[molpro-user] State Averaged CASSCF calculations

[Tom Oliver]

Benj,

I had found this work-around; but I want to have more control over my basis 
sets. The two gradient programs in Molpro: Alaska and Cadpac have their 
individual downfalls. Alaska can use contracted basis sets, but cannot 
state average, Cadpac cannot use contracted basis sets, but can state 
average.

Tom

--On 19 November 2008 07:56 -0600 Benj FitzPatrick <benjfitz at uchicago.edu> 
wrote:

> Tom,
> I found this website useful for just that purpose.
>
> http://setani.sci.hokudai.ac.jp/sapporo/Order.do;jsessionid=6726EB76B5048
> DA7442C277E7B3E8D41
>
> Cheers,
> Benj
>
> ---- Original message ----
>> Date: Tue, 18 Nov 2008 14:50:57 +0000
>> From: Tom Oliver <t.a.a.oliver at bristol.ac.uk>
>> Subject: [molpro-user] State Averaged CASSCF calculations
>> To: molpro-user at molpro.net
>>
>> Dear All,
>>
>> I am currently trying to optimize state-averaged CASSCF excited states
>> for  small heteroaromatic systems using CPMCSCF which requires segmented
>> basis  sets. I would like to use contracted basis sets (as given in the
>> molpro  basis library), but cannot do this. Is there an alternative way
>> of  calculating optimised state-averaged CASSCF excited states? Or at
>> least  using segmented basis sets in a more meaningful way?
>>
>> I would be most grateful if anyone could help.
>>
>> Many thanks
>>
>> Tom
>>
>> My input looks like:
>>
>> memory,30,m
>>
>> geometry={
>> ANGSTROM
>> C
>> C                  1              B1
>> N                  1              B2    2              A1
>> H                  3              B3    1              A2    2
>> D1
>> H                  1              B4    3              A3    2
>> D2
>> H                  2              B5    1              A4    3
>> D3
>> C                  2              B6    1              A5    3
>> D4
>> H                  7              B7    2              A6    1
>> D5
>> N                  7              B8    2              A7    1
>> D6}
>>
>> B1=1.37914855
>> B2=1.35799100
>> B4=1.07917398
>> B5=1.07872848
>> B6=1.41281136
>> B7=1.08068329
>> B8=1.32916568
>> A1=106.10333601
>> A2=127.85599586
>> A3=121.83299143
>> A4=127.26010479
>> A5=104.50045469
>> A6=128.63998526
>> A7=112.00582846
>> D1=180.00000000
>> D2=180.00000000
>> D3=180.00000000
>> D4=0.00000000
>> D5=180.00000000
>> D6=0.00000000
>>
>>
>> B3a=[1.0]
>> B3=B3a(1)
>>
>> basis={
>> spd,c,aug-cc-pvtz;c
>> s,h,aug-cc-pvtz;c
>> spd,n,aug-cc-pvtz;c
>> }
>>
>>
>> {hf
>> occ,15,3
>> orbprint
>> wf,36,1,0
>> }
>>
>> do i=1,#B3a
>> B3=B3a(i)
>>
>> {casscf
>> occ,17,5,
>>
>> closed,13,0,
>>
>> natorb,,ci
>> wf,36,2,0;state,2
>> CPMCSCF,GRAD,1.2,2.2}
>>
>> {optg
>> inactive,B3}
>>
>> {casscf
>> occ,17,5,
>>
>> closed,13,0,
>>
>> natorb,,ci
>> wf,36,2,0;state,2
>> CPMCSCF,GRAD,1.2,2.2}
>>
>> {rs2c,shift=0.3
>> occ,17,5
>> closed,13,0
>> core,5,0
>> state,1,2,
>> wf,36,2,0}
>>
>> ept2_root2(i)=emp2
>>
>> enddo
>>
>> And the output ends:
>>
>> PROGRAM * OPT (Geometry optimization)     Authors: F. Eckert and H.-J.
>> Werner
>>
>>
>> *** Long output written to logfile /Users/stephen/Downloads/steve.log ***
>>
>>
>> Geometry optimization using default procedure for command CASSCF
>>
>> Inactive variables:    B3
>>
>> Active variables:      B1 B2 A1 A2 D1 B4
>>                        A3 D2 B5 A4 D3 B6
>>                        A5 D4 B7 A6 D5 B8
>>                        A7 D6
>>
>> Geometry written to block  1 of record 700
>>
>> GENERAL BASIS CONTRACTION NOT IMPLEMENTED IN CADPAC GRADIENTS
>>
>> Please try with segmented basis set
>>
>>
>> ----------------------
>> Tom Oliver
>> Laser Group
>> School of Chemistry
>> University of Bristol
>> BS8 1TS
>>
>> Email: t.a.a.oliver at bristol.ac.uk
>> Tel: 0117 928 9938
>> _______________________________________________
>> Molpro-user mailing list
>> Molpro-user at molpro.net
>> http://www.molpro.net/mailman/listinfo/molpro-user



----------------------
Tom Oliver
Laser Group
School of Chemistry
University of Bristol
BS8 1TS

Email: t.a.a.oliver at bristol.ac.uk
Tel: 0117 928 9938