A Machine Learning Approach to Zeolite Synthesis Enabled by Automatic Literature Data Extraction

Zeolites are porous, aluminosilicate materials with many industrial and "green" applications. Despite their industrial relevance, many aspects of zeolite synthesis remain poorly understood requiring costly trial and error synthesis. In this paper, we create natural language processing techniques and text markup parsing tools to automatically extract synthesis information and trends from zeolite journal articles. We further engineer a data set of germanium-containing zeolites to test the accuracy of the extracted data and to discover potential opportunities for zeolites containing germanium. We also create a regression model for a zeolite's framework density from the synthesis conditions. This model has a cross-validated root mean squared error of 0.98 T/1000 Å3, and many of the model decision boundaries correspond to known synthesis heuristics in germanium-containing zeolites. We propose that this automatic data extraction can be applied to many different problems in zeolite synthesis and enable novel zeolite morphologies.

Synchrotron XRD and NMR evidence of germanium redistribution during silylation of BEC-type germanosilicate

Post-synthesis silylation of BEC-type germanosilicate followed by synchrotron XRD measurements of the Si/Ge ratios in various T-positions reveals site-specific isomorphous substitution.

Locating Organic Guests in Inorganic Host Materials from X-ray Powder Diffraction Data

Can the location of the organic structure-directing agent (SDA) inside the channel system of a zeolite be determined experimentally in a systematic manner? In an attempt to answer this question, we investigated six borosilicate zeolites of known framework structure (SSZ-53, SSZ-55, SSZ-56, SSZ-58, SSZ-59, and SSZ-60), where the location of the SDA had only been simulated using molecular modeling techniques in previous studies. From synchrotron powder diffraction data, we were able to retrieve reliable experimental positions for the SDA by using a combination of simulated annealing (global optimization) and Rietveld refinement. In this way, problems arising from data quality and only partially compatible framework and SDA symmetries, which can lead to indecipherable electron density maps, can be overcome. Rietveld refinement using geometric restraints were then performed to optimize the positions and conformations of the SDAs. With these improved models, it was possible to go on to determine the location of the B atoms in the framework structure. That is, two pieces of information that are key to the understanding of zeolite synthesis-the location of the organic SDA in the channel system and of the positions adopted by heteroatoms in the silicate framework-can be extracted from experimental data using a systematic strategy. In most cases, the locations of the SDAs determined experimentally compare well with those simulated with molecular modeling, but there are also some clear differences, and the reason for these differences can be understood. The approach is generally applicable, and has also been used to locate organic guests, linkers, and ligands in metal-organic compounds.