COF cell optimize exploded

Linfeng Gan

unread,

Apr 4, 2020, 10:27:13 PM4/4/20

to cp2k

Hello all,

I need some help about the covalent-organic frameworks (COF) 2D sheet cell optimize. After I optimized the COF with a 2X2X2 supercell, the cell size was totally exploded. I had kept the cell vectors and symmetry constant, changed different basis sets and different XC functions but the cell exploded anyway.

I had tried basis sets DZVP-MOLOPT-GTH and DZVP-MOLOPT-SR-GTH with PADE, PBE and BLYP. And the DFTD3 correction was also add to each calculation. The GAPW method I also had tried, but none of them worked.

Any suggestion is welcome, thanks!

Linfeng

The COF-5-2X2X2 supercell vectors should be:

CELL_TOP| Volume [angstrom^3]: 21253.270

CELL_TOP| Vector a [angstrom 60.040 0.000 0.000 |a| = 60.040

CELL_TOP| Vector b [angstrom -30.020 51.996 0.000 |b| = 60.040

CELL_TOP| Vector c [angstrom 0.000 0.000 6.808 |c| = 6.808

CELL_TOP| Angle (b,c), alpha [degree]: 90.000

CELL_TOP| Angle (a,c), beta [degree]: 90.000

CELL_TOP| Angle (a,b), gamma [degree]: 120.000

But after cell optimized calculation the output cell vectors are:

*******************************************************************************

*** GEOMETRY OPTIMIZATION COMPLETED ***

*******************************************************************************

Reevaluating energy at the minimum

CELL| Volume [angstrom^3]: 25653.623

CELL| Vector a [angstrom]: 60.345 0.000 0.000 |a| = 60.345

CELL| Vector b [angstrom]: -30.173 52.260 0.000 |b| = 60.345

CELL| Vector c [angstrom]: 0.000 -0.000 8.135 |c| = 8.135

CELL| Angle (b,c), alpha [degree]: 90.000

CELL| Angle (a,c), beta [degree]: 90.000

CELL| Angle (a,b), gamma [degree]: 120.000

COF-5-2X2X2.xyz

opt_cell.inp

hut...@chem.uzh.ch

unread,

Apr 6, 2020, 3:32:56 AM4/6/20

to cp...@googlegroups.com

Hi

I assume your system is weakly bound in z-direction. Small force
changes in that direction can cause large changes.
The only obvious problem I could spot is the definition of the
reference cell. In order to minimize the effect of a change in
volume, the reference cell should be larger than the initial
and final cell. I would set c=10 or 12 and I would also increase
the cutoff.
You could test this by just scan the energy as function of c and
correctly adapt the geometry of course.

regards

Juerg Hutter
--------------------------------------------------------------
Juerg Hutter Phone : ++41 44 635 4491
Institut für Chemie C FAX : ++41 44 635 6838
Universität Zürich E-mail: hut...@chem.uzh.ch
Winterthurerstrasse 190
CH-8057 Zürich, Switzerland
---------------------------------------------------------------

-----cp...@googlegroups.com wrote: -----
To: "cp2k" <cp...@googlegroups.com>
From: "Linfeng Gan"
Sent by: cp...@googlegroups.com
Date: 04/05/2020 04:27AM
Subject: [CP2K:13056] COF cell optimize exploded

--
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+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/2da4af3b-2dcf-46c3-a49e-3dc62b170405%40googlegroups.com.

[attachment "COF-5-2X2X2.xyz" removed by Jürg Hutter/at/UZH]
[attachment "opt_cell.inp" removed by Jürg Hutter/at/UZH]

Linfeng Gan

unread,

Apr 7, 2020, 4:11:46 AM4/7/20

to cp2k

Thanks Juerg. I will try as you suggest and update a few days later.

To unsubscribe from this group and stop receiving emails from it, send an email to cp...@googlegroups.com.

Linfeng Gan

unread,

May 14, 2020, 6:48:20 AM5/14/20

to cp2k

I had carefully tested the CUTOFF and the CELL_REF. The test results showed that increased the CUTOFF had most helpful.

So, the problem solved

shixun sun

unread,

May 14, 2020, 6:54:53 AM5/14/20

to cp...@googlegroups.com

Hi, Gan

I had the same problem with you now, and could you tell me how much you increase your cutoff. and did you change your c axis, like you want 6, so you increase it to 10 in your input

Millions of thanks for your help

best

Linfeng Gan <linfe...@gmail.com> 于2020年5月14日周四下午6:48写道：

To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/232917c9-b4fa-4094-874a-51d9730e9db2%40googlegroups.com.

Linfeng Gan

unread,

May 14, 2020, 8:29:29 PM5/14/20

to cp2k

Hi, Sun

I would like to share some of my experience with you.

(1) CUTOFF

As you can see from the cutoff test, you must increase cutoff at least up to 1000 Ry. I had test cutoff ranged from 600 to 2500, the highest the best. At 1200 Ry, the cell optimization result is very close to the experimental data.

