Spectroscopic and theoretical study of the o-vanillin hydrazone of the mycobactericidal drug isoniazid

Article abstract of DOI:10.1016/j.molstruc.2011.10.026

A complete and detailed study of the hydrazone obtained from condensation of antituberculous isoniazid (hydrazide of the isonicotinic acid, INH) and o-vanillin (2-hydroxy-3-methoxybenzaldehyde, o-HVa) is performed. It includes structural and spectroscopic

Full text of DOI:10.1016/j.molstruc.2011.10.026

Journal of Molecular Structure 1007 (2012) 95–101
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Structure
journal homepage: www.elsevier.com/locate/molstruc
Spectroscopic and theoretical study of the o-vanillin hydrazone
of the mycobactericidal drug isoniazid
a
a
b
Ana C. González-Baró a, , Reinaldo Pis-Diez , Beatriz S. Parajón-Costa , Nicolás A. Rey
⇑
a CEQUINOR (UNLP/CONICET), Fac. Cs. Exactas, CC962, 1900 La Plata, Argentina
b Laboratório de Síntese Orgânica e Química de Coordenação Aplicada a Sistemas Biológicos (LABSO-BIO), Chemistry Department,
PUC-Rio, Rua Marquês de São Vicente, 225 – 22453-900 Gávea, RJ, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 23 August 2011
Received in revised form 11 October 2011
Accepted 13 October 2011
Available online 25 October 2011
A complete and detailed study of the hydrazone obtained from condensation of antituberculous isoniazid
(hydrazide of the isonicotinic acid, INH) and o-vanillin (2-hydroxy-3-methoxybenzaldehyde, o-HVa) is
performed. It includes structural and spectroscopic analyses, comparing experimental and theoretical
results. The compound was obtained as a chloride of the pyridinic salt (INHOVA+Clꢀ) but it will be
referred as INHOVA for the sake of simplicity. The conformational space was searched and optimized
geometries were determined both in gas phase and including solvent effects. Vibrational (IR and Raman),
electronic and NMR spectra were registered and assigned with the help of computational methods based
on the Density Functional Theory. Isoniazid hydrazones are good candidates for therapeutic agents
against tuberculosis with conserved efficiency and lower toxicity and resistance than parent INH.
Ó 2011 Elsevier B.V. All rights reserved.
Keywords:
Isoniazid
Hydrazone
o-Vanillin
Spectroscopy
DFT calculations
1. Introduction
used for analytical determination of INH in blood by formation of a
colored compound (phtivazide or vanicide) [5]. This colorimetric
Isoniazid, the hydrazide of isonicotinic acid (INH, see Scheme 1)
has been recognized as an effective antituberculous agent since the
middle of the 20th century. It is still employed in the treatment
and prevention of this disease, not only as a single drug but also
combined with others. Unfortunately, the actual formulations
show several undesired collateral effects and can even cause irre-
versible damage in the liver in chronic patients. This fact, together
with the resistance that microorganisms develop against these
drugs, encouraged the search for new compounds with therapeutic
effects [1,2]. The combination of INH with some hydroxyaldehydes
leads to the formation of stable hydrazones that show conserved
activity and less toxicity, due to the inactivation of the NH2 group
of INH [3]. Particularly, a group of hydrazones has been determined
to be more effective than INH itself as antituberculous agents in
macrophages [4]. This result is important, since the goal of a ther-
apy against tuberculosis is to eradicate the Mycobacterium both in
the extracellular and intracellular environment.
method is still used nowadays for quantification of INH in pharma-
ceutical formulations [6]. Both HVa and its isomer o-HVa (2-hydro-
xy-3-methoxybenzaldehyde, see Scheme 2) have antioxidant
activity, due to their ability to capture free radicals. This behavior
is related to the presence of the OH moiety in their structure [7].
The low toxicity and therapeutic properties of these aldehydes
make them good candidates for the preparation of hydrazones
with a better pharmacological profile. Taking into account that
tuberculosis, as other endemic diseases caused by parasites,
mainly affects population of low economical level, the develop-
ment of low-cost drugs is also an important objective in the re-
search field [8].
It is well known that the biological activity of a compound is
strongly dependent on its physicochemical, structural and elec-
tronic properties. Moreover, INHOVA could be a suitable ligand
for the development of coordination compounds because of its po-
tential ability to coordinate several metal ions of biological and
pharmacological interest through its N and O donor atoms [9,10].
Thus, the detailed physicochemical characterization of organic
compounds that can be used in the development of therapeutic
agents is an important tool in the understanding of its mechanism
of interaction both with biological substrates as with biometallic
centers. Additionally, complementary studies on the solution
chemistry of such a drug candidate, even when performed in an
Among the hydroxyaldehydes that can react with INH, vanillin
(4-hydroxy-3-methoxybenzaldehyde, HVa, see Scheme 2) has been
⇑
Corresponding author. Tel./fax: +54 221 4240172.
