Experimental and theoretical study of the dissociation enthalpy of the N-O bond on 2-hydroxypyridine N-oxide: Theoretical analysis of the energetics of the N-O bond for hydroxypyrydine N-oxide isomers

Article abstract of DOI:10.1016/j.jct.2003.10.001

The standard (p° = 0.1 MPa) molar enthalpy of formation of crystalline 2-hydroxypyridine N-oxide was measured, at T = 298.15 K, by static bomb calorimetry and the standard molar enthalpy of sublimation, at T = 298.15 K, was obtained using Calvet microcalorimetry. These values were used to derive the standard molar enthalpy of formation of 2-hydroxypyridine N-oxide in gaseous phase, and to evaluate the dissociation enthalpy of the N-O bond. Additionally, high-level density functional theory calculations using the B3LYP hybrid exchange-correlation energy functional have been performed for the three isomers of hydroxypyridine N-oxide in order to confirm the experimental trend for the dissociation enthalpy of the (N-O) bond.

Full text of DOI:10.1016/j.jct.2003.10.001

J. Chem. Thermodynamics 36 (2004) 107–113
www.elsevier.com/locate/jct
Experimental and theoretical study of the dissociation enthalpy of
the N–O bond on 2-hydroxypyridine N-oxide: theoretical analysis
of the energetics of the N–O bond for hydroxypyrydine
N-oxide isomers
a,
Maria D.M.C. Ribeiro da Silva , M. Agostinha R. Matos ,
a
*
a
Margarida S. Miranda , Victor M.F. Morais , W.E. Acree Jr.
b
c
a
Centro de Investigacß a~ o em Qu ꢀı mica, Departamento de Qu ꢀı mica, Faculdade de Ci e^ ncias da Universidade do Porto,
Rua do Campo Alegre, 687, P-4169-007 Porto, Portugal
b
Instituto de Ci e^ ncias Biom eꢀ dicas Abel Salazar, Universidade do Porto, P-4099-003 Porto, Portugal
Department of Chemistry, University of North Texas, Denton, TX 76203-5070, USA
c
Received 12 June 2003; accepted 3 October 2003
Abstract
ꢀ
The standard ( p ¼ 0:1 MPa) molar enthalpy of formation of crystalline 2-hydroxypyridine N-oxide was measured, at
T ¼ 298:15 K, by static bomb calorimetry and the standard molar enthalpy of sublimation, at T ¼ 298:15 K, was obtained using
Calvet microcalorimetry. These values were used to derive the standard molar enthalpy of formation of 2-hydroxypyridine N-oxide
in gaseous phase, and to evaluate the dissociation enthalpy of the N–O bond. Additionally, high-level density functional theory
calculations using the B3LYP hybrid exchange-correlation energy functional have been performed for the three isomers of hy-
droxypyridine N-oxide in order to confirm the experimental trend for the dissociation enthalpy of the (N–O) bond.
Ó 2004 Elsevier Ltd. All rights reserved.
Keywords: Enthalpy; Combustion; Sublimation; N–O bond; Theoretical calculations
1
. Introduction
The thermochemical study of pyridine N-oxide de-
rivatives has been one of our interests, in order to
enlarge the data needed to evaluate the dissociation
Amine N-oxides have been used for a large variety of
ꢀ
oxidations, since they can readily lose oxygen to mild
reducing agents. Pyridines are readily oxidized to the N-
oxides and they have been used to activate pyridine rings
for substitution chemistry followed by deoxygenating.
Pyridine N-oxides derivatives are also employed in the
oxidation of alkyl halides to aldehydes, due to the rel-
ative ease deoxygenating compared to deoxygenating of
tertiary amine oxides.
enthalpy of the N–O bond, D (N–O) [1–3], as well as
the effect of substitution of various groups into the
m
pyridinic ring on the N–O bond energy.
Previously, it was found that the dissociation enthalpy
for the N–O bond in the 3-hydroxypyridine N-oxide
ꢀ
ꢃ1
D ðN–OÞ ¼ ð255:0 ꢁ 4:8Þ kJ ꢂ mol
[2]. The present
m
work reports experimental results that allow to derive the
values of the standard molar enthalpies of formation, in
the gaseous state, for 2-hydroxypyridine N-oxide, and the
ꢀ
correspondent D ðN–OÞ value. In addition to the ex-
m
perimental work, we have performed theoretical calcu-
lations for the three hydroxypyridine N-oxide isomers
using density functional theory with the B3LYP func-
tional and the 6-31G* and 6-311G** extended basis sets.
*
Corresponding author. Tel.: +351-22-6082-838; fax: +351-22-6082-
822.
E-mail address: mdsilva@fc.up.pt (M.D.M.C. Ribeiro da Silva).
