Inconsistency for space 1e (XMS-RS2C)

[Clément S]

Dear Molpro users,

I am performing a XMS(10)-CASPT2 scan on a small system with Molpro (NO + O2) and I am facing an error that I don't understand. The first point run just fine, but the second point fail due to "Inconsistency for space 1e !!". Does anyone know what cause this, especially while it worked for the first point of the scan ? This error occurs both with Molpro 2021.3 and 2022.1.

Best regards,

Clément Soulié

Please see the output below:

 PROGRAM * RS2C (Multireference RS Perturbation Theory)     Authors: H.-J. Werner (1993), P. Celani (1998)

           XMS-CASPT2 extension: T. Shiozaki (2011)

 IPEA shift= 0.25

 Convergence thresholds:  THRVAR = 1.00D-08  THRDEN = 1.00D-08

 Number of optimized states: 10  Roots:   1   2   3   4   5   6   7   8   9  10
 Number of reference states: 10  Roots:   1   2   3   4   5   6   7   8   9  10

 Using rs2c

 Reference symmetry:                   2   Doublet
 Number of electrons:                 31
 Maximum number of shells:             6
 Maximum number of spin couplings:   132

 Reference space:     1144 conf     3024 CSFs
 N elec internal:     2304 conf     6048 CSFs
 N-1 el internal:     2907 conf    12852 CSFs
 N-2 el internal:     3134 conf    23390 CSFs

 Number of electrons in valence space:                     23
 Maximum number of open shell orbitals in reference space:  7
 Maximum number of open shell orbitals in internal spaces: 11


 Number of core orbitals:           4 (   4   0 )
 Number of closed-shell orbitals:   6 (   6   0 )
 Number of active  orbitals:        9 (   5   4 )
 Number of external orbitals:     145 (  97  48 )

 Molecular orbitals read from record     2143.2  Type=MCSCF/CANONICAL (state averaged)

 Integral transformation finished. Total CPU:   0.56 sec, npass=  1  Memory used:   1.87 MW

 Number of p-space configurations:  25

 Reference wavefunction optimized for reference space (refopt=1)

 State     Reference Energy
     1      -278.95515650
     2      -278.92760734
     3      -278.92760719
     4      -278.90291825
     5      -278.79015510
     6      -278.77178455
     7      -278.76633302
     8      -278.76633272
     9      -278.76264216
    10      -278.76088108

 Number of blocks in overlap matrix:     4   Smallest eigenvalue:  0.12D-06
 Number of redundant functions:         19
 Number of N-2 electron functions:    2250
 Number of N-1 electron functions:   12852

 Number of internal configurations:                 3024
 Number of singly external configurations:        931280
 Number of doubly external configurations:      12131560
 Total number of contracted configurations:     13065864
 Total number of uncontracted configurations:  123651184

 Weight factors for SA-density in H0:    0.100000  0.100000  0.100000  0.100000  0.100000  0.100000  0.100000  0.100000  0.100000
                                         0.100000

 FIMAX=  0.29D-02  FXMAX=  0.11D-01  DIAG=  F  F  NOREF=1  NOINT=0  IHPPD=2

 Nuclear energy:                      77.47456689
 Core energy:                       -256.98759545
 Zeroth-order valence energy:        -14.50832181    -14.49255831    -14.49255801    -14.47631696    -14.49561882
                                     -14.20851917    -14.20851984    -14.20851912    -14.47984320    -14.20851956
 Zeroth-order total energy:         -194.02135037   -194.00558688   -194.00558658   -193.98934553   -194.00864739
                                    -193.72154774   -193.72154840   -193.72154768   -193.99287177   -193.72154812
 First-order energy:                 -84.93380613    -84.92202047    -84.92202061    -84.91357272    -84.78150771
                                     -85.05023681    -85.04478462    -85.04478503    -84.76977040    -85.03933296

