Synthesis, spectroscopic and theoretical study of isoquinoline alkaloid dehydrosalsolinol base and its triacetylated derivative

Article abstract of DOI:10.1016/j.molstruc.2007.07.035

Dehydrosalsolinol and its triacetylated derivative i.e. are synthesized, isolated, and the spectra and structure are elucidated. Solid-state linear-dichroic infrared spectroscopy, ¹H and ¹³C NMR, TGV and DTA methods are applied. The potential energy surface of both compounds are explored with DFT correlation functional B3LYP method and MP2, using the 6-311++G^** basis set. Single point calculations are performed at levels up to B3LYP/6-311++G^**// MP2/6-311++G^** and MP2/6-311++G^**. Two isomers of dehydrosalsolinol are located and the relative energies are determined. The planar quinoidal structure with strong intramolecular OH...O hydrogen bond and dipole moment of 7.392 D is more stable than the diphenolic form. Calculated vibrational frequencies are used to determine the type of molecular motions associated with each of the experimental IR-bands observed. IR-spectroscopic data show that dehydrosalsolinol exists in solid-state as the quinoide form in contrast to solution, where ¹H NMR data indicate a keto-diphenolic equilibrium. Comparison between calculated and experimental IR-frequencies of triacetyl derivative shows the presence of two phenolic ester carbonyl and one-amide groups and an exocyclic double bond. The same conclusions are obtained by corresponding ¹H NMR data.

Full text of DOI:10.1016/j.molstruc.2007.07.035

Available online at www.sciencedirect.com
Journal of Molecular Structure 878 (2008) 1–9
www.elsevier.com/locate/molstruc
Synthesis, spectroscopic and theoretical study of isoquinoline alkaloid
dehydrosalsolinol base and its triacetylated derivative
a,b,
a
Tsonko M. Kolev
*, Plamen Angelov
a
Plovdiv University, Department of Organic Chemistry, 24 Tzar Assen Street, Plovdiv 4000, Bulgaria
Institute of Organic Chemistry, Bulgarian Academy of Sciences, Acad. G. Bonchev Str. Build. 9, 1113 Sofia, Bulgaria
b
Received 14 December 2006; received in revised form 20 June 2007; accepted 24 July 2007
Available online 3 August 2007
Abstract
Dehydrosalsolinol and its triacetylated derivative i.e. are synthesized, isolated, and the spectra and structure are elucidated. Solid-state
linear-dichroic infrared spectroscopy, 1H and 13C NMR, TGV and DTA methods are applied. The potential energy surface of both com-
pounds are explored with DFT correlation functional B3LYP method and MP2, using the 6-311++G** basis set. Single point calcula-
tions are performed at levels up to B3LYP/6-311++G**// MP2/6-311++G** and MP2/6-311++G**. Two isomers of dehydrosalsolinol
are located and the relative energies are determined. The planar quinoidal structure with strong intramolecular OH. . .O hydrogen bond
and dipole moment of 7.392 D is more stable than the diphenolic form. Calculated vibrational frequencies are used to determine the type
of molecular motions associated with each of the experimental IR-bands observed. IR-spectroscopic data show that dehydrosalsolinol
1
exists in solid-state as the quinoide form in contrast to solution, where H NMR data indicate a keto-diphenolic equilibrium. Compar-
ison between calculated and experimental IR-frequencies of triacetyl derivative shows the presence of two phenolic ester carbonyl and
1
one-amide groups and an exocyclic double bond. The same conclusions are obtained by corresponding H NMR data.
ꢀ 2008 Published by Elsevier B.V.
Keywords: Dehydrosalsolinol; Solid-state IR–LD spectroscopy; Ab initio and DFT calculations; Vibrational analysis; 1H NMR; TGA and DSC
1. Introduction
bromide salt [3]. The approach presented herein is
experimentally simple and gives the free base 4 directly.
It has been demonstrated earlier that 3,4-dimethoxyphenyl-
ethyl amine reacts readily with 1,3-dicarbonyl compounds
to give enaminones which can later cyclize to 1,1-disubsti-
tuted tetrahydroisoquinolines in acidic media [4]. Extend-
ing these studies to different phenylethyl amines, it is not
surprising that dopamine also gives enaminone 2 upon
reaction with ethyl acetoacetate (Scheme 1).
