From owner-nwchem-users@emsl.pnl.gov Thu Dec 2 11:19:33 2004 Received: from odyssey.emsl.pnl.gov (localhost [127.0.0.1]) by odyssey.emsl.pnl.gov (8.12.10/8.12.10) with ESMTP id iB2JJX0h004601 for ; Thu, 2 Dec 2004 11:19:33 -0800 (PST) Received: (from majordom@localhost) by odyssey.emsl.pnl.gov (8.12.10/8.12.10/Submit) id iB2JJXgU004600 for nwchem-users-outgoing; Thu, 2 Dec 2004 11:19:33 -0800 (PST) Date: Thu, 02 Dec 2004 11:19:31 -0800 From: "Bylaska, Eric J" Subject: RE: default CUTOFF values for library pseudopotentials To: Lev Gelb Cc: nwchem-users@emsl.pnl.gov Message-id: MIME-version: 1.0 X-MIMEOLE: Produced By Microsoft Exchange V6.5.7226.0 Content-type: text/plain; charset=US-ASCII Thread-Topic: default CUTOFF values for library pseudopotentials Thread-Index: AcTYm8ND+hjyIMGIRt+80CTUQ8KHSgAAz3Zg Content-class: urn:content-classes:message X-MS-Has-Attach: X-MS-TNEF-Correlator: X-OriginalArrivalTime: 02 Dec 2004 19:19:32.0060 (UTC) FILETIME=[D9EBE1C0:01C4D8A3] Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by odyssey.emsl.pnl.gov id iB2JJW0h004597 Sender: owner-nwchem-users@emsl.pnl.gov Precedence: bulk Lev, The cutoff radii of the psps are always specified in the pspw input data block, i.e. elements involved in the cluster: 1: Si core charge: 4.0 lmax= 2 comment : Hamann pseudopotential pseudpotential type : 0 highest angular component : 2 local potential used : 2 number of non-local projections: 4 cutoff = 1.059 1.286 1.286 2: C core charge: 4.0 lmax= 2 comment : Parameterized (J.Phys.Chem., vol 100, page 6966) Hamman psp pseudpotential type : 0 highest angular component : 2 local potential used : 2 number of non-local projections: 4 cutoff = 0.800 0.850 0.850 For Hamann psp's the default radii are defined as follows: For the potentials of the bound states 1s, 2p, 3d, ..., rc=0.4*Outer_most_peak_in_the_all_electron_wavefunction For the potentials of other bound states, i.e. 2s,3s,3p..., rc=0.6*Outer_most_peak_in the_all_electron_wavefunction For the potentials of scattering state, rc = Max(other rc's) These rules are are based on the fact that for ae-wavefunctions with no radial nodes the pseudowavefunction will be converged to the ae-wavefunction at a distance of ~2.5*rc, (note that 0.4=1/2.5), for ae-wavefunction with radial nodes convergence is expected to be a slightly quicker. These arguments don't always work for defining cutoff, and in general these default Hamann rules will produce cutoffs that result in very stiff pseudopotentials. The output for psp generation is printed to stdout by running "task pspw energy", with "print debug"...the ae results for Si are as follows. ------- Grid information ------- Zcharge : 1.400000e+01 amesh : 1.005000e+00 R_max : 4.470524e+01 Ngrid : 2955 ------------------------------------------------------------ n l population Eigenvalue Outer Peak 1 s 2.00 -6.535740e+01 7.252218e-02 2 s 2.00 -5.098976e+00 4.545499e-01 2 p 6.00 -3.513807e+00 3.855685e-01 3 s 2.00 -3.998215e-01 1.765023e+00 3 p 2.00 -1.529635e-01 2.144010e+00 ------------------------------------------------------------ electronic charge = -1.400000e+01 electronic core charge = -1.000000e+01 total atom charge = 1.332268e-13 Using this information, you can compute the cutoff radii for Si ouputed in the pspw input block. rcut(3s) = rcut(0) = 0.6 * 1.765023e+00 = 1.05901 rcut(3p) = rcut(1) = 0.6 * 2.144010e+00 = 1.28641 rcut(3d) = rcut(2) = Max(rcut(i)) = rcut(1) = 1.28641 For TM potentials the default cutoff radii make a little more sense since the matching is done via the wavefunctions, rather than the potentials (i.e. Hamann). In this case, the default cutoff radii are defined to be the Outer most peak in the all electron wavefunction. Hope this helps Eric -----Original Message----- From: owner-nwchem-developers@emsl.pnl.gov [mailto:owner-nwchem-developers@emsl.pnl.gov] On Behalf Of Lev Gelb Sent: Thursday, December 02, 2004 10:08 AM To: nwchem-users@emsl.pnl.gov Subject: default CUTOFF values for library pseudopotentials I've notice that the default hamann-type pseudopotentials in "library1" don't have "cutoff" (rcut) blocks specified; these are: Al, As, Be, Li, S, and Si What are the default values used when generating these pseudopotentials, then? Thanks, Lev Gelb ------------------------------------------------------------------------ -- Lev Gelb http://www.chemistry.wustl.edu/~gelb gelb@wustl.edu Department of Chemistry, Washington University, Campus Box 1134 St. Louis, MO 63130 USA ph:(314)935-5026 fax:(314)935-4481 ------------------------------------------------------------------------ -- "71.3841674321 percent of statistics are made up on the spot." --- J. A. Paulos