Claire S. Adjiman

Towards crystal structure prediction of complex organic compounds – a report on the fifth blind test

The results of the fifth blind test of crystal structure prediction, which show important success with more challenging large and flexible molecules, are presented and discussed.

A group contribution method for associating chain molecules based on the statistical associating fluid theory (SAFT-γ)

A predictive group-contribution statistical associating fluid theory (SAFT-gamma) is developed by extending the molecular-based SAFT-VR equation of state [A. Gil-Villegas et al. J. Chem. Phys. 106, 4168 (1997)] to treat heteronuclear molecules which are formed from fused segments of different types. Our models are thus a heteronuclear generalization of the standard models used within SAFT, comparable to the optimized potentials for the liquid state OPLS models commonly used in molecular simulation; an advantage of our SAFT-gamma over simulation is that an algebraic description for the thermodynamic properties of the model molecules can be developed. In our SAFT-gamma approach, each functional group in the molecule is modeled as a united-atom spherical (square-well) segment. The different groups are thus characterized by size (diameter), energy (well depth) and range parameters representing the dispersive interaction, and by shape factor parameters (which denote the extent to which each group contributes to the overall molecular properties). For associating groups a number of bonding sites are included on the segment: in this case the site types, the number of sites of each type, and the appropriate association energy and range parameters also have to be specified. A number of chemical families (n-alkanes, branched alkanes, n-alkylbenzenes, mono- and diunsaturated hydrocarbons, and n-alkan-1-ols) are treated in order to assess the quality of the SAFT-gamma description of the vapor-liquid equilibria and to estimate the parameters of various functional groups. The group parameters for the functional groups present in these compounds (CH(3), CH(2), CH(3)CH, ACH, ACCH(2), CH(2)=, CH=, and OH) together with the unlike energy parameters between groups of different types are obtained from an optimal description of the pure component phase equilibria. The approach is found to describe accurately the vapor-liquid equilibria with an overall %AAD of 3.60% for the vapor pressure and 0.86% for the saturated liquid density. The fluid phase equilibria of some larger compounds comprising these groups, which are not included in the optimization database and some binary mixtures are examined to confirm the predictive capability of the SAFT-gamma approach. A key advantage of our method is that the binary interaction parameters between groups can be estimated directly from an examination of pure components alone. This means that as a first approximation the fluid-phase equilibria of mixtures of compounds comprising the groups considered can be predicted without the need for any adjustment of the binary interaction parameters (which is common in other approaches). The special case of molecular models comprising tangentially bonded (all-atom and united-atom) segments is considered separately; we comment on the adequacy of such models in representing the properties of real molecules.

Accurate statistical associating fluid theory for chain molecules formed from Mie segments

A highly accurate equation of state (EOS) for chain molecules formed from spherical segments interacting through Mie potentials (i.e., a generalized Lennard-Jones form with variable repulsive and attractive exponents) is presented. The quality of the theoretical description of the vapour-liquid equilibria (coexistence densities and vapour pressures) and the second-derivative thermophysical properties (heat capacities, isobaric thermal expansivities, and speed of sound) are critically assessed by comparison with molecular simulation and with experimental data of representative real substances. Our new EOS represents a notable improvement with respect to previous versions of the statistical associating fluid theory for variable range interactions (SAFT-VR) of the generic Mie form. The approach makes rigorous use of the Barker and Henderson high-temperature perturbation expansion up to third order in the free energy of the monomer Mie system. The radial distribution function of the reference monomer fluid, which is a prerequisite for the representation of the properties of the fluid of Mie chains within a Wertheim first-order thermodynamic perturbation theory (TPT1), is calculated from a second-order expansion. The resulting SAFT-VR Mie EOS can now be applied to molecular fluids characterized by a broad range of interactions spanning from soft to very repulsive and short-ranged Mie potentials. A good representation of the corresponding molecular-simulation data is achieved for model monomer and chain fluids. When applied to the particular case of the ubiquitous Lennard-Jones potential, our rigorous description of the thermodynamic properties is of equivalent quality to that obtained with the empirical EOSs for LJ monomer (EOS of Johnson et al.) and LJ chain (soft-SAFT) fluids. A key feature of our reformulated SAFT-VR approach is the greatly enhanced accuracy in the near-critical region for chain molecules. This attribute, combined with the accurate modeling of second-derivative properties, allows for a much improved global representation of the thermodynamic properties and fluid-phase equilibria of pure fluids and their mixtures.