E-mail addresses: agb@quimica.unlp.edu.ar (A.C. González-Baró), pis_diez@
quimica.unlp.edu.ar (R. Pis-Diez), beatrizp@quimica.unlp.edu.ar (B.S. Parajón-Costa),
nicoarey@puc-rio.br (N.A. Rey).
0022-2860/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.molstruc.2011.10.026

96
A.C. González-Baró et al. / Journal of Molecular Structure 1007 (2012) 95–101
heating. Drops of concentrated hydrochloric acid (Merck) were
added to reach a pH value of 5. The starting pale yellow solution
turned immediately to bright yellow and a precipitated was formed.
The system was left in digestion for 2 h and the solid was filtered and
washed with methanol. This procedure differs from the previously
reported by Chen et al. [11]. The product corresponds to the hydra-
zone hydrochloride hemi-hydrate: INHOVAꢁHClꢁ½H2O (Yield: 78%;
m.p: 228–229 °C; Anal. Calcd. for C14H15O3.5N3Cl: C, 53.04%; H,
4.74%; N, 13.26%. Found: C, 52.24%; H, 4.59%; N, 12.98%).
Scheme 1. Schematic representation of isoniazid (INH).
2.2. Spectroscopic analysis
IR spectra were recorded with a Bruker 113v FTIR instrument,
using the KBr pellet technique. Raman spectra were obtained with
a Perkin–Elmer Raman Station 400, using the 785 nm line for exci-
tation. Carbon, hydrogen and nitrogen determinations were
performed in an elemental analyzer EA 1110 from CE Instruments.
Melting point was determined with a MQAPF-302 Micro-Química
instrument. The electronic absorption spectrum of the compound
(5 ꢂ 10ꢀ5 M) was measured on methanolic solution in the
200–800 nm spectral range. It was recorded with a Hewlett–Pack-
ard 8452-A diode array spectrometer, using 10 mm quartz cells. 1H
and 13C NMR spectra were obtained at room temperature in a
Varian Mercury 200 spectrometer (4.7 T, 200 MHz for the 1H nu-
clei) using a 5 mm probe and d6-DMSO as solvent. Calibration
was made with the solvent residual peaks as references (2.50
and 39.52 ppm for hydrogen and carbon, respectively) [12].
Scheme 2. Schematic representation of vanillin (HVa) and its isomer o-vanillin
(o-HVa).
2.3. Computational methods
The conformational space of the protonated form of INHOVA
was investigated with the aid of the semiempirical PM3 method
[13]. Several starting geometries derived from selected variations
in the C5AC4AC7AN9, C4AC7AN9AN10 and N9AN10AC11AC12
dihedral angles were considered to investigate how rings orientate
relative to each other, see Scheme 3 for atom labeling. Moreover,
dihedral angles involving the OH and OCH3 groups were also ac-
counted for to achieve an adequate description of the conforma-
tional space of the molecule under study.
Scheme 3. Schematic representation of the hydrazone obtained from condensation
of INH and o-HVa (INHOVA+).
Geometries were further optimized using the hybrid meta-GGA
M06–2X exchange–correlation density functional [14] with a tri-
ple-zeta 6–311+G(d,p) basis set [15,16]. Numerical integrations
were carried out using a grid containing 96 radial points and 590
angular points around each atom. The critical points found after
optimization were characterized by the sign of the eigenvalues of
the Hessian matrix of the total electronic energy with respect to
the nuclear coordinates. When the critical point corresponded to
a minimum on the potential energy surface of the molecule, the
eigenvalues were converted to harmonic vibrational frequencies
and a set of thermodynamic functions were calculated. Vibrational
frequencies were scaled by a factor of 0.982 to ease the comparison
with experimental values [14].
organic solvent, could shed light on the system and give us some
information about the nature of the species occurring in INHOVA
solutions.
In this work we report a detailed study of the hydrazone obtained
from condensation of INH and o-HVa in methanol, differing from the
procedure described in the literature for its synthesis [11]. The com-
pound was obtained as a chloride of the pyridinic salt (INHOVA+Clꢀ)
but it will be referred as INHOVA (see Scheme 3) for the sake of sim-
plicity. The work includes experimental and theoretical results.
Vibrational (IR and Raman), electronic and NMR spectra were regis-
tered and assigned with the help of computational methods based
on the Density Functional Theory.
Solvent effects (methanol) were included implicitly through the
Conductor-like Polarizable Continuum Model [17,18] to obtain the
free energy of solvation. Geometries were kept frozen to their gas-
phase structures, thus only wavefunction polarization effects were
taken into account.
2. Experimental methods
2.1. Synthesis
Electronic transitions were calculated within the framework of
the Time-Dependent Density Functional Theory [19,20]. In this
case, a coarser grid was used for numerical integrations containing
96 radial points and 302 angular points around each atom.
Geometry optimizations, Hessian matrix calculation and diago-
nalization, and electronic transition calculations were performed
with the GAMESS-US program [21].
