Error message of the raaxff

LAMMPS LAMMPS General Discussion
,
1

Dear all,
I attempted to conduct a simulation using the Si/C/Li reaxff force field. I directly downloaded the txt file provided by the author of the literature and used the following approach.

pair_style reaxff NULL lgvdw yes
pair_coeff  * * LiSiC.reaxff Li Si C
...
fix 1 all qeq/reaxff 50 0.0 10.0 1.0e-4 reaxff

The error message is as follows:

ERROR on proc 0: Not a valid floating-point number: 'Si' (src/REAXFF/reaxff_ffield.cpp:584)
Last command: pair_coeff  * * C.reaxff Li Si C

I would like to inquire about the reason for the error. The link to the original document is as follows.

LiSiC.reaxff (7.1 KB)
https://pubs.acs.org/doi/abs/10.1021/acs.jpcc.2c07773#_i32

2

The force field name was incorrectly entered in the second codg(c.reaxff), please forgive my mistake :pray:

3

That ReaxFF force field file is bogus. It cannot work.

For example, it has

  7    ! Nr of atoms; cov.r; valency;a.m;Rvdw;Evdw;gammaEEM;cov.r2;#            
            alfa;gammavdW;valency;Eunder;Eover;chiEEM;etaEEM;n.u.               
            cov r3;Elp;Heat inc.;n.u.;n.u.;n.u.;n.u.                            
            ov/un;val1;n.u.;val3,vval4

But then there are only 3 entries following:

 Li   2.1097   1.0000   6.9410   2.1461   0.3726   0.8651  -0.1000   1.0000     
      9.0000   1.2063   1.0000   0.0000   0.0000  -6.2351  12.7757   0.0000     
     -1.0000   0.0000  37.5000   5.4409   6.9107   0.1973   0.8563   0.0000     
    -17.1659   2.2989   1.0338   1.0000   2.8103   1.3000   0.2000  13.0000     
 Si   2.0057   4.0000  28.0600   1.7098   0.2683   0.3041   1.2635   4.0000     
     12.5005   5.2054   4.0000  21.7115 139.9309   3.8858   5.3952   0.0000     
     -1.0000   0.0000 104.0000  11.7786  23.8188   0.8381   0.8563   0.0000     
     -2.5000   1.8326   1.0338   4.0000   2.5791   1.4000   0.2000  13.0000     
 C    1.4620   4.0000  12.0000   2.0931   0.2002   0.9751   1.2239   4.0000     
      9.0000   1.5000   4.0000  30.0000  79.5548   3.4759  13.2596   0.0000     
      1.1080   0.0000 181.0000  14.6589  24.4406   6.7313   0.8563   0.0000     
     -8.2623   5.0000   1.0564   4.0000   2.9663   1.0000   0.1000  10.0000

And this continues throughout the file.

Furthermore, this is not a parameter file compatible with lgvdw yes
It does not have enough data per atom for that. This is why LAMMPS is trying to read the next number in the per-element block, but that is then the “Si” element label starting the next block.

Finally, the last line is bogus, so the number of hydrogen bonds must be 0

Overall, considering these gross mistakes, I would not trust this parameter file or the corresponding publication at all.

1 Like