Report on the sixth blind test of organic crystal structure prediction methods

The results of the sixth blind test of organic crystal structure prediction methods are presented and discussed, highlighting progress for salts, hydrates and bulky flexible molecules, as well as on-going challenges.

A Rigorous Global Optimization Algorithm for Problems with Ordinary Differential Equations

The optimization of systems which are described by ordinary differential equations (ODEs) is often complicated by the presence of nonconvexities. A deterministic spatial branch and bound global optimization algorithm is presented in this paper for systems with ODEs in the constraints. Upper bounds for the global optimum are produced using the sequential approach for the solution of the dynamic optimization problem. The required convex relaxation of the algebraic functions is carried out using well-known global optimization techniques. A convex relaxation of the time dependent information is obtained using the concept of differential inequalities in order to construct bounds on the space of solutions of parameter dependent ODEs as well as on their second-order sensitivities. This information is then incorporated in the convex lower bounding NLP problem. The global optimization algorithm is illustrated by applying it to four case studies. These include parameter estimation problems and simple optimal control problems. The application of different underestimation schemes and branching strategies is discussed.

Successful prediction of a model pharmaceutical in the fifth blind test of crystal structure prediction

The range of target structures in the fifth international blind test of crystal structure prediction was extended to include a highly flexible molecule, (benzyl-(4-(4-methyl-5-(p-tolylsulfonyl)-1,3-thiazol-2-yl)phenyl)carbamate, as a challenge representative of modern pharmaceuticals. Two of the groups participating in the blind test independently predicted the correct structure. The methods they used are described and contrasted, and the implications of the capability to tackle molecules of this complexity are discussed.

Group contribution methodology based on the statistical associating fluid theory for heteronuclear molecules formed from Mie segments.

A generalization of the recent version of the statistical associating fluid theory for variable range Mie potentials [Lafitte et al., J. Chem. Phys. 139, 154504 (2013)] is formulated within the framework of a group contribution approach (SAFT-γ Mie). Molecules are represented as comprising distinct functional (chemical) groups based on a fused heteronuclear molecular model, where the interactions between segments are described with the Mie (generalized Lennard-Jonesium) potential of variable attractive and repulsive range. A key feature of the new theory is the accurate description of the monomeric group-group interactions by application of a high-temperature perturbation expansion up to third order. The capabilities of the SAFT-γ Mie approach are exemplified by studying the thermodynamic properties of two chemical families, the n-alkanes and the n-alkyl esters, by developing parameters for the methyl, methylene, and carboxylate functional groups (CH3, CH2, and COO). The approach is shown to describe accurately the fluid-phase behavior of the compounds considered with absolute average deviations of 1.20% and 0.42% for the vapor pressure and saturated liquid density, respectively, which represents a clear improvement over other existing SAFT-based group contribution approaches. The use of Mie potentials to describe the group-group interaction is shown to allow accurate simultaneous descriptions of the fluid-phase behavior and second-order thermodynamic derivative properties of the pure fluids based on a single set of group parameters. Furthermore, the application of the perturbation expansion to third order for the description of the reference monomeric fluid improves the predictions of the theory for the fluid-phase behavior of pure components in the near-critical region. The predictive capabilities of the approach stem from its formulation within a group-contribution formalism: predictions of the fluid-phase behavior and thermodynamic derivative properties of compounds not included in the development of group parameters are demonstrated. The performance of the theory is also critically assessed with predictions of the fluid-phase behavior (vapor-liquid and liquid-liquid equilibria) and excess thermodynamic properties of a variety of binary mixtures, including polymer solutions, where very good agreement with the experimental data is seen, without the need for adjustable mixture parameters.

