Posted by Marco Marazzi on February 01, 2010 at 19:13:53:

Dear Molcas users,

I have problems in the calculation of the Lagrangian multipliers (&MCLR) that I need for a CASSCF minimum energy path with analytical gradients.

The message shown is:

"No convergence for perturbation no: 1. Increase Iter."

Even increasing the number of iterations to 1000, the system does not converge.

My input is the following:

>>> Do while

&SEWARD

>>> If ( Iter = 1 ) <<<

&SCF

Charge

1

UHF

>>> EndIf <<<

&RASSCF

SPIN

1

CHARGE

1

NACTEL

6 0 0

FROZEN

0

INACTIVE

23

RAS2

6

CIROOT

2 2

1 2

1 1

RLXRoot

2

&MCLR

SALA

2

Iterations

1000

&ALASKA

&SLAPAF

Constraints

a = Sphere

Value

a = 0.1

End of Constraints

MEP-search

NMEP

200

>>> EndDo

With numerical calculation of the gradients (no &MCLR required) no problems are found, but it needs much more time.

Are there other ways to calculate CASSCF analytical gradients?

Thanks in advance