Identification of 2-chloropyrazine oxidation products and several derivatives by multinuclear magnetic resonance

Article abstract of DOI:10.1002/mrc.1232

¹H, ¹³C, ¹⁴N and ¹⁵N NMR chemical shifts were used to prove the structures of the products of 2-chloropyrazine oxidation. It was shown that oxidation by hydrogen peroxide in acetic acid or m-chloroperbenzoic acid leads to the N4-oxide, whereas potassium persulfate in sulfuric acid gives the N1-oxide as the main product. Additionally, the results of NMR measurements of products from the nucleophilic substitution of the chlorine atom by azide anion, yielding the respective azides, and ethylation reactions of both 2-chloropyrazine N-oxides leading to the N-ethyl salts confirm the structures of both isomeric N-oxides. Protonation studies of the compounds obtained are also reported. The favoured protonation site is found to be the N atom that is not hindered by any substituents, and in some cases probably the oxygen atom of the N-oxide function. Copyright

Full text of DOI:10.1002/mrc.1232

MAGNETIC RESONANCE IN CHEMISTRY
Magn. Reson. Chem. 2003; 41: 693–698
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mrc.1232
Identification of 2-chloropyrazine oxidation products
and several derivatives by multinuclear magnetic
resonance
Piotr Cmoch∗
Institute of Organic Chemistry, Polish Academy of Sciences, Kasprzaka 44/52, 01-224 Warsaw and Pharmaceutical Research Institute, Rydygiera 8,
01-793 Warsaw, Poland
Received 28 February 2003; Revised 26 May 2003; Accepted 27 May 2003
1H, 13C, 14N and 15N NMR chemical shifts were used to prove the structures of the products of
2-chloropyrazine oxidation. It was shown that oxidation by hydrogen peroxide in acetic acid or m-
chloroperbenzoic acid leads to the N4-oxide, whereas potassium persulfate in sulfuric acid gives the
N1-oxide as the main product. Additionally, the results of NMR measurements of products from the
nucleophilic substitution of the chlorine atom by azide anion, yielding the respective azides, and
ethylation reactions of both 2-chloropyrazine N-oxides leading to the N-ethyl salts confirm the structures
of both isomeric N-oxides. Protonation studies of the compounds obtained are also reported. The favoured
protonation site is found to be the N atom that is not hindered by any substituents, and in some cases
probably the oxygen atom of the N-oxide function. Copyright  2003 John Wiley & Sons, Ltd.
KEYWORDS: NMR; 1H NMR; 13C NMR; 14N NMR; 15N NMR; pyrazine; oxidation; ethylation; protonation
2-chloropyrazine N-oxides, prepared both isomeric azides
and described them by means of H NMR data. Jovanovic
INTRODUCTION
1
During the synthesis of pyrazine derivatives, which are used
for the preparation of differently substituted azidopyrazines,
it was observed that 2-chloropyrazine N-oxide obtained
under various reaction conditions has different 1H NMR
characteristics. This observation is not surprising since 2-
chloropyrazine (1) contains two nitrogen atoms and both
may be oxygenated. An application of different oxidation
mixtures produces two different isomers: the N1- and the
N4-oxide, which are known in the literature.1,2 According
to Mixan and Pews, the treatment of 1 with potassium
persulfate in concentrated sulfuric acid at room temperature
gives the N1-oxide,1 whereas others suggested that oxidation
of 1 with hydrogen peroxide in glacial acetic acid at
also presented 15N NMR chemical shifts of differently
substituted pyrazines and their N-oxides, but the results for
both 2-chloropyrazine N-oxides were surprisingly identical.7
Taking into account the above results, there are still
doubts about the structures of both 2-chloropyrazine N-
oxides produced under different oxidation conditions. To
clarify these uncertainties, all the reactions were carried out
again and a detailed analysis of the multinuclear magnetic
resonance spectra was performed. The structure of both
possible N-oxide isomers was additionally confirmed by
analysis of the NMR spectra of the N-ethyl salts, the products
of ethylation of both isomeric N-oxides, and the azide
derivatives of pyrazine N-oxides resulting from nucleophilic
substitution of the chlorine atom by azide ion in both N-
oxides. The results of the protonation of the compounds
studied in trifluoroacetic acid (TFA) are also presented.
1,2
°
70 C results in the isomeric N4-oxide (Scheme 1). [The
numbering of atoms in all compounds studied throughout
this paper is according to 2-chloropyrazine (1).].
