Poisson solver issue for spin-unrestricted calculation of heme
79 views
Skip to first unread message
Lukas Krieger
Nov 29, 2022, 7:47:58 AM11/29/22
to cp2k
Hi all,
my name is Lukas and I get severe problems while simulating a heme molecule without a ligand and with Fe^2+ in a quintet state. I always get the warning that the charge density does not vanish at the edges of the box. Increasing the box to 35 Angström has no effect what is surprising as the spin-restricted calculation for the singlet Fe^2+-heme runs even with a 20 A box! I'll also submit my inp and out files. I stopped the energy calculation after a couple of steps as the values are completly out of range. I'll also submit a spin-restricted calculation (Bound_state_3) to compare.
Jürg Hutter
Nov 30, 2022, 7:27:02 AM11/30/22
to cp...@googlegroups.com
Hi
it seems (cannot verify without full input) that the problem is the initial guess
of your quintet calculation. The initial energy is +5000 instead about -500.
regards
JH
________________________________________
From: 'Lukas Krieger' via cp2k <cp...@googlegroups.com>
Sent: Tuesday, November 29, 2022 1:47 PM
To: cp2k
Subject: [CP2K:18111] Poisson solver issue for spin-unrestricted calculation of heme
Hi all,
my name is Lukas and I get severe problems while simulating a heme molecule without a ligand and with Fe^2+ in a quintet state. I always get the warning that the charge density does not vanish at the edges of the box. Increasing the box to 35 Angström has no effect what is surprising as the spin-restricted calculation for the singlet Fe^2+-heme runs even with a 20 A box! I'll also submit my inp and out files. I stopped the energy calculation after a couple of steps as the values are completly out of range. I'll also submit a spin-restricted calculation (Bound_state_3) to compare.
--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com<mailto:cp2k+uns...@googlegroups.com>.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/308e2e71-a67e-4692-bd93-2cef04ecd662n%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/308e2e71-a67e-4692-bd93-2cef04ecd662n%40googlegroups.com?utm_medium=email&utm_source=footer>.
it seems (cannot verify without full input) that the problem is the initial guess
of your quintet calculation. The initial energy is +5000 instead about -500.
regards
JH
________________________________________
From: 'Lukas Krieger' via cp2k <cp...@googlegroups.com>
Sent: Tuesday, November 29, 2022 1:47 PM
To: cp2k
Subject: [CP2K:18111] Poisson solver issue for spin-unrestricted calculation of heme
Hi all,
my name is Lukas and I get severe problems while simulating a heme molecule without a ligand and with Fe^2+ in a quintet state. I always get the warning that the charge density does not vanish at the edges of the box. Increasing the box to 35 Angström has no effect what is surprising as the spin-restricted calculation for the singlet Fe^2+-heme runs even with a 20 A box! I'll also submit my inp and out files. I stopped the energy calculation after a couple of steps as the values are completly out of range. I'll also submit a spin-restricted calculation (Bound_state_3) to compare.
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com<mailto:cp2k+uns...@googlegroups.com>.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/308e2e71-a67e-4692-bd93-2cef04ecd662n%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/308e2e71-a67e-4692-bd93-2cef04ecd662n%40googlegroups.com?utm_medium=email&utm_source=footer>.
Lukas Krieger
Nov 30, 2022, 7:34:44 AM11/30/22
to cp...@googlegroups.com
Hi Jürg,
thank you a lot for your answer. Do you have an idea how to choose a better guess for Fe while keeping the quintet multiplicity? I really need the quintet as it is experimentally determined that unbound heme contains quintet-Fe^2+.
Sincerely
Lukas
PS: You have said that my input is not complete. Which sections are missing there?
You received this message because you are subscribed to a topic in the Google Groups "cp2k" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cp2k/YxAw86EsOpw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ZR0P278MB0759B015F15D70EEE61D74669F159%40ZR0P278MB0759.CHEP278.PROD.OUTLOOK.COM.
Jürg Hutter
Nov 30, 2022, 7:43:14 AM11/30/22
to cp...@googlegroups.com
Hi
the coordinate file is missing.
I would start with
LSD
Charge -2
Multiplicity 5
without any specific BS input for the atoms.
regards
JH
________________________________________
From: 'Lukas Krieger' via cp2k <cp...@googlegroups.com>
Sent: Wednesday, November 30, 2022 1:34 PM
To: cp...@googlegroups.com
Subject: Re: [CP2K:18116] Poisson solver issue for spin-unrestricted calculation of heme
Hi Jürg,
thank you a lot for your answer. Do you have an idea how to choose a better guess for Fe while keeping the quintet multiplicity? I really need the quintet as it is experimentally determined that unbound heme contains quintet-Fe^2+.
Sincerely
Lukas
PS: You have said that my input is not complete. Which sections are missing there?
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ZR0P278MB0759B015F15D70EEE61D74669F159%40ZR0P278MB0759.CHEP278.PROD.OUTLOOK.COM.