0
021-9614/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jct.2003.10.001

108
M.D.M.C. Ribeiro da Silva et al. / J. Chem. Thermodynamics 36 (2004) 107–113
ꢀ
ꢃ1
2
. Experimental
empirical formula CH1:686O0:843, Dcu ¼ )16250 J ꢂ g
[7]: this value was confirmed in our laboratory. The
corrections for nitric acid formation were based on
The crystalline 2-hydroxypyridine N-oxide was syn-
ꢃ1
ꢃ3
thesized using the method Newbold and Spring [4]. The
procedure has involved a first step consisting on the
oxidation of 2-methoxypyridine with an acetic acid hy-
drogen peroxide mixture. The sample of 2-methoxy-
pyridine N-oxide produced was then refluxed in an
aqueous hydrochloric acid solution to give 2-hydroxy-
pyridine N-oxide. The sample was purified by three
crystallizations from benzene. Immediately before the
combustion measurements, the sample was submitted to
further purification by repeated vacuum sublimation.
The purity of the compound was confirmed by the av-
erage ratio of the mass of carbon dioxide recovered after
combustion to that calculated from the mass of sample:
)59.7 kJ ꢂ mol [8], for the molar energy of formation
of 0.1 mol ꢂ dm HNO3(aq) from N2, O2, and H2O(l).
The amount of substance used in each experiment was
determined from the total mass of carbon dioxide pro-
duced after allowance was made for that formed from
the cotton thread fuse and auxiliary of combustion.
The density at T ¼ 298:15 K for 2-hydroxypyridine N-
ꢃ3
oxide was estimated as q ¼ 1:0 g ꢂ cm . An estimated
pressure coefficient of massic energy: ðou=opÞ ¼ ꢃ0:2
T
ꢃ1
ꢃ1
J ꢂ g ꢂ MPa at T ¼ 298:15 K, a typical value for most
organic compounds, was assumed. For each compound,
ꢀ
the massic energy of combustion D u was calculated by
c
the procedure given by Hubbard et al. [9]. The relative
atomic masses used throughout this paper were those
recommended by the IUPAC Commission in 1999 [10].
The enthalpy of sublimation of 2-hydroxypyridine N-
oxide was measured by the ‘‘vacuum sublimation’’ drop
microcalorimetric method [11]. From the observed en-
thalpy of sublimation, the standard molar enthalpy of
sublimation at T ¼ 298:15 K was calculated using
(1.0002 ꢁ 0.0004), where the uncertainty is the standard
deviation of the mean.
The combustion experiments were performed with a
static bomb calorimeter, with a twin valve bomb, type
1
108, Parr Instrument Company. The calorimeter, sub-
sidiary apparatus and technique have been described
5,6]. Benzoic acid (Bureau of Analyzed Samples,
[
T
ꢀ
Thermochemical Standard, BCS-CRM-190 n) was used
for calibration of the bomb. Its massic energy of com-
bustion is )(26432.3 ꢁ 3.8) J ꢂ gꢃ1, under certificate
conditions. The calibration results were corrected to give
the energy equivalent e(calor) corresponding to the av-
erage mass of water added to the calorimeter: 3119.6 g.
D
H ðgÞ estimated by a group method based on
2
98:15 K
m
the values of Stull et al. [12]. The microcalorimeter
(Calvet High Temperature Microcalorimeter, SETA-
RAM HT 1000) was calibrated in situ for these mea-
surements making use of the reported molar enthalpy of
sublimation, at T ¼ 298:15 K, of naphthalene, C10H8,
ꢃ1
From six calibration experiments, e(calor) ¼ (16007.3 ꢁ
ð72:51 ꢁ 0:01Þ kJ ꢂ mol [13].
