Substituted naphthalene reaction rates with peroxy-acid treatment: prediction of reactivity using PEST.

Persistent organic contaminants in the environment pose an environmental risk due to widespread occurrence and toxic properties. Advanced oxidation processes (AOPs) are treatment methods that have been used to successfully degrade organic contaminants in water, soil, sediments and sludge. Reaction rate constants (k) for peroxy acid treatment of 10 substituted naphthalene compounds were determined. The treatment method utilized acetic acid, hydrogen peroxide and a sulphuric acid catalyst to degrade the polyaromatic structures found in the compounds. Molecular structures of the selected compounds were derived at the B3LYP/6-31G* level of theory. Property-encoded surface translator (PEST) descriptors were calculated from B3LYP/6-31G* optimized structures and were found to have significant levels of correlation with k. Models using minimum local ionization potential (PIP.MIN) and a histogram [bin] of the gradient of the K electronic kinetic energy normal to the isosurface (DKN) were evaluated and found to agree within 10% of experimentally derived values of k in most instances. Results show that a combination of PEST descriptors could be used to predict reactivity by the peroxy-acid process. The PEST technology could prove to be a valuable asset for effective remediation design by predicting reaction outcomes for substituted naphthalene compounds and possibly other hydrophobic organic compounds (HOCs).

A quantitative structure-activity relationship to predict efficacy of granular activated carbon adsorption to control emerging contaminants.

A quantitative structure-activity relationship was developed to predict the efficacy of carbon adsorption as a control technology for endocrine-disrupting compounds, pharmaceuticals, and components of personal care products, as a tool for water quality professionals to protect public health. Here, we expand previous work to investigate a broad spectrum of molecular descriptors including subdivided surface areas, adjacency and distance matrix descriptors, electrostatic partial charges, potential energy descriptors, conformation-dependent charge descriptors, and Transferable Atom Equivalent (TAE) descriptors that characterize the regional electronic properties of molecules. We compare the efficacy of linear (Partial Least Squares) and non-linear (Support Vector Machine) machine learning methods to describe a broad chemical space and produce a user-friendly model. We employ cross-validation, y-scrambling, and external validation for quality control. The recommended Support Vector Machine model trained on 95 compounds having 23 descriptors offered a good balance between good performance statistics, low error, and low probability of over-fitting while describing a wide range of chemical features. The cross-validated model using a log-uptake (qe) response calculated at an aqueous equilibrium concentration (Ce) of 1 µM described the training dataset with an r(2) of 0.932, had a cross-validated r(2) of 0.833, and an average residual of 0.14 log units.

Polycyclic aromatic hydrocarbon reaction rates with peroxy-acid treatment: prediction of reactivity using local ionization potential.

Property-Encoded Surface Translator (PEST) descriptors were found to be correlated with the degradation rates of polycyclic aromatic hydrocarbons (PAHs) by the peroxy-acid process. Reaction rate constants (k) in hr(-1) for nine PAHs (acenaphthene, anthracene, benzo[a]pyrene, benzo[k]fluoranthene, fluoranthene, fluorene, naphthalene, phenanthrene, and pyrene) were determined by a peroxy-acid treatment method that utilized acetic acid, hydrogen peroxide, and a sulphuric acid catalyst to degrade the polyaromatic structures. Molecular properties of the selected nine PAHs were derived from structures optimized at B3LYP/6-31G(d) and HF/6-31G(d) levels of theory. Properties of adiabatic and vertical ionization potential (IP), highest occupied molecular orbitals (HOMO), HOMO/lowest unoccupied molecular orbital (LUMO) gap energies and HOMO/singly occupied molecular orbital (SOMO) gap energies were not correlated with rates of peroxy-acid reaction. PEST descriptors were calculated from B3LYP/6-31G(d) optimized structures and found to have significant levels of correlation with k. PIP Min described the minimum local IP on the surface of the molecule and was found to be related to k. PEST technology appears to be an accurate method in predicting reactivity and could prove to be a valuable asset in building treatment models and in remediation design for PAHs and other organic contaminants in the environment.

New Group IV chemical motifs for improved dielectric permittivity of polyethylene.

