2-Amino-3-methylpyridinium, 2-amino-4-methylbenzothiazolium and 2-amino-5-chloropyridinium salts. Experimental and theoretical findings

Article abstract of DOI:10.1016/j.molstruc.2020.128914

Reaction between 2-amino-3-methylpyridine, 2-amino-4-methylbenzothiazole with salicylic acid, and 2-amino-5-chloropyridine with 3-chlorobenzoic acid were carried out and respective molecular salts 1–3 were obtained. During reactions, proton transfer from acid to endocyclic N of the base was observed. Structural elucidation of all compounds was carried out with the help of FT-IR and X-ray Diffraction for single crystals. The acid/base pairs in a molecular salt are held together with the help of H- and charge-assisted bonding and additional non-covalent bonding extend the supramolecular structure. All hydrogen bonds and secondary interactions have been discussed and a detailed comparison of experimental data with theoretical calculations through AIM, NBO and Wiberg bond index analysis has been made. The non-covalent character was mainly typified by positive values of Laplacian of electronic density. The theoretical studies demonstrate that other than classical H-bonds, the non-classical and secondary non-covalent interactions are very important to direct supramolecular structure of the respective molecular salt. The DFT calculations with solvent model SMD proved a good alternative to deduce the molecular salt formation when the 0>ΔpKa>3. Moreover, global reactivity descriptors (GRD) have been calculated utilizing the energies of FMOs. The calculated values for HOMO of compounds 1–3, -5.469, -5.642 and -6.435, respectively indicate that the compounds adopt the numerical order in terms of electron donation/Lewis basicity. The molecular salt formation is more feasible for heterosynthon 1 and least for 3.

Full text of DOI:10.1016/j.molstruc.2020.128914

Journal Pre-proof
2-Amino-3-methylpyridinium, 2-amino-4-methylbenzothiazolium and
2-amino-5-chloropyrinium salts. Experimental and theoretical findings
Muhammad Hanif , Ezzat Khan , Muhammad Khalid ,
ˆ
ˆ
Muhammad Nawaz Tahir , Sara Figueiredo de Alcantara Morais ,
Ataualpa Albert Carmo Braga
PII:
S0022-2860(20)31239-4
DOI:
Reference:
https://doi.org/10.1016/j.molstruc.2020.128914
MOLSTR 128914
To appear in:
Journal of Molecular Structure
Received date:
Revised date:
Accepted date:
26 March 2020
12 July 2020
15 July 2020
Please
cite
this
article
as:
Muhammad Hanif ,
Ezzat Khan ,
Muhammad Khalid ,
ˆ
ˆ
Muhammad Nawaz Tahir , Sara Figueiredo de Alcantara Morais , Ataualpa Albert Carmo Braga ,
2-Amino-3-methylpyridinium, 2-amino-4-methylbenzothiazolium and 2-amino-5-chloropyrinium
salts. Experimental and theoretical findings, Journal of Molecular Structure (2020), doi:
https://doi.org/10.1016/j.molstruc.2020.128914
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier B.V.

Highlights
Molecular salts of N-heterocyclic bases with salicylic and benzoic acid were obtained as crystals.
Molecular structure of three salts was determined by X-ray diffraction.
Non-covalent interactions in synthons have been discussed in details.
The AIM analysis of non-covalent interactions was carried out.
The O∙∙∙O, O∙∙∙S, Cl∙∙∙Cl, etc collaborate to direct the supramolecular arrangements.

2-Amino-3-methylpyridinium, 2-amino-4-methylbenzothiazolium and 2-
amino-5-chloropyrinium salts. Experimental and theoretical findings
Muhammad Hanifa, Ezzat Khan*a, Muhammad Khalid*b, Muhammad Nawaz Tahirc, Sara
Figueirêdo de Alcântara Moraisd, Ataualpa Albert Carmo Bragad
aDepartment of Chemistry, University of Malakand, Chakdara 18800, Khyber Pakhtunkhwa,
Pakistan
bDepartment of Chemistry, Khwaja Farid University of Engineering & Information Technology,
Rahim Yar Khan-64200, Pakistan
cDepartment of Physics, University of Sargodha, 40100 Punjab, Pakistan.
dDepartamento de Química Fundamental, Instituto de Química, Universidade de São Paulo, Av.
Prof. Lineu Prestes, 748, São Paulo, 05508-000, Brazil
Corresponding author’s E-mail address: ekhan@uom.edu.pk (Prof. Dr. Ezzat Khan),
muhammad.khalid@kfueit.edu.pk; Khalid@iq.usp.br (Muhammad Khalid)

Abstract
Reaction between 2-amino-3-methylpyridine, 2-amino-4-methylbenzothiazole with
salicylic acid, and 2-amino-5-chloropyridine with 3-chlorobenzoic acid were carried out and
respective molecular salts 1-3 were obtained. During reactions proton transfer from acid to
endocyclic N of the base was observed. Structural elucidation of all compounds was carried out
with the help of FT-IR and X-ray Diffraction for single crystals. The acid/base pairs in a
molecular salt are held together with the help of H- and charge assisted bonding and additional
non-covalent bonding extend the supramolecular structure. All hydrogen bonds and secondary
interactions have been discussed and a detailed comparison of experimental data with theoretical
calculations through AIM, NBO and Wiberg bond index analysis has been made. The non-
covalent character was mainly typified by positive values of Laplacian of electronic density. The
theoretical studies demonstrate that other than classical H-bonds, the non-classical and secondary
non-covalent interactions are very important to direct supramolecular structure of the respective
molecular salt. The DFT calculations with solvent model SMD proved a good alternative to
deduce the molecular salt formation when the 0>ΔpKa>3. Moreover, global reactivity
descriptors (GRD) have been calculated utilizing the energies of FMO. The calculated values for
HOMO of compounds 1-3, -5.469, -5.642 and -6.435, respectively indicate that the compounds
adopt the numerical order in terms of electron donation/Lewis basicity. The molecular salt
formation is more feasible for heterosynthon 1 and least for 3.
Keywords: Molecular salts; salicylic acid; 3-chlorobenzoic acid; 2-amino-3-methylpyridine; 2-
amino-4-methylbenzothiazole; Secondary/noncovalent interactions.

