Skip to content

Latest commit

 

History

History
127 lines (111 loc) · 4.35 KB

README.md

File metadata and controls

127 lines (111 loc) · 4.35 KB

Overview

Beyond-rpa performs electronic single-point energy calculations using the random-phase approximation (RPA) methods with higher-order corrections expressed via simplified coupled-cluster amplitudes. While beyond-rpa can be applied to any system, the algorithms and numerical thresholds have been hand-tuned for a numerically-stable calculation of 100s or 1000s of small energy terms in the many-body expansion of the crystal lattice energy:

  • two-body noncovalent interaction energies,
  • nonadditive energies of molecular trimers,
  • nonadditive energies of molecular tetramers.

You can use beyond-rpa like any other electronic structure software, but the simplest way to carry out automated fragment-based lattice energy calculations is to use it in combination with the mbe-automation companion program available at github.

Compilation

  1. Clone the repository
git clone https://github.com/modrzejewski/beyond-rpa.git
  1. Go to ./beyond-rpa/src and run the compilation script
./build.py -np 4 ifort-I64

The compiler flags, here it is ifort-I64, should correspond to one of the subdirectories in ./src/CompilerFlags. Choose a set appropriate for your system among the available sets or make your own. You can create your own subdirectory and add the command lines in the compiler and linker text files.

Usage

Running beyond-rpa

Execute the launcher script

./beyond-rpa/bin/run -nt 16 example.inp

where

  • -nt 16 specifies the number of concurrent threads used by the program. To get the best efficiency, the number of threads should be equal to the number of physical cores available for your calculations. For example, if you reserved a single node with shared memory and 2 CPUs, each having 18 physical cores, then the optimal setting is -nt 36.
  • example.inp is the input text file which contains the definition of the physical system and the requested level of theory. Example inputs are given below.

Examples

1. Noncovalent interaction energy of a water dimer

jobtype uks rpa
basis aug-cc-pVDZ
scf
 xcfunc HF
end

rpa
 accuracy default
 TheoryLevel JCTC2024
end

xyz
3 3
O      1.531750     0.005922    -0.120880
H      0.575968    -0.005249     0.024966
H      1.906249    -0.037561     0.763218
O     -1.396226    -0.004990     0.106766
H     -1.789372    -0.742283    -0.371009
H     -1.777037     0.777638    -0.304264
end

2. Nonadditive 3-body interaction energy of a formaldehyde trimer

jobtype uks rpa
basis aug-cc-pVDZ
scf
 xcfunc HF
end

rpa
 accuracy default
 TheoryLevel JCTC2024
end

xyz
4 4 4
O   2.074088   0.498855   1.635515
C   2.169454   0.154280   2.793163
H   2.970226  -0.523067   3.124239
H   1.470340   0.502963   3.564302
O   2.074088   0.498855   6.109515
C   2.169454   0.154280   7.267163
H   2.970226  -0.523067   7.598239
H   1.470340   0.502963   8.038302
O   2.074088   0.498855  10.583515
C   2.169454   0.154280  11.741163
H   2.970226  -0.523067  12.072239
H   1.470340   0.502963  12.512302
end

Authors

  • Marcin Modrzejewski (main author)

with contributions from:

  • Dominik Cieśliński (direct-ring amplitudes)
  • Aleksandra Tucholska (coupled-cluster 2-RDM)
  • Grzegorz Czekało (reference code for the quadratic corrections)
  • Krystyna Syty (decomposition of amplitudes)
  • Khanh Ngoc Pham (finding bugs)

Literature

You can use beyond-rpa to replicate the numerical results from the following publications:

  1. Cieśliński, D., Tucholska, A., Modrzejewski, M., J. Chem. Theory Comput. 19, 6619 (2023); doi: 10.1021/acs.jctc.3c00496
  2. Pham, K.N., Modrzejewski, M., Klimeš, J., J. Chem. Phys. 160, 224101 (2024); doi: 10.1063/5.0207090
  3. Pham, K.N., Modrzejewski, M., Klimeš, J., J. Chem. Phys. 158, 144119 (2023); doi: 10.1063/5.0142348
  4. Modrzejewski, M., Yourdkhani, S., Śmiga, Sz., Klimeš, J., J. Chem. Theory Comput. 17, 804 (2021); doi: 10.1021/acs.jctc.0c00966
  5. Modrzejewski, M., Yourdkhani, S., Klimeš, J., J. Chem. Theory Comput. 16, 427 (2020); doi: 10.1021/acs.jctc.9b00979

License

This program is freely available for use, modification, and integration into other software, provided that proper attribution is given to the original authors. Additionally, users are requested to cite the relevant publications that describe the methods implemented in this program. For full licensing details, please refer to LICENSE.txt.