All chemicals were of analytical grade and were used as pur-
chased. The compound was prepared according to the reported pro-
cedure for the reaction of HVa and INH employed as analytical
method [6], replacing ethanol by methanol and HVa by o-HVa. Ten
milliliter of methanolic solution containing 3 mmol of o-HVa (Sig-
ma) was drop-wise added to a solution of 1.5 mmol of INH (Fluka)
in 16 mL of methanol (Merck), with continuous stirring and slight
Isotropic magnetic shieldings of 13C and 1H were calculated at
the B3LYP/6–311+G(2d,p)//B3LYP/6–31Gꢃ level of theory as

A.C. González-Baró et al. / Journal of Molecular Structure 1007 (2012) 95–101
97
Table 1
suggested by Cheeseman and co-workers, using the Gauge-Includ-
ing Atomic Orbital method [22–24]. Isotropic magnetic shieldings
were turned into chemical shifts by subtracting the corresponding
isotropic magnetic shieldings of TMS, which were calculated at the
same level of theory as above. Solvent effects (DMSO) were in-
cluded implicitly through the Conductor-like Polarizable Contin-
uum Model [25]. Isotropic magnetic shieldings were calculated
with the Gaussian 03 package [26].
All drawings in this work were prepared with Gabedit [27].
The calculated electronic spectrum of INHOVA was simulated
from the theoretical results to ease the comparison with experi-
mental data. The simulation was performed with the following
function:
Comparison of calculated geometrical parameters of INHOVA with experimental
values. Calculated values are shown for isomers 1 and 2. Relative free energies are
also shown. Bond lengths are in Å and bond and dihedral angles are in degrees. Free
energies are in kcal/mol. See Scheme 3 for atom labeling.
Expa
Calc. – 1
Calc. – 2
Bond lengths
1.223(4)
1.346(5)
1.369(4)
1.280(5)
1.362(4)
1.430(5)
1.363(4)
2.632b
C7AO8
C7AN9
1.212
1.359
1.362
1.280
1.354
1.421
1.345
2.650
1.802
3.919
4.704
1.217
1.356
1.365
1.278
1.357
1.424
1.359
3.133
4.078
2.757
1.968
N9AN10
N10AC11
C16AO17
O17AC18
C19AO20
N10AO20
N10ꢁ ꢁ ꢁHO20
O20AN9
O20ꢁ ꢁ ꢁHN9
1.914b
X
1
p
r
fi
3.909b
IðEÞ ¼
2
4.577b
r2 þ ðE ꢀ ei
Þ
i
Bond angles
120.9(4)
116.8(3)
118.9(3)
116.9(3)
120.8(4)
125.0(4)
117.5(3)
117.5(3)
122.0(3)
145.5(3)
35.5b
where I is the peak intensity at a given energy E,
ei are the calculated
C4AC7AO8
C4AC7AN9
120.0
114.7
119.0
117.5
121.3
125.0
117.3
117.0
123.6
143.0
35.3
118.9
119.7
122.1
119.2
131.3
126.3
117.7
119.0
120.3
11.1
electronic transitions, is the band width and fi is the oscillator
r
C7AN9AN10
N9AN10AC11
N10AC11AC12
C15AC16AO17
C16AO17AC18
C16AC19AO20
C12AC19AO20
N10AHAO20
O20AHAN9
strength of the electronic transition.
3. Results and discussion
3.1. Conformational analysis and geometry optimization
Two isomers of INHOVA were found after the conformational
search. They differ mainly in the conformation around the
N10AC11 double bond, see Fig. 1. Gas-phase geometry optimiza-
tions show that form 2 is more stable than 1 by about 8 kcal/mol.
When solvent effects are included, free energy of solvation of 1 is
greater than that of 2 by about 12 kcal/mol, thus form 1 becomes
more stable than 2 by about 4 kcal/mol in solution.
Both isomers present an intramolecular hydrogen bond, the
atoms involved in those bonds depending on the value of the
N9AN10AC11AC12 dihedral angle. The N10ꢁ ꢁ ꢁHO20 hydrogen
bond is observed for isomer 1, whereas a bond between O20 and
the HN9 group is predicted for 2. Selected experimental [11] and cal-
culated structural parameters of both stable isomers are shown in
Table 1. Despite the expected deviations between experimental
and calculated bond lengths, it is worth noting the excellent agree-
ment between the experimental N10ꢁ ꢁ ꢁO20 and N10ꢁ ꢁ ꢁHO20 bond
distances and the calculated values for isomer 1. The apparent dis-
crepancy in the N10ꢁ ꢁ ꢁHO20 bond distance is due to the fact that
experimental Hꢁ ꢁ ꢁO20 distance is reported to be 0.820 Å, whereas
the calculated value is 0.979 Å. Calculated values of N10ꢁ ꢁ ꢁO20 and
N10ꢁ ꢁ ꢁHO20 for isomer 2, on the other hand, are about 0.5 and
132.0
Dihedral angles
ꢀ175.1(3)
ꢀ178.9(3)
ꢀ1.3(5)
C5AC4AC7AN9
ꢀ151.8
180.0
ꢀ0.7
ꢀ151.5
1.0
N9AN10AC11AC12
C15AC16AO17AC18
N10AN9AC7AO8
C4AC7AN9AN10
N10AC11AC12AC13
C16AC19AO20AH
3.7
2.9(6)
ꢀ1.0
ꢀ170.3
11.2
136.3
0.7
0.0
+3.88
ꢀ177.3(3)
174.7(4)
178.3b
178.7
178.4
ꢀ179.4
+8.63
0.0
D
D
G (gas-phase)
G (methanol)
a
Values taken from Ref. [11].
