From 23721e0abb1e72860cd3ab2779eab75578eb0af9 Mon Sep 17 00:00:00 2001 From: Yury <38009554+PolyachenkoYA@users.noreply.github.com> Date: Thu, 21 Aug 2025 16:46:02 -0400 Subject: [PATCH] Update RELAX_MULTIPLICITY description --- src/input_cp2k_dft.F | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/src/input_cp2k_dft.F b/src/input_cp2k_dft.F index bbe6ae6df1..e25c37825e 100644 --- a/src/input_cp2k_dft.F +++ b/src/input_cp2k_dft.F @@ -217,7 +217,15 @@ CONTAINS "of alpha and beta MOs due to the initially "// & "defined multiplicity, but rather follow the Aufbau principle. "// & "A value greater than zero activates this option. "// & - "Larger tolerance values increase the probability for a spin flip. "// & + "If alpha/beta MOs differ in energy less than this tolerance, "// & + "then alpha-MO occupation is preferred even if it is higher "// & + "in energy (within the tolerance). "// & + "Such spin-symmetry broken (spin-polarized) occupation is used "// & + "as SCF input, which (is assumed to) bias the SCF "// & + "towards a spin-polarized solution. "// & + "Thus, larger tolerance increases chances of ending up "// & + "with spin-polarization. "// & + "More details in a GitHub `thread `. "// & "This option is only valid for unrestricted (i.e. spin polarised) "// & "Kohn-Sham (UKS) calculations.", & usage="RELAX_MULTIPLICITY 0.00001", &