ꢃ1
0
.7) J ꢂ K , where the uncertainty quoted is the stan-
dard deviation of the mean. For all experiments, ignition
was made at T ¼ ð298:150 ꢁ 0.001) K. Combustion ex-
periments were performed in oxygen at p ¼ 3:04 MPa,
3. Experimental results
3
with 1 cm of water added to the bomb. n-Hexadecane,
Table 1 lists results for individual combustion ex-
periments, where Dm(H2O) is the deviation of the mass
of water added to the calorimeter from the average mass
assigned to e(calor), and DUR is the correction to the
standard state. The remaining terms are as previously
described [9]. As samples were ignited at T ¼ 298:15 K,
ꢀ
ꢃ1
Dcu ¼ ꢃð47158:7 ꢁ 3:2Þ J ꢂ g , was used as an auxiliary
in the combustion experiments. The electrical energy for
ignition was determined from the change in potential
difference across a capacitor when discharged through
the platinum ignition wire. For the cotton-thread fuse,
TABLE 1
Results of combustion experiments of 2-hydroxypyridine N-oxide, at T ¼ 298:15 K (p ¼ 0:1 MPa)
ꢀ
m(CO
m(cpd)/g
m(fuse)/g
m(n-Hexadecane)/g
2
, total)/g
1.74444
0.55400
0.00356
0.20627
1.38708
16.48
2.13810
0.65707
0.00372
0.23233
1.60736
16.83
1.71141
0.53700
0.00365
0.20643
1.36434
16.49
2.12135
0.74072
0.00376
0.20844
1.65472
16.83
1.97016
0.67467
0.00349
0.20203
1.54441
16.69
1.76824
0.52814
0.00377
0.20231
1.33994
16.29
DTad/K
ꢃ1
e
f
/(J ꢂ K
)
O)/g
Dm(H
2
)0.3
0.1
)0.1
0.1
0.2
0.1
ꢃDU(IBP)/J
22223.39
9727.21
57.81
25756.02
10956.27
60.41
21860.22
9734.77
59.28
26515.02
9829.67
61.06
24747.71
9527.48
56.68
21470.06
9540.70
61.22
ꢃDU(n-Hexadecane)/J
ꢃDU(fuse)/J
ꢃDU(HNO
3
)/J
ꢃDU(carbon)/J
38.63
0.00
46.69
0.00
39.76
0.00
46.69
0.00
48.48
0.00
56.54
0.00
DU(ign.)/J
1.13
12.88
22358.95
1.20
15.38
22337.45
1.11
12.54
22372.20
1.12
16.79
22357.72
1.19
15.27
22381.02
1.15
12.12
22341.58
ꢃDU
R
/J
ꢃD
c
uꢀ/(J ꢂ gꢃ1
)

M.D.M.C. Ribeiro da Silva et al. / J. Chem. Thermodynamics 36 (2004) 107–113
109
DUðIBPÞ¼ꢃfeðcalorÞþcpðH2O;lÞDmðH2OÞþe gDT þ
calculated at the optimum B3LYP/6-31G* conforma-
tions. These harmonic vibrational frequencies, after
scaling by the factor 0.9613 [18] were used to obtain
estimates of the thermal factors at T ¼ 298 K properly
corrected for the anharmocity. More accurate energies
were also obtained from single-point calculations at the
most stable B3LYP/6-311G** geometries, using the
triple-zeta correlation consistent basis set, cc-PVTZ [19]
and the B3LYP functional.
f
ad
DU(ign), where DU(IBP) is the energy associated to
the isothermal bomb process, ef is the energy of the
bomb contents after ignition and DTad is the adiabatic
temperature raise. The mean value of the standard
ꢀ
ꢃ1
massic energy of combustion, ꢃhD u i/(J ꢂ g ) ¼
c
(
22358.2 ꢁ 6.9), at T ¼ 298:15 K, the derived standard
molar energy and enthalpy of combustion and the
standard molar enthalpy of formation for the
compound, in the condensed state, are, respectively,
The obtained electronic energies for all isomers are
given in table 3. All calculations were performed using
the UK version of GAMESS [20,21].
ꢀ
ꢃ1
ꢀ
ꢃDcH ðcrÞ/
m
ꢃDcU ðcrÞ/(kJ ꢂ mol ) ¼ (2484.0 ꢁ 1.6),
m
ꢃ1
ꢀ
ꢃ1
(
kJ ꢂ mol )¼ (2483.4 ꢁ 1.6), and ꢃDf H ðcrÞ/(kJ ꢂ mol )¼
m
(
198.7 ꢁ 1.7). In accordance with normal thermochemi-
cal practice, the uncertainties assigned to the standard
molar enthalpies of combustion and formation are twice
the overall standard deviation of the mean and include
the uncertainties in calibration and in the auxiliary
4.1. Optimum conformations
The –OH substituent is found to be coplanar with the
aromatic ring in the most stable conformations of all
studied systems, a result which evidences the greater
effectiveness of the interaction of the p lone-electronic-
pair of the oxygen atom with the aromatic ring p orbi-
tals relative to the corresponding interaction involving
the oxygen r lone-electronic-pair. Moreover, for the
ortho substituted isomers the hydrogen atom of the hy-
droxyl group is found to be oriented cis relative to the
nitrogen atom of the ring, a fact which possibly facili-
tates the occurrence of further stabilizing interactions
(hydrogen bonds) between the nitrogen or the oxygen
ꢀ
ꢀ
quantities used. To derive Df H from DcH the stan-
m
m
dard molar enthalpy of formation for H2OðlÞ:
ꢃ1
ꢃð285:83 ꢁ 0:04Þ kJ ꢂ mol and for CO2ðgÞ: ꢃð393:51ꢁ
ꢃ1
0
:13Þ kJ ꢂ mol , were used [14].