The cyclization of 2, however, does not require acidic
catalysis and proceeds under solvent-free conditions at
80–100 ꢁC. The tetrahydroisoquinoline 3 formed in this
way is not isolated as it spontaneously eliminates ethyl ace-
tate in a retro-Mannich manner and gives dehydrosalsolin-
ol (4) in 60% overall yield (Scheme 2).
Dehydrosalsolinol (4) as a metabolite of the salsolinol-
1-carboxylic acid [1], which is a central nervous system
alkaloid, might be related to chronic alcohol consumption
and Parkinson’s disease [2,3], when it is present in the
mammalian brain. This, along with the interesting struc-
tural features, prompted us to study a new approach for
its synthesis and to carry out some computational struc-
tural investigations.
Dehydrosalsolinol has previously been synthesized by
demethylation of 3,4-dihydro-6,7-dimethoxy-1-methyl iso-
quinoline, which yields the target compound as its hydro-
*
Corresponding author. Address: Institute of Organic Chemistry,
Present research is focused on the spectroscopic and the-
oretical characterization of 4 and 5 (Scheme 2) by means of
solid-state linear-polarized IR-spectroscopy of oriented
Bulgarian Academy of Sciences, Acad. G. Bonchev Str. Build. 9, 1113
Sofia, Bulgaria. Tel.: +359 2 9606 106.
E-mail address: kolev@orgchm.bas.bg (T.M. Kolev).
0022-2860/$ - see front matter ꢀ 2008 Published by Elsevier B.V.
doi:10.1016/j.molstruc.2007.07.035

2
T.M. Kolev, P. Angelov / Journal of Molecular Structure 878 (2008) 1–9
Et3N
HO
HO
HO
HO
CH3COCH2COOEt
NH2.HCl
NH
O
1
2
H3C
OEt
Scheme 1. Synthesis scheme.
HO
HO
HO
HO
HO
o
100 C
N
N
NH
O
H
HO
- CH3COOEt
H3C
H3C
O
CH3
OEt
4
OEt
O
3
2
NH
CH3
HO
o
Ac2O, 100 C, 15 min
O
O
O
O
N
O
CH2
5
Scheme 2. Synthesis scheme.
1
samples as suspension in nematic liquid crystal, H NMR,
and quantum chemical calculations at ab initio (MP2)
and DFT (B3LYP) levels of theory and the 6-311++G**
basis set.
down to r.t. and the solidified mass was triturated with
Et2O:MeOH 2:1 (v:v) to give 4 as a yellow powder
(0.106 g, 60%). Elemental analysis (Found: C, 68.15; H,
5.75; N, 7.96; [C10H11NO2] calcd.: C, 68.13; H, 5.79; N,
7.93%). M.p. 232 (with decomposition). 1H NMR
(DMSO-d6, d ppm): 6.81 (s, 1H), 6.25 (s, 1H), 3.41 (t, 2H,
J = 7.7), 2.54 (t, 2H, J = 7.7), 2.23 (s, 3H).
2. Experimental
Unless otherwise noted, all the reagents and solvents
were purchased from Sigma–Aldrich and were used as
commercially supplied.
2.1.3. Acetic acid 7-acetoxy-2-acetyl-1-methylene-1,2,3,4-
tetrahydro-isoquinolin-6-yl ester (5)
Acetic anhydride (0.6 ml) was added to dehydrosalsolin-
ol (0.1000 g, 0.565 mmol) and the mixture was heated at
100 ꢁC for 15 min. Then the mixture was allowed to cool
down to r.t. and with vigorous magnetic stirring, water
(6 ml) was added to hydrolyze the excess acetic anhydride.
NaHCO3 was then carefully added to adjust the pH to 7–8,
and the mixture was extracted with DCM (2· 50 ml). The
organic layers were dried (Na2SO4) and dry-loaded onto
silica. Column chromatography in Et2O afforded the prod-
uct (5) as colourless crystals (0.130 g, 76%). Elemental
analysis (Found: C, 63.36; H, 5.66; N, 4.64; [C16H17NO5]
calcd.: C, 63.36; H, 5.65; N, 4.62%.). M.p. 123 (with
2.1. Synthesis
2.1.1. Synthesis of the enaminone (2)
To a suspension of dopamine hydrochloride (0.5689 g,
3 mmol) in methanol (7 ml), triethylamine (0.3036 g,
3 mmol) was added, followed by ethylacetoacetate
(0.3904 g, 3 mmol). The mixture was stirred at r.t. for
24 h. Then water (100 ml) was added and extraction with
dichloromethane (2· 30 ml) was performed. The organic
layer was dried with Na2SO4 and the solvent was distilled
to yield 2 as a clean to slightly brown oil (0.5870 g, 73%).