# Grid cutoff vs total energy

# Date: Tue Apr 7 10:28:38 2020

# rel_cutoff = 60

50 -511.8311527382 -145.0795204885 1191545 55218 480 0 0

100 -498.0188507854 -141.1643970471 1191545 55218 480 0 0

150 -498.1333770910 -141.1968597477 1084277 152208 10758 0 0

200 -498.0596390179 -141.1759585497 1011045 180500 55218 480 0

250 -498.0576533189 -141.1753956995 894935 264998 86030 1280 0

300 -498.0561376050 -141.1749660675 839389 277782 127992 2080 0

350 -498.0556031621 -141.1748145786 784888 299389 152208 10758 0

400 -498.0546430297 -141.1745424273 743542 307599 172215 23887 0

450 -498.0543833523 -141.1744688212 715246 305115 187987 38415 480

500 -498.0547962093 -141.1745858463 664938 346107 180500 55218 480

550 -498.0548813984 -141.1746099933 648732 310339 223070 63822 1280

600 -498.0548597218 -141.1746038491 597516 325211 250210 73026 1280

650 -498.0549993006 -141.1746434129 558534 336401 264998 86030 1280

700 -498.0548179649 -141.1745920130 513592 356057 290284 85230 2080

750 -498.0547555671 -141.1745743262 499950 340239 294170 110804 2080

800 -498.0547613057 -141.1745759528 452398 386991 277782 127992 2080

850 -498.0547622822 -141.1745762296 415918 410299 267688 147908 5430

900 -498.0547014016 -141.1745589729 403034 388334 292909 157536 5430

950 -498.0547014005 -141.1745589726 390150 394738 299389 152208 10758

1000 -498.0546629837 -141.1745480833 374096 381572 318409 162408 10758

1050 -498.0546627095 -141.1745480056 372626 379024 308511 172231 14851

1100 -498.0546627489 -141.1745480167 333972 409570 307599 172215 23887

1150 -498.0546599853 -141.1745472334 291540 452002 287140 192674 23887

1200 -498.0546599398 -141.1745472205 283078 459664 282659 188911 32931

1250 -498.0546664857 -141.1745490759 265352 449894 305115 187987 38895

1300 -498.0546695315 -141.1745499393 257340 430648 331573 180200 47482

1350 -498.0546694852 -141.1745499261 249304 433344 336433 172464 55698

1400 -498.0546619268 -141.1745477837 249304 415634 346107 180500 55698

1450 -498.0546657092 -141.1745488558 237064 427874 330835 195772 55698

1500 -498.0546657265 -141.1745488607 237064 419238 322207 203632 65102

1550 -498.0546649445 -141.1745486391 232240 416492 310339 223070 65102

1600 -498.0546621702 -141.1745478527 232240 400820 312383 236698 65102

1650 -498.0546580786 -141.1745466929 224234 392132 329077 228294 73506

1700 -498.0546579920 -141.1745466684 224234 373282 325211 250210 74306

1750 -498.0546626389 -141.1745479855 224234 368338 318615 255886 80170

1800 -498.0546626425 -141.1745479866 216222 347172 339757 263922 80170

1850 -498.0546624400 -141.1745479292 216222 342312 336401 264998 87310

1900 -498.0546624669 -141.1745479368 196998 345386 332173 285376 87310

1950 -498.0546619107 -141.1745477791 153670 372770 343209 290284 87310

2000 -498.0546645404 -141.1745485245 127866 385726 356057 290284 87310

2050 -498.0546619783 -141.1745477983 107134 406458 356057 285100 92494

2100 -498.0546619823 -141.1745477994 82148 431444 339841 296252 97558

2150 -498.0546619947 -141.1745478029 77348 422602 340239 294170 112884

2200 -498.0546617280 -141.1745477273 69312 400926 369951 294170 112884

2250 -498.0546617309 -141.1745477282 64488 392818 382883 285546 121508

2300 -498.0546623124 -141.1745478930 64488 387910 386991 277782 130072

2350 -498.0546623561 -141.1745479054 53548 378466 399339 285818 130072

2400 -498.0546621789 -141.1745478552 53548 370382 407423 270648 145242

2450 -498.0546621769 -141.1745478546 43948 371970 410299 267688 153338

2500 -498.0546611280 -141.1745475573 35936 379982 402203 274984 154138

(2) CELL_REF

I had tested the CELL_REF vectors A, B, and C from 1 to 3 times of origin CELL corresponding vectors with 0.2 interval. I think 1.2 time of the CELL vectors ABC is enough which has a good balancing the computational time and accuracy.

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/232917c9-b4fa-4094-874a-51d9730e9db2%40googlegroups.com.

shixun sun

unread,

May 15, 2020, 3:11:32 AM5/15/20

to cp...@googlegroups.com

Dear Gan

That helps a lot. Thank you!

Best

Linfeng Gan <linfe...@gmail.com> 于2020年5月15日周五上午8:29写道：

To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ed5e69f2-5ea6-4c8c-8a9b-612e661ae157%40googlegroups.com.

Yingchun Zhang

unread,

May 15, 2020, 2:26:45 PM5/15/20

to cp2k

Hi Linfeng,

can you test the cutoff with EPS_DEFAULT of 1.0e-12 or 1.0e-14?

based on my experience, a cutoff of 1200 is too large for DZVP and CHO systems;

Best,

Yingchun

Linfeng Gan

unread,

May 16, 2020, 9:07:59 PM5/16/20

to cp2k

Hi Yingchun,
I don’t think decrease the EPS_DEFAULT value would significantly help to solve the explosion problem. Normally, a CUTOFF of 1200 is definitely too large for a system that only have CHO elements. As I posed above 600Ry seemed was a not bad choice, right. But it didn’t work. I would try a calculation with EPS_DEFAULT of 1.0e-14 a few days later. I don’t have time and resources right now. And I will pose results to let you know.

Reply all

Reply to author

Forward