g;T
cr;298:15 K
ꢀ
The observed enthalpy, D
H , at T ¼ 386 K,
m
3
86 K
ꢀ
was corrected to T ¼ 298:15 K using D
H ðgÞ/
m
298:15 K
ꢃ1
(
kJ ꢂ mol ) ¼ 11.4, estimated from the values of Stull
g
cr
ꢀ
m
et al. [12]. The enthalpy of sublimation, D H (298.15
ꢃ1
K)/(kJ ꢂ mol ) ¼ (89.4 ꢁ 0.9), was determined from six
ꢀ
experiments, thus enabling the calculation of ꢃDf H ðgÞ/
m
ꢃ1
(
kJ ꢂ mol ) ¼ (109.3 ꢁ 1.9).
TABLE 3
Calculated total energies (hartree) at the optimum B3LYP/6-311G**
geometries for hydroxypyridines and hydroxypyridines N-oxide
isomers
4
. Theoretical calculations and discussion
B3LYP/6-311G**
B3LYP/cc-PVTZ
The most stable conformations of the hydroxy-
pyridine and hydroxypyridine N-oxide isomers were
obtained from density functional theory (DFT) calcu-
lations based on the Becke 3-parameter hybrid exchange
Py
2-OH-Py
)248.349880
)323.609439
)323.600229
)323.593330
)323.597032
)323.535475
)398.795287
)398.773626
)398.778107
)398.776888
)248.377426
)323.645538
)323.637318
)323.629915
)323.633677
)323.572605
)398.841186
)398.820265
)398.824336
)398.823243
2-OH-Py (trans)
3-OH-Py
4-OH-Py
[
15] and Lee–Yang–Parr [16] correlation density func-
tional (B3LYP) and the PopleÕs split valence 6-31G* and
-311G** extended basis sets [17]. The most relevant
Py-NO
2-OH-Py-NO
2-OH-Py-NO (trans)
6
optimum geometrical parameters obtained from the
calculations using the larger basis set are displayed in
table 2. Harmonic vibrational frequencies were then
3
-OH-Py-NO
-OH-Py-NO
4
TABLE 2
Calculated B3LYP/6-311G** bond lengths (nm), r, electronic density (atomic units) at the bond critical point, qb and bond order, n, for the C–N,
N–O and C–O bonds of hydroxypyridines and hydroxypyridines N-oxide isomers
C–N
C–N
N–O
C–O
r
qb
n
r
qb
n
r
qb
n
r
qb
n
Py
0.1336
0.1340
0.1335
0.1339
0.1374
0.1356
0.1372
0.1369
0.341
0.334
0.339
0.341
0.308
0.313
0.307
0.310
1.54
1.48
1.54
1.53
1.28
1.32
1.27
1.30
1.336
1.326
1.332
1.335
1.374
1.382
1.372
1.373
0.341
0.356
0.343
0.344
0.308
0.313
0.308
0.313
1.54
1.66
1.55
1.55
1.28
1.32
1.30
1.32
2
3
4
-OH-Py
-OH-Py
-OH-Py
1.352
1.363
1.357
0.298
0.286
0.290
1.20
1.14
1.16
Py-NO
1.270
1.300
1.270
1.278
0.442
0.409
0.442
0.433
1.43
1.32
1.43
1.40
2
3
4
-OH-Py-NO
-OH-Py-NO
-OH-Py-NO
1.322
1.359
1.362
0.323
0.290
0.287
1.33
1.15
1.14

110
M.D.M.C. Ribeiro da Silva et al. / J. Chem. Thermodynamics 36 (2004) 107–113
atom for pyridine or pyridine N-oxide derivatives, re-
spectively. As a result of the N–O bond formation an
increase of the ring C–N bond lengths is observed, this
effect being most important for the C–N bond closer to
the O–H substituent. On the other hand, when com-
pared to non-substituted pyridine N-oxide, the C–N
bonds in the substituted pyridine N-oxide isomers are
generally shortened (the only exception occurs for
the C–N bond closer to the O–H substituent of 2-
hydroxypyridine N-oxide). This behavior suggests that
these bonds become stronger upon the O–H substitu-
tion. As a consequence of this strengthening of the C–N
bonds, the N–O bond is expected to become weaker in
the substituted pyridine N-oxide isomers. This follows
from the fact that a constant bond order sum must be
maintained around the nitrogen atom. We must, how-
ever, note that this weakening of the N–O bond may not
be always apparent since in the case of 2-hydroxypyri-
dine N-oxide the large stabilization afforded by the hy-
drogen bond formation can become an overshadowing
factor.
