Cyclic anhydride ring opening reactions: theory and application.

The development of a zero net shrinkage dental restorative material based upon a polymer-bioactive-glass composite requires a second-phase material that expands. This study details the mechanisms of organic cyclic anhydride ring expansion via hydrolysis. Six cyclic anhydrides were used to represent potential side groups, each of which could be an expanding phase or component. Maleic, 4META, tetrahydrophthalic, norbornene, itaconic, and succinic anhydrides were modeled using the Austin method (AM1), a semi-empirical molecular orbital method. The reaction pathways were determined for the anhydride ring opening reaction to form an acid for each case. The activation barriers (Ea) for the ring openings were found from the transition state geometries wherein only one imaginary eigen value in the vibration spectrum existed (a true saddle point). In each case the reaction pathway included the hydrogen bonding of a H2O molecule to the ring, weakening of the C-O bridging bonds of the ring, and, finally, the dissociation of the H2O, forming two carboxyl groups and opening the ring. The activation for the ring openings are +34.3, +36.9, +40.6, +43.1, +45.9, and +47.7 kcal/mol, respectively. The volumetric expansion of the anhydrides was estimated based upon the dilation of C-O-C atomic distances. The dimensional change was found to be 24.0%, 24.0%, 19.1%, 20.3%, 20.8%, and 17.9% for the anhydride rings, respectively. Finally, it was found that a linear correlation exists between the cyclic anhydride C-O asymmetric rocking (as-v) vibration and the activation energy (Ea) for hydrolysis to an acid. This may be used as an experimental indicator of a cyclic anhydride's activity.

Molecular orbital models of ring expansion mechanisms in the silica-carbon monoxide system.

The development of a zero net shrinkage dental restorative material based upon a polymer-bioactive glass composite requires a second-phase material that expands. This study details the mechanisms of silica ring expansion by reaction with carbon monoxide. Carbon monoxide was used as a model adduct to represent potentially active sites on the polymer phase of the dental restorative. Silica rings were used to model the bioactive-glass phase of the composite. The 3-, 4-, 5-, and 6-"member" silica rings have been modeled using the Austin Method (AM1) semi-empirical molecular orbital calculations. The reaction pathways were determined for carbon monoxide (CO) reaction addition to each of the rings. The activation barriers (Ea) for the ring expansions were determined from the transition state geometries wherein only one imaginary eigenvalue in the vibration spectrum existed (a true saddle point). In each case the reaction pathway included the hydrogen bonding of CO with a silicon, exothermic pentacoordinate bonding to silicon by the CO and weakening of the Si-O bridging bonds of the ring, and, finally, the incorporation of CO into the ring, forming a silica-carbonate ring. The activation for the ring expansions are +4.3, +6.1, +7.0, and -2.9 Kcal/mol for 3-, 4-, 5-, and 6-"member" silica rings, respectively. The volumetric expansion of the silica was estimated based upon the dilation of adjacent silicon-silicon atomic distances. The dimensional change was calculated to be 3.9%, 21.3%, 19.4%, and 24.2% for 3-, 4-, 5-, and 6-membered silica-carbonate rings, respectively.

Theoretical analysis of hydrolysis of polydimethylsiloxane (PDMS).

Silicones (polydimethylsiloxane, PDMS) are the materials currently used in most breast implants. ICP and FTIR analysis of the tissue capsule around aged breast implants and in vitro models show that Si-containing material is leaking from the PDMS implants. In this study, the hydrolysis of PDMS has been theoretically modeled using a semiempirical quantum mechanical method called AM1. The activation barrier for removing a methanol monomer was found to be +82 Kcal/mol while the removal of a methane monomer was +41 Kcal/mol. Using the same AM1 method, hydrolysis of the identical to Si-O-Si identical to bond also has been modeled for pentasilicic acid and, in this study, for 1,1,3,3,-fetramethyldisiloxane-1,3-diol. The barrier to the removal of a silicon-containing tetrahedron for both studies was found to be +27 Kcal/mol. This is approximately one and a half times smaller than the energy of that needed to remove a methyl group. The pentacoordinated silicon-activated transition state for hydrolysis of PDMS may provide an energetically favorable pathway for development of a surface that will enhance chemisorption of charged protein molecules, and such a pathway may show up in NMR studies of the hydrolysis of PDMS.

Molecular modeling study of adsorption of poly-L-lysine onto silica glass.

A research program was initiated with both experimental and computational chemistry based molecular modeling components to investigate specific amino acid-surface interactions. The experimental portion of this study, with details reported elsewhere, investigated the adsorption of selected molecular weights of poly(L-lysine) onto silica glass microspheres with the adsorption enthalpy per adsorbed mer determined to be -0.23 +/- 0.13 kcal/mol (mean +/- 95% confidence interval). Molecular modeling of this system was then conducted using two approaches: an AM1 semiempirical molecular orbital method to predict L-lysine/glass interaction energy and an MM2 molecular mechanics method to investigate the structural configuration for poly(L-lysine). The modeling predicted a minimum energy configuration of a rotational backbone structure for poly(L-lysine) with approximately one full rotation occurring about every 8 mers, and that the amine side chains of the L-lysine will hydrogen bond with the silica surface with an average adsorption energy of approximately -0.34 kcal/mol/mer. The molecular modeling results are in good agreement with the experimentally measured value and provide insights into possible molecular-level behavior which would be very difficult to determine by experimental analyses alone. This work demonstrates the use of molecular modeling in conjunction with experimental studies to investigate complex molecular interactions.

AM-1 molecular orbital calculations of silica-alanine-nitrogen interaction.

Chemical binding of proteins with bioactive surfaces is modeled using a semi-empirical molecular orbital theory (AM-1). The model calculates the optimized molecular structures of an amino acid (L-alanine) interacting with a cyclotetrasiloxane silica cluster (a four-membered hydrated silica ring). The calculated heats of formation for various orientations of alanine show +5 kcal/mol difference for binding via the -NH2 group following a condensation reaction with a pentacoordinate Si intermediate. Hydrogen bonding of the alanine via the -COOH group occurs with +13 to +15 kcal/mole differences in heats of formation and imposes a highly specific geometric orientation on the amino acid. Association of a diatomic N2 molecule with the silica cluster before interaction with alanine inhibits formation of an intermolecular bond, as is observed experimentally in studies of silica-alanine epitaxy.