Wrong scf solution in the presence of effective fragments

Questions
#1

Hello,

I have successfully set up a calculation with user-defined effective fragments, as previously discussed. The system I am describing involves a small dye molecule (-2 1) in the ground state configuration, obtained using molecular dynamics simulation in an explicit organic solvent.

I created the QChem input with the dye molecule as the QM region, a 5A solvent (ionic liquid) sphere using EFs, and the remaining atoms as point charges. I was able to compute excited states using the RPA approach successfully for the dye molecule alone, as well as in the presence of only point charges.

However, even with just two EFs, my virtual MO orbitals have some negative values. I interpret this as an incorrect SCF solution, even though it is converged at defaults for excited state calculation (10-8). I have tried several different SCF algorithms, namely RCA_DIIS, DIIS, and DIIS_GDM, but the result remains the same.

I attempted to use molecular orbitals from one of the said calculations to start a NEWTON_CG, but it crashed with the following error: “Illegal MetDIIS in DIISController constructor”. I am unsure why I am receiving a DIIS-related error when I am attempting the Newton algorithm.

Additionally, I attempted an INTERNAL_STABILITY analysis. However, for reasons unknown (possibly not implemented in the presence of EFPs), the calculation executed without running the stability analysis.

I am currently at a loss for ideas on how to correctly calculate this system. It appears that the EFPs are somehow affecting the SCF space, resulting in a convergence at the triplet ground state.

I would greatly appreciate any advice to help me progress with my work. I have attached an input I used at the start, without modifying much of the REM parameters.

Thank you.

$molecule
-2 1
C      32.327000   30.492001   29.794001
O      32.702999   30.445000   28.631001
C      31.875000   31.677999   30.542000
C      30.976000   30.884001   31.261999
O      29.988001   31.222000   31.830000
C      31.238001   29.674999   30.403000
C      31.573000   25.892000   26.868000
H      31.724001   25.202000   26.044001
C      32.674999   26.535000   27.497999
H      33.729000   26.455999   27.249001
C      32.334999   27.349001   28.613001
H      33.162998   27.826000   29.135000
N      31.070000   27.684999   29.011000
C      30.014999   27.093000   28.329000
C      30.261999   26.148001   27.302999
H      29.471001   25.551001   26.858000
C      30.541000   28.551001   29.907000
C      29.212999   28.441999   29.997999
C      28.766001   27.570000   29.032000
C      27.378000   27.194000   28.648001
H      26.747999   27.385000   29.524000
H      27.344999   26.125999   28.405001
H      27.098000   27.764999   27.754999
C      28.587000   28.375000   31.202000
C      31.625000   36.557999   32.445999
H      31.462999   37.535999   32.888000
C      31.430000   35.355999   33.102001
H      31.188999   35.390999   34.160000
C      31.575001   34.091999   32.411999
H      31.313999   33.167999   32.924999
N      31.615000   34.125000   31.072001
C      31.945000   35.317001   30.408001
C      31.898001   36.520000   31.051001
H      31.896999   37.415001   30.434999
C      31.802999   33.061001   30.233000
C      31.690001   33.618999   28.988001
C      31.743999   35.006001   29.097000
C      31.101999   35.853001   28.035000
H      31.475000   35.632000   27.028000
H      31.379999   36.887001   28.268999
H      30.028999   35.712002   28.208000
C      28.986000   27.521000   32.251999
H      29.889000   26.948000   32.062000
C      28.202999   27.417999   33.408001
H      28.496000   26.666000   34.133999
C      27.059000   28.188999   33.569000
C      26.677999   28.951000   32.500999
H      25.714001   29.445000   32.575001
C      27.398001   29.125000   31.362000
H      27.165001   29.871000   30.608000
O      26.350000   28.246000   34.743000
C      26.809999   27.393999   35.833000
H      26.841000   26.334999   35.550999
H      27.837000   27.712999   36.049000
C      25.764999   27.681000   36.952000
H      25.799999   26.952000   37.770000
H      24.775999   27.479000   36.523998
C      25.783001   29.200001   37.513000
H      26.024000   30.009001   36.813999
H      24.716000   29.445000   37.575001
C      26.506001   29.549000   38.896999
H      26.110001   28.815001   39.609001
H      26.365000   30.612000   39.127998
S      28.299000   29.263000   38.862000
O      28.687000   29.612000   40.216999
O      28.862000   30.122999   37.827000
O      28.415001   27.847000   38.500000
C      31.306999   32.866001   27.761999
C      31.899000   33.101002   26.527000
H      32.580002   33.936001   26.386000
C      31.559000   32.212002   25.441000
H      32.171001   32.358002   24.556000
C      30.816999   31.002001   25.631001
C      30.125000   30.863001   26.834999
H      29.326000   30.139999   26.966000
C      30.243000   31.840000   27.851999
H      29.622000   31.794001   28.742001
O      30.534000   30.084000   24.645000
C      31.099001   30.122999   23.319000
H      32.193001   30.208000   23.337000
H      30.757000   31.028999   22.802000
C      30.655001   28.906000   22.509001
H      29.586000   28.989000   22.282000
H      31.170000   28.806999   21.547001
C      30.593000   27.563000   23.349001
H      29.657000   27.534000   23.919001
H      31.500999   27.441000   23.952000
C      30.371000   26.421000   22.393000
H      29.302000   26.306999   22.173000
H      30.875000   26.716000   21.465000
S      30.889000   24.841000   22.861000
O      30.437000   24.100000   21.688999
O      32.313999   25.052999   22.974001
O      30.214001   24.555000   24.087000
$end

