From owner-chemistry {*at*} ccl.net Thu Aug 20 11:46:01 2015 From: "Hao-Bo Guo guohaobo!^!gmail.com" To: CCL Subject: CCL:G: UV-Visible spectrum Message-Id: <-51602-150820114416-9237-l3HXqVfz85jSYuyRUniL6g(-)server.ccl.net> X-Original-From: Hao-Bo Guo Content-Type: multipart/alternative; boundary=001a113a765cc1d6eb051dc00546 Date: Thu, 20 Aug 2015 11:44:11 -0400 MIME-Version: 1.0 Sent to CCL by: Hao-Bo Guo [guohaobo:-:gmail.com] --001a113a765cc1d6eb051dc00546 Content-Type: text/plain; charset=UTF-8 Hi Mr. Robert Molt, That's the reason I said an assessment of the functionals, basis sets etc is needed for TDDFT... I did not claim what is true or false, but based on my previous assessment the range-separated functionals that I used did show better results compared to others. In the assessment I also chose EOM-CCSD results as the gas-phase references; however, the spectra may typically be measured in solution(s) and the solvation effect must be taken into account. You can check this publication if interested: http://pubs.acs.org/doi/abs/10.1021/jp3084293 On Thu, Aug 20, 2015 at 12:14 AM, Robert Molt r.molt.chemical.physics]^[ gmail.com wrote: > There are many ways to calculated electronic spectra. > > CIS is a very cheap, reasonable wavefunction methodology. > > You can also use TD-DFT (someone stated that range-separated functionals > are better; this is not true for excited states, as the ground and excited > states will have the same error and it will be subtracted off very > commonly). As DFT is is often heavily parameterized, it will fail in some > cases, but not others. Hence, sometimes people encourage comparing to > experiment (although this somewhat defeats the point). > > The best methods are EOM-CCSD or MR-CI excited state calculations, > depending on if dynamic or static correlation is expected to be the greater > issue. These methods have the advantage is that they do not require any > notion of comparing to experiment; they have no parameters, and are thus > equally likely to do well on any system (aside the issue of static vs. > dynamic correlation) because they are based on systematically moving to the > exact wavefunction answer. > > Given that you are doing benzene, a highly symmetric molecule with few > atoms, you could easily afford to do EOM-CCSD or MR-CI if you wish to. > > Remember to use diffuse functions in your basis set, always, to describe > excited states. > > > On 8/19/15 10:45 PM, Hao-Bo Guo guohaobo{:}gmail.com wrote: > > Try TDDFT. Several functionals may do an excellent job for benzene, > especially the range-separated functionals CAM-B3LYP, wB97XD, LC-wPBE. > Others such as doubly hybrid B2PLYPD, and meta-GGA hybrid M06-2X, may also > do a good job for benzene. The basis set you used may be too small, but > acceptable. Your really need an assessment for the functionals, basis set > etc, using the experimental spectra as the reference. > ----From my experience in TDDFT calculations of the benzoic acid > derivatives (JPCA 2012). Hao-Bo Guo > > On Wed, Aug 19, 2015 at 5:11 PM, Kaci Tiziouzou kaci.tiziouzou/./gmail.com > wrote: > >> Hi all, >> >> Can anyone share with me the procedure that I can use to get a UV-VIS >> spectrum for Benzene. I am currently using the TD-SCF in gaussian 09 with a >> 6-31G basis set (Hartree Fock) and my value is way below the experimental. >> >> I must be doing something wrong!! >> >> Thanks in advance >> >> K.T >> > > > -- > Dr. Robert Molt Jr.r.molt.chemical.physics/a\gmail.com > Nigel Richards Research Group > Department of Chemistry & Chemical Biology > Indiana University-Purdue University Indianapolis > LD 326 > 402 N. Blackford St. > Indianapolis, IN 46202 > > --001a113a765cc1d6eb051dc00546 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
Hi Mr. Robert Molt,
That's the reason I said an ass= essment of the functionals, basis sets etc is needed for TDDFT... I did not= claim what is true or false, but based on my previous assessment the range= -separated functionals that I used did show better results compared to othe= rs. In the assessment I also chose EOM-CCSD results as the gas-phase refere= nces; however, the spectra may typically be measured in solution(s) and the= solvation effect must be taken into account. You can check this publicatio= n if interested:
http://pubs.acs.org/doi/abs/10.1021/jp3084293

On Thu, Aug 20, 2015 at 12:14 = AM, Robert Molt r.molt.chemical.physics]^[gmai= l.com <owner-chemistry .. ccl.net> wrote:
=20 =20 =20
There are many ways to calculated electronic spectra.

CIS is a very cheap, reasonable wavefunction methodology.

You can also use TD-DFT (someone stated that range-separated functionals are better; this is not true for excited states, as the ground and excited states will have the same error and it will be subtracted off very commonly). As DFT is is often heavily parameterized, it will fail in some cases, but not others. Hence, sometimes people encourage comparing to experiment (although this somewhat defeats the point).

The best methods are EOM-CCSD or MR-CI excited state calculations, depending on if dynamic or static correlation is expected to be the greater issue. These methods have the advantage is that they do not require any notion of comparing to experiment; they have no parameters, and are thus equally likely to do well on any system (aside the issue of static vs. dynamic correlation) because they are based on systematically moving to the exact wavefunction answer.

Given that you are doing benzene, a highly symmetric molecule with few atoms, you could easily afford to do EOM-CCSD or MR-CI if you wish to.

Remember to use diffuse functions in your basis set, always, to describe excited states.


On 8/19/15 10:45 PM, Hao-Bo Guo guohaobo{:}gmail.com wrote:
Try TDDFT. Several functionals may do an excellent job for benzene, especially the range-separated functionals CAM-B3LYP, wB97XD, LC-wPBE. Others such as doubly hybrid B2PLYPD, and meta-GGA hybrid M06-2X, may also do a good job for benzene. The basis set you used may be too small, but acceptable. Your really need an assessment for the functionals, basis set etc, using the experimental spectra as the reference.
----From my experience in TDDFT calculations of the benzoic acid derivatives (JPCA 2012). Hao-Bo Guo

On Wed, Aug 19, 2015 at 5:11 PM, Kaci Tiziouzou kaci.tiziouzou/./gmail.com <owner-chemistry()ccl.net> wrote:
Hi all,

Can anyone share with me the procedure that I can use to get a UV-VIS spectrum for Benzene. I am currently using the TD-SCF in gaussian 09 with a 6-31G basis set (Hartree Fock) and my value is way below the experimental.

I must be doing something wrong!!

Thanks in advance

K.T


--=20
Dr. Robert Molt Jr.
=
r.molt.chemical.physics/a\gmail.com
Nigel Richards Research Group
Department of Chemistry & Chemical Biology
Indiana University-Purdue University Indianapolis
LD 326
402 N. Blackford St.
Indianapolis, IN 46202

--001a113a765cc1d6eb051dc00546--