1
decomposition). H NMR (CDCl3, d ppm): 7.55 (s, 1H),
2.1.2. Synthesis of dehydrosalsolinol (4)
The enaminone 2 (0.2653 g, 1 mmol) was heated at
100 ꢁC for 4 h in a sealed vial. It was then allowed to cool
6.98 (s, 1H), 5.66 (s, 1H), 5.08 (s, 1H), 3.96 (t, 2H, J
= 6.0) 2.87 (t, 2H, J = 6.0), 2.30 (s, 3H), 2.28 (s, 3H),
2.20 (s, 3H).

T.M. Kolev, P. Angelov / Journal of Molecular Structure 878 (2008) 1–9
3
The most intensive signals in the mass spectra of the
compounds studied are that of the peak at m/z 178.01
and 304.11, corresponding to the singly charged cations
[C10H12NO2]+ and [C16H18NO5]+, with molecular weights
of 178.24 and 304.16, respectively. These findings confirm
the synthesis and isolation of 4 and 5.
2.2. Materials and methods
IR-spectroscopic measurements within 4000–400 cmÀ1
range were obtained on a Bruker 113v FT-IR spectrometer
(resolution 2 cmÀ1, 250 scans). A Specac wire-grid pola-
rizer was used. Oriented solid samples were obtained as a
suspension in a nematic liquid crystal of the 4’-cyano-4’-
alkylbicyclohexyl type (ZLI 1695, Merck), mesomorphic
at r.t. Its weak IR-spectrum permits recording of the
guest-compound IR-bands in the whole 4000–400 cmÀ1
range. The isolated nitrile stretching IR-band at
2236 cmÀ1 serves additionally as an orientation indicator.
The effective orientation of the samples was achieved
through the following procedure: 5 mg of the compound
to be studied was mixed with the liquid crystal substance
until a slightly viscous suspension was obtained. The phase
thus prepared was pressed between two KBr-plates for
which, in advance, one direction had been rubbed out by
means of fine sandpaper. The grinding of the mull in the
rubbing direction promotes an additional orientation of
the sample. The validation of this new orientation solid-
state method [5] used in linear-dichroic infrared (IR–LD)
spectroscopy for accuracy, precision and the influence of
the liquid crystal medium on peak positions and integral
absorbances of the guest molecule bands, as well as the
optimization of experimental conditions and an experimen-
tal design for quantitative evaluation of the impact of four
input factors, have been presented [6,7]. The number of
scans, the rubbing-out of KBr-pellets, the amount of stud-
ied compounds included in the liquid crystal medium and
the ratios of Lorentzian to Gaussian peak functions in
the curve-fitting procedure on the spectroscopic signal at
five different frequencies has been studied [6,7]. It has been
found that the procedure for the position (mi) and integral
absorbancies (Ai) determination for each i-peak have been
carried out by deconvolution and curve-fitting procedures
at 50:50% ratio of Lorentzian to Gaussian peak functions,
v2 factors within 0.00066–0.00019 and 2000 iterations [6,7].
The means of the two treatments were compared by Stu-
dent t-test. The experimental IR-spectral patterns have
been acquired and processed by GRAMS/AI 7.01 IR spec-
troscopy (Thermo Galactic, USA) and STATISTICA for
Windows 5.0 (StatSoft, Inc., Tulsa, OK, USA) program
packages. The nature and balance of the forces in the
nematic liquid crystal suspension system, the mathematical
model for their clearance, morphology of the suspended
particles and the influence of the space system types on
the degree of orientation i.e. ordering parameter have been
shown [8] using five liquid crystal and 15 compounds.
Applicability of the last approach for experimental
Scheme 3. Optimized geometry of (4).