$efp_fragments
octanoate_u
   A01C1     26.433001    34.216000    44.223999
   A02H2     27.096001    33.445999    43.813999
   A03H3     27.014000    35.092999    44.532001
octanoate_u
   A01C1     37.349998    30.384001    20.291000
   A02H2     37.019001    31.427999    20.330999
   A03H3     37.535000    30.033001    21.312000
octanoate_u
   A01C1     36.212002    31.044001    30.388000
   A02H2     36.812000    31.209000    29.486000
   A03H3     36.743000    31.179001    31.337000
choline_u
   A01C1     27.985001    32.818001    35.766998
   A02H2     28.938999    33.348000    35.743999
   A03H3     27.039000    33.351002    35.868000
choline_u
   A01C1     33.973999    41.403999    27.903000
   A02H2     34.080002    41.214001    26.834000
   A03H3     32.993999    41.200001    28.337000
octanoate_u
   A01C1     39.362999    29.687000    28.285000
   A02H2     40.325001    29.344999    28.683001
   A03H3     38.618000    29.587999    29.083000
choline_u
   A01C1     35.980000    18.576000    27.514999
   A02H2     35.801998    17.726000    26.855000
   A03H3     36.667999    19.271000    27.028999
octanoate_u
   A01C1     37.375999    33.118999    34.810001
   A02H2     37.174000    33.813000    35.633999
   A03H3     37.665001    32.167999    35.271999
choline_u
   A01C1     31.961000    30.077000    16.712999
   A02H2     32.436001    29.171000    17.091999
   A03H3     30.879000    30.059999    16.570999
octanoate_u
   A01C1     26.441000    27.900999    45.688000
   A02H2     26.062000    28.441999    46.563999
   A03H3     27.535000    27.974001    45.705002
choline_u
   A01C1     22.156000    30.417000    31.671000
   A02H2     23.065001    30.709000    31.143000
   A03H3     22.329000    30.379000    32.748001
octanoate_u
   A01C1     23.812000    25.427000    32.028000
   A02H2     23.730000    25.712999    33.082001
   A03H3     23.334000    24.445000    31.933001
octanoate_u
   A01C1     32.872002    42.282001    34.841000
   A02H2     31.981001    42.546001    35.422001
   A03H3     32.766998    41.202000    34.685001
choline_u
   A01C1     25.044001    36.417000    27.851000
   A02H2     23.996000    36.672001    27.688000
   A03H3     25.318001    37.147999    28.614000
choline_u
   A01C1     26.480000    39.012001    39.951000
   A02H2     26.577000    38.173000    40.641998
   A03H3     27.476000    39.451000    39.875999
choline_u
   A01C1     25.458000    22.483000    24.693001
   A02H2     25.045000    23.336000    24.150999
   A03H3     25.165001    21.610001    24.108000
choline_u
   A01C1     33.998001    36.259998    34.945000
   A02H2     34.185001    35.771000    35.902000
   A03H3     33.000000    36.685001    34.827000
choline_u
   A01C1     38.063000    34.070000    26.320999
   A02H2     37.368999    33.252998    26.115999
   A03H3     38.966999    33.976002    25.719000
choline_u
   A01C1     33.139999    36.120998    22.900000
   A02H2     33.679001    37.042999    22.674000
   A03H3     33.644001    35.541000    23.674000
octanoate_u
   A01C1     31.027000    19.851000    22.677000
   A02H2     30.745001    19.055000    23.375000
   A03H3     30.101999    20.343000    22.356001
octanoate_u
   A01C1     25.444000    23.229000    19.197001
   A02H2     25.187000    22.750000    18.245001
   A03H3     25.962999    24.160000    18.938999
octanoate_u
   A01C1     35.082001    31.579000    24.141001
   A02H2     34.181000    31.091000    23.752001
   A03H3     35.787998    31.013000    23.523001
octanoate_u
   A01C1     23.027000    21.625999    30.333000
   A02H2     22.586000    22.273001    31.100000
   A03H3     24.121000    21.576000    30.318001
octanoate_u
   A01C1     22.172001    27.096001    38.048000
   A02H2     22.938000    26.658001    37.397999
   A03H3     22.681999    27.983999    38.437000
choline_u
   A01C1     30.077000    26.106001    36.556000
   A02H2     29.344999    26.691000    35.997002
   A03H3     30.039000    25.063000    36.237000
choline_u
   A01C1     30.884001    32.574001    44.119999
   A02H2     30.900000    31.973000    45.030998
   A03H3     29.917999    32.532001    43.615002
octanoate_u
   A01C1     22.312000    28.455999    26.723000
   A02H2     21.434999    28.924000    27.184999
   A03H3     22.736000    29.139999    25.979000
octanoate_u
   A01C1     25.506001    38.021000    31.495001
   A02H2     25.809000    37.320000    30.709000
   A03H3     24.812000    38.804001    31.167999
octanoate_u
   A01C1     23.014999    23.686001    35.764999
   A02H2     23.125999    24.774000    35.838001
   A03H3     23.325001    23.301001    34.786999
choline_u
   A01C1     27.601000    29.577999    24.195999
   A02H2     28.485001    29.822001    23.606001
   A03H3     28.020000    29.243000    25.146000
octanoate_u
   A01C1     25.313999    41.598999    34.408001
   A02H2     25.888000    41.706001    33.480000
   A03H3     24.993999    40.556000    34.507999
choline_u
   A01C1     32.896999    29.368999    44.321999
   A02H2     32.117001    29.363001    45.084999
   A03H3     33.570000    30.198000    44.544998
choline_u
   A01C1     28.629000    36.513000    18.041000
   A02H2     29.694000    36.382000    18.239000
   A03H3     28.308001    36.043999    17.110001
choline_u
   A01C1     31.611000    33.526001    40.620998
   A02H2     32.259998    33.021999    41.338001
   A03H3     31.931999    34.518002    40.299000
choline_u
   A01C1     35.313999    26.069000    22.952000
   A02H2     35.507999    26.639000    23.862000
   A03H3     35.641998    26.635000    22.079000
octanoate_u
   A01C1     38.236000    24.739000    30.292000
   A02H2     38.348999    24.617001    29.208000
   A03H3     37.886002    23.761000    30.639999
octanoate_u
   A01C1     21.864000    31.652000    44.653999
   A02H2     22.958000    31.704000    44.640999
   A03H3     21.775000    30.766001    45.293999
octanoate_u
   A01C1     27.610001    28.565001    18.933001
   A02H2     27.844000    29.427999    19.566000
   A03H3     28.469999    27.886000    18.915001
octanoate_u
   A01C1     25.236000    39.186001    21.962000
   A02H2     25.298000    39.356998    20.881001
   A03H3     24.417999    39.791000    22.370001
choline_u
   A01C1     34.403999    21.954000    33.858002
   A02H2     34.383999    22.478001    32.901001
   A03H3     34.741001    20.917000    33.866001
choline_u
   A01C1     31.733999    24.754999    30.677000
   A02H2     32.573002    25.313999    30.260000
   A03H3     30.840000    25.372000    30.566999
octanoate_u
   A01C1     31.037001    39.411999    39.873001
   A02H2     31.622000    40.306000    40.119999
   A03H3     31.782000    38.608002    39.886002
choline_u
   A01C1     28.429001    25.601000    43.033001
   A02H2     27.796000    26.306000    43.573002
   A03H3     27.902000    24.646000    43.026001
choline_u
   A01C1     31.197001    22.743000    18.752001
   A02H2     31.846001    21.872999    18.643999
   A03H3     30.993999    22.886000    19.813999
$end