Introduction
Non-covalent interactions play a vital role in establishing supramolecular chemistry. The
design and growth of functional materials is possible because of these interactions.[1, 2] These
forces make the backbone of supramolecular chemistry and are consisting of classical and non-
classical hydrogen bonding, electrostatic interactions and charge assisted/transfer interactions.[3-
5] Intermolecular non-covalent interactions can be used as a key molecular recognition element
in the multi-component crystalline systems. Multicomponent crystalline systems are classified
into molecular salts, solvates and cocrystals. Many active pharmaceutical ingredients (APIs) with
improved physical profile have been reported in the form of molecular salts, solvates and/or co-
crystals.[6, 7] Hydrogen bonding is a significant non-covalent interaction and plays an important
role in molecular recognition, crystal engineering, material science and biological systems.[8-10]
It is also a significant driving force in organic salts in terms of its short separation distance,
selectivity and directionality.[11-16] Generally hydrogen bonding interactions in molecular salts
have shorter intermolecular distances as compared with those in co-crystals.[17, 18] Though
molecular salts and co-crystals both can be described as multicomponent crystals but molecular
salt can be distinguished from a co-crystal based on proton transfer from an acid to a base.[19,
20] The proton transfer depends upon the basicity and acidity of interacting functional groups
that participate in hydrogen bond formation.[21, 22] An easy way for knowing the nature of
multicomponent crystals comes from looking at the difference in pKa values of acid and the
base. ―Rule of 3‖ is generally operative to predict the fate of multicomponent material, which
categorize acid base pairs into two main groups. Pairs with ΔpKa = pKa (base) – pKa (acid)> 3
will undergo proton transfer and the resultant material are molecular salts. Similarly, if the value
is less than 0, then co-crystal is the predominant product. All the components present in co-
crystal do not alter in terms of their chemical structure and function.[23] However, the rule is
improper for predicting molecular salt or co-crystal formation when the ΔpKa value is between 0
and 3 [24] where both type of multicomponent compounds are possible. In this narrow region
molecular salts and co-crystals overlap and the type of resulting complex appears to be random,
signifying that pKa is not a good predicting means in this region.[25] The molecular structure of
acids and base and position of H-acceptor atom has to be considered in this regard.
The strategy of molecular salts is frequently used to modify certain properties like
solubility, stability and hygroscopicity of a drug in the desired way.[14, 26] The carboxylic acid

functionality is very efficient in establishing the hydrogen bonding with a suitable base function
and is therefore widely in practice in crystal engineering.[27] Besides acidic groups, CH3, Cl,
NO2 groups are also active in constructing organic crystalline solids through non-covalent
bonds.[28] It is remarkable to utilize the strong and directional identification of carboxylic acid
with nitrogen containing molecules.[29-33] For this purpose the most commonly used bases are
aminopyridine derivatives. Various molecular salts involving aminopyridine/carboxylate non-
covalent interactions have been reported in literature recently.[34-37] 2-Aminopyridine and
related bases possess wide spread applications including some medicinal applications[38] and
supramolecular chemistry.[39] Such salts frequently lead to supramolecular systems which play
an important role in crystal engineering.[40, 41]
A comprehensive literature survey reveals that behaviour of 2-amino-3-methylpyridine,
2-amino-4-methylbenzothiazole and 2-amino-5-chloropyridine particularly non-covalent
interactions are unexplored. In continuation to our previous work[42] on non-covalent
interactions in molecules, here such interactions between the above stated bases with salicylic
acid and a chlorobenzoic acid (Scheme 1) are unexplored in solid state and the same have been
theoretically reproduced and validated by DFT. An experimental and theoretical detailed study
of non-covalent interactions is carried out.
Scheme 1.
Starting synthons, three bases (upper) and two acids (lower) used in this study.
Experimental section

General consideration and spectroscopic studies
Chemical reagents such as 2-amino-3-methylpyridine, 2-amino-4-methylbenzothiazole
and 2-amino-5-chloropyridine, salicylic acid, acetic acid and 3-chlorobenzaldehyde are
commercial products (TCI, Japan) and were used without further purification. Handling of
chemicals was carried out in open air, at room temperature. The melting point of compounds was
determined in sealed capillary tubes using Stuart SMP-10 (Japan) and are uncorrected. FT-IR
spectra for KBr pellets were recorded by SHIMADZU FT-IR Model 18400, in the range 4000-
400 cm-1. Single X-ray diffraction data (Mo-Kα, λ = 0.71073 Å) were collected by using Bruker
kappa APEXII CCD diffractometer at room temperature. The crystal structure solution and
refinements were accomplished by SIR97,[43] SHELXL97,[44] WinGX31[45] and
PLATON.[46]
Syntheses of heterosynthones 1, 2 and 3
2-Amino-3-methylpyridinium 2-hydroxybenzoate (1)
Molecular salt, 1 was prepared by adopting the literature procedure,[37] by treating 2-
amino-3-methylpyridine (0.313 g, 0.291 mL, 2.172 mmol) solution in 10 mL ethanol which was
drop wise added to solution (EtOH) containing equimolar amount of salicylic acid (0.300 g,
2.172 mmol). The reaction mixture was heated to reflux for 4 h, and was allowed to cool to room
temperature. White precipitates were separated by filtration, washed with EtOH and were
dissolved in ethylacetate. The solution was allowed to evaporate slowly at room temperature,
after several days colorless crystals in the solution were grown. Crystals were separated, melting
point was determined and FT-IR and single X-ray data were collected. m. p. = 171-172°C; FT-IR
data (cm-1) = 3379br, 3088br, 2975w, 1667s, 1578s, 1455s, 1252s, 1049m, 855s, 759s, 665s.
2-Amino-4-methylbenzothiazolium 2-hydroxybenzoate (2)
The proposed molecular salt 2 was prepared in the same way as molecular salt 1, by
treating 2-amino-4-methylbenzothiazole (0.2378 g, 1.448 mmol) and equimolar amount of
salicylic acid (0.200 g, 1.448 mmol) in ethanol. The expected compound was obtained as white
precipitates, which was dissolved in THF and crystals were obtained at room temperature in few
days. m. p. = 181-182°C; FT-IR data (cm-1) = 3166, 3082, 2921, 2726 (overlapped broad),
1627s, 1595s, 1451s, 1350s, 1246s, 772s, 718s.

