[CP2K:5012] Periodic RESP charges for paracetamol form I crystal
Dorothea Golze
dorothe... at googlemail.com
Sun Mar 2 21:35:55 UTC 2014
Hi Mike,
there are a few things you should check on
1) Search for the Root-mean-square (RMS) and Relative root-mean-square
(RRMS) error in the output file. They give you hints on the quality of the
fit. Check J . Phys. Chem. 1993,97, 10269-10280
to get an idea in which range the values should be for ESP.
2) You can directly compare the potentials (original and fitted one).
Enable in the respective print sections &V_HARTREE_CUBE and &V_RESP_CUBE
Sometimes the fitting is good, but charges are unphysical. This can
theoretically happen with ESP;
To make the charges "more physical" you have to employ restraints
(restrained ESP -> RESP)
3) check if the fitting points were collected as you intended by switching
on &COORD_FIT_POINTS in the print section. Fix the first line in the
generated .xyz (number atoms + number fitting points) to view in vmd
4) if the collection of fitting points is correct and the fitting is bad
according to the RRMS values, please send me your latest resp input file;
It should work with monoclinic cells. However, I never explicitly tested it
(only orthorhombic and hexagonal cells)
5) Sometimes the potential is "too complicated" and the fitting is just
bad. This can also happen, even for small molecules, see J . Phys. Chem.
1993,97, 10269-10280
Kind regards,
Dorothea
2014-03-02 20:35 GMT+01:00 mbellucci <mbell... at gmail.com>:
>
>
>
>
> Hi Dorothea,
>
> Thanks for the reply. I tried what you suggested, i.e. using the
> NONPERIODIC_SYS keyword instead of the PERIODIC_SYS keyword, but it only
> seems to give physically meaningful charges when I make the cell dimensions
> very large (25.0 25.0 25.0 cubic box) so that all 4 molecules that make up
> the unit cell are completely within the box. However, these very large
> cubic cell dimensions don't correspond to the actual dimensions of the
> monoclinic unit cell, and so, the periodicity with this large cell doesn't
> correspond the periodicity of the actual crystal. When I try to use the
> cell dimensions of the unit cell (see previously posted input file) with
> the NONPERIODIC_SYS keyword, once again I end up with non-physical RESP
> charges. I've also tried a range of different values for RMIN and RMAX in
> the NONPERIODIC_SYS section, but this doesn't seem to help very much.
> Perhaps I am missing something simple, and so any suggestions would be
> appreciated.
>
> Mike
>
>
>
> On Saturday, March 1, 2014 3:19:58 PM UTC-5, Dorothea Golze wrote:
>
>> Hi,
>>
>> the keywords "RANGE" and "LENGTH" make no sense in your case. They are
>> for slab-like periodic systems. The section "PERIODIC_SYS" and
>> "NONPERIODIC_SYS" refer to the way the fitting points are collected.
>> PERIODIC_SYS sets the parameter for collecting fitting points for
>> slab-like systems. In "NONPERIODC_SYS" fitting points are collected in
>> spheres around the atoms. That is what you need here. Whether a periodic
>> RESP fit is required or not is solely decided on the usage of a periodic or
>> non-periodic Poisson solver.
>> I admit the names of these subsection are maybe a little bit misleading.
>> I should change them....
