From owner-chemistry@ccl.net Sat Dec 6 06:53:01 2014 From: "Lars Goerigk lars.goerigk_+_unimelb.edu.au" To: CCL Subject: CCL: Calculations for excited states Message-Id: <-50796-141206064943-7456-8jiVtsqljuZtRc5Gs2Jbrw%%server.ccl.net> X-Original-From: Lars Goerigk Content-Language: en-US Content-Type: multipart/alternative; boundary="_000_114F18CB0E024EDD92879BE87512CCE0unimelbeduau_" Date: Sat, 6 Dec 2014 11:49:27 +0000 MIME-Version: 1.0 Sent to CCL by: Lars Goerigk [lars.goerigk|unimelb.edu.au] --_000_114F18CB0E024EDD92879BE87512CCE0unimelbeduau_ Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable Dear James, if I understand you correctly, you define "TDDFT/CIS(D)/CC2/SOPPA=94 to be = one one =93level" and you ask if the methods belonging to this =93level" pr= ovide results with different accuracies. The short answer is yes. TDDFT results depend on the type of functional and generally for valence ex= citations double-hybrid functionals surpass all other functionals. They sometimes also outperform CC2, which definitely outperforms CIS(D). In= fact time-dependent double-hybrids are a mixture between TD-DFT and CIS(D)= . The average accuracy for those methods is around 0.14 eV, which means the= y are close to the chemical-accuracy limit of 0.1 eV. Unfortunately, only one program allows such calculations at the moment, whi= ch is Frank Neese=92s ORCA. The good news however is that this program is f= reely available for academic researchers. More information on TD double-hybrids and a comparison with other density f= unctionals, with CC2 and with CIS(D) is given in: J. Chem. Phys. 132, 184103 (2010); http://dx.doi.org/10.1063/1.3418614 and in Phys. Chem. Chem. Phys. 11, 4611-4620 (2009); http://dx.doi.org/10.1039/B90= 2315A I hope this answered your question somehow. With best wishes, Lars --- Dr. Lars Goerigk ARC DECRA Fellow School of Chemistry The University of Melbourne VIC 3010 Australia Research profile: http://www.chemistry.unimelb.edu.au/dr-lars-goerigk List of my publications: http://www.researcherid.com/rid/D-3717-2009 On 5 Dec 2014, at 12:39 pm, James Mao xjamesmao+*+gmail.com > wrote: Sent to CCL by: "James Mao" [xjamesmao_+_gmail.com] Dear colleagues, I am recently trying to carry out calculations to simulate UV spectra of relative small molecules (less than 30 atoms). To plot the spectra the excitation energy as well as oscillator strength are needed from calculatio= n. I noticed that actually there are several candidates which are capable to do this. After some practice and comparison it seems to me that it can be done= in two different levels: TDDFT/CIS(D)/CC2/SOPPA produce quite reasonable and similar results, and EOM-CCSD/SAC-CI/CCSD produce similar results at next/higher level. My question: in principle, for methods at the same level= , is any of them expected to produce more accurate data than others? I am not talking about the difference of computation cost here since my molecules ar= e all small. Any comments/suggestion are welcome and will be greatly appreciated. Best Regards, James -=3D This is automatically added to each message by the mailing script =3D-E-mail to subscribers: CHEMISTRY*_*ccl.net or use:E-mail to administrators: CHEMISTRY-REQUEST*_*ccl.net or usehttp://www.ccl.net/chemistry/sub_unsub.shtmlhttp://www.ccl.net/spammers.txt--_000_114F18CB0E024EDD92879BE87512CCE0unimelbeduau_ Content-Type: text/html; charset="Windows-1252" Content-ID: <0B9DF90BF70F0D44A4812874F718C4D3*_*unimelb.edu.au> Content-Transfer-Encoding: quoted-printable Dear James,
if I understand you correctly, you define "TDDFT/CIS(D)/CC2/SOPPA= =94 to be one one =93level" and you ask if the methods belonging to th= is =93level" provide results with different accuracies.
The short answer is yes.
TDDFT results depend on the type of functional and generally for valen= ce excitations double-hybrid functionals surpass all other functionals.
They sometimes also outperform CC2, which definitely outperforms CIS(D= ). In fact time-dependent double-hybrids are a mixture between TD-DFT and C= IS(D). The average accuracy for those methods is around 0.14 eV, which mean= s they are close to the 
chemical-accuracy limit of 0.1 eV.

Unfortunately, only one program allows such calculations at the moment= , which is Frank Neese=92s ORCA. The good news however is that this program= is freely available for academic researchers.

More information on TD double-hybrids and a comparison with other dens= ity functionals, with CC2 and with CIS(D) is given in:

J. Chem. Phys. 132, 184103 (2010); http://dx.doi.org/10.1063/= 1.3418614

and in

Phys. Chem. Chem. Phys. 1= 1, 4611-4620 (2009); http://dx.doi.org/= 10.1039/B902315A       

I hope this answered your= question somehow.

With best wishes,

Lars      =
---
Dr. Lars Goerigk
ARC DECRA Fellow
School of Chemistry
The University of Melbourne
VIC 3010
Australia

Research profile: 
List of my publications:

On 5 Dec 2014, at 12:39 pm, James Mao xjamesmao+*+gmail.com <owner-chemistry*_*ccl.net> wrote:


Sent to CCL by: "James  Mao" [xjamesmao_+_gmail.com]
Dear colleagues,

I am recently trying to carry out calculations to simulate UV spectra of relative small molecules (less than 30 atoms). To plot the spectra the
excitation energy as well as oscillator strength are needed from calculatio= n. I
noticed that actually there are several candidates which are capable to do =
this. After some practice and comparison it seems to me that it can be done= in
two different levels: TDDFT/CIS(D)/CC2/SOPPA produce quite reasonable and <= br> similar results, and EOM-CCSD/SAC-CI/CCSD produce similar results at
next/higher level. My question: in principle, for methods at the same level= , is
any of them expected to produce more accurate data than others? I am not talking about the difference of computation cost here since my molecules ar= e
all small.
Any comments/suggestion are welcome and will be greatly appreciated.

Best Regards,
James



-=3D This is automatically added to each message by the mailing script =3D-=
E-mail to subscribers: CHEMISTRY*_*ccl.n= et or use:
     http://www.ccl.net/cgi-bin/ccl/send_ccl_message

E-mail to administrators: CHEM= ISTRY-REQUEST*_*ccl.net or use
     http://www.ccl.net/cgi-bin/ccl/send_ccl_message
     http://www.ccl.net/chemistry/sub_unsub.shtml

Before posting, check wait time at: http://w= ww.ccl.net

Job: http://www.ccl.net/jobs
Conferences: http://server.ccl.net/chemistry/announcements/conferences/

Search Messages: http://www.ccl.net/chemistry/searchccl/index.shtml
     h= ttp://www.ccl.net/spammers.txt

RTFI: http:= //www.ccl.net/chemistry/aboutccl/instructions/



--_000_114F18CB0E024EDD92879BE87512CCE0unimelbeduau_--