Table 1
˚
Selected geometrical parameters (bond lengths [A] and angles [ꢁ]) of 4 and
5, obtained at the MP2/6-311++G** level of theoretical approximation,
using atom numbering Schemes 3 and 4
Name definition
Values
Name definition
Values
4
R(1,2)
R(1,3)
R(1,4)
R(2,5)
R(6,7)
R(7,10)
R(8,12)
R(10,11)
R(11,12)
1.482
1.250
1.460
1.384
1.522
1.537
1.387
1.476
1.364
A(2,1,3)
A(3,1,4)
A(1,2,9)
A(5,2,9)
A(1,4,6)
A(4,6,7)
A(4,6,8)
A(6,7,10)
A(6,8,12)
A(9,8,12)
D(7,6,8,12)
D(6,8,9,2)
D(6,8,12,11)
D(9,8,12,13)
D(7,10,11,12)
D(10,11,12,8)
122.6(5)
123.3(7)
122.0(8)
124.6(7)
124.0(3)
121.9(7)
120.5(4)
110.1(4)
119.8(1)
122.9(3)
5.7
D(3,1,2,5)
D(4,1,2,9)
D(2,1,4,6)
D(1,2,9,8)
D(1,4,6,8)
D(4,6,7,10)
D(8,6,7,10)
D(4,6,8,9)
0.0
0.7
1.2
0.6
1.6
9.2
6. 4
41.8
0.3
148.5
36.8
1.1
10.6
5
R(1,2)
R(1,9)
R(3,6)
R(4,8)
R(5,15)
R(8,18)
R(8,20)
1.391
1.396
1.388
1.396
1.396
1.510
1.222
A(1,2,4)
A(3,2,4)
A(2,3,5)
A(6,7,9)
A(4,8,18)
A(4,8,20)
A(18,8,20)
A(1,9,10)
125.7(5)
114.7(8)
114.4(2)
118.9(3)
107.5(2)
124.7(6)
127.7(1)
122.0(8)
107.8(1)
107.5(0)
5.6
153.7
28.2
139.2
61.2
141.3
2.4
5.0
2.5
A(9,10,11)
A(5,15,21)
D(14,7,9,10)
D(6,7,14,12)
D(9,7,14,12)
D(1,9,10,11)
D(9,10,11,12)
D(10,11,12,13)
D(11,12,13,19)
D(14,12,13,17)
D(11,12,14,7)
D(13,12,14,16)
D(9,1,2,3)
D(2,1,9,7)
D(1,2,3,6)
D(4,2,3,5)
D(1,2,4,8)
D(6,3,5,15)
D(2,3,6,7)
D(2,4,8,20)
D(3,6,7,9)
D(6,7,9,10)
0.9
0.0
0.5
0.5
0.7
1.0
0.7
0.1
1.6
176.3
7.1

4
T.M. Kolev, P. Angelov / Journal of Molecular Structure 878 (2008) 1–9
IR-spectroscopic band assignment and especially for
obtaining structural information, has been demonstrated
in a series of organic systems and metal complexes [9],
Cu(II) complexes [10], polymorphs [11], codeine derivatives
[12], peptides and their Au(III) complexes, and hydrochlo-
rides and hydrogensquarates [13–15]. It has been found
that enough (15–20%) for reasonable linear-polarized IR-
spectra orientation is achieved when 5 1 wt% of the com-
pound with particle size within 0.3–0.9 lm are mixed with
appropriate for IR-spectroscopy nematic liquid crystal sub-
stance. The KBr-plates, which are in advance in one direc-
tion, are rubbed out with size 5 lm. The KBr-plates are
moved repeatedly with 3 lm/s for 100 times. The optimal
cell thickness is 100 lm. The theoretical backgrounds of
IR–LD spectroscopy are described in [16–19].
6-311++G** basis set. In the second case, the B3LYP
method, which combines Becke’s three-parameter non-
local exchange functional with the correlation function of
Lee et al. [22,23], is applied. Molecular geometries of the
studied species were fully optimized by the force gradient
method using Bernys’ algorithm [24]. For every structure,
the stationary points found on the molecule potential
energy hypersurfaces were characterized using standard
analytical harmonic vibrational analysis. Absence of the
imaginary frequencies and negative eigenvalues of the sec-
ond-derivative matrix confirmed that the stationary points
correspond to minima of the potential energy hypersurfac-
es. The calculation of vibrational frequencies and infrared
intensities were checked to see which kind of calculation
performed agreed best with the experimental data. In cases
of 4 and 5, the DFT method provided more accurate vibra-
tional data, as the calculated standard deviations are
9 cmÀ1 (B3LYP) and 21 cmÀ1 (MP2), respectively.
