Spectrochimical, ab initio and density functional studies on the conversion of 2-hydroxybenzonitrile (o-cyanophenol) into the oxyanion

Article abstract of DOI:10.1016/j.saa.2003.12.040

The spectral and structural changes caused by the conversion of 2-hydroxybenzonitrile (o-cyanophenol) into the corresponding oxyanion have been followed by IR spectra, ab initio and density functional force field calculations. In agreement between theory and experiment, the conversion is accompained by a 29 cm^-1 frequency decrease of the cyano stretching band, 2.7-fold increase in its integrated intensity, 5.8-fold (total value) intensification of the aromatic skeletal bands of Wilson's 8 and 19 types, and other essential spectral changes. According to the calculations, the strongest structural changes are the shortening of the Ph-O bond with 0.10 Angstroem, lengthenings of the adjacent CC bonds in the phenylene ring with 0.06 Angstroem and bond angle variations near the oxyanionic center. All these changes are connected with the formation of a quasi-ortho-quinonoidal structure of the o-phenylene ring in the oxyanion. According to the electronic density analysis, 0.41 e^- (Mulliken) or 0.56 e^- (natural bond orbital, NBO) of the anionic charge remain localized at the oxyanionic center. Conformations and hydrogen bonds have also been discussed on the basis of experimental and theoretical data.

Full text of DOI:10.1016/j.saa.2003.12.040

Spectrochimica Acta Part A 60 (2004) 2601–2610
Spectrochemical, ab initio and density functional studies on the
conversion of 2-hydroxybenzonitrile (o-cyanophenol)
into the oxyanion
∗
Yuri I. Binev , Miglena K. Georgieva, Lalka I. Daskalova
Institute of Organic Chemistry, Bulgarian Academy of Sciences, 1113 Sofia, Bulgaria
Received 7 August 2003; accepted 18 December 2003
Abstract
The spectral and structural changes caused by the conversion of 2-hydroxybenzonitrile (o-cyanophenol) into the corresponding oxyanion
have been followed by IR spectra, ab initio and density functional force field calculations. In agreement between theory and experiment, the
−1
conversion is accompanied by a 29 cm frequency decrease of the cyano stretching band, 2.7-fold increase in its integrated intensity, 5.8-fold
total value) intensification of the aromatic skeletal bands of Wilson’s 8 and 19 types, and other essential spectral changes. According to the
(
calculations, the strongest structural changes are the shortening of the Ph–O bond with 0.10 Å, lengthenings of the adjacent CC bonds in the
phenylene ring with 0.06 Å and bond angle variations near the oxyanionic center. All these changes are connected with the formation of a
−
quasi-ortho-quinonoidal structure of the o-phenylene ring in the oxyanion. According to the electronic density analysis, 0.41 e (Mulliken)
−
or 0.56 e (natural bond orbital, NBO) of the anionic charge remain localized at the oxyanionic center. Conformations and hydrogen bonds
have also been discussed on the basis of experimental and theoretical data.
©
2004 Elsevier B.V. All rights reserved.
Keywords: IR; Ab initio; DFT; 2-Hydroxybenzonitrile (o-cyanophenol); Anion; Conformations; Hydrogen bonds
1
. Introduction
tions [13–16]. Gebicki and Krantz found that the formation
of CO · · · H–O–C H4CN hydrogen bonds in solid argon ma-
6
−
1
−1
Finkelstein and Willems reported as early as 1929 that
-hydroxybenzonitrile, alone or with appropriate addi-
tives, can be used as a disinfectant, preservative, fungi-
cide, parasiticide, etc. [1]. The general toxic action of
trices resulted in a 63 cm decrease in ␯O–H and a 15 cm
2
increase in ␯C=O [17]. Complexes of 2-hydroxybenzonitrile
with water and methanol have recently been studied using
IR/UV double resonance and other techniques [18]. Detailed
vibrational assignments of the three isomeric hydroxyben-
zonitriles were reported by Ram et al., quite a long time ago
[19].
Vibrational spectra of the 2-hydroxybenzonitrile oxyanion
have not been studied, so we believe the IR data for this
anion will be of a special interest. We did not find in the
literature any ab initio or DFT force field studies on the title
species.
2-hydroxybenzonitrile has been repeatedly studied, e.g.
[
2–4], and herbicidal and insecticidal activities of its deriva-
tives have also been proven [5,6]. There is over a hundred of
patents describing 2-hydroxybenzonitrile as a precursor in
the synthesis of medicines, e.g. for the treatment of throm-
boembolic disorders [7–9] and to enhance the anti-tumor
activity of other cytotoxic agents [10], as drug delivery
systems [11], modulators of neuropatic pains [12], etc.
Starting by Puttnam’s work of 1960 [13], the IR stud-
ies of 2-hydroxybenzonitrile dealt mainly with its intra- and
intermolecular hydrogen bondings in connection with the
s-cis/s-trans preference of the molecule under various condi-
2. Experimental
We prepared 2-hydroxybenzonitrile 1 from salicylalde-
∗
hyde 2, according to the classical Meyer’s procedure [20]
(Scheme 1). The reaction of salicylaldoxime 3 with acetan-
Corresponding author. Tel.: +359-2-739998; fax: +359-2-8700225.
E-mail address: yb@orgchm.bas.bg (Y.I. Binev).
1
386-1425/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.saa.2003.12.040

2
602
Y.I. Binev et al. / Spectrochimica Acta Part A 60 (2004) 2601–2610
Scheme 1. Preparation of 2-hydroxybenzonitrile 1 from salicylaldehyde 2.
hydride resulted in simultaneous conversion of the aldoxime
group into cyano and the phenolic hydroxy one into ace-
toxy. Hydrolysis of the ester 4 by 10 wt.% KOH/aq. gave
the potassium salt 5, which reacted with 10 wt.% H2SO4/aq.
to release the end product 1. The reaction mixture was
extracted by ether, the ethereal layer was washed, dried
and the solvent evaporated. The residue was crystallized
indices; the theoretical spectra were simulated by means of
the SimOpus [27] program.
4. Results and discussion
4.1. Energy analysis
◦
◦
(
m.p. 65 C) after an overnight stay at 4 C. Recrystalliza-
◦
The 2-cyanobenzonitrile molecule can exist as two con-
tion from ethanol gave white crystals with m.p. 91 C; their
◦
◦
+
formers, s-cis 1a and s-trans 1b (Scheme 2). Their total en-
ergies and that of their oxyanion 6, calculated by the five
methods used, are compared in Table 1. Here are some com-
ments on the data in Table 1:
m.p. increased to 96 C after a 2-day storage at −18 C. The
2
-hydroxybenzonitrile oxyanion 6 (counter ion Na ) was
−
1
prepared by adding 0.15 and 0.30 mol l DMSO/DMSO-d6
solutions of 1 to an excess of dry sodium methoxide-d0 and
-
d3 respectively, under argon. The conversion was practi-
1. According to all the calculations performed, 1a is more
−1
cally complete: no bands of 1 were seen in the spectra after
metalation. The dry sodium salt (6, counter ion Na ) was
prepared by reacting equivalent amounts of 1 and NaOH in
water under argon, followed by evaporation of the solvent
in vacuum. The IR spectra were measured on a Bruker IFS
stable than 1b by 9.31–11.18 kJ mol . This result is
due to the formation of weak intra-molecular hydro-
gen bonds, known from experimental IR and NMR data
[13–16], CNDO/2 [14], MNDO [28] and DFT [29] cal-
culations. However, the two conformers coexist in solu-
tions ([13–16] and the next sections). Having in mind the
dipole moment value of 3.2 Debye [30], the s-cis con-
former 1a should be prevalent in carbon tetrachloride so-
lutions (sections 4.2 and 4.3.1), and in the solid state, the
conformation is firmly s-trans [31].
