A meta-analysis of catalytic literature data reveals property-performance correlations for the OCM reaction

Schmack, Roman; Friedrich, Alexandra; Kondratenko, Evgenii V.; Polte, Jörg; Werwatz, Axel; Kraehnert, Ralph

doi:10.1038/s41467-019-08325-8

Cited by 76 publications

(79 citation statements)

References 67 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…analyzed results of various catalytic reactions obtained from the literature, including CO oxidation, water gas shift (WGS), and transesterification for biodiesel production . It is worthy of note that a recent report by Schmack and coworkers described a meta‐analysis of experimental results of the OCM reaction . The proposed method incorporates general textbook knowledge about fundamental material properties and the experienced intuition of a chemist or material scientist about possible property‐performance correlations in addition to the experimental data reported in literature.…”

Section: Introductionmentioning

confidence: 99%

Statistical Analysis and Discovery of Heterogeneous Catalysts Based on Machine Learning from Diverse Published Data

et al. 2019

View full text Add to dashboard Cite

The literature provides insights for catalyst design and discovery. Effective analysis of reported data using machine learning (ML) methods offers the ability to gain valuable information. However, utilizing the literature in this way has obstacles such as lack of compositional overlaps, bias from prior published data, and low sample counts for many elements. The present study describes an ML approach that considers elemental features as input representations instead of inputting catalyst compositions directly. This ML method has the potential for catalyst discovery, including catalytic reactions with limited catalyst composition overlap in the available data. Oxidative coupling of methane (OCM), water gas shift (WGS), and CO oxidation reactions were chosen to confirm the effectiveness of the proposed method by analysis using several state‐of‐the‐art ML methods. Among the ML methods tested, gradient boosting regression with XGBoost (XGB) provided the best results, and prediction accuracy was improved by the proposed approach for all three reaction types. In addition, a quantitative value of “feature importance score” was calculated to evaluate the most influential input variables on catalyst performance. Finally, catalyst optimization was explored using ML as a “surrogate” model, and the top 20 promising candidate catalysts were identified for the OCM reaction based on the optimization. The advantages of ML in catalysis analysis as well as the difficulties and limitations originating from the complexity of heterogeneous catalysis were explored.

show abstract

Section: Introductionmentioning

confidence: 99%

Statistical Analysis and Discovery of Heterogeneous Catalysts Based on Machine Learning from Diverse Published Data

et al. 2019

View full text Add to dashboard Cite

show abstract

“…So even by looking at just the data without any prior knowledge about OCM we would be able to make an educated guess for some OCM catalysts. Looking at the hypothesis from Schmack et al [1] about one stable oxide support and one active species that could form a carbonate we could also guess a support like Alumina and active metals like Li, Mg, Ca Sr and Ba from the visualizations we just did. Schmack et al also verified their hypothesis experimentally and an initial guess based on the cluster mapping would be a valuable experimental starting point.…”

Section: Unsupervised Learningmentioning

confidence: 92%

“…From the original study [1] the authors did not only publish the manuscript and an electronic supplementary file but there is also an Excel file available online. This link is put now into the variable url.…”

Section: Data Preprocessingmentioning

confidence: 99%

“…[10] about the key elements and from Ref. [1] for the combinations of support and active components. For example Ca on Alumina would be a good experimental candidate and the experimental data from Ref.…”

Section: Random Forestsmentioning

confidence: 99%

“…In a first publication based on the original data Kondratenko et al [11] just used a fraction of the dataset to gain more insights. Finally Schmack et al [1] came to an even more curated dataset based on the original version. Based on their statistical analysis they showed that a good OCM catalyst contains at least two elements and one of it must be able to form a carbonate at the reaction temperature of the OCM reaction.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Primer about Machine Learning in Catalysis – A Tutorial with Code

Palkovits

2020

ChemCatChem

View full text Add to dashboard Cite

Based on a well‐edited dataset from literature by Schmack et al.[1] this manuscript provides a tutorial‐like introduction to Machine Learning (ML) and Data Science (DS) based on the actual programming code in the Python programming language. The study will not only try to illustrate a ML workflow, but will also show important tasks like hyperparameter tuning and data pre‐processing which often cover much of the time of an actual study. Moreover, the study spans from classical ML methods to Deep Learning with Neural Networks.

show abstract

Machine Learning Descriptors for Data‐Driven Catalysis Study

Mou

Han

Smith³

et al. 2023

Advanced Science

View full text Add to dashboard Cite

Traditional trial-and-error experiments and theoretical simulations have difficulty optimizing catalytic processes and developing new, better-performing catalysts. Machine learning (ML) provides a promising approach for accelerating catalysis research due to its powerful learning and predictive abilities. The selection of appropriate input features (descriptors) plays a decisive role in improving the predictive accuracy of ML models and uncovering the key factors that influence catalytic activity and selectivity. This review introduces tactics for the utilization and extraction of catalytic descriptors in ML-assisted experimental and theoretical research. In addition to the effectiveness and advantages of various descriptors, their limitations are also discussed. Highlighted are both 1) newly developed spectral descriptors for catalytic performance prediction and 2) a novel research paradigm combining computational and experimental ML models through suitable intermediate descriptors. Current challenges and future perspectives on the application of descriptors and ML techniques to catalysis are also presented.

show abstract

A meta-analysis of catalytic literature data reveals property-performance correlations for the OCM reaction

Cited by 76 publications

References 67 publications

Statistical Analysis and Discovery of Heterogeneous Catalysts Based on Machine Learning from Diverse Published Data

Statistical Analysis and Discovery of Heterogeneous Catalysts Based on Machine Learning from Diverse Published Data

A Primer about Machine Learning in Catalysis – A Tutorial with Code

Machine Learning Descriptors for Data‐Driven Catalysis Study

Contact Info

Product

Resources

About