B3LYP/6-311++G** data are presented for the above dis-
cussed modes, where for better correspondence between the
experimental and theoretical values, a modification of the
results using the empirical scaling factor 0.9614 [25] was
made. In the case of MP2 calculations, the scaling factor
of 0.8929 was used.
The elemental analysis was carried out according to the
standard procedures for C and H (as CO2 and H2O) and N
(by the Dumas method).
The thermogravimetric and calorimetry study were per-
formed on a Stanton Redcroft 1500 under argon with heat-
ing at 10 ꢁC/min.
1H measurements were made at 298 K with a Bruker
DRX-250 spectrometer using CDCl3 or DMSO-d6 as the
solvent.
The FAB mass spectra were recorded on a Fisons VG
autospect instrument employing 3-nitrobenzylalcohol
(Sigma–Aldrich) as the matrix.
3. Results and discussion
Quantum chemical calculations were performed using
GAUSSIAN 98 program package [20] and the output fails
were visualized by means of ChemCraft program [21].
Geometries of 4 and 5 (Scheme 2) were optimized at two
levels of theory: second-order Moller–Pleset perturbation
theory (MP2) and density functional theory (DFT) using
3.1. Theoretical data
3.1.1. Molecule structures
According to quantum chemical calculations at the
MP2 and DFT levels of theory, two distinct stable con-
formers of 4 exist, differing by internal rotation around
Scheme 4. Optimized geometry of (5).

T.M. Kolev, P. Angelov / Journal of Molecular Structure 878 (2008) 1–9
5
the C2–O5 bond (Scheme 3) with C1–C2–O5–H15 of
0.1ꢁ and 177.6ꢁ (MP2/6-311++G**), respectively. When
energies are corrected by zero point vibrational energy
contribution, the first conformer is more stable with
4 kJ/mol. The geometry is characterized with an intramo-
aim, we are focused on the values obtained by MP2/6-
311++G**. Molecule 4 is absolutely flat excepting both
CH2 groups, which are deviated from co-planarity at
angles of 4.3ꢁ and 13.9ꢁ, respectively (Scheme 3).
The calculated geometrical parameters of 5 are shown in
Table 1. Similar to 4, the geometry of 5 is flat excepting
both CH2 groups disposed at angles of 2.6ꢁ and 5.6ꢁ
towards the plane of the molecule (Scheme 4). Carboxylic
groups are co-planar with the plane of the benzene ring
with the maximum deviation of 0.6ꢁ. The amide C@O
and C@C double bonds are co-linear, disposed (5.1ꢁ),
and lie in a plane supposing the co-linearity of correspond-
ing stretching mC=O (Amide I) and mC=C modes.
˚
lecular OH. . .O hydrogen bond with length of 2.781 A
and OHO angle of 111.2(6)ꢁ, respectively. Predicted geo-
metrical parameters are summarized in Table 1. The
comparison between MP2 and DFT data shows that
the ab initio approach gives better structural parameters
[22,23] with a difference of 9% in some cases. For the
purpose of the next IR/LD spectral analysis, where the
geometrical prediction of compounds studied is a main
(2)
(1)
1800
1600
1400
1200
1000
800
600
Absorbance / Wavenumber (cm-1)
(2)
(1)
1600
1400
1200
1000
800
600
400
Absorbance / Wavenumber (cm-1)
Fig. 1. Non-polarized IR-(1) and difference IR–LD (2) spectra of 4 (down) and 5 (up) in nematic liquid crystal suspension.

6
T.M. Kolev, P. Angelov / Journal of Molecular Structure 878 (2008) 1–9
3.1.2. Vibrational analysis
pearance of characteristic maxima for 1,2,4,5-
tetrasubstituted benzene and the observation of typical
peaks for conjugated double bonds is also obtained. The
maxima at 955 and 944 cmÀ1 correspond to out-of-plane
(o.p.) frequencies of trisubstituted double bonds in the
molecule of 4.
