Dear Dr. David Dubbeldam
I met something wrong when I studied the effect of flexibility of MOFs on gas separation. There was a segmentation fault occuring in my result when the pressure ranges from 0.1 bar to 10 bar.
_cell_length_a: 16.733000
_cell_length_b: 13.038000
_cell_length_c: 6.812000
_cell_length_alpha: 90.000000
_cell_length_beta: 90.000000
_cell_length_gamma: 90.000000
_symmetry_space_group_name_Hall: -I 2 2a found space group: 347
_symmetry_space_group_name_H-M: I m c m found space group: 347
_symmetry_Int_Tables_number: 343
space group found from symmetry elements: 347 (nr elements: 16)
End reading cif-file
Number of bonds: 0 0 1
Shift all potentials
Writing Crash-file!: 0
/var/spool/torque/mom_priv/jobs/794.localhost.SC: line 16: 19513 Segmentation fault (core dumped) $RASPA_DIR/bin/simulate simulation.input
The MOFs proposed in the study are MIL-53 series, including MIL-53(Al), MIL-53(Cr) and so on. Refering to the example (IRMOF-1) you gave to us, I defined the information of framework. Taking MIL-53 (Cr) as example,
the cif file we defined is:
data_MIL-53ht
_audit_creation_method RASPA-1.0
_audit_creation_date 2011-3-9
_audit_author_name '?'
_citation_author_name 'C. Serre, F. Millange, C. Thouvenot, M. Nogues, G. Marsolier, D. Louer, and G. Ferey'
_citation_title 'Very large breathing effect in the first nanoporous chromium(III)-based solids: MIL-53 or Cr-III(OH).{O2C-C6H4-CO2}.{HO2C-C6H4-CO2H}(x).H2Oy'
_citation_journal_abbrev 'J. Am. Chem. Soc.'
_citation_journal_volume 124
_citation_journal_number 45
_citation_page_first 13519
_citation_page_last 13526
_citation_year 2002
_cell_length_a 16.733
_cell_length_b 13.038
_cell_length_c 6.812
_cell_angle_alpha 90
_cell_angle_beta 90
_cell_angle_gamma 90
_cell_volume 1486.14
_symmetry_cell_setting orthorhombic
_symmetry_space_group_name_Hall '-I 2 2a'
_symmetry_space_group_name_H-M 'I m c m'
_symmetry_Int_Tables_number 74
loop_
_symmetry_equiv_pos_as_xyz
'x,y,z'
'-x+1/2,y,-z'
'x+1/2,-y,-z'
'-x,-y,z'
'-x,-y,-z'
'x+1/2,-y,z'
'-x+1/2,y,z'
'x,y,-z'
'x+1/2,y+1/2,z+1/2'
'-x,y+1/2,-z+1/2'
'x,-y+1/2,-z+1/2'
'-x+1/2,-y+1/2,z+1/2'
'-x+1/2,-y+1/2,-z+1/2'
'x,-y+1/2,z+1/2'
'-x,y+1/2,z+1/2'
'x+1/2,y+1/2,-z+1/2'
loop_
_atom_site_label
_atom_site_type_symbol
_atom_site_fract_x
_atom_site_fract_y
_atom_site_fract_z
_atom_site_charge
Cr1 Cr 0.25000 0.75000 0.25 0.574
O1 O 0.25000 0.686 0 -0.486
O2 O -0.168 0.161 0.167 -0.211
C3 C -0.033 0.032 0.179 -0.073
C2 C -0.066 0.063 0 -0.043
C1 C -0.137 0.132 0 0.083
Ho1 H 0.25000 0.611 0 0.121
H1 H -0.060 0.058 0.329 0.035
The structure as well as the corresponding atom names are shown in the attach file, named MIL-53(Cr)ht.bmp
The flexiblity of framework was defined as follow:
#CoreShells bond BondDipoles UreyBradley bend inv tors improper-torsion bond/bond bond/bend bend/bend stretch/torsion bend/torsion
0 5 0 0 7 0 8 3 0 0 0 0 0
#bond stretch atom n1-n2, equilibrium distance, bondforce-constant
C3 H1 HARMONIC_BOND 366001.13136396 0.95
C3 C3 HARMONIC_BOND 483413.91047488 1.36
C2 C3 HARMONIC_BOND 483413.91047488 1.36
C1 C2 HARMONIC_BOND 353750.919316375 1.42
O2 C1 HARMONIC_BOND 543840.64928424 1.25
#bond bending atom n1-n2-n3, equilibrium angle, bondforce-constant
