Synthesis and experimental versus theoretical research on spectroscopic and electronic properties of 3-methyl-4-nitroisothiazole

Article abstract of DOI:10.1016/j.molstruc.2019.06.028

The alternative way of synthesis of valuable 5-amino-3-methylisothiazole and 3-methyl-4-nitroisothiazole has been presented and described. It appeared that new, undescribed 3,3′-dimethyl-4,4′-dinitro-5,5′-bisisothiazole has been produced as a side-product during synthesis. Both experimental and calculated spectral and electronic properties of 3-methyl-4-nitroisothiazole have been extensively researched and discussed. The complete vibrational assignments were made on the basis of potential energy distributions PED. The fully anharmonic infrared and Raman spectra with calculated anharmonic intensities for fundamental bands, overtones as well as combination bands are presented. Vibrational spectra predicted by anharmonic approximation agree very well with experimental data. The stability of molecules of 3-methyl-4-nitroisothiazole arising from hyper conjugative interaction has been estimated using natural bond orbital (NBO 6.0) analysis. The combined frontier molecular orbital and NBO analysis of canonical molecular orbitals was used to predict the most probable sites of 3-methyl-4-nitroisothiazole molecule where single electron can be accepted to and detached from. Measured reduction potential of 3-methyl-4-nitroisothiazole and calculated electron affinity point out that this compound is susceptible to reduction at the similar level of 1-methyl-3-nitropyrazole and this susceptibility is much more than respective 1-methyl-4-nitropyrazole. UV-Vis spectra analysis reveals a nature of valence electron excitation and electron transition of 3-methyl-4-nitroisothiazole. Besides, unambiguous assignment of NMR signal shifts of carbon atoms of isothiazole ring of 3-methyl-4-nitroisothiazole, 5-amino-3-methylisothiazole and intermediate and side products of synthesis was conducted thanks to full detailed analysis of ¹H, ¹³C NMR spectra and their two dimensional (2D) variants. Additionally, molar enthalpy of vaporization (ΔH_vap) of 3-methyl-4-nitroisothiazole has been estimated (ΔH_vap = 52.2 kJ/mol).

Full text of DOI:10.1016/j.molstruc.2019.06.028

Accepted Manuscript
Synthesis and experimental versus theoretical research on spectroscopic and
electronic properties of 3-methyl-4-nitroisothiazole
Andrzej Regiec, Piotr Wojciechowski
PII:
S0022-2860(19)30742-2
DOI:
https://doi.org/10.1016/j.molstruc.2019.06.028
MOLSTR 26670
Reference:
To appear in: Journal of Molecular Structure
Received Date: 9 May 2019
Accepted Date: 6 June 2019
Please cite this article as: A. Regiec, P. Wojciechowski, Synthesis and experimental versus theoretical
research on spectroscopic and electronic properties of 3-methyl-4-nitroisothiazole, Journal of Molecular
Structure (2019), doi: https://doi.org/10.1016/j.molstruc.2019.06.028.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. 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.

ACCEPTED MANUSCRIPT
Synthesis and experimental versus theoretical research on spectroscopic and electronic
properties of 3-methyl-4-nitroisothiazole
1*
2
Andrzej Regiec and Piotr Wojciechowski
1
Department of Organic Chemistry, Faculty of Pharmacy, Wrocław Medical University,
11A Borowska Street, 50-556 Wrocław, Poland
2
2
Wroclaw University of Science and Technology, Faculty of Chemistry, 27 Wybrzeże Wyspiańskiego
Street, 50-370 Wrocław, Poland
*Corresponding author, e-mail: Andrzej Regiec (andrzej.regiec@umed.wroc.pl)
1

ACCEPTED MANUSCRIPT
Abstract
The alternative ways of synthesis of valuable 5-amino-3-methylisothiazole and 3-methyl-4-
nitroisothiazole has been presented and described. It appeared that new, undescribed 3,3’-
dimethyl-4,4’-dinitro-5,5’-bisisothiazole has been produced as a side-product during
synthesis. Both experimental and calculated spectral and electronic properties of 3-methyl-4-
nitroisothiazole have been extensively researched and discussed. The complete vibrational
assignments were made on the basis of potential energy distributions PED. The fully
anharmonic infrared and Raman spectra with calculated anharmonic intensities for
fundamental bands, overtones as well as combination bands are presented. Vibrational spectra
predicted by anharmonic approximation agree very well with experimental data. The stability
of molecules of 3-methyl-4-nitroisothiazole arising from hyper conjugative interaction has
been estimated using natural bond orbital (NBO 6.0) analysis. The combined frontier
molecular orbital and NBO analysis of canonical molecular orbitals was used to predict the
most probable sites of 3-methyl-4-nitroisothiazole molecule where single electron can be
accepted to and detached from. Measured reduction potential of 3-methyl-4-nitroisothiazole
and calculated electron affinity point out that this compound is susceptible to reduction at the
similar level of 1-methyl-3-nitropyrazole and this susceptibility is much more than respective
1-methyl-4-nitropyrazole. UV-Vis spectra analysis reveals a nature of valence electron
excitation and electron transition of 3-methyl-4-nitroisothiazole. Besides, unambiguous
assignment of NMR signal shifts of carbon atoms of isothiazole ring of 3-methyl-4-
nitroisothiazole, 5-amino-3-methylisothiazole and intermediate and side products of synthesis
1
13
was conducted thanks to full detailed analysis of H, C NMR spectra and their two
dimensional (2D) variants. Additionally, molar enthalpy of vaporization (ꢀHvap) of 3-methyl-
4-nitroisothiazole has been estimated (ꢀHvap = 52.2 kJ/mol).
Keywords: 3-Methyl-4-nitroisothiazole; 5-Amino-3-methylisothiazol; Synthesis; NMR-
Vibrational-UV-Vis spectral analyses; Reduction potential; NBO analysis
2

ACCEPTED MANUSCRIPT
1. 1.Introduction
N
S
Cl
Isothiazole (1,2-thiazole)
O
N
H
N
O
N
S
NH S
NH2
N
S
NH
N
H
S
O
Brassilexin
Sulphasomizole
O
Vratizolin (Denotivir)
O
O
Cl
NH
N
O
N
N
N
N
N
N
S
Perospirone
S
Ziprasidone
Fig. 1. Isothiazole and their biologically active derivatives.
In general, isothiazole, i.e. 1,2-thiazole (see Fig. 1), is very interesting heterocycle
aromatic system, known from 1956, when Slack and Adams have described its synthesis [1].
Simple monocycle isothiazole derivatives are unknown in the nature up to date but in 1988
Davys et al. [2, 3] informed that they had isolated from plant Brassica juncea (L.) (Indian
mustard, brown mustard) - tricyclic compound called brassilexin (see Fig. 1), which includes
isothiazole ring condensed with indole moiety. Brassilexin has appeared to be a phytoalexin
having antimicrobial activity, especially antifungal one. Many compounds, which have
isothiazole ring in their structure, possess strong biological activity such as mentioned
antimicrobial [4, 5], antiviral [6, 7], antiinflammatory [8], immunosuppressive [9, 10],
anticancer [11, 12], antipsychotic [13, 14] one. Furthermore, a few of them are drugs such as
3

ACCEPTED MANUSCRIPT
antibacterial Sulphasomisole (see Fig. 1) [4], antiviral and antiinflammatory Denotivir
(Vratizolin) (Fig.1) [6, 8, 15], antipsychotic Ziprasidone (Fig.1) [13] and Perospirone
(Fig.1) [14] which were or are used in medical treatment.
This work is continuation of our investigation of spectroscopic and electrochemical
properties of nitroazoles with application of computational methods [16, 17] to evaluate their
usefulness to designing of potentially bioreductive drugs, particularly effective in anticancer
therapy. Especially, nitropirazoles and nitroisoxazoles have appeared to be very useful ligands
to prepare valuable platinum complexes with significant anticancer activity [18, 19]. This
time, our research has focused on 3-methyl-4-nitroisothiazole (1), (see Fig.2), which is known
compound [1, 20] but little studied and obscurely described so far, what may be caused by its
poor availability and laborious synthesis.
O
N
H C
O
3
N
S
1
Fig. 2. Structure of 3-methyl-4-nitroisothiazole (
. Experimental details
The melting (m.p.) and boiling (b.p.) points of the compounds were measured on the
1)
2
Büchi M560 melting point apparatuses (BÜCHI Labortechnik AG, CH-9230 Flawil/SG,
Switzerland) and were uncorrected. Elemental analysis made by Microlaboratory of the
Department of Pharmacy, Wroclaw Medical University with Carlo Erba series NA 1500
elemental analyzer (Carlo Erba, Milan, Italy). Electrospray Mass Ionization (ESI) spectra
were recorded with EVOQ™ Triple Quadrupole Mass Spectrometer (Bruker Daltonics) in the
2
following solvents: water (H O), methanol (MeOH), acetonitrile (AcCN) or in their mixtures.
Thin-layer chromatography method (TLC) was applied to monitor the reaction course as well
as to confirm the purity of the synthesized compound. TLC-PET foils with fluorescent
indicator 254 nm, silica gel matrix plates (Fluka) for TLC were used, eluting medium was
chloroform-acetone (9:1 vol. ratio) or chloroform-methanol (9:1 vol ratio), detection of the
compounds on the chromatograms was done with UV light and/or by treatment with iodine
vapors.
2.1. IR and Raman spectra measurements.
4