2-Amino-5-chloropyridinium 3-chlorobenzoate (3)
Reaction between 3-chlorobenzaldehyde (0.5466 g, 3.889 mmol, in 10 mL ethanol) and
2-amino-5-chlropyridine (0.5000 g, 3.889 mmol) in the presence of 2 mL acetic acid was carried
out with the intent to get the respective Schiff base, under 6 h reflux conditions. During the
course of reaction, oxidation of aldehyde lead to in situ formation of 3-chlorobenzoic acid which
further reacted with 2-amino-5-chloropyridine via proton transfer to afford corresponding
molecular salt, 3. The clear solution was allowed to stand at room temperature wherein after few
days colourless crystals were grown, separated and studied. m.p. = 128-130°C; FT-IR data (cm-
1) = 3196w, 3070w, 2980w, 2752w, 2357m, 2355m, 1666s, 1481s, 1419s, 1354 s, 1149m, 846s,
738s, 655s.
Theoretical studies
To analyze the intermolecular interactions in the molecular salts 1-3, theoretical
calculations using Density Functional Theory (DFT) were performed. The DFT calculations
were carried out using Gaussian 09[47] package with functional B3LYP-D3[48-50] and basis set
6-311+G(d,p)[51-53] by implicit solvent SMD model with ethanol as solvent. Structures of
molecular salts 1-3 were obtained from their respective crystallographic information files (CIFs).
Single point calculations (in gas phase) of structures 1, 2 and 3 and their respective dimers were
performed to analyse the non-covalent interactions in the packed structure of the crystals.
Furthermore, optimization calculations (gas phase and with implicit solvent SMD model) of
structures 1-3 were also performed to analyse the intermolecular interactions, especially the
hydrogen bonds. The time-dependent DFT applying the TD-B3LYP-D3/6-311+G(d,p) method
was used to calculated the global reactivity parameters. The non-covalent interactions were
analysed through topological analysis using the Quantum Theory of Atoms in Molecules
(AIM)[54, 55] and natural bond orbitals (NBO) analyses.[56] The AIM analysis of structures 1-3
were carried out using AIM, all package (Version 19.10.12)[57] from their wave function file
(wfx), obtained from their calculations (optimization and single point) performed with Gaussian
09 package. The NBO analysis was performed to evaluate their intermolecular interactions of
compounds 1-3 and the NBOs were calculated with Gaussian 09 using NBO Version 3.1 and the
surfaces were plotted with ChemCraft package.[58]

Results and discussion
2-Amino-3-methylpyridine,
2-amino-4-methylbenzothiazole
and
2-amino-5-
chloropyridine possess pKa values 7.24, 4.7 and 4.67, respectively. These bases bear two
potential nucleophilic centres wherein at least one can be proton acceptor. These sites are easily
accessible to establish such interactions with a suitable reagent particularly an acid to afford
heterosynthons. We have recently reported the behaviour of analogous base derivatives with
terephthalic acid, where molecular salts were exclusively obtained.[37] The carboxylic acids
particularly benzoic acid derivatives have contributed tremendously towards the designing of
multicomponent crystals. They have been used to improve poor profile of APIs. We dealt with
several bases as ligands bearing at least two nucleophilic centres to explore the efficiency of
these centres for their coordinating ability. In this regard we have reported a number of metal
complexes wherein the endocyclic N has been proved to be more reactive/basic than the
exocyclic N atom.[59-61] While ligating property of endocyclic N (aromatic) was found to be
less than phosphorus (P) in other derivatives.[62] The basicity of various groups present in the
same molecule was explored with selected two acids salicylic acid (pKa = 2.98) and 3-
chlorobenzoic acid (pKa = 3.82). On the basis of ―rule of 3‖ these pairs give pKa = 4.26, 1.72
and 0.85 for 2-amino-3-methylpyridine/salicylic acid, 2-amino-4-methylbenzothiazole/salicylic
acid and 2-amino-5-chloropyridine/3-chlorobenzoic acid pairs, respectively. The heterosynthons
were expected to be molecular salt, for Acid/Base pair (hereafter denoted by A/B pairs) where
the difference was greater than 3 and cocrystals for other A/B two pairs. We exclusively obtained
molecular salts for all three A/B pairs, aiming to get deeper insights in the validity of the above
stated rules.
Vibrational studies
The FT-IR data of compounds 1-3 (spectra given at the end of supporting file) indicate
that molecular salts have been afforded as a result of proton transfer from acid to base. The
presence of NH2 group of the amine moiety indicates that this group is intact and does not accept
proton because of its less basicity. All molecules 1-3, show stretching bands in the range 3000-
3100 cm-1, which correspond to CH bonds.[63] In our study the IR bands at 3088 cm-1 in
molecular salt 1, at 2921 cm-1 in molecular salt 2 and at 3070-2895 cm-1 in 3, correspond to
aromatic C-H stretching vibrations. The C-H stretching vibration of methyl group in salt 1

appears at 2975 cm-1 and in salt 2 at 2855 cm-1. According to literature study the N-H stretching
as a typically broad appears in the range of 3500-3200 cm-1.[64] The N-H stretching was
observed at 3302 cm-1 in salt 1, at 3082, 3030 cm-1 in salt 2 and at 3210 and 3196 cm-1 in salt 3.
There is observed a broad absorption at 3379 cm-1 in salt 1 and at 3166 cm-1 in salt 2 associated
with stretching vibration of OH group in these compounds. Shift in the vibrational frequencies of
some CH bonds, NH and OH reveal their involvement in secondary interactions, leading to
supramolecular structure. Similarly literature study shows that C=C stretching of aromatic ring is
mostly observed in the range 1610-1558 cm-1.[65] The C-S stretching vibrations are normally
observed in finger print region in the expected range of 800-600 cm-1.[66] In this study, bands at
1578 cm-1 in salt 1, at 1595 cm-1 in salt 2 and at 1539 cm-1 in salt 3 can be assigned to C=C
stretching of aromatic rings. In salt 2 the C-S stretching vibrations appears at 772-718 cm-1. The
strong peak appearing at 738 cm-1was assigned to C-Cl stretching in salt 3. Generally the
carboxylate group (COO-) shows two types of stretching, symmetric and asymmetric stretching
in FT-IR spectrum. The symmetric stretching for carboxylate group is observed at 1425-1393
cm-1 while the asymmetric stretching appeared in the region of 1605-1585 cm-1.[67] In molecular
salt 1 symmetric stretching frequencies of C-O appeared at 1610 cm-1 while the asymmetric
stretching frequency appeared at 1667 cm-1. In molecular salt 2 symmetric stretching appeared at
1627 cm-1 while the asymmetric stretching at 1680 cm-1. Similarly in molecular salt 3 symmetric
stretching of C-O appeared at 1665 while asymmetric at 1666 cm-1. These stretching vibrations
of carboxylate group (COO-) confirms the molecular salts formation by proton transfer from
carboxylic functional group (COOH) to endocyclic nitrogen of aromatic ring of amine
derivatives. The least difference in symmetric and asymmetric frequencies of CO groups indicate
that the electron density is equally distributed within the OCO fragment, giving almost uniform
C-O bond in terms of strength.
Single crystal structure description of 1-3
Crystal structure solution and refinements parameters pertaining to compounds 1-3 are
summarized in Table 1. Compound 1 crystallizes as monoclinic system in the space group P21/c.
The asymmetric unit of compound 1 is composed of a cation of 2-amino-3-methylpyridinium
and an anion of 2-hydroxybenzoate, thus affording a molecular salt, as shown in Figure 1. The
molecular structure clearly indicates proton transfer from acid to the base. In the base unit a
comparison of N nucleophilic centres (endo- and exocyclic) shows that aromatic N is more basic