C1 C2 C3 HARMONIC_BEND 34926.5543205787 120.0
C2 C3 H1 HARMONIC_BEND 37263.15559911 120.0
C3 C3 H1 HARMONIC_BEND 37263.15559911 120.0
C3 C2 C3 HARMONIC_BEND 90640.10821404 120.0
C3 C3 C2 HARMONIC_BEND 90640.10821404 120.0
O2 C1 O2 HARMONIC_BEND 135960.162321060 130.0
O2 C1 C2 HARMONIC_BEND 54882.4848123699 115.0
#torsion atom n1-n2-n3-n4,
O2 C1 C2 C3 TRAPPE_DIHEDRAL 0.0 0.0 1258.890391861 0.0
C1 C2 C3 H1 TRAPPE_DIHEDRAL 0.0 0.0 1510.668470234 0.0
C1 C2 C3 C3 TRAPPE_DIHEDRAL 0.0 0.0 1510.668470234 0.0
H1 C3 C3 H1 TRAPPE_DIHEDRAL 0.0 0.0 1510.668470234 0.0
C2 C3 C3 H1 TRAPPE_DIHEDRAL 0.0 0.0 1510.668470234 0.0
C2 C3 C3 C2 TRAPPE_DIHEDRAL 0.0 0.0 1510.668470234 0.0
H1 C3 C2 C3 TRAPPE_DIHEDRAL 0.0 0.0 1510.668470234 0.0
C3 C2 C3 C3 TRAPPE_DIHEDRAL 0.0 0.0 1510.668470234 0.0
#improper torsion atom n1-n2-n3-n4,
C2 C1 O2 O2 TRAPPE_IMPROPER_DIHEDRAL 0.0 0.0 5035.561567446 0.0
C3 C2 C3 C1 TRAPPE_IMPROPER_DIHEDRAL 0.0 0.0 5035.561567446 0.0
C2 C3 C3 H1 TRAPPE_IMPROPER_DIHEDRAL 0.0 0.0 186.3157779955 0.0
There is a question here: Is it necessary to define the bond, bend, torsion information between Cr(III) and surrounding O or C? And how to obtain these information and change them into the form that RASPA can identify
The force_field_mixing_rules.def file was defined as follow:
# general rule for shifted vs truncated
shifted
# general rule tailcorrections
no
# number of defined interactions
28
# type interaction
Al_ Lennard-jones 254.09 4.01
Br_ Lennard-jones 126.29 3.73
C_ Lennard-jones 52.83 3.43
Ca_ Lennard-jones 119.75 3.03
Cl_ Lennard-jones 114.21 3.52
Cr_ Lennard-jones 7.55 2.69
F_ Lennard-jones 25.16 3
Fe_ Lennard-jones 6.54 2.59
Ga_ Lennard-jones 208.81 3.9
H_ Lennard-jones 22.14 2.57
In_ Lennard-jones 301.39 3.98
N_ Lennard-jones 34.72 3.26
O_ Lennard-jones 30.19 3.12
Os_ Lennard-jones 18.62 2.78
V_ Lennard-jones 8.05 2.8
CH4_sp3 lennard-jones 148.0 3.72
CH3_sp3 lennard-jones 98 3.76
CH2_sp3 lennard-jones 46 3.96
CH_sp3 lennard-jones 17.0 4.67
C_sp3 lennard-jones 0.8 6.38
O_co2 lennard-jones 79 3.05
C_co2 lennard-jones 27.0 2.8
Ow lennard-jones 81.9 3.16
Hw lennard-jones 0 0
Lw lennard-jones 0 0
S_S lennard-jones 122 3.6
H_S lennard-jones 50 2.5
M_S lennard-jones 0 0
# general mixing rule for Lennard-Jones
Lorentz-Berthelot
The pseudo_atoms.def file was defined as follow:
#number of pseudo atoms
23
#type print as chem oxidation mass charge polarization B-factor radii connectivity anisotropic anisotropic-type tinker-type
Cr1 yes Zn Zn 0 65.37 0.574 0 1.0 1.6 0 0 relative 0
O1 yes O O 0 15.9994 -0.486 0 1.0 0.68 2 0 relative 0
O2 yes O O 0 15.9994 -0.211 0 1.0 0.68 2 0 relative 0
C3 yes C C 0 12.0107 -0.073 0 1.0 0.720 0 0 relative 0
C2 yes C C 0 12.0107 -0.043 0 1.0 0.720 0 0 relative 0
C1 yes C C 0 12.0107 0.083 0 1.0 0.720 0 0 relative 0
Ho1 yes H H 0 1.00794 0.121 0 1.0 0.320 0 0 relative 0
H1 yes H H 0 1.00794 0.035 0 1.0 0.320 0 0 relative 0
He yes He He 0 4.002602 0 0 1.0 1 0 0 relative 0
CH4_sp3 yes C C 0 16.04246 0 0 1.0 1 0 0 relative 0
CH3_sp3 yes C C 0 15.03452 0 0 1.0 1 0 0 relative 0
CH2_sp3 yes C C 0 14.02658 0 0 1.0 1 0 0 relative 0
CH_sp3 yes C C 0 13.01864 0 0 1.0 1 0 0 relative 0
CH2_sp2 yes C C 0 14.02658 0 0 1.0 1 0 0 relative 0