Klein et al.2 characterized the product obtained only by
elemental analysis and melting-point but based on these data
alone the structure could not be determined unambiguously.
A few years later, Mixan and Pews1 proposed a mechanism
for both reactions but they did not give any further analytical
data confirming the structure of both N-oxides, except for
RESULTS AND DISCUSSION
Oxidation of 2-chloropyrazine (1)
Three methods of oxidation of 2-chloropyrazine (1) were
used: potassium persulfate in concentrated sulfuric acid
(A), hydrogen peroxide in glacial acetic acid (B) and m-
chloroperbenzoic acid in methylene chloride (C) (Scheme 1).
Although in all three cases white crystalline solids were
obtained, the comparison of the 1H and 13C NMR spectra of
all reaction products (Table 1) clearly proves that reactions
B and C yield the same compound (probably 3), whereas
reaction A produces another isomer (probably 2).
1
the H NMR chemical shifts. Uchimaru et al. presented the
electron ionization mass spectra of both 2-chloropyrazine N-
oxides, but nothing was said about their fragmentation.3 A
few years later, Stanovnik et al.4 and Jovanovic5,6, using both
ŁCorrespondence to: Piotr Cmoch, Institute of Organic Chemistry,
Polish Academy of Sciences, Kasprzaka 44/52, 01-224 Warsaw,
Poland. E-mail: piocmo@icho.edu.pl
Copyright  2003 John Wiley & Sons, Ltd.

694
P. Cmoch
CH2CH3
4 N+
N4
N1
4 N
5
6
3
(C2H5)3OBF4
5
6
3
(C2H5)3OBF4
5
6
3
-
-
BF4
Cl
BF4
Cl
+
2
2
N
2
Cl
N1
1
7
1
CH3CH2
6
A. K2S2O8, H2SO4
B. H2O2, CH3COOH
C. m-ClPhCOOOH, CH2Cl2
O
O
O
4 N
N4
N 4
5
6
4
4 N
3
2
N
-
5
3
5
6
3
N3
1. NH2NH2
3
5
6
3
5
6
2. NO2 , H+
-
acetone, aq
6
N
N 1'
N 1'
2
N
O
N
O
2
Cl
2
1
N
1
N
N
1'
2'
2
1
N1
Cl
1
N
N
2
2'
2'
N
3
5T
3'
4
5A
(C2H5)3OBF4
CH2CH3
N 3'
3'
N
(C2H5)3OBF4
CH2CH3
O
4N+
4 N
O
4 N+
4 N
5
6
3
5
6
3
-
-
BF4
Cl
BF4
Cl
5
6
3
5
3
-
-
BF4
Cl
+
+
BF4
Cl
N
O
2
2
N
1
+
6
1
2
N1
2
N1
9
8
O
CH2CH3
CH3CH2
10
11
Scheme 1. Scheme of different reactions of 2-chloropyrazine 1 and its N-oxides 2 and 3. Throughout this paper atoms of all
compounds are numbered in accordance to 1.
Additionally, in the 1H NMR spectrum of the product of
reaction A, small quantities of the product of reaction B D C
(ca 7%) are also present. To explain which isomeric N-oxide
was produced under the respective oxidation conditions,
detailed analysis of the multinuclear magnetic resonance
chemical shifts and coupling constants, presented in Table 1,
was performed.
Table 1. 1H, 13C, 14N and 15N NMR data (in acetone, 303 K)
for 1 and the products of reactions A, B and C (2 and 3)
Product of
reaction A: 2
Product of
reactions B and C: 3
1b
H3
H5
H6
J(H3,H5)
J(H5,H6)
J(H3,H6)
C2
C3
C5
8.69
8.60
8.47
0.45
2.52
1.45
8.75
8.45
8.39
—
4.05
0.60
139.7
148.5
146.3
135.9
193.5
188.8
194.4
67.7
8.39
8.21
8.35
1.50
4.20
1
The assignment of H, 13C and nitrogen NMR chemical
d
shifts for both N-oxides obtained in reactions A, B and C
was made using results of 2D 1H–13C g-HSQC, 1H–13C and
1H–15N g-HMBC experiments (for NMR nomenclature, see
Experimental.), and also comparison of the multinuclear
magnetic resonance data for N-oxides obtained with 2-
chloropyrazine (1).