$rem
   JOBTYPE           SP
   METHOD            wB97X-D
   BASIS             GEN
   PURECART          1111
   CIS_N_ROOTS       5
   CIS_SINGLETS      true
   CIS_TRIPLETS      false
   RPA               true
   EFP               true
   EFP_FRAGMENTS_ONLY   false
   EFP_COORD_XYZ     true
   EFP_DIRECT_POLARIZATION_DRIVER   true
   THRESH            14
   VARTHRESH         0
   INCFOCK           0
   SCF_ALGORITHM     DIIS
   #SCF_GUESS         READ
   MAX_SCF_CYCLES    1000
   MEM_TOTAL         16000   ! Specify memory in megabytes (16 GB)
$end

$basis
H     0
def2-SVPD
****
C     0
S   5   1.00
   1238.4016938              0.54568832082D-02
    186.29004992             0.40638409211D-01
     42.251176346            0.18025593888
     11.676557932            0.46315121755
      3.5930506482           0.44087173314
S   1   1.00
      0.40245147363          1.0000000
S   1   1.00
      0.13090182668          1.0000000
S   1   1.00
      0.67053540256D-01      1.0000000
P   3   1.00
      9.4680970621           0.38387871728D-01
      2.0103545142           0.21117025112
      0.54771004707          0.51328172114
P   1   1.00
      0.15268613795          1.0000000
D   1   1.00
      0.8000000              1.0000000
****
N     0
S   5   1.00
   1712.8415853             -0.53934125305D-02
    257.64812677            -0.40221581118D-01
     58.458245853           -0.17931144990
     16.198367905           -0.46376317823
      5.0052600809          -0.44171422662
S   1   1.00
      0.58731856571          1.0000000
S   1   1.00
      0.18764592253          1.0000000
S   1   1.00
      0.96171241529D-01      1.0000000
P   3   1.00
     13.571470233           -0.40072398852D-01
      2.9257372874          -0.21807045028
      0.79927750754         -0.51294466049
P   1   1.00
      0.21954348034          1.0000000
D   1   1.00
      1.0000000              1.0000000
****
O     0
S   5   1.00
   2266.1767785             -0.53431809926D-02
    340.87010191            -0.39890039230D-01
     77.363135167           -0.17853911985
     21.479644940           -0.46427684959
      6.6589433124          -0.44309745172
S   1   1.00
      0.80975975668          1.0000000
S   1   1.00
      0.25530772234          1.0000000
S   1   1.00
      0.76572453250D-01      1.0000000
P   3   1.00
     17.721504317            0.43394573193D-01
      3.8635505440           0.23094120765
      1.0480920883           0.51375311064
P   1   1.00
      0.27641544411          1.0000000
P   1   1.00
      0.69002276350D-01      1.0000000
D   1   1.00
      1.2000000              1.0000000
****
S     0
S   5   1.00
   9184.9303010             -0.22294387756D-02
   1381.5105503             -0.16683029937D-01
    313.87147580            -0.75262436116D-01
     88.053870623           -0.19376827038
     27.039914905           -0.17718020803
S   3   1.00
     45.648731303           -0.10736062573
      4.9664522326           0.65066293018
      2.0116242047           0.59712155354
S   1   1.00
      0.35661077013          1.0000000
S   1   1.00
      0.13507221477          1.0000000
S   1   1.00
      0.44999823506D-01      1.0000000
P   5   1.00
    261.98233439            -0.92729929822D-02
     61.306894736           -0.66547669241D-01
     19.103729887           -0.24828595903
      6.6567720378          -0.48703847402
      2.3959635161          -0.39337850312
P   1   1.00
      0.61776161679          1.0000000
P   1   1.00
      0.16993376871          1.0000000
P   1   1.00
      0.41706793676D-01      1.0000000
D   1   1.00
      0.5500000              1.0000000
****
$end
#2