C H N–OH þ C H ! C H N þ C H OH
ð1Þ
ð2Þ
5
4
6
6
5
5
6
5
C
5
H
4
NO–OH þ C
6
H
6
! C
5
H
5
NO þ C
6 5
H OH
whose energy variations represent the stabilizing effect
of the hydroxyl group on the pyridine or pyridine N-
oxide, relative to its effect on the stability of benzene.
From the calculated reaction enthalpies, at T ¼ 298:15
K, and the known experimental enthalpies of formation
of pyridine, benzene and phenol (all the necessary ex-
perimental values were taken from reference [22]), esti-
mates of the enthalpies of formation for all isomers can
be obtained. These are shown in table 4 under the
heading Set I together with the correspondent experi-
mental values. The theoretical estimates obtained from
this type of reaction are in very good agreement with the
experimental observed ones, the largest error obtained
ꢃ1
being only 12.7 kJ ꢂ mol for the 6-311G** results and
ꢃ1
1
3.7 kJ ꢂ mol for the cc-PVTZ calculations. The cor-
ꢃ1
responding mean unsigned errors are 5.4 kJ ꢂ mol and
ꢃ1
5
.9 kJ ꢂ mol , respectively.
Also the following isodesmic reactions (Set II):
The effect of such hydrogen bond has been estimated
by optimizing the geometrics of 2-hydroxypyridine and
C
C
5
5
H
H
5
4
NO þ CH
4
! C
5
H
5
N þ CH
3
OH
ð3Þ
ð4Þ
ð5Þ
2
-hydroxypyridine N-oxide with the restriction that the
N–OH þ CH
4
! C
5
H
5
N þ CH
3
OH
O–H bond is frozen at the trans conformation. The re-
sulting final energies are also shown in table 3. The effect
of the interaction between the hydrogen atom of the
hydroxyl substituent and the nitrogen atom in 2-
C H NO–OH þ 2CH ! C H N þ 2CH OH
5
4
4
5
5
3
were considered in this work, the energy variations of
which represent the effect of the introduction of the
hydroxyl group on the pyridine or pyridine N-oxide
relative to its effect on methane. The resulting enthalpies
of formation, at T ¼ 298:15 K, are also shown in table 4.
These results also reproduce very closely the experi-
mental values, with errors not higher than 17.4
ꢃ1
hydroxypyridine is then estimated to be 24.2 kJ ꢂ mol
while the effect of the corresponding interaction with the
oxygen atom of 2-hydroxypyridine N-oxide is estimated
as 56.9 kJ ꢂ mol . The net effect upon the N–O bond in
ꢃ1
ꢃ1
that isomer should then be about 32.7 kJ ꢂ mol
.
ꢃ1
ꢃ1
kJ ꢂ mol and 11.3 kJ ꢂ mol , respectively, for the 6-
4
.2. Energetics
.2.1. Stability
311G** and cc-PVTZ calculations. The corresponding
ꢃ1
mean unsigned errors are now 7.7 kJ ꢂ mol and 7.2
ꢃ1
4
kJ ꢂ mol , respectively.
In order to study the energetic stability of these sys-
tems we have considered two different types of isodesmic
reactions. The first one, denoted by Set I in table 4, in-
volves the reactions:
4.2.2. N–O bond dissociation enthalpies
The effect of the hydroxyl substituents on the N–O
bond strengths of the substituted pyridine N-oxide
TABLE 4
Enthalpies of formation (kJ ꢂ molꢃ1) of hydroxypyridines and hydroxypyridines N-oxide isomers
ꢀ
f
Experimental D H
m
Set I
Set II
ꢀ
ꢀ
D
f
H
D
f
H
m
m
6
-311G**
cc-PVTZ
6-311G**
cc-PVTZ
2
3
4
-OH-Py
-OH-Py
-OH-Py
)78.6
)37.5
)46.6
)77.1
)37.2
)46.5
)79.7
)40.2
)49.4
111.4
)74.6
)36.4
)45.7
113.8
)(79.7 ꢁ 1.7) [23]
)(43.7 ꢁ 1.7) [23]
)(40.8 ꢁ 2.1) [23]
(124.7 ꢁ 1.9) [24]
)(109.3 ꢁ 1.9)
Py-NO
2
3
4
-OH-Py-NO
-OH-Py-NO
-OH-Py-NO
)96.6
)50.8
)47.8
)95.6
)50.8
)48.0
)112.6
)66.9
)63.8
)105.6
)60.8
)58.1
)49.5 ꢁ 4.6 [2]

M.D.M.C. Ribeiro da Silva et al. / J. Chem. Thermodynamics 36 (2004) 107–113
111
molecules must reflect the corresponding effect upon the
stability of the substituted pyridines and the homolo-
gous substituted pyridine N-oxide species. The homo-
lytic N–O bond dissociation enthalpy, at T ¼ 298:15 K,
perimental measurements presumably refer to a mixture
containing almost exclusively the most stable tautomer
simply correcting our theoretical estimates for the above
energy differences must be sufficient to produce good
comparisons with experiment. This correction brings
our estimates to very close agreement with the experi-
mental N–O bond dissociation energy of 2-hydroxy-
pyridine-N-oxide. In the case of 2-hydroxypyridine, the
oxo tautomer is also predicted to be the most stable
ꢀ
D (N–O), is defined as the enthalpy variation associated
m
with the following bond breaking reactions:
Å
C
C
5
5
H
4
NO ! C
5
H
4
N þ O
ð6Þ
ð7Þ
Å
H
4
NO–OH ! C
5
H
4
N–OH þ O
ꢃ1
form, but just by 4.0 kJ ꢂ mol (from B3LYP/6-311G**)
We have calculated such property directly from the
ꢃ1
and 2.7 kJ ꢂ mol (B3LYP/cc-pVTZ), respectively.