>> There were questions about the periodic RESP charges before, see
>> my last post there:
>> https://groups.google.com/forum/#!msg/cp2k/G3THGbvvIcU/CgGZxvtbGrsJ
>>
>> Kind regards,
>> Dorothea
>>
>>
>> 2014-03-01 21:02 GMT+01:00 mbellucci <mbel... at gmail.com>:
>>
>>>
>>> Dear CP2K Users,
>>>
>>> I am trying to calculate the RESP charges for a unit cell of the form I
>>> crystal of paracetamol using periodic boundary conditions using version 2.5
>>> of CP2K. I have tried many different things but in every calculation, the
>>> RESP charges seem to be unphysical, i.e. the magnitudes of the charges seem
>>> to very large and most have the wrong sign. The things I have tried in
>>> order to fix this include:
>>>
>>> increasing the cutoff
>>> changing the functional
>>> increasing the size of the basis set
>>> changing the values of the LENGTH and RANGE keywords in the PERIODIC_SYS
>>> section of RESP
>>> changing the SURF_DIRECTION in the PERIODIC_SYS section of RESP
>>> I've also experimented with the keywords WIDTH,
>>> RESTRAIN_HEAVIES_STRENGTH, and RESTRAIN_HEAVIES_TO_ZERO in RESP
>>>
>>> Nothing seems to work for the RESP charges. However, when I turn off
>>> the periodic boundary conditions the calculations produce RESP charges that
>>> make physical sense. Also, for calculations with or without periodic
>>> boundary conditions, the DDAP charges make physical sense, but I'm hesitant
>>> to use them in MD simulations since they are less frequently used in MD
>>> simulations in the literature. If anyone could help identify an error I am
>>> making when calculating these RESP charges, I would greatly appreciate it!
>>> Also, if anyone could comment on the reliability of DDAP charges in
>>> general, I would appreciate it as well.
>>>
>>> For the calculation, the coordinates of the form I crystal as well as
>>> the dimensions of the unit cell were obtained from the cambridge structural
>>> database. The input file I'm using, the outputted RESP charges, and the
>>> outputted DDAP charges are the following:
>>>
>>> &GLOBAL
>>> PROJECT form_1
>>> RUN_TYPE ENERGY_FORCE
>>> PRINT_LEVEL MEDIUM
>>> WALLTIME 86400
>>> &END GLOBAL
>>>
>>> &FORCE_EVAL
>>> METHOD Quickstep
>>> &DFT
>>> BASIS_SET_FILE_NAME EMSL_BASIS_SETS
>>> POTENTIAL_FILE_NAME POTENTIAL
>>> CHARGE 0
>>> MULTIPLICITY 1
>>> &MGRID
>>> CUTOFF 300
>>> REL_CUTOFF 50
>>> &END MGRID
>>>
>>> &QS
>>> METHOD GAPW
>>> EPS_DEFAULT 1.0E-12
>>> MAP_CONSISTENT TRUE
>>> EXTRAPOLATION ASPC
>>> EXTRAPOLATION_ORDER 3
>>> &END QS
>>>
>>> &SCF
>>> MAX_SCF 500
>>> SCF_GUESS ATOMIC
>>> EPS_SCF 1.0E-7
>>> &OT ON
>>> MINIMIZER DIIS