ACCEPTED MANUSCRIPT
-1
The Attenuated Total Reflectance IR (ATR-IR) spectra (4000-450 cm ) of 3-methyl-
-nitroisothiazole was measured with the Nicolet iS50 FT-IR spectrophotometer (Thermo
4
Fisher Scientific Inc.) using clean liquid form of compound (on the diamond crystal surface,
-1
3
2 scans, resolution: 1 cm , measurement temperature: 23 °C). IR Spectra were recorded
without and with ATR intensity correction. It is known, that the relative intensity of bands in
an ATR spectrum increases with wavelength causing the distortion of relative peak intensities
in comparison with the classical transmission experiment [21] Omnic Specta software
(Thermo Fisher Scientific Inc.) was used for IR spectra analysis. Baselines of all IR spectra
were corrected with autocorrection (fit order 1). The FT-Raman spectrum of liquid compound
–
1
–1
(
1) was recorded in the 3600 to 50 cm range with a 4 cm spectral resolution on the Bruker
MultiRAM FT–Raman spectrometer equipped with a Nd:YAG laser (emitting at 1064 nm)
and a germanium detector cooled with liquid nitrogen. Spectrum of sample was collected with
laser power of 500 mW and 512 scans. Then, correction of baseline and normalization of
Raman spectrum was conducted.
2.2. NMR spectra measurements.
1
13
H-NMR (300.15 MHz), C-NMR (75.47 MHz, broadband full decoupling method),
1
1
1
H- H 2D COSY (Correlation Spectroscopy) NMR (both channels F2 and F1: 300.15 MHz),
13
1
13
H- C 2D COSY NMR (channel F2: 300.15 MHz, channel F1: 75.47 MHz), 2D H- C
HSQC (Heteronuclear Single Quantum Coherence=Heteronuclear Single Quantum
1
13
Correlation) and 2D H- C HMBC (Heteronuclear Multiple Bond Correlation) spectra were
recorded using 5mm tubes and concentration about 20 mg of tested compounds in 0.6 mL of
deuterated chloroform (CDCl
3 6
) and/or dimethyl sulfoxide DMSO-d with Bruker Fourier
spectrometer (Bruker Analytische Messtechnik GmbH, Rheinstetten, Germany).
2
.3. Potential of reduction measurement.
The potential of reduction of 3-methyl-4-nitroisothiazole was established using digital
puls strippinig voltammetry method with PA3 Analyzer (Laboratorni P istroje, Praha 1983).
ř
Hanging mercury drop electrode (HMDE) was used as working electrode and saturated
calomel electrode (SCE) as an internal reference. The measurements were carried out at 22˚C
in 0.1 M NaCl aqueous solution saturated with argon. The final result (-0.7350 V) was
corrected to standard hydrogen electrode (SHE) giving value -0.3994 V. Potential of calomel
electrode versus SHE taken to correction is +0.3356V (25 °C, 0.1M KCl) [22].
5

ACCEPTED MANUSCRIPT
2.4. UV-Vis spectra measurements.
UV-Vis spectra of 3-methyl-4-nitroisothiazole were recorded with double beam U-
2
800 Spectrophotometer (Hitachi - Science & Technology) in water and n-hexane solutions in
-4
-5
two concentrations, i.e., c=1.00ꢁ10 M and 5.00ꢁ10 M. Quartz cuvettes (path length =10.0
mm) were used. The measurement was conducted three or four times for each concentration.
Preliminary measurement range was set at 190-1000 nm (sampling interval 1 nm, scans speed
4
2
00 nm/minute) and final range was set at 190-400 nm (sampling interval 0.5 nm, scans speed
00 nm/minute). The base lines were corrected with pure water and n-hexane reference
solvents, respectively. Estimation of molar attenuation coefficients (formerly molar extinction
coefficient) were estimated on the basis of Beer-Lambert law and its statistics was conducted
with DellStatistica ver.13.1 software.
2
.1. Synthesis of 3-methyl-4-nitroisothiazole (1)
O
O
N
O
OH
NaNO2
H SO
HNO3
2
4 (conc.)
N
CuBr
N
N
N
∆T
Br
NH2
Br
NH2
S
oleum 20%
S
S
S
2
3
4
5
CuCN
O
N
O
O
O
O
N
N
O
N
O
O
NaNO2
H SO
∆
T
S
OH
N
N
N
N
N
pyrolysis
C
S
S
1
S
2
4 (conc.)
S
N
N
O
+ O
9
6
O
N
N
8
O
O
NaNO2
H SO
NH2
2
4 (conc.)
S
O
7
Fig. 3. Scheme of synthesis 3-methyl-4-nitroisothiazole (1)
.
3
-Methyl-4-nitroisothiazole
feasible and efficient decarboxylation process of 3-methyl-4-nitro-5-isothiazolecarboxylic
acid ( ) during its dry pyrolysis at 120-130 °C (Fig. 3). A starting material of total synthesis
was known 5-amino-3-methyl-4-isothiazolecarboxylic acid [23] or its -benzoyl derivatives
8] which were transformed into known 5-amino-3-methylisothiazole [1] in efficient way by
decarboxylation with hot concentrated sulphuric acid. In the case of -benzoyl derivatives [8]
1, used in experiments, was obtained as a result of easily
9
2
N
[
3
N
deacylation and decarboxylation processes occur simultaneously during this reaction. Then,
6

ACCEPTED MANUSCRIPT
amine
3, in Sandmayer reaction type, gave known 5-bromo-3-methylisothiazole 4 [1] which
was undergone nitration process resulting in known 5-bromo-3-methyl-4-nitroisothiazole (
. It is worthy to note, that the use of 20% oleum instead of concentration sulfuric acid in
nitrating mixture gives better results. In turn, the compound was transformed into known 3-
methyl-4-nitro-5-isothiazolecarbonitrile [24] in reaction with cuprous cyanide (CuCN) in
5)
[1]
5
6
heptane. It appeared that, apart the main product
described 3-methyl-4-nitro-5-isothiazolecarboxamide (
6
, side-products, namely, previously
7
) [24] and unknown up to date - 3,3’-
dimethyl-4,4’-dinitro-5,5’-bisisothiazole (
mixture during purification process. Creation of 3,3’-dimethyl-4,4’-dinitro-5,5’-bisisothiazole
) can point out that the reaction has radical nature. However, Small amounts of 3-methyl-4-
nitro-5-isothiazolecarboxamide ( ) can probably arise during purification on silica-gel
through hydrolysis of 3-methyl-4-nitro-5-isothiazolecarbonitrile . Then, nitrile has been
hydrolysed into 3-methyl-4-nitro-5-isothiazolecarboxylic acid [24 with 95% H SO in the
presence of sodium nitrite. Acid can be also successfully obtained through denitrogenation
of 3-methyl-4-nitro-5-isothiazolecarboxamide ( ) under influence of nitrous acid generated in
situ from nitrites, where nitrite agent fulfills crucial role.
8) have been also separated from final reaction
(8
7
6
6
9
]
2
4
9
7
2.1.1. Preparation and experimental properties of 5-amino-3-methylisothiazole
(3), Mmol
=
114.168 g/mol
The mixture of 10 g (63.2 mmol) of 5-amino-3-methyl-4-isothiazolecarboxylic acid
2 [23]
and 21 ml of fresh concentrated sulphuric acid (98%) was heated at 100-110 °C on oil bath
until carbon dioxide stopped releasing (about 5 h). Then, a cooled mixture was poured on an
ice and alkalyzed with a cooled aqueous solution of sodium hydroxide. Solid contaminations
were filtered off and the filtrate was extracted with chloroform (100 ml, 3 times). Collected
chloroform solution of 5-amino-3-methylisothiazol was dried over anhydrous magnesium
sulphate (MgSO
solidified giving 6.2 g (86%) of final crude red colored product
recrystallized from n-hexane what resulted in 5.36 g (74%) of light yellow 5-amino-3-
methylisothiazole was obtained, m.p. = 49-51 °C (ref. 51-53 °C [1]). Compound can be also
obtained using 5-benzoylamino-3-methyl-4-isothiazolecarboxylic acids [8] instead of 5-
amino-3-methyl-4-isothiazolecarboxylic acid as a starting material but in this case, the
higher temperature (130-150 C) and longer time (about 15 h) is required in order to conduct
4
), filtered off and condensed to dryness. After cooling, an oily residue
1
(m.p.=42-44 °C), which was
1
2
°
the decarboxylation with releasing of benzoic acids as side products. Free 5-amino-3-
7