definition and the resulting estimates are collected in
table 5, where, for comparison purposes the corre-
sponding experimental estimates have also been in-
cluded in the last column.
Previously, it was found that the experimental disso-
ciation enthalpy for the (N–O) bond in the 3-hydroxy-
The electronic charge density transfers for all systems
have been calculated using the natural bond orbitals
(
NBO) formalism [26] with the B3LYP/6-311G** wave
functions.
The NBO analysis for the N-oxide systems describes
the oxygen atom involved in the N–O bond as having
three lone electronic pairs, thus evidencing that the N–O
bond is mostly a dative bond which is formed using the
lone electronic pair of the nitrogen atom, while the ox-
ygen atom is described having three lone electronic
pairs. The r electronic lone pair orbitals of this atom
remain essentially doubly occupied for all systems while
the corresponding lone electronic pair orbital having p
symmetry is predicted to have electronic populations of
ꢀ
ꢃ1
pyridine N-oxide, D ðN–OÞ ¼ ð255:0 ꢁ 4:8Þ kJ ꢂ mol
,
m
ꢃ1
is within the region (260.0 ꢁ 10) kJ ꢂ mol , the normal
value for the compounds with this kind of chemical
environment [2] (exception made to the nicotinamide N-
oxide and isocotinamide N-oxide [3]). For the 2-hy-
droxypyridine N-oxide, the compound reported in this
ꢀ
work, the experimental value for D (N–O) is
m
ꢃ1
(
the 2- and 3-pyridinecarboxylic acids N-oxides [2], where
278.8 ꢁ 2.3) kJ ꢂ mol . A similar trend was observed for
1.610, 1.731, 1.607 and 1.675 electrons, respectively, for
ꢀ
the D (N–O) experimental values are, respectively,
m
pyridine N-oxide and 2-, 3- and 4-hydroxypyridine N-
oxide molecules. These populations, being considerably
lower (ca. 13–20%) than its maximum occupation of two
electrons, suggest the involvement of the p lone elec-
tronic pair of the oxygen atom in the global delocalized
electronic system. Since such involvement must occur
through the N–O bond, this bond is then expected to
display a considerable double bond character.
A more accurate description of the double bond
character of the N–O bond can be achieved through a
topological analysis of the electronic density in the
framework of the theory of atoms in molecules (AIM)
ꢃ1
ꢃ1
(
275.9 ꢁ 5.1) kJ ꢂ mol and (252.8 ꢁ 2.7) kJ ꢂ mol . In
both 2-substituted-pyridine N-oxides the high values for
the dissociation enthalpy for the N–O bond can be par-
tially explained by an intramolecular hydrogen bond.
As can be observed from table 5, the predicted esti-
mates reproduce quite well the experimental values, the
ꢃ1
largest deviation, of 16.4 kJ ꢂ mol , being obtained for
2
-hydroxypyridine N-oxide. We must, however, note
that, since tautomeric equilibria occurs for the N-oxides
of hydroxypyridines, proper account of the relative
populations of the involved tautomers must be consid-
ered when comparisons between calculated and experi-
mental values of energetic properties are to be made. In
particular, we have found that 1-hydroxypyridine-2-one
is more stable than 2-hydroxypyridine-N-oxide by about
[
27]. Using the B3LYP/6-311G** wave functions, the
electronic density at the bond critical point, q , of the
b
most relevant bonds has been calculated, the results
being displayed in table 2. Correlations can then be es-
tablished for these bonds between the bond order and
qb, as [28]
ꢃ1
ꢃ1
1
8.2 kJ ꢂ mol
and 15.0 kJ ꢂ mol
from B3LYP/6-
11G** and from B3LYP/cc-pVTZ calculations, re-
3
spectively, a finding which agrees with some previous
ab initio results [25] on the same system. Since the ex-
n ¼ exp½Aðq ꢃ bÞꢄ:
ð8Þ
b
TABLE 5
Estimated N–O bond dissociation enthalpies
ꢀ
ꢃ1
)
Definition
D (N–O)/(kJ ꢂ mol
B3LYP/6-311G**
m
B3LYP/cc-PVTZ
Experimental
Py-NO
250.6
252.8
248.3
236.1
259.6
262.4
257.5
245.4
(264.9 ꢁ 2.0) [2]
(278.8 ꢁ 2.3)
2-OH-Py-NO
3-OH-Py-NO
4-OH-Py-NO
(255.0 ꢁ 4.8) [2]

112
M.D.M.C. Ribeiro da Silva et al. / J. Chem. Thermodynamics 36 (2004) 107–113
The parameters A and b for the N–O bond correlation
have been obtained from a fitting using the electronic
density at the bond critical points of HONO. For the
C–N and C–O bonds we used, respectively, the pairs of
support to Centro de Investigaßc ~a o em Qu ꢀı mica of the
University of Porto.
