Dear All,
I would like to know whether the dimer method works well for the fixed substrates and the molecule alone relaxed? In my calculation I have the substrate fixed and the 6 atoms of the molecule relaxed. I have started the new version of VTST tools with the approximate POSACR close to saddle point derived from NEB. Also I have used exactly the same INCAR as posted on your website. The calculation have been running since a long time without any convergence of force to zero.
Could any of you please help me to solve this problem?
Thanking you in advance
Will dimer method work for fixed substrates
Moderator: moderators
No, I have not used this script to generate POSCAR and MODECAR. I have just taken the highest energy point I have obtained so far and which has two imaginary frequencies. I have generated MODECAR file using modemake.pl
DIMCAR is as follows
Step Force Torque Energy Curvature Angle
1 0.58593 4.73833 -113.57518 0.17641 5.39813
1 0.58593 3.92963 -113.57518 0.28389 5.99695
1 0.58593 3.26786 -113.57518 -0.09804 3.89166
1 0.58593 1.25774 -113.57518 -0.16851 1.60162
2 0.15410 1.84870 -113.57836 -0.65149 2.23031
2 0.15410 2.31080 -113.57836 -0.18084 2.42119
2 0.15410 2.03897 -113.57836 -0.27761 2.00529
2 0.15410 3.09949 -113.57836 -0.21811 2.44410
3 0.16641 3.04452 -113.57862 -0.79193 3.17446
3 0.16641 3.48608 -113.57862 -0.22266 2.96166
3 0.16641 2.00520 -113.57862 -0.41552 1.56009
3 0.16641 2.89197 -113.57862 -0.57129 3.10367
4 0.20294 2.07665 -113.57845 -0.97640 1.78263
4 0.20294 2.58352 -113.57845 -0.45502 2.68744
4 0.20294 3.82830 -113.57845 -0.66946 3.17654
4 0.20294 3.03303 -113.57845 -0.64545 2.56333
5 0.56561 1.51455 -113.57268 -0.50091 1.93325
5 0.56561 1.25784 -113.57268 -0.16276 1.64874
5 0.56561 2.92476 -113.57268 -0.11383 2.54735
5 0.56561 3.51436 -113.57268 -0.25447 3.11061
6 0.37482 2.19668 -113.57305 -0.24320 2.01683
6 0.37482 2.87121 -113.57305 0.06273 3.15540
6 0.37482 3.60676 -113.57305 0.15630 3.30218
6 0.37482 3.02275 -113.57305 -0.07678 2.70807
7 0.39548 2.27706 -113.57601 -0.59953 2.30143
7 0.39548 2.45941 -113.57601 -0.28934 2.46077
7 0.39548 2.41906 -113.57601 -0.29361 2.17738
7 0.39548 2.25087 -113.57601 -0.42244 1.79675
8 0.40685 1.32146 -113.58241 -0.59732 1.36703
8 0.40685 1.94183 -113.58241 -0.42303 1.60742
8 0.40685 4.02167 -113.58241 -0.15710 3.13217
8 0.40685 4.26936 -113.58241 -0.41720 4.45740
9 0.49114 3.18476 -113.58541 -0.53707 2.99694
9 0.49114 1.79161 -113.58541 -0.24211 1.56717
9 0.49114 2.29448 -113.58541 -0.42686 1.88111
9 0.49114 3.09017 -113.58541 -0.17943 2.48089
10 0.27115 2.54581 -113.59187 -0.57609 2.63544
10 0.27115 1.45025 -113.59187 -0.10979 1.32309
10 0.27115 1.98431 -113.59187 -0.22606 1.89318
10 0.27115 3.22304 -113.59187 -0.16600 2.56151
11 0.36852 1.23393 -113.59287 -0.53317 1.32145
11 0.36852 1.39791 -113.59287 -0.14478 1.28253
11 0.36852 3.00724 -113.59287 -0.36202 2.61957
11 0.36852 3.24555 -113.59287 -0.18496 2.66762
12 1.15556 1.68075 -113.59961 -0.59350 1.81347
12 1.15556 2.14588 -113.59961 -0.18584 2.02631
12 1.15556 2.44508 -113.59961 -0.32023 2.02572
12 1.15556 3.13653 -113.59961 -0.31882 2.46100
13 0.51330 2.15487 -113.61004 -0.61738 2.54055
13 0.51330 2.12847 -113.61004 -0.15817 2.11387
13 0.51330 1.74706 -113.61004 -0.37510 1.60621
13 0.51330 2.75493 -113.61004 -0.22790 2.46601
14 0.58998 1.85989 -113.62465 -0.65768 1.64342
14 0.58998 1.58646 -113.62465 -0.29045 1.50964
14 0.58998 2.71437 -113.62465 -0.48721 2.21683
14 0.58998 2.67654 -113.62465 -0.30715 2.29428
And the calculation is still running.