thus accepting proton from the counter acid molecule. There are three oxygen atoms in the
molecule of salicylic acid which give different bond lengths, worth explaining here for better
understanding of difference in their chemical behaviour. The distance between C1-O1 is 1.271
Å, C1-O2 is 1.249 Å and between C3-O3 is 1.354 Å (Table S1). The separation distance between
N1…O1 is 2.882 and between N2…O2 is 2.699 Å (see Table S1 for more details). Literature data
regarding C-O distance in carboxylic acids fall in the region 1.237-1.245 and 1.265-1.273 Å,
where slight variation can be explained because of intramolecular H-bonding. The bonding
between acid/base pair is different than our recently reported compounds[37] where the
interaction was established between H2N…O=C and HN+…-O-C pairs, while in the instant
compound 1, the interaction has been found between C-O-…NH2and C=O…+NH pairs. The data
indicate that the electron density is partially delocalized on O1-C1-O2 fragment but still it is easy
to decide that the negative charge is localized sufficiently on O1. The electron rich nature of O1
makes it suitable acceptor site for H-bonding, thus making an intramolecular hydrogen bonding
with OH present at ortho position to it. Further, the same centre is also involve to establish a
hydrogen bonding with NH2 group of 2-amino-3-methylpyridine resulting the formation of a
cyclic hydrogen bonded 8 membered ring, (8). This ring pattern is strong and is one of the 24
most regularly observed bimolecular cyclic hydrogen bonded ring in organic crystal
structures.[8, 68, 69]
Figure 1. Crystal structure of compound 1, thermal ellipsoids are drawn at 50% level, with
numbering scheme for non-hydrogen atoms, Intramolecular and intermolecular
hydrogen bonds are shown by dotted lines. Selected bond lengths and angles are
given in Table S1. Proton transferred to N is shown in the molecule.

Hydrogen bonding in molecules of compound 1, are given in Table S2, O3 and N1 play
an important role in establishing a 1D supramolecular structure as depicted in Figure 2. The
molecules containing (8) type ring units are linked together with the help of H-bonding
between O3 and N1. The molecules are arranged in a parallel manner with a twist angle of 103°
and are extended in a 1D fashion. The separation distance between two layers i.e., O3…N1 is
3.038 Å. Some other short contacts between C9…O1, C9…O3 and C10…O2 are also present with
a separation distance of 3.448, 3.170 and 3.549, respectively which extends the supramolecular
structure in 2D fashion (not shown).
Table 1.
Crystal refinements and solution parameters of compounds 1-3
Compound
1
2
3
Empherical formula
Formula mass
Crystal system
Space group
Temperature
Unit cell parameters
a(Å)
C13H14N2O3
246.26
Monoclinic
P21/c
C15H14N2O3S
302.34
C12H10Cl2N2O2
285.12
Orthorhombic
P212121
Monoclinic
P21/c
296
10.433(6)
8.903(5)
13.450(9)
90
6.970(11)
8.857(9)
11.755(9)
12.203(14)
90
b(Å)
14.544(2)
c(Å)
14.570(3)
α( )
90
β( )
103.277(2)
90
90
101.798(4)
90
γ( )
90
V (Å3)
1215.87(13)
1477.0 (4)
1243.6 (2)
Z
4
1.355
D (g cm-3)
(mm-1)
Crystal size
1.345
1.523
0.10
0.23
0.52
0.46×0.40×0.34
0.42×0.32×0.28
0.38×0.26×0.22

Absorption correction
Rint
Multi-scan
0.041
0.038
0.094
Number of measured/
independent/observed
7583/2659/1790
8362/3217/2351
7787/2848/1994
Reflections [I>2σ(1)]
Number of reflection/
Restraints/parameter
Δρmax, Δρmin (e A-3)
Theta range ( )
2659/0/165
3217/0/192
2848/0/172
0.15, -0.15
2.8 – 27.0
520
0.14, -0.16
2.0 – 27.0
628
0.27, -0.30
2.4-27.6
F(000)
584
R[F2>2σ(F2)],wR(F2),S
0.045, 0.134, 1.04
0.043, 0.113, 1.01
0.049, 0.125, 1.04
Figure 2. 1D supramolecular chain of compound 1, stabilized by H-bonding between N1 and
( )
O3 of neighbouring acid/base pairs. The type molecules in two layers chain are
parallel with respect to each other. Hanging contacts, other short contacts and H-
atoms are omitted for clarity.
Structural description of compound 2
Molecular salt 2 with formula [(2A4MBT)+(SA)-] crystallizes as orthorhombic colourless
crystals in the space group P212121. The structure of compound 2 with partial atomic numbering
scheme is shown in Figure 3. Compound 2 reveals similar H-bonding as in compound 1. The
synthon (salicylic acid) undergoes proton transfer to 2-amino-4-methylbenzothiazole (2A4MBT)

where the endocyclic N acts as proton acceptor. Donor/acceptor atoms in reacting species
interact to afford (8) ring system. The base molecule, 2-amino-4-methylbenzothiazole
possesses three nucleophilic centres where both N are involved in H-bonding and charge assisted
interactions, while S was not found to be involved in any sort of interaction as has been reported
in several compounds reported by our research group.[59, 70, 71] The bond lengths C1-O1
1.244(4), C1-O2 1.261(4) and C3-O3 1.331(5), shown in Table S1, are slightly different than
those in compound 1. The electron density is relatively equally distributed over O1-C1-O2
fragment with a difference in C-O bonds equal to 0.017 Å in comparison to the difference of
0.022 Å in compound 1. The C1-C2 distance is exactly equal to each other in both the
compounds 1 and 2. The C3-O3 bond is slightly shorter in comparison to compound 1. All
structural features of the counter molecule are very close to its structural analogue.[72]
Figure 3.
Molecular structure of compound 2, with numbering scheme for C, N, S and O
atoms, all hydrogen except N+-H are omitted for clarity, Hydrogen bonding (intramolecular and
intermolecular) and charge assisted interaction are shown by dotted lines. For details pertaining
to bond lengths and angles see Table S1 and S2.
The 2-amino-4-methylbenzothiazolium interacts with carboxylate group of 2-
hydroxybenzoate ion via non-covalent interactions in the same way as in compound 1. The
C1=O1 of salicylic acid with slightly more double bond characters interacts with NH2 group with
a separation distance O1…N 2.764 Å, and the same centres of the parallel layers are also linked
together (2.751 Å). The C1-O2 bond and N2 are linked through charge assisted interaction CO-