+
1
13v FTIR spectrophotometer in a CaF2 cell of 0.13 mm
−
1
path length (for 0.15 and 0.30 mol l DMSO/DMSO-d so-
lutions), a KBr cell of 0.18 mm path length (for 0.09 mol l
CHCl3/CDCl3 solutions), and KBr and CsI discs, at a reso-
6
−
1
−
1
lution of 1 cm and 50 scans.
2
. The deprotonation energy ED of 2-hydroxybenzonitrile
can be calculated as a difference between the ZPVE-corr-
ected total energies of the anion 6 and that of the
3
. Computations
more stable conformer 1a of the molecule, ED
=
The ab initio force field computations were performed
E
anion
− E
. The four methods used gave
molecule
D
by using the GAMESS software [21] (AIX version, 2000)
at the HF [22] and MP2 [23] 6-31G(d) levels. The GAUS-
SIAN 98 program package [24] was used for the density
functional BLYP [25], B3LYP [26] 6-31G(d) and B3LYP
within the 6-31G(d) basis E values in the interval of
1410.24–1430.44 kJ mol 1. The ED values are related
to the gas-phase acidities of Broensted acids and could
be used as an approximate measure of their pKa values
in polar aprotic solvents, which solvate cations mainly,
−
6
-31 + G(d,p) force field computations. All the computa-
D
tions were performed on a Pentium 2 PC, at a full geometry
optimization. No scaling in the ab initio or density functional
force fields was done. A standard least-square program was
used for calculating the single-parameter linear regression
e.g. DMSO: higher E ’s → lower acidities → higher
pKa’s, e.g. [32–39]. For example, there is a fair linear
correlation between gas-phase ꢀGacid and DMSO pKa
values for 97 C–H acids of various structures [40].

Y.I. Binev et al. / Spectrochimica Acta Part A 60 (2004) 2601–2610
2603
Scheme 2. Mulliken MP2 6-31G(d) (in italics) and natural bond orbital (NBO) B3LYP 6-31G(d) (in bold) total net electric charges over fragments of
the s-cis 1a, s-trans 1b 2-hydroxybenzonitrile molecules of their oxyanion 6.
mean deviation in kJ mol 1): HF, 46.8; MP2, 28.5; BLYP,
2
−
Protic solvents, e.g. water, solvate both cations and an-
ions, and modify strongly the Broensted acidity [41,42].
The pKa value of 4-hydroxybenzonitrile in DMSO is
1.7; and B3LYP, 28.7. The corresponding results, ob-
tained on the basis of semi-empirical calculations were
not very different from these just cited: MINDO/3, 96.4;
MNDO, 29.0; and AM1, 17.3 [49]. It should be empha-
sized here that the ED values of 1382.4 kJ mol obtained
for the title 2-hydroxybenzonitrile by means of the B3LYP
1
3.01–13.18 [38–40,43–45], however, we did not find in
the literature of any pKa data for DMSO solvent of the
−1
studied 2-hydroxybenzonitrile. So, we can compare the
D
E
values for the three hydroxybenzonitriles only with
the corresponding pKa’s, measured in water, viz.: the pKa
6
-31 + G(d,p) method (cf. Table 1, last column), practi-
D
value of 7.86 [41] for the studied o-isomer (see its E
cally coincided with the experimental one (deviation of
D
−
1
above) is close to that of 7.97 [41], for the p-isomer (E =
−
1.9 kJ mol only, cf. above).
−
1
1
394.47–1419.37 kJ mol , by the same methods [46]),
quite lower than pKa of 8.61 [47] and for the m-isomer
4
.2. Correlation analysis
D
−1
(
[
E = 1414.99–1445.09 kJ mol , using the same methods
46]).
In order to check which one among the methods used
D
Fortunately, however, the experimental E ’s of the iso-
gave better description of the vibrational frequencies of
the species studied, we compared their frequency pre-
dictions with our experimental data (sections 4.3.1 and
meric hydroxybenzonitriles were known for a long time.
They were determined by gas-phase proton-transfer equi-
libria, using a pulsed electron beam high-pressure mass
spectrometer [48]. The following values were obtained
4
.3.2). On the other hand, the correlation obtained could
give some information about the prevalent conformer
1a or 1b) of 2-hydroxybenzonitrile, like in the case of
-amino-2-propenenitrile [38]. The results are summarised
D
−1
in this manner (isomer, E in kJ mol ): ortho, 1384.3;
meta, 1392.3; and para, 1378.0 [48]. These values could
directly be compared (they should coincide) with the the-
(
3
in Table 2 and we can see there, that:
D
oretical E ’s. In fact, however, the ab initio and DFT
methods, applied within the 6-31G(d) basis set (this work
and [46]), overestimated the deprotonation energies of hy-
droxybenzonitriles (cf. the above given E intervals) by the
1. As it can be expected [34–39,50,51], the DFT correla-
tions are much better than the ab initio ones: they show
largest R (correlation coefficient) smallest standard de-
viation (S.D.) and ρ (slope) parameters closed to 1, i.e.
D
following mean deviations for the three isomers (method,
Table 1
Native and ZPVE-corrected (values in italics) total energies (in hartree), calculated for the s-cis 1a and s-trans 1b 2-hydroxybenzonitrile conformers, and
for their oxyanion 6
No.
Structure
HF 6-31G(d)
MP2 6-31G(d)
BLYP 6-31G(d)
B3LYP 6-31G(d)
B3LYP 6-31 + G(d,p)
1
2
3
4
5
6
1a
−397.295980
−397.184515
−397.292272
−397.180968
−396.736625
−396.639545
−398.507809
−398.405105
−398.504005
−398.401535
−397.955184
−397.865900
−399.579164
−399.478825
−399.574697
−399.474565
−399.028426
−398.941554
−399.711788
−399.608025
−399.707406
−399.603866
−399.158462
−399.068427
−399.741249
−399.637834
−399.737270
−399.634037
−399.200739
−399.111074
1b
6

2
604
Y.I. Binev et al. / Spectrochimica Acta Part A 60 (2004) 2601–2610
Table 2
Correlations between theoretical and experimental IR frequencies of s-cis 1a and s-trans 1b 2-hydroxybenzonitrile conformers, and of their oxyanion 6,
according to the equation νexp. = ρνtheor. + b (cm−
1
)
a
bb
26.9
Rc
S.D.d
ne
No.
