Qchem fatal error

Questions
#1

Hello everyone,

I am running EDA on nickel complexes and have received the following message at the end of my output file.

Q-Chem fatal error occurred in module libalmo/libalmo/scfmi/subrot_rss.C, line 566:

QUADRATIC EQUATION FOR X NOT SATISFIED IN RS PT

Has anyone come across this error message and how did you solve the problem. I have also attached my input file in case this may help. I am running my calculations on Q-Chem 5.3.

$molecule
0 2

0 1
Ni -0.196933000000 1.868158000000 -2.576557000000
C -1.350391000000 2.975666000000 -3.650642000000
C -3.756542000000 4.284431000000 -4.309107000000
C -1.602108000000 4.290207000000 -3.224533000000
C -2.341721000000 2.331943000000 -4.415737000000
C -3.527591000000 2.973274000000 -4.738916000000
C -2.791286000000 4.936503000000 -3.544424000000
H -0.851229000000 4.835107000000 -2.650317000000
H -2.200908000000 1.297366000000 -4.743914000000
H -4.289175000000 2.448961000000 -5.326052000000
H -2.961949000000 5.958066000000 -3.193426000000
C 2.055420000000 2.425187000000 -0.942165000000
C 2.523884000000 4.586735000000 -0.408718000000
C 1.075420000000 4.433579000000 -0.910748000000
N 0.985700000000 3.032372000000 -1.337832000000
H 3.157346000000 5.150591000000 -1.117244000000
H 0.919103000000 5.074424000000 -1.796316000000
C 1.638035000000 -0.944322000000 -2.217576000000
C 2.518706000000 -1.269666000000 -0.991650000000
C 2.114936000000 0.963618000000 -1.166246000000
N 1.397665000000 0.497028000000 -2.132480000000
H 0.664973000000 -1.466497000000 -2.150327000000
H 3.438800000000 -1.820019000000 -1.253795000000
H 2.594724000000 5.079140000000 0.575887000000
H 1.982427000000 -1.851800000000 -0.222549000000
C -0.011525000000 4.743539000000 0.126305000000
H -0.966672000000 4.461860000000 -0.360433000000
C 2.264509000000 -1.301743000000 -3.571873000000
H 1.529102000000 -0.967739000000 -4.329771000000
N 3.001948000000 3.196880000000 -0.345428000000
N 2.838473000000 0.060584000000 -0.454227000000
C 0.118019000000 3.913981000000 1.391629000000
H -0.688658000000 4.161072000000 2.100830000000
H 0.037646000000 2.835362000000 1.175021000000
H 1.076082000000 4.094072000000 1.914108000000
C -0.048594000000 6.233156000000 0.419542000000
H -0.186581000000 6.823685000000 -0.502465000000
H -0.876216000000 6.480434000000 1.103944000000
H 0.884582000000 6.578351000000 0.900153000000
C 3.565452000000 -0.559336000000 -3.828490000000
H 3.970646000000 -0.811759000000 -4.822184000000
H 3.422305000000 0.535447000000 -3.799600000000
H 4.343257000000 -0.818094000000 -3.085990000000
C 2.432919000000 -2.804815000000 -3.710541000000
H 1.483620000000 -3.337685000000 -3.534059000000
H 2.783221000000 -3.069910000000 -4.721274000000
H 3.176718000000 -3.201667000000 -2.997040000000
C 3.311997000000 0.223624000000 0.869618000000
C 4.560001000000 -0.298704000000 1.218994000000
C 2.526601000000 0.866723000000 1.830074000000
C 5.029293000000 -0.153124000000 2.521083000000
H 5.165075000000 -0.798706000000 0.455586000000
C 3.012525000000 1.028741000000 3.124656000000
H 1.528689000000 1.228068000000 1.560127000000
C 4.263335000000 0.520427000000 3.473113000000
H 6.009378000000 -0.557361000000 2.790773000000
H 2.397310000000 1.540661000000 3.870789000000
H 4.639435000000 0.641261000000 4.493102000000
C 4.382930000000 2.905424000000 -0.269321000000
C 5.108649000000 3.348370000000 0.839651000000
C 5.033205000000 2.223181000000 -1.302872000000
C 6.473402000000 3.089939000000 0.922481000000
H 4.587041000000 3.878340000000 1.643164000000
C 6.394852000000 1.950355000000 -1.202854000000
H 4.471686000000 1.908680000000 -2.189218000000
C 7.118439000000 2.382214000000 -0.091970000000
H 7.035622000000 3.433421000000 1.795674000000
H 6.895979000000 1.408505000000 -2.010688000000
H 8.189589000000 2.173173000000 -0.019912000000
I -1.899535000000 0.600107000000 -0.986311000000
C -5.022919000000 4.959520000000 -4.715802000000
F -5.164492000000 6.166858000000 -4.168603000000
F -5.091474000000 5.127533000000 -6.047621000000
F -6.104652000000 4.245360000000 -4.381668000000