…HN+ separated by a distance of 2.715 Å (for detailed H-bonding see Table S3 and S4).These
interactions stabilize the supramolecular structure in a 1D fashion as shown in Figure 4. There
are several other interactions (C-H…π, C-H…O, N-H…C, C-H…C) which further cause to extend
the structure in a 3D fashion. The S centre of benzothiazol moiety was not involved in any sort
of interaction as has been reported in related studies[59, 73].
Figure 4. 1D chain of molecules of compound 2stabilizedthrough H-bonding (ellipsoid and
capped sticks are given for easy understanding). The (8) rings are shown shaded
in parallel layers of molecules for easy understanding.
Structural description of compound 3
Compound 3(molecular salt) was serendipitously obtained during the formation of a proposed
Schiff base under aerobic conditions as discussed in experimental part (vide supra). The same
molecular salt was successfully obtained in a reaction between the respective acid and the base.
The molecule is monoclinic having space group P21/c. The compound 3 (as shown in Figure 5)
consists of 2-amio-5-chloropyridinium and 3-chlorobenzoate ions. The proton transfer from 3-
chlorobenzoic acid to the ring nitrogen atom of 2-amino-5-chloropyridine was observed. The
same ring system as discussed above for molecular salts 1 and 2 was achieved for compound 3

( (8)). Details pertaining to selected bond lengths and angles are given in Table S1 and S2 due
to deprotonation, the resultant negative charge is distributed throughout the OCO fragment in a
way to impart bond distances O1-C1 and C1-O2 1.238(3) and 1.271(2), respectively (see Table
S1 for more details). Oxygen atom bonded to carbon in the latter case reveals comparatively
more single bond characters thus indicating sufficient electron density (polarity). The O2 and N2
are associated with each other via charge assisted interaction (CO-…+HN 2.623 Å). Base-cation
and acid-anion orient themselves in such a way that their additional donor/acceptor centres
establish H-bonding thus affording a planar Base-Acid type molecular salt. The separation
distance between O1…N1 is 2.755 Å, the data clearly differentiate hydrogen bonded sites from
charge assisted bound.
Figure 5. X-ray structure of Molecular salt 3 with complete numbering scheme (for non-
hydrogen atoms), thermal ellipsoids are drawn at 50% probability level, hydrogen
atoms are omitted for clarity and non-covalent interactions are shown by dotted
lines.
Supramolecular structure of compound 3, as a result of various secondary interactions is
shown in Figure 6. The Acid/Base pairs are linked through H-bonding and charge assisted
interactions in a molecular salt. Molecular salts or heterosynthons further interact with each other
wherein both oxygen atoms are involved. The O1 is linked to Cl1 (3.240 Å) while O2 interacts

with N1 with a separation distance of 2.926 Å, Shown in Figure 6. Because of these interactions
the supramolecular structure is stabilized in a 3D manner. Other noncovalent interactions were
also found as were present in compounds 1 and 2.
Figure 6
Chains of molecular salt 3, stabilized by secondary interactions. On the top,
molecular salts are linked through NH…O hydrogen bonding, O…Cl hanging contacts are also
shown which extend the structure further. Middle, cations and anions are shown, where 2-
aminopyridinuium cations exist in a zigzag manner. Lower, acids moieties are linked via O…Cl
and Acid/base pairs via O…N interactions. Selected hanging contacts are shown with each acid
and base fragment.
Theoretical studies
The DFT based (B3LYP-D3/6-311+G(d,p)) optimized geometries of molecular structures
(in gas and solvent phases) 1-3 were found in close agreement to their respective experimental

structures (Table S1-S4). The optimization in gas phase reveals that formation of molecular salt
is predominant for 1 and 2, while a cocrystal for compound 3. However, the optimization with
solvent model (SMD/DFT) indicates a salt formation for all structures 1-3. Because of that, the
optimized structural parameters with SMD/DFT were found in best agreement to experimental
data. Therefore, the optimized geometries of structures 1-3 with SMD model were found more
suitable hence used to calculate the global reactivity descriptors (GRDs) with TD-B3LYP-D3/6-
311+G(d,p) method to evaluate the reactivity of the structures 1-3.
The global reactivity descriptors were obtained applying the highest occupied molecular
orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energies (in eV) of structures
1-3, calculated with the following equations. The ionization potential (IP) and electron affinity
(EA) were calculated using Eq 1 and Eq. 2. [74]
Eq. 1
Eq. 2
Koopman’s theorem was used to calculate the global hardness (η)[75], electronegativity
(Χ)[76]and chemical potential (μ)[77] with following equations:[78]
[
]
Eq. 3
Eq. 4
Eq. 5
[
]
The global electrophilicity index (ω) [79, 80] was calculated using the Eq. 6 reported by
Parr et. al. as:
Eq. 6
To calculate the global softness (σ) [81], Eq. 7 was used:

Eq. 7
The frontier molecular orbitals (FMO) analysis (Table 2) of structures 1-3 showed that
the followed trend of ΔEHOMO-LUMO gap is 1<3<2. Frontier molecular orbitals (HOMO and
LUMO) are important to the reactivity description of molecules. The HOMO values can
demonstrate the electron donating characteristic while the LUMO energies demonstrate
characteristic of electrons acceptance. Furthermore, higher ΔEHOMO-LUMO indicate chemical
stability. Thus, the observed trends of ΔEHOMO-LUMO gap indicating a higher chemical stability for
structure 2, followed by 3 and 1, respectively. Although, the structures 1-3 showed small energy
differences in the FMO analysis characterizing them with similar chemical stability. The plot of
HOMO and LUMO surfaces are shown in the Figure 7. The global reactivity descriptors (GRDs)
calculated from HOMO and LUMO energies of structures 1-3 which are arranged in Table 3.
Table 2.
Frontier molecular orbitals energies for structures 1-3
Structures MOs
Energy
ΔE
HOMO -5.469
LUMO -2.456
HOMO -5.642
LUMO -1.771
HOMO -6.435
LUMO -2.786
1
2
3
3.013
3.871
3.649

