Gaussian 09, as of d.01, has several algorithms for SCF calculations. I paste here the manual descriptions:
- The default SCF procedure uses a combination of EDIIS [Kudin02] and CDIIS, with no damping or Fermi broadening
- QC Calls for the use of a quadratically convergent SCF procedure [Bacskay81]. By default this involves linear searches when far from convergence and Newton-Raphson steps when close (unless the energy goes up). This method is slower than regular SCF with DIIS extrapolation but is more reliable.
- XQC Add an extra SCF=QC step in case the first-order SCF has not converged.
- YQC Provides a new algorithm that is useful for difficult SCF convergence cases involving very large molecules. It does steepest descent and then scaled steepest descent as in QC, but then switches to regular SCF instead of quadratic convergence, using the quadratic algorithm only if the regular SCF fails to converge.
The system I'm currently working on is amine-metal clusters with 1 Cu or 1 Fe atom with ECPs and up to a total 76 atoms. The calculations are unrestricted DFT b3lyp.
I've used normal SCF and had some convergence errore here and there. QC seems to be overkill in cases when normal SCF works, so I tried XQC and recently YQC. Neither are havin convergence problems but lately each calculation is taking about a week, so I wonder if the slowdown is the systems I'm using or the switch to YQC.
Has anyone done or seen a benchmark of both method on the same systems? Or can anyone guess from the descriptions if that is my problem? Thanks!
Maybe SCF=XQC is the best to save time because it use expensive QC option only when your system is not converged by default method.
Hi Federico. If it is still actual, try to use the pseudopotential basis sets for Cu and Fe atoms. It usually accelerates calculation speed at the cost of metal's core electrones contribution correction, preestablished as "superposition-like" single potential value.
Yes I am, Alex. LANL2DZ basis + ECP for optimizations and LANL2TZ basis + ECP for energy. The latest is not in G09 by default and I downloaded it from the EMSL basis set exchange.
I am also using scf=xqc for metal complex dft calculation and having good result.
- Badges
- Science topic
- Similar topics
- Correction