0.70
149.8
151.9
134.3
134.6
147.3
199.2
195.4
191.5
74.4
145.4
143.6
144.8
193.0
186.3
187.0
64.0
C6
It is known that after oxidation of different substituted
pyridines, nitrogen nuclei are shielded by ca 30 ppm, whereas
shieldings of carbon nuclei directly bonded to pyridine
nitrogen increase by ca 10 ppm in comparison with the
parent compounds.8 Although assignment of the 1H and
13C NMR chemical shifts of the products of reactions A,
J(C3,H3)
J(C5,H5)
J(C6,H6)
J(C2,C3)a
J(C5,C6)a
N1
53.4
59.1
ꢀ74.6e
59.8
ꢀ55.5
[220]c
ꢀ35.7
[230]c
ꢀ85.1c
[370]c
ꢀ65.8
[30]c
B and C was straightforward using results of the 1H–13
C
g-HSQC/g-HMBC experiments and the above-mentioned
information, the assignment of nitrogen signals was not an
easy task. To assign the nitrogen signals unambiguously,
the 14N NMR signal half-width was used. It is known
that the half-width of 14N NMR signals of the N-oxide
function is ca 10 times smaller than that of ‘non-oxygenated’
atoms.9,10 In the case of the product of reaction A, both 14N
signals overlap and it is difficult to determine their half-
width accurately, but visually the signal at ꢀ74.6 ppm is
sharper than that at ꢀ72.8 ppm. This means that the signal at
N4
ꢀ72.8e
a 1J(13C,13C) coupling constants measured in DMSO.
b 3J(C3,C6) D 14.3, 3J(C2,C5) D 18.9, 2J(C3,C5) D 7.0, 2J(C2,C6) D
6.0 Hz.
c Half-widths of the 14N NMR signals.
d Not determined owing to broadening of the signal caused by
the influence of oxygen.
e Signals overlap in the 14N NMR spectrum.
Copyright  2003 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2003; 41: 693–698

Identification of 2-chloropyrazine oxidation products
695
ꢀ74.6 ppm belongs to nitrogen possessing an oxygen atom,
whereas that at ꢀ72.8 ppm corresponds to the free nitrogen
atom. For the product of reactions B and C, the previously
mentioned procedure can be easily repeated. The nitrogen
signal of product B D C at ꢀ65.8 ppm (half-width ca 30 Hz)
corresponds to oxygenated nitrogen, whereas the signal at
ꢀ85.1 ppm (half-width ca 370 Hz) originates from the ‘non-
oxygenated’ nitrogen atom. The half-widths of 14N NMR
signals and results of 1H–15N g-HMBC experiments for both
N-oxides only support the assignment of the nitrogen shifts,
but do not allow a differentiation between both N-oxide
structures obtained.
To determine the structure of both oxidation products
of 1, a detailed analysis of the 13C NMR data was carried
out. Comparison of 13C NMR chemical shifts of 1 and the
products of reactions A and B D C leads to the following
conclusion. Nuclei of the quaternary C-2 and tertiary C-6
atoms of the product of reaction A and both tertiary C-3
and C-5 carbons of the product of reactions B D C are more
shielded by ca 10 ppm than the same nuclei in 1.
of reactions B and C) or sodium azide (product of reaction
A), yield white crystalline solids, which first were measured
in acetone solution by 1H NMR spectroscopy. In the 1H NMR
spectrum of 4 obtained from the product of reaction A only
one set of signals (three signals) was observed, confirming
the absence of the azide–tetrazole equilibrium in this case. In
the case of 5 obtained from the product of reactions B and C,
the existence of the azide–tetrazole equilibrium is obvious,
1
as the H NMR spectrum contains two sets of signals (six
signals). The existence of both tautomeric forms (azide A and
tetrazole T) is confirmed unambiguously by analysis of the
13C, 14N and 15N NMR spectra. Especially the positions of
the 15N NMR signals are of great significance (Table 2). The
presence of signals at ꢀ276.1 and C24.9 ppm clearly proves
the existence of both forms in acetone solution.15,16 The
same compound dissolved in chloroform produces similar
multinuclear spectra, but the integration of 1H, 13C and 15
N
signals is different to that obtained in acetone solution. A
1
comparison of the H NMR integrals of 5 taken in acetone-
d6 and CDCl3 indicates that in the less polar chloroform
the content of the azide form is higher than in acetone
solution. This phenomenon relating to solvent polarity is
well known.13–16
From the point of view of NMR parameter changes,
substitution of the chlorine atom by the azide anion causes
a shielding increase of the respective nuclei. In the case of
4 and the azide form of 5, nuclei C-3 and N-1 are more
shielded than the parent chloropyrazine N-oxides 2 and 3
by ca 8 and 10–20 ppm, respectively. A closure of the azide
Based on this information and the conclusions from
nitrogen NMR, it can be assumed that the product of reaction
A has the structure of the 2-chloropyrazine N1-oxide 2,
whereas the two other products (from reactions B and C)
have the same structure of the 2-chloropyrazine N4-oxide 3.
