scholarly journals Enumeration of de novo inorganic complexes for chemical discovery and machine learning

2020 ◽  
Vol 5 (1) ◽  
pp. 139-152 ◽  
Author(s):  
Stefan Gugler ◽  
Jon Paul Janet ◽  
Heather J. Kulik
Keyword(s):  
Machine Learning ◽  
Transition Metal ◽  
Metal Complexes ◽  
Transition Metal Complexes ◽  
Spin State ◽  
De Novo ◽  
Machine Learning Model ◽  
Inorganic Complexes ◽  
Model Training ◽  
Unique Spin

Enumerated, de novo transition metal complexes have unique spin state properties and accelerate machine learning model training.

scholarly journals Seeing is Believing: Experimental Spin States from Machine Learning Model Structure Predictions

2020 ◽  
Author(s):  
Michael Taylor ◽  
Tzuhsiung Yang ◽  
Sean Lin ◽  
Aditya Nandy ◽  
Jon Paul Janet ◽  
...  
Keyword(s):  
Machine Learning ◽  
Electronic Structure ◽  
Transition Metal ◽  
Metal Complexes ◽  
Transition Metal Complexes ◽  
Ground State ◽  
Spin States ◽  
Machine Learning Model ◽  
Metal Ligand ◽  
Ligand Bond

<p>Determination of ground-state spins of open-shell transition metal complexes is critical to understanding catalytic and materials properties but also challenging with approximate electronic structure methods. As an alternative approach, we demonstrate how structure alone can be used to guide assignment of ground-state spin from experimentally determined crystal structures of transition metal complexes. We first identify the limits of distance-based heuristics from distributions of metal–ligand bond lengths of over 2,000 unique mononuclear Fe(II)/Fe(III) transition metal complexes. To overcome these limits, we employ artificial neural networks (ANNs) to predict spin-state-dependent metal–ligand bond lengths and classify experimental ground state spins based on agreement of experimental structures with the ANN predictions. Although the ANN is trained on hybrid density functional theory data, we exploit the method-insensitivity of geometric properties to enable assignment of ground states for the majority (ca. 80-90%) of structures. We demonstrate the utility of the ANN by data-mining the literature for spin-crossover (SCO) complexes, which have experimentally-observed temperature-dependent geometric structure changes, by correctly assigning almost all (> 95%) spin states in the 46 Fe(II) SCO complex set. This approach represents a promising complement to more conventional energy-based spin-state assignment from electronic structure theory at the low cost of a machine learning model. </p>

scholarly journals Seeing is Believing: Experimental Spin States from Machine Learning Model Structure Predictions

2020 ◽  
Author(s):  
Michael Taylor ◽  
Tzuhsiung Yang ◽  
Sean Lin ◽  
Aditya Nandy ◽  
Jon Paul Janet ◽  
...  
Keyword(s):  
Machine Learning ◽  
Electronic Structure ◽  
Transition Metal ◽  
Metal Complexes ◽  
Transition Metal Complexes ◽  
Ground State ◽  
Spin States ◽  
Machine Learning Model ◽  
Metal Ligand ◽  
Ligand Bond

<p>Determination of ground-state spins of open-shell transition metal complexes is critical to understanding catalytic and materials properties but also challenging with approximate electronic structure methods. As an alternative approach, we demonstrate how structure alone can be used to guide assignment of ground-state spin from experimentally determined crystal structures of transition metal complexes. We first identify the limits of distance-based heuristics from distributions of metal–ligand bond lengths of over 2,000 unique mononuclear Fe(II)/Fe(III) transition metal complexes. To overcome these limits, we employ artificial neural networks (ANNs) to predict spin-state-dependent metal–ligand bond lengths and classify experimental ground state spins based on agreement of experimental structures with the ANN predictions. Although the ANN is trained on hybrid density functional theory data, we exploit the method-insensitivity of geometric properties to enable assignment of ground states for the majority (ca. 80-90%) of structures. We demonstrate the utility of the ANN by data-mining the literature for spin-crossover (SCO) complexes, which have experimentally-observed temperature-dependent geometric structure changes, by correctly assigning almost all (> 95%) spin states in the 46 Fe(II) SCO complex set. This approach represents a promising complement to more conventional energy-based spin-state assignment from electronic structure theory at the low cost of a machine learning model. </p>

scholarly journals Strategies and Software for Machine Learning Accelerated Discovery in Transition Metal Chemistry