Efficient Handling of Molecular Flexibility in Lattice Energy Minimization of Organic Crystals

This paper presents a novel algorithm, CrystalOptimizer, for the minimization of the lattice energy of crystals formed by flexible molecules. The algorithm employs isolated-molecule quantum mechanical (QM) calculations of the intramolecular energy and conformation-dependent atomic multipoles in the course of the lattice energy minimization. The algorithm eliminates the need to perform QM calculations at each iteration of the minimization by using Local Approximate Models (LAMs), with a minimal impact on accuracy. Additional computational efficiencies are achieved by storing QM-derived components of the lattice energy model in a database and reusing them in subsequent calculations whenever possible. This makes the approach particularly well suited to applications that involve a sequence of lattice energy evaluations, such as crystal structure prediction. The algorithm is capable of handling efficiently complex systems with considerable conformational flexibility. The paper presents examples of the algorithms application ranging from single-component crystals to cocrystals and salts of flexible molecules with tens of intramolecular degrees of freedom whose optimal values are determined by the interplay of conformational strain and packing forces. For any given molecule, the degree of flexibility to be considered can vary from a few torsional angles to relaxation of the entire set of torsion angles, bond angles, and bond lengths present in the molecule.

SAFT-γ Force Field for the Simulation of Molecular Fluids: 2. Coarse-Grained Models of Greenhouse Gases, Refrigerants, and Long Alkanes

In the first paper of this series [C. Avendaño, T. Lafitte, A. Galindo, C. S. Adjiman, G. Jackson, and E. A. Müller, J. Phys. Chem. B2011, 115, 11154] we introduced the SAFT-γ force field for molecular simulation of fluids. In our approach, a molecular-based equation of state (EoS) is used to obtain coarse-grained (CG) intermolecular potentials that can then be employed in molecular simulation over a wide range of thermodynamic conditions of the fluid. The macroscopic experimental data for the vapor-liquid equilibria (saturated liquid density and vapor pressure) of a given system are represented with the SAFT-VR Mie EoS and used to estimate effective intermolecular parameters that provide a good description of the thermodynamic properties by exploring a wide parameter space for models based on the Mie (generalized Lennard-Jones) potential. This methodology was first used to develop a simple single-segment CG Mie model of carbon dioxide (CO2) which allows for a reliable representation of the fluid-phase equilibria (for which the model was parametrized), as well as an accurate prediction of other properties such as the enthalpy of vaporization, interfacial tension, supercritical density, and second-derivative thermodynamic properties (thermal expansivity, isothermal compressibility, heat capacity, Joule-Thomson coefficient, and speed of sound). In our current paper, the methodology is further applied and extended to develop effective SAFT-γ CG Mie force fields for some important greenhouse gases including carbon tetrafluoride (CF4) and sulfur hexafluoride (SF6), modeled as simple spherical molecules, and for long linear alkanes including n-decane (n-C10H22) and n-eicosane (n-C20H42), modeled as homonuclear chains of spherical Mie segments. We also apply the SAFT-γ methodology to obtain a CG homonuclear two-segment Mie intermolecular potential for the more challenging polar and asymmetric compound 2,3,3,3-tetrafluoro-1-propene (HFO-1234yf), a novel replacement refrigerant with promising properties. The description of the fluid-phase behavior and the prediction of the other thermophysical properties obtained by molecular simulation using our SAFT-γ CG Mie force fields are found to be of comparable quality (and sometimes superior) to that obtained using the more sophisticated all-atom (AA) and united-atom (UA) models commonly employed in the field. We should emphasize that though the focus of our current work is on simple homonuclear models, the SAFT-γ methodology is based on a group contribution methodology which is naturally suited to the development of more sophisticated heteronuclear models.

Global optimization of dynamic systems

Many chemical engineering systems are described by differential equations. Their optimization is often complicated by the presence of nonconvexities. A deterministic spatial branch and bound global optimization algorithm is presented for problems with a set of first-order differential equations in the constraints. The global minimum is approached from above and below by generating converging sequences of upper and lower bounds. Local solutions, obtained using the sequential approach for the solution of the dynamic optimization problem, provide upper bounds. Lower bounds are produced from the solution of a convex relaxation of the original problem. Algebraic functions are relaxed using well-known convex underestimation techniques. The convex relaxation of the dynamic information is achieved using a new convex relaxation procedure. Parameter independent as well as parameter dependent bounds on the dynamic system are utilized. The global optimization algorithm is illustrated by applying it to case studies relevant to chemical engineering, where affine functions of the optimization variables are used as a relaxation of the dynamic system.