DIMCAR is as follows
Step Force Torque Energy Curvature Angle
1 0.58593 4.73833 -113.57518 0.17641 5.39813
1 0.58593 3.92963 -113.57518 0.28389 5.99695
1 0.58593 3.26786 -113.57518 -0.09804 3.89166
1 0.58593 1.25774 -113.57518 -0.16851 1.60162
2 0.15410 1.84870 -113.57836 -0.65149 2.23031
2 0.15410 2.31080 -113.57836 -0.18084 2.42119
2 0.15410 2.03897 -113.57836 -0.27761 2.00529
2 0.15410 3.09949 -113.57836 -0.21811 2.44410
3 0.16641 3.04452 -113.57862 -0.79193 3.17446
3 0.16641 3.48608 -113.57862 -0.22266 2.96166
3 0.16641 2.00520 -113.57862 -0.41552 1.56009
3 0.16641 2.89197 -113.57862 -0.57129 3.10367
4 0.20294 2.07665 -113.57845 -0.97640 1.78263
4 0.20294 2.58352 -113.57845 -0.45502 2.68744
4 0.20294 3.82830 -113.57845 -0.66946 3.17654
4 0.20294 3.03303 -113.57845 -0.64545 2.56333
5 0.56561 1.51455 -113.57268 -0.50091 1.93325
5 0.56561 1.25784 -113.57268 -0.16276 1.64874
5 0.56561 2.92476 -113.57268 -0.11383 2.54735
5 0.56561 3.51436 -113.57268 -0.25447 3.11061
6 0.37482 2.19668 -113.57305 -0.24320 2.01683
6 0.37482 2.87121 -113.57305 0.06273 3.15540
6 0.37482 3.60676 -113.57305 0.15630 3.30218
6 0.37482 3.02275 -113.57305 -0.07678 2.70807
7 0.39548 2.27706 -113.57601 -0.59953 2.30143
7 0.39548 2.45941 -113.57601 -0.28934 2.46077
7 0.39548 2.41906 -113.57601 -0.29361 2.17738
7 0.39548 2.25087 -113.57601 -0.42244 1.79675
8 0.40685 1.32146 -113.58241 -0.59732 1.36703
8 0.40685 1.94183 -113.58241 -0.42303 1.60742
8 0.40685 4.02167 -113.58241 -0.15710 3.13217
8 0.40685 4.26936 -113.58241 -0.41720 4.45740
9 0.49114 3.18476 -113.58541 -0.53707 2.99694
9 0.49114 1.79161 -113.58541 -0.24211 1.56717
9 0.49114 2.29448 -113.58541 -0.42686 1.88111
9 0.49114 3.09017 -113.58541 -0.17943 2.48089
10 0.27115 2.54581 -113.59187 -0.57609 2.63544
10 0.27115 1.45025 -113.59187 -0.10979 1.32309
10 0.27115 1.98431 -113.59187 -0.22606 1.89318
10 0.27115 3.22304 -113.59187 -0.16600 2.56151
11 0.36852 1.23393 -113.59287 -0.53317 1.32145
11 0.36852 1.39791 -113.59287 -0.14478 1.28253
11 0.36852 3.00724 -113.59287 -0.36202 2.61957
11 0.36852 3.24555 -113.59287 -0.18496 2.66762
12 1.15556 1.68075 -113.59961 -0.59350 1.81347
12 1.15556 2.14588 -113.59961 -0.18584 2.02631
12 1.15556 2.44508 -113.59961 -0.32023 2.02572
12 1.15556 3.13653 -113.59961 -0.31882 2.46100
13 0.51330 2.15487 -113.61004 -0.61738 2.54055
13 0.51330 2.12847 -113.61004 -0.15817 2.11387
13 0.51330 1.74706 -113.61004 -0.37510 1.60621
13 0.51330 2.75493 -113.61004 -0.22790 2.46601
14 0.58998 1.85989 -113.62465 -0.65768 1.64342
14 0.58998 1.58646 -113.62465 -0.29045 1.50964
14 0.58998 2.71437 -113.62465 -0.48721 2.21683
14 0.58998 2.67654 -113.62465 -0.30715 2.29428
And the calculation is still running.
There does seem to be a convergence problem here, but one that I think can be easily fixed. When things are working properly, the Torque should drop systematically, and the curvature also drop to a converged value, for the rotational iterations within each step. The first step doesn't look too bad, but the second is not good (the torque increases, and the curvature also rises).
These calculations require accurate forces. If there is noise in the forces on the order of the difference in force between the dimer images, things will not converge.
Can you try the following changes, starting from the same initial conditions. I think it will be clear when the method is working, and it would be interesting to know what fixes it.
1. increasing the accuracy in the forces: ediff=1e-8
2. increasing the dimer separation: DdR=0.01
3. increase the precision: PREC=Acc
There are a few other parameters which could increase efficiency once things are working, but first, it's most important to see smooth convergence of the lowest mode. One other check is to increase DRotMax=10, and make sure the Torque drops to a low value at each step.
If these changes don't work, we would really like to see your input files and understand what is going on.
These calculations require accurate forces. If there is noise in the forces on the order of the difference in force between the dimer images, things will not converge.
Can you try the following changes, starting from the same initial conditions. I think it will be clear when the method is working, and it would be interesting to know what fixes it.
1. increasing the accuracy in the forces: ediff=1e-8
2. increasing the dimer separation: DdR=0.01
3. increase the precision: PREC=Acc
There are a few other parameters which could increase efficiency once things are working, but first, it's most important to see smooth convergence of the lowest mode. One other check is to increase DRotMax=10, and make sure the Torque drops to a low value at each step.
If these changes don't work, we would really like to see your input files and understand what is going on.