ACCEPTED MANUSCRIPT
methylisothiazole should be stored in darkness at 4 °C in oxygen free container because it is
sensitive to oxidizing and can slowly decompose. For long time storage 5-amino-3-
methylisothiazole should be transformed into its salt (e.g. hydrochloride)
4 6 2
Elemental analysis for formula C H N S (Mmol = 114.168 g/mol).
Calculated (%): C 42.08, H 5.30, N 24.54.
Found (%): C 41.74, H 5.68, N 24.14.
Analysis of Mass Spectrum (ESI-MS) of 3: The calculated value of the parent peak mass for
the formula C
4
H
6
N
2
2
S is 114.0251 u. Positive Electrospray Ionization in H O, m/z [u/e]
+
(
relative signal intensity): 117.0283 (4.8%) [M+2+H] - second isotope quasi-molecular ion
+
peak, 116.0346 (6.2%) [M+1+H] - first isotope quasi-molecular ion peak, 115.0324
+
(
100.0%) [M+H] - quasi-molecular ion and base peak.
1
H-NMR of 5-amino-3-methylisothiazole in deuterated chloroform: (CDCl , 300.15 MHz,
3
T=296.7 K):
3 2
δ 2.28 (s; 3H, -CH ), 4.44 (s, 2H, low and broad, NH ) 6.10 (s; 1H, proton at
position 4 of isothiazole ring) ppm.
1
3
C-NMR of 5-amino-3-methylisothiazole in deuterated chloroform.
T= 296.5 K): 19.4 (- , the carbon of methyl group), 106.2 (C4), 167.1 (C3, quaternary
carbon), 171.8 (C5, quaternary carbon) ppm.
3
: (CDCl , 75.47 MHz,
δ
CH
3
1
1
H- H 2D COSY NMR of 3-methyl-5-aminoisothiazole (CDCl3, 300.15 MHz, T=296.7 K)
shows weak coupling between protons of methyl group (2.28 ppm) and proton at position 4 of
isothiazole ring (6.10 ppm).
1
13
H- C 2D COSY NMR of 3-methyl-5-aminoisothiazole (CDCl3, channel 2: 300.15 MHz,
channel 1: 75.47 MHz, T=301.3 K) shows strong coupling between protons of methyl group
2.28 ppm) and methyl carbon (19.4 ppm) and strong coupling between proton at position 4 of
isothiazole ring (6.10 ppm) and C4 carbon (106.2 ppm).
(
1
13
H- C 2D HMBC NMR (CDCl channel 2: 300.15 MHz, channel 1: 75.47 MHz, T=296.4
3
,
K) shows strong long range coupling through two bonds between methyl protons (2.28 ppm)
and quaternary carbon C3 (167.1 ppm), little less intensive signal that points the coupling
through three bonds between methyl protons (2.28 ppm) and quaternary carbon C4 (106.2
ppm). Also, strong signals of long range coupling through two bonds between proton at
position 4 of isothiazole ring (6.10 ppm) and quaternary carbons C3 (167.1 ppm) and C4
(
106.2 ppm) appear. Besides, the least intensive signal responsible for the coupling through
three bonds between proton at position 4 of isothiazole ring (6.10 ppm) and methyl carbon
19.4 ppm) is also visible.
(
8

ACCEPTED MANUSCRIPT
2.1.2. Preparation and experimental properties of 3-methyl-4-nitroisothiazole (1), Mmol
=144.151 g/mol
4
.00 g (21.3 mmol) of dry 3-methyl-4-nitro-5-isothiazolecarboxylic acid
9 was solventlessly
heated at 120-130 °C in a round bottom flask under a condenser until carbon dioxide stopped
releasing (aobut 15 min.). Brown liquid crude product (2.52 g, 82%), which solidifies during
cooling below 20 °C, was purified by distillation under reduced pressure. It was distilled at
b.p. 83-84 °C (at 5 mmHg = 6.67 mba (reference, b.p. 108 °C at 14 mmHg = 18.67 mba [1]))
giving 1.98 g (65%) of pale yellowish liquide of pure 3-methyl-4-nitroisothiazole
solidified into colorless needles after cooling in a refrigerator. Melting point (m.p.) and
boiling point (b.p.) of 3-methyl-4-nitroisothiazole , measured by us, are 18.5-20 °C and 225-
27 °C (at 1019 mba (764.32 mmHg) of pressure) respectively. Hence, average molar
1, which
1
2
enthalpy of vaporization (ꢀH ) of 3-methyl-4-nitroisothiazole can be roughly estimated on
vap
the basis of Clausius-Clapeyron equation for temperature range 83-227 °C (356.15 - 500.15
K), 52.2 kJ/mol (assuming that it is constant at this temperature range).
Analysis of Mass Spectrum (ESI-MS): The calculated value of the parent peak mass for the
vap
ꢀH ≈
formula C
4
H
4
N
2
O
2
2
S is 143.999353 u. Positive Electrospray Ionization in H O, m/z [u/e]
+
(
relative signal intensity): 147.0048 (4.90%) [M+2+H] - second isotope quasi-molecular ion
+
peak, 146.0094 (6.06%) [M+1+H] - first isotope quasi-molecular ion peak, 145.0075 (100%)
+
[
M+H] - quasi-molecular ion peak.
1
3
H-NMR of 3-methyl-4-nitroisothiazole in deuterated chloroform: (CDCl , 300.15 MHz,
T=300.0 K):
is necessary to point that shifts assignments of protons for 3-methyl-4-nitroisothiazole in
CCl , which were done by Staab et al., are 2.74 and 9.54 ppm, respectively after recalculation
from old scale [25].
C-NMR of 3-methyl-4-nitroisothiazole in deuterated chloroform
02.0 K): 19.83 (- H ), 145.44 (C4, nitro substituted quaternary carbon, very low intensity
signal due to high relaxation time), 151.35 (C5), 162.16 (C3, quaternary carbon) ppm.
3
δ 2.78 (s; 3H, -CH ), 9.51 (s; 1H, proton at position 5 of isothiazole ring) ppm. It
4
τ
1
3
:
3
(CDCl , 75.47 MHz, T=
3
δ
C
3
1
13
H- C 2D HSQC NMR (CDCl Bruker, channel 2: 300.15 MHz, channel 1: 75.47 MHz,
3
,
T=299.0 K) shows strong coupling between protons of methyl group (2.76 ppm) and methyl
carbon (19.8 ppm) and strong coupling between proton at position 5 of isothiazole ring (9.51
ppm) and C5 ring carbon (151.3 ppm).
1
13
H- C 2D HMBC NMR (CDCl3, channel 2: 300.15 MHz, channel 1: 75.47 MHz, T=300.1
K) shows strong long range coupling through two bonds between methyl protons (2.76 ppm)
and quaternary carbon C3 (162.2 ppm), less intensive signal that points the coupling through
9

ACCEPTED MANUSCRIPT
three bonds between methyl protons (2.76 ppm) and quaternary carbon C4 (145.4 ppm) and
finally, the least intensive signal responsible for the coupling through four bonds between
methyl protons (2.76 ppm) and C5 carbon (151.3 ppm). Besides, strong long range coupling
3
J
H5,C3 through 3 bonds between proton H5 (9.51 ppm) and carbon C3 (162.16 ppm) appears
2
due to favorable dihedral angle (180 °) between these atoms. Signal of coupling
J
H5,C4
between H5 (9.51 ppm) and C4 (145.44 ppm) is also visible but it has much less intensity.
2
.1.3. Separation, purification, some physical and spectral data of 3,3’-dimethyl-4,4’-
dinitro-5,5’-bisisothiazole (8), Mmol =286.286 g/mol
,3’-dimethyl-4,4’-dinitro-5,5’-bisisothiazole ( ) is created in small quantities (variously, up
to 10% of total mass of product mixture) as one of side products during reaction of 5-bromo-
-methyl-4-nitroisothiazole ( ) with cuprous cyanide in heptane in the presence of
dimethylformamide (DMF). It was separated and purified from compounds and with
3
8
3
5
6
7
chromatographic column packed with silica gel 60 (0.04-0063 mm, 230-400 mesh ASTM,
Merck KGaA). Gradient elution was used starting form chloroform, then mixture of
chloroform-methanol (9:1 vol) and finally chloroform-methanol (8:2 vol). The nitrile
eluted as a first one, then 3,3’-dimethyl-4,4’-dinitro-5,5’-bisisothiazole ( ) and finally, the
amide The compound forms white fine crystals, m.p. = 223-224 °C and is soluble in
6 has
8
7
8
chloroform and DMSO but insoluble in water.
Analysis of Mass Spectrum (ESI-MS) of 8: The calculated value of the parent peak mass for
the formula C
8
H
6
N
4
O
4
S
2
is 285.983056 u. Positive Electrospray Ionization in acetonitrile, m/z
+
[
u/e] (relative signal intensity): 288.9847 (10.5%) [M+2+H] - second isotope quasi-
+
molecular ion peak, 287.9903 (12.4%) [M+1+H] - first isotope quasi-molecular ion peak,
+
2
86.9881 (100.0%) [M+H] - quasi-molecular ion and base peak.
1
H-NMR of 3,3’-dimethyl-4,4’-dinitro-5,5’-bisisothiazole (8) in deuterated DMSO
:
(DMSO-d
6
, 300.15 MHz, T=300.0 K): 2.73 (s; 3H, -C
δ
3
H ) ppm. Spectrum data in deuterated
chloroform (CDCl
3
): 2.83 (s; 3H, -CH ) ppm
δ
3
1
3
C-NMR of 3,3’-dimethyl-4,4’-dinitro-5,5’-bisisothiazole (8) in deuterated DMSO
(
DMSO-d ): 20.0 (- ), 141.8 (C4, nitro substituted quaternary carbon, relatively low
6
δ
CH
3
intensity signal due to high relaxation time), 153.5 (C5, quaternanry carbon), 162.6 (C3,
quaternary carbon) ppm. Spectrum data in deuterated chloroform (CDCl , Bruker, 75.47
MHz, T= 300.0 K): 20.6 (- ), 142.0 (C4, quaternary carbon, very low intensity signal
due to high relaxation time), 152.6 (C5), 164.1 (C3, quaternary carbon) ppm.
:
3
δ
CH
3
1
0