C_sp3 yes C C 0 12 0 0 1.0 1 0 0 relative 0
C_co2 yes C C 0 12.0107 0.7 1.508 1.0 0.720 0 0 relative 0
O_co2 yes O O 0 15.9994 -0.35 0.9475 1.0 0.68 0 0 relative 0
Ow yes O O 0 15.9994 0 0 1.0 0.5 2 0 relative 0
Hw yes H H 0 1.00794 0.524 0 1.0 1 1 0 relative 0
Lw no L - 0 0 -1.048 0 1.0 1 1 0 relative 0
S_S yes S S 0 32.06 0 0 1.0 1 0 0 relative 0
H_S yes H H 0 1 0.21 0 1.0 1 0 0 relative 0
M_S no M - 0 0 -0.42 0 1.0 1 0 0 relative 0
No extra force field defined in the force_field.def
As for the input file, we used the hybrid MCMD method to calculate the adsorption capacity of CH4 in MIL-53(Cr)
SimulationType MonteCarlo
NumberOfCycles 100000
NumberOfInitializationCycles 100000
PrintEvery 5000
RestartFile no
ContinueAfterCrash no
WriteBinaryRestartFileEvery 5000
ChargeMethod Ewald
Forcefield MIL-53-Cr-ht-pri
CutOffVDW 12.5
RemoveAtomNumberCodeFromLabel yes
Framework 0
FrameworkName MIL-53-Cr-ht-pri
ChargeFromChargeEquilibration yes
UnitCells 2 2 4
HeliumVoidFraction 0.5093
FrameworkDefinitions MIL-53-Cr-ht-pri
ExternalTemperature 298.15
ExternalPressure 61000
FlexibleFramework yes
HybridMCMDMoveProbability 1.0
Movies yes
WriteMoviesEvery 5000
Component 0 MoleculeName CH4
StartingBead 0
MoleculeDefinition gwq_trappe
TranslationProbability 1.0
RotationProbability 1.0
SwapProbability 1.0
RegrowProbability 1.0
IdealGasRosenbluthWeight 1.0
CreateNumberOfMolecules 0
I would be very very grateful if you can give me some suggestions. ;D ;D ;D ;D ;D ;D
In addition, I'd like to ask you some questions:
In the first example listed in the Advanced Example, you said the adsorption of CO2 in a totally flexible IRMOF-1 using a hybrid MC and MD simulation in μVT ensemble . However, you listed another example to illustrate the CO2 adsorption in flexible IRMOF-1 in osmotic ensemble.
1. If I want to obtain the gas separation capability of a flexible MOF, which method is better for us to use? In which condition should I used the instruction of the first example, and in which condition I should use the instruction in the second example.
2. According to Sven M. J. Rogge (Adv. Theory Simul. 2019, 2, 1800177, doi: 10.1002/adts.201800177), the hybrid MCMD is the best way to predict the adsorption behaviors in a flexible MOF in a restricted osmotic ensemble, which seems to be the combination of the first and second example (Scheme 4). In that work, he did a MD at the very beginning, but in the case of hybridMCMD method in RASPA, it seems no MD relative defination involved. So, how does it manange to do MD simulation.
[/color]
I am really looking forward to your respond ;D ;D ;D
Sincere
Monica
I had the same problem. I wonder if you have solved this problem?
In case of a segmentation fault, run it under a debugger (like gdb, or lldb) to figure out at which line in the code it crashes. That can help to pinpoint what the problem might be.
About modeling flexibility, different types of flexibility requires different treatments. So it depends on the situation. For flexibility around equilibrium positions simple atom-displacements might suffice. For large scale movements you will need to use MC/MD-hybride moves since it is collective motion on longer time scales.