Species
Method
ρ
1
2
3
4
5
6
7
8
9
1a
HF 6-31G(d)
MP2 6-31G(d)
BLYP 6-31G(d)
B3LYP 6-31G(d)
B3LYP6-31 + G(d,p)
HF 6-31G(d)
0.8786
0.9203
0.9725
0.9409
0.9353
0.8723
0.9140
0.9662
0.9344
0.9291
0.9106
0.9356
0.9922
0.9622
0.9609
0.9992
0.9986
0.9992
0.9995
0.9996
0.9990
0.9990
0.9996
0.9997
0.9997
0.9989
0.9989
0.9997
0.9996
0.9997
35.6
48.3
35.9
27.9
25.7
40.4
41.0
25.9
22.0
23.9
41.8
40.6
21.9
24.1
22.0
34
34
34
34
34
34
34
34
34
34
26
26
26
26
26
70.1
31.6
31.2
42.5
37.8
81.0
42.2
42.2
52.8
−13.7
50.8
15.5
10.8
18.0
1b
6
MP2 6-31G(d)
BLYP 6-31G(d)
B3LYP 6-31G(d)
B3LYP6-31 + G(d,p)
HF 6-31G(d)
10
11
12
13
14
15
MP2 6-31G(d)
BLYP 6-31G(d)
B3LYP 6-31G(d)
B3LYP6-31 + G(d,p)
a
b
c
d
e
Slope.
Intercept.
Correlation coefficient.
Standard deviation.
Number of data points.
the native (unscaled) DFT frequencies are closer to the
experimental values.
the first scaling correlation equations appeared within
1992–1996 for AM1 [53] and ab initio [54–56] frequen-
cies [53–56] and intensities [55,56]. This type of scaling
is nowadays used frequently, e.g. [36–38,57,58].
2
. The frequencies, calculated for the s-trans conformer
b, correlate better with the experimental ones (excep-
1
tion: HF), than those for 1a. The sum of the S.D. values
ꢀ
−
1
1
for the five methods used
is essentially smaller than
(
S.D. (1b) = 147.9 cm
The correlations between calculated and measured inte-
grated intensities are poor, with R of 0.31–0.51 for 1a and 1b,
and 0.32–0.45 for 6. The S.D.’s of ca. 12.0 (1a), 11.6 (1b)
and 51.3 (6) km mol are larger than the intensities of many
bands in the corresponding spectra, so these “correlations”
could be used for some orientation only.
ꢀ
−
S.D. (1a) = 162.3 cm
cf. Table 2). The separate correlations for solvent
DMSO/DMSO-d (10 data points) and CHCl3/CDCl3
28 data points) (not included in Table 2) are also infor-
−
1
6
(
mative, as follows:
ꢁ
S.D. (1b) = 89.1 cm−
1
;
solvent DMSO/DMSO-d6 :
4.3. Spectral analysis
ꢁ
S.D. (1a) = 109.0 cm−
1
The IR spectra (1100–2300 cm 1) of the molecule and
−
anion studied are compared on Fig. 1. Let us first consider
separately the spectra of the species studied, which will make
it possible to follow the spectral changes, caused by the
conversion of the 2-hydroxybenzonitrile molecule 1b (see
above) into the corresponding oxyanion 6.
ꢁ
S.D. (1b) = 159.3 cm−1
solvent CHCl3/CDCl3 :
ꢁ
;
S.D. (1a) = 176.1 cm−
1
All these results can be explained in the following
way: like in both cyclohexane and carbon tetrachlo-
ride solutions [13–16], 1a and 1b coexist also in the
chloroform solutions studied (see also Section 4.3.1).
In DMSO/DMSO-d6 solutions 1b should be strongly
prevalent, due to the formation of strong intermolecular
hydrogen bonds with the solvent mainly.
4.3.1. The 2-hydroxybenzonitrile molecule 1b
We compare the theoretical and experimental IR data for
this molecule in Table 3. A good agreement can be seen
there between experimental and theoretical frequencies. The
mean absolute deviation (m.d.) between them for the ab ini-
−
1
tio calculations is ca. 27 cm , which value is not away
−
1
3
. We shall use the ρ and b indices of correlations 6-15
from the 15–25 cm interval of deviations, typical for ni-
trile molecules [37] (and references therein). For the DFT
(Table 2) for correlational scaling of the corresponding
−1
theoretical frequencies. As it was stated in Alcolea’s
review article [52], the use of scaling equations instead
of scale factors gave better results, especially in the
low-frequency region. Again according to Alcolea [52],
calculations this deviation is ca. 18 cm . This value is es-
sentially smaller than the ab initio one and lies within the
−
1
deviation interval of 9–20 cm , found in cases of DFT cal-
culations for five nitrile molecules [36–38,46].

Y.I. Binev et al. / Spectrochimica Acta Part A 60 (2004) 2601–2610
2605
Table 3
Theoretical (HF 6-31G(d), MP2 6-31G(d), BLYP 6-31G(d) and B3LYP 6-31G(d), 6-31 + G(d,p) force field) and experimental (solvent DMSO/DMSO-d6)
IR data for the s-trans-2-hydroxybenzonitrile molecule 1b
No.
1
HF
MP2
ν (cm
3540
BLYP
ν (cm
3525
B3LYP
B3LYPa
B3LYPa
A (km mol
80.5
Approximate descriptiond
Experimentalb
−
1 c
−1 c
−1 c
−1 c
−1 c
−1
−1
−1
A (km mol )
ν (cm
3650
)
)
)
ν (cm
)
ν (cm
)
)
ν (cm
)
3550
3609
νO–H
3586e,f
13.6
36.6
189.7
9.6
1.5
5.0
w
e,g
3
3
3076
559
311
e,h
e
2
3
4
5
6
7
8
9
3012
2995
2983
2979
2313
1617
1599
1501
1451
1313
1274
1211
1181
1159
1107
1069
1030
1010
982
874
837
778
751
722
609
591
562
3075
3057
3045
3038
2102
1622
1610
1514
1464
1421
1319
1294
1224
1192
1188
1112
1068
886
875
862
816
759
756
625
587
577
3088
3075
3063
3039
2213
1586
1569
1492
1443
1344
1298
1264
1202
1171
1168
1091
1042
945
914
841
831
751
733
726
601
584
3058
3045
3034
3012
2238
1596
1579
1493
1444
1329
1297
1264
1201
1165
1160
1091
1042
959
929
844
842
758
736
729
608
590
3046
3034
3025
3005
2223
1588
1569
1481
1433
1321
1289
1255
1194
1157
1155
1087
1040
973
943
853
847
764
740
735
612
594
6.4
5.8
4.6
νPh–H
νPh–H
νPh–H
νPh–H
3040e
e
3021
9.2
2960i
2223
1603
1590
1506
1457
1368
1275
1246
1215
1182
1160
1104
40.4
46.0
25.4
36.2
62.9
30.1
89.4
27.0
27.9
20.4
25.1
57.5
2.8
0.0
1.7
0.7
20.6
79.4
2.4
0.1
2.7
0.4
1.4
12.6
3.9
0.0
8.4
116.1
0.3
2.5
0.1
νC , νC–CN
24.6
20.0
25.7
6.6
47.4
19.9
13.0
16.2
48.4
0.2
5.7
3.8
4.5
0.7
4.2
0.1
5.9
s
≡
N
νCC, δPhH, δCCC
νCC, δPhH, δPhoH
δPhH, νCC, δCCC
δPhH, δCCC, νCC
δPhH, δPhH
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
νPh–O, δPhH, νCC
δPhH, δPhOH, νCC
δPhH, νC–CN, δCCC
δPhH, νCC, δPhOH
νCC, δPhH, δPhOH
νCC, δPhH, δPhOH
ν, δPhH
c
1034
c
νCC, δPhOH, δPhH
δPhH
945
877
859
847
767i
754
732i
598
567
560
496
468
463
400
378
c
c
γPhH
c
γPhH
δCCC
i
γPhH
s
m
γPhH
e
δCCC
1.7
1.7
8.1
0.1
0.4
12.1
w
e
τsk, γC–C≡N
δCCC, δC–C≡N
τsk, γC–C≡N
δCCC, δPhO, δC–C≡N
τsk, γC–C≡N
δCCC, δC–C≡N
γPhOH
e
510
497
473
421
409
384
281
209
565
510
474
411
407
384
257
174
570
518
479
418
391.2
391.0
266
183
163
575
523
486
424
398
377
272
192
e
523
472
418
382
329
261
182
157
e
e
e
j
k
τsk, γC–C≡N
δCCC, δPhO, δC–C≡N
τsk, γPhO
m
190j
k
187
154
172
a
b
c
d
B3LYP 6-31 + G(d,p).
