CCL:G: geometry optimization of O2 (singlet delta)



 Sent to CCL by: Sathya Perumal [sathya**theochem.kth.se]
 Hi,
      The first excited singlet ^1 Delta_g (placed 0.97 eV  above the
 ^3 Sigma_g^-1 ) is described by linear combination of two degenerate
 pi^* orbitals and therefore two determinants for normal iteration
 procedure. However it's possible to describe in a single determinant
 (as in RHF) by employing complex orbitals. and hence it's necessary to
 specify 'complex' keyword for obtaining 1^Delta_g in linear
 representation. I am not sure whether the analytical derivatives works
 for complex orbitals and so for frequencies for minimum location. But
 it's possible to check the instability of the obtained Complex RHF
 wavefunctions using the keyword stable=(crhf,1opt). Ofcourse immediate
 other possible alternatives would be casscf, or perhaps with guess=mix.
 regards
 Sathya
 ---
 KTH, Stockholm
 On 15 December 2010 21:43, Roy Jensen JensenRH/./MacEwan.ca
 <owner-chemistry : ccl.net> wrote:
 >
 > Sent to CCL by: Roy Jensen [JensenRH{=}MacEwan.ca]
 > I found that G03 Test Job 001 calculates the singlet delta energy of
 > oxygen at fixed geometry.
 > ********************
 > #P HF/STO-3G COMPLEX
 >
 > Gaussian Test Job 01
 > GROUND STATE STO-3G//STO-3G DIOXYGEN
 >
 >  0 1
 >  O
 >  O 1 R
 >
 > R 1.22
 > ********************
 >
 > Whereas removing the 'COMPLEX' keyword calculates the singlet sigma
 > state. However, every time I attempt to do a geometry optimization,
 > with both OPT and COMPLEX, it crashes.
 > ********************
 >  (Enter C:\G03W\l401.exe)
 >  Initial guess read from the read-write file:
 >  Guess basis will be translated and rotated to current coordinates.
 >  Unable to project occupied orbitals !
 >  Error termination via Lnk1e in C:\G03W\l401.exe at Mon Dec 13
 > 11:23:37 2010.
 > ********************
 >
 > According to the Gaussian manual, the COMPLEX keyword is very
 > limiting. Is there another way to optimized to the singlet delta state
 > of oxygen and oxygen analogues (NF specifically)?
 >
 > Thanks,
 >  Dr. Roy Jensen
 > (==========)-----------------------------------------¤
 >  Chemistry, Grant MacEwan University
 >  Room 5-172J, 10700-104 Avenue
 >  Edmonton, AB    T5J 4S2
 >  780.633.3915>      http://www.ccl.net/cgi-bin/ccl/send_ccl_message>;  
    http://www.ccl.net/cgi-bin/ccl/send_ccl_message>;  
    http://www.ccl.net/chemistry/sub_unsub.shtml>;  
    http://www.ccl.net/spammers.txt>;
 >
 >