A comparison between calculated gas-phase and exper-
imental IR-frequencies in solid-state are made for 4 and
5. The calculated IR-spectrum of 4 is characterized with
the highest absorption maximum at 3622 cmÀ1 assigned
to stretching mode of OH group (mOH) included in intramo-
lecular hydrogen bonding. The band at 3100 cmÀ1 corre-
sponds to mNH mode. The untypical low-frequency value
could be explained with the conjugation of the unpaired
electronic couple with quinoide structure in the molecule
of 4. The IR-bands within 3500–2500 cmÀ1 correspond to
symmetric and asymmetric stretching modes of CH, CH2
and CH3 groups. The 1650–1500 cmÀ1 IR-spectroscopic
region shows a series of bands with mixed character start-
ing with the 1630 cmÀ1 peak with predominant mC=O char-
acter. The bands at 1609 and 1510 cmÀ1 correspond to
mC=C stretch, while the band at 1457 cmÀ1 to stretching
mC=C and bending dNH modes. Low-frequency shifting of
the mC=O shows a stabilization of the quinoide structure
in solid-state, due to a conjugation with significant single
charge redistribution and distortion of the aromatic char-
acter of the phenyl ring in enolic form (Scheme 2). Disap-
The highest frequency peak in the predicted IR-spec-
trum of 5 at 3211 cmÀ1 is assigned to the mas mode of
CH2
the double bond. The corresponding msCH maximum is pre-
2
dicted at 3002 cmÀ1. The maxima within 3500–2500 cmÀ1
IR-spectroscopic region correspond to asymmetric and
symmetric CH2 and CH3 stretching frequencies, as well
as the in-plane (i.p.) vibrations of the aromatic fragment
(pseudo 1,2,4,5-tetrasubstituted phenyl [26]). The peaks at
1772 and 1768 cmÀ1 are assigned to symmetric and asym-
metric stretching vibrations of the carboxylic C@O groups,
while the peaks at 1662 and 1650 cmÀ1 correspond to
Amide I, and mC=C corresponds to a peak at 1650 cmÀ1
.
The band at 980 cmÀ1 corresponds to the o.p. mode of
disubstituted C@C bond. Since the molecule structure
excludes the strong intermolecular interactions due to the
absence of protondonation groups, excellent correlation
(2)
(1)
1600
1400
1200
1000
800
600
Absorbance / Wavenumber (cm-1)
Fig. 2. Non-polarized IR-(1) and reduced IR–LD (2) spectra of 4 after elimination of the peak at 1600 cmÀ1
.

T.M. Kolev, P. Angelov / Journal of Molecular Structure 878 (2008) 1–9
7
between the calculated and experimentally obtained data
are established using the B3LYP/6-311++G** approach.
The calculated standard deviation in the data of 5 is
9 cmÀ1. It is important to note that the last value is
obtained in the cases of the molecular motion connected
with the fragments, where intramolecular or intermolecular
hydrogen bonding is absent. When the MP2/6-311++G**
approach is used, the obtained difference of 23% between
both theoretical methods is connected namely with the last
discussed modes.
3.2. Experimental conventional and linear-polarized IR-data
Experimental IR-spectroscopic patterns of 4 and 5 are
preliminary deconvolution and curve fitting. The number
of peaks and their positions are obtained. The experimental
confirmation of IR-characteristic band assignment is
achieved by means of linear-polarized IR-spectroscopic
spectra and the application of reducing-difference
procedure.
The difference IR–LD spectra (Fig. 1) of 4 and 5 show a
good orientation of the crystalline samples in the liquid
crystal medium, resulting to adequate conclusions of the
data obtained by the IR–LD spectra interpretation, using
the criteria of [6–8].
The application of the reducing-difference procedure to
polarized IR–LD spectra of both compounds resulted in
the following data:
1640
1630
1620
1610
1600
1590
1580
1570
Absorbance / Wavenumber (cm-1)
Fig. 3. Non- and deconvoluted (c factor of 3.8, increment of 1) IR-
spectroscopic pattern of 4 in 1650–1550 cmÀ1 region.