 Performing an extended multi-state calculation

 Feff
              1             2             3             4             5             6             7             8
    1   -14.5083218    -0.0000002    -0.0000001    -0.0000002     0.0000039     0.0000001    -0.0000093    -0.0000029
    2    -0.0000002   -14.4925583    -0.0000000     0.0000052    -0.0000007    -0.0000106     0.0000001     0.0000000
    3    -0.0000001    -0.0000000   -14.4925580     0.0000025    -0.0000002    -0.0000046     0.0000000     0.0000000
    4    -0.0000002     0.0000052     0.0000025   -14.4763170    -0.0000006     0.0000000     0.0000001     0.0000000
    5     0.0000039    -0.0000007    -0.0000002    -0.0000006   -14.4956188    -0.0000013    -0.0000350    -0.0000074
    6     0.0000001    -0.0000106    -0.0000046     0.0000000    -0.0000013   -14.2085192    -0.0000003    -0.0000001
    7    -0.0000093     0.0000001     0.0000000     0.0000001    -0.0000350    -0.0000003   -14.2085198    -0.0000001
    8    -0.0000029     0.0000000     0.0000000     0.0000000    -0.0000074    -0.0000001    -0.0000001   -14.2085191
    9    -0.0000005     0.0000017     0.0000043    -0.0000007    -0.0000010     0.0000616     0.0000147     0.0000016
   10    -0.0000001    -0.0000001    -0.0000000    -0.0000001     0.0000016     0.0000002     0.0000026     0.0000004
              9            10
    1    -0.0000005    -0.0000001
    2     0.0000017    -0.0000001
    3     0.0000043    -0.0000000
    4    -0.0000007    -0.0000001
    5    -0.0000010     0.0000016
    6     0.0000616     0.0000002
    7     0.0000147     0.0000026
    8     0.0000016     0.0000004
    9   -14.4798432     0.0000182
   10     0.0000182   -14.2085196

 Rotated reference energies:

 E0(1) =   -194.02135037     Eref(1) =   -278.95515648     Eigenvector:  1.0000  0.0000  0.0000  0.0000 -0.0003 -0.0000  0.0000  0.0000  0.0000  0.0000
 E0(2) =   -194.00864739     Eref(2) =   -278.79015512     Eigenvector: -0.0003 -0.0002 -0.0001 -0.0000 -1.0000 -0.0000 -0.0001 -0.0000 -0.0001  0.0000
 E0(3) =   -194.00558688     Eref(3) =   -278.92760733     Eigenvector: -0.0000  0.9985  0.0551 -0.0003 -0.0002  0.0000 -0.0000 -0.0000 -0.0002  0.0000
 E0(4) =   -194.00558658     Eref(4) =   -278.92760717     Eigenvector:  0.0000  0.0551 -0.9985  0.0001  0.0001 -0.0000  0.0000  0.0000  0.0003 -0.0000
 E0(5) =   -193.99287178     Eref(5) =   -278.76264219     Eigenvector: -0.0000  0.0001  0.0003  0.0002 -0.0001 -0.0002 -0.0001 -0.0000  1.0000 -0.0001
 E0(6) =   -193.98934552     Eref(6) =   -278.90291825     Eigenvector: -0.0000  0.0003  0.0002  1.0000 -0.0000 -0.0000 -0.0000 -0.0000 -0.0002  0.0000
 E0(7) =   -193.72155097     Eref(7) =   -278.76383674     Eigenvector: -0.0000 -0.0000 -0.0000  0.0000 -0.0001  0.1013  0.7163  0.0924  0.0000 -0.6842
 E0(8) =   -193.72154771     Eref(8) =   -278.76974571     Eigenvector:  0.0000  0.0000  0.0000 -0.0000  0.0000 -0.7934  0.1324 -0.5912 -0.0002 -0.0588
 E0(9) =   -193.72154766     Eref(9) =   -278.76827431     Eigenvector: -0.0000  0.0000  0.0000 -0.0000 -0.0000 -0.5999 -0.0761  0.7941 -0.0001 -0.0612
 E0(10) =  -193.72154558     Eref(10) =  -278.76347460     Eigenvector: -0.0000  0.0000  0.0000 -0.0000 -0.0001 -0.0194  0.6809  0.1065  0.0001  0.7243

 ?WARNING: SMALL EIGENVALUE OF PAIR OVERLAP MATRIX:  0.1317D-06

 Number of blocks in overlap matrix:     4   Smallest eigenvalue:  0.12D-06
 Number of N-2 electron functions:    2231
 Number of N-1 electron functions:   12852

 Number of internal configurations:                 3024
 Number of singly external configurations:        931280
 Number of doubly external configurations:      12018909
 Total number of contracted configurations:     12953213
 Total number of uncontracted configurations:  123651184


 Using contracted singles

 Number of contracted N-1 electron functions:      10527
 Inconsistency for space 1e !!
 ISymIJ,ISymA,le1,le2 1 1      1671   36

 GLOBAL ERROR fehler on processor   0

[qia...@theochem.uni-stuttgart.de]

Would you please provide the input file so that we can investigate the problem?