2018 ◽  
Author(s):  
Aditya Nandy ◽  
Chenru Duan ◽  
Jon Paul Janet ◽  
Stefan Gugler ◽  
Heather Kulik
Keyword(s):  
Machine Learning ◽  
Transition Metal ◽  
Metal Complexes ◽  
Transition Metal Complexes ◽  
Density Functional ◽  
Open Shell ◽  
Frontier Molecular Orbital ◽  
Data Set ◽  
Homo Level ◽  
Homo Lumo

<p>Machine learning the electronic structure of open shell transition metal complexes presents unique challenges, including robust and automated data set generation. Here, we introduce tools that simplify data acquisition from density functional theory (DFT) and validation of trained machine learning models using the molSimplify automatic design (mAD) workflow. We demonstrate this workflow by training and comparing the performance of LASSO, kernel ridge regression (KRR), and artificial neural network (ANN) models using heuristic, topological revised autocorrelation (RAC) descriptors we have recently introduced for machine learning inorganic chemistry. On a series of open shell transition metal complexes, we evaluate set aside test errors of these models for predicting the HOMO level and HOMO-LUMO gap. The best performing models are ANNs, which show 0.15 and 0.25 eV test set mean absolute errors on the HOMO level and HOMO-LUMO gap, respectively. Poor performing KRR models using the full 153-feature RAC set are improved to nearly the same performance as the ANNs when trained on down-selected subsets of 20-30 features. Analysis of the essential descriptors for HOMO and HOMO-LUMO gap prediction as well as comparison to subsets previously obtained for other properties reveals the paramount importance of non-local, steric properties in determining frontier molecular orbital energetics. We demonstrate our model performance on diverse complexes and in the discovery of molecules with target HOMO-LUMO gaps from a large 15,000 molecule design space in minutes rather than days that full DFT evaluation would require. </p>

scholarly journals Should External Heavy-Atom Effects be Expected in Transition Metal Complexes?

2018 ◽  
Author(s):  
Justin Lomont ◽  
Son Nguyen ◽  
Charles Harris
Keyword(s):  
Transition Metal ◽  
Metal Complexes ◽  
Transition Metal Complexes ◽  
Spin State ◽  
Computational Study ◽  
Heavy Atom ◽  
Minimum Energy ◽  
Free Energy Barrier ◽  
Heavy Atom Effect ◽  
Atom Effect

Spin state changes frequently play a key role in the reactivity of transition metal complexes. The rates of spin-forbidden reactions are mediated both by the free energy barrier connecting reactants and products, as well as the strength of spin-orbit coupling (SOC) between the relevant electronic states. Since the 1950’s, there have been numerous demonstrations of external heavy-atom effects on organic molecules, in which a heavy atom, not chemically bonded to the molecule undergoing a change in spin state, perturbs the strength of SOC via an intermolecular effect. However, the potential role of external heavy atom effects on the rates of reactions involving transition metal complexes remains almost entirely unexplored. We report a computational investigation into the changes in SOC that occur along a bimolecular reaction coordinate when an incoming atom coordinates to the prototypical triplet reaction intermediate Fe(CO)4. The calculated changes in SOC are compared for molecules containing atoms ranging in atomic number from Z = 8 to Z = 53 approaching the Fe center (ZFe =26). No evidence for an external heavy atom effect was found, and the changes in SOC with the approach of each incoming group were similar in magnitude. In fact, when taking into account the different minimum energy crossing point geometries for the different incoming groups, the opposite of an external heavy atom effect trend is predicted for this reaction. The results of this computational study suggest that external heavy atom effects are unlikely to have a significant effect on the rates of spin-forbidden reactions for transition metal complexes. <br>

Sign in / Sign up

Export Citation Format

Share Document