SCF doesn't convergence

Hi, PSI4 communities,
When I ran the SAPT0, I met the problem of SCF doesn’t convergence, and here are my questions.

  1. Is PSI4 could increase the iterations or add an outer SCF if the first circle doesn’t perform well?
  2. Could I turn down the [E_CONVERGENCE] like 10-8, or 10-7
    I attached my inp file with this email, could someone help me, please!
    Any reply will be appreciated!

memory 4 GB

molecule dimer {
0 1
C 7.00480000 5.45620000 5.47264000
C 8.23840000 6.17340000 5.49440000
C 5.77420000 7.58760000 5.48752000
C 7.00000000 8.30500000 5.59024000
C 4.54700000 9.71440000 5.50928000
C 5.77800000 10.41760000 5.62336000
C 3.31760000 11.84440000 5.50480000
C 4.55500000 12.55320000 5.53504000
C 2.08840000 13.97720000 5.48784000
C 3.32220000 14.68899900 5.47472000
C 9.47160000 5.45620000 5.47056000
C 10.70060000 6.16700000 5.44944000
C 8.23820000 7.58540000 5.56384000
C 9.47660100 8.30520000 5.58672000
C 6.98760000 9.70660000 5.71648000
C 8.23920100 10.43120000 6.12784000
C 5.77620000 11.84820100 5.61376000
C 7.01800000 12.56020000 5.60768000
C 4.55080000 13.97900000 5.49088000
C 5.78040000 14.68780000 5.45936000
C 11.93020000 5.45600000 5.42848000
C 13.16000000 6.16620000 5.40960000
C 10.70240000 7.58780000 5.48352000
C 11.93120000 8.29580000 5.46624000
C 9.48880000 9.70680000 5.71312000
C 10.69880000 10.41780000 5.61984000
C 8.23840000 11.87400000 5.70832000
C 9.45880000 12.56040000 5.60704000
C 7.01060000 13.97880000 5.49056000
C 8.23840000 14.68840000 5.45696000
C 14.38820000 5.45700000 5.41200000
C 15.61660000 6.16620000 5.41072000
C 13.16040000 7.58540000 5.41904000
C 14.38840000 8.29500000 5.42208000
C 11.92939900 9.71440000 5.50592000
C 13.15860000 10.42540000 5.47936000
C 10.70060000 11.84820100 5.61200000
C 11.92180100 12.55340000 5.53472000
C 9.46600100 13.97880000 5.49040000
C 10.69620000 14.68780000 5.46048000
C 16.84640100 5.45580000 5.43040000
C 18.07600000 6.16700000 5.45280000
C 15.61619900 7.58520000 5.42080000
C 16.84540000 8.29580000 5.46992000
C 14.38820000 9.71580000 5.44960000
C 15.61800000 10.42520000 5.48080000
C 13.15899900 11.84460100 5.50432000
C 14.38840000 12.55300000 5.49264000
C 11.92580000 13.97900000 5.49200000
C 13.15460000 14.68899900 5.47584000
O 8.24119900 10.52940000 7.58752000
H 7.19520000 10.55580000 7.95808000
H 7.21220000 9.85380000 4.60848000

0 1
O 8.23280000 7.95340000 8.65664000
H 7.86280000 8.69340000 9.39664000
H 9.34280000 7.95340000 8.65664000
}

set {
reference rohf
guess sad
scf_type DF
freeze_core True
basis jun-cc-pVDZ
}

energy(‘sapt0’)

Please follow our guidelines for posting topics, especially the one about backticks.

I don’t understand your questions. I don’t know what an “outer SCF” is, or what you mean by “is PSI4 could increase the iterations”. Do you mean “Can I increase the number of iterations spent in the SCF cycle before Psi4 stops gives up trying to converge it?”

As for your second question, are you asking how can I reduce e_convergence or should I reduce e_convergence, or something else?

Hi, jmisiewicz

Thanks for your reply, I will follow the guidelines for posting topic

I’ve tried to change the MAXITER to 1000 to increase the iterations, and I follow the manual to reduce D/E_CONVERGENCE to 1e-6 for this DF calculation. but it still convergence slow and stop the calculation without any feedback.My PSI4 version is 1.3.2.

Here’s is my inp file and the out file. Could you help me to accelerate the convergence of it? A lot of thanks!