ACCEPTED MANUSCRIPT
1
13
H- C 2D HSQC NMR (DMSO-d6, channel 2: 300.15 MHz, channel 1: 75.47 MHz, T=
98.7 K) shows strong coupling between protons of methyl group (2.73 ppm) and methyl
carbon (20.0 ppm).
2
1
13
H- C 2D HMBC NMR (DMSO-d
T=297.6 K) shows strong long range coupling through two bonds between methyl protons
2.76 ppm) and quaternary carbon C3 (162.6 ppm), less intensive signal that points the
coupling through three bonds between methyl protons (2.76 ppm) and quaternary carbon C4
141.8 ppm) and finally, the least intensive signal responsible for the coupling through four
bonds between methyl protons (2.76 ppm) and C5 carbon (153.5 ppm).
6
, channel 2: 300.15 MHz, channel 1: 75.47 MHz,
(
(
2.1.4. Preparation, spectral data and some properties of the other intermediate and side-
products 4,5,6,7,9 of the synthesis have been described in detail in supplementary material.
3. Computational details
For compatibility with our previous calculations for 1-methyl-4-nitropyrazole [16], 1-
methyl-3-nitropyrazole and 1-methyl-5-nitropyrazole [17], a full geometry optimization has
been performed using the B3LYP [27, 28] hybrid density functionals with the extended 6-
3
11++G(df,pd) basis set and a molecule of 3-methyl-4-nitroisothiazole was treated as isolated
in vacuum state. As it is shown in the supplementary materials, the 6-311++G(df,pd) basis set
proved to be the best in the case of harmonic spectrum simulations (also better than a larger 6-
3
11++G(3df,3pd) one). Moreover, based on crystallographic data for 4-hydroxymethyl-
isothiazole-3-carboxylic acid [29], we confirmed the correctness of usage this theory level for
optimizing the structures of aromatic isothiazole ring (see Table S1 in supplementary
material). Subsequently, harmonic and anharmonic vibrational frequencies and natural
population analysis were computed using the Gaussian 2016 revision A.03 software at the
same level [30]. To possibly get most accurate approximation the keywords, i.e.
Fopt=(VeryTight,CalcAll), SCF=(Direct,VeryTight) and Integral(Grid=SuperFineGrid) were
used in calculation. The ‘Stable’ keyword was used to test the stability of DFT wavefunctions.
The absence of imaginary wavenumbers on the calculated vibrational spectrum confirms that
the final structure and presented results correspond to a minimum on the potential energy
surface. The fully anharmonic IR spectrum with anharmonic frequencies and anharmonic
intensities for all fundamental bands as well as for overtones and combination bands were
computed with the DVPT2 approach [31-35]. The Natural Bond Orbital (NBO) analysis with
1
1

ACCEPTED MANUSCRIPT
natural resonance analysis was performed and atomic charges were calculated by using the
GENNBO 6.0 stand alone program [36]. Moreover, the NBO analysis and potential energy
surface scan studies have been carried out to explain the coplanarity of isothiasole ring and
nitro group.
The non redundant set of 33 internal coordinates was defined and the theoretical
normal modes were analyzed on the basis of the procedure described in refs. [37] and [38].
The matrices of force constants were transformed from Cartesian coordinates to the internal
coordinates and potential energy distribution (PED) matrices were calculated using the
equation given in refs. [37] and [38]. For comparative purposes with previous research on
nitropyrazoles [16, 17], four harmonic frequencies, which correspond to the CH stretching
vibrations, were scaled by a factor of 0.958, while the remaining twenty-nine harmonic
-1
frequencies, below 2000 cm , were scaled by 0.983. The harmonic frequencies of
combination bands are also scaling with respect to these factors. These scaling factors were
successfully used in the previous theoretical studies and well correlated with data obtained by
anharmonic approximation [16, 17, 39, 40]. The Chemcraft program [41] was used to
illustrate the computed results and list of frequencies in which the S1 and H13 atoms are
mostly involved, whereas VMS-Draw tool [42] was used to visualize the spectroscopic
properties of 3-methyl-4-nitroisothiazole (
1). Because Raman scattering activities cannot be
observed explicitly, they must be converted to experimentally observable Raman cross-
R
4
-1
section. Hence, the calculated Raman scattering activities A (Å
ꢁamu ) produced by
Gaussian software have been converted to simulated molar Raman cross-section per solid
csR
2
-1
-1
angle unit,
I
(cm ꢁsr ꢁmol ) using following eq 1, which is based on Placzek’s
polarizability theory [43-45]:
2
~
~
4
R
2
π ⋅ h
(
ν − ν
)
Ai
csR
Ii
−9
0
i
=
6.022140857
⋅
10 ⋅ N ⋅
⋅
⋅
⋅TCoeff
(1)
A
~
c
ν
45
i
~
-1
th
where
~
v
is the calculated harmonic spectroscopic wavenumber (cm ) of the
i
normal mode,
i
-1
R
4
-1
v
is the excitation wavenumber of the laser line (cm ) and Ai (Å
ꢁ
amu ) is the calculated
0
th
Raman scattering activity of
i
mode. It is necessary to remember that a.m.u.
-24
(
1.660539040(20)·10 g) is mass unit that relates to one molecule and can be expressed as
2
3
1
g/NAn, where NAn is dimensionless Avogadro number (6.022140857(74)·10 ) what finally
-24
results in the value of a.m.u=1.660539040(20)·10 g. Because a value of Raman activity,
1
2

ACCEPTED MANUSCRIPT
which is calculated with Gaussian, relates to one unit of mass of one molecule of the
considered compound, it must be converted into Raman activity expressed by mole amount of
2
3
molecules of the considered substance. Hence, Avogadro constant
A
N (6.022140857(74)·10
-1
-27
mol ) is appeared in the eq 1. The symbol
h
means Planck’s constant (6.626070040(81)·10
2
-2
10
-1
erg
ꢁ
s, where erg=g
ꢁ
cm
ꢁ
s ) and
c
is speed of light in vacuum (2.99792458ꢁ10 cmꢁs ). A
-9
value 6.022140857·10 is conversion factor to transform the Gaussian Raman scattering
4
-1
4
-1
activity unit, i.e. Å
ꢁ
amu into cm
ꢁ
g
unit. The Tcoeff in eq 1 is Boltzman distribution
~
expressed with 1/(1-exp((-h·c·
v
)/(k·T))), which includes the effect of temperature on
0
-34
Raman intensity. In which, Planck’s constant - h (6.6260·10
Jꢁ
s), speed of light - c
2
-1
-23
-1
(
299792458
ꢁ
10 cm
ꢁ
s ), Boltzman’s constant - k (1.3806·10
Jꢁ
K ) are fundamental
constants and T is temperature expressed in Kelvin degree. Calculations were performed for
the temperature T = 298.15 K which corresponds to the temperature of measurement.
Besides, Gaussian software can also perform anharmonic calculation producing Raman
R
6
activity A (Å ) with respect to the dimensionless normal coordinates. In this case, Raman
R
6
csR
2
-
activities A (Å ) should be converted to simulated molar Raman cross-section, I (cm
ꢁ
sr
1
-1
ꢁ
mol ) using eq 2, which is implemented in VMS-Draw Tool [42].
−
48
R
N ⋅10
⋅ Ai
CsR
A
~
~
4
Ii
=
(
2π⋅
(
ν − ν ))
(2)
0
i
4
5
2
3
−1
~
-1
where
of the
N
A
is Avogadro constant (6.0221·10 mol ),
~
v
is the calculated wavenumber (cm )
i
th
-1
i
normal mode,
v
is the excitation wavenumber of the laser line (cm ) [46].
0
Additionally, it is necessary to clearly state that output file from Gaussian computation can
contain two tables with calculated Raman activities. The table connected with calculated
harmonic frequencies in Gaussian output file contains Raman scattering activities computed
4
with respect to mass-weighted normal coordinates given in Å /amu (which should be used in
(1) eq.), whereas the part of Gaussian output, which is described as ‘Anharmonic Raman
6
Spectroscopy’, contains the Raman activities given in Å , which were calculated with respect
to the dimensionless normal coordinates and have been converted to Raman intensities
according to eq 2. It is worthy of emphasizing that both units of Raman scattering activity can
6
16
be easily converted to each other, using following expression, Å = 10
ꢁ
(h/(c
amu ). Besides, it is necessary to take note that activity expressed in Å cannot be
directly used to calculate Raman intensity for scaled harmonic wavenumber because such
·
v~
i
2
4
-1
6
·8π ))ꢁ(Å ꢁ
1
3

