Multiwfn forum

mkedwards

Member

Registered: 2024-11-22

Posts: 10

Multicenter Bond Index for aromatic macrocycles

Thank you for Multiwfn!  It's amazing!

I observe that there is an implementation of the Multicenter Bond Index, but it is limited to 10 atoms.  Is there any way around this?  I have an 18-atom macrocycle for which I would like to calculate MBI to verify aromaticity (in comparison to a related non-aromatic molecule).

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,856

Website

Re: Multicenter Bond Index for aromatic macrocycles

Please do not use very old version. Due to my special algorithm for Multicenter Bond Index (see "Appendix: The extremely efficient implementation of MCBO in Multiwfn" in Section 3.11.2 of latest Multiwfn manual), current version of Multiwfn is able to calculate Multicenter Bond Index for arbitrarily large rings (even more than 100 atoms!)

mkedwards

Member

Registered: 2024-11-22

Posts: 10

Re: Multicenter Bond Index for aromatic macrocycles

I'm on the current dev version, I think; I installed it on MacOS through Homebrew (which builds from source using cmake).  Is there a way to check the date on the source code?

mkedwards

Member

Registered: 2024-11-22

Posts: 10

Re: Multicenter Bond Index for aromatic macrocycles

(base) MKE-Air:orca medwards$ multiwfn
Warning: "settings.ini" was found neither in current folder nor in the path defined by "Multiwfnpath" environment variable. Now using default settings instead

Multiwfn -- A Multifunctional Wavefunction Analyzer
Version 3.8(dev), update date: 2024-Nov-13

mkedwards

Member

Registered: 2024-11-22

Posts: 10

Re: Multicenter Bond Index for aromatic macrocycles

Maybe the command line version applies an artificial constraint?  Or maybe I'm using the wrong option?

I can calculate a multi-center bond order:

            ================ Bond order analysis ===============
-1 Define fragment 1 and 2 for options 1,3,4,7,8,10 (to be defined)
0 Return
1 Mayer bond order analysis
2 Multicenter bond order analysis
-2 Multicenter bond order analysis in NAO basis
3 Wiberg bond order analysis in Lowdin orthogonalized basis
4 Mulliken bond order (Mulliken overlap population) analysis
5 Decompose Mulliken bond order between two atoms to orbital contributions
6 Orbital occupancy-perturbed Mayer bond order
7 Fuzzy bond order analysis (FBO)
8 Laplacian bond order (LBO)
9 Decompose Wiberg bond order in NAO basis as atomic orbital pair contribution
10 Intrinsic bond strength index (IBSI)
11 AV1245 index (approximate multicenter bond order for large rings) and AVmin
20 Bond order density (BOD) and natural adaptive orbital (NAdO) analyses
2
Calculating PS matrix, please wait...
Note: Because the PS matrix may be not symmetric, the result may be dependent of inputting order of atomic indices. Settings the iMCBOtype in settings.ini to 1 is recommended, in this case the results corresponding to forward and reverse inputting orders will be automatically averaged

Input atom indices, e.g. 3,4,7,8,10   (number of centers is arbitrary)
Note: The input order must be in consistency with atomic connectivity. You can also input e.g. 4-10 if the indices are contiguous
Input -3/-4/-5/-6 will search all possible three/four/five/six-center bonds
Input 0 can return to upper level menu
1,2,3,4,5,6,22,9,10,11,12,13,14,15,16,24,19,20
The multicenter bond order:    0.1980697050
The normalized multicenter bond order:    0.9139750416

But I can't calculate a multi-center delocalization index:

                 ======== Fuzzy atomic space analysis ========
-10 Set X,Y,Z of reference point, current:    0.00000   0.00000   0.00000 Bohr
-6 Choose type of integration grid for AOM, current: Atomic grid
-5 Define the atoms to be considered in options 1, 2, 13, current: all atoms
-4 Adjust reference parameter for FLU
-3 Set the number of iterations for Becke partition, current:  3
-2 Select radius definition for Becke partition, current: Modified CSD
-1 Select method for partitioning atomic spaces, current: Becke
0 Return
1 Perform integration in fuzzy atomic spaces for a real space function
2 Calculate atomic and molecular multipole moments and <r^2>
3 Calculate and output atomic overlap matrix (AOM) in current folder
33 Calculate and output fragment overlap matrix (FOM) in current folder
4 Calculate localization (LI) and delocalization index (DI)
44 Calculate fragment LI (FLI) and interfragment DI (IFDI)
5 Calculate PDI (Para-delocalization index)
6 Calculate FLU (Aromatic fluctuation index)
7 Calculate FLU-pi
8 Perform integration in fuzzy overlap region for a real space functions
9 Calculate condensed linear response kernel (CLRK)
10 Calculate PLR (Para linear response index)
11 Calculate multi-center delocalization index
12 Calculate information-theoretic aromaticity index (ACS Omega, 3, 18370)
13 Calculate atomic effective volume, free volume, polarizability and C6 coefficient
11
Note: The highest  2565 virtual orbitals will not be taken into account
Allocating memory...
Radial points:   45    Angular points:  170   Total:      7650 per center
Please wait...