0 2
C 0.487725000000 2.531236000000 -4.382329000000
H 1.428879000000 2.300971000000 -3.837612000000
C 0.300505000000 1.492221000000 -5.469532000000
H 1.250749000000 1.305238000000 -5.999182000000
H -0.029628000000 0.529069000000 -5.041896000000
H -0.443777000000 1.797583000000 -6.220748000000
O 2.186161000000 3.866180000000 -5.980617000000
O -0.211881000000 4.506288000000 -6.087657000000
C 3.451606000000 3.575928000000 -5.385709000000
H 3.563362000000 2.476395000000 -5.309953000000
C -0.443332000000 5.861198000000 -6.485122000000
H 0.442718000000 6.235230000000 -7.031750000000
O 1.215622000000 5.210353000000 -3.994081000000
P 0.918193000000 4.158496000000 -5.007762000000
H -0.559544000000 6.488260000000 -5.582524000000
H 3.484937000000 3.983196000000 -4.356572000000
C -1.674253000000 5.894991000000 -7.341915000000
H -2.549814000000 5.523470000000 -6.784594000000
H -1.885366000000 6.924878000000 -7.667959000000
H -1.547635000000 5.268561000000 -8.239095000000
C 4.545707000000 4.169682000000 -6.225748000000
H 5.531509000000 3.925096000000 -5.800890000000
H 4.507103000000 3.776789000000 -7.254173000000
H 4.450652000000 5.266003000000 -6.271740000000
$end

$rem
JOBTYPE EDA
SCF_PRINT_FRGM TRUE
EDA2 1
EDA_PCT_A TRUE
EDA_CLS_DISP TRUE
FRGM_METHOD GIA
FRGM_LPCORR RS_EXACT_SCF
SYMMETRY FALSE
SYM_IGNORE TRUE
METHOD M06
BASIS def2-SVP
ECP def2-ECP
DFT_D D3
EDA_COVP TRUE
EDA_PRINT_COVP 2
MAKE_CUBE_FILES TRUE
PLOTS TRUE
MAX_SCF_CYCLES 200
SCF_ALGORITHM DIIS
MAX_SCF_CYCLES 150
MEM_TOTAL 30000
MEM_STATIC 2000
SOLVENT_METHOD smd
$end

$plots
grid_points 100 100 100
$end

$smx
solvent THF
$end

Thanks

Q-Chem fatal error
#2

I ran this input file successfully with Q-Chem 5.4.2, runs to completion. I also don’t find that precise error message anywhere in the code, suggesting perhaps a bug that has been fixed. Can you please post a little more of the Q-Chem output before the crash so that we can see what is going on?

#3

Hello Jherbert,

Sure thing!