ACCEPTED MANUSCRIPT
value of activity covers implicite own wavenumber attributed. Moreover, for alternative
simulation of Raman spectrum, both Freq=ROA and CPHF=RDFREQ keywords were used to
compute the analytic Raman optical activity intensities. For all of the above simulations of
~
-1
Raman spectra, the incident light wavenumber
v
was 9398.5 cm (frequency
s × 9398.5 cm = 281.76 THz), which corresponded to a wavelength of
Nd:YAG excitation laser.
However, more useful for comparative purposes is dimensionless relative Raman
0
v =
0
-1
-1
2
9979245800 cmꢁ
csR
Rel
cross-section (
I
) calculated as a percentage with following eq 3
csR
Ii
csR
IRel
=
⋅100%
(3)
csR
I
max
csR
th
,
where Ii is the calculated (with eq 1) Raman cross-section for the harmonic
i
normal
csR
mode and
I
is the highest calculated value of Raman cross-section of whole considered
max
range of spectrum.
Electron affinities have been calculated as difference of negative value of energy (-
between the energy of anionradical of considered compound and its electric inert molecule
according to formula E = - E = -(E(anionradical)-E(inert molecule)) = E(inert molecule)-
∆
E)
A
∆
E(anionradical), where E(inert molecule) is an energy of a molecule before electron
acceptance and E(anionradical) is an energy of a molecule after electron acceptance.
Following calculation parameters were used to energetical optimization of molecules of tested
compounds, namely Fopt=(VeryTight, CalcAll), SCF=VeryTight, Integral=SuperFineGrid.
The Conductor-like Polarizable Continuum Model (CPCM) [47] was used for simulation of
aqueous environment with Gaussian 16 default set of prarameters (Scaling factor for van der
Waals radius for all atoms, ie., alpha=1.1).
Excited states were computed with TD (time dependent) DFT [48] method with
B3LYP functional and different functional basis sets using CPCM [47] for simulation of
solvent environment (see details in chapter 4.3 and also in supplementary material, page
….).
Gaussian 16 (Revision A.03) default sets of CPCM parameters were used [30_G16
Reference]. All excited states were calculated with TD method for singlet excited states
(Singlets option) and 15 states (NStates=15). On the basis of the results of excited states
1
4

ACCEPTED MANUSCRIPT
computations, UV-VIS spectra were simulated using following eq 4 according to following
reference [49]:
ꢌ
ꢅ
ꢊꢂ - ꢊꢂ ꢆ
ꢄ
ꢆ
ε ꢁ νꢂ ꢃ = 1.3062974 ∙ 10 ∙ ∙ exp ꢈ- ꢉ ꢋ ꢍ
(4)
ꢀ
ꢇ
ꢇ
-1
-1
~
i
where
ε
i
( νꢂ ) is molar attenuation coefficient in L
·mol
·
cm ,
v
is the excitation energy (in
-1
wavenumbers expressed in nm ) corresponding to the electronic excitation of interest that
-
corresponds to appropriate oscillator strength f (dimensionless value), a value value νꢂ (in nm
i
1
)
represents any single wavenumber of whole searched spectral range of interest, σ is the
-1
standard deviation in wavenumbers expressed in cm , which is related to the width of the
simulated band (an appropriate value for 0.4 eV was set).
Isotropic chemical shieldings of NMR were calculated using GIAO (Gauge-
Independent Atomic Orbital) method [50] with following combinations functional/basis set:
B3LYP/6-311+G(d,p), B3LYP/6-311++G(df,pd) and B3LYP/aug-cc-pVTZ with application
of CPCM [47] with default prarameters for simulation of solvent environment. Following
parameters were used nmr(All, PrintEigenvectors, mixed, Susceptibility) (see details in
chapter 4.4).
In supplementary materials we have included detailed information related to the
indication of the best methods for spectra simulation of 3-metyl-4-nitroisothiazole. Our
general recommendation is presented in Table 1.
Table 1. Recommended levels of theory.
Simulated spectrum
Method / basis set
Short comment:
Anharmonic spectrum B3LYP/(d)aug-cc-pVTZ The B3LYP/aug-cc-pVTZ computation
(IR/Raman)
give similar deviation to B3LYP/daug-
cc-pVTZ one, but it is more faster.
Harmonic spectrum B3LYP/6-311++G(df,pd) This method we used in previous
(IR/Raman)
computations.
UV-VIS spectrum
B3LYP/aug-cc-pVTZ
B3LYP/6-31+G(d,p
Same accuracy as in B3LYP/daug-cc-
pVTZ computation, but much faster.
Larger basis sets give worse results
1
H and 13C-NMR
)
4
4
. Results and discussion
.1. Geometry and NBO analysis
1
5

ACCEPTED MANUSCRIPT
Fig. 4. Atom numbering, calculated bond lengths and ring valence angles of 3-methyl-4-
nitroisothiazole (1).
The full optimized molecular structure with atom numbering and some geometrical
parameters is presented in Figure 4. Five predominant resonance structures of 3-methyl-4-
nitroisothiazole (1) are shown in the Figure 5. Additionally, the detailed geometrical
parameters are collected in Table S3 in supplementary. The isothiazole ring is considerably
distorted from a regular pentagon and it is coplanar with nitro group. Only H10 and H11
atoms of tetrahedral methyl group are beyond the plane of the molecule. These both hydrogen
atoms are respectively symmetrically set above and below the plane of symmetry what results
in reflection symmetry (C
S
point group) of whole molecule. The Natural Bond Critical Point
(NBCP) analysis finds 14 characteristic critical points of (3,-1) saddle-type topology [51].
Thirteen of them correspond to thirteen σ bonds in a molecule. The last (3,-1) bond critical
point lies on the straight line connecting the C6 and O8 atoms. This critical point is defined by
2
electron density ρ = 0.0118; Laplacian density ∇ ρ = 0.0481 and the three Hessian
eigenvalues: 0.0578, -0.0046, -0.0050, respectively. Main NBO-based contributions to this
bond critical point are O8 lone pair (34.3%), σC6-H10 bond (15.9%) and σC6-H11 bond (15.9%).
The presence of this (3, -1) critical point in electron density topology confirms the impact of
long-distance interaction which stabilizes the “gauche conformation” of methyl groups and
nitro. This interaction is one of the main reasons, why the rotation barrier of methyl group in
3
-methyl-4-nitroisothiazole (1) (1.67 kcal/mol) is four times higher than in 3-
methylisothiazole, which is (0.41 kcal/mol). On the other hand, natural bond-bond
polarizability (NBBP) analysis shows that this effect is relatively weak. The bond-bond
polarizability between LP2(O8) and C3-C4 bonds is 0.0009, whereas between LP2(O8) and
gauche C6-H10 bond as well as between LP2(O8) and gauche C6-H11 one is only -0.0002.
16

ACCEPTED MANUSCRIPT
-
O
N
O
+
+
-
H3
C
H3
C
N
O
O
N
N
H
H
S
S
I 19.45%
II 19.00%
-
O
O
+
+
-
H3
C
N
H
3
C
N O
O
-
-
C
C
N
N
H
H
S
S
+
+
IV 6.74%
V 6.52%
O
H3
C
N
O
-
C
N
H
S
+
III 9.19%
Fig. 5. The predominating resonance structures of 3-methyl-4-nitroisothiazole (1) with their
contributions (>5%) computed with B3LYP/6-311++G(df pd) theory level.
,
Table 2. Natural atomic charges (
e
), Natural Atomic Valencies, selected second order
, kcal/mol) between donor → acceptor NBOs, and selected NBOs with
) and orbital energy (u.a) calculated at the B3LYP/6-311++G(df,pd)
2
interaction energy (
E
electron occupancies (
level of theory.
e
Natural Atomic
Charge (e)
0.672
Natural Atomic
Valencies
2.378
Atom
S1
N2
C3
-0.614
0.224
2.877
3.926
C4
C5
C6
N7
-0.017
-0.352
-0.600
0.491
3.640
3.887
3.986
3.762
O8
O9
H10
H11
H12
H13
-0.377
-0.383
0.233
0.233
0.226
1.634
1.627
0.992
0.992
0.997
0.995
0.264
2
Donor → acceptor
LP(2)S1 → π*N2-C3
LP(2)S1 → π*C4-C5
πC4-C5 → π*N2-C3
πC4-C5 → π*N7-O8
Energy (E , kcal/mol)
15.84
30.20
23.37
27.52
12.03
19.47
LP(2)O8 → σ*C4-N7
LP(2)O8 → σ*N7-O9
17

ACCEPTED MANUSCRIPT
LP(2)O9 → σ*C4-N7
12.76
LP(2)O9→ σ*N7-O8
LP(3)O9→ π*N7-O8
19.20
171.23
NBO
Occupancy (e)
Orbital energy (u.a.)
-0.47864
-0.39660
-0.35928
-0.33308
-0.31742
-0.31607
-0.29990
-0.28615
πN7-O8
LP(1)N2
πN2-C3
πC4-C5
LP(2)O8
LP(2)O9
LP(3)O9
LP(2)S1
1.98302
1.94332
1.89863
1.79612
1.90031
1.89946
1.45695
1.56269
On the basis of crystallographic and other data of 45 monoring isothiazole derivatives
it has been observed that the S-N bond length ranges from 1.617Å to 1.707Å, the S-C one
ranges from 1.679Å to 1.742Å, whereas the N-S-C valence angle ranges from 91.32 º to
9
7.37º (detailed data are collected in Table S2 in supplementary materials). As it is shown in
Fig. 4 and Table S2, the values computed for 3-methyl-4-nitroisothiazole are close to the
averages of these ranges. With reference to the previously described 1-methyl-nitropyrazoles
[
16], the increasing of the heteroatomic S1-N2, S1-C5 and C3-C4 bond lengths and
decreasing of N2-C3 and C4-C5 bond lengths are simultaneously observed in comparison to
analogous bond lengths in pirazole ring (in which sulfur atom is replaced by nitrogen one).
The similar trend was observed for unsubstituted isothiazole, in which the S1-N2 and S1-C5
bond lengths are 1.646Å and 1.702Å for microwave studies [52] and 1.642(5)Å and
1
.702(5)Å for electron diffraction [53], respectively. In the case of 3-methyl-4-
nitroisothiazole (1), molecule bond lengths, in which the sulfur atom is involved, are longer
about 0.3Å than the equivalent bond lengths, which nitrogen atom of pyrrole type creates in 1-
N
-methylnitropyrazoles. Therefore, the N2-S1-C5 angle (95.0º) is smaller for isothiazole ring
than reported for analogous angle in pirazole ring. This value of N2-S1-C5 angle in 3-methyl-
-nitroisothiazole (1) is similar to computed one for isothiazole at MP2/6-31G* (96.64º) [54]
4
and at B3LYP/6-31G* (94.93º) [54, 55] and at MP2/TZVP level of theory (95.4º) [56].
Obviously, reducing the N2-S1-C5 angle increases the remaining ring angles. Therefore other
bond angles of aromatic ring are higher for isothiazole than for pyrazoles, as it is shown in
Table S3 in the supplementary material.
The S1-N2 bond length (1.676Å) in 3-methyl-4-nitroisothiazole is longer than
reported for unsubstituted isothiazole, which is 1.642Å (electron diffraction structure [53])
and 1.6706Å (theoretical ground state [56]). The σS1-N1 natural bond orbital (NBO) occupied
5
.51
by 1.9708e is formed from an sp hybrid (83.78% p-character) of sulfur S1 interacting with
18