Progress: [##################################################]   100.0 %     -
Calculation took up      99 seconds wall clock time
Error of AOM is    0.00280439
Maximum diagonal deviation to 1:     0.000345 at orbital    95
Maximum nondiagonal deviation to 0:  0.000149 between orbitals    70   112

Warning: The integration is not very accurate
To improve accuracy, please try one or some of following treatments:
(1) Enlarge "radpot" and "sphpot" in settings.ini
(2) Set "radcut" in settings.ini to 0
(3) Choose option -6 to change to much more expensive molecular grid
(4) If diffuse functions were heavily employed, remove them

Input atom indices, e.g. 3,4,7,8,10    (Up to 10 atoms)
Input q can return to upper level menu
1,2,3,4,5,6,22,9,10,11,12,13,14,15,16,24,19,20
Please wait...
Multicenter DI:    0.0000000

Multicenter DI in normalized form:     0.0000000

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,856

Website

Re: Multicenter Bond Index for aromatic macrocycles

For a large ring, please only use multi-center bond order (alternatively, use AV1245), do not use multicenter DI (only supports up to 10 atoms). Your Multiwfn version is new enough.

mkedwards

Member

Registered: 2024-11-22

Posts: 10

Re: Multicenter Bond Index for aromatic macrocycles

AV1245 is a great suggestion.  Thank you!

mkedwards

Member

Registered: 2024-11-22

Posts: 10

Re: Multicenter Bond Index for aromatic macrocycles

I don't really understand the numbers I'm getting out of AV1245, though.

1,2,3,4,5,6,22,9,10,11,12,13,14,15,16,24,19,20
Number of selected atoms:    18
Atomic sequence:
     1     2     3     4     5     6    22     9    10    11    12    13
    14    15    16    24    19    20

4-center electron sharing index of     1     2     4     5:    0.17676882
4-center electron sharing index of     2     3     5     6:   -0.50162138
4-center electron sharing index of     3     4     6    22:    0.36387824
4-center electron sharing index of     4     5    22     9:   -0.16439330
4-center electron sharing index of     5     6     9    10:   -0.49054669
4-center electron sharing index of     6    22    10    11:   -0.81549232
4-center electron sharing index of    22     9    11    12:   -0.14739340
4-center electron sharing index of     9    10    12    13:   -0.45451047
4-center electron sharing index of    10    11    13    14:   -0.39040796
4-center electron sharing index of    11    12    14    15:    0.12690133
4-center electron sharing index of    12    13    15    16:    0.45539971
4-center electron sharing index of    13    14    16    24:   -0.25926207
4-center electron sharing index of    14    15    24    19:    0.14457754
4-center electron sharing index of    15    16    19    20:    0.11225174
4-center electron sharing index of    16    24    20     1:    0.17069799
4-center electron sharing index of    24    19     1     2:   -0.06304604
4-center electron sharing index of    19    20     2     3:   -0.17882162
4-center electron sharing index of    20     1     3     4:    0.20749270

AV1245 times 1000 for the selected atoms is  -94.86262021
AVmin times 1000 for the selected atoms is    63.046041 (   24   19    1    2)

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,856

Website

Re: Multicenter Bond Index for aromatic macrocycles

Please read Section 3.11.10 of Multiwfn manual, the theoretical background has been carefully introduced.

mkedwards

Member

Registered: 2024-11-22

Posts: 10

Re: Multicenter Bond Index for aromatic macrocycles

Thank you; that's very helpful.  It looks like we'll be wanting the accuracy of MCBO with NBO (we're using a basis set that includes diffuse functions), so we have a direction in which to go.  Thank you for Multiwfn; we'll be citing you!

mkedwards

Member

Registered: 2024-11-22

Posts: 10

Re: Multicenter Bond Index for aromatic macrocycles

Do you think that feeding ORCA's "MP2 natural orbitals" into Multiwfn would allow the computation of an MCBO without the problems that MCBO "option 2" (no external NBO program) encounters with basis sets that include diffuse functions?

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,856

Website

Re: Multicenter Bond Index for aromatic macrocycles

The two factors are independent of each other. Standard definition of MCBO is incompatible with diffuse functions, irrespective of the choice of theoretical method.

mkedwards

Member

Registered: 2024-11-22

Posts: 10

Re: Multicenter Bond Index for aromatic macrocycles

Thank you.  Do you happen to know whether JANPA fixes the problem in the way that NBO does?  I have no particular objection to NBO myself, but my PI would like to know what tools exist in this area.

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,856

Website

Re: Multicenter Bond Index for aromatic macrocycles

I know noting about JANPA.

If you have Gaussian, you can freely use the NBO 3.1 embedded in it. In addition, licence of NBO 7.0 is merely 100 USD.

mkedwards

Member

Registered: 2024-11-22

Posts: 10

Re: Multicenter Bond Index for aromatic macrocycles

Just to follow up:  AVmin (computed on orbitals re-optimized using a basis set with no diffuse functions) was exactly the metric I needed.  Thank you!