Jens Abildskov

Protein Dynamics in Organic Media at Varying Water Activity Studied by Molecular Dynamics Simulation

In nonaqueous enzymology, control of enzyme hydration is commonly approached by fixing the thermodynamic water activity of the medium. In this work, we present a strategy for evaluating the water activity in molecular dynamics simulations of proteins in water/organic solvent mixtures. The method relies on determining the water content of the bulk phase and uses a combination of Kirkwood-Buff theory and free energy calculations to determine corresponding activity coefficients. We apply the method in a molecular dynamics study of Candida antarctica lipase B in pure water and the organic solvents methanol, tert-butyl alcohol, methyl tert-butyl ether, and hexane, each mixture at five different water activities. It is shown that similar water activity yields similar enzyme hydration in the different solvents. However, both solvent and water activity are shown to have profound effects on enzyme structure and flexibility.

Computer-aided polymer design using group contribution plus property models

The preliminary step for polymer product design is to identify the basic repeat unit structure of the polymer that matches the target properties. Computer-aided molecular design (CAMD) approaches can be applied for generating the polymer repeat unit structures that match the required constraints. Polymer repeat unit property prediction models are required to calculate the properties of the generated repeat units. A systematic framework incorporating recently developed group contribution plus (GC+) models and an extended CAMD technique to include design of polymer repeat units is highlighted in this paper. The advantage of a GC+ model in CAMD applications is that a very large number of polymer structures can be considered even though some of the group parameters may not be available. A number of case studies involving different polymer design problems have been solved through the developed framework. The paper highlights three such case studies.

A general model for membrane-based separation processes

A separation process could be defined as a process that transforms a given mixture of chemicals into two or more compositionally distinct end-use products. One way to design these separation processes is to employ a model-based approach, where mathematical models that reliably predict the process behaviour will play an important role. In this paper, modelling of membrane-based processes for separation of gas and liquid mixtures are considered. Two general models, one for membrane-based liquid separation processes (with phase change) and another for membrane-based gas separation are presented. The separation processes covered are: membrane-based gas separation processes, pervaporation and various types of membrane distillation processes. The specific model for each type of membrane-based process is generated from the two general models by applying the specific system descriptions and the corresponding modelling assumptions. Analyses of the generated models, together with their validation and application in process design/analysis are highlighted through several case studies.

Towards the development of a second-order approximation in activity coefficient models based on group contributions

Abstract A simple modification of group contribution based models for estimation of liquid phase activity coefficients is proposed. The main feature of this modification is that contribution estimated from the present first-order groups in many instances are found insufficient since the first-order groups do not provide all the important molecular structural information. Therefore, addition of second-order terms, which provide the needed extra molecular structural information, is proposed. The scope of the new approach is demonstrated through the well-known UNIFAC model in terms of improved correlation/prediction capabilities, distinction between isomers and ability to overcome proximity effects.

Accurate Kirkwood–Buff integrals from molecular simulations

A method is proposed for obtaining thermodynamic properties via Kirkwood–Buff (KB) integrals from molecular simulations. In order to ensure that the KB integration converges, the pair distribution function is extrapolated to large distances using the extension method of Verlet, which enforces a theoretical limiting behaviour on the corresponding direct correlation function. The method is evaluated for the pure Lennard-Jones and Stockmayer fluids. The results are verified by comparing pure fluid isothermal compressibilities obtained from the KB integrals with values from derivatives of equations of state fitted to simulation results. Good agreement is achieved for both fluids at densities larger than 1.5 times the critical density.

Analysis of Infinite Dilution Activity Coefficients of Solutes in Hydrocarbons from UNIFAC

Molecular structural effects on infinite dilution activity coefficients of solutes in n-alkanes and other hydrocarbons are studied within the UNIFAC model. Characteristic chain-length dependencies and other structural relationships imbedded in the model are discussed with emphasis to the consequences this has for model development. The cases treated have subtle but major implications for the correlation of activity coefficients and derivatives since these imply that combinatorial terms may not be small and they can be essential to the success of correlations based on UNIFAC. We have examined a number of infinite dilution properties and find that current expressions do not adequately describe these and other cases. The analysis is described and comparisons of the expressions with data for important systems are presented. New models are not presented, but improvements with either modified group definitions or revised relationships are discussed. The importance of such adjustments, for adding new terms to the existing equations, is stressed.

Computer aided polymer design using multi-scale modelling

The ability to predict the key physical and chemical properties of polymeric materials from their repeat-unit structure and chain-length architecture prior to synthesis is of great value for the design of polymer-based chemical products, with new functionalities and improved performance. Computer aided molecular design (CAMD) methods can expedite the design process by establishing input-output relations between the type and number of functional groups in a polymer repeat unit and the desired macroscopic properties. A multi-scale model-based approach that combines a CAMD technique based on group contributionplus models for predicting polymer repeat unit properties with atomistic simulations for providing first-principles arrangements of the repeat units and for predictions of physical properties of the chosen candidate polymer structures, has been developed and tested for design of polymers with desired properties. A case study is used to highlight the main features of this multi-scale model-based approach for the design of a polymer-based product.

Generation of thermodynamic data for organic liquid mixtures from molecular simulations

Fluctuation solution theory (FST) is employed to analyze results of molecular dynamics (MD) simulations of liquid mixtures. The objective is to generate parameters for macroscopic thermodynamic property models. Two benchmark systems, benzene–methyl acetate at 303.15 K and benzene–ethanol at 298.15 K, are used. MD simulations are performed in the isobaric–isothermal ensemble (NPT) at the respective temperatures and at a pressure of 1 atm. We use the CHARMM27 force field at different mixing ratios. We sample positions to determine the binary (between the centers-of-mass of molecules of a pair) radial distribution functions (RDFs). The RDFs are integrated to give the total correlation function integrals (TCFIs). Errors in TCFIs due to uncertainties in RDFs from simulation have been overcome by introducing a simple expression to reproduce the indirect interactions allowing reliable extrapolation to infinite distances. We compare the results of our computations with measured data on both systems studied. The results for activity coefficients agree to within experimental uncertainty.

The Solvent Selection framework: solvents for organic synthesis, separation processes and ionic- liquids solvents

This paper presents a systematic integrated framework for solvent selection and solvent design. The framework is divided into several modules, which can tackle specific problems in various solvent-based applications. In particular, three modules corresponding to the following solvent selection problems are presented: 1) solvent selection and design for organic synthesis, 2) solvent screening and design of solvent mixtures for pharmaceutical applications and 3) ionic liquids selection and design as solvents. The application of the framework is highlighted successfully through case studies focusing on solvent replacement problem in organic synthesis and solvent mixture design for ibuprofen respectively.

Beyond Basic UNIFAC

A new approach is proposed to extend UNIFAC to more complex substances. While first-order solution-of-groups methods such as UNIFAC can successfully represent measured phase equilibria for structurally simple systems such as mixtures of n-alkanes and linear alkanols with good accuracy, they are not as good for branched chain and polyfunctional substances such as secondary or tertiary alcohols and diols with branched or cyclic alkanes. Our method models departure from first-order behavior by adding second-order contributions to first-order correlations. The second-order contributions are derived from perturbations with respect to structural and energetic parameters. The basis of the extension and its ability to correlate and predict effects due to structural differences in VLE, SLE, and activity coefficients at infinite dilution (γ∞) for binary and multicomponent systems is described. Addition of a few second-order group parameters to the UNIFAC tables has improved the results for essentially all cases where the molecular structure justifies including such effects.