ACCEPTED MANUSCRIPT
4
.90
an sp
terms as σS1-N2 = 0.6392(sp
0.7690) corresponds with the higher electronegativity of nitrogen N2 in comparison to sulfur
S1 (0.6392). The natural charges are presented in the Table 2 and also confirm the much
higher electronegativity of nitrogen N2. The natural charge of S1 is 0.672 vs -0.614 for N1.
This is consistent with calculated natural electron configurations of S1 and N2, which are
hybrid (82.78%
p
5
-character) of nitrogen N2, which should be expressed in NBO
.51
4.90
)S1 + 0.7690(sp
)N2. The larger polarization coefficient of N2
(
e
e
1
.65 3.63
2
1.51 4.08
[
Ne]3s 3p and 1s 2s 2p , respectively.
Table 3. The comparison of HOMO/LUMO of 3-methyl-4-nitroisothiazole in various
environment (vacuum and water). The 3D shape frontier molecular orbital vizualizations with
energies, molecular orbital diagrams with normalized sum of squares of atomic orbital
2
coefficients (eigenvectors) c (on the basis of LCAO) and NBO analysis of canonical
molecular orbitals computed at B3LYP/6-311++G(df
,
pd) level for vacuum and water solution.
HOMO
LUMO
EHOMO(vacuum) = -8.123 eV
ELUMO(vacuum) = -3.096 eV
Energy gap (vacuum): ELUMO(vacuum) - EHOMO(vacuum) = 5.027 eV
0.014
O
0.147
O
0.000
0.000
N
N
O
0.005
O
0.250
H
H
0
.017
0.166
0.017
0.214
0.151
0.057
0.018
0.008
H
H
0
.018
H
0.008
S
0.556
S
0.122
H
0.149
0.015
0
.021
N
0.006
N
H
H
0.023
0.007
0.000
0.000
Normalization coefficient = 1.693
Normalization coefficient = 1.095
NBO analysis of frontier Molecular Orbitals with NBO contributions to MO (vacuum
)
*
φ
HOMO ≅ 0.701·LP(2)S1 + 0.385·π*C4-C5
+
φ
LUMO ≅ - 0.670·π N7-O8 + 0.438·LP(3)O9 +
*
*
-
0.383·πN2-C3 - 0.260·π N2-C3
+ 0.436·π C4-C5 - 0.297·LP(2)S1
19

ACCEPTED MANUSCRIPT
2
5
2
5.9% bonding character (14.7% N2-C3)
1.2% nonbonding character (49.1% S1)
2.9% antibonding c. (14.8% C4-C5 + 6.8% N2-C3)
5.2% bonding character
30.8% nonbonding character (19.2% O9 + 8.8% S1)
63.9% antibonding c. (44.9 % N7-O8 + 19.0% C4-C5)
EHOMO(water) = -7.896 eV
ELUMO(water) = -3.123 eV
Energy gap (water): ELUMO(water) - EHOMO(water) = 4.773 eV
0.017
O
0.165
O
0.000
0.000
N
N
O
0.004
O
0.304
H
H
0
.019
0.184
0.019
0.174
0.145
0.028
0.018
0.007
H
H
0
.018
H
0.007
S
0.570
S
0.105
H
0.147
0.021
0
.020
N
0.005
N
H
H
0.023
0.002
0.000
0.000
Normalization coefficient = 1.714
Normalization coefficient = 1.138
NBO analysis of frontier Molecular Orbitals with NBO contributions to MO (water
)
*
φ
HOMO ≅ 0.701·LP(2)S1 + 0.390·π*C4-C5
+
φ
LUMO ≅ 0.699·π N7-O8 - 0.449·LP(3)O9 +
*
*
-
0.377·πN2-C3 - 0.261·π N2-C3
- 0.380·π C4-C5 + 0.297·LP(2)S1
2
5
2
4.8% bonding character (14.2% N2-C3)
1.4% nonbonding character (49.2% S1)
3.7% antibonding c. (15.2% C4-C5 + 6.8% N2-C3)
5.5% bonding character
30.7% nonbonding character (20.1% O9 + 7.8% S1)
63.9% antibonding c. (48.9 % N7-O8 + 14.5% C4-C5)
The S1 atom also forms σS1-C5 natural bond orbital occupied by 1.9782e, which should
4
.46
2.74
be noted as σS1-C5 = 0.6814 (sp
)S1 + 0.7319 (sp
)C5. In this the σS1-C5 NBO is form from an
4
.46
2.74
sp
hybrid (81.07%
p
-character) on S1 interacting with an sp
hybrid (73.07% p-
character) on C5. The sulfur hybrid on σS1-C5 bond is bent away from line of S1-C5 centers by
.2°, whereas the carbon C5 hybrid is bent away by 5.7°. The difference between polarization
6
coefficients shows that the σS1-N2 is more than twice as polarized bond than σS1-C4 one. As a
result, the S1-N2 bond (1.676Å) is shorter than the S1-C5 one (1.699Å). The natural binding
indices (Natural Cluster Unit strength parameters) for S1-N2 and S1-C5 bonds are 1.0558 and
1
.1428, respectively. Similar relationship is provided by Wiberg bond orders in the natural
atomic orbital NAO basis, which are 1.1148 and 1.3060 for S1-N2 and S1-C5 bonds,
respectively. The natural bond order of S1-N2 NBO is 1.0591 (with 0.8110 covalent and
0
.2482 ionic contributions to bond order), and of S1-C5 NBO is 1.3191 (with 1.0682 covalent
and 0.2509 ionic one). Finally Natural Localized Molecular Orbital (NLMO) bond order S1-
N2 is 0.8670 and S1-C5 is 1.0805. Although, S1-C5 bond is longer than the S1-N2 one, four
above mathematical model of bond order description, suggest that the S1-C5 bond is stronger
20

ACCEPTED MANUSCRIPT
than the S1-N2 one. Moreover, the natural resonance theory module of NBO 6.0 program
shows that S1-C5 bond has more covalent character than S1-N2 one
Molecular structure of unsubstituted isothiazole was extensively described by Schultz
[
53] and also Palmer and Gordon [56]. In the case of 3-methyl-4-nitroisothiazole, the structure
*
of isothiazole ring is strongly stabilized by πC4C5 → π
interaction and strong electron
NBO orbitals. According to
N2C3
*
*
donation from sulfur lone pair to antibonding π N2C3 and π
C4C5
NBO 6.0 analysis, the second-order stabilization energy between πC4C5 (donor) and antibond
*
π
(acceptor) is 23.37 kcal/mol. The strongest stabilization, 30.20 kcal/mol is given from
N2C3
*
*
LP(2)S1 → π C4C5 interaction between the sulfur lone pair LP(2)S1 and antibond π C4C5 non-
Lewis orbital. The sulfur lone pair LP(2)S1 also electron donate antibond π*N2C3 NBO orbital,
*
and LP(2)S1 → π N2C3 delocalization energy is 15.84 kcal/mol, as it is shown in Table 2. D.
Kaur et al. reported quite analogues stabilized interaction for isothiazol, for which second
(
2)
*
*
order stabilization energies calculated for E LP(2)S1 → π C4C5 and LP(2)S1 → π N2C3 are
0.12 and 32.79 kcal/mol, respectively (se Table 3 at ref. [57]). Also similar stabilization
5
donor-acceptor (bond-antibond) interactions were found for pyrazole ring [16]. For example,
in case 1-methyl-4-nitropyrazole, the comparable second order delocalization energies are
*
*
2
9.19, 47.79 and 26.67 kcal/mol respectively for πC4C5 → π N2C3, LP(1)N1 → π C4C5 and
*
LP(1)N1 → π N2C3
.
The C4-N7 bond is relatively week so C4-N7 bond length is 1.454Å and C4-N7
natural bond order is only 0.9633 (with 0.7220 covalent and 0.2412 ionic components). The
1
.77
2.15
hybridization of N7 is sp on σC4N7 NBO and sp on both σN7O8 and σN7O8 NBOs, therefore
nitro group is flat. As it is shown in Table 2, the electron donating from second lone pair of
both oxygen atoms is significant. The acceptors of this electron densities are three antibound
σ* orbitals (σ*C4-N7, σ*N7-O8, σ*N7-O9). All of these σ-orbitals lie in the plane of the molecule
2
and these donor→acceptor interaction stabilized planar structure of NO group. The O8-N7-
O9 angle (124.8º) is a little smaller than in 1-methyl-nitropyrazoles, as an effect of described
previously interaction O8 with methyl group and attractive Coulomb interactions of O9 with
H13. For connecting the oxygen lone pair LP2(O9) with the C5-H13 bond, the off-diagonal
NBBP element is 0.0023, evidence of their attraction. It should be added that the NBCP
analysis clearly shows also that there is no intramolecular hydrogen bond between O9 and
…
H13, although the O9 H13 distance (2.579Å) is smaller than sum of Van der Waals radius
for hydrogen and oxygen atoms (2.72Å).
21

