I'm running GaAs calculation using HSE06 and ADMM basis sets, and running into some issues. <div><br></div><div>I'm trying to initialize with a non hybrid calculation and use SCREEN_ON_INITIAL_P. This is because at 128 atoms (minimum size required for the HSE06 pot.), the memory requirements get to be pretty large even compared to other systems.</div><div><br></div><div>Problem is that while initializing with PBE may sometimes lead to a tighter screening threshold, initializing with PBE in this case invariable leads to a destabilized SCF, with no trend to convergence. I thought that, since PBE leads to very low band gap, it might be too poor of an initial guess. So I tried initializing it with (a) r2scan, (b) LDA+OPTX, and (c) LDA+TB09. These each give better estimates of the band gap and so I would expect them to provide better initial density matrices, but all of them failed to create a stable calculation. </div><div><br></div><div>Are there any guesses for what could be done and why this is happening? Does GaAs require no initial guess? That would require some pretty large memory nodes, but is theoretically doable I guess.</div><div><br></div><div>Calculation details:</div><div><br></div><div><p>&QS</p>
<p> METHOD GPW</p>
<p> EPS_DEFAULT 1e-12</p>
<p> EXTRAPOLATION ASPC </p>
<p> EPS_PGF_ORB 1e-16 </p>
<p> &END QS</p>
<p><br></p>
<p> ! Parameters needed to perform an SCF run.</p>
<p> &SCF</p>
<p> MAX_SCF 20 ! Max number of steps for an inner SCF loop</p>
<p> EPS_SCF 1e-06 ! Convergence threshold for SCF</p>
<p> SCF_GUESS RESTART ! How to initialize the density matrix</p>
<p> MAX_ITER_LUMO 400 ! Iterations for solving for unoccupied levels when running OT</p>
<p><br></p>
<p> &OT<br></p>
<p> MINIMIZER diis</p>
<p> PRECONDITIONER FULL_SINGLE_INVERSE</p>
<p> ENERGY_GAP -1</p>
<p> ALGORITHM IRAC</p>
<p> LINESEARCH 2PNT</p>
<p> ROTATION True</p>
<p> OCCUPATION_PRECONDITIONER False</p>
<p> &END OT</p>
<p> &OUTER_SCF</p>
<p> MAX_SCF 20</p>
<p> EPS_SCF 1e-06</p>
<p> &END OUTER_SCF</p>
<p> &END SCF</p>
<p> &MGRID</p>
<p> CUTOFF 350.0 </p><p> REL_CUTOFF 50 </p>
<p> NGRIDS 5 </p>
<p> PROGRESSION_FACTOR 3</p></div><div><p>&XC</p>
<p> &XC_FUNCTIONAL</p>
<p> &PBE</p>
<p> PARAMETRIZATION ORIG</p>
<p> SCALE_C 1</p>
<p> SCALE_X 0</p>
<p> &END PBE</p>
<p> &XWPBE</p>
<p> omega .11</p>
<p> scale_x -.25</p>
<p> scale_x0 1</p>
<p><br></p>
<p> &end</p>
<p> &END XC_FUNCTIONAL</p>
<p> &HF</p>
<p> FRACTION 0.25</p>
<p> &SCREENING</p>
<p> EPS_SCHWARZ 1e-07</p>
<p> EPS_SCHWARZ_FORCES 1e-07</p>
<p> SCREEN_ON_INITIAL_P T</p>
<p> SCREEN_P_FORCES False</p>
<p> &END SCREENING</p>
<p> &INTERACTION_POTENTIAL</p>
<p> CUTOFF_RADIUS 6.63</p>
<p> T_C_G_DATA t_c_g.dat</p>
<p> omega .11</p>
<p> POTENTIAL_TYPE shortrange</p>
<p> &END INTERACTION_POTENTIAL</p>
<p> &LOAD_BALANCE</p>
<p> RANDOMIZE True</p>
<p> &END LOAD_BALANCE</p>
<p> &MEMORY</p>
<p> EPS_STORAGE_SCALING 0.1</p>
<p> MAX_MEMORY 2000</p>
<p> &END MEMORY</p>
<p> &END HF</p>
<p> &END XC</p></div>
<p></p>
-- <br />
You received this message because you are subscribed to the Google Groups "cp2k" group.<br />
To unsubscribe from this group and stop receiving emails from it, send an email to <a href="mailto:cp2k+unsubscribe@googlegroups.com">cp2k+unsubscribe@googlegroups.com</a>.<br />
To view this discussion on the web visit <a href="https://groups.google.com/d/msgid/cp2k/62f09ab5-01d6-497e-8c27-f11b2693756dn%40googlegroups.com?utm_medium=email&utm_source=footer">https://groups.google.com/d/msgid/cp2k/62f09ab5-01d6-497e-8c27-f11b2693756dn%40googlegroups.com</a>.<br />