(2)
(1)
1600
1400
1200
1000
800
600
400
Absorbance / Wavenumber (cm-1)
Fig. 4. Non-polarized IR-(1) and reduced IR–LD (2) spectra of 4 after elimination of the peak at 1600 cmÀ1
.

8
T.M. Kolev, P. Angelov / Journal of Molecular Structure 878 (2008) 1–9
105
5
4: It can be seen the co-linear disposition of transition
moments of mC=O and mC=C of both the trisubstituted
double bonds suppose an elimination of corresponding
maxima at 1625, 1600, 1504 and 1459 cmÀ1 in equal
dichroic ratio, which is obtained experimentally in Fig. 2.
In parallel an elimination of 927, 933 and 466 cmÀ1 peaks
belonging to o.p. cCH of disubstituted C@C and OH
groups is also obtained. As far as in the frame of one
molecule is impossible to be obtained the elimination of
i.p. and o.p. maxima in the same dichroic ratio, this result
assumes the presence of pairs of molecules in the unit cell
of compound studied mutual near to perpendicular
oriented. Similar data are observed in the monoclinic
polymorphs of paracetamol [11] with space group P21/n
and Z = 4. The presence of two co-planar by pairs mole-
cules in the unit cell, oriented at an angles about 90ꢁ leads
to a simultaneous elimination of i.p. and o.p. modes of
unsaturated structural fragments in molecules. The proce-
dure gives rise to a second peak at 1631 cmÀ1 (deconvo-
luted IR-spectroscopic pattern is given in Fig. 3), which
100
95
90
85
80
75
70
0
-5
wt. loss:
°
-10
-15
-20
from252 to 303 C -8.25%
from303 to 345 C -13.12%
°
272.1
100
200
Temperature, o
300
400
C
Fig. 6. TGV and DTA data of dehydrosalsolinol.
analysis shows the same i.p. symmetry class of both elimi-
nated maxima.
5: The data of the obtained vibrational analysis suppose
a co-linear disposition of transition moments of Amide I
and mC=C in the molecule of 5, which explains the obtained
simultaneous elimination of the peaks at 1656 and
1644 cmÀ1 (Fig. 5(2)). On the other hand, the obtained
reduction of the peaks at 923, 840 and 677 cmÀ1, are all
predicted as o.p. bending modes of 1,2,4,5-tetrasubstituted
benzene [26].
reduction does not affect the IR-bands in 1700–400 cmÀ1
.
Only the disappearance of the peak at 1377 cmÀ1 is
obtained (Fig. 4), supposing a combination character of
the band at 1631 cmÀ1 (1377 + 254). Moreover, vibrational
3.3. TGV and DTA
The TGV and DTA data of dehydrosalsolinol are
presented in Fig. 6. As can be seen in the 100–400 ꢁC
temperature range, a weight loss of 8.25% is established
which corresponds to an effect of –19.2 lV at 272.1 ꢁC.
The next weight loss of 13.12% between 303 and 345 ꢁC
can be explained with decomposition of the compound
studied. The high melting point of dehydrosalsolinol is in
accordance with significant dipole moment of 7.392 D
derived from quantum chemical calculations and the pres-
ence of strong intramolecular hydrogen bond in the
molecule.
(2)
The complete acetylation of this compound leads to a
product which has a higher molecular weight, but lack of
hydrogen bond, which results in a significantly lower melt-
ing point of 122 ꢁC and dipole moment of 5.085 D.
4. Conclusion
Dehydrosalsolinol and its completely acetylated product
were synthesized, and theoretically and experimentally elu-
cidated. On the basis of optical and magnetic spectral and
theoretical data, the following essential conclusions were
made: (i) the molecule of dehydrosalsolinol exists in
solid-state as quinoidal form; (ii) in the solution, a keto-
enolic equilibrium is found; (iii) the predicted electronic
structures of the keto form of dehydrosalsolinol and its
completely acetylated product are planar in the frame of
the conjugate systems; (iv) it is proven that during the
process of the acetylation, an isomerization takes place,
(1)
1680
1660
1640
1620
1600
1580
Absorbance / Wavenumber (cm-1
)
Fig. 5. Non-polarized IR-(1) and reduced IR–LD (2) spectra of 5 after
elimination of the peak at 1644 cmÀ1
.