--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com<mailto:cp2k+uns...@googlegroups.com>.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAGwKOOZjyTZHNchSXENSO3L-%3DFCNScH2odTNoT13qgbxY0POEQ%40mail.gmail.com<https://groups.google.com/d/msgid/cp2k/CAGwKOOZjyTZHNchSXENSO3L-%3DFCNScH2odTNoT13qgbxY0POEQ%40mail.gmail.com?utm_medium=email&utm_source=footer>.
the coordinate file is missing.
I would start with
LSD
Charge -2
Multiplicity 5
without any specific BS input for the atoms.
regards
JH
________________________________________
From: 'Lukas Krieger' via cp2k <cp...@googlegroups.com>
To: cp...@googlegroups.com
Subject: Re: [CP2K:18116] Poisson solver issue for spin-unrestricted calculation of heme
Hi Jürg,
thank you a lot for your answer. Do you have an idea how to choose a better guess for Fe while keeping the quintet multiplicity? I really need the quintet as it is experimentally determined that unbound heme contains quintet-Fe^2+.
Sincerely
Lukas
PS: You have said that my input is not complete. Which sections are missing there?
Jürg Hutter <hut...@chem.uzh.ch<mailto:hut...@chem.uzh.ch>> schrieb am Mi., 30. Nov. 2022, 13:27:
Hi
it seems (cannot verify without full input) that the problem is the initial guess
of your quintet calculation. The initial energy is +5000 instead about -500.
regards
JH
________________________________________
From: 'Lukas Krieger' via cp2k <cp...@googlegroups.com<mailto:cp...@googlegroups.com>>
Hi
it seems (cannot verify without full input) that the problem is the initial guess
of your quintet calculation. The initial energy is +5000 instead about -500.
regards
JH
________________________________________
Sent: Tuesday, November 29, 2022 1:47 PM
To: cp2k
Subject: [CP2K:18111] Poisson solver issue for spin-unrestricted calculation of heme
Hi all,
my name is Lukas and I get severe problems while simulating a heme molecule without a ligand and with Fe^2+ in a quintet state. I always get the warning that the charge density does not vanish at the edges of the box. Increasing the box to 35 Angström has no effect what is surprising as the spin-restricted calculation for the singlet Fe^2+-heme runs even with a 20 A box! I'll also submit my inp and out files. I stopped the energy calculation after a couple of steps as the values are completly out of range. I'll also submit a spin-restricted calculation (Bound_state_3) to compare.
--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com<mailto:cp2k%2Bunsu...@googlegroups.com><mailto:cp2k+uns...@googlegroups.com<mailto:cp2k%2Bunsu...@googlegroups.com>>.
To: cp2k
Subject: [CP2K:18111] Poisson solver issue for spin-unrestricted calculation of heme
Hi all,
my name is Lukas and I get severe problems while simulating a heme molecule without a ligand and with Fe^2+ in a quintet state. I always get the warning that the charge density does not vanish at the edges of the box. Increasing the box to 35 Angström has no effect what is surprising as the spin-restricted calculation for the singlet Fe^2+-heme runs even with a 20 A box! I'll also submit my inp and out files. I stopped the energy calculation after a couple of steps as the values are completly out of range. I'll also submit a spin-restricted calculation (Bound_state_3) to compare.
--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/308e2e71-a67e-4692-bd93-2cef04ecd662n%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/308e2e71-a67e-4692-bd93-2cef04ecd662n%40googlegroups.com?utm_medium=email&utm_source=footer>.
--
You received this message because you are subscribed to a topic in the Google Groups "cp2k" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cp2k/YxAw86EsOpw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cp2k+uns...@googlegroups.com<mailto:cp2k%2Bunsu...@googlegroups.com>.
--
You received this message because you are subscribed to a topic in the Google Groups "cp2k" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cp2k/YxAw86EsOpw/unsubscribe.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ZR0P278MB0759B015F15D70EEE61D74669F159%40ZR0P278MB0759.CHEP278.PROD.OUTLOOK.COM.
--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com<mailto:cp2k+uns...@googlegroups.com>.
Lukas Krieger
Nov 30, 2022, 7:48:51 AM11/30/22
to cp...@googlegroups.com
Hi,
ah true, I'll send it to you in a couple of minutes. My first idea was also to set multiplicity to 5 but then I recognized that 2 of the nitrogens around the iron are negatively charged so that the charge compensates to 0 in the inner region (qtot=-2 comes from the carboxy-groups). As nitrogen has 3 unpaired P-electrons due to Hund's rule I thought that I have to give a start configuration at least for Fe^2+, N and N^-.
Sincerely
Lukas
To unsubscribe from this group and all its topics, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ZR0P278MB0759DE797A4FA4283C7949A89F159%40ZR0P278MB0759.CHEP278.PROD.OUTLOOK.COM.
0 new messages