Measured after having decomposed the complex bands into components.
Scaled, according to the correlation equations Nos. 6, 7, 8, 9 and 10 (Table 2), respectively.
Vibrational modes: ν, stretching; δ, bendings, except out-of-plane ones; γ, out-of-plane bendings; τ, torsion; subscript sk, skeletal. For the atom
numbering see Scheme 1.
e
Solvent CDCl3.
A shoulder corresponding to free OH groups.
f
g
OH groups, bonded intra-molecularly O–H · · · N.
h
A broad band, corresponding to OH groups, bonded intermolecularly O–H · · · O.
i
KBr disc. s, strong; m, moderate; w, weak.
j
CsI disc.
The band was not detected.
k
matrices this band appears at 3576 cm 1 (1a) [17]. The con-
centrations of our CHCl3/CDCl3 solutions were relatively
high (see Section 2), so we found three νO–H absorptions: a
−
According to the calculations, vibration 1 (Table 3) is a
pure νO–H mode. This vibration was studied repeatedly in
connection with the conformers of various o-substituted phe-
nols, including the title molecule [13–17]. For example, in
very diluted CCl4 solutions νO–H gives rise to two bands,
−
1
−1
shoulder at 3586 cm (1b, free), a band at 3559 cm (1a,
intra-molecularly bonded) and a broad band at 3311 cm
−1
,
−
1
−1
,
at 3595 cm , weak (s-trans 1b, i.e. free) and 3559 cm
strong (s-cis 1a, intra-molecularly bonded) [15], and in argon
corresponding to intermolecularly bonded O–H· · · O. The
relative intensities of these bands (Table 3) cannot give in-

2
606
Y.I. Binev et al. / Spectrochimica Acta Part A 60 (2004) 2601–2610
tegrated intensity ACN (1b) however, is essentially lower
than ACN (p-isomer): predicted (B3LYP) and measured
.7-fold (Table 3 and [46,60]). For the corresponding an-
ions the ortho/para variations are qualitatively similar, viz.
1
−
1
a νCN increase with: predicted 14 cm
(B3LYP), mea-
−
1
sured 7 cm ; ACN decrease: predicted 1.9-fold (B3LYP),
measured 2.0-fold (Table 4 and [46,60]).
According to Katritzky and Topsom [61], the intensities
of the aromatic skeletal bands 8a,b and 19a,b (Wilson’s
notation) can be used as a quantitative measure of the
intra-molecular electronic interactions. For 1b these are
bands Nos. 7 (8a), 8 (8b) and 9 (19), both predicted and
experimentally found (Table 3). Their total integrated inten-
−
1
sity of 52.3 kJ mol is intermediate with respect to those of
−
1
−1
the meta (46.4 km mol [46]) and para (142.0 km mol
60]) isomeric hydroxybenzonitrile molecules.
The ν coordinate dominates in vibration 12, with pre-
dicted frequency 1274–1319 cm (ab initio, DFT, Table 3).
We found the corresponding band at 1275 cm (DMSO-d6)
[
Ph–O
−
1
−
1
Fig. 1. (A) Experimental IR spectra (0.15 mol l
in DMSO) of
−1
2
-hydroxybenzonitrile (1b, shaded) and of its oxyanion (6, counter ion
−
1
+
and 1304 cm (CDCl and KBr). These frequencies are not
away from the values of 1291, 1278 and 1285 cm , mea-
Na ). (B) B3LYP 6-31 + G(d,p) (scaled by correlation equations nos.
3
−1
1
0 and 15 (Table 2), respectively) spectra of the same species. Simulat-
ing the theoretical spectra by means of our program SimOpus [27] we
postulated half-widths of 10 cm for all the bands.
sured for the p-isomer under the same conditions [60]. The
−1
corresponding vibration is also dominated by the ν
co-
Ph–O
−
1
ordinate, but the intensity of the latter band (56.1 mol l
,
formation about the 1a/1b concentration ratio, as the stretch-
ing bands of hydrogen-bonded O–H, N–H, etc. groups are
more intense (in cases of intermolecular hydrogen bonds
they are very broad and much more intense) than the bands
of the corresponding free (non hydrogen-bonded) groups
solvent DMSO-d [60]) is 4.3-fold higher than A (1b)
6
Ph–O
(cf. Table 3). All these values are reasonably higher (due
to the HO/CN resonance, see Section 4.4.1) than the value
−
1
of 1249 cm , measured by Pinchas for the unsubstituted
phenol in H O, and assigned as ν on the basis of the
2
Ph–O
−
1 16
18
[59]. We shall mention that if two o-hydroxybenzonitrile
15 cm O/ O isotopic shift [62]. Normal vibrations with
essential contributions of the ν
molecules form an intermolecular hydrogen bond, at least
coordinate have been
Ph–O
−
1
one of them has s-trans conformation. The 1b νO–H fre-
predicted near 1270 and 1256 cm , and found at 1293
−
1
−1
quency of 3586 cm (shoulder) is very close to that of the
and 1243 cm (solvent DMSO-d ) in the IR spectrum of
6
−
1
p-isomer (3580 cm [60]) and practically coincides with
4-hydroxynitrobenzene [63].
−
1
νO–H of the m-isomer (3587 cm [46]); their sharp bands
were measured under the same conditions. The molecules of
4.3.2. The 2-hydroxybenzonitrile oxyanion 6
3
- and 4-hydroxybenzonitriles cannot form intra-molecular
The theoretical and experimental IR data for the anion
6 are compared in Table 4. As above (Section 4.3.1), we
can find there a good agreement between experimental and
scaled theoretical frequencies. The m.d. between them for
hydrogen bonds.
