Advanced search
1 file | 8.86 KB

Normal mode calculations with the QM/MM full Hessian and the Mobile Block Hessian (MBH) method

Author
Organization
Abstract
We have implemented the full Hessian evaluation in QM/MM simulations, as well as the approximate Mobile Block Hessian (MBH). The Hessian is the 3Nx3N matrix containing the second derivatives of the potential energy surface with respect to the 3N nuclear coordinates, and needs to be diagonalized when calculating the frequencies and normal modes. In extended systems, however, its calculation, storage and diagonalization is an expensive computational task. Note that even in case of a small QM region, the numerous derivatives of the QM/MM interaction terms still form a bottleneck in the frequency calculation. Recently, the Mobile Block Hessian (MBH) method was developed in order to reduce the dimensionality of the Hessian. The main concept is the introduction of blocks, which move as rigid bodies during the vibrational analysis. The blocks can also be linear or have atoms in common (leading to adjoined blocks). This block concept is now combined with the QM/MM scheme. The reduced computational cost opens the path to a broad range of applications of normal mode analysis.

Downloads

  • 2009 abstract cccc7.pdf
    • full text
    • |
    • open access
    • |
    • PDF
    • |
    • 8.86 KB

Citation

Please use this url to cite or link to this publication:

Chicago
Ghysels, An, Lee H. Woodcock, Yihan Shao, Bernard R. Brooks, Veronique Van Speybroeck, Dimitri Van Neck, and Michel Waroquier. 2009. “Normal Mode Calculations with the QM/MM Full Hessian and the Mobile Block Hessian (MBH) Method.” In Abstracts 7th Canadian Computational Chemistry Conference, 83.
APA
Ghysels, A., Woodcock, L. H., Shao, Y., Brooks, B. R., Van Speybroeck, V., Van Neck, D., & Waroquier, M. (2009). Normal mode calculations with the QM/MM full Hessian and the Mobile Block Hessian (MBH) method. Abstracts 7th Canadian Computational Chemistry Conference (p. 83). Presented at the 7th Canadian Computational Chemistry Conference.
Vancouver
1.
Ghysels A, Woodcock LH, Shao Y, Brooks BR, Van Speybroeck V, Van Neck D, et al. Normal mode calculations with the QM/MM full Hessian and the Mobile Block Hessian (MBH) method. Abstracts 7th Canadian Computational Chemistry Conference. 2009. p. 83.
MLA
Ghysels, An, Lee H. Woodcock, Yihan Shao, et al. “Normal Mode Calculations with the QM/MM Full Hessian and the Mobile Block Hessian (MBH) Method.” Abstracts 7th Canadian Computational Chemistry Conference. 2009. 83. Print.
@inproceedings{765735,
  abstract     = {We have implemented the full Hessian evaluation in QM/MM simulations, as well as the approximate Mobile Block Hessian (MBH). The Hessian is the 3Nx3N matrix containing the second derivatives of the potential energy surface with respect to the 3N nuclear coordinates, and needs to be diagonalized when calculating the frequencies and normal modes. In extended systems, however, its calculation, storage and diagonalization is an expensive computational task. Note that even in case of a small QM region, the numerous derivatives of the QM/MM interaction terms still form a bottleneck in the frequency calculation.

Recently, the Mobile Block Hessian (MBH) method was developed in order to reduce the dimensionality of the Hessian. The main concept is the introduction of blocks, which move as rigid bodies during the vibrational analysis. The blocks can also be linear or have atoms in common (leading to adjoined blocks). This block concept is now combined with the QM/MM scheme. The reduced computational cost opens the path to a broad range of applications of normal mode analysis.},
  author       = {Ghysels, An and Woodcock, Lee H. and Shao, Yihan and Brooks, Bernard R. and Van Speybroeck, Veronique and Van Neck, Dimitri and Waroquier, Michel},
  booktitle    = {Abstracts 7th Canadian Computational Chemistry Conference},
  language     = {eng},
  location     = {Halifax, Canada},
  title        = {Normal mode calculations with the QM/MM full Hessian and the Mobile Block Hessian (MBH) method},
  year         = {2009},
}