Using NBO to try to characterize H-bond interaction is not a good idea. The main nature of typical H-bond is electrostatic, however this vital component cannot be faithfully revealed by E(2) analysis. In addition, even the charge-transfer component in H-bond can also not be correctly reflected by E(2) energy, as demonstrated in J. Phys. Chem. A, 2017, 121 (7), pp 1531–1534.

There are lots of much better ways of demonstrating existence and revealing strength of H-bond, such as Atoms-in-molecules (AIM), noncovalent interaction (NCI), Independent Gradient Model (IGM), electrostatic potential on van der Waals surface and so on. They are fully supported by my wavefunction analysis code Multiwfn (http://sobereva.com/multiwfn). If you are a Gaussian user, the .fch or .wfn/wfx file can be directly used as input file. Please check Section 4.A.5 of Multiwfn manual for a brief review of all methods supported by Multiwfn that may be used to analyze weak interactions.

Using NBO to try to characterize H-bond interaction is not a good idea. The main nature of typical H-bond is electrostatic, however this vital component cannot be faithfully revealed by E(2) analysis. In addition, even the charge-transfer component in H-bond can also not be correctly reflected by E(2) energy, as demonstrated in J. Phys. Chem. A, 2017, 121 (7), pp 1531–1534.

There are lots of much better ways of demonstrating existence and revealing strength of H-bond, such as Atoms-in-molecules (AIM), noncovalent interaction (NCI), Independent Gradient Model (IGM), electrostatic potential on van der Waals surface and so on. They are fully supported by my wavefunction analysis code Multiwfn (http://sobereva.com/multiwfn). If you are a Gaussian user, the .fch or .wfn/wfx file can be directly used as input file. Please check Section 4.A.5 of Multiwfn manual for a brief review of all methods supported by Multiwfn that may be used to analyze weak interactions.

Thank you Tian Lu! I will check the alternatives

Hi Dear

How much E2 value in the lines LP ( 1) O a /***. LP*( 1) H c?

If E2 greater than 80kcal/mol, there is a bond between O-H, for smaller than value, there is donor acceptor interaction.

Hello Fazlolah,

Considering a molecule that has 3 h-bonds, I get 3 energies for 3 LP ( 1) O a /***. LP*( 1) H c pairs: 15 kcal, 12 kcal and 3 kcal. However, the results from some of the methodologies proposed by Tian Lu suggest the 3 h-bonds are similar, so I assume these values cannot correspond to E(2) from the interactions. I still haven't figure out why NBO analysis didn't output any BD*( 1) O - H lines.

You should reduce occupancy value in NBO input at end coordinate of atoms in following form:

0 1

C 0.1234 0.986 1.265

...

....

...

$NBO Occupancy=1.89 $end

Also, you can use atomic in molecule (QTAIM) software. If this software does not show bond, and this option does not work, probably, your molecular structure is wrong.

So, you must redesign and recalculate NBO and AIM jobs.

Best regards.

Fazlolah Eshghi turns out that using pop=NBO in gaussian applied the RESONANCE keyword to the analysis. Like you said, changing the occupancy threshold allowed to recover the BD* ( 1) lines I was looking for. Than you!

Tian Lu

As you suggested, NBO is not good to characterize Hydrogen bond, What do you suggest for halogen bonding, as this interaction is sometimes referred to as electrostatic, sometime dispersive?

Prasanta Bandyopadhyay Commonly I suggest using these ways to characterize Halogen bonding:

1 Electrostatic potential (ESP) on monomer vdW surface (colored isosurface map) and monomer ESP contour line map

2 Density difference between dimer and monomer

3 NCI analysis or IGM analysis

4 SAPT energy decomposition analysis

Other analyses may be also useful, such as mutual penetration distance analysis, AIM analysis, valence electron density analysis and so on.

I agree with Prof Tian Lu , AIM and NCI are useful approaches to characterise H-Bonds. You can do this using Multiwfn software, based on Gaussian output. All details are described in Multiwfn manual (attached PDF file)

Prasanta Bandyopadhyay It is not decided that NBO calculations do not describe H-bonds . First, many papers shows the charge tansfer nature of H-Bonds, precisely described by NBO calculations [1-3]. Second, SAPT theory seems to present numerical anomalies [4], and third, AIM analyzes do not appear to be as reliable for determining H-bonds strength [5]. @Tia lu, For Dr. A.J stone's article, Frank Weinhold published a "overwhelming" response [6].

[1]- Z. Zhang ,D. Li,W. Jiang & Z. Wang, The electron density delocalization of hydrogen bond systems, ADVANCES IN PHYSICS: X 2018, VOL. 3, NO. 1, 1428915

[2]- Weinhold, F. Resonance character of hydrogen-bonding interactions in water and other H-bonded species. In Peptide Solvation and H-Bonds: Advances in Protein Chemistry; Baldwin, R. L.; Baker, D., Eds.; Elsevier: Amsterdam, The Netherlands, 2006; Vol. 72, pp 121− 155.

[3]- Gilli, P.; Bertolasi, V.; Gilli, G.; et al. Covalent nature of the strong homonuclear hydrogen-bond – Study of the O-H···O system by crystal-structure correlation methods. J. Am. Chem. Soc. 1994, 116, 909−915.

[4]- Lao, K. U.; Schaffer, R.; Jansen, G.; Herbert, J. M. Accurate ̈ description of intermolecular interactions involving ions using symmetry-adapted perturbation theory. J. Chem. Theory Comput. 2015, 11, 2473−2486

[5]- S. Shabazian, Why Bond Critical Points AreNot “Bond” CriticalPointsChem. Eur.J. 2018,24 ,5401 –5405

[6]- F. Weinhold and E. D. Glendening , Comment on “Natural Bond Orbitals and the Nature of the Hydrogen Bond” , J. Phys. Chem. A 2018, 122, 724−732