>>> PRECONDITIONER FULL_ALL
>>> ENERGY_GAP 0.001
>>> &END OT
>>>
>>> &OUTER_SCF
>>> EPS_SCF 1.0E-5
>>> MAX_SCF 100
>>> &END OUTER_SCF
>>> &END SCF
>>>
>>> &XC
>>> &XC_FUNCTIONAL PBE
>>> &END XC_FUNCTIONAL
>>> &END XC
>>> &END DFT
>>>
>>>
>>> &SUBSYS
>>> &CELL
>>> ABC 12.93 9.40 7.10
>>> ALPHA_BETA_GAMMA 90.00 115.90 90.00
>>> PERIODIC XYZ
>>> &END CELL
>>>
>>> &COORD
>>> C 9.510000 3.334000 5.417000
>>> C 10.770000 3.348000 4.827000
>>> C 11.771000 2.497000 5.288000
>>> C 11.518000 1.636000 6.343000
>>> C 10.266000 1.639000 6.953000
>>> C 9.279000 2.483000 6.489000
>>> C 8.231000 4.763000 3.830000
>>> C 6.944000 5.542000 3.720000
>>> H 10.930000 3.957000 4.145000
>>> H 12.675000 2.557000 4.918000
>>> H 10.101000 1.062000 7.690000
>>> H 8.407000 2.482000 6.891000
>>> H 12.348000 0.517000 7.562000
>>> H 7.724000 4.239000 5.582000
>>> H 6.614000 5.480000 2.836000
>>> H 6.512000 5.574000 4.484000
>>> H 7.181000 6.589000 3.692000
>>> N 8.426000 4.170000 5.017000
>>> O 12.536000 0.800000 6.745000
>>> O 9.035000 4.678000 2.901000
>>> C 6.784000 8.034000 0.970000
>>> C 5.524000 8.048000 1.560000
>>> C 4.523000 7.197000 1.099000
>>> C 4.776000 6.336000 0.044000
>>> C 6.028000 6.339000 -0.567000
>>> C 7.015000 7.183000 -0.102000
>>> C 8.063000 9.463000 2.557000
>>> C 9.350000 10.242000 2.667000
>>> H 5.364000 8.657000 2.242000
>>> H 3.618000 7.257000 1.469000
>>> H 6.193000 5.762000 -1.303000
>>> H 7.887000 7.182000 -0.505000
>>> H 3.945000 5.217000 -1.175000
>>> H 8.570000 8.939000 0.805000
>>> H 9.680000 10.180000 3.551000
>>> H 9.782000 10.274000 1.903000
>>> H 9.113000 11.289000 2.695000
>>> N 7.867000 8.870000 1.370000
>>> O 3.758000 5.500000 -0.358000
>>> O 7.259000 9.378000 3.486000
>>> C 0.319000 6.066000 0.970000
>>> C -0.941000 6.052000 1.560000
>>> C -1.942000 6.903000 1.099000
>>> C -1.689000 7.764000 0.044000
>>> C -0.437000 7.761000 -0.567000
>>> C 0.550000 6.917000 -0.102000
>>> C 1.598000 4.637000 2.557000
>>> C 2.885000 3.858000 2.667000
>>> H -1.101000 5.443000 2.242000
>>> H -2.847000 6.843000 1.469000
>>> H -0.272000 8.338000 -1.303000
>>> H 1.422000 6.918000 -0.505000
>>> H -2.520000 8.883000 -1.175000
>>> H 2.105000 5.161000 0.805000
>>> H 3.215000 3.920000 3.551000
>>> H 3.317000 3.826000 1.903000
>>> H 2.648000 2.811000 2.695000
>>> N 1.402000 5.230000 1.370000
>>> O -2.707000 8.600000 -0.358000
>>> O 0.794000 4.722000 3.486000
>>> C 3.045000 1.366000 5.417000
>>> C 4.305000 1.352000 4.827000
>>> C 5.306000 2.203000 5.288000
>>> C 5.053000 3.064000 6.343000
>>> C 3.801000 3.061000 6.953000
>>> C 2.814000 2.217000 6.489000
>>> C 1.766000 -0.063000 3.830000
>>> C 0.479000 -0.842000 3.720000