Done Job For Fragment 1
calling putenv with: QCSCRATCH=/expanse/lustre/scratch/rayz97/temp_project/job_8067888
calling putenv with: QCTHREADS=30
Doing General SCFMI
Doing General SCFMI
Exchange: 0.2700 Hartree-Fock + 1.0000 M06
Correlation: 1.0000 M06
Using SG-3 standard quadrature grid
Dispersion: Grimme D3
performing fock build before subrot guess so that all methods have access
Printing subrot_rss_gen Method Specification:
Occupied Subspace Method is 0
Virtual Subspace Method is 0
Printing subrot_rss_gen Method Specification:
Occupied Subspace Method is 0
Virtual Subspace Method is 0
Energy of the spin-aligned supersystem initial determinant -4327.056285003309
Solvent energy: -0.048070376385
Energy of the spin-aligned supersystem initial determinant w/o solvent -4327.008214626925
Exchange: 0.2700 Hartree-Fock + 1.0000 M06
Correlation: 1.0000 M06
Using SG-3 standard quadrature grid
Dispersion: Grimme D3
Hartree-Fock
Doing General SCFMI
Doing General SCFMI
Exchange: 0.2700 Hartree-Fock + 1.0000 M06
Correlation: 1.0000 M06
Using SG-3 standard quadrature grid
Dispersion: Grimme D3
performing fock build at end of subrot_u(subrot,subrot) constructor so that all methods have access
<S^2> of initial wavefunction before polarization iterations 0.7546842107
Beginning iterations for the determination of the polarized wavefunction
Energy prior to optimization (guess energy) = -4327.056284965167
Begin Timing: Total SCF Calculation
begin iterations for algorithm: DIIS
the SCF tolerance is set to 1.00e-08
1 -4327.0562849652 2.02e-04 00000
2 -4327.0906409913 1.55e-04 00000
3 -4327.0957794162 1.85e-04 00000
4 -4327.1052118911 1.07e-04 00000
5 -4327.1124021417 5.66e-05 00000
6 -4327.1173381651 3.60e-05 00000
7 -4327.1214910170 2.85e-05 00000
8 -4327.1236849588 2.43e-05 00000
9 -4327.1249533938 2.08e-05 00000
10 -4327.1259134143 1.56e-05 00000
11 -4327.1268742767 1.31e-05 00000
12 -4327.1277123100 9.85e-06 00000
13 -4327.1282421187 7.78e-06 00000
14 -4327.1285937115 6.73e-06 00000
15 -4327.1288402296 4.27e-06 00000
16 -4327.1289654140 2.83e-06 00000
17 -4327.1290016907 1.72e-06 00000
18 -4327.1290215858 1.30e-06 00000
19 -4327.1290276035 8.60e-07 00000
20 -4327.1290305339 6.07e-07 00000
21 -4327.1290319250 4.87e-07 00000
22 -4327.1290327317 3.25e-07 00000
23 -4327.1290333582 2.75e-07 00000
24 -4327.1290338064 2.07e-07 00000
25 -4327.1290340786 1.45e-07 00000
26 -4327.1290341194 1.25e-07 00000
27 -4327.1290341931 8.66e-08 00000
28 -4327.1290343328 6.41e-08 00000
29 -4327.1290343498 4.92e-08 00000
30 -4327.1290342846 3.95e-08 00000
31 -4327.1290342227 2.52e-08 00000
32 -4327.1290343174 1.23e-08 00000
33 -4327.1290342401 8.48e-09 00000 Convergence criterion met
write pseudocanon occs and virts to disk
write pseudocanon orbital energies to disk
Timing for Total SCF: 2837.00s (wall), 49752.50s (cpu)
<S^2> of polarized wavefunction after polarization iterations 1.4794795615
relaxing scfmi constraint to allow all rotations in the entire AO span
relaxing scfmi constraint to allow all rotations in the entire AO span
Doing General SCFMI
Doing General SCFMI
Exchange: 0.2700 Hartree-Fock + 1.0000 M06
Correlation: 1.0000 M06
Using SG-3 standard quadrature grid
Dispersion: Grimme D3
performing fock build at end of subrot_u(subrot,subrot) constructor so that all methods have access
<S^2> of polarized wavefunction before ct-allowed iterations 1.4794795617
Beginning iterations for the determination of the CT-Allowed wavefunction
Energy prior to optimization (guess energy) = -4327.129034274451
Begin Timing: Total SCF Calculation
begin iterations for algorithm: DIIS
the SCF tolerance is set to 1.00e-08
1 -4327.1290342745 3.76e-04 00000
2 -4327.2474082753 1.01e-04 00000
3 -4327.2520180393 2.12e-04 00000
4 -4327.2656426432 6.61e-05 00000
5 -4327.2696502354 4.64e-05 00000
6 -4327.2721715056 1.65e-05 00000
7 -4327.2729761886 1.31e-05 00000
8 -4327.2734455218 6.89e-06 00000
9 -4327.2736220137 5.12e-06 00000
10 -4327.2737057993 3.88e-06 00000
11 -4327.2737400031 1.96e-06 00000
12 -4327.2737503025 9.45e-07 00000
13 -4327.2737540674 6.53e-07 00000
14 -4327.2737564132 5.00e-07 00000
15 -4327.2737582434 4.05e-07 00000
16 -4327.2737597887 3.18e-07 00000
17 -4327.2737610026 2.31e-07 00000
18 -4327.2737616652 1.76e-07 00000
19 -4327.2737620033 1.38e-07 00000
20 -4327.2737620898 7.56e-08 00000
21 -4327.2737621270 3.05e-08 00000
22 -4327.2737620824 1.92e-08 00000
23 -4327.2737621746 1.14e-08 00000
24 -4327.2737620803 9.95e-09 00000 Convergence criterion met
write pseudocanon occs and virts to disk
write pseudocanon orbital energies to disk
Timing for Total SCF: 2014.00s (wall), 36196.02s (cpu)
<S^2> of ct_allowed wavefunction after ct-allowed iterations 0.7803423132
Perturbative CTA
Number of subspaces: 2
Appending the orthogonal complement of the current SCFMI subspace
Number of subspaces in CTA becomes 2
Number of subspaces: 2
Appending the orthogonal complement of the current SCFMI subspace
Number of subspaces in CTA becomes 2
Doing General SCFMI
Doing General SCFMI
Exchange: 0.2700 Hartree-Fock + 1.0000 M06
Correlation: 1.0000 M06
Using SG-3 standard quadrature grid
Dispersion: Grimme D3
performing fock build at end of subrot_u(subrot,subrot) constructor so that all methods have access
Doing General SCFMI
Exchange: 0.2700 Hartree-Fock + 1.0000 M06
Correlation: 1.0000 M06
Using SG-3 standard quadrature grid
Dispersion: Grimme D3
Doing General SCFMI
Exchange: 0.2700 Hartree-Fock + 1.0000 M06
Correlation: 1.0000 M06
Using SG-3 standard quadrature grid
Dispersion: Grimme D3
PCT method is RS
Max deviation in Fvo + FvvXvo - XvoFoo - XvoFovXvo is 4.58e-09
X in RS PT satisfies the quadratic equation
Max deviation in Fvo + FvvXvo - XvoFoo - XvoFovXvo is 1.13e-08