memory 4 GB

molecule dimer {
0 1
C 7.00480000 5.45620000 5.47264000
C 8.23840000 6.17340000 5.49440000
C 5.77420000 7.58760000 5.48752000
C 7.00000000 8.30500000 5.59024000
C 4.54700000 9.71440000 5.50928000
C 5.77800000 10.41760000 5.62336000
C 3.31760000 11.84440000 5.50480000
C 4.55500000 12.55320000 5.53504000
C 2.08840000 13.97720000 5.48784000
C 3.32220000 14.68899900 5.47472000
C 9.47160000 5.45620000 5.47056000
C 10.70060000 6.16700000 5.44944000
C 8.23820000 7.58540000 5.56384000
C 9.47660100 8.30520000 5.58672000
C 6.98760000 9.70660000 5.71648000
C 8.23920100 10.43120000 6.12784000
C 5.77620000 11.84820100 5.61376000
C 7.01800000 12.56020000 5.60768000
C 4.55080000 13.97900000 5.49088000
C 5.78040000 14.68780000 5.45936000
C 11.93020000 5.45600000 5.42848000
C 13.16000000 6.16620000 5.40960000
C 10.70240000 7.58780000 5.48352000
C 11.93120000 8.29580000 5.46624000
C 9.48880000 9.70680000 5.71312000
C 10.69880000 10.41780000 5.61984000
C 8.23840000 11.87400000 5.70832000
C 9.45880000 12.56040000 5.60704000
C 7.01060000 13.97880000 5.49056000
C 8.23840000 14.68840000 5.45696000
C 14.38820000 5.45700000 5.41200000
C 15.61660000 6.16620000 5.41072000
C 13.16040000 7.58540000 5.41904000
C 14.38840000 8.29500000 5.42208000
C 11.92939900 9.71440000 5.50592000
C 13.15860000 10.42540000 5.47936000
C 10.70060000 11.84820100 5.61200000
C 11.92180100 12.55340000 5.53472000
C 9.46600100 13.97880000 5.49040000
C 10.69620000 14.68780000 5.46048000
C 16.84640100 5.45580000 5.43040000
C 18.07600000 6.16700000 5.45280000
C 15.61619900 7.58520000 5.42080000
C 16.84540000 8.29580000 5.46992000
C 14.38820000 9.71580000 5.44960000
C 15.61800000 10.42520000 5.48080000
C 13.15899900 11.84460100 5.50432000
C 14.38840000 12.55300000 5.49264000
C 11.92580000 13.97900000 5.49200000
C 13.15460000 14.68899900 5.47584000
O 8.24119900 10.52940000 7.58752000
H 7.19520000 10.55580000 7.95808000
H 7.21220000 9.85380000 4.60848000

0 1
O 8.23280000 7.95340000 8.65664000
H 7.86280000 8.69340000 9.39664000
H 9.34280000 7.95340000 8.65664000
}

set {
guess sad
MAXITER 1000
D_CONVERGENCE 1e-6
E_CONVERGENCE 1e-6
reference rohf
scf_type DF
freeze_core True
basis jun-cc-pVDZ
damping_percentage 40
}

energy(‘sapt0’)

@DF-ROHF iter 103: -2043.04823294325979 4.93727e-05 5.92103e-07 DIIS/DAMP=40%
@DF-ROHF iter 104: -2043.04803962594087 1.93317e-04 5.65661e-07 DIIS/DAMP=40%
@DF-ROHF iter 105: -2043.04868482006623 -6.45194e-04 5.64652e-07 DIIS/DAMP=40%
@DF-ROHF iter 106: -2043.04889561005893 -2.10790e-04 5.69422e-07 DIIS/DAMP=40%
@DF-ROHF iter 107: -2043.04863848304558 2.57127e-04 5.66919e-07 DIIS/DAMP=40%
@DF-ROHF iter 108: -2043.04805477998070 5.83703e-04 5.65644e-07 DIIS/DAMP=40%
@DF-ROHF iter 109: -2043.04841666184075 -3.61882e-04 5.53349e-07 DIIS/DAMP=40%
@DF-ROHF iter 110: -2043.04813163678500 2.85025e-04 5.79283e-07 DIIS/DAMP=40%
@DF-ROHF iter 111: -2043.04872487157991 -5.93235e-04 5.78141e-07 DIIS/DAMP=40%
@DF-ROHF iter 112: -2043.04910896020738 -3.84089e-04 5.88053e-07 DIIS/DAMP=40%
@DF-ROHF iter 113: -2043.04867750853009 4.31452e-04 5.87519e-07 DIIS/DAMP=40%
@DF-ROHF iter 114: -2043.04855170501105 1.25804e-04 5.85296e-07 DIIS/DAMP=40%
@DF-ROHF iter 115: -2043.04853703915705 1.46659e-05 5.91871e-07 DIIS/DAMP=40%
@DF-ROHF iter 116: -2043.04802064875730 5.16390e-04 5.83342e-07 DIIS/DAMP=40%
@DF-ROHF iter 117: -2043.04862215818412 -6.01509e-04 5.96505e-07 DIIS/DAMP=40%

Without any feedback, it stopped.

Best,
Sun

Update
I tried the soscf, it does help, but it suddenly stopped without any error.
Does anyone know what’s going on? Really appreciate your help!
PSI version 1.3.2
inp.inp.dat (16.4 KB)

As I said in the other post. Fix your structure!
You don’t have a closed shell molecule with all those cut bonds. It’s a highly reactive fantasy molecule. Saturate the dangling bonds, possibly enlarge the sheet to reduce edge effects.

Millions of thanks fo your help! It really does work.

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.