In DMSO/DMSO-d solutions the 1b molecules form
6
very strong O–H · · · O= S(CH3)2 bonds, manifested by
−1
a νO–H multiplet (over 10 bands) between 3200 and
the ab initio methods used is 31.0 cm , larger than the
−1
−1
2
500 cm , like νO–H of the m- [46] and the p- [60] isomers
14–28 cm
interval of deviations, typical for nitrile an-
under the same conditions.
ions [37] (and references therein). For the DFT calculations
−
1
According to the calculations, the δPhOH coordinate
is strongly delocalized, taking parts in vibrations 8, 11,
this deviation is ca. 16 cm . This value is almost two-fold
smaller than the ab initio one and also lies within the DFT
−
1
1
3, 15-17 (Table 3). On the contrary, γPhOH is well lo-
deviation interval of 9–24 cm , found for five nitrile anions
[36–38,46].
calized: vibration 33 is almost a pure γPhOH mode. This
vibration was studied in detail by Carlson and coworkers
The conversion of the 2-hydroxybenzonitrile molecule 1b
into the oxyanion 6 results in very essential changes in the
IR spectrum (cf. Tables 3 and 4); some of them are well
seen in Fig. 1. The changes are in a qualitative agreement
between theory and experiment, e.g.:
[
15,16], who measured the corresponding bands (satu-
−
1
rated, <0.01 mol l cyclohexane solutions) of 1a at 392,
3
−1
−1
(Fermi-doublet) and of 1b at 343 cm , sepa-
76 cm
−
1
rately. For 1b (Cs disc) we found this band at 378 cm (in
the solid state the conformation is s-trans [31]).
The frequency of the cyano stretching band νCN of 1b
is a bit higher than νCN of the p-isomer, with: predicted
1. Decrease in the cyano stretching frequency νCN: pre-
−
1
dicted 22 cm
(B3LYP 6-31 + (d,p)), measured
−1
−1
−1
5
cm (B3LYP), measured 2 cm ; the corresponding in-
29 cm
.

Y.I. Binev et al. / Spectrochimica Acta Part A 60 (2004) 2601–2610
2607
Table 4
Theoretical (HF 6-31G(d), MP2 6-31G(d), BLYP 6-31G(d) and B3LYP 6-31G(d), 6-31 + G(d,p) force field) and experimental (DMSO/DMSO-d6) IR
+
data for the 2-hydroxybenzonitrile oxyanion 6 (counter ion Na )
No.
HF
MP2
BLYP
B3LYP
B3LYPa
B3LYPa
Approximate descriptiond
Experimentalb
−
1 c
−1 c
−1 c
−1 c
−1 c
−1
−1
−1
A (km mol )
ν (cm
)
ν (cm
)
ν (cm
)
ν (cm
)
ν (cm
)
A (km mol
)
␯ (cm
)
1
2
3
4
5
6
7
8
9
3061
3041
3015
3006
2302
1639
1585
1518
1467
1382
1246
1209
1143
1124
1056
985
978
952
837
835
738
701
684
594
3072
3060
3033
3016
2101
1649
1567
1548
1481
1387
1343
1256
1176
1159
1098
1042
867
856
821
783
728
697
621
611
585
3085
3071
3050
3017
2188
1614
1550
1502
1434
1376
1276
1224
1157
1139
1081
1016
915
874
853
817
729
698
692
598
579
3075
3063
3043
3014
2218
1627
1562
1507
1461
1373
1275
1227
1154
1133
1075
1012
923
886
849
820
729
698
692
597
578
3078
3066
3047
3021
2201
1595
1528
1499
1448
1361
1272
1227
1149
1131
1076
1012
937
905
853
823
728
709
704
600
581
44.0
55.4
21.3
41.6
278.2
358.8
220.6
107.7
54.5
5.8
71.3
2.0
21.9
27.3
0.9
16.6
0.3
2.1
4.8
17.0
0.8
70.7
2.6
1.9
1.2
2.1
νPh–H
νPh–H
νPh–H
νPh–H
3075
3058
3042
3016
2194
1595
1522
1496
1453
1364
1276
1246
1169
1146
1107
4.2
2.4
3.6
8.5
65.8
84.0
13.7
207.1
14.1
46.2
25.0
1.6
4.3
26.8
5.1
w
νC≡N, νC–CN
νCC, νPhO
νCC, νPh–O
νCC, δPhH
νCC, δCCC, δPhH
δPhH, δCCC
δPhH, νCC
δCCC, δPhH, νCC
δPhH, νCC
δPhH, νCC
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
γPhH, δPhH
νCC, δPhH
γPhH
e
1033
f
−
e
γPhH, τsk
863
f
m
δCCC, νCC
γPhH, γPhO
γPhH, γPhO, τsk
νCC, δCCC
γPhH, γPhO
δCCC, δC–C≡N
γC–C≡N, τsk
δCCC, νCC
γPhH, τsk, γPh
δCCC, δPhO, δC–C≡N
τsk, γC–C≡N
−
e
841
f
m
−
748
728
598
570
506
479
454
407
m
s
m
m
m
m
m
m
577
543
500
456
528
486
415
413
546
499
464
410
560
500
462
407
566
507
470
411
19.2
6.7
0.2
g
395
a
B3LYP 6-31 + G(d,p).
b
c
d
Measured after decomposing the complex band into components.
Scaled, according to the correlation equations Nos. 11, 12, 13, 14 and 15 (Table 2), respectively.
Vibrational modes: ν, stretching; δ, bendings, except out-of-plane ones; γ, out-of-plane bendings; τ, torsion; subscript sk, skeletal. For the atom
numbering, see Scheme 1.
e
KBr disc.
The band was not detected.
Followed by four lower-frequency vibrations.
f
g
2
. Essential increase in the corresponding intensity ACN:
predicted 6.9-fold (same method), measured 2.7-fold.
. Strong enhancement of the intensity of the aromatic
skeletal bands of Wilson’s 8 and 19 types; for 6 these are
Nos. 6 (8), 7 (19a) and 8 (19b), both predicted and mea-
sured (Table 4). The sum of their integrated intensities of
IR frequencies and intensities [65–67] (and references
therein) are the σ constants of anionic substituents
+
3
[68,69].
4. Unlike the preceding case (1b, ν
at 1275 cm 1), the
−
Ph–O
ν
coordinate of 6 is delocalized, participating to, but
Ph–O
not dominating in vibrations 6 and 7, predicted within the
−
1
−1
−1
3
04.8 km mol is again intermediate between the values
1614–1649 cm and 1528–1585 cm intervals (ab ini-
tio, DFT) and found at 1595 and 1522 cm (in fact these
−
1
−1
of 237.1 [46] and 421.0 [60] km mol , measured for
the m- and p-hydroxybenzonitrile anions, respectively.