In the assignment of the 1H, 13C and nitrogen NMR chem-
ical shifts with the use of 1H–13C g-HSQC, 1H–13C, 1H–15
N
g-HMBC and additionally INADEQUATE experiments for
both N-oxides 2 and 3, it was observed that the values of
4J(H,H) are larger than 5J(H,H) and behave in the oppo-
site way to those for 1 and other substituted pyrazines.11
Additionally, production of the 2-chloropyrazine N-oxides
is associated with changes of other 13C NMR parameters. The
introduction of an oxygen on to the nitrogen atom causes an
Table 2. 1H, 13C, 14N and 15N NMR data (in acetone, 303 K)
for the azide derivatives of the pyrazine N-oxides 4 and 5
4
5A
7.88
8.30
8.00
1.50
4.20
0.70
5T
1
increase in the 1J(13C,13C) and the J(1H,13C) couplings of ca
4–10 and ca 3–8 Hz, respectively (Table 1).
H3
H5
H6
8.25
8.28
8.22
—
4.10
0.75
9.13
9.23
8.06
1.80
5.90
0.85
In order to determine the structures of both N-oxides
unambiguously, the detailed analysis of the multinuclear
data for products of nucleophilic substitution of the chlorine
atom and the ethylation reaction of both N-oxides were used.
a
J(H3,H5)
J(H5,H6)
J(H3,H6)
C2
Azide derivatives of the 2-chloropyrazine N-oxides
2 and 3
141.5
156.2
148.6
C3
C5
140.2
143.3
125.9
132.1
126.1
132.0
The 2-chloropyrazine N-oxides 2 and 3, similarly to 2-
halopyridines and halopyrazines, react with nucleophiles,
yielding the corresponding substituted pyrazine N-oxides.
For instance, both N-oxides 2 and 3 react with hydrazine
hydrate yielding the hydrazines, and with sodium azide
producing the respective azides (Scheme 1). Treatment of
hydrazines with sodium nitrite in acidic solutions leads
to azide derivatives, which may exhibit azide–tetrazole
tautomerism.12–16 Analysis of both N-oxide structures stud-
ied (2 and 3) suggests that after the above-mentioned
reactions such equilibrium could exist only in the case of
the derivative of 2-chloropyrazine N4-oxide (3). In the case
of the azide derivative of 2 (2-chloropyrazine N1-oxide) such
tautomerism is not possible, because the nitrogen atom is
blocked by oxygen (Scheme 1).
C6
134.1
146.5
122.9
N1
ꢀ84.5
ꢀ108.5
[220]b
ꢀ62.7
[60]b
ꢀ141.8
[110]b
ꢀ76.4
[30]b
[50]b
N4
ꢀ69.8
[—]c
N10
N20
N30
ꢀ295.3
[550]b
ꢀ146.0
[15]b
ꢀ276.1
[380]b
ꢀ142.6
[10]b
ꢀ67.7
[—]c
C24.9
[380]b
ꢀ30.1
[370]b
ꢀ134.3
ꢀ138.8
[80]b
[—]c
a Not determined owing to broadening of the signal caused by
the influence of oxygen.
b The half-widths of the 14N NMR signals.
c Not determined because of 14N signal overlap.
The products of oxidation of 1, after treatment with
hydrazine followed by reaction with sodium nitrite (product
Copyright  2003 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2003; 41: 693–698

696
P. Cmoch
group to the tetrazole ring (in the case of azide 5) causes a
shielding increase of N-1 and N-4 nuclei by ca 30 and 15 ppm,
respectively. Additionally, the closure of the azide group to
the tetrazole moiety is associated with an increased shielding
(by ca 23 ppm) of the C-6 nucleus.
The above analysis of the multinuclear data for the
product of nucleophilic substitution of the chlorine atom
in both isomeric N-oxides completely confirms the path of
oxidation reactions A, B and C for 1, i.e. potassium persulfate
in sulfuric acid at room temperature yields the N1-oxide 2
whereas hydrogen peroxide and m-chloroperbenzoic acid
give the N4-oxide 3.
