Michael T. Klein

Computer generated reaction modelling: Decomposition and encoding algorithms for determining species uniqueness

Abstract The concept of computer generated reaction modelling was broadened through the development of a general planar graph algorithm for determination of isomorphism. The previous capability was limited by its inability to determine the uniqueness of ring-containing species unambiguously, restricting the application of automatic network generation to non-cyclic species or cyclic species where the ring was not involved in the chemical transformation. In this work, the systematic identification of both noncyclic and cyclic species was carried out by constructing the structurally explicit decomposition tree, an assembly of the biconnected components of the graph, from which a graph invariant unique string code was obtained by iteratively encoding and ordering the subtrees of the decomposition tree. A lexicographical comparison of the unique string code of the candidate species with the string codes of all previously generated species with the same empirical formula allowed unambiguous determination of species uniqueness.

Molecular representation of complex hydrocarbon feedstocks through efficient characterization and stochastic algorithms

Abstract A stochastic method was devised to transform efficient sets of analytical characterizations into molecular representations of complex petroleum feedstocks. Important structural attributes of petroleum molecules (e.g. number of aromatic rings, number of naphthenic rings, number and length of aliphatic side chains) were assembled into molecules according to quantitative probability density functions for each attribute. The outcome was the atomic detail of a large ensemble of representative molecular structures from which both molecular and global product properties were deduced. Critical steps in the stochastic method were the generation of a chemical logic diagram, the compilation of cumulative probability functions for the structural attributes, stochastic sampling of each distribution, and the molecular construction. A general Monte Carlo algorithm provided an unbiased sampling of the probability density functions. The method was applied to three different complex petroleum feedstock fractions: an offshore California asphaltene, a Kern River heavy oil, and sour import heavy gas oil. The asphaltene example predicted the defining solubility protocol to within 1%. The heavy oil and gas oil simulations reproduced boiling point fractionation curves to within standard deviations of 20.8 and 25°C.

Computer generated reaction networks: on-the-fly calculation of species properties using computational quantum chemistry

Abstract An algorithm for the translation of the chemically rich two-dimensional molecular graph into a three-dimensional internal coordinate representation of the relative positions of the atoms in space is described. This enables on-the-fly calculations of species properties during the computer generation of reaction networks. The algorithm systematically species bond distances, bond angles and dihedral angles. Geometric parameters for atoms belonging to cycles required the specification of predefined geometrical templates. Templates for six-membered rings of the types aromatic, cyclohexadienyl, cyclohexenyl and cyclohexyl were therefore developed and optimized. An algorithm was also developed for the identification of tight cycles and their member atoms. This was critical to categorize atoms as noncyclic or cyclic.

Monte Carlo simulation of complex reaction systems : molecular structure and reactivity in modelling heavy oils

Abstract The problem of asphaltene precipitation during their thermolysis in a hypothetical inert oil motivated development of a stochastic model of asphaltene structure, reactions, and product identification. This amounted to stochastic construction of 10,000 asphaltene molecules as a representation of those found in a real resid. The reactions of each molecule, deduced from related model compound reaction pathways and kinetics, provided the reaction of the asphaltene representation after Monte Carlo simulation. A regular solution theory based thermodynamic model assembled the 10,000+ product molecules into global solubility-based product fractions. The simulation data were in good agreement with laboratory data for an off-shore California crude-derived asphaltene.

Hydrolysis in supercritical water: Solvent effects as a probe of the reaction mechanism

Abstract The extreme pressure dependence of the solvent properties—density, dielectric constant, and solubility parameter—has been exploited in the scrutiny of the mechanism of hydrolysis in supercritical water. Molecules containing the structural moiety of a saturated carbon attached to a heteroatom-containing leaving group undergo parallel pyrolysis and hydrolysis reactions in supercritical water. Herein, kinetics, labelling studies, and salt effects are used to assess likely mechanisms. Under isothermal conditions, the selectivity to the hydrolysis pathway increased with water density. Moreover, the rate constant for hydrolysis also increased with water density and with the addition of salts. An SN2 mechanism with H2O as the nucleophile appears to be the most likely candidate.

The effect of salts on hydrolysis in supercritical and near-critical water: Reactivity and availability☆

Abstract The effect of salt concentration on the rate of hydrolysis of dibenzyl ether (DBE) and benzyl phenyl amine (BPA) in supercritical (SC) and near-critical water was examined. The addition of salts to the reaction mixture increased the hydrolysis rate at low salt concentration, while having no effect on a back ground pyrolysis reaction rate. This is consistent with a polar hydrolysis reaction mechanism wherein the rate constant would be increased with increases in the solvent polarity. At higher salt concentrations, the hydrolysis rate reached a maximum then decreased, approaching the rate observed in the absence of salts. This is interpreted in terms of the solution phase behavior. Two extremes were considered — a macroscopic two-phase limit and a molecular, local concentration model.

