skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Motif based Hessian matrixfor ab initio geometry optimization ofnanostructures

Abstract

A simple method to estimate the atomic degree Hessian matrixof a nanosystem is presented. The estimated Hessian matrix, based on themotif decomposition of the nanosystem, can be used to accelerate abinitio atomic relaxations with speedups of 2 to 4 depending on the sizeof the system. In addition, the programing implementation for using thismethod in a standard ab initio package is trivial.

Authors:
; ;
Publication Date:
Research Org.:
Ernest Orlando Lawrence Berkeley NationalLaboratory, Berkeley, CA (US)
Sponsoring Org.:
USDOE Director. Office of Science. Advanced ScientificComputing Research
OSTI Identifier:
926286
Report Number(s):
LBNL-59974
R&D Project: K11116; BnR: KJ0101030; TRN: US200807%%623
DOE Contract Number:
DE-AC02-05CH11231
Resource Type:
Journal Article
Resource Relation:
Journal Name: Physical Review B; Journal Volume: 73; Journal Issue: 19; Related Information: Journal Publication Date: 05/2006
Country of Publication:
United States
Language:
English
Subject:
36; GEOMETRY; IMPLEMENTATION; NANOSTRUCTURES; OPTIMIZATION; geometry optimization motif Hessian matrix preconditioned CGnanostructures convergence acceleration atomic relaxation

