Spin Constraint in CDFT

Hi,
I am doing a spin constraint DFT for the neutral TCNQ system and I want to apply spin constraint (-1/2 and 1/2) to make a singlet diradical of TCNQ. (input is shown below)

The constraint converged well. But I am confused with the net spin of CDFT Becke Populations. I want to get net spin zero having first constraint 0.5 for spin-up and -0.5 for spin down But the calculation gives 1 and -1 for up and down spin with the total spin zero.
Can I use this spin constraint to explain the diradical nature of the TCNQ?
If anyone explains this it will be helpful.

<S^2> = 0.8051
CDFT Becke Populations

        Atom      Excess Electrons     Population (a.u.)    Net Spin
   1       C       -0.078697        5.921303         -0.114195
   2       C       -0.506393        5.493607          0.242580
   3       C       -0.506393        5.493607          0.242580
   4       C       -0.057956        5.942044          0.356097
   5       C       -0.126469        5.873531         -0.040869
   6       N        0.204501        7.204501          0.150111
   7       C       -0.126469        5.873531         -0.040869
   8       N        0.204501        7.204501          0.150111
   9       H        0.496116        1.496116          0.027227
  10       H        0.496116        1.496116          0.027227
  11       C       -0.506393        5.493607         -0.242580
  12       C       -0.078697        5.921303          0.114195
  13       C       -0.506393        5.493607         -0.242580
  14       C       -0.057956        5.942044         -0.356097
  15       C       -0.126469        5.873531          0.040869
  16       N        0.204501        7.204501         -0.150111
  17       C       -0.126469        5.873531          0.040869
  18       N        0.204501        7.204501         -0.150111
  19       H        0.496116        1.496116         -0.027227
  20       H        0.496116        1.496116         -0.027227

INNPUT
$molecule
0 1
C 0.0000000000000000 0.0000000000000000 1.4152006028486606
C 0.0000000000000000 1.2420365992696554 0.6686836631384167
C 0.0000000000000000 -1.2420365992438873 0.6686836631297544
C 0.0000000000000000 0.0000000000000000 2.7793485122575290
C 0.0000000000000000 -1.2074202754390841 3.5387234605255946
N 0.0000000000000000 -2.1770323450305376 4.1546371203769299
C 0.0000000000000000 1.2074202754340271 3.5387234605373572
N 0.0000000000000000 2.1770323450173050 4.1546371204008974
H 0.0000000000000000 -2.1780997469321841 1.2088931031797505
H 0.0000000000000000 2.1780997469621197 1.2088931031781973
C 0.0000000000000000 1.2420365992505300 -0.6686836631278923
C 0.0000000000000000 0.0000000000000000 -1.4152006028459807
C 0.0000000000000000 -1.2420365992623186 -0.6686836631353219
C 0.0000000000000000 0.0000000000000000 -2.7793485122552455
C 0.0000000000000000 -1.2074202754366481 -3.5387234605358873
N 0.0000000000000000 -2.1770323450207361 -4.1546371203986903
C 0.0000000000000000 1.2074202754315717 -3.5387234605311662
N 0.0000000000000000 2.1770323450179561 -4.1546371203897978
H 0.0000000000000000 -2.1780997469557679 -1.2088931031761225
H 0.0000000000000000 2.1780997469376153 -1.2088931031769796
$end

$rem
JOBTYPE OPT
exchange OmegaB97X-D
correlation none
basis cc-pVTZ
MAX_SCF_CYCLES 500
SCF_ALGORITHM DIIS
SCF_CONVERGENCE 8
!SYMMETRY FALSE
BECKE_SHIFT UNSHIFTED
SYM_IGNORE TRUE
CDFT TRUE
MEM_TOTAL 8000
MEM_STATIC 2000
CDFT_BECKE_POP 1
solvent_method pcm
$end

$pcm
Theory cpcm
Method SWIG
Solver Inversion
HeavyPoints 194
HPoints 194
Radii Bondi
vdwScale 1.2
$end

$cdft
1
0.5 1 10 s
-0.5 11 20 s
$end

$solvent
Dielectric 20.7
$end

Thanks,
Roshan

The $cdft input section seems to be incorrect. The manual states that for a spin constraint, the spin-up and spin-down densities contribute with opposite signs (CA α − CA β = CA), resulting in a measure of the net spin on atom A. The spin constraint should be set to 1, not 0.5. Like this

$cdft
2
1 1 10 s
-1 11 20 s
$end

You should take a closer look at the manual and its examples, and try experimenting more.

2 Likes

If there is an error in the manual, could you please be more explicit?

@Chandler Thank you for your reply.
From the input you suggest,
$cdft
2
1 1 10 s
-1 11 20 s
$end
Which means total spin 2 is constraint to the system, right?
I get total CDFT net spin =0 (1 net spin (up) and -1 net spin (down). That satisfies the constraint,
As I understand, Net spin (from Q-chem oup put) gives the spin contribution to the particular atom from the applied constraint.

But, my confusion is what is the spin value for spin-up(1/2 or 1)? Can you please explain to me.

I want to apply total 1 spin constraint to my system considering 1/2 up and 1/2 down.

Q-chem manual says we can use any spin constraint.