<html><head><meta http-equiv="content-type" content="text/html; charset=us-ascii"></head><body style="overflow-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;"><div>Dear Asif,</div><div><br></div><div>there are two separate issues here: defining the functional consistently and</div><div>converging it.</div><div><br></div><div>First, alpha is not merely an SCF parameter. Standard HSE06 uses alpha = 0.25</div><div>and omega = 0.11 bohr^-1. If alpha is tuned, the short-range HF term and the</div><div>complementary short-range PBE exchange must be changed together.</div><div><br></div><div>When HYB_GGA_XC_HSE06 from LibXC is used, changing only &HF/FRACTION leaves</div><div>the semilocal HSE mixing at its original HSE06 value. This therefore no longer</div><div>defines a consistent tuned HSE functional.</div><div><br></div><div>A transparent native CP2K definition for alpha = 0.20 is:</div><div><br></div><div>&XC</div><div> &XC_FUNCTIONAL</div><div> &PBE</div><div> SCALE_C 1.0</div><div> SCALE_X 0.0</div><div> &END PBE</div><div> &XWPBE</div><div> OMEGA 0.11</div><div> SCALE_X -0.20</div><div> SCALE_X0 1.0</div><div> &END XWPBE</div><div> &END XC_FUNCTIONAL</div><div> &HF</div><div> FRACTION 0.20</div><div> &INTERACTION_POTENTIAL</div><div> OMEGA 0.11</div><div> POTENTIAL_TYPE SHORTRANGE</div><div> &END INTERACTION_POTENTIAL</div><div> &END HF</div><div>&END XC</div><div><br></div><div>The value in FRACTION and the negative value in XWPBE/SCALE_X must contain</div><div>the same alpha. Your existing screening and ADMM settings can be retained.</div><div><br></div><div>For convergence, I recommend continuation instead of jumping directly from</div><div>alpha = 0.15 to a fresh alpha = 0.20 calculation:</div><div><br></div><div>1. Converge PBE or alpha = 0.15.</div><div>2. Restart successively with alpha = 0.17, 0.19 and 0.20.</div><div>3. Use the converged wavefunction from each step as the initial guess for the</div><div> next one.</div><div>4. First converge only the occupied ground state with OT, for example:</div><div><br></div><div>&SCF</div><div> SCF_GUESS RESTART</div><div> EPS_SCF 1.0E-7</div><div> MAX_SCF 25</div><div> &OT</div><div> MINIMIZER CG</div><div> PRECONDITIONER FULL_ALL</div><div> ENERGY_GAP 0.001</div><div> LINESEARCH 3PNT</div><div> &END OT</div><div> &OUTER_SCF</div><div> EPS_SCF 1.0E-7</div><div> MAX_SCF 10</div><div> &END OUTER_SCF</div><div>&END SCF</div><div><br></div><div>5. After the occupied state has converged, perform a separate restart using</div><div> conventional diagonalization and ADDED_MOS to calculate the conduction</div><div> states and band gap. Do not combine OT and ADDED_MOS.</div><div><br></div><div>A NaN in the first SCF step is not ordinary slow convergence. Changing the</div><div>mixing method cannot repair invalid arithmetic. Remove a stale or incompatible</div><div>restart wavefunction, retry once with SCF_GUESS ATOMIC, and inspect the complete</div><div>output for overlap, basis-set or integration-grid problems.</div><div><br></div><div>I would also verify the ADMM auxiliary basis and repeat a smaller calculation</div><div>without ADMM as a reference.</div><div><br></div><div>Finally, do not calibrate alpha only until one Gamma-point supercell</div><div>eigenvalue difference equals 6.1 eV. First converge cutoff and REL_CUTOFF,</div><div>supercell or k-point sampling, ADMM, and the structural parameters. The</div><div>experimental gap also depends on temperature, electron-phonon effects and</div><div>whether it is being compared with a Kohn-Sham or quasiparticle gap.</div><div><br></div><div>If you attach the complete input and the output up to the first NaN, the</div><div>immediate numerical cause can be diagnosed more specifically.</div><div><br></div><div>Best regards,</div><div>Thomas</div><div><br><blockquote type="cite"><div>Am 29.07.2026 um 06:32 schrieb Asif Billah <asifbillah010@gmail.com>:</div><br class="Apple-interchange-newline"><div><p>Hello CP2K community,</p><p>I am calculating the bulk AlN band gap using CP2K 2024.1 with HSE, ADMM, a 72-atom 3x3x2 wurtzite supercell, and Gamma point sampling. My target band gap is approximately 6.1 eV.</p><p>Using standard diagonalization, Broyden mixing, ADMM_PURIFICATION_METHOD NONE, and alpha=0.15, I obtained a converged gap of:</p>Band gap = 5.267211 eV
<p>However, when I used alpha=0.20, the SCF initially approached about 5.7 eV but later diverged catastrophically. A retry with DIRECT_P_MIXING and a level shift produced a NaN in the first SCF step.</p><p>Could you please advise: What is the recommended stable procedure or strategy so that the calculated bandgap reaches 6.1 eV using HSE calculations with ADMM? </p><p>Any guidance would be greatly appreciated.</p><p>Thank you.</p><div><br class="webkit-block-placeholder"></div>
-- <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 visit <a href="https://groups.google.com/d/msgid/cp2k/4d47782f-5994-47fa-9366-63bda159a137n%40googlegroups.com?utm_medium=email&utm_source=footer">https://groups.google.com/d/msgid/cp2k/4d47782f-5994-47fa-9366-63bda159a137n%40googlegroups.com</a>.<br>
</div></blockquote></div><br></body></html>
<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 visit <a href="https://groups.google.com/d/msgid/cp2k/28367854-F59D-48DE-BA07-ED7CC8DA0B0A%40gmail.com?utm_medium=email&utm_source=footer">https://groups.google.com/d/msgid/cp2k/28367854-F59D-48DE-BA07-ED7CC8DA0B0A%40gmail.com</a>.<br />