>>> H 4.465000 0.743000 4.145000
>>> H 6.210000 2.143000 4.918000
>>> H 3.636000 3.638000 7.690000
>>> H 1.942000 2.218000 6.891000
>>> H 5.883000 4.183000 7.562000
>>> H 1.259000 0.461000 5.582000
>>> H 0.149000 -0.780000 2.836000
>>> H 0.047000 -0.874000 4.484000
>>> H 0.716000 -1.889000 3.692000
>>> N 1.961000 0.530000 5.017000
>>> O 6.071000 3.900000 6.745000
>>> O 2.570000 0.022000 2.901000
>>> &END COORD
>>>
>>>
>>> &KIND H
>>> BASIS_SET 6-31++Gxx
>>> POTENTIAL GTH-PBE-q1
>>> &END KIND
>>>
>>> &KIND C
>>> BASIS_SET 6-31++Gxx
>>> POTENTIAL GTH-PBE-q4
>>> &END KIND
>>>
>>> &KIND N
>>> BASIS_SET 6-31++Gxx
>>> POTENTIAL GTH-PBE-q5
>>> &END KIND
>>>
>>> &KIND O
>>> BASIS_SET 6-31++Gxx
>>> POTENTIAL GTH-PBE-q6
>>> &END KIND
>>>
>>> &END SUBSYS
>>>
>>> &PROPERTIES
>>> &RESP
>>> &PERIODIC_SYS
>>> ATOM_LIST 1..80
>>> LENGTH 3.0
>>> RANGE 3.0 5.0
>>> SURF_DIRECTION Z
>>> &END PERIODIC_SYS
>>> &PRINT
>>> &V_RESP_CUBE
>>> &END V_RESP_CUBE
>>> &RESP_CHARGES_TO_FILE
>>> &END RESP_CHARGES_TO_FILE
>>> &PROGRAM_RUN_INFO
>>> &END
>>> &COORD_FIT_POINTS
>>> &END
>>> &END PRINT
>>>
>>> &END RESP
>>>
>>> &FIT_CHARGE
>>> &END FIT_CHARGE
>>>
>>> &END PROPERTIES
>>>
>>> &END FORCE_EVAL
>>>
>>>
>>>
>>>
>>> RESP charges:
>>> Type | Atom | Charge
>>>
>>> RESP 1 C 0.070078
>>> RESP 2 C 0.903292
>>> RESP 3 C 1.113722
>>> RESP 4 C -0.120125
>>> RESP 5 C 1.058323
>>> RESP 6 C 0.655164
>>> RESP 7 C 0.200902
>>> RESP 8 C 2.666416
>>> RESP 9 H -1.222912
>>> RESP 10 H -1.170830
>>> RESP 11 H -0.983979
>>> RESP 12 H -0.818745
>>> RESP 13 H -1.214944
>>> RESP 14 H -1.366249
>>> RESP 15 H -1.224385
>>> RESP 16 H -1.119378
>>> RESP 17 H -1.137280
>>> RESP 18 N 1.259042
>>> RESP 19 O 1.713353
>>> RESP 20 O 0.738109
>>> RESP 21 C 0.041413
>>> RESP 22 C 0.956097
>>> RESP 23 C 1.040182
>>> RESP 24 C -0.082712
>>> RESP 25 C 1.051482
>>> RESP 26 C 0.595410
>>> RESP 27 C 0.202160
>>> RESP 28 C 2.662283
>>> RESP 29 H -1.166123
>>> RESP 30 H -1.126516
>>> RESP 31 H -0.984189
>>> RESP 32 H -0.800838
>>> RESP 33 H -1.188989
>>> RESP 34 H -1.401436
>>> RESP 35 H -1.215709
>>> RESP 36 H -1.113423
>>> RESP 37 H -1.128041
>>> RESP 38 N 1.271056
>>> RESP 39 O 1.679915
>>> RESP 40 O 0.705805
>>> RESP 41 C 0.116651
>>> RESP 42 C 0.813652
>>> RESP 43 C 1.108160
>>> RESP 44 C -0.087035
>>> RESP 45 C 1.066794
>>> RESP 46 C 0.598212
>>> RESP 47 C 0.201601
>>> RESP 48 C 2.665972
>>> RESP 49 H -1.193708
>>> RESP 50 H -1.119339
>>> RESP 51 H -0.987625
>>> RESP 52 H -0.794927
>>> RESP 53 H -1.193920
>>> RESP 54 H -1.403616
>>> RESP 55 H -1.227170
>>> RESP 56 H -1.113105
>>> RESP 57 H -1.132964
>>> RESP 58 N 1.272201
