[CP2K-user] Simulate reactions with MC
Yike Huang
ykhua... at gmail.com
Sat May 9 05:57:31 UTC 2020
Hi, cp2k users and experts,
Recently I read an interesting paper (doi: 10.1038/s41467-019-11856-9) where researchers find structure of oxidized nanoparticle by GEMC, which makes me curious because finding "correct" structure of nanoparticle is always a hard task. I want to find structure of oxidized Pt nanoparticle by GEMC too, and to compare energies of structures given by MD and GEMC.
However, it is the first time I use MC, so I read traditional MC and GEMC run_examples (ice_h and water liquid-vapor equilibrium) provided on cp2k website and learned fundamental concepts of MC. Unfortunately I am still confused about several basic points:
1. In examples, topology of H2O molecule is clearly defined, all parameters controlling motions can be understood with clear physical meaning. Note that examples provided don't contain any reactions, I can hardly write MC input files correctly to simulate even the simplest reaction, 2H→H2, where there're changes in the number of molecule and molecule type, because I have no idea about how to set parameters if I want to simulate reactions in MC.
2. I tried to turn off CONNECTIVITY and all parameters controlling motions of molecule in MC section, in order to avoid cp2k complains about inconsistency of the number and type of molecule. What I expect is that, if I don't define molecule, MC will directly sample/presample motions of atoms, evaluate energy and force by QS (just like if one deletes connectivity information in Gaussian, I turn off LBIAS indeed), but it seems not. cp2k told me MOLECULE must be explicitly defined, and parameters pertaining to molecule are compulsory.
3. What's worse, even for a one-atom toy system which I want to run it in MC, cp2k crashes after the first scf converges. cp2k complains segmentation fault, and there's no error printed out in stdout file. (I attach this input file at the end)
In brief, I think it must be of help if someone is glad to write a short example input illustrating how to simulate 2H→H2 reaction in MC. If so, I think my questions asked above will be perfectly solved.
Thanks in advance.
Huang Yike, PhD candidate.
Dalian Institute of Chemical Physics, CAS, China.
-------------toy-system input----------------
&FORCE_EVAL
METHOD Quickstep
&DFT
BASIS_SET_FILE_NAME BASIS_SET
POTENTIAL_FILE_NAME POTENTIAL
&MGRID
CUTOFF 280
&END MGRID
&SCF
SCF_GUESS ATOMIC
EPS_SCF 1.0E-5
MAX_SCF 50
&MIXING
METHOD BROYDEN_MIXING
NBUFFER 8
ALPHA 0.2
&END MIXING
&OT
MINIMIZER DIIS
PRECONDITIONER FULL_SINGLE_INVERSE
ALGORITHM IRAC
ENERGY_GAP 0.5
&END OT
&END SCF
&XC
&XC_FUNCTIONAL BLYP
&END XC_FUNCTIONAL
&END XC
UKS = .TRUE.
&END DFT
&SUBSYS
&CELL
ABC 10 10 10
&END CELL
&COORD
O 5 5 5
&END COORD
&KIND O
BASIS_SET 6-31G*
POTENTIAL GTH-BLYP-q6
&END KIND
&TOPOLOGY
&MOL_SET
&MOLECULE
NMOL 1
CONN_FILE_NAME topology_atom_O.psf
&END
&END
CONNECTIVITY MOL_SET
&END TOPOLOGY
&END SUBSYS
&END FORCE_EVAL
&GLOBAL
PROJECT ATOMS_MC
RUN_TYPE MC
PRINT_LEVEL LOW
&END GLOBAL
&MOTION
&MC
ENSEMBLE TRADITIONAL
TEMPERATURE 100.0
PRESSURE 1.013
IPRINT 1
LSTOP no
NMOVES 8
NSTEP 5
ETA 0.0
NSWAPMOVES 0
&MOVE_PROBABILITIES
PMSWAP 0.0
PMTRAION 0.30
PMTRANS 0.60
PMVOLUME 0.05
PMHMC 0.0
PMAVBMC 0.0
&MOL_PROBABILITIES
PMAVBMC_MOL 1.0
PMSWAP_MOL 1.0
PMTRAION_MOL 1.0
PMTRANS_MOL 1.0
PMROT_MOL 1.0
&END MOL_PROBABILITIES
&BOX_PROBABILITIES
PMVOL_BOX 1.0
PMHMC_BOX 1.0
&END BOX_PROBABILITIES
&END MOVE_PROBABILITIES
&MOVE_UPDATES
IUPTRANS 100
IUPVOLUME 100
&END MOVE_UPDATES
&MAX_DISPLACEMENTS
&MOL_DISPLACEMENTS
RMDIHEDRAL 3.0
RMANGLE 3.0
RMBOND 0.074
RMROT 26.0
RMTRANS 0.38
&END MOL_DISPLACEMENTS
&BOX_DISPLACEMENTS
RMVOLUME 100.5
&END BOX_DISPLACEMENTS
&END MAX_DISPLACEMENTS
&AVBMC
AVBMC_ATOM 1
AVBMC_RMIN 1.0
AVBMC_RMAX 5.0
PBIAS 0.5
&END AVBMC
&END MC
&END MOTION
More information about the CP2K-user
mailing list