2 3 2 3 2
molecules CH NH, CH NH , and CH OH, H CO in
References
the fitting procedures. In all cases the electronic densities
of the simpler molecules have also been calculated at the
B3LYP/6-311G** level. These correlations allowed the
calculation of the bond order estimates for all the rele-
vant bonds of the studied systems, which are displayed
in table 2. These bond orders show that the N–O bonds
have indeed some double-bond character, giving sup-
port to our conjecture that indicates the involvement of
the oxygen p lone electronic pair in the N–O bond. In
addition, the observed general trend of bond order de-
crease upon OH substitution is compatible with a sim-
ilar decrease of the N–O bond strength.
On the other hand, if we consider the intrinsic N–O
bond dissociation enthalpy of 2-hydroxypyridine N-
oxide, obtained by subtracting from the calculated bond
dissociation enthalpy the net contribution of the hydro-
gen bond formation as estimated above, the effect of the
O–H substituent is to weaken the N–O bond of the
hydroxypyridine N-oxide, relative to that of the non-
substituted pyridine N-oxide. In addition a good corre-
lation with the calculated bond orders is now obtained.
The decrease of the N–O bond dissociation enthalpy can
be interpreted as a result of the interaction of the p elec-
tron donor ability of both O–H and O substituents when
attached to the same aromatic heterocyclic ring. As it is
well known, the topological properties of the lowest un-
occupied p molecular orbitals imply that they have large
[
[
[
[
1] W.E. Acree Jr., S.A. Tucker, M.D.M.C. Ribeiro da Silva, M.A.R.
Matos, J.M. Gonßcalves, M.A.V. Ribeiro da Silva, G. Pilcher, J.
Chem. Thermodyn. 27 (1995) 391–398.
2] M.D.M.C. Ribeiro da Silva, M.A.R. Matos, M.C. Vaz,
L.M.N.B.F. Santos, G. Pilcher, W.E. Acree Jr., J.R. Powell, J.
Chem. Thermodyn. 30 (1998) 869–878.
3] M.D.M.C. Ribeiro da Silva, J.M. Gonßcalves, S.C.C. Ferreira,
L.C.M. Silva, M.J. Sottomayor, G. Pilcher, W.E. Acree Jr., L.E.
Roy, J. Chem. Thermodyn. 33 (2001) 1263–1275.
4] G.T. Newbold, F.S. Spring, J. Chem. Soc. (1948) 1864–1866.
[5] M.D.M.C. Ribeiro da Silva, M.A.V. Ribeiro da Silva, G. Pilcher,
J. Chem. Thermodyn. 16 (1984) 1149–1155.
[6] M.A.V. Ribeiro da Silva, M.D.M.C. Ribeiro da Silva, G. Pilcher,
Rev. Port. Qu ꢀı m. 26 (1984) 163–172.
[
7] J. Coops, R.S. Jessup, K. Van Nes, in: F.D. Rossini (Ed.),
Experimental Thermochemistry, Interscience, New York, 1956
(Chapter 3).
[
[
8] The NBS Tables of Chemical Thermodynamic Properties, J. Phys.
Chem. Ref. Data 11 (Suppl. 2) (1982).
9] W.N. Hubbard, D.W. Scott, G. Waddington, in: F.D. Rossini
(
Ed.), Experimental Thermochemistry, Interscience, New York,
956 (Chapter 5).
[10] T.B. Coplen, J. Phys. Chem. Ref. Data 30 (2001) 701–712.
1
[
[
[
[
11] F.A. Adedeji, D.L.S. Brown, J.A. Connor, M.L. Leung, I.M. Paz-
Andrade, H.A. Skinner, J. Organomet. Chem. 97 (1975) 221–228.
12] D.R. Stull, E.F. Westrum, G.C. Sinke, The Chemical Thermody-
namics of Organic Compounds, Wiley, New York, 1969.