These values are obtained after converting fractional coordinates provided in
the CIF file of Ref. [11] to Cartesian coordinates. Thus, no standard uncertainties are
given.
b
2.0 Å larger than the experimental distances. It is also evident from
the table that experimental O20ꢁ ꢁ ꢁN9 and O20ꢁ ꢁ ꢁHN9 bond
istances suggest that no hydrogen bond is formed between those
atoms. Experimental bond angles are reasonably well reproduced
by the two stable isomers. The calculated N10AC11AC12 bond angle
for isomer 2 seems to be the exception as it is more that 10° larger
than the experimental value. The N10AHAO20 and O20AHAN9
bond angles, which are used as geometric parameters to describe
intramolecular hydrogen bonds, show an excellent agreement be-
tween experimental values and calculated values for isomer 1. Dihe-
dral angles are the most important geometric parameters as they
could describe conformation features in detail. The experimental va-
lue of the C5AC4AC7AN9 angle clearly indicates that the pyridine
ring of the INH moiety re-orients itself when the solid is formed from
the methanolic solution, changing that dihedral angle more than
20°. Finally, the experimental values of the N10AN9AC7AO8,
C4AC7AN9AN10, N10AC11AC12AC13 and C16AC19AO20AH
dihedral angles are very well described by the values found for iso-
mer 1. Dihedral angles calculated for isomer 2, on the other hand, ex-
hibit large deviations from experimental data.
Present results seem to suggest that solvent plays a dominant
role in determining the stable structure of INHOVA, which, in turn,
modifies slightly during crystallization. Taking these results into
account, calculation of both solid state and solution properties
were performed on isomer 1.
Fig. 1. Optimized geometries of the two stable isomers of INHOVA found in the
present work.

98
A.C. González-Baró et al. / Journal of Molecular Structure 1007 (2012) 95–101
Table 2
Complete assignment of the IR and Raman spectra of INHOVA. Experimental data of INH and o-HVA are included for comparison. Frequencies are given in cmꢀ1. Calculated
frequencies correspond to isomer 1.
INH
IR
o-HVA
INHOVA
IR
Calculated
Assignment
Raman
IR
Raman
Raman
3578 w
3428 m
3205 w
3157 w
m
m
m
m
m
m
m
m
m
m
m
m
m
m
m
m
m
as(H2O)
s(H2O)
NH(Py)
NH(hydrazone)
NH2
CH(o-HVA)ip
CH(o-HVA)op
CH(INH)ip
CH(INH)op
asCH(methyl)
OH
asCH(methyl)
CH(aldehyde)
sCH(methyl)
CH(aldehyde)
C@O
3513
3492
3304 m
3111 vs
3111/3093 w
3066/3054 m
3088 vw
3069 vw
3083 vw
3073 vw
3088 m
3088 vw
3186
3173
3050 w
3013 w
3039 vw
3014 w
2973 w
2939 w
2884 w
2839 w
3041 vw
3016 w
2976 vw
2942 vw
2887 vw
2849 vw
3050 w
3014 w
2985 vw
2947 vw
3133
3594
3061
3022
3014 vvw
2948 vw
2844 vw
1688 vs
2846 vw
1683 w
1667 vs
1670 m
1795
1645 vs
1641 s
C@O(o-HVA)
1634 s
1602 m
1643 w
1602 vs
dNH2
(m
ring + dCH)INH
1606 vs
1636 w
1576 w
1564 m
1501 m
1464 s
1604 vs
1637 mw
1556 s
1689
1669
1657
1614
1671
1482
1460
m
m
m
m
C@N
ring(INH) +
ring(o-HVA)
ring(INH) +
m
ring(o-HVA)
1591 sh
1589 w
m
ring(o-HVA)
1556 s
1412 s
1550 w
1518 sh
1467 m
dNH(Py) + (
dasCH3
dsCH3
(dCH +
dOH
(dCH +
qrNH2
m
ring + dCH) INH
1471 m
1455 s
1470 m
1458 m
1440 sh
m
m
ring)INH
ring)INH
1388 s
1374 m
1368 vw
1332 w
1338
1361
1334 s
1321 sh
1332 vs
1322 sh
1327 s
1257 s
1217 m
1330 s
1324 m-w
1283 m
1256 vs
1323 w
1281 s
1260 sh
1230 m
1316
1284
1263
1256
m
ArAOH + dCH(ring + aldehyde) o-HVA
ArAOCH3 + dCH(ring) + dOH) o-HVA
ArAOCH3 + dCH(ring + aldehyde) o-HVA
(m
1275 vw
1218 m
1187 vw
m
1221 m
1218 ms
1187 s
dNH(hydrazone + Py) + dCH(INH)
CAC + dCH)aldehyde + dCHring
NAN + dCH + CX
dCH + rCH3
rCH3
dCH +
(m
1192 vw
m
m
1184 sh
1163 m-w
1188 vw
1149 m
1181/1197
1166
q
q
1142 m
1098 sh
1132 w
1095 w
m
ring +