The intensity enhancement, caused by the 1b → 6 con-
version is: predicted 6.4-fold (B3LYP 6-31 + G(d,p)),
measured 5.8-fold (Tables 3 and 4 and Fig. 1). The
essential intensity enhancements of the 8 and 19 type
bands, caused by conversions of molecules into carban-
ions, azanions and oxyanions [37,39,46,51,55,56] (and
references therein) are examples for the strong effects of
anionic substituents on these intensities [64]. Approxi-
mate quantitative measures of these effects on various
are Wilson’s 8a and 19a modes, but modified by the sub-
stantial participation of the ν
coordinate, Table 4).
Ph–O
Like in the case of the 4-hydroxybenzonitrile oxyanion
[60] however, this shift of the ν coordinate to higher
Ph–O
frequencies is obviously due to the strong shortening
(hence strengthening) of the Ph–O bond, accompany-
ing the conversion of the 2-hydroxybenzonitrile molecule
into the oxyanion (cf. 4.4). For example, according to the
calculations, the ν
bration (nominally Wilson’s 19a) of the 4-nitrophenolate
coordinate dominates in the vi-
Ph–O

2
608
Y.I. Binev et al. / Spectrochimica Acta Part A 60 (2004) 2601–2610
anion, predicted near 1552 cm 1 and found at 1532 cm
−
−1
ues can usually [34–39,52,60] answer quantitatively for
the question which one among the theoretical methods
used gives the best description. For comparison: sim-
(
solvent DMSO-d ) [63].
6
◦
ilar m.d.’s of 0.006-0.024 Å and 0.94–1.23 have re-
4
.4. Structural analysis of the species studied
cently been found for 4-hydroxybenzonitrile (same meth-
ods [46]). Of course nobody can expect a full coinci-
dence between calculated (for isolated species) and X-ray
We shall follow the changes in both the steric and
electronic structures, caused by the conversion of the
-hydroxybenzonitrile molecule 1b into the corresponding
oxyanion 6.
(
measured in the crystal state) geometry parameters.
2
¯
2
. The m.d. value (∆i) could give information on failures
of most or all the methods used in the determination of
a given structural element [38]. The ∆i values in Table 5
show that the methods used overestimate all the bond
lengths, but the strong disagreement in the O–H case only
¯
4
.4.1. Steric structure
Certain calculated and measured geometry parameters of
the species studied are compared in Table 5. As we men-
tioned above, in the monocrystal 2-hydroxybenzonitrile ex-
ists as infinite chains of molecules in s-trans conformation
31]. So, we included in Table 5 the geometry parameters,
calculated for 1b only. The following comments can be done
¯
∆i = 0.054 Å) can be classified as a real failure. In fact,
(
it is known that both ab initio and density functional cal-
culations overestimate essentially the element–hydrogen
bond lengths [70,71]. As the methods used give good de-
scriptions of the structure of the molecule 1b, we hope
their structural predictions for the oxyanion 6 would also
be adequate. So we should be able to follow the structural
changes, accompanying the conversion of 1b into 6. The
[
on the corresponding data:
1
. In fact all the methods used give good descriptions of
the steric structure of the molecule 1b. The m.d. val-
Table 5
Theoretical (HF 6-31G(d), MP2 6-31G(d), BLYP 6-31G(d) and B3LYP 6-31G(d), 6-31 + G(d,p)), and experimental (X-ray) bond lengths (Å) and bond
angles (degrees) in the s-trans-2-hydroxybenzonitrile molecule 1b and in its oxyanion 6
Speciesa
Methods
Molecule
HF
Anion
HF
¯
∆i
d
B3LYPb
MP2
BLYP
1.435
B3BYP
B3LYPb
X-rayc
1.431
MP2
1.428
BLYP
1.427
B3LYP
1
2
C –C
1.441
1.136
1.394
1.391
1.387
1.383
1.387
1.381
1.341
0.947
1.433
1.183
1.406
1.402
1.396
1.395
1.396
1.393
1.366
0.974
1.431
1.163
1.411
1.405
1.398
1.394
1.397
1.391
1.358
0.970
1.432
1.164
1.411
1.406
1.399
1.395
1.399
1.392
1.359
0.967
0.003
0.029
0.015
0.013
0.011
0.023
0.019
0.020
0.011
0.053
1.433
1.422
1.170
1.468
1.409
1.456
1.378
1.414
1.388
1.253
1.424
1.170
1.465
1.411
1.454
1.381
1.416
1.390
1.261
2
3
C ≡N
1.177
1.424
1.417
1.408
1.404
1.408
1.401
1.372
0.981
1.136
1.394
1.391
1.387
1.371
1.378
1.372
1.348
0.915e
1.143
1.444
1.399
1.452
1.360
1.412
1.372
1.232
1.187
1.456
1.406
1.448
1.384
1.408
1.390
1.265
1.183
1.485
1.420
1.467
1.389
1.424
1.399
1.266
1
4
8
5
6
7
8
C C
1
C C
4
C C
5
C C
6
C C
7
C C
4
9
C –O
9
14
g
g
g
g
g
O –H
–
–
–
–
–
m.d.f
0.0052
0.0180
0.0264
0.0156
0.0191
0.0000
1
2
3
C C N
178.05
120.36
118.34
121.30
119.46
119.23
120.01
120.94
119.06
117.76
122.78
111.25
0.687
178.28
179.21
179.18
120.35
120.23
119.42
119.59
120.72
120.19
120.65
119.44
117.36
123.06
109.44
0.700
178.37
120.35
120.22
119.43
119.65
120.69
120.16
120.64
119.44
117.36
122.98
110.45
0.765
179.15
−0.53
0.79
0.28
−1.04
1.20
0.54
0.06
−0.93
0.18
−0.33
−0.86
–
176.93
119.31
118.50
122.20
113.37
122.10
122.75
122.18
117.42
124.11
122.50
177.73
118.09
118.90
123.00
112.81
121.10
123.39
121.50
118.18
123.75
123.40
178.04
119.21
118.97
121.82
113.22
121.72
123.31
121.55
118.38
123.76
123.02
178.88
119.13
118.84
122.03
113.21
121.75
123.20
121.69
118.11
123.81
122.99
177.41
119.53
118.63
121.84
113.68
121.71
122.99
121.61
118.16
123.68
122.63
2
1
4
8
8
5
7
6
7
8
9
C C C
119.74
120.30
120.00
119.52
120.20
120.11
120.52
119.70
116.80
123.70
108.82
0.633
120.50
120.33
119.17
119.69
120.80
120.22
120.58
119.55
117.20
123.11
108.63
0.789
119.47
119.60
120.90
118.38
119.79
120.08
121.60
119.26
117.63
123.99
2
1
C C C
4
1
C C C
1
4
C C C
1
8
C C C
4
5
C C C
5
6
C C C
6
7
C C C
1
4
C C O
5
4
9
C C O
4
9
14
h
g
g
g
g
g
C O H
m.d.f
–
–
–
–
–
–
–
–
–
–
–
0.000
–
a
For atom numbering see Scheme 2.
B3LYP 6-31 + G(d,p).
b
c
d
e
f
X-ray diffraction of a monocrystal [31].
Mean algebraic deviation between theoretical and experimental values for a given structural element.