>>> RESP 59 O 1.669491
>>> RESP 60 O 0.737908
>>> RESP 61 C 0.017005
>>> RESP 62 C 0.961044
>>> RESP 63 C 1.064429
>>> RESP 64 C -0.116028
>>> RESP 65 C 1.044203
>>> RESP 66 C 0.665757
>>> RESP 67 C 0.203199
>>> RESP 68 C 2.660164
>>> RESP 69 H -1.159143
>>> RESP 70 H -1.166104
>>> RESP 71 H -0.983609
>>> RESP 72 H -0.830205
>>> RESP 73 H -1.208912
>>> RESP 74 H -1.380887
>>> RESP 75 H -1.210223
>>> RESP 76 H -1.114933
>>> RESP 77 H -1.129467
>>> RESP 78 N 1.264612
>>> RESP 79 O 1.718065
>>> RESP 80 O 0.706396
>>> Total 0.000000
>>>
>>>
>>> DDAP FULL DENSITY charges:
>>> Atom | Charge
>>>
>>> 1 C 0.164768
>>> 2 C 0.038789
>>> 3 C -0.067304
>>> 4 C 0.359095
>>> 5 C -0.273509
>>> 6 C -0.132861
>>> 7 C 0.453357
>>> 8 C -0.073176
>>> 9 H -0.008341
>>> 10 H -0.007149
>>> 11 H 0.160840
>>> 12 H 0.127025
>>> 13 H 0.128967
>>> 14 H 0.094082
>>> 15 H -0.056650
>>> 16 H -0.027392
>>> 17 H -0.004180
>>> 18 N -0.249484
>>> 19 O -0.299466
>>> 20 O -0.327283
>>> 21 C 0.164803
>>> 22 C 0.039171
>>> 23 C -0.067954
>>> 24 C 0.359126
>>> 25 C -0.273851
>>> 26 C -0.132660
>>> 27 C 0.453316
>>> 28 C -0.072910
>>> 29 H -0.008512
>>> 30 H -0.006834
>>> 31 H 0.161121
>>> 32 H 0.126815
>>> 33 H 0.128910
>>> 34 H 0.094007
>>> 35 H -0.056723
>>> 36 H -0.027522
>>> 37 H -0.004288
>>> 38 N -0.249623
>>> 39 O -0.299372
>>> 40 O -0.327146
>>> 41 C 0.164786
>>> 42 C 0.039184
>>> 43 C -0.067967
>>> 44 C 0.359125
>>> 45 C -0.273855
>>> 46 C -0.132651
>>> 47 C 0.453282
>>> 48 C -0.072882
>>> 49 H -0.008520
>>> 50 H -0.006829
>>> 51 H 0.161118
>>> 52 H 0.126809
>>> 53 H 0.128911
>>> 54 H 0.094000
>>> 55 H -0.056734
>>> 56 H -0.027529
>>> 57 H -0.004294
>>> 58 N -0.249597
>>> 59 O -0.299364
>>> 60 O -0.327128
>>> 61 C 0.164787
>>> 62 C 0.038773
>>> 63 C -0.067297
>>> 64 C 0.359102
>>> 65 C -0.273505
>>> 66 C -0.132867
>>> 67 C 0.453392
>>> 68 C -0.073203
>>> 69 H -0.008330
>>> 70 H -0.007154
>>> 71 H 0.160838
>>> 72 H 0.127032
>>> 73 H 0.128964
>>> 74 H 0.094090
>>> 75 H -0.056640
>>> 76 H -0.027383
>>> 77 H -0.004176
>>> 78 N -0.249513
>>> 79 O -0.299471
>>> 80 O -0.327306
>>> Total 0.000000
>>>
>>>
>>> --
>>> You received this message because you are subscribed to the Google
>>> Groups "cp2k" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to cp2k+... at googlegroups.com.
>>> To post to this group, send email to cp... at googlegroups.com.
>>>
>>> Visit this group at http://groups.google.com/group/cp2k.
>>> For more options, visit https://groups.google.com/groups/opt_out.
>>>
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20140302/5c7b92f5/attachment.htm>
More information about the CP2K-user
mailing list