m
CN
1148 w
m
NAN + dCH
dCH +
m
ring +
m
CN
1096 w
1078 m
1067 m-w
1030 vw
1005 m
968 m
1097 w
1079 w
1066 w
1105
1134
1067
1091
1011
1000
962
dCH (o-HVa + INH)
CAO(OCH3) + dOH + dNNC + dCH(o-HVA)
(dCH + dring) INH
(NANAC@O)
dring(INH)
dring + CAO(OACH3)
CH(o-HVA)
1070 s
1092 w
946 vw
m
1062 w
995 s
1056 w
1003 vs
m
1007 m
971 w
891 w
947 s
895 w
m
893 m
c
c
c
888 m-w
845 s
887 ms
849 w
CH +
CH +
c
c
CX
CX
853 w
839 m
780 m
905
612
842
dNAC@O
OH
dring(o-HVA)
CH (ring + aldehyde)o-HVA
CH ip + NH + CO)INH
CH(o-HVA)ip
838 m
779 w
763 s
837 w
837 w
c
c
746 m
751 m
731 s
752 vvw
731 w
790
775
(c
c
c
737 m
717 s
740 vw
718 m
c
dring + dCH aldehyde + d(ArAOCH3)
dring + CX + dNAC@O
(dring + dCH)INH
675 s
660 m
682 m-w
666 m
m
686 w
686 vw
692
631 sh/
608 m,b
633 w
434 sh
675
cCAC(O)AN + cring(INH)
453 vw
453 vw
443
cNH(hidrazone)
X = COANHANH2, vs: very strong; s: strong; m: medium; w: weak; vw: very weak; b: broad; sh: shoulder; ip: in-phase; op: out-of phase.
3.2. Vibrational spectroscopy
presented in Table 2 and compared with the bands of the INH and
o-HVa IR spectra. Fig. 2 shows the experimental IR and Raman spec-
The solid-state vibrational properties of INHOVA were examined
by infrared and Raman spectroscopies. The spectral assignments
were done based on reported data [28,29] and with the aid of the
calculated normal mode frequencies. The proposed assignment is
tra in the 1800–400 cmꢀ1 range.
The bands at higher frequencies are assigned to crystallization
water molecules that were not considered in the calculations.
The IR band assigned to NAH stretching at 3205 cmꢀ1 denotes

A.C. González-Baró et al. / Journal of Molecular Structure 1007 (2012) 95–101
99
involved in hydrogen bonds. This effect also explains the fact that
the calculated frequency of the out-of-plane OH bending is more
than 200 cmꢀ1 lower than the experimental value. On the other
hand, calculated stretchings of both NH groups and the CO group
differ from experimental values from 100 cmꢀ1 to almost
400 cmꢀ1. It is argued that those groups are involved in hydrogen
bonds with both hydration water molecules and the chloride anion
according to the crystal packing reported in Ref. [11]. As present
results are obtained for the isolated INHOVA molecule, every pos-
sible intermolecular hydrogen bond is merely ignored in the
calculations.
3.3. NMR spectroscopy
Table 3 displays the 1H and 13C NMR chemical shifts for
INHOVA, along with the signals assignment. 1H NMR spectrum of
INHOVA was already reported by Hingorani and Agarwala [10].
Two well-defined aromatic spin systems can be observed for the
molecule: pyridine hydrogens, more de-shielded, appear above d
8.00, while the phenolic ring hydrogens are observed in the range
6.75–7.25. The absorption of the HAC@N Schiff base hydrogen, a
very important signal to characterize the formation of INHOVA, is
present at d 8.76. Our data are not completely consistent with
those published previously. Some hydrogens seem to be more
de-shielded than reported; namely, H2, H6, H3, H5, HAC@N,
HAO, and OCH3. In the case of H2, H6, H3 and H5 a possible expla-
nation resides in the protonation of INHOVA at the pyridine ring,
which withdraws electron density from these hydrogen atoms.
On the other hand, the reason for the discrepancies concerning
HAC@N, HAO, and OCH3 is unclear. The 13C NMR spectrum of INH-
OVA is reported here for the first time. 12 signals are observed for a
total of 14 carbon atoms. The pairs C2/C6 and C3/C5 are chemically
equivalent in solution. Carbonyl signal is observed at d 160.1 and
that of the HAC@N group, at d 149.4.
Fig. 2. Experimental IR and Raman spectra of INHOVA in the 1800–400 cmꢀ1
spectral range.