Figure 7
Table 3.
Frontier molecular orbitals calculated for structures 1-3.
Global reactivity descriptors calculated for structures 1-3
Structures IP
EA
χ
η
μ
ω
Σ
5.469 2.456
5.642 1.771
6.435 2.786
3.963 1.507
3.707 1.936
4.610 1.825
-3.963
-3.707
-4.610
5.211 0.332
3.549 0.258
5.824 0.274
1
2
3
Comparing the global reactivity descriptors (Table 3) among the structures 1-3 it was
verified that the structure 3 has the higher ionization potential (IP = 6.435) followed by structures
2 and 1 (IP = 5.642 and 5.469, respectively). Although the IP values demonstrate that
compounds 1-3 possess larger stability and chemical inertness. A mutual comparison indicates
that compound 3 possess higher stability than compounds 1 and 2. Beyond that, the structure 3
also exhibits highest electronic affinity (EA = 2.786), than 1 and 2, indicating that the structure 3
accepts electrons more easily than structures 1 and 2 (EA = 2.456 and 1.771, respectively). The
global electronegativity (χ = 4.610) of structure 3 was also higher than structures 1 and 2 (χ =
3.963 and 3.707, respectively), which is expected due the presence of Cl atoms in the structure 3.
On the other hand, the global hardness descriptor of structure 1 (which has the higher energy gap
i.e., ΔEHOMO-LUMO) showed higher value (η = 1.936) than structures 3 and 2 (η = 1.825 and 1.507,
respectively). Moreover, the structure 3 exhibited lowest global chemical potential value (μ = -
4.610) followed by structures 1 and 2 (μ = -3.963 and -3.707, respectively). That is, associated
with the higher global electronegativity, the compound 3 exhibited a higher tendency to attract
electrons and hold them firmly in comparison to structures 1 and 2. The structure 3 also has
higher global electrophilicity (ω = 5.824) than structures 1 and 2 (ω = 5.211 and 3.549,

respectively). Nevertheless, the structure 1 has ω value closer to 3 indicating that these
compounds have higher electrophilic character than 2. Finally, the structure 1 showed higher
global softness value (σ = 0.332) than structures 3 and 2 (σ = 0.274 and 0.258, respectively).
Analysing the data as discussed is in the opinion that compound 1 is softest among the three.
AIM and NBO Analysis
Topological AIM analysis is widely used to characterize non-covalent interactions
through the Quantum Theory of Atoms in Molecules (QTAIM) developed by Richard Bader.[55,
82, 83] The QTAIM is based on the topological analysis of electronic density ( ), to describe
the AIM properties such as presence and strength of non-covalent interactions in a molecule. In
broad terms, the attractors (atoms) defines the distribution of gradient vector, ( ), which
establishes the fundamental conditions for topological analysis of molecular systems. The
evaluation of Laplacian of electronic density, ( ), curvature characterize the critical points
( )
( )
(CP). Thereby, the trace of Hessian Matrix of density, H
provides the CPs that are featured by rank (number of eigenvalues not null) and
signature (sum of ). The most important CP in our analysis is the bond critical point (BCP)
with rank +3 and signature -1, a topological saddle point (+3,-1) between two attractors. The
path lines that connect two attractors starting from same BCP is called bond path (BP) and it
features an interaction between two attractors, which can be a bond (covalent or ionic), a non-
covalent interaction etc. Moreover, negative Laplacian of density values ( ( )) reveals
converge of ( ), which indicates covalent interactions, while the positive values of ( )
reveals divergence of ( ) and is an indication of charge separation at the BCP that can reveal
charge assisted and other non-covalent interactions. Moreover, the ellipticity evaluates the
electronic density distribution symmetry in the BCP, thus indicates a σ bond or a triple-
bond and indicates a double bond (π-bond). Finally, the density of potential energy ( ( ))
its related with the Laplacian of density (in atomic units) by the virial theory in the Eq. 8, in
which ( ) is the density of kinetic energy.[84-86]
( ) ( ) ( )
(Eq. 8)

It was observed that all structures 1-3 containing extensive hydrogen bonding network by
AIM analysis (in gas phase or with solvent model SMD). These interactions were primarily
characterized by the presence of Bond Paths (BPs) connecting the atoms involved in the non-
covalent interactions followed by analysis of AIM data (Table 4). Moreover, the data of
optimized structures in gas phase are summarized in the supplementary material (Tables S5 and
S6).
The molecular graphs of optimized structures 1-3 are presented in the Figures 8, 9 and 10
and the respective AIM data can be seen in the Table 4. Further, the AIM data of some other
highlighted bonds of optimized structures 1-3 (in gas phase or with SMD/DFT) is arranged in the
Tables S5 and S6. The non-covalent interactions were mainly characterized by their positive
values which indicates a non-covalent interaction and its strengthening trend was determined
comparing values of electronic density ( ) (higher values indicate stronger interactions),
ellipticity ( )value closer to zero indicates more directional interactions and density of potential
energy( ) with negative values indicate greater electron stability in the interaction.
Figure 8 Molecular graph of AIM analysis of optimized structure 1 (with SMD/DFT method).
BCPs (green spheres) and BPs (solid and black lines).

Figure 9 Molecular graph of AIM analysis of optimized structure 2 (with SMD/DFT method).
BCPs (green spheres) and BPs (solid and black lines).
Figure 10 Molecular graph of AIM analysis of optimized structure 3 (with SMD/DFT
method). BCPs (green spheres) and BPs (solid and black lines).
Table 4. AIM data of highlighted bonds and non-covalent interactions of optimized structures
1-3 with SMD/DFT method. Electronic density ( ), Laplacian of density ( ),
ellipticity ( ) and density of potential energy ( ).
Bond
(e/a3)
(hartree.e/a3)
5
Structures
(e/a )