[Clément S]

Sure:

 ***,ON O2 scan angle
 memory,900,m
 gprint,orben,orbital,civector
 gthresh,energy=1.d-8,step=1.d-4

 orient,noorient
 angstrom

 rno=                 1.0634    ANGSTROM
 ro2=                 1.2075    ANGSTROM
 ango2=         5 DEGREE
 angno=               0.00000000 DEGREE
 dihedral=            0.00000000 DEGREE

 r= 15 ANGSTROM

 orbref=2140.2
 nocc=19
 nclosed=10

 basis,t-aug-cc-pvdz
 symmetry,x

 if(ango2.eq.0) then
 ango2=0.01
 elseif(ango2.eq.180) then
 ango2=179.99
 endif

 if(angno.eq.0) then
 angno=0.01
 elseif(angno.eq.180) then
 angno=179.99
 endif

 geometry={
 N1,
 Q1,N1,(15.999/(14.007+15.999))*rno,
 Q2,Q1,r,N1,angno,
 O3,Q1,(14.007/(14.007+15.999))*rno,Q2,180-angno,N1,180,
 O1,Q2,0.5*ro2,Q1,ango2,N1,dihedral,
 O2,Q2,0.5*ro2,Q1,180-ango2,O1,180
 }
 {multi,
 orbital,orbref+1
 orbprint,1
 maxit,40
 wf,31,1,1
 occ,nocc-4,4
 closed,nclosed,0
 state,10
 weight,4[1],6[3]
 wf,31,2,1
 occ,nocc-4,4
 closed,nclosed,0
 state,10
 weight,4[1],6[3]
 noextra
 }

 R_list=[15,6,4,3.5,3.4,3.3,3.2,3.1,3,2.9,2.8,2.7,2.65,2.6,2.55,2.5,2.45,2.4,2.3,2.2,2.1,2,1.9,1.8]

 do i=1,#R_list
 r=R_list(i)        ANGSTROM

 {multi,
 start,orbref+i
 canorb,orbref+i+1
 orbprint,4
 maxit,40
 {iterations
 dont,internal,1,to,2}
 wf,31,2,1
 occ,nocc-4,4
 closed,nclosed,0
 state,10
 weight,4[1],6[3]
 wf,31,1,1
 occ,nocc-4,4
 closed,nclosed,0
 state,10
 weight,4[1],6[3]
 noextra
 }

 {rs2c,mix=10,xms=1,ipea=0.25
 orbital,orbref+i+1
 wf,31,2,1
 state,10}

 D2(i)=msenergy(1)
 D4(i)=msenergy(2)
 D6(i)=msenergy(3)
 D8(i)=msenergy(4)
 D9(i)=msenergy(5)
 D11(i)=msenergy(6)
 D13(i)=msenergy(7)
 D15(i)=msenergy(8)
 D17(i)=msenergy(9)
 D19(i)=msenergy(10)

 enddo

 table,R_list,D2,D4,D6,D8,D9,D11,D13,D15,D17,D19

---

Thank you for looking into it!

Kind regards,

Clément

[qia...@theochem.uni-stuttgart.de]

From my initial investigation this is most likely to be a bug in Molpro related to the linear-dependence in the singles excitation. We are working on it and will post here again when the bug is fixed.

[Clément S]

Thanks a lot for your reactivity!

Cheers,

Clément

[qia...@theochem.uni-stuttgart.de]

We have fixed the problem in the source code. The patch will be alive in the next binary release.

Note that you may want to use the "noprop" option in MS-MR-RS2C calculations with many states. The property code was not designed to run efficiently in this case and it will slow things down noticeably.

[Clément S]

Many thanks, that's amazing.

Yes, I tried the noprop keyword in the lastest version, and it increadibly speed things up.

Will the program work with more additional diffuse functions, necessary to compute anions?

Such as q-aug-cc-pVTZ?

Thanks,

Clément

[qia...@theochem.uni-stuttgart.de]

In general Molpro does not restrict what basis set one can use, although you may need to define the basis yourself if it is not in Molpro basis library.

[Clément S]

Thanks for the reply.

My concern was more about the linear dependency problem issued here.
Was this problem related in some way to the diffuseness of some basis functions?

Cheers,

Clément

[qia...@theochem.uni-stuttgart.de]

No, this problem is due to an improper algorithm in removing the linear dependence. It has no direct connection to the basis set you use. After the patch it should be fine.

[Clément S]

Super, many thanks!