Why are you convinced that EFPs should not affect the electronic structure? The point of QM/MM is that the QM solution that you get may depend on the MM environment.

Negative virtual orbitals do not necessarily indicate a problem, although have you taken a look at them? I would expect functional and basis set to have a (potentially) large impact on whether bound virtuals are obtained, EFP not so much. I could be wrong about that, but if the gas-phase calculation produces no bound virtuals but the QM/EFP calculation does, this suggests that the EFP molecules are stabilizing a bound anion. Possibly, if (for example) the QM/EFP has a much larger dipole moment than the gas-phase QM but that seems unlikely for this system.

By the way, there is no way for your calculation to converge to a triplet since you’ve set the multiplicity to singlet. So, I think you don’t completely understand the electronic structure of the singlet that you’re getting, and trying to understand that may provide insight about what step to take next.

#3

I expected my electronic structure to be changed when the EFP contribution is added. Indeed I calculated triplet excitation using TDDFT/TDA and found out that all energies are negative. This made me realize that the converged electronic structure is not the global minima. As you put forth, the gas phase anion is well stabilized by the EFPs and as a result this is possible (There are close to 20 cation/anion pairs sourounding the QM region). So what I don’t understand is what else I could do to obtain the global minima since for TDDFT/RPA you need the B matrix to be positive. I have tried changing the starting configuration, scf algorithm and functional (wB97x-V).

#4

I can’t run this job without the EFP library files. What do the orbital energies look like? Have you tried some kind of analogous QM/MM calculation with external point charges, just to see if the problem is EFP-specific?

#5

I attempted a mixed basis approach with STO-3G to express all solvent although I had trouble converging the scf earlier.

When I did all solvent with single point charges scf converged and I obtained all excited states expected. Following is a snippet when EFPs are not included.

 Alpha MOs
 -- Occupied --
-89.0336 -88.9870 -19.1763 -19.1364 -19.0600 -19.0580 -19.0307 -19.0267
-19.0266 -19.0047 -18.9864 -18.9797 -14.3959 -14.3879 -10.2458 -10.2397
-10.2280 -10.2260 -10.2212 -10.2184 -10.2181 -10.2153 -10.2123 -10.2114
-10.2024 -10.1986 -10.1964 -10.1950 -10.1933 -10.1901 -10.1863 -10.1829
-10.1822 -10.1812 -10.1777 -10.1702 -10.1694 -10.1679 -10.1674 -10.1670
-10.1625 -10.1624 -10.1622 -10.1610 -10.1596 -10.1594 -10.1575 -10.1548
-10.1542 -10.1533 -10.1526 -10.1511 -10.1433 -10.1411 -10.1174 -10.0968
 -8.0526  -8.0079  -6.0070  -6.0068  -6.0066  -5.9626  -5.9620  -5.9612
 -1.0347  -1.0249  -0.9989  -0.9934  -0.9891  -0.9877  -0.9705  -0.9481
 -0.8774  -0.8701  -0.8536  -0.8434  -0.8370  -0.8183  -0.8123  -0.8031
 -0.7903  -0.7881  -0.7834  -0.7727  -0.7422  -0.7350  -0.7254  -0.7177
 -0.7157  -0.7043  -0.7029  -0.6902  -0.6878  -0.6747  -0.6628  -0.6540
 -0.6472  -0.6413  -0.6123  -0.6066  -0.5963  -0.5895  -0.5809  -0.5696
 -0.5622  -0.5519  -0.5482  -0.5426  -0.5255  -0.5249  -0.5165  -0.5056
 -0.5008  -0.4931  -0.4810  -0.4760  -0.4707  -0.4675  -0.4614  -0.4585
 -0.4547  -0.4435  -0.4413  -0.4371  -0.4324  -0.4310  -0.4279  -0.4221
 -0.4175  -0.4146  -0.4128  -0.4116  -0.4061  -0.4007  -0.3995  -0.3981
 -0.3939  -0.3931  -0.3908  -0.3864  -0.3818  -0.3807  -0.3794  -0.3757
 -0.3697  -0.3678  -0.3674  -0.3653  -0.3650  -0.3625  -0.3619  -0.3518
 -0.3514  -0.3477  -0.3427  -0.3412  -0.3345  -0.3325  -0.3300  -0.3281
 -0.3275  -0.3237  -0.3221  -0.3206  -0.3179  -0.3144  -0.3141  -0.3058
 -0.3040  -0.3021  -0.2998  -0.2935  -0.2897  -0.2872  -0.2834  -0.2819
 -0.2807  -0.2780  -0.2665  -0.2601  -0.2559  -0.2546  -0.2441  -0.2342
 -0.2341  -0.2301  -0.2260  -0.2227  -0.2000  -0.1985  -0.1943  -0.1925
 -0.1899  -0.1897  -0.1866  -0.1865  -0.1817  -0.1763  -0.1732  -0.1614
 -0.1579  -0.1448  -0.1421  -0.1390  -0.1356  -0.1299  -0.1232  -0.1074
 -0.0869
 -- Virtual --
  0.0913   0.1043   0.1334   0.1520   0.1580   0.1667   0.1743   0.1813
  0.1887   0.1941   0.1967   0.2021   0.2058   0.2101   0.2124   0.2168
  0.2202   0.2222   0.2233   0.2278   0.2294   0.2335   0.2354   0.2452
  0.2499   0.2542   0.2582   0.2608   0.2638   0.2682   0.2704   0.2724
  0.2776   0.2789   0.2834   0.2849   0.2864   0.2899   0.2906   0.2922
  0.2950   0.2968   0.3017   0.3038   0.3066   0.3092   0.3105   0.3134
  0.3164   0.3174   0.3199   0.3217   0.3233   0.3241   0.3277   0.3304
  0.3322   0.3331   0.3369   0.3382   0.3396   0.3431   0.3438   0.3451
  0.3494   0.3496   0.3520   0.3548   0.3563   0.3584   0.3610   0.3618
  0.3630   0.3673   0.3681   0.3694   0.3705   0.3726   0.3738   0.3750
  0.3764   0.3791   0.3803   0.3817   0.3846   0.3865   0.3893   0.3920
  0.3934   0.3950   0.3977   0.3991   0.4018   0.4043   0.4047   0.4054
  0.4081   0.4093   0.4115   0.4131   0.4147   0.4157   0.4165   0.4180
  0.4194   0.4213   0.4242   0.4273   0.4276   0.4308   0.4324   0.4337
  0.4351   0.4362   0.4380   0.4384   0.4402   0.4436   0.4439   0.4468
  0.4477   0.4488   0.4496   0.4517   0.4557   0.4560   0.4579   0.4593
  0.4598   0.4605   0.4618   0.4634   0.4647   0.4679   0.4690   0.4707