Hydrogenation of polynuclear aromatic hydrocarbons. 2. quantitative structure/reactivity correlations

Abstract A consistent data base of reaction pathways, kinetics, and mechanisms for catalytic hydrogenation of one-, two-, three-, and four-fused aromatic ring compounds allowed for correlation of their Langmuir-Hinshelwood-Hougen-Watson (LHHW) rate law parameters with molecular structure. A total of 68 hydrogenation and dehydrogenation rate law parameters for 28 aromatic and hydroaromatic compounds were summarized into 7 parameters for quantitative structure/reactivity correlations (QS/RC) that characterized the associated set of series of homologous reactions, i.e. reaction families. Evaluation of the 28 LHHW adsorption constants was accomplished by imposing a correlation betwen the adsorption constant and the number of aromatic rings and the number of saturated carbons. Surface reaction rate constants correlated with the enthalpy of hydrogenation and the highest bond order in the aromatic ring being saturated. Semiempirical molecular orbital calculations provided acceptable estimates of the enthalpy of reaction, which, via compensation, provided estimates of the entropy of reaction, and thus equilibrium constants. The overall parity between measured parameters and those predicted by the 10 QS/RC parameters was very good, and allowed for 88% reduction in the number of parameters needed to model the saturation kinetics of polynuclear aromatic hydrocarbons.

Effect of pressure on the rate of butyronitrile hydrolysis in high-temperature water

Abstract The reactions of butyronitrile were investigated in high-temperature water at 330 °C and various pressures ranging from 128 to 2600 bar. Residence times ranged from 5 to 180 min. The product spectrum included butanamide, butyric acid and ammonia; gas formation was negligible. A four-step autocatalytic rate model incorporating the product acid as the catalytic species fitted the data well. This allowed the estimation of apparent activation volumes for each of the reaction steps. These were in turn decomposed into electrostatic and non-electrostatic contributions. At the lowest pressure studied (128 bar), the electrostatic activation volume ΔV es ≠ was twice as large as the hydrostatic activation volume ΔV hs ≠ . These activation volumes were comparable in value at all other pressures studied.

Catalytic cracking of alkylbenzenes: Modeling the reaction pathways and mechanisms

Abstract The catalytic cracking reaction pathways and mechanisms of 1-phenylhexane, 1-phenyloctane and 1-phenyldecane were studied. Experiments at 500°C with a rare earth Y (REY) catalyst indicated that the dominant reactions included dealkylation and cracking in the alkyl side chain. To describe these results, a mechanistic model of the catalytic cracking of 1-phenyloctane was developed using a novel mechanism-based lumping scheme that exploits the chemical similarities within reaction families. The formal application of 17 reaction family matrices, which correspond to 15 reaction family classes, to the matrix representations of the reactants and derived products generated the model. The reaction family concept was further exploited to constrain the kinetics within each reaction family to follow a quantitative structure/reactivity Polanyi relationship. Ultimately, eight Polanyi relationship parameters, one catalyst specific parameter and three coking/deactivation parameters were optimized using experimental data. The resulting model correlations were excellent, as the overall parity between experimental and model values was y Model =0.00131+0.994y Exp with a correlation coefficient of 0.998.

Asphaltene reaction pathways—v. Chemical and mathematical modeling

The pyrolysis of a generic, fully hydrocarbon asphaltene was simulated by combining model-compound-deduced thermolysis kinetics and pathways with asphaltene structural information. The latter was used in a Monte Carlo simulation of 10,000 prototype unit sheets, whose ensemble average conformed with the distribution and averages of structural probability density functions. Asphaltene reactions were thus mathematically equivalent to changes in the distributions and average of asphaltene structural information. The model-coumpound-based reaction pathways and kinetics of the unit sheet provided the temporal variations of the structural information, and the mathematical model was thus truly a priori with respect to asphaltene prolysis kinetics. Reaction products were assembled using the same Monte Carlo method as for the reactant but with reaction-altered distribution functions. Average structural parameters predicted by the model for the reactant asphaltene were consistent with literature values, and the temporal variation of the yields of solubility-based products fraction from simulated pyrolyses were in good accord with experimental result. These observations demonstrate that reaction engineering data derived from model compounds can be used to provide reasonable estimates of the reactivity of complex substrates such as petroleum asphaltenes.