Q-Chem fatal error occurred in module libalmo/libalmo/scfmi/subrot_rss.C, line 566:

QUADRATIC EQUATION FOR X NOT SATISFIED IN RS PT

Please submit a crash report at Q-Chem Crash Reporter

#4

Your job is crashing in the orthogonalization step just before the COVP analysis. It’s also clear that this code has been updated between the version that you have and the present one (my output is slightly different, and I did find your error message in an older version of the source code). If you’re not trying to do COVP then you might bypass this and be okay but presumably that’s in your input for a reason.

Basically that error message (about Fvo + FvvXvo - XvoFoo - XvoFovXvo) is telling you that the matrix equations have not been solved to the required precision, not clear to me why not. Only suggestion that I have is to try changing value of THRESH to see if you can sneak around this. (That’s a shot in the dark.) Since SCF_CONVERGENCE=8, you are using THRESH=11 by default in this calculation.

Maybe someone at Q-Chem or someone more familiar with the EDA code has better suggestions.

#5

Thank you for your suggestions. I will make an attempt at trying to get this calculation in the current version I am using and if all else fails I will just try in the latest version of Q-Chem.

#6

Would it be possible to ask for the output file?

my email is trevinoramon97@gmail.com

Thanks

#7

Hello ramon_trevino,

As jherbert pointed out, the related code has been improved for 5.4. With 5.4, you can manually set the convergence criteria through KONSCF (in $rem section) while the criteria is fixed at 10^(-8) in 5.3. I would recommend requesting a copy of 5.4 to confirm that the issue is fixed for 5.4. Feel free to hit us with an email (support@q-chem.com) if you have additional questions.

#8

FYI, “KONSCF” is aliased to SCF_CONVERGENCE in the user-facing code, so what Kuan-Yu really means is SCF_CONVERGENCE.