Multiwfn forum

Pregabalin

Member

Registered: 2020-05-30

Posts: 6

Bond Paths

Hi,

I am a newcomer to both computational chemistry and this software.

I have performed topological analysis on silylene iridacycle as shown in the attached images.

One of my objectives is to investigate the curvature/distortion of the bond path associated in Ir-Si-Se motif
using the difference between bond path length and interatomic distance/distance between the corresponding nuclear attractors
in which I expected this parameter to positive for all cases.

However, as shown in the result below,
I have obtained a negative value of the parameter for Ir-Se and Ir-Si whose BP is straight and slightly curve respectively.

So, Is there anything wrong with my method that I obtained such unexpected results? or perhaps I miscalculated? or I misunderstand the concept of BP in the first place?

--------------------

Method:
1.wB97XD/jorge-DZP-DKH int=DKHSO : to generate .47 file
2. used .47 file to generate .31 to .40 file
3. used .31 and .39 as an input in the multiwfn program

----------------------------
Multiwfn Output

!!!
Ir-Se
# 105 CP: 161 (3,-1) ---> CP: 30 (3,-3) Length: 2.34000
# 106 CP: 161 (3,-1) ---> CP: 61 (3,-3) Length: 2.34000
c30 c61
The distance is 4.752838 Bohr ( 2.515094 Angstrom)

!!!
Ir-Si
# 109 CP: 164 (3,-1) ---> CP: 30 (3,-3) Length: 2.73000
# 110 CP: 164 (3,-1) ---> CP: 52 (3,-3) Length: 1.47000
c30 c52
The distance is 4.275782 Bohr ( 2.262647 Angstrom)

Se-Si
# 127 CP: 175 (3,-1) ---> CP: 52 (3,-3) Length: 2.07000
# 128 CP: 175 (3,-1) ---> CP: 61 (3,-3) Length: 2.70000
c61 c52
The distance is 4.504549 Bohr ( 2.383705 Angstrom)

I await your response
Sincerely,
Pregabalin

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,776

Website

Re: Bond Paths

The reason is that in old version, during generation of bond paths, if current position nearly reaches a critical point, the generation will be stopped, and thus there is a very small distance between the final point of the path and the critical point.

I just updated Multiwfn 3.7(dev). In the new version, the nearly reached critical point is added to the path as the final point, thus the reported path length now become more reasonable, and correspondingly, path length must be equal or larger than the distance between the two critical points. Please download the new version and have a try.

BTW: It is strange why .31 and .39 are used as input file. Commonly, for Gaussian users, .fch/.wfn/.wfx should be used as input file for analysis of real space functions. I cannot 100% guarantee that AIM analysis based on .31 with .39 is always reasonable.

Pregabalin

Member

Registered: 2020-05-30

Posts: 6

Re: Bond Paths

Hi, sir

First of all, I wholeheartedly appreciated your reply.

The reason I applied .31 and .39 files as input is that I aim to study NLMOs contribution at the BCPs.

Is it a "sacrilegious" way of doing?

Or Is it better to perform NBCP implemented in NBOPro to investigate the NLMO contribution?

Sincerely,
Pregabalin

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,776

Website

Re: Bond Paths

According to your statement, using .31+.39 is fully reasonable for you aim. The analysis result should be correct.

However, since this usage is relatively uncommon, I suggest you input below commands after loading .31 and .39:
100
4
1
This task will integrate electron density over the whole space. If the result is very close to actual number of electrons (e.g. error less than 0.01), that implies all kinds of analysis of real space function should be completely correct.

Pregabalin

Member

Registered: 2020-05-30

Posts: 6

Re: Bond Paths

Hi again,

Firstly, Thank you very much for your advice. I have performed the integration in which I found the total number of the electron from .31/.39 to be somewhat higher than .fchk [ 466.0143907251 and 466.000933 respectively ]. Does this mean the results from .31/.39 are unreasonable? Although, both of them are fairly equal; What do you think?

-------------------------------------------------------------------------------------------

.31/.39
rho lap G V
Ir-Se 0.075776034 0.147724563 0.054669412 -0.072407683
Ir-Si 0.108848363 -0.0470047152 0.063333872 -0.138417862
Se-Si 0.072182714 -0.015621025 0.030569424 -0.065044104
RCP 0.067685053 0.079127 0.045450094 -0.06443654

.fchk
rho lap G V
Ir-Se 0.075720075 0.148735938 5.49E-02 -0.072652482
Ir-Si 0.108832057 -0.047047366 0.063330011 -0.138421864
Se-Si 0.072193593 -0.015434965 0.030554198 -0.064967138
RCP 0.067797297 0.105374383 0.045293538 -0.064243481

Also, having tried the new version of Multiwfn, my BP length issue still persists for both .fchk input and .31/.39 input;

-------------------------------------------------------------------------------------------
.31/.39

Ir Si
# 109 CP: 164 (3,-1) ---> CP: 30 (3,-3) Length: 2.73000
# 110 CP: 164 (3,-1) ---> CP: 52 (3,-3) Length: 1.47000

a1 a3
The distance is 4.275782 Bohr ( 2.262646 Angstrom)
c30 c52
The distance is 4.275782 Bohr ( 2.262647 Angstrom)

Ir Se
# 105 CP: 161 (3,-1) ---> CP: 30 (3,-3) Length: 2.34000
# 106 CP: 161 (3,-1) ---> CP: 61 (3,-3) Length: 2.34000

a1 a2
The distance is 4.752838 Bohr ( 2.515094 Angstrom)
c30 c61
The distance is 4.752838 Bohr ( 2.515094 Angstrom)
-------------------------------------------------------------------------------------------

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,776

Website

Re: Bond Paths

If you do not quite focus on data accuracy, the marginal error of .31+.39 result should be tolerable.

It seems that I uploaded a wrong file yesterday. I just uploaded again, please re-download.

Pregabalin

Member

Registered: 2020-05-30

Posts: 6

Re: Bond Paths

It worked!

I wholeheartedly appreciated your help.

Sincerely,
Pregabalin