Theoretical calculations are in good agreement with experimen-
tal data. From Table 3 it can be noticed that both H and C atoms in
positions 2 and 6, and 3 and 5 are equivalent in solution, whereas
calculated values indicate they are slightly nonequivalent. This
could be explained by considering that calculations were done for
1, which is a static conformation, while in solution atoms can freely
rotate along a single bond and thus, observed chemical shifts are
protonation of the pyridinic nitrogen atom. The calculated fre-
quency for this mode is higher than the calculated for the NAH
stretching of the hydrazone moiety observed at 3157 cmꢀ1, accord-
ing to the expecting shift to lower frequencies compared with the
NAH stretching mode of the hydrazide moiety in the IR spectra of
INH (3304 cmꢀ1). Calculations show that the in-plane bendings of
pyridinic and hydrazonic NAH are coupled with other modes. The
band at 1501 cmꢀ1 (IR) is assigned to the first while the band at
1230 cmꢀ1 (Ra) has a higher contribution of the second one. The
hydrazonic NAH in-plane bending mode is not observed in the
experimental IR spectrum because the corresponding band is over-
lapped with the very strong band at 1256 cmꢀ1. The bands related
to the NH2 group of INH (3111, 1634 and 1321 cmꢀ1) are absent in
the INHOVA spectra and the very strong band at 1606 cmꢀ1 (IR),
1604 cmꢀ1 (Ra) is assigned to the stretching mode of the C@N bond
in the hydrazone. Other bands related to the NNCO group appear in
the spectra at lower frequencies (below 1030 cmꢀ1). The
vibrational modes of the OH group of o-HVa are conserved in the
INHOVA spectra. The weak band at 1148 cmꢀ1, assigned to NAN
stretching mode, denotes a shift to higher frequencies upon
hydrazone formation if compared to the band in the INH spectra.
It is worth mentioning that some calculated harmonic vibra-
tional frequencies are much higher than the corresponding exper-
imental values. The discrepancy of more than 500 cmꢀ1 observed
in the OH stretching mode could be attributed to the fact that even
when this group is involved in hydrogen bond with N10, the calcu-
lated OH distance remains almost unaltered. Then, the correspond-
ing harmonic vibrational frequency recalls that of an OH group not
Table 3
1H (200 MHz) and 13C (50 MHz) NMR chemical shifts, in ppm, for INHOVA (d6-DMSO).
Calculated chemical shifts for isomer 1 are also shown. See Scheme 3 for atom
labeling.
C/H
dC (Exp.)
dC (Calc.)
dH (Exp.)
dH (Calc.)
12
19
16
15
14
13
HAO
OCH3
HAC@N
HAN
C@O
2
3
4
5
6
114.1
148.0
147.2
119.2
119.0
120.3
–
55.87
149.4
–
160.1
146.8
123.5
143.8
123.5
146.8
123.4
158.0
156.4
119.2
125.4
129.7
–
57.19
161.1
–
167.4
149.9
134.8
159.0
130.8
150.1
–
–
–
–
–
–
7.22 (d, 1H, 3JHH = 8.0)
6.88 (t, 1H, 3JHH = 8.0)
7.07 (d, 1H, 3JHH = 8.0)
12.46 (s, 1H)
7.4479
7.4587
7.4557
11.56
3.98
8.91
10.3
–
9.18
9.01
–
8.65
9.19
12.48
3.83 (s, 3H)
8.76 (s, 1H)
10.70*
–
8.90 (d, 2H, 3JHH = 6.0)
8.04 (d, 2H, 3JHH = 6.0)
–
8.04 (d, 2H, 3JHH = 6.0)
8.90 (d, 2H, 3JHH = 6.0)
–
NH(Py)
The signals multiplicity, as well as their respective integration and coupling con-
stants (J, Hz) are in parenthesis.
*
Very broad singlet of difficult observation. It cannot be possible to observe the
signal referent to the protonated pyridinic nitrogen (see text).

100
A.C. González-Baró et al. / Journal of Molecular Structure 1007 (2012) 95–101
Table 4
average values of different conformations. Only two signal inver-
sions were observed in the 13C spectrum when comparing theory
and experiment. They involve the nuclei in positions 4 and 15. In
the case of C4, theory predicts that this nucleus should be the more
de-shielded aromatic one. On the other hand, calculations indicate
that C15 should be the more shielded aromatic nucleus. In both
cases, theoretical predictions do not support the experimental data.
Electronic spectra of 5 ꢂ 10ꢀ5 M methanolic solution of INHOVA. Calculated elec-
tronic transitions for isomer 1 are also shown. Only those calculated electronic
transitions relevant for the assignments are listed. Absorption maxima are given in
nm. Oscillator strengths of calculated transitions, shown in parenthesis, are in atomic
units.
Exp.
204
Calc.
Assignment
211 (0.533)
HOMO ? LUMO + 7 (64%)
HOMO ? LUMO + 8 (11%)
HOMO ꢀ 1 ? LUMO + 2 (79%)
HOMO ꢀ 1 ? LUMO (66%)
HOMO ? LUMO (15%)
3.4. Electronic spectroscopy
222
304
250 (0.332)
306 (0.295)
The electronic spectrum of a methanolic (5 ꢂ 10ꢀ5 M) solution
of INHOVA was registered in the 200–800 nm spectral range and
it is shown in Fig. 3. The simulated spectrum obtained from
calculated electronic transitions is also depicted in the figure.
