Search

A surface hopping method with a Monte Carlo procedure for deciding whether to hop at each step along the classical trajectories used in the semiclassical calculation is discussed. It is shown for a simple one-dimensional model problem that the numerical efficiency of the method can be improved by averaging over several copies of the sections of each trajectory that span the interaction regions. The use of Sobol sequences in the selection of the initial momentum for the trajectories is also explored. It is found that accurate results can be obtained with relatively small trajectory samples.

The accuracy of a semiclassical surface-hopping expansion of the time-independent wave function for problems in which the nonadiabatic coupling is peaked in the classically forbidden regions is studied numerically for a one-dimensional curve-crossing problem. This surface-hopping expansion has recently been shown to satisfy the Schrodinger equation to all orders in h and all orders in the nonadiabatic coupling. It has also been found to provide very accurate transition probabilities for problems in which the crossing points of the diabatic energy surfaces are classically allowed. In the numerical study reported here, transition probabilities are evaluated for energies well below the crossing point energy. It is found that the expansion provides accurate results for transition probabilities as small as 10(-11).

The quantum mechanical relaxation rate for a high-frequency vibrational mode is evaluated for a one-dimensional model system having two diatomic molecules involved in a collinear collision. The thermally averaged rate is obtained as an integral over energies for the relative translation of the two molecules. These calculations show that energies several times K(B)T make the largest contributions to the rate. Several orders of magnitude of cancellation due to phase interference is found in the evaluation of the coupling matrix elements between the initial and final states, and this is one of the main factors leading to the very small value for the relaxation rate. The region near the classical turning point in the relative translational motion of the colliding molecules dominates the calculation of the contribution to the rate at each energy. Calculations using low-order expansions of the translational potential energy and the interstate coupling about this turning point provide good approximations to the exact quantum mechanical rate. This suggests a possible method for performing calculations of the rate by means of realistic simulations of liquid systems.

A model for the plateau in the relaxation modulus is presented. This model treats the relaxation on successively longer length scales. The extent of relaxation in the strain-induced excess free energy for each length scale is evaluated by minimizing the free energy through motions on that length scale, while constraining the system such that there is no relaxation on all longer length scales. The extent of relaxation is determined by the number of other chain segments in contact with a relaxing chain segment. The residual excess free energy, which remains after the relaxation on a length scale, is further reduced by the subsequent relaxation on the longer length scales. This analysis predicts that the entanglement chain length for relaxation should scale as Ne ∼ ρ-2 b-3b1-3, where ρ is the monomer density, b is the monomer bond length, and b1 is the Kuhn length. Two methods for the evaluation of the time dependent relaxation modulus, which are based on this length scale dependent analysis of the relaxatio...

A semiclassical surface hopping Herman−Kluk (HK) propagator is developed for multisurface, multidimensional nonadiabatic problems. Its application for two-surface model problems is presented for two-dimensional collinear systems. Both noncrossing and crossing cases are investigated. The calculation on the basis of Monte Carlo method shows that the general multisurface semiclassical HK propagator method is efficient and accurate.

Indeno[1,2-b]pyrazin-2,3-diones have been identified as a novel series of potent ligands on the glycine site of the NMDA receptor. To improve their in vivo activities, an acetic acid-type side chain was introduced to the 5-position, giving water-soluble compounds when formulated as the sodium salt (>10 mg/mL). Introduction of a chlorine atom in the 8-position led to a dramatic improvement of anticonvulsant activity and this was surprising since this change did not improve binding affinity. A plausible explanation is a reduced recognition by a Na + ,K + -ATPase active transport system responsible for the excretion of these compounds from the brain and kidney. This promising new chemical series led to the optically active isomer (-)-10i (RPR 118723), a glycine/NMDA antagonist with nanomolar binding affinity and in vivo activity in animal model of convulsions and electrophysiology at doses in the range of 2-3 mg/kg following iv administration.

Superparamagnetic microspheres are prepared by incorporating nanometer-sized iron oxide crystals into micron-sized phenolic polymer particles. The synthesis of ferrite particles is conducted in the structured environment of water-in-oil microemulsions (reversed micelles). This is followed by the enzymatic polymerization of p-ethylphenol. The polymer precipitates in spherical morphologies and during precipitation incorporates ferrite nanocrystals into the polymeric matrix. SQUID (superconducting quantum interference device) generated magnetic hysteresis loops exhibit zero remanence and coercivity for the ferrite−polymer composite at room temperature, indicating the superparamagnetic nature of the composite. At very low temperatures, hysteretic effects, nonzero remanence, and coercivity are seen. Spin glass behavior is apparent at low temperatures from dc magnetic susceptibility measurements. Thermal and isothermal remanent magnetization measurements confirm the existence of the spin glass state. The compos...

A numerical simulation model is considered to study the relaxation of polymer melts within the context of the lateral motion model. This model assumes that the chain motion is primarily lateral motion along the backbones of neighboring chains. It is unclear whether a plateau would be observed in the relaxation modulus, if this is the mechanism of the chain dynamics. The simulation model considers the relaxation of a many-chain system. The beads of a given chain ere paired with the beeds of other chains. The two beads in a pair are constrained to move in unison. The chains are also allowed to slide along the backbones of the chains with which they have contacts through moves that interchange adjacent bead pairing

A model is presented for the lateral motion of linear polymer chains in a monodisperse melt. The model incorporates the constraining effect that neighboring chains have on lateral motion of a given chain, although this feature is not included in its full complexity. The cooperative motion of the various chains results in a lateral chain motion. The behavior which results from this lateral motion for the bead mean-squared displacement, terminal relaxation time, center of mass diffusion constant and the relaxation modulus are explored within this model and found in excellent agreement with experimental results

A model for stress relaxation of linear chain polymer melts in the linear viscoelastic regime is proposed. This model is developed in the context of the lateral diffusion model of polymer dynamics, which assumes that polymers diffuse in the melt by sliding laterally along the curvilinear contours of neighboring chains. Calculations are performed for monodisperse and bidisperse melts and the results are found to be in very good agreement with published experimental results

A model is presented for the dynamics of dense systems of linear monodisperse and polydisperse polymers which is based upon the assumption that the primary mechanism for chain motion is a sliding of each chain along the contours of neighboring chains. A time-dependent efffective single-bead friction coefficiant is derived which accounts for the constraining nature of the neighboring chains on the motion of a given chain.