O1–H1
O2–H2
O1–H3
O1–C1
O2–C1
O1–H1
O2–H2
O2–H3
O2–H15
O1–C1
O2–C1
O1–H1
O2–H2
O1–C1
O2–C1
0.0350
0.0494
0.0569
0.3553
0.3703
0.0448
0.0511
0.0556
0.0054
0.3737
0.3512
0.0362
0.0606
0.3717
0.3625
+0.1137
+0.1326
+0.1512
-0.4887
-0.4359
+0.1282
+0.1364
+0.1515
+0.0216
-0.4241
-0.4916
+0.1140
+0.1397
-0.4310
-0.4538
0.0615
0.0415
0.0331
0.0293
0.0364
0.0453
0.0457
0.0360
1.3263
0.0390
0.0250
0.0522
0.0376
0.0367
0.0289
-0.0289
-0.0459
-0.0579
-0.9977
-1.0824
-0.0403
-0.0485
-0.0565
-0.0033
-1.1015
-0.9774
-0.0299
-0.0603
-1.0906
-1.0402
1
2
3
The H-bonds in optimized structures (1-3) were verified between O1 or O2 atoms of
carboxylic groups (SA- for 1 and 2 and 3CBA-for 3) and H1–N1- or H2–N2- of their respective
molecular cations (2A3MP, 2A4MBT or 2A5CP), as shown in the Figures 8-10. However, for
the structure 3, optimized in gas phase (Figure S.3), was verified a H-bond between N2 and H2-
O2- bond, instead O2 and H2-N2-, thus the structure 3 (optimized in gas phase) proved a co-
crystal not a molecular salt. Demonstrating the importance of solvent model use for the
molecular salt calculations because, in this case, it led to results closer to the experimental data.
Comparing the strength of H-bonds for structures 1-3 through AIM analysis (Table 4) it
was verified that the H-bond N2-H2∙∙∙O2 (ρ = 0.0494, 0.0511 and 0.0606 for 1-3, respectively) is
stronger than N1-H1∙∙∙O1 (ρ = 0.0350, 0.0448 and 0.0362 for 1-3, respectively). The strength of
the same H-bonds among the structures 1-3 was observed as 2>3≈1 for N1-H1∙∙∙O1and 3>2≈1
for N2-H2∙∙∙O2. Therefore, for both cases the structure 1 has the weakest H-bonds in the series.
To evaluate this behaviour the bond N2-H2 of molecular cations of structures 1 and 2 (2A3MP+

and 2A4MBT+, respectively) was verified that the bond N2-H2 of 2A3MP+ (ρ = 0.3344) is
stronger than N2-H2 of 2AMBT+ (ρ = 0.3305) which agrees with the expected for 2A3MP that
has a higher pKa than 2A4MBT (Table S7). The optimized molecular graphs of AIM analysis of
molecular ions 2A3MP+, 2A4MBT+, 2A5CP+, SA-_1 (SA- of structure 1), SA-_2 (SA- of
structure 2) and 3CBA- optimized with SMD/DFT is arranged in the Figure S.4. The molecular
graphs of molecular ions optimized with DFT in gas phase were not showed because in very
similar, but their AIM data is arranged in the Table S7.
Further, the nature of O-C1 bonds for structures 1-3 was evaluated thus, highest
and proved that O-C1 bonds have higher π-character than other bonds. It was verified that the
O2-C1 bond in structure 1 bears higher value of electronic density (ρ = 0.3703) and ellipticity (
= 0.0364) values than O1-C1 bond (ρ = 0.3553 and = 0.0293) in the same molecule. For
structure 1 the O2-C1 bond has highest -bond characters, hence the negative charge of
carboxylic group is more localized on O1 atom, as discussed in experimental results. Therefore,
the H-bonds between molecular ion pairs of structure 1 was established through C1–O1-∙∙∙H1–
N1H1B, a non-usual interaction as discussed in Experimental Section. Moreover, it was
observed that O1-C1 bond in compound 2 and 3 bear higher electronic densities, ρ = 0.3737 and
0.3717, respectively and their ellipticity values were found as = 0.0390 and 0.0367,
respectively. However, O2-C1 bond in compound 2 and 3 contains electronic densities ρ =
0.3512 and 0.3625 with = 0.0250 and 0.0289, respectively. The above-mentioned findings
provide enough insights that the O1-C1 bond of structures 2 and 3 has the highest π-bond
characters, as the negative charge of carboxylic group is relatively more localized at O2 atom.
Thus, the H-bonds between molecular ion pairs of structures 2 and 3 were established through
C1=O1∙∙∙H1–N1H1B and C1–O2-∙∙∙H2–+N2 interactions in agreement with experimental results.
Furthermore, the NBO charge (Table 5) and Wiberg bond index analysis (Table 7) also
demonstrate the same tendency. These analyses for other highlighted bonds, H-bonds and the
atoms involved in these interactions with optimized with SMD/DFT or gas phase is described in
supplementary material (Tables S9 and S10).
Table 5. NBO charges of O1 and O2 atoms of structures 1-3 optimized with SMD/DFT.
NBO charges
Bonds
1
2
3

O1
O2
-0.813
-0.784
-0.776
-0.811
-0.790
-0.796
Both more negative charged O atoms of structures 1 and 2 (O1 or O2 respectively) has
neighboured by anO3-H3 bond which interacts through an intramolecular H-bond O1∙∙∙H3-O3 (ρ
= 0.0569) or O2∙∙∙H3-O3 (ρ = 0.0556) for structures 1 and 2, respectively. It was verified that H-
bond O1∙∙∙H3-O3 of structure 1 is slightly stronger than analogous interaction O2∙∙∙H3-O3 of
structure 2. This behaviour agrees with experimental observations. Comparing the strength of H-
bonds O(1 or 2)∙∙∙H3-O3 and O2∙∙∙H2-N2 among structures 1 and 2, was observed through AIM
analysis (Table 4 above) that the O(1 or 2)∙∙∙H3-O3 interaction is slightly stronger than O2∙∙∙H2-
N2. This behaviour can be explained by the distance between H∙∙∙A atoms, that is short for O(1
or 2)∙∙∙H3-O3 than O2∙∙∙H2-N2 for both structures (experimentally and theoretically as can be
seen in Table S3 and S4).
However, through NBO (Table 6) and Wiberg bond index analysis (Table 7) was
observed an inverse behaviour of strength (but again with slight differences). It was verified that
the O2∙∙∙H2-N2 is slightly stronger than O(1 or 2)∙∙∙H3-O3 interaction for both 1 and 2 structures.
This invention of trend can be explained by the bond angle of O2∙∙∙H2-N2 which is higher than
O(1 or 2)∙∙∙H3-O3 (Table S4) that allows a better orbital alignment to a higher donor-acceptor
interaction for O2∙∙∙H2-N2 (as can be observed in the NBO surfaces plots for these interactions
in given in Figures 11 and 12 for O2∙∙∙H2-N2 beyond Figures S.8 and S.11 for O(1 or 2)∙∙∙H3-
O3). Another data that reinforces this explanation is the analysis of Laplacian of density ( )
from AIM analysis that presented positive values for both interactions, but higher for O(1 or
2)∙∙∙H3-O3 ( = +0.1512 and +0.1515, for structures 1 and 2, respectively) than O2∙∙∙H2-N2
interaction ( = +0.1326 and +0.1364, for structures 1 and 2 respectively), which can indicates
an interaction more dispersive for O(1 or 2)∙∙∙H3-O3 than O2∙∙∙H2-N2. These non-covalent
interactions (O∙∙∙H3-O3), as discussed in experimental results, contributes to direct the
supramolecular structure of their respective molecular salts, especially for structure1 that will be
discussed in molecular salt dimer structures (vide infra).
Table 6.
Second Order Perturbation Theory Analysis of Fock Matrix in NBO basis of main
donor-acceptor interactions of 1-3 optimized with SMD/DFT method.