Absorption maxima are compared with calculated values in
Table 4. The corresponding assignments are also provided in the
table.
336
335 (0.240)
HOMO ? LUMO (62%)
HOMO ꢀ 1 ? LUMO (21%)
HOMO ? LUMO + 2 (14%)
Results of the calculations show that many-body effects are far
from negligible for INHOVA, a fact that it is reflected in that most
transitions are described by two or more one-electron excitations.
Thus, only the dominant transitions, chosen according to their oscil-
lator strength, are used to assign the observed bands for simplicity. It
can be seen from Table 4 that calculated transitions describe very
well the experimental electronic spectrum. The only exception is
the band observed at 222 nm, which is assigned to a transition cal-
culated at 250 nm. The observed band at 304 nm and the shoulder
at 336 nm are dominated by one-electron excitations from the
HOMO ꢀ 1 and HOMO (Highest Occupied MO) to the LUMO (Lowest
Unoccupied MO), respectively, in spite of minor contributions from
other one-electron excitations. The observed band at 222 nm is as-
signed to a sole one-electron excitation from the HOMO ꢀ 1 to
LUMO + 2. Finally, the observed band at 204 nm is assigned to a
dominant one-electron excitation from the HOMO to the LUMO + 7,
with a minor contribution from a HOMO ? LUMO + 8 excitation. It
is worthnotingthat theobserved band at 204 nm is assigned to a cal-
culated transition at 211 nm, which corresponds to the twelfth ex-
cited state of INHOVA. This is the reason why two unoccupied
MO’s lying very high in energy are involved in the one-electron exci-
tations. The MO’s mainly involved in the electronic transitions used
to assign observed bands are depicted in Fig. 4. It can be seen from
the figure that both the HOMO and HOMO ꢀ 1 are mainly localized
on the o-HVa moiety of INHOVA. They present a
pbonding character
in the carbon atoms in the ring and a non-bonding character in the
oxygen atoms of OCH3 and OH groups. Moreover, the HOMO ꢀ 1 also
Fig. 4. Molecular orbitals involved in the electronic transitions of isomer 1 of
INHOVA. The energy scale is only qualitative and does not represent the actual
energy of the molecular orbitals.
shows some non-bonding contributions from nitrogen and oxygen
atoms from the C@NANACAO moiety. The LUMO is mainly localized
on the INH ring. It presents a
p bonding character in the carbon
atoms in the ring and a non-bonding character in nitrogen and oxy-
gen atoms. The other three MO’s involved in the assignment are
higher in energy and are mainly localized on the o-HVa moiety of
the molecule. The LUMO + 2 exhibits a non-bonding character in
some carbon atoms in the ring, the oxygen atom of the OH group
and the C@N group, besides a
LUMO + 7 and LUMO + 8 share a very similar localization pattern.
Both MO’s are anti-bonding in some carbon atoms in the ring of
p bonding between C11 and C12.
Fig. 3. Electronic spectra of 5 ꢂ 10ꢀ5 M methanolic solution of INHOVA. Experi-
mental (—) and calculated (——) spectra of isomer 1 are shown.
p

A.C. González-Baró et al. / Journal of Molecular Structure 1007 (2012) 95–101
101
o-HVa and are non-bonding in the oxygen atoms of OCH3 and OH
groups.
According to the above analysis, it can be argued that observed
bands at 204 and 222 nm correspond to transitions involving o-
HVa ? o-HVa fragment of INHOVA. On the other hand, the band
at 304 nm is clearly an o-HVa ? INH transition. Band at 336 is
mainly dominated by a o-HVa ? INH transition with minor contri-
bution from an o-HVa ? o-HVa transition.
References
[1] C.M. Nolan, S.V. Goldberg, Int. J. Tuberc. Lung Dis. 6 (2002) 952.
[2] S. Keshavjee, I.Y. Gelmanova, P.E. Farmer, S.P. Mishustin, A.K. Strelis, Y.G.
Andreev, A.D. Pasechnikov, S. Atwood, J.S. Mukherjee, M.L. Rich, J.J. Furin, E.A.
Nardell, J.Y. Kim, S.S. Shin, Lancet 372 (2008) 1403.
[3] A. Zubrys, C.O. Siebenmann, Can. J. Chem. 33 (1955) 11.
[4] R. Maccari, R. Ottanà, M.G. Vigorita, Bioorg. Med. Chem. Lett. 15 (2005) 2509.
[5] E.N. Deeb, G.R. Vitagliano, J. Am. Pharm. Assoc. 44 (1955) 182.
[6] E.F. Oga, Int. J. Pharm. Pharm. Sci. 2 (2010) 55.
[7] S. Santosh Kumar, K. Priyadarsini, K.B. Sainis, Redox Rep. 7 (2002) 35.
[8] WHO, Guidelines for the Programmatic Management of Drug-resistant
4. Conclusions
Tuberculosis:
Emergency
Update,
2008
<http://whqlibdoc.who.int/
publications/2008/9789241547581_eng.pdf> (25.09.08) (accessed 17.03.09).