But when EFPs are included I get large number of negative virtual orbital energies.

 Alpha MOs
 -- Occupied --
-91.5932 -91.5838 -21.8093 -21.7747 -21.7624 -21.7477 -21.7431 -21.7425
-21.7235 -21.7116 -21.7083 -21.6717 -16.9784 -16.9498 -12.9334 -12.8637
-12.8508 -12.8333 -12.8302 -12.8251 -12.8242 -12.8211 -12.8208 -12.8202
-12.8078 -12.8073 -12.8058 -12.8039 -12.7965 -12.7925 -12.7900 -12.7826
-12.7745 -12.7737 -12.7732 -12.7722 -12.7624 -12.7582 -12.7577 -12.7511
-12.7511 -12.7496 -12.7447 -12.7391 -12.7144 -12.7127 -12.7052 -12.6979
-12.6967 -12.6833 -12.6820 -12.6731 -12.6607 -12.6524 -12.5954 -12.5930
-12.5922 -10.6224 -10.6125  -8.5767  -8.5755  -8.5749  -8.5671  -8.5654
 -8.5644  -6.3933  -6.0826  -4.8776  -4.0000  -3.8397  -3.7336  -3.6539
 -3.6368  -3.6163  -3.6111  -3.6028  -3.5910  -3.5746  -3.5681  -3.5454
 -3.5377  -3.4981  -3.4837  -3.4814  -3.4806  -3.4574  -3.4334  -3.4218
 -3.3994  -3.3959  -3.3608  -3.3485  -3.3400  -3.3247  -3.3128  -3.2978
 -3.2870  -3.2737  -3.2625  -3.2508  -3.2403  -3.2169  -3.2152  -3.2111
 -3.2074  -3.2007  -3.1938  -3.1735  -3.1698  -3.1589  -3.1560  -3.1418
 -3.1333  -3.1266  -3.1202  -3.1097  -3.1065  -3.0984  -3.0882  -3.0837
 -3.0812  -3.0591  -3.0503  -3.0472  -3.0437  -3.0370  -3.0357  -3.0301
 -3.0162  -3.0108  -3.0029  -3.0001  -2.9912  -2.9872  -2.9862  -2.9820
 -2.9746  -2.9705  -2.9638  -2.9629  -2.9582  -2.9564  -2.9520  -2.9513
 -2.9492  -2.9427  -2.9378  -2.9370  -2.9335  -2.9316  -2.9283  -2.9221
 -2.9164  -2.9134  -2.9116  -2.9066  -2.9010  -2.8942  -2.8921  -2.8880
 -2.8813  -2.8796  -2.8762  -2.8737  -2.8652  -2.8552  -2.8546  -2.8470
 -2.8436  -2.8399  -2.8378  -2.8300  -2.8262  -2.8233  -2.8222  -2.8202
 -2.8192  -2.8156  -2.8107  -2.8079  -2.8049  -2.8012  -2.7953  -2.7905
 -2.7870  -2.7787  -2.7775  -2.7764  -2.7727  -2.7711  -2.7643  -2.7628
 -2.7621  -2.7554  -2.7506  -2.7495  -2.7433  -2.7396  -2.7391  -2.7365
 -2.7291  -2.7244  -2.7238  -2.7234  -2.7218  -2.7100  -2.7035  -2.7011
 -2.6798
 -- Virtual --
 -2.5846  -2.5656  -2.5617  -2.5534  -2.5362  -2.5289  -2.5189  -2.5147
 -2.5119  -2.5080  -2.5001  -2.4947  -2.4798  -2.4721  -2.4519  -2.4353
 -2.4225  -2.4190  -2.3958  -2.3756  -2.3680  -2.3501  -2.3405  -2.3348
 -2.3300  -2.3180  -2.3137  -2.3076  -2.2947  -2.2843  -2.2742  -2.2646
 -2.2582  -2.2440  -2.2312  -2.2213  -2.2126  -2.2028  -2.1880  -2.1827
 -2.1728  -2.1675  -2.1671  -2.1607  -2.1506  -2.1484  -2.1424  -2.1403
 -2.1309  -2.1307  -2.1262  -2.1190  -2.1151  -2.1132  -2.1121  -2.1072
 -2.1069  -2.1023  -2.1004  -2.0939  -2.0927  -2.0920  -2.0905  -2.0836
 -2.0795  -2.0770  -2.0754  -2.0741  -2.0712  -2.0690  -2.0677  -2.0603
 -2.0585  -2.0579  -2.0509  -2.0497  -2.0432  -2.0419  -2.0398  -2.0368
 -2.0357  -2.0345  -2.0337  -2.0282  -2.0258  -2.0233  -2.0217  -2.0188
