What is ExCorr about?
A popular approach to protect metals from corrosion is the incorporation of chemical substances into the service environment, thus influencing the corrosion rate. Countless experiments were reported in literature to investigate the effect different compounds have on certain base materials, and there are many more experiments to come. The ExCorr database is a collection of these experiments, providing a platform for researchers to explore available data, contribute new data, and build machine learning models to predict the corrosion inhibition performance of small (organic) compounds.
How to use ExCorr?
ExCorr comes with three major components: a structure landscape (1), a table of already tested compounds (2) and a table of available experimental data (3).
Structure landscape
Each point in the structure landscape stands for an individual chemical compound. The landscape is a two-dimensional projection of all molecular similarities within the database, as calculated by the smooth overlap of atomic position (SOAP). Two points that lie close to each other in the landscape are structurally similar, while points that are far apart are structurally different. By clicking on a point, you can access the structural information in the structure table as well as associated experimental data in the experiment table.
Structure table
In the structure table, you can click on a row to jump to the respective structure in the structure landscape and access the associated experimental data in the experiment table. The structure table contains IUPAC names (if available), CAS numbers (if available) and SMILES strings of the included compounds. Using the search field, you can search for individual compounds, e.g., using a CAS number. You can get further information on the compounds by clicking on the information icon. Additionally, you can look for similar compounds to the one selected within the database by clicking on the lense icon. The gear icon allows you to control which columns are shown in the experiment table.
Experiment table
The experiment table includes all available experimental associated to the selected compound. You can control the shown columns by clicking on the gear icon next to the structure table. You can also filter the data by clicking on the filter button above the structure table. Filter settings include the base material, the alloy, the method used to collect the data, the pH and the minimal recorded inhibition efficiency.
How to contribute?
You want to contribute your experimental data to ExCorr, so the community can benefit from it? You are awesome! Please use this template, fill in the data and send it to XXX. We will process it and upload it to the database so everyone can access it.
If you need help to fill out the template or have any questions, do not hesitate to contact us.
API
Machine learning models work best when provided with high-quality and versatile data. To facilitate easy access to the ExCorr database to build machine learning models, we provide an API that allows you to use the data in a structured way. If you use the data in your work, we would appreciate if you could cite the accompanying review paper:
D. A. Winkler, A. E. Hughes, C. Özkan, A. Mol, T. Würger, C. Feiler, D. Zhang, and S. V. Lamaka. Impact of automation, computational models, and inhibition mechanisms on the discovery of organic corrosion inhibitors. Progress in Materials Science, 2024.
The API is available at https://excorr.web.app/database/api/experiments. You can directly use the data in your machine learning workflow using the following Python code:
1import urllib.request, json2import pandas as pd3with urllib.request.urlopen("https://excorr.web.app/database/api/experiments") as url:4data = json.load(url)56allExperiments = pd.DataFrame(data["allExperiments"])