffs between the charge transport mobility and electronic energy levels for n‐type OFETs.

show abstract

“…This is particularly relevant in a high-throughput setting, e. g. when a large chemical reaction network with many intermediates and transition states is to be explored, or a large chemical space is of interest. [10][11][12][13] The wide range of ML methods that have emerged in this context raises the question which one should be used for a given application. Since the atomization energy (AE) has a long tradition as the foremost benchmark property to judge the accuracy of quantum chemical approximations, [14][15][16] it has also become one of the standard targets to illustrate the accuracy of novel ML methods.…”

Section: Introductionmentioning

confidence: 99%

“…Importantly, the performance of the models is compared across different size-ranges both within the QM9 database and between databases going up to molecules with more than 80 heavy atoms. [11,22]…”

Section: Introductionmentioning

confidence: 99%

Size‐Extensive Molecular Machine Learning with Global Representations

et al. 2020

Self Cite

View full text Add to dashboard Cite

Machine learning (ML) models are increasingly used in combination with electronic structure calculations to predict molecular properties at a much lower computational cost in highthroughput settings. Such ML models require representations that encode the molecular structure, which are generally designed to respect the symmetries and invariances of the target property. However, size-extensivity is usually not guaranteed for so-called global representations. In this contribution, we show how extensivity can be built into global ML models using, e. g., the Many-Body Tensor Representation. Properties of extensive and non-extensive models for the atomization energy are systematically explored by training on small molecules and testing on small, medium and large molecules. Our results show that non-extensive models are only useful in the size-range of their training set, whereas extensive models provide reasonable predictions across large size differences. Remaining sources of error for extensive models are discussed.

show abstract

Finding the Right Bricks for Molecular Legos: A Data Mining Approach to Organic Semiconductor Design

Cited by 53 publications

References 104 publications

Chemical diversity in molecular orbital energy predictions with kernel ridge regression

Chemical diversity in molecular orbital energy predictions with kernel ridge regression

Machine Learning for Understanding the Relationship between the Charge Transport Mobility and Electronic Energy Levels for n‐Type Organic Field‐Effect Transistors

Size‐Extensive Molecular Machine Learning with Global Representations

Contact Info

Product

Resources

About