 -2.0180  -2.0149  -2.0123  -2.0093  -2.0048  -2.0045  -1.9993  -1.9963
 -1.9943  -1.9919  -1.9900  -1.9865  -1.9859  -1.9846  -1.9807  -1.9774
 -1.9764  -1.9731  -1.9701  -1.9683  -1.9660  -1.9641  -1.9603  -1.9588
 -1.9560  -1.9544  -1.9493  -1.9480  -1.9458  -1.9430  -1.9417  -1.9396
 -1.9385  -1.9349  -1.9318  -1.9310  -1.9300  -1.9286  -1.9272  -1.9255
 -1.9212  -1.9207  -1.9184  -1.9180  -1.9153  -1.9139  -1.9108  -1.9093
 -1.9086  -1.9066  -1.9053  -1.9018  -1.8994  -1.8980  -1.8953  -1.8935
 -1.8909  -1.8901  -1.8887  -1.8861  -1.8847  -1.8826  -1.8806  -1.8777
 -1.8766  -1.8749  -1.8724  -1.8703  -1.8695  -1.8638  -1.8624  -1.8610
 -1.8578  -1.8573  -1.8554  -1.8524  -1.8493  -1.8454  -1.8445  -1.8426
 -1.8419  -1.8388  -1.8353  -1.8341  -1.8327  -1.8293  -1.8278  -1.8250
 -1.8244  -1.8208  -1.8186  -1.8181  -1.8156  -1.8121  -1.8115  -1.8095

Thats why I assumed this is due to how EFPs perturb the quntum hamiltonian.

Furthermore, I have doubt about how I prepared the EFP parameters given that I didn’t include damping as given in the example for user-defined EFPs. Following is my sample input I used for one of the fragments.

!
! GAMESS input for EFP generation
!
 $contrl units=angs local=boys runtyp=makefp coord=cart icut=11 ICHARG=  1 $end
 $system timlim=99999 mwords=200 $end
 $scf dirscf=.t. soscf=.f. diis=.t. conv=1.0d-06 $end
 $basis gbasis=n31 ngauss=6 ndfunc=3
   diffs=.t. $end
 $stone
   bigexp=0.0
 $end
 $data
  octanoate C8H15O2- (GEOMETRY: MP2/cc-pVTZ)
c1
 C1      6.0    -1.44874    -0.39047    -0.00001
 H2      1.0    -1.40548    -1.04608    -0.87502
 H3      1.0    -1.40548    -1.04608     0.87501
 C4      6.0    -0.22836     0.52223    -0.00001
 H5      1.0    -0.27328     1.17906    -0.87536
 H6      1.0    -0.27329     1.17908     0.87532
 C7      6.0    -2.77387     0.36060    -0.00000
 H8      1.0    -2.81766     1.01615     0.87539
 H9      1.0    -2.81768     1.01612    -0.87542
 C10     6.0     1.09368    -0.23602    -0.00000
 H11     1.0     1.14365    -0.89610     0.86833
 H12     1.0     1.14366    -0.89613    -0.86831
 C13     6.0    -3.99144    -0.55467     0.00002
 H14     1.0    -3.94687    -1.20844    -0.87430
 H15     1.0    -3.94688    -1.20838     0.87437
 C16     6.0    -5.30695     0.21444    -0.00000
 H17     1.0    -5.37899     0.85471     0.87959
 H18     1.0    -6.16912    -0.45247     0.00000
 H19     1.0    -5.37898     0.85468    -0.87962
 C20     6.0     2.31343     0.66853    -0.00000
 H21     1.0     2.29708     1.32853     0.87266
 H22     1.0     2.29708     1.32851    -0.87269
 C23     6.0     3.68341    -0.07669     0.00000
 O24     8.0     4.68446     0.67936     0.00002
 O25     8.0     3.62761    -1.33249    -0.00001
 $end

Also I did try turning on internal stability with gen_scfman=true and default davison interations were not enough to converge.