An enhanced dielectric permittivity of polyethylene and related polymers, while not overly sacrificing their excellent insulating properties, is highly desirable for various electrical energy storage applications. In this computational study, we use density functional theory (DFT) in combination with modified group additivity based high throughput techniques to identify promising chemical motifs that can increase the dielectric permittivity of polyethylene. We consider isolated polyethylene chains and allow the CH2 units in the backbone to be replaced by a number of Group IV halides (viz., SiF2, SiCl2, GeF2, GeCl2, SnF2, or SnCl2 units) in a systematic, progressive, and exhaustive manner. The dielectric permittivity of the chemically modified polyethylene chains is determined by employing DFT computations in combination with the effective medium theory for a limited set of compositions and configurations. The underlying chemical trends in the DFT data are first rationalized in terms of various tabulated atomic properties of the constituent atoms. Next, by parametrizing a modified group contribution expansion using the DFT data set, we are able to predict the dielectric permittivity and bandgap of nearly 30,000 systems spanning a much larger part of the configurational and compositional space. Promising motifs which lead to simultaneously large dielectric constant and band gap in the modified polyethylene chains have been identified. Our theoretical work is expected to serve as a possible motivation for future experimental efforts.

Transferable atom equivalent multicentered multipole expansion method.

The transferability of atomic and functional group properties is an implicit concept in chemistry. The work presented here describes the use of Transferable Atom Equivalents (TAE) to represent molecular electrostatic potential fields through the use of integrated atomic multipole moments that are associated with each TAE atom type used in the reconstruction. TAE molecular surface distributions of electrostatic potentials are compared with analytical ab initio and empirical (Gasteiger) partial charge reference models for several conformations of test peptides. Surface electrostatic potential distributions computed using TAE multipole representations were found to converge at the octopole level, with incremental improvement observed when hexadecapoles were included. Molecular electrostatic potential fields that were produced using the TAE method were observed to be responsive to conformational changes and to compare well with ab initio reference distributions. Generation of TAE atom types and their associated multipoles does not involve fitting to sample electrostatic potential fields, but rather utilizes integrated AIM atomic electron density distributions within representative chemical environments. The RECON program was used for TAE reconstruction. RECON is capable of processing 5,000 drug-sized molecules or 25 proteins per minute per 1.7 GHz P4 Linux processor.

Prediction of protein retention in ion-exchange systems using molecular descriptors obtained from crystal structure.

In this paper, a novel approach is described for the a priori prediction of protein retention in ion exchange systems. Quantitative structure retention relationship (QSRR) models based on a genetic algorithm/partial least squares approach were developed using experimental chromatographic data in concert with molecular descriptors computed using protein crystal structures. The resulting QSRR models were well-correlated, with cross-validated r2 values of 0.938 and 0.907, and the predictive power of these models was demonstrated using proteins not included in the derivation of the models. Importantly, these models were able to predict selectivity reversals observed with two different stationary phase materials. To our knowledge, this is the first published example of predictive QSRR models of protein retention based on crystal structure data.

Data strip mining for the virtual design of pharmaceuticals with neural networks.

A novel neural network based technique, called "data strip mining" extracts predictive models from data sets which have a large number of potential inputs and comparatively few data points. This methodology uses neural network sensitivity analysis to determine which predictors are most significant in the problem. Neural network sensitivity analysis holds all but one input to a trained neural network constant while varying each input over its entire range to determine its effect on the output. The least sensitive variables are iteratively removed from the input set. For each iteration, model cross-validation uses multiple splits of training and validation data to determine an estimate of the model's ability to predict the output for data points not used during training. Elimination of variables through neural network sensitivity analysis and predicting performance through model cross-validation allows the analyst to reduce the number of inputs and improve the model's predictive ability at the same time. This paper illustrates this technique using a cartoon problem from classical physics. It then demonstrates its effectiveness on a pair of challenging problems from combinatorial chemistry with over 400 potential inputs each. For these data sets, model selection by neural sensitivity analysis outperformed other variable selection methods including forward selection and a genetic algorithm.

The response of electrons to structural changes.

The properties of a molecule are determined by the distribution of its electrons. This distribution can be described by the charge density, which is readily obtained from the wave functions derived by ab initio molecular orbital calculations. The charge density may be analyzed in a number of different fashions to give information about the effects of substituents, structural changes, and electronic excitation on the properties of molecules; one common procedure makes use of projection density or charge difference plots. Charge density also may be partitioned among atoms, and by numerical integration over appropriate volume elements one may obtain atomic charges, dipoles, kinetic energies, and other properties of the atoms in a molecule. Many chemical phenomena have been analyzed in terms of charge densities.