From owner-chemistry@ccl.net Thu Feb 2 15:15:00 2006 From: "makowskm|*|chemia.uj.edu.pl" To: CCL Subject: CCL: excited states of polyene cations Message-Id: <-30757-060202110000-26437-Xz6VXRcPom54T6vxWSY95g,server.ccl.net> X-Original-From: makowskm++chemia.uj.edu.pl Content-Transfer-Encoding: 8bit Content-Type: text/plain;charset=iso-8859-2 Date: Thu, 2 Feb 2006 02:46:47 +0100 (CET) MIME-Version: 1.0 Sent to CCL by: makowskm!^!chemia.uj.edu.pl Dear Ulrike, TDDFT usually gives wrong description of relative enegies of A_g and B_u states of neutral polyenes. There was a paper by Head-Gordon showing that Tamm-Dancoff approximation to TDDFT works better in this case than popular RPA, but still this does not seem to be a method of choice for these systems, probably. CASSCF worked quite fine as long as large active space can be used. I remember one paper pointing that you need sth like double-zeta active space (for octatetraene it would be 4 occupied, 8 virtual) to obtain very good accuracy. I can search for reference if you are interested. It was so as I've checked up to C10. I was testing some time ago CIS(D) and SAC-CI but they still had big problems with A_g state (the one dominated by double excitation). The similar problems might happen for cations - this is out of my knowledge (I did not make any calculations on them). Yours, Marcin Makowski -- Dr Marcin Makowski Jagiellonian University Department of Theoretical Chemistry email:makowskm^-^chemia.uj.edu.pl or Kyushu University email:marcin^-^cube.kyushu-u.ac.jp