Structure
Donor
n(1)O1
n(2)O2
n(2)O1
n(2)O1
n(2)O2
n(2)O2
n(2)O1
n(2)O2
Acceptor
σ*N1-H1
σ*N2-H2
σ*O3-H3
σ*N1-H1
σ*N2-H2
σ*O3-H3
σ* N1-H1
σ* N2-H2
E(2) kcal/mol
8.85
E(j)-E(i) a.u.
1.14
F(i,j) a.u.
0.090
0.120
0.115
0.110
0.096
0.097
0.093
0.143
25.44
0.69
1
23.31
0.69
21.22
0.70
17.14
0.66
2
3
16.39
0.69
14.88
0.70
35.99
0.70
Delocalization energy (E(2)), Energy difference between i (donor) and j (acceptor) NBO orbitals (E(j)-
E(i)) and Fock matrix element i and j NBO orbital (F(i,j)).

Table 7:
Wiberg bond index of atoms involved in the main H-bonds between molecular
ions of structures 1-3 optimized with SMD/DFT method.
Wiberg bond index
Bond
1
2
3
0.062
0.128
0.623
-
O1–H1
O2–H2
N2–H2
O1–H3
O2–H3
0.055
0.099
0.655
0.096
-
0.085
0.101
0.635
-
0.091
-
Meanwhile, the structure 3, which not exhibit this type of intramolecular interaction,
showed a stronger O1∙∙∙H1–N1 interaction than structure 1 but weaker than 2. The presence of Cl
atoms play an important role in the supramolecular 3 arrangement through the non-classical non-
covalent interaction i.e., Cl∙∙∙Cl beyond polarizing the electronic density of structure 3 through
withdrawal inductive effect (apart from the donor effect by resonance), and this polarization
reflects on its chemical reactivity, as demonstrated in the GRD analysis, which in the other hand
reflects on its non-covalent interactions. Especially for the O2∙∙∙H2–N2 which is stronger than
structures 1 and 2. This, non-covalent interaction combined with its chemical reactivity must be
responsible for the structure 3 is a molecular salt. Moreover, the electronic availability of Cl1
(from acid) and Cl2 (from base) was evaluated though NBO charge (Table S10) analysis which
demonstrate that Cl1 presented a negative charge (-0.026) while Cl2 presented a positive charge
(0.010). Beyond that, the Cl1presented a weaker bond with C4 atom (Cl1-C4) than Cl2 with C11
atom (Cl2-C11). This behaviour was verified by AIM analysis (Table S6) through which it was
verified a ρ = 0.1862 for Cl1-C4 and ρ = 0.1937 for Cl2-C11) and by Wiberg index analysis
(Table S9) through which it was verified a bond index 1.036 for Cl1-C4 and 1.050 for Cl2-C11.
Thus, the weaker bond Cl1-C4 than Cl2-C11 combined with the negative charge of Cl1 against
the positive charge of Cl2 demonstrated that Cl1 atom from acid is able to interact through
secondary interactions while that Cl2 from base was found inert to this type of interaction. These
observations corroborate the experimental results which demonstrated that just Cl1 from acid
interacts though secondary interactions while Cl2 was not present in the non-covalent

interactions in extended structures. The DFT calculations (in gas phase) demonstrate the same
trend observed with SMD/DFT for AIM analysis, but inverse with Wiberg analysis beyond the
both Cl atoms showed positive charge. Demonstrating again the importance of solvent model use
for these molecular salts calculation.
These observed trends for H-bonding among the structures 1-3 (2>3≈1 for N1-H1∙∙∙O1
and 3>2≈1 for N2-H2∙∙∙O2) thought AIM analysis were also observed by NBO analysis. Thus,
the NBO analysis showed (generally) higher delocalization energies (E(2)) for the stronger H-
bonds of structures 1-3.Theverified trends of NBO delocalization energies (donor-acceptor) were
2>3>1 for N1-H1∙∙∙O1 and 3>1>2 for N2-H2∙∙∙O2 (Tables 6 and S8). The NBO surfaces plot of
orbitals involved in the more energetic donor-acceptor interactions of H-bonds of structures 1-3
are shown in the Figures 11, 12 and 13. The plots of other NBO surfaces are given in the
supplementary material (Figures S.5-S.15).
Figure 11.
NBO surfaces plot of structure 1 optimized with SMD/DFT method. Non-bonding
orbitals n(1)O1 and n(2)O2 which interact with anti-bonding orbitals σ*N1-H1 and σ*N2-H2
.

Figure 12.
NBO surfaces plot of structure 2 optimized with SMD/DFT method. Non-bonding
orbitals n(2)O1 and n(2)O2 which interact with anti-bonding orbitals σ*N1-H1 and σ*N2-H2
.
Figure 13.
NBO surfaces plot of structure 3 optimized with SMD/DFT method. Non-bonding
orbitals n(2)O1 and n(2)O2 which interact with anti-bonding orbitals σ*N1-H1 and σ*N2-H2
.
It was observed that the N2-H2∙∙∙O2 bond of structure 1 the most important delocalization
energy (n(2)O2→σ*N2-H2) of this H-bond has higher E(2)(25.44 kcal/mol) than 2 (17.14 kcal/mol),
indicating a stronger H-bond for 1 than 2. However it is important emphasized that for structure
2 the interaction n(1)O2→σ*N2-H2has higher E(2) (15.13 kcal/mol) than 1 (6.69 kcal/mol), in the
same direction of AIM analysis (Table S8). Wiberg bond index analysis was performed to
further investigate these H-bonds and this analysis revealed that trends of H-bonds strength were
observed equivalent to AIM analysis: 2>3>1 for N1-H1∙∙∙O1 and 3>2≈1 for N2-H2∙∙∙O2 (Table