All calculation I have done so far points to stabilization through efps and as a result not converging to a global minima.

I am happy to email you the two efp fragment files if that would help suggest more applicable alternatives.

Thanks!

#6

I don’t have any experience with generating the EFP parameters, so it’s not at all clear to me where this goes wrong, but clearly it has. Those virtual orbital energies are in Hartree, so your LUMO is -2.6 Eh or 70 eV below vacuum level, and you have hundreds of levels that are that strongly bound. What even does the LUMO look like? When you said “some bound virtuals” I thought maybe a few that were bound like creating a molecular anion, but this is totally (catastrophically) unrealistic. What is the dipole moment of your system? Could it be that your undamped EFP is creating huge dipole moments (in what is supposed to be classical solvent)? I might also be a bit wary of the use of diffuse functions.

#7

I have not saved the Fchk files in any of my calculations, so sadly I couldn’t visualize what the orbitals look like. I was only able to compare my solvated systems (water and ionic liquid) with respective to the vacuum system. I removed all diffuse D functions in the basis set to ease the convergence though I still wanted to have the S and P functions given the anionic nature of the molecule. The dipole moment of my vacuum system is around 20 debye which could be expected given the anionic nature and size of my molecule (94 atoms including O, N and S). However, in the presence of EFP waters + point charges and EFP ionic liquids + point charges the dipole moment jumps to a rather high values. Below snippet shows the virtual orbitals energies and dipole moment outputs for the three systems.

vacuum

 -0.2539  -0.2499  -0.2403  -0.2362  -0.2301  -0.2210  -0.2168  -0.2166
 -0.2148  -0.2032  -0.2000  -0.1992  -0.1922  -0.1892  -0.1844  -0.1754
 -0.1722  -0.1715  -0.1707  -0.1606  -0.1547  -0.1511  -0.1444  -0.1290
 -0.1154
 -- Virtual --
  0.0718   0.1264   0.1334   0.1533   0.1671   0.1785   0.1824   0.1867
  0.1900   0.1955   0.1993   0.2082   0.2104   0.2194   0.2220   0.2254
  0.2296   0.2322   0.2353   0.2389   0.2432   0.2440   0.2453   0.2484
  0.2510   0.2518   0.2541   0.2569   0.2585   0.2589   0.2607   0.2640
  0.2649   0.2706   0.2711   0.2734   0.2759   0.2782   0.2788   0.2806
cont...

 -----------------------------------------------------------------
                    Cartesian Multipole Moments
 -----------------------------------------------------------------
    Charge (ESU x 10^10)
                -9.6064
    Dipole Moment (Debye)
         X       7.8187      Y      16.6505      Z      -8.1627
       Tot      20.1246

5A EFP Water + point charges

   -0.37342    -0.37161    -0.36275    -0.35894    -0.33100    -0.32031
   -0.31147    -0.30784    -0.29702    -0.29218    -0.28811    -0.28462
   -0.28429
 -- Virtual --
   -0.11208    -0.10142    -0.06978    -0.06345    -0.04199    -0.03796
   -0.03719    -0.03653    -0.03131    -0.02204    -0.01190    -0.00948
   -0.00771    -0.00687    -0.00363    -0.00276    -0.00087     0.00342
    0.00896     0.01246     0.01295     0.01509     0.01673     0.01962
    0.02319     0.02529     0.02695     0.02727     0.02833     0.03064
    0.03253     0.03332     0.03510     0.03687     0.03977     0.03999
    0.04098     0.04323     0.04781     0.05323     0.05609     0.05660
cont....

 -----------------------------------------------------------------
                    Cartesian Multipole Moments
 -----------------------------------------------------------------
    Charge (ESU x 10^10)
                 0.0000
    Dipole Moment (Debye)
         X    -196.0009      Y      75.6479      Z      44.4294
       Tot     214.7392

5A EFP ionic liquid + point charges

 -0.6887  -0.6821  -0.6804  -0.6797  -0.6769  -0.6689  -0.6675  -0.6654
 -0.6641  -0.6571  -0.6527  -0.6484  -0.6427  -0.6402  -0.6392  -0.6346
 -0.6299  -0.6289  -0.6210  -0.6172  -0.6029  -0.6005  -0.5925  -0.5875
 -0.5706
 -- Virtual --
 -0.4963  -0.4802  -0.4582  -0.4463  -0.4368  -0.4355  -0.4304  -0.4187
 -0.4125  -0.3958  -0.3855  -0.3764  -0.3560  -0.3373  -0.3048  -0.3007
 -0.2657  -0.2582  -0.2518  -0.2424  -0.2379  -0.2194  -0.2147  -0.2033
 -0.1879  -0.1855  -0.1780  -0.1712  -0.1662  -0.1520  -0.1451  -0.1424
 -0.1355  -0.1344  -0.1316  -0.1281  -0.1233  -0.1196  -0.1167  -0.1052
 -0.1024  -0.0980  -0.0903  -0.0864  -0.0821  -0.0782  -0.0750  -0.0693
 -0.0670  -0.0645  -0.0623  -0.0601  -0.0566  -0.0545  -0.0496  -0.0463
 -0.0430  -0.0395  -0.0365  -0.0343  -0.0330  -0.0327  -0.0282  -0.0248
 -0.0246  -0.0226  -0.0208  -0.0193  -0.0165  -0.0111  -0.0086  -0.0063
 -0.0050  -0.0033  -0.0026  -0.0009   0.0013   0.0034   0.0056   0.0071
cont...