ACCEPTED MANUSCRIPT
Two leading resonance I and II structures of compound 1 are shown at the top of
Figure 5, which represent a resonance hybrid of two equivalent zwitter ionic polar structures.
The resonance weight for these two structures are not exactly equal, because interactions of
N7O8 and N7O9 with remainder are different for the reason that electron density in
isothiazole ring is not regular. Also computed bonds lengths are not exactly identical, and the
N7-O8 bond length (1.22526Å) of the C-N is slightly higher than the N7-O9 one (1.22510Å).
Fifteen distinct natural resonance structures provided by Natural Resonance Theory (with
threshold > 1%) are collected in Table S4 in supplementary materials. An interesting is third
resonance structure III with single O8-O9 bond, which is powerfully stabilized by N7 electron
donation. In this case, the LP(1)N7 → σO8-O9 interaction between nitrogen N9 lone pair and
the σO8-O9 bond is seen to give the extremely strong stabilization, 948.51 kcal/mol. The
resonance structures of isothiazole ring could be also simple described by ω-bond. The 3-
center, 4-electron hyperbond search has shown a strong hybrid of the two localized Lewis
structure. This resonance hybrid has 65.0% character described by localized Lewis structure
with double C4=C5 bond and single S1-C5 bond, and it has 35.0% character of alternative
structure with single C4-C5 bond and double S1=C5 bond (see resonant structures IV and V).
Also electron structure of nitro group can be described as 4-electron hyperbond. In this case
the resonance hybrid has 50.4% character described by localized Lewis structure with double
O8=N7 bond and single N7-O9 bond and it has 49.6% character of alternative structure with
single O8-N7 bond and double N7=O9 bond. This ω-bond is formally occupied by 4.0752
electrons, with corresponds to the negative charge of the nitro group, and it is a result of
strongly σ electron-withdrawing from isothiazole ring. One of the effects is the change in
electron density at the (alpha) C5−H13 bonds adjacent to the nitro group. The calculated
natural atomic charge for H13 is 0.264
e
, whereas for unsubstituted isothiazole ring the
. Because of these properties, the H13 hydrogen
corresponding natural atomic charge is 0.230
e
1
is acidic, which is consistent with H-NMR (δ = 9.52 ppm). Fifteen distinct natural resonance
structures provided by Natural Resonance Theory (with threshold > 1%) are collected in
Table S4 in supplementary material.
In the bottom of Table 2, eight occupied NBOs with the highest orbital energies (all σ,
π and LP NBOs are included in Table S5 in supplementary) are collected. As it shown, a lone
pair orbital localized upon the sulfur is the highest occupied NBO with 1.56269e occupancy
and -0.2861 u.a. orbital energy. This low occupancy of sulfur lone pair is a result of electron
donation LP(2)S1 to π*C4-C5, π*N2-C3 described previously and it is one of the reason that
22

ACCEPTED MANUSCRIPT
natural charge computed for S1 is 0.672e. In consequence, a partial positive charge forms on
the S1 atom, which is represent by third resonance structure in Figure 5. In contrast the energy
of LP(1)N2 NBO (-0.3966 u.a.) is lower than LP(2)S1 NBO, therefore electron occupancy of
nitrogen N2 lone pair is much higher (1.94332e) than sulfur lone pair. When considering an
aromatic substitution, the π-orbital of the highest energy plays a crucial role. Typically
electrophilic agents should attack the ring in a place where there is the greatest density of π
electrons. As it follows from Table 2, from two π-orbitals in izothiazole ring, the orbital with
the highest π electron density is πN2-C3 NBO. The polarization coefficients for these bonds,
0
.7685 of N2 and 0.6398 of C3, show that the highest π density in aromatic ring is on nitrogen
atom N2. Together, the πC4-C5 NBO in aromatic ring has the higher orbital energy than πN2-
C3 NBO, therefore π electron density on πC4-C5 is lower than on πN2-C3. Nonetheless, we
found an increased electron density on C5, which is related to sulfur electron donate and
mesomeric coupling of a nitro group. Additionally, the positive charge on N7 (0.491) makes it
a strong electron withdrawing and exerts a strong pull on neighboring electrons, especially
from oxygen O9 and O8 lone pairs. For comparison, the natural charges, computed for 3-
methylisothiazole, are 0.605, -0.630, 0.197, -0.297 and -0.384 for following S1, N2, C3, C4
and C5 atoms of isothiazole ring respectively. All these values of the natural charges
computed for 3-methyl-isothiazole are much smaller than corresponding ones calculated for
considered 3-methyl-4-nitroisothiazole (1), what simply confirms the electron-withdrawing
character of nitro group. It must be stressed that presented here “high occupied NBOs” can
not be equated with HOMO, HOMO-1, etc. Although all eight NBOs from bottom of Table 3
are partially involved in HOMO orbital.
4
.2. Reduction Potential and Electron Affinity of 3-methyl-4-nitroisothiazole in the light
of Frontier molecular orbitals and NBO analysis of canonical molecular orbitals
The frontier molecular orbitals are crucial to explanation of the electric properties,
chemical reactions (e.g. redox) and stability of molecule. The comparison of highest occupied
molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) with
corresponding NBO analysis of Canonical Molecular Orbitals (CMO) is shown in Table 3. As
it follows from NBO contributions to MO, the HOMO derives primarily (~49%) from the
sulfur LP(2)S1 lone pair, with weaker contribution from πN2-C3 bond (~15%) and two anti
*
*
bonds: π C4-C5 (~15%) and π N2-C3 (~7%). As a result, HOMO character is about 49% non-
bonding (on S1) and 29% bonding (mostly on adjacent N2-C3 bond). Also, the calculated
23

ACCEPTED MANUSCRIPT
coefficients of molecular orbital show that HOMO is composed exclusively of the 11 pz basis
functions on S1. The LP(2)S1 → π*C4-C5 interaction with 30.20 kcal/mol second order
stabilization energy, is responsible for delocalization electron density from LP(2)S1 and
deflection lobe of this lone pair in direction of the carbon C5, as it is shown on 3D
z
visualization (shape) for HOMO in Table 3. (The LP(2)S1 NBO is transformed from 2p AO
of S1 and it creates π-cloud). The molecular diagram with sums of squared atomic orbital
2
coefficients (eigenvectors) c of HOMO, clearly shows that the highest value of squared
2
eigenvectors c is for sulfur (Table 3), what points out that the highest probability of electron
2
finding is at S1 atom on HOMO. Squared eigenvectors c are important determinant of
probability of electron finding at appropriate position of a molecule. This remark is fully
compatible with presented NBO analysis, where we show that a lone pair orbital localized
upon the sulfur is the highest occupied NBO. In other side, the NBO analysis shows that π-
type NBO with the highest energy (-0.33308 a.u.) is πC4-C5 NBO, but lobes on C4 and C5
have opposite sign what weakens this π bond in consequence.
It is common knowledge that HOMO is acceptor of electrophic agents and fulfils
essential part in oxidation reactions of a molecule as electron donor. As it is apparent from
MO theory, the every N2, C3, C4 and C5 atoms give one and S1 two electrons to π-sextet of
aromatic ring. Obviously, these six electrons of π-sextet can create three bonding π-orbitals in
unsubstituted aromatic ring of isothiazole. But in case of 3-methyl-4-nitroisothiazole, analysis
of MO’s (LCAO conducted for vacuum) revealed six bonding π-orbitals (MO37=HOMO,
MO35=HOMO-2,
MO33=HOMO-4,
MO29=HOMO-8,
MO27=HOMO-10
and
MO24=HOMO-13) (see details in Tables S6, S7 and S8 in supplementary material). It results
from the reason that substituents of the isothiazole ring are involved in delivering additional
bonding π-orbitals, especially through overlapping of non hybridized atomic p -orbitals of
z
nitro group (particularly in MO33 (96%) and MO24 (91%)). Some contribution of non
hybridized orbital p
32%) and MO27 (19%), what results in addtitional stabilization of isothiazole ring (see Table
S8 in supplementary). Albeit, mainly the electrons of following MO, i.e., HOMO(MO37)-
0%, HOMO-2(MO35)-86%, HOMO-10(MO27)-57% and HOMO-8(MO29)-44%, give
essential contribution to stabilization of aromaticity (π-sextet) of isothiazole ring of 3-methyl-
-nitroisothiazole.
Formally, the NBO analysis assigned two electrons of sulfur atom to LP(2)S1,
z
of carbon C6 of methyl group can play similar role, especially in MO29
(
9
4
whereas in context of MO theory these electrons are involved in π-orbital (mainly in
HOMO(MO37), HOMO-2(MO35) and HOMO-10(MO27)). In both HOMO-2 and HOMO-8
24