Not included in the m.d. calculation.
Mean absolute deviation (m.d.) between theoretical and experimental values for a given method.
g
h
14
There is no H in the anion.
No experimental data available.

Y.I. Binev et al. / Spectrochimica Acta Part A 60 (2004) 2601–2610
2609
¯
very essential IR spectral and structural changes. These
changes, however, can be described and predicted ade-
quately by both ab initio and density functional force field
calculations.
strong disagreement in the O-H case (ꢀ = +0.053 Å)
i
may not be failure of the methods used. According to
J. Hvoslof (Acta Crystallogr. Sect. B 24 (1968) 1431),
the X-ray diffraction O-H bonds are systematically short-
ened by the decrement of 0.148 Å, compared to the neu-
¯
tron diffraction ones. So the real ꢀ for the O-H bond
i
Acknowledgements
length becomes −0.095 Å and can be considered as a
manifestaion of the intermolecular hydrogen bonds in the
monocrystals, like in the case of 4-hydroxybenzonitrile
The authors thank Prof. I.G. Binev, D.Sc, for the useful
discussion. The financial support by the Bulgarian Council
of Scientific Research for contract Chem.-1213 is also po-
litely acknowledged.
[46].
3
. Among the bond length changes, caused by the conver-
sion of 1b into 6, the strongest are the shortening of the
Ph–O bond with 0.10 Å and lengthenings of the adjacent
CC bonds in the phenylene ring with 0.059 and 0.055 Å,
respectively (mean values for the five methods used).
. The conversion causes changes in all bond angles, but
the anion 6 remains planar. The strongest bond angle
References
4
[1] H. Finkelstein, J. Willems, I.G. Farbenind, A.G., German Patent
538,319 (1929).
◦
changes take place at the oxyanionic center (−6.41 to
[2] S. Ren, Environ. Toxicol. 17 (2002) 119.
◦
◦
◦
+
6.58 , mean values) and its o-positions (2.16 –3.03 ,
[3] S.B. Mekapati, C. Hansch, J. Chem. Inform. Comput. Sci. 42 (2002)
956.
mean values).
5
. Both the bond length and bond angle changes discussed
are connected with the formation of a quasi-ortho-quinonoidal
structure of the phenylene ring in the oxyanion.
[4] A.O. Aptula, T.I. Netzeva, I.V. Valkova, M.T.D. Cronin, T.W. Schltz,
R. Kuhne, G. Schuurmann, Quant. Struct.: Activity Relat. 21 (2002)
12.
[5] K. Carpenter, H.J. Cottrell, B.J. Heywood, W.G. Leeds, Mededel
Landbonwhogesh. Opz., St. Gent 29 (1964) 644.
4
.4.2. Electronic structures
The MP2 6-31G(d) (Mulliken, in italics) and B3LYP 6-31
G(d,p) (NBO, in bold) net electric charges q in the species
[6] H. Mizuno, N. Sakamoto, Y. Kinoshita, PCT Int. Appl. WO 02
4663 (Cl. C07D239/00), 28 March 2002.
2
[7] R.R. Wexler, C.G. Clack, J.M. Fevig, R.A. Galemmo, Q. Han, I.G.
Jacobson, P.Y.S. Lam, H.Y. Li, D.Y.P. Pinto, PCT Int. Appl. WO 00
+
studied, divided into fragments (cyano groups, phenylene
38683 (Cl. A61K31/495), 6 July 2000.
−
rings and HO/O groups) are shown in Scheme 2. Here are
some comments on the corresponding data:
[
8] K. Ueno, A. Ogawa, Y. Ohta, Y. Nomoto, K. Takaki, H. Kusaka, H.
Yano, C. Suzuki, S. Nakanishi, PCT Int. Appl. WO 01 47931 (Cl.
C07D487/14), 5 July 2001.
1
. The dipole moment of 2-hydroxybenzonitrile (mainly 1a
in CCl4 solutions) is 3.2 Debye [30]. The MP2 calcula-
tions performed give the value of 3.26 Debye for 1a, and
larger values, up to 3.7 Debye are obtained by the other
methods used. So, we think that among the Mulliken
charges, the MP2 ones are closest to the real charges in
the title species.
[9] T. Sakai, T. Senga, T. Furuta, A. Miwa, PCT Int. Appl. WO 01
7890 (Cl. C07D215/14), 5 July 2001.
10] L.D. Arnold, J.W. Coe, T. Kaneko, M.P. Moyer, US 6130217 (Cl.
14-253 A61K31/495), 10 October 2000.
11] D. Moye-Sherman, D. Cschneidner, PCT Int. Appl. WO 02 20466
Cl. C07C255/00), 14 May 2002.
4
[
[
5
(
[12] S.-K. Choi, P.R. Fatheree, D.C. Green, D. Marquess, PCT Int. Appl.
WO 02 57215 (Cl. C07C217/32), 25 July 2002.
[13] N.A. Puttnam, J. Chem. Soc. (1960) 5100.
2
. The net charge changes ꢀq = q
− q
are usu-
anion
molecule
[14] J.P. Segnin, F. Guillame-Vilport, R. Uzan, J.P. Poucet, J. Chem. Soc.,
Perkin Trans. 2 (1986) 773.
ally quite informative to show the distribution of the new
carbanionic, oxyanionic, etc.) negative charge between
(
[15] G.L. Carlson, W.G. Fateley, J. Phys. Chem. 77 (1973) 1157.
individual fragments of the anions [37] (and many ref-
[16] W.G. Fateley, G.L. Carlson, F.F. Bentley, J. Phys. Chem. 79 (1975)
99.
1
erences therein). In the anion 6 studied, ꢀq = q(6) − q
[
[
17] J. Gebicki, A. Krantz, J. Am. Chem. Soc. 106 (1984) 8097.
18] M. Broguier, F. Lahmani, A. Zehnacker-Rentien, V. Brenner, Ph.
Millie, A. Peremans, J. Phys. Chem. A 105 (2001) 6841.
19] S. Ram, B.K. Gupta, J.S. Yadaw, Indian J. Pure Appl. Phys. 19
(1981) 1110.
(
0
1a) is distributed as follows (NBO ꢀq’s in bold): 0.06,
−
−
.06 e in the cyano group, 0.53, 0.38 e delocalize over
−
the phenylene ring and 0.41, 0.56 e remain in the oxyan-
[
[
ionic center. For comparison: in the oxyanionic centers
20] B. Meyer, Ber. Deutsch. Chem. Ges. 26 (1893) 1254.
of the 3- and 4-hydroxybenzonitrile anions remain 0.44,
−
[21] 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)
0
.55 and 0.49, 0.53 e [46], respectively (same methods)
of the oxyanionic charges.
1347.
[22] C.C. Roothan, Rev. Mod. Phys. 23 (1951) 69.
[
23] M.J. Frisch, M. Head-Gordon, J.A. Pople, Chem. Phys. Lett. 166
5
. Conclusion
(1990) 275.
[
24] 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. Strat-
mann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels, K.N.