[9] B. Bottari, R. Maccari, F. Monforte, R. Ottana, E. Rotondo, M.G. Vigorita, Bioorg.
Med. Chem. Lett. 10 (2000) 657–660.
[10] S. Hingorani, B.V. Agarwala, Trans. Metal Chem. 18 (1993) 576.
[11] Shao-Wen Chen, Han-Dong Yin, Da-Qi Wang, Xia Kong, Xiao-Fang Chen, Acta
Cryst. E62 (2006) o2043.
[12] H.E. Gottlieb, V. Kotlyar, A. Nudelman, J. Org. Chem. 62 (1997) 7512.
[13] J.J.P. Stewart, J. Comput. Chem. 10 (1989) 209.
[14] Y. Zhao, D.G. Truhlar, Theor. Chem. Acc. 120 (2008) 215.
[15] R. Krishnan, J.S. Binkley, R. Seeger, J.A. Pople, J. Chem. Phys. 72 (1980) 650.
[16] T. Clark, J. Chandrasekhar, P.V.R. Schleyer, J. Comput. Chem. 4 (1983) 294.
[17] V. Barone, M. Cossi, J. Phys. Chem. A 102 (1998) 1995.
[18] M. Cossi, N. Rega, G. Scalmani, V. Barone, J. Comput. Chem. 24 (2003) 669.
[19] A. Dreuw, M. Head-Gordon, Chem. Rev. 105 (2005) 4009.
[20] P. Elliott, F. Furche, K. Burke, Rev. Comp. Chem. 26 (2009) 91.
[21] GAMESS-US version 1 October 2010 (R1). M.W. Schmidt, K.K. Baldridge, J.A.
Boatz, S.T. Elbert, M.S. Gordon, J.H. Jensen, S. Koseki, N. Matsunaga, K.A.
Nguyen, S.J. Su, T.L. Windus, M. Dupuis, J.A. Montgomery, J. Comput. Chem. 14
(1993) 1347.
[22] J.R. Cheeseman, G.W. Trucks, T.A. Keith, M.J. Frisch, J. Chem. Phys. 104 (1996)
5497.
[23] R. Ditchfield, Mol. Phys. 27 (1974) 789.
[24] K. Wolinski, J.F. Hinton, P.J. Pulay, Am. Chem. Soc. 112 (1990) 8251.
[25] J. Tomasi, B. Mennucci, E. Cancès, J. Mol. Struct. THEOCHEM 464 (1999) 211.
[26] Gaussian 03, Revision D.01, M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E.
Scuseria, M.A. Robb, J.R. Cheeseman, J.A. Montgomery, Jr., T. Vreven, K.N.
Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V. Barone, B. Mennucci,
M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H. Nakatsuji, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H.
Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian, J.B. Cross, V. Bakken, C. Adamo,
J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C.
Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J.
Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels, M.C. Strain, O. Farkas,
D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G.
Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A.
Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong,
C. Gonzalez, J.A. Pople, Gaussian, Inc., Wallingford CT, 2004.
The INHOVA Schiff base was obtained in good yield by a con-
densation reaction between o-HVa and INH. The conformational
search of the compound led to two stable isomers. The geometric
features of the most stable isomer in solution agrees very well with
crystallographic data, suggesting that crystallization from metha-
nolic solutions takes place with small structural changes. Charac-
teristic IR and Raman bands of Schiff bases denote the hydrazone
formation and the NAH stretching mode of protonated pyridinic
ring is consistent with the cationic species of INHOVA, present in
the solid state.
1H and 13C NMR spectroscopy clearly shows the presence of the
C@N hydrazonic linkage, in accordance to vibrational spectroscopy.
De-shielding of the H2, H6, H3, and H5 nuclei, when compared to
the values reported in literature, suggest that protonation of the
pyridinic ring is maintained upon dissolution in DMSO. 13C NMR
spectrum of INHOVA is reported here for the first time. Theoretical
values for the chemical shifts obtained through the GIAO method-
ology are in good agreement with experiment.
The electronic spectrum of INHOVA shows three bands and a
shoulder in the range from 200 nm to 400 nm. Calculated transitions
allow the assignment of the two bands at 204 nm and 222 nm as
o-HVa ? o-HVa transitions. The band at 304 nm and the shoulder
at 336 nm are assigned mainly as o-HVa ? INH transitions.
Acknowledgments
This work has been supported by CONICET (CCT-La Plata),
National University of La Plata, Argentina, and PUC-Rio of Brazil.
A.C.G.B., R.P.D., and B.S.P.C. are members of the Research Career
of CONICET. The authors thank Prof. Maria Isabel Pais da Silva
and M.Sc. Rosana Garrido Gomes (Department of Chemistry,
PUC-Rio) for NMR facilities.
[27] A.R. Allouche, J. Comput. Chem. 32 (2011) 174.
[28] A.C. González-Baró, R. Pis-Diez, C.A. Franca, M.H. Torre, B.S. Parajón-Costa,
Polyhedron 29 (2010) 959.
[29] E. Akalin, S. Akyuz, J. Mol. Struct. 834–836 (2007) 492.