ACCEPTED MANUSCRIPT
the π-cloud of aromatic ring is also delivered from electron density of πC4-C5 and πN2-C3
bonds. In the case of HOMO, the one lobe of π-cloud covers S1 and C5 atoms and second
lobe of π-cloud contains N2, C3 and C4 atoms in opposite sign. These electrons of highest π-
orbital plays crucial role in electrophilic substitutions. An increased total electron charge on
carbon C5 may suggest that this position of ring may be susceptible for initial addition of
electrophilic agent in electrophilic aromatic substitution in consequence of which the
hydrogen H13, which is attached to carbon C5, could be replaced by an electrophile.
However, calculated (e.g. for water environment at B3LYP/6-311++G(df
densities of HOMO (Table 3) for aromatic ring atoms S1, N2, C3, C4 and C5 atoms are
.1409e, 0.0455e, 0.2939e, 0.2899e and 0.0383e respectively. Similar results are for vacuum
calculation, which are following 1.1320e (S1), 0.0466e (N2), 0.2976e (C3), 0.3033e (C4) and
.0338e (C5). It shows that all positions of aromatic ring, apart from sulfur S1, have very low
,pd) level) π-electron
1
0
π-electron densities on HOMO. And additionally, calculated (at B3LYP/aug-cc-pVTZ level)
overall π-electron densities (for water environment), which come from all six bonding
determined as π-orbitals (MO37, MO36, MO33, MO29, MO27 and MO23) (see Table S7 and
S8 in supplementary material), assigned to S1, N2, C3, C4 and C5 atoms are 1.8802e,
1
.3235e, 0.7502e, 1.0105e and 0.8571e, respectively, what reveals strong deficiency of
overall π-electron density at C3 and C5 aromatic carbons. Similar results were obtained for
calculations (for water environment) at B3LYP/6-311++G(df pd) level, which are as it
,
follows: 1.9963e (S1), 1.3042 (N2), 0.7970e (C3), 1.0365e (C4) and 0.8780e (C5). Therefore,
because of low density of π-cloud, the electrophilic aromatic substitution reaction is unlikely
in this case. The 3D plot of those MOs with orbital energies, leading NBO contributions to
MO and bonding character are gathered in Table S6 in supplementary material.
In opposite to HOMO, the LUMO plays important role in reduction process of a
molecule (as an electron acceptor) and as a nuclephile agent acceptor. According to calculated
molecular orbital coefficients, the high contributions to LUMO is delivered from the 4pz basis
functions on C5 carbon atom and N7 nitrogen (with opposite sign). This is consistent with
expectations for the formally unoccupied antibounding π*C4-C5 and antibounding π*N7-O8
2
NBO orbitals, which computed occupancies are 0.35378e and 0.63522e, respectively. The c
values (as result of LCAO) of LUMO indicates a probability of location of new electron on
this molecular orbital, and as follows from Table 3, hence, the nitrogen N7 and both oxygen
atoms of nitro group are the most probable sites (56% and 65% for vacuum and water
solution, respectively) of electron acceptance. Furthermore, the NBO 6.0 analysis of canonical
25

ACCEPTED MANUSCRIPT
molecular orbitals confirms that LUMO is heavily localized on nitro group (about 65% of
LUMO) and on π*C4-C5 antibond (19%).
Presented data has been completed by measured reduction potential of 3-methyl-4-
A
nitroisothiazole, value of which is -0.3994 V. The calculated electron affinities (E ) for 3-
methyl-4-nitroisothiazole amount 1.351 eV (31.16 kcal/mol) and 3.484 eV (80.33 kcal/mol)
for vacuum and aqueous solution, respectively. These positive values of electron affinities
show that energy is released during electron acceptance. The calculated electron affinity of
aqueous solution is significantly much larger, what suggests that acceptance of additional
electron by 3-methyl-4-nitroisothiazole is much easier in polar aqueous solvent than in
vacuum. More detailed information is collected in Table S9 in the supplementary.
After electron acceptance, the unpaired (alpha) electron is mainly localized on nitro
group in radical anion (Fig. 6). The greater electron affinity E , the greater the reduction
A
potential is, what means the molecule can be easier reduced. The measured reduction potential
(-0.3994 V) is within the range specified for the nitropyrazoles [16, 17], for which there are -
0
.5044 V, -0.4094 V, and -0.2994 V for 1-methyl-4-nitropyrazole, 1-methyl-3-nitropyrazole
and 1-methyl-5-nitropyrazole, respectively. This notice may be supplemented by computation
for half-filled HOMO (e.i. SOMO - single occupied molecular orbital) of a radical anion
arising from the acceptance of free electron by 3-methyl-4-nitroisothiazole in water solution.
in vacuo
in water solution
0
.022
0.141
0.005
0
.181
H
O
H
O
0
.022
0.005
0.000
0
.187
0.295
H
H
H
0
.004
0
.011
0.014
N
N
0.018
C
O
C
O
0
.000
0.025
H
0
.128
0.166
0
.070
0
.256
0.178
0
.011
0.001
N
N
H
0.000
H
0.000
S
.139
S
0
0.121
MO Energy = +0.356 eV
Norm. coeff. = 1.069
MO Energy = -3.851 eV
Norm. coeff = 1.211
2
Fig. 6. Molecular diagrams with normalized squared eigenvectors values (c ) of HOMO
SOMO) of radical anion after electron acceptance by 3-methyl-4-nitroisothiazole molecule in
various environment (B3LYP/6-311++G(df pd)).
(
,
In case of this radical anion, the HOMO (SOMO) has nonbonding character. Leading NBO
contributions to HOMO are 26.5% of LP(1)N7 NBO, 21.7% of LP(2)O8 NBO and 20.7% of
LP(3)O9 NBO, which shows that electron density of HOMO (SOMO) orbital in radical anion
is mainly delivered from lone pairs of atoms of nitro group. Surprising that calculation for
26

ACCEPTED MANUSCRIPT
half-filled HOMO (SOMO) of a radical anion for gas phase (vacuum), shows that the highest
2
value of normalized squared eigenvectors c (probability of electron finding) is for carbon C5
(0.256). And secondly, the energy of SOMO has positive value for vacuum what suggests that
radical anion system is very unstable in this condition whereas polar solvent stabilizes it
significantly. It explains somewhat the creation of side product 8. It is apparent from Table 3,
the total electron density on nitro group for HOMO is somewhat higher in polar aqueous
solvent than in vacuum. Moreover, the total energy of the molecule of the radical anion
computed for a vacuum state (Evacuum = -813.071846705 hartree) is higher than computed for
water solution (Ewater = -813.159738982 hartree), what suggests that radical anion should be
more stable (∆E = 49.17 kcal/mol) in water than in vacuum. The energy gap for anion radical,
th
calculated as energy of LUMO (unoccupied 38 MO for beta electrons) minus HOMO (half-
th
filled 38 MO for alpha electrons), is 2.19 eV and 2.40eV, respectively for computation of
vacuum and water-solution environment. Differently, the energy difference between the
HOMO and LUMO calculated for 3-methyl-4-nitroisothiazole (1) is higher for gas phase
(5.027 eV) than for aqua solutions (4.773 eV). The energy gaps between frontier orbitals are
generally the lowest energy electronic excitation that is possible in a molecule. The detailed
comparison of HOMO/LUMO energies, electron affinities EEA and reduction potentials for 3-
methyl-4-nitroisothiazole (1) and 1-methyl-nitropyrazoles are gathered in Table S9 in
supplementary material.
4
.3. Analysis of experimental UV-Vis spectra of 3-methyl-4-nitroisothiazole with the
theoretical methods.
Considering the best compatibility of the experimental UV-Visible spectra with the
simulated one (Fig. 7) and taking into account the computation time, we recommend the
B3LYP/aug-cc-pVTZ method to excited states calculations (see discussion in supplementary
materials). For this reason, we discuss just this topic at B3LYP/aug-cc-pVTZ theory level.
While the other computations of properties of 3-methyl-4-nitroisothiazole were mainly
conducted at B3LYP/6-311++G(df
,pd) level in order to compare the results with previous
calculations for nitropyrazoles [16, 17].
27

ACCEPTED MANUSCRIPT
Fig. 7. Comparison of the experimental UV-Vis spectra of 3-methyl-4-nitroisothiazole 1
-4
(
c = 1.00·10 M) with the simulated ones on the basis of B3LYP/aug-cc-pVTZ theory level.
4
.3.2. Water solution of 3-methyl-4-nitroisothiazole
Calculation of excited states and their appropriate oscillator strengths (f) for water
solution of 3-methyl-4-nitroisothiazole shows five very highly probable excitations at 190-
00 nm wave lengths range: 302.74 nm (f=0.2289), 270.50 nm (f=0.1862), 216.42 nm
4
(f=0.1382), 210.13 nm (f=0.2287), 196.74 nm (f=0.1995) (see supplementary material)
Analysis of eigenvectors (sum of squared c orbital coefficients) of considered MO’s
reveals that the most probability of electron finding on HOMO (MO37) is at sulphur region
57%) (see Table 4). On LUMO (MO38), the excited electron can appear mainly at nitro
(
group region, N7 (34%), O8 (20%), O9 (18%) what results in above 70% of probability of
electron finding there. Next, bonding MO36 has following positions with the most probable of
28

ACCEPTED MANUSCRIPT
electron finding, mainly at the region of N2 (37%), C4 (21%) and C5 (26%) atoms involved
in π electron sextet of isothiazole ring. Antibonding MO39 can accept excited electron mainly
at region of following atoms S1 (20%), N2 (26%), C3 (16%) and C5 (17%) of isothiazole
ring.
29