Similar analysis of the results of ethylation of both
2-chloropyrazine N-oxides 2 and 3 suggests creation of
four isomers, 8–11 (Scheme 1), but the NMR data indicate
existence of only three N-ethyl salts for both N-oxides. After
ethylation of 2 only one set of signals is observed in the 1H,
13C and 15N NMR spectra. The increase of the C-3 and C-5
shieldings by ca 6 ppm and the strong 15N shielding increase
of the N-4 nucleus by ca 100 ppm indicate the existence of
N4-ethyl-2-chloropyrazinium tetrafluoroborate N1-oxide (9).
No traces of the second isomeric N-ethyl salt 8 are observed
in multinuclear spectra of the ethylation product because
the bulky chlorine atom probably hinders access of the ethyl
group to the oxygen atom.
In the second case, where 3 (product of reactions B
and C) was ethylated, two sets of signals were present in
the multinuclear spectra of the ethylation product. Based
on the 1H NMR integrals, the percentages of both salts
were determined. The comparison of the differences in the
multinuclear NMR data presented in Table 3 for the less
abundant form (25%) with those of 3 and their comparison
with differences in multinuclear data between 1 and 7,
presented above, indicate the existence of the N1-ethyl
salt 11.
The analysis of multinuclear NMR data of the second,
more abundant salt (75%) leads to the conclusion that 2-
chloropyrazine N4-oxide (3) is also ethylated at the oxygen
atom. Direct proof of this statement is obtained from strong
changes in the shieldings of several nuclei that are not
comparable to the remaining N-ethyl salts 7, 9 and 11. The
decrease in the shieldings of N-1 by ca 50 ppm, of the CH2
carbon by ca 25 ppm and a modest shielding increase for
the N-4 nucleus by ca 45 ppm, when compared with those
of 3, strongly indicate the presence of the second isomer,
N4-ethoxy-2-chloropyrazinium tetrafluoroborate (10).
Ethylation of the 2-chloropyrazine N-oxides 2 and 3
Additional evidence for the 2-chloropyrazine N-oxide struc-
tures 2 and 3 is obtained from the analysis of the multinuclear
spectra of ethylation products of the 2-chloropyrazine N-
oxides with triethyloxonium tetrafluoroborate (Scheme 1).
Ethylation of 1 leads theoretically to two different
products, the N1-ethyl (6) and N4-ethyl (7) salts. A detailed
analysis of the 1H, 13C and 15N NMR chemical shifts of
the product (Table 3) indicates that only one compound is
obtained. The strong shielding increase of the N-4 nucleus
in this compound by ca 100 ppm and the decrease in
shielding of the N-1 nucleus by ca 20 ppm, compared with
the 15N NMR results for 1 (Table 1), prove the existence
of the N4-ethyl 2-chloropyrazinium salt 7. The ethylation
causes also a decrease in the shielding of all protons by ca
0.6–0.8 ppm, a shielding increase for carbon nuclei (C-3 and
C-5) directly bonded to ethylated nitrogen by ca 6 ppm and
simultaneously a shielding decrease for two other carbon
nuclei by the same value.
Table 3. 1H, 13C and 15N NMR data (in acetone 303 K) for the
N-ethyl salts 7 and 9–11
Protonation of 2-chloropyrazine (1), its N-oxides 2
and 3 and the azide derivatives of the pyrazine
N-oxides 4 and 5
7
9
10
11
Compounds 1–5 were dissolved in TFA in order to determine
their protonation sites. Each of them contains at least two
nitrogen atoms and thus may be protonated at two different
positions. Compounds 2–5 additionally possess an oxygen
atom, which may also be protonated. To determine the site
of protonation, multinuclear spectra were registered and the
results are given in Table 4.
The NMR data, especially the 15N NMR chemical shifts,
indicate that 1 and 2 undergo protonation at N-4. This is
demonstrated by the strong shielding increase of the N-
4 nucleus by ca 80–90 ppm compared with its position in
aprotic solution, whereas the N-1 nucleus experiences only a
small opposite effect (ca 7–10 ppm).
In the case of protonation of 3 (N4-oxide), both nuclei
are only slightly shielded compared with their positions in
acetone (by ca 1 and 15 ppm for N-1 and N-4, respectively).