 -----------------------------------------------------------------
                    Cartesian Multipole Moments
 -----------------------------------------------------------------
    Charge (ESU x 10^10)
                -9.6064
    Dipole Moment (Debye)
         X    -369.5552      Y    -241.1487      Z    -342.5025
       Tot     558.5980

So as you would notice the dipole moment increases from 20 to 200 to 500 when changing the environment. The behaviour is the same in terms of dipole moment if you remove all EFP fragments and just use point charges. However, in such scenario virtual orbital are as expected, positive.

 -0.2807  -0.2780  -0.2665  -0.2601  -0.2559  -0.2546  -0.2441  -0.2342
 -0.2341  -0.2301  -0.2260  -0.2227  -0.2000  -0.1985  -0.1943  -0.1925
 -0.1899  -0.1897  -0.1866  -0.1865  -0.1817  -0.1763  -0.1732  -0.1614
 -0.1579  -0.1448  -0.1421  -0.1390  -0.1356  -0.1299  -0.1232  -0.1074
 -0.0869
 -- Virtual --
  0.0913   0.1043   0.1334   0.1520   0.1580   0.1667   0.1743   0.1813
  0.1887   0.1941   0.1967   0.2021   0.2058   0.2101   0.2124   0.2168
  0.2202   0.2222   0.2233   0.2278   0.2294   0.2335   0.2354   0.2452
  0.2499   0.2542   0.2582   0.2608   0.2638   0.2682   0.2704   0.2724
  0.2776   0.2789   0.2834   0.2849   0.2864   0.2899   0.2906   0.2922
cont...

 -----------------------------------------------------------------
                    Cartesian Multipole Moments
 -----------------------------------------------------------------
    Charge (ESU x 10^10)
                -9.6064
    Dipole Moment (Debye)
         X    -272.6213      Y    -271.0193      Z    -301.6716
       Tot     488.6508

So that’s why I am struggling to understand what really is going on here. I appreciate any input you could give.

Thanks!

#8

I’m not sure that I’ve ever seen a dipole moment that large. This is a liquid configuration? Even if individual molecules have sizable dipole moments, I would expect that even in a snapshot from the liquid, those would approximately cancel, not add to something like hundreds of Debye. This has to be something wrong with the EFP fragments, since the virtual orbital eigenvalues are much more typical with point charges. I would investigate the options for generating EFP fragments from GAMESS. It would also be interesting to visualize those virtual orbitals; do they put electrons onto the EFP fragments, or are the localized on the QM parts?

#9

Could you clarify what you meant as a liquid configuration? I did obtain the initial coordinates of the solvated system using a molecular dynamics snapshot. For water, I used the QChem library fragment and for the ionic liquid I generated the effective fragments separately for cation and anion. I followed the exact protocol for parameter generation except using damping function for screening of the charge term in the distributed multipole expansion.

I will also rerun a calculation to obtain an output to visualize orbitals.

#10

That is what I meant by “liquid configuration”- extracted from MD and thus representative of a liquid. Lack of screening of the charge term seems like it could be the culprit, yes?

#11

That was my thought as well. However, I did try some more calculations with screening charges for my anion in the liquid. Result is the same. I am still working to find a reasoning behind why my EFPs influence this kind of behaviour.

To further expand my thoughts could please clarify what I have shared below is a normal phenomena or not? As you would notice, scf has converged in to a rather higher energy than the one before the last step.

 Compute EFP polarization integrals 
   69   -3717.6052586465      1.51E-08

 Compute EFP polarization integrals 
   70   -3717.6052586526      1.38E-08

 Compute EFP polarization integrals 
   71   -3717.6052586608      1.20E-08

 Compute EFP polarization integrals 
   72   -3717.6052586712      1.11E-08

 Compute EFP polarization integrals 
   73   -3670.3592402290      9.56E-09 Convergence criterion met
#12

That is indeed odd. I have very little (approaching zero) experience with QM/EFP so I don’t know what to tell you at this point. You could try contacting Q-Chem support (please include your user-generated EFP files) to see if they can reproduce the issue or if something seems obviously amiss to them.

#13

Sadly, I think I found the issue that could be pertaining to all these problems.
It is that the current implimentation of Q-Chem does not incooperate CT interaction term in the libefp. Review by Gordon ( Phys. Chem. Chem. Phys. , 2019,21 , 16878-16888) specifically discusses there are up to 20% differences between the EFP and SAPT2 + 3 predicted interaction energies, with the charge transfer energy component having the largest percent differences. I believe given the highly ionic nature of my liquid solvent, EFP2 approach incorrectly models the interaction needed to drive the SCF to real global minima. With that been said I am looking to using other fragmentation approaches like density embedding methods.

#14

It’s true that Q-Chem’s implementation doesn’t have the CT terms and that probably does adversely affect the EFP-EFP interactions in a system such as yours. However, I’m skeptical that this accounts for what looks like a severe (EFP-induced) problem on the QM side.