We show in this work that the conversion of 2-hydroxybe-
nzonitrile molecule into the corresponding oxyanion causes

2
610
Y.I. Binev et al. / Spectrochimica Acta Part A 60 (2004) 2601–2610
Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. [43] K. Kalfus, J. Socha, M. Vecera, Coll. Czech. Chem. Commun. 39
Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochter-
ski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Mal-
ick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski,
J.V. Ortiz, A.G. Baboul, B.B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, D.J. Fox, T.
Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez,
M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong,
J.L. Andres, C. Gonzalez, M. Head-Gordon, E.S. Replogle and J.A.
Pople, Gaussian 98, Revision A.7, Gaussian, Inc., Pittsburgh, PA,
(1974) 275.
[44] M.K. Chantooni Jr., I.M. Kolthoff, J. Phys. Chem. 80 (1976) 1306.
[45] J.K. Ellington, E.M. Arnett, J. Am. Chem. Soc. 110 (1988) 7778.
[46] M.K. Georgieva, P.N. Angelova, I.G. Binev, J. Mol. Struct., 2004,
in press.
[47] A. Bryson, R.W. Matthews, Aust. J. Chem. 16 (1963) 401.
[48] T.B. McMahan, P. Kebarle, J. Am. Chem. Soc. 99 (1977) 2222.
[49] R. Voets, J.-P. Francois, J.M.L. Martin, J. Mullens, J. Yperman, L.C.
van Pouke, J. Comput. Chem. 11 (1990) 269.
1
998.
[50] A.P. Scot, L. Radom, J. Phys. Chem. 100 (1996) 16502.
[51] B.A. Stamboliyska, E.A. Velcheva, I.G. Binev, Spectrochim. Acta
Part A 59 (2003) 1541.
[
[
[
25] A.D. Becke, Phys. Rev. A 38 (1988) 3098.
26] A.D. Becke, J. Chem. Phys. 97 (1992) 9173.
27] Y.I. Binev, Ph.D. Thesis, Bulgarian Academy of Sciences, Sofia,
[52] M. Alcolea Palafox, Rec. Res. Dev. Phys. Chem. 2 (1998) 213.
[53] M. Alcolea Palafox, Spectrosc. Lett. 27 (1994) 1165.
[54] G. Leroy, M. Sana, C. Willante, D.R. Temsaman, J. Mol. Struct. 259
(1992) 369.
[55] I.G. Binev, J.A. Tsenov, E.A. Velcheva, I.N. Juchnovski, J. Mol.
Struct. 344 (1995) 205.
2
000.
28] N.C. Santhosh, A.C. Nair, P.C. Mishra, Indian. J. Pure Appl. Phys.
6 (5) (1998) 264.
[
[
3
29] F. Himo, L.A. Eriksson, M.R.A. Blomberg, P.E.M. Siegbahn, Int. J.
Quant. Chem. 76 (2000) 714.
[
[
30] C. Curran, E. Chaput, J. Am. Chem. Soc. 69 (1947) 1134.
31] C. Beswick, M. Kubicki, P.W. Codding, Acta Crystallogr., Sect. C
C52 (1996) 3171.
[56] I.G. Binev, J.A. Tsenov, E.A. Velcheva, V.B. Radomirska, I.N. Juch-
novski, J. Mol. Struct. 378 (1996) 133.
[57] I.D. Reva, S.V. Ilieva, R. Fausto, Phys. Chem. Chem. Phys. 3 (2001)
4235.
[58] O. Sala, N.S. Goncalves, L.K. Noda, J. Mol. Struct. 565–566 (1997)
411.
[59] G.N. Andreev, Molecular Spectroscopy (Bulg.), P. Khilendarski Uni-
versity Press, Plovdiv, 1999, p. 69.
[60] Y.I. Binev, J. Mol. Struct. (Theochem) 535 (2001) 93.
[61] A.R. Katritzky, R.D. Topsom, Chem. Rev. 77 (1977) 639.
[62] S. Pinchas, Spectrochim. Acta Part A 28 (1972) 801.
[63] I.G. Binev, R.R. Petrova, Y.I. Binev, Compt. Rend. Acad. Bulg. Sci.
50 (1997) 39.
[64] I.G. Binev, P.J. Vassileva, J. Mol. Struct. 114 (1984) 375.
[65] P.J. Vassileva, I.G. Binev, I.N. Juchnovski, Spectrochim. Acta Part
A 39 (1983) 709.
[66] I.G. Binev, E.A. Velcheva, Spectrosc. Lett. 32 (1999) 793.
[67] E.A. Velcheva, I.N. Juchnovski, I.G. Binev, Spectrochim. Acta Part
A 59 (2003) 1745.
[68] I.G. Binev, R.B. Kuzmanova, J. Kaneti, I.N. Juchnovski, J. Chem.
Soc. Perkin Trans. II (1982) 1533.
[
[
[
[
[
[
[
[
[
32] I.G. Binev, B.A. Stamboliyska, E.A. Velcheva, Spectrochim. Acta
Part A 52 (1996) 1135.
33] I.G. Binev, B.A. Stamboliyska, Y.I. Binev, E.A. Velcheva, J.A.
Tsenov, J. Mol. Struct. 513 (1999) 231.
34] Y.I. Binev, C.T. Petkov, L. Pejov, Spectrochim. Acta Part A 56 (2000)
1949.
35] B.A. Stamboliyska, Y.I. Binev, V.B. Radomirska, J.A. Tsenov, I.N.
Juchnovski, J. Mol. Struct. 516 (2000) 237.
36] Y.I. Binev, J.A. Tsenov, I.N. Juchnovski, I.G. Binev, J. Mol. Struct.
(Theochem) 625 (2003) 207.
37] Y.I. Binev, M.K. Georgieva, S.I. Novkova, Spectrochim. Acta Part
A 59 (2003) 3041.
38] S.I. Novkova, M.K. Georgieva, Y.I. Binev, Ch.Ts. Petkov, I.G. Binev,
J. Mol. Struct. (Theochem) (2003) 167.
39] E.A. Velcheva, B.A. Stamboliyska, Spectrochim. Acta Part A 60
(2004), in press.
40] I.A. Koppel, J. Koppel, V. Pihl, I. Leito, M. Mishima, V.M. Vlassov,
L.M. Yagupolski, R.W. Taft, J. Chem. Soc. Perkin Trans. 2 (2000)
1
125.
41] A. Fini, P. De Maria, A. Guarnjeri, L. Varoli, J. Pharm. Sci. 76
1987) 48.
[69] I.G. Binev, E.A. Velcheva, V.B. Radomirska, I.N. Juchnovski, Com-
put. Rend. Acad. Bulg. Sci. 47 (1994) 33.
[70] A.A. El-Azhary, H.U. Suter, J. Phys. Chem. 100 (1996) 15056.
[71] T.M. Kolev, B.A. Stamboliyska, Spectrochim. Acta Part A 58 (2002)
3127.
[
[
(
42] R.M. Terkka, I.-H. Um, H.-Q. Xie, A.-P. Chatorouse, F. Terrier, S.
Hoz, E. Buncel, J. Phys. Org. Chem. 11 (1998) 847.