These small effects suggest protonation at the oxygen rather
than at nitrogen, although in the case of the derivatives of
the pyridine N-oxides, the nitrogen nucleus is shifted by ca
40 ppm after protonation.8
H3
H5
H6
CH2
CH3
J(H3,H5)
J(H5,H6)
J(H3,H6)
C2
C3
C5
C6
CH2
CH3
N1
9.46
9.22
9.36
4.94
9.55
9.05
9.03
4.76
1.73
2.0
9.86
9.53
9.47
5.05
1.57
0.8
3.9
1.6
154.8
135.5
133.2
152.2
81.6
9.20
8.81
9.10
4.83
1.67
2.2
1.80
a
—
3.2
5.5
5.6
a
a
a
—
—
—
154.0
139.4
136.9
151.2
59.8
143.5
142.3
140.0
140.1
58.0
147.2
140.4
139.2
142.5
55.8
15.9
15.7
13.2
14.4
ꢀ35.2
[720]b
ꢀ141.7
[50]b
ꢀ63.5
ꢀ37.3
[800]b
ꢀ110.3
[310]b
ꢀ181.1
[120]b
ꢀ53.9
[90]b
[70]b
N4
ꢀ179.0
[45]b
a Not determined owing to broadening of the signal caused by
the oxygen or introduction of the ethyl group.
b The half-widths of the 14N NMR signals.
Compound 5 in TFA solution exists in azide–tetrazole
equilibrium. Based on 1H NMR integrals, the content of
Copyright  2003 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2003; 41: 693–698

Identification of 2-chloropyrazine oxidation products
697
Table 4. 1H, 13C and 15N NMR chemical shifts (in TFA, 303 K)
for the protonated compounds 1–5
Moreover, the presence of the bulky atoms such as chlorine
at position 2 of pyrazine is responsible for differentiation
of the nitrogen shielding effects after oxidation (ca 20 and
40 ppm for both nitrogens of 2, whereas for 3 these changes
are similar and ca 30 ppm).
Large and regular changes of the NMR parameters,
especially 13C and 15N NMR chemical shifts, resulting from
the structural modifications caused by alkylation and the
nucleophilic substitution of the chlorine atom allow the
determination of the structural of the derivatives obtained.
They also, indirectly, confirm the structural differentiation
among the N-oxides, depending on the conditions of the
oxidation of the differently substituted pyrazines.
1
2
3
4
5A
8.10
8.20
8.57
5T
9.59
8.30
9.08
H3
H5
H6
C2
C3
C5
C6
N1
8.65
8.57
8.78
8.97
8.70
8.72
8.52
8.44
8.60
8.42
8.41
8.61
151.9
144.3
152.0
146.9
131.0
132.4
157.1
145.2
138.6
135.8
146.9
ꢀ48.6
140.9
138.0
138.1
ꢀ67.3
134.3
133.3
146.2
ꢀ86.1
ꢀ81.0
—
124.7
128.3
146.5
ꢀ94.5 ꢀ138.6
129.7
129.6
122.7
137.6
ꢀ75.9
ꢀ179.6
ꢀ287.8
ꢀ149.2
N2 ꢀ116.9 ꢀ166.0
ꢀ93.5
ꢀ272.8
ꢀ149.1
ꢀ136.3
ꢀ94.8
N10
N20
N30
—
—
—
—
—
—
—
b
EXPERIMENTAL
—
—
C14.4
ꢀ29.2
a
—
Compounds
The compounds studied were prepared according to known
methods presented in Scheme 1.
a Not observed in the 14N NMR spectrum and the 1H–15
N
2-Chloropyrazine N1-oxide (2) was obtained according to the
g-HMBC experiment.
literature1 and purified by liquid chromatography [C6H14 –CHCl3
b Not observed in the 15N NMR invgate and the 14N NMR
°
(70 : 30)]. EI-MS (at 33 C): m/z 132 (33.6%), 130 (100%), 116 (3.7%),
spectra, or in the 1H–15N g- HMBC experiment.
114 (11.1%), 79 (15.8%), 77 (6.5%), 75 (15.4%), 68 (12.7%), 60 (15.7%),
52 (15.3%). IR (CHCl3ꢀ: 3113 (m), 1581 (m), 1498 (w), 1450 (s), 1399
(s), 1318 (s), 1282 (w), 1179 (m), 1153 (w), 1121 (m), 1043 (m), 912 (w),
875 (m), 868 (m), 577 (w), 565 (m), 545 cmꢀ1 (m). Analysis: calculated
for C4H3N2OCl, H 2.30, C 36.78, N 21.46; found, H 2.15, C 36.45, N
21.22%. The 1H, 13C, 14N and 15N NMR data in acetone-d6 and TFA
solutions are collected in Tables 1 and 4, respectively.
both forms determined as 21% azide and 79% tetrazole
form. This observation is in agreement with the known
tendency of the tetrazole ring to open in acidic conditions.
