Multiwfn forum

sonalis

Member

Registered: 2023-02-25

Posts: 13

CDA of aromatic organic compounds

I wanted to carry out CDA analysis, for which I read sections 3.19 and 4.16. I have some confusions regarding my calculation. The molecule (M) shows charge transfer properties. A part of the structure has a benzene ring connected to a -NMe2. The -NMe2 part acts as the donor (fragment 1) and rest of the aromatic part acts as the acceptor (fragment 2).

1. M is a closed shell molecule. But for fragments 1 and 2, should I use them as open shell or cap each of them with a hydrogen atom to make them closed shell?

2. If I have to use them as closed shell fragments, then can I use restricted B3LYP for all of them?

3. As all of my other calculations for M has been done in B3LYP/6-311+g(d,p) (which is suitable for my molecule after screening), can I use the same for CDA calculations?

4. If yes, is the following input correct (for Gaussian user)?
   M (closed shell): #p rb3lyp/6-311+g(d,p) density pop=NO IOp(3/33=1) nosymm
   Fragments (if open shell): #p ub3lyp/6-311+g(d,p) density pop=NOAB IOp(3/33=2) nosymm
   Fragments (if closed shell): #p rb3lyp/6-311+g(d,p) density pop=NO IOp(3/33=2) nosymm

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,830

Website

Re: CDA of aromatic organic compounds

1、2 Please check the CH3NH2 CDA example in Section 4.16.2 of Multiwfn manual, your situation is very similar with this case, namely there are two open-shell fragments to be defined while the whole system is closed shell.

3 There is no reason to add diffuse function in this case, also the diffuse functions will greatly break the physical meaning of CDA. Please use 6-311G* or the better def2-TZVP instead.

4 You just use DFT, in this case pop=NO or NOAB and "density" are fully redundant. Please check input files in Section 4.16.2.

sonalis

Member

Registered: 2023-02-25

Posts: 13

Re: CDA of aromatic organic compounds

1. So, I do not need to cap hydrogens and proceed with them as open shell fragments.
2. Thank you so much for informing about the choice of basis sets.
3. I'll use the similar input as CH3NH2 [# b3lyp/def2-tzvp nosymm] along with ub3lyp for fragments and b3lyp for the whole system. Is that correct Sir? Should not IOp(3/33=2) be used?

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,830

Website

Re: CDA of aromatic organic compounds

1 Yes
3 Correct. Note that it is def2TZVP rather than def2-TZVP, the bar should be removed when writing Gaussian keyword.
If you use .fch file as input file for CDA analysis, then this IOp is fully redundant.

sonalis

Member

Registered: 2023-02-25

Posts: 13

Re: CDA of aromatic organic compounds

Thank you so much for your kind help Sir.

sonalis

Member

Registered: 2023-02-25

Posts: 13

Re: CDA of aromatic organic compounds

When I am running the calculations, I am getting the following error. Please help.

Error: segmentation violation
   rax 0000000000000000, rbx ffffffffffffffff, rcx ffffffffffffffff
   rdx 0000000000007c6f, rsp 00007fff581275a8, rbp 00007fff58127b20
   rsi 000000000000000b, rdi 0000000000007c6f, r8  0000000000000020
   r9  00007fff58126cb8, r10 00007fff581269a0, r11 0000000000000202
   r12 00007fff58127b68, r13 0000000000000000, r14 0000000000000000
   r15 00000000000003e6
  /lib64/libpthread.so.0(+0xf5d0) [0x2aaaaced25d0]
  /lib64/libc.so.6(kill+0x7) [0x2aaaad4174d7]
  /lfs/sware/g16/l9999.exe() [0x43f8f0]
  /lfs/sware/g16/l9999.exe() [0x4565b1]
  /lfs/sware/g16/l9999.exe() [0x485f37]
  /lfs/sware/g16/l9999.exe() [0x47984f]
  /lfs/sware/g16/l9999.exe() [0x4705a5]
  /lfs/sware/g16/l9999.exe() [0x4250cb]
  /lfs/sware/g16/l9999.exe() [0x40de97]
  /lfs/sware/g16/l9999.exe() [0x409138]
  /lfs/sware/g16/l9999.exe() [0x4048c0]
  /lfs/sware/g16/l9999.exe() [0x4047fb]
  /lib64/libc.so.6(__libc_start_main+0xf5) [0x2aaaad4033d5]
  /lfs/sware/g16/l9999.exe(sched_setaffinity+0xc1) [0x404729]

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,830

Website

Re: CDA of aromatic organic compounds

If the task is geometry optimization, l9999 error usually means the optimization was not converged. You need to carefully check the end of Gaussian output file.

sonalis

Member

Registered: 2023-02-25

Posts: 13

Re: CDA of aromatic organic compounds

I was doing single point calculation required for CDA analysis. (# b3lyp/def2-tzvp nosymm)

sonalis

Member

Registered: 2023-02-25

Posts: 13

Re: CDA of aromatic organic compounds

With the input # b3lyp/def2tzvp nosymm

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,830

Website

Re: CDA of aromatic organic compounds

Check end of output file to find reason.

sonalis

Member

Registered: 2023-02-25

Posts: 13

Re: CDA of aromatic organic compounds

I had not added geom=connectivity in the input file, which was the reason for the error.

I did the CDA calculations and plotted the fragment MO diagrams. I have attached the diagrams for alpha (left side) and beta (right side) electrons. However, I am unable to understand the cause of the differences in these two diagrams. The charge transfer takes place from fragment 2 to fragment 1, and I have flipped the electrons for fragment 1 here.

Thank you

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,830

Website

Re: CDA of aromatic organic compounds

This map doesn't directly exhibit how electrons transfer, but exhibit how fragment orbitals mix to complex orbitals.
To study electron transfer due to orbital mixing, you need to pay attention to d and b terms of CDA, see CDA examples in Multiwfn manual. Also Multiwfn is able to decompose d and b terms into contributions of interfragment orbital interactions to provide deeper insight.