T.M. Kolev, P. Angelov / Journal of Molecular Structure 878 (2008) 1–9
[13] B.B. Ivanova, J. Mol. Struct. 782 (2006) 122.
9
leading to the formation of exocyclic double bound. This
was proven unambiguously by means of theoretical and
experimental data; (v) the theoretical vibrational and
polarized IR-spectroscopy of dehydrosalsolinol show a
mixed character of vibrational modes of the quinoline part
of the molecule; (vi) the existence of strong intramolecular
hydrogen bonds between the OH and O@C groups in the
keto form of dehydrosalsolinol, established by IR-spectros-
copy, is confirmed by non-empirical quantum chemical
calculations.
[14] B.B. Ivanova, T. Kolev, S.Y. Zareva, Biopolymers 82 (2006) 587.
[15] Ts. Kolev, Biopolymers 83 (2006) 39.
[16] B. Jordanov, B. Schrader, J. Mol. Struct. 347 (1995) 389.
[17] B. Jordanov, R. Nentchovska, B. Schrader, J. Mol. Struct. 297 (1993)
292.
[18] J. Michl, E.W. Thulstrup, Spectroscopy with Polarized Light. Solute
Alignment by Photoselection, in Liquid Crystals, Polymers, and
Membranes, VCH Publishers, NY, 1986.
[19] E.W. Thulstrup, J.H. Eggers, Chem. Phys. Lett. 1 (1996) 690.
[20] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb,
J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery Jr., R.E.
Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels,
¨
K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi,
References
R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J.
Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, P.
Salvador, J.J. Dannenberg, D.K. Malick, A.D. Rabuck, K. Raghav-
achari, J.B. Foresman, J. Cioslowski, J.V. Ortiz, A.G. Baboul, B.B.
Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Koma´romi, R.
Gomperts, 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, J.L. Andres, C. Gonzalez, M. Head-Gordon,
E.S. Replogle, J.A. Pople, Gaussian 98, Gaussian, Inc., Pittsburgh,
PA, 1998.
[1] P. Dostert, M. Strolin Benedetti, V. Bellotti, C. Allievi, G. Dordain, J.
Neural Transm. General Section 81 (3) (1990) 215.
[2] P. Dostert, M. Strolin Benedetti, F. Della Vedova, C. Allievi, R. La
Croix, G. Dordain, D. Vernay, F. Durif, Adv. Neurol. 60 (1993) 218.
[3] F. Zhang, G. Dryhurst, J. Med. Chem. 36 (1993) 11.
[4] A.P. Venkov, P.A. Angelov, Synthetic Commun. 33 (2003) 3025.
[5] B.B. Ivanova, M.G. Arnaudov, P.R. Bontchev, Spectrochim. Acta
60A (2004) 855.
[6] B.B. Ivanova, D.L. Tsalev, M.G. Arnaudov, Talanta 69 (2006) 822.
[7] B.B. Ivanova, V.D. Simeonov, M.G. Arnaudov, D.L. Tsalev,
Spectrochim. Acta A, in press.
[21] G.A. Zhurko, D.A. Zhurko, ChemCraft: Tool for treatment of
chemical data, Lite version build 08 (freeware) (2005).
[22] D. Backe, J. Chem. Phys. 98 (1993) 5648.
[23] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B37 (1988) 785.
[24] C. Peng, P.Y. Ayala, H.B. Schlegel, M.J. Frisch, J. Comp. Chem. 17
(1996) 49.
[25] A.P. Scott, L.J. Radom, J. Phys. Chem. 100 (1996) 16502.
[26] G. Varsanyi, Vibrational Spectra of Benzene Derivatives, Academy
Press, Budapest, 1969, pp. 1–413.
[8] B.B. Koleva, Ts. Kolev, V. Simeonov, T. Spassov, M. Spiteller, J.
Colloid Interface Sci., in preparation.
[9] B.B. Ivanova, Spectrochim. Acta A 62 (1–3) (2005) 5.
[10] B.B. Ivanova, H. Mayer-Figge, J. Coord. Chem. 58 (2005) 653.
[11] B.B. Ivanova, J. Mol. Struct. 738 (1–5) (2005) 233.
[12] R. Bakalska, B.B. Ivanova, Ts. Kolev, Cent. Eur. J. Chem. 4 (3)
(2006) 533.