However, small changes of the 15N NMR chemical shifts
of the N-1 and N-4 nuclei in both tautomeric forms provide
the proof that proton–nitrogen interaction is very weak. This
statement is also supported by small changes of the 13C NMR
chemical shifts for 5 after protonation. For 1 and 2, where
the proton–nitrogen interaction is very strong, the increase
in 13C shieldings (of carbon nuclei in direct neighborhood of
‘protonated’ nitrogen N-4) is very significant (ca 8 ppm). In
the case of 3 and its azide derivative 5 in both tautomeric
forms A and T, this increase is small (ca 3 ppm). Based on
these observations, it can be stated that 3 and both tautomeric
forms of 5 are protonated at the oxygen atom of the N—O
function rather than not protonated at all.
2-Chloropyrazine N4-oxide (3) was obtained according to the
2
°
literature. EI-MS (at 33 C): m/z 132 (34.1%), 130 (100%), 116 (5.3%),
114 (15.7%), 79 (20.8%), 77 (12.6%), 75 (30.0%), 68 (6.7%), 60 (12.3%),
52 (24.3%). IR (CHCl3ꢀ: 3123 (m), 1582 (s), 1543 (w), 1496 (m), 1463
(m), 1436 (s), 1410 (s), 1330 (s), 1317 (m), 1236 (m), 1173 (w), 1115
(s), 1092 (s), 1001 (s), 939 (s), 844 (m), 823 (w), 615 (w), 543 cmꢀ1
(w). Analysis: calculated for C4H3N2OCl, H 2.30, C 36.78, N 21.46;
found, H 2.21, C 36.56%, N 21.34%. The 1H, 13C, 14N and 15N NMR
data in acetone-d6 and TFA solutions are collected in Tables 1 and 4,
respectively.
2-Azidopyrazine N1-oxide (4): the product obtained according
to the literature6 was contaminated with 2 (50 : 50), so it was
characterized only by NMR spectroscopy. The 1H, 13C, 14N and
15N NMR data in acetone-d6 and TFA solutions are collected in
Tables 2 and 4, respectively.
2-Azidopyrazine N4-oxide (5A) was obtained according to the
literature.4,5 EI-MS (at 109 C): m/z 137 (100%), 83 (2.4%), 82 (52.7%),
°
66 (10.8%), 65 (2.8%), 53 (33.7%), 52 (45.8%). IR (CHCl3ꢀ: 2400 (w),
2247 (w), 2205 (w), 2161 (w), 2141 (s), 2089 (w), 1650 (w), 1593 (s),
1505 (m), 1489 (w), 1451 (s), 1427 (m), 1343 (w), 1289 (m), 1239 (s),
1110 (m), 1005 (m), 967 (m), 930 (w), 842 (m), 643 (w), 627 cmꢀ1
(w). Analysis: calculated for C4H3N5O, H 2.19, C 35.04, N 51.09;
found, H 2.10, C 34.86, N 50.86%. The 1H, 13C, 14N and 15N NMR
data in acetone-d6 and TFA solutions are collected in Tables 2 and 4,
respectively.
In contrast to weak protonation of both tautomeric forms
of the 2-azidopyrazine N4-oxide 5, the 2-azidopyrazine N1-
oxide 4 is fully N-protonated. Direct evidence of protonation
is provided by the strong 15N shielding increase of the N-4
nucleus by ca 100 ppm. Protonation is also supported by a
strong 13C shielding increase of the C-3 and C-5 nuclei by ca
8 ppm.
Preparation of the N-ethyl salts [7, 9–11] was carried out
according to the literature.17,18 The 1H, 13C, 14N and 15N NMR
data in acetone-d6 are collected in Table 3.
Conclusions
The analysis of NMR data presented above provides
sufficient evidence to conclude that oxidation of the 2-
chloropyrazine with potassium persulfate (reaction A)
leads to the 2-chloropyrazine N1-oxide 2, whereas the
two remaining reactions (with hydrogen peroxide and m-
chloroperbenzoic acid) lead to the N4-oxide isomer 3.
Compared with the substrate, the N-oxides can be
characterized by the shielding increase of the nitrogen nuclei
(by ca 30 ppm), the shielding increase of the carbon nuclei of
the atom located next to the oxygenated nitrogen atom (by ca
10 ppm) and the decrease in the value of the 14N NMR signal
half-width for the oxygenated nitrogen nucleus (4–10-fold).