13] C.G. de Kruif, T. Kuipers, J.C. van Mittenburg, R.C.F. Schaake,
G. Stevens, J. Chem. Thermodyn. 13 (1981) 1081–1086.
14] J.D. Cox, D.D. Wagman, V.A. Medvedev, CODATA Key Values
for Thermodynamics, Hemisphere, New York, 1989.
15] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.
contributions from the p atomic orbitals of the ring at-
p
oms which are ortho or para relative to nitrogen and very
small contributions from the p atomic orbitals of the ring
[
[
16] C.T. Lee, W.T. Yang, R.G. Parr, Phys. Rev. B 37 (1998) 785.
p
atoms which are para relative to the nitrogen atom. So, for
[17] P.C. Hariharan, J.A. Pople, Theoret. Chim. Acta 28 (1973) 213.
[18] A.P. Scott, L. Radom, J. Phys. Chem. 100 (1996) 16502.
[19] T.H. Dunning Jr., J. Chem. Phys. 90 (1989) 1007;
D.E. Woon, T.H. Dunning Jr., J. Chem. Phys. 98 (1993)
2
- and 4-hydroxypyridine N-oxide each substituentÕs p
electron donor tendency is attenuated by the presence of
the other substituent. This implies that the double char-
acter of the N–O bond results attenuated in those cases
and the bond dissociation enthalpy becomes much lower
than that of the non-substituted pyridine N-oxide. On the
other hand, for 3- hydroxypyridine N-oxide the conflict-
ing p electron donor tendencies are not so effective since
the substituents try to donate p electronic charge to dif-
ferent ring p molecular orbitals. In this case a much
smaller decrease of the N–O double bond character is
expected and this bond remains essentially as strong as the
N–O bond of pyridine N-oxide.
1
358;
D.E. Woon, T.H. Dunning Jr., J. Chem. Phys. 100 (1994)
975;
2
A.K. Wilson, D.E. Woon, K.A. Peterson, T.H. Dunning Jr., J.
Chem. Phys. 110 (1999) 7667.
[
20] GAMESS-UK is a package of ab initio programs written by M.F.
Guest, J.H. van Lenthe, J. Kendrick, K. Schoffel, P. Sherwood,
with contributions from R.D. Amos, R.J. Buenker, H.J.J. van
Dam, M. Dupuis, N.C. Handy, I.H. Hillier, P.J. Knowles, V.
Bonacic-Koutecky, W. von Niessen, R.J. Harrison, A.P. Rendell,
V.R. Saunders, A.J. Stone, A.H. de Vries. The package is derived
from the original GAMESS code due to M. Dupuis, D. Spangler,
J. Wendoloski, NRCC Software Catalog, Vol. 1, Program No.
QG01 (GAMESS), 1980.
[
[
21] The initial DFT module within GAMESS-UK was developed by
Dr. P. Young under the auspices of EPSRCÕs Collaborative
Computacional Project No. 1 (CCP1) (1995–1997). Subsequent
developments have been undertaken by staff at the Daresbury
Laboratory.
Acknowledgements
22] J.B. Pedley (Ed.), Thermochemical Data and Structures of
Organic Compounds, vol. I, Thermodynamics Research Center,
College Station, TX, 1994.
Thanks are due to Fundaßc ~a o para a Ci ^e ncia e a
Tecnologia, F.C.T., Lisbon, Portugal, for financial

M.D.M.C. Ribeiro da Silva et al. / J. Chem. Thermodynamics 36 (2004) 107–113
113
[
[
23] S. Suradi, N. el Saiad, G. Pilcher, H.A. Skinner, J. Chem.
Thermodyn. 14 (1982) 45.
[26] J.P. Foster, F. Weinhold, J. Am. Chem. Soc. 102 (1980)
7211;
24] W.V. Steele, R.D. Chirico, W.B. Collier, J.A. Hossenlopp, A.
Nguyen, M.M. Strube H., Thermochemical and Thermophysical
Properties of Organic Nitrogen Compunds Found in Fossil Fuels,
NIPER, 188. Published by DOE Fossil Energy, Bartervillle
Project Office. Available from NTIS Order No DE- 87001204.
November, 1986.
A.E. Reed, F. Weinhold, J. Chem. Phys. 78 (1983) 4066.
[27] R.W.F. Bader, M. Carroll, J.R. Cheeseman, C. Chang, J. Am.
Chem. Soc. 109 (1987) 7968.
[28] R.W.F. Bader, T.S. Slee, D. Kremer, E. Kraka, J. Am. Chem.
Soc. 105 (1983) 5061.
JCT 03/075
[
25] G. LaManna, F. Biondi, J. Mol. Struct. (Theochem)287 (1993) 127.