Citation Formats

Zhao, Zhengji, Wang, Lin-Wang, and Meza, Juan. Motif based Hessian matrixfor ab initio geometry optimization ofnanostructures. United States: N. p., 2006. Web. doi:10.1103/PhysRevB.73.193309.
Zhao, Zhengji, Wang, Lin-Wang, & Meza, Juan. Motif based Hessian matrixfor ab initio geometry optimization ofnanostructures. United States. doi:10.1103/PhysRevB.73.193309.
Zhao, Zhengji, Wang, Lin-Wang, and Meza, Juan. Wed . "Motif based Hessian matrixfor ab initio geometry optimization ofnanostructures". United States. doi:10.1103/PhysRevB.73.193309. https://www.osti.gov/servlets/purl/926286.
@article{osti_926286,
title = {Motif based Hessian matrixfor ab initio geometry optimization ofnanostructures},
author = {Zhao, Zhengji and Wang, Lin-Wang and Meza, Juan},
abstractNote = {A simple method to estimate the atomic degree Hessian matrixof a nanosystem is presented. The estimated Hessian matrix, based on themotif decomposition of the nanosystem, can be used to accelerate abinitio atomic relaxations with speedups of 2 to 4 depending on the sizeof the system. In addition, the programing implementation for using thismethod in a standard ab initio package is trivial.},
doi = {10.1103/PhysRevB.73.193309},
journal = {Physical Review B},
number = 19,
volume = 73,
place = {United States},
year = {Wed Apr 05 00:00:00 EDT 2006},
month = {Wed Apr 05 00:00:00 EDT 2006}
}
  • The cost of calculating nuclear hessians, either analytically or by finite difference methods, during the course of quantum chemical analyses can be prohibitive for systems containing hundreds of atoms. In many applications, though, only a few eigenvalues and eigenvectors, and not the full hessian, are required. For instance, the lowest one or two eigenvalues of the full hessian are sufficient to characterize a stationary point as a minimum or a transition state (TS), respectively. We describe here a method that can eliminate the need for hessian calculations for both the characterization of stationary points as well as searches for saddlemore » points. A finite differences implementation of the Davidson method that uses only first derivatives of the energy to calculate the lowest eigenvalues and eigenvectors of the hessian is discussed. This method can be implemented in conjunction with geometry optimization methods such as partitioned-rational function optimization (P-RFO) to characterize stationary points on the potential energy surface. With equal ease, it can be combined with interpolation methods that determine TS guess structures, such as the freezing string method, to generate approximate hessian matrices in lieu of full hessians as input to P-RFO for TS optimization. This approach is shown to achieve significant cost savings relative to exact hessian calculation when applied to both stationary point characterization as well as TS optimization. The basic reason is that the present approach scales one power of system size lower since the rate of convergence is approximately independent of the size of the system. Therefore, the finite-difference Davidson method is a viable alternative to full hessian calculation for stationary point characterization and TS search particularly when analytical hessians are not available or require substantial computational effort.« less
  • In this work we examine the consequences of incorporating the ab-initio derived monomer potential energy surface and non-linear dipole surface of Partridge and Schwenke [J. Chem. Phys. 106, 4618 (1997)] into the previously developed TTM2-R model of Burnham et al. [J. Chem. Phys. xx. yyyy (2001)] in order to develop a new, all-atom polarizable, flexible model for water (TTM2-F). We found that the use of the non-linear dipole surface is essential in modeling the change in the internal geometry of interacting water molecules and, in particular, the increase in the internal H-O-H bend angle with cluster size. This is themore » first demonstration of a flexible model which shows an increase in the bending angle in clusters. An explanation for this behavior is presented using the concept of geometric polarizabilities . The model furthermore reproduces the n=2-6 cluster binding energies to within an RMS deviation of 0.05 kcal/mol per hydrogen bond with respect to the MP2 complete basis set estimates.« less
  • The optimized geometry, harmonic vibrational frequencies, and infrared absorption intensities of the lithium trifluoromethanesulfonate (triflate) ion pair, CF[sub 3]SO[sub 3]-Li have been investigated using the ab initio self-consistent Hartree-Fock and correlated second-order Moller-Plesset perturbation theory with the 6-31G* and lower basis sets. In the optimized structure the lithium cation is bound to two of the oxygens of the SO[sub 3] group forming a bidentate complex with C[sub s] symmetry. A local minimum with a monodentate structure was obtained in the HF/3-21G* calculations. The energy difference between the mono- and bidentate structures of the complex is predicted to be nearly 39more » kJmol[sup [minus]1] in this basis. A splitting of 230 and 158 cm[sup [minus]1] is obtained for the antisymmetric SO[sub 3] stretching for the bi- and monodentate coordination of the lithium cation with the free anion, respectively. The infrared spectrum of lithium triflate in poly(propylene oxide) shows a splitting of 43 cm[sup [minus]1]. The strong interaction of the metal cation with the anion in the 1:1 complex thus overemphasizes the [open quotes]splitting behavior[close quotes] observed for lithium triflate dissolved in polymers. In the bidentate (MP2/6-31G*) complex the symmetric SO[sub 3] stretching shows a downshift of 38 cm[sup [minus]1], in contrast to an upshift of 47 cm[sup [minus]1] for the monodentate complex. 22 refs., 2 figs., 6 tabs.« less
  • Based on {ital ab initio} quantum-chemical methods, accurate calculations on small boron clusters B{sub n} (n=2{endash}14) were carried out to determine their electronic and geometric structures. The geometry optimization with a linear search of local minima on the potential-energy surface was performed using analytical gradients in the framework of the restricted Hartree-Fock self-consistent-field approach. Most of the final structures of the boron clusters (n{gt}9) are composed of two fundamental units: either of hexagonal or of pentagonal pyramids. Proposing an {open_quotes}Aufbau principle{close_quotes} one can easily construct various highly stable boron species. The resulting quasiplanar and convex structures can be considered asmore » fragments of planar surfaces and as segments of nanotubes or hollow spheres, respectively. {copyright} {ital 1997} {ital The American Physical Society}« less
  • We investigate the liquid structure, ion hydration, and some thermodynamic properties associated with the rigid geometry approximation to water by applying ab initio molecular dynamics simulations (AIMD) with the Perdew-Burke-Ernzerhof (PBE) exchange-correlation functional at T = 320 K. We vary the rigid water geometry in order to locate a class of practical water models that yield reasonable liquid structure and dynamics, and to examine the progression of AIMD-predicted water behavior as the OH bond length varies. Water constrained at the optimal PBE gas phase geometry yields reasonable pair correlation functions. The predicted liquid phase pressure, however, is large ({approx}8.0 kbar).more » Although the O-H bond in water should elongate when transferred from gas to the condensed phase, when it is constrained to 0.02, or even just 0.01 {angstrom} longer than the optimal gas phase value, liquid water is predicted to be substantially overstructured compared to experiments. Zero temperature calculations of the thermodynamic properties of cubic ice underscore the sensitivity toward small variations in the O-H bond length. We examine the hydration structures of potassium, chloride, and formate ions in one rigid PBE water model. The results are in reasonable agreement with unconstrained AIMD simulations.« less