Spectra
1
The H, 13C, 14N and 15N NMR spectra were measured at
303 K on a Bruker DRX 500 spectrometer equipped with
a TBI 500SB H-C/BB-D-05 Z-G probehead, operating at
500.133, 125.773, 36.141 and 50.690 MHz for 1H, 13C, 14N and
15N nuclei, respectively. Standard procedures were used to
record the 1H, 13C, 14N and 15N NMR spectra employing,
among others, power-gated and inverse-gated (invgate)
decoupling sequences. The J(13C,13C) spin couplings were
obtained using the INADEQUATE sequence.
Two-dimensional 1H–13C g-HSQC (gradient-selected
heteronuclear single quantum coherence; C,H correlation
1
Copyright  2003 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2003; 41: 693–698

698
P. Cmoch
3. Uchimaru F, Okada S, Kosasayama A, Konno T. J. Heterocycl.
Chem. 1971; 8: 99.
4. Stanovnik B, Tisˇler M, Trcˇek N, Vercˇek B. Vest. Slov. Kem. Drus.
1981; 28: 45.
5. Jovanovic M. Heterocycles 1983; 20: 1987.
6. Jovanovic MV. Heterocycles 1984; 22: 1115.
7. Jovanovic MV. Spectrochim. Acta, Part A 1984; 40: 637.
8. Yavari I, Roberts JD. Org. Magn. Reson. 1979; 12: 87.
9. Witanowski M, Stefaniak L, Webb GA. Annu. Rep. NMR
Spectrosc. 1981; 11B: 84.
10. Witanowski M, Stefaniak L, Webb GA. Annu. Rep. NMR
Spectrosc. 1986; 18: 135.
11. Palamidessi G, Vigevani A, Zarini F. J. Heterocycl. Chem. 1974;
11: 607.
12. Butler N. Chem. Ind. (London) 1973; 371.
13. Tisˇler M. Synthesis 1973; 123.
14. Pocionok VJ, Avramenko LF, Grigorienko TF, Skopienko WN.
Usp. Khim. 1975; 44: 1028.
via double INEPT transfer in the phase-sensitive mode) and
1
1H–13C as well as H–15N g-HMBC (gradient-selected het-
eronuclear multiple bond coherence; long-range correlation
experiment) were performed using standard Bruker soft-
ware and the following parameters: the spectral widths in
1
F2 and F1 were ca 10 ppm for H, 80–160 ppm for 13C and
100–400 ppm for 15N. The relaxation delay was usually ca
2.0 s, the refocusing delay in the g-HSQC experiment was
ca 1.20 ms and delays for long-range evolutions were ca 80
1
1
and 80–320 ms for H/13C g-HMBC and H/15N g-HMBC
experiments, respectively. The 2D spectra were acquired as
2048 ð 512 or 1024 ð 256 hypercomplex files, with 4–8 tran-
sients for each 512 or 256 time increments, using appropriate
90 and 180 pulse widths for the H, 13C and 15 N channels.
1
°
°
For the H and 13C spectra in acetone-d6, internal TMS
1
15. Cmoch P, Wiench JW, Stefaniak L, Webb GA. J. Mol. Struct. 1999;
510: 165.
16. Cmoch P, Korczak H, Stefaniak L, Webb GA. J. Phys. Org. Chem.
1999; 12: 470.
17. Messmer A, Hajos G, Juhasz-Riedl Z, Sohar P. J. Org. Chem. 1988;
53: 973.
18. Cmoch P, Wiench JW, Stefaniak L, Sitkowski J. J. Mol. Struct.
1999; 477: 119.
was used as the chemical shift standard, whereas external
nitromethane was applied as the standard for the 14N and
15N NMR measurements. In the case of the 1H and 13C
spectra in TFA solutions, external DMSO-d6 was used as the
chemical shift standard. The concentrations of all solutions
were between 0.05 and 0.2 mol dmꢀ3 and only in the case
of INADEQUATE experiments were the concentrations ca
2 mol dmꢀ3
.
REFERENCES
1. Mixan CE, Pews RG. J. Org. Chem. 1977; 42: 1869.
2. Klein B, Hetman NE, O’Donnell ME. J. Org. Chem. 1963; 28: 1682.
Copyright  2003 John Wiley & Sons, Ltd.
Magn. Reson. Chem. 2003; 41: 693–698