A study of the thermal decomposition of 2-azidoethanol and 2-azidoethyl acetate by ultraviolet photoelectron spectroscopy and matrix isolation infrared spectroscopy

Article abstract of DOI:10.1021/jp020625e

The thermal decomposition of 2-azidoethanol and 2-azidoethyl acetate were studied by matrix isolation IR spectroscopy and real-time UV photoelectron spectroscopy. The products detected in a flow system at different temperatures (CH₂NH, H₂CO, N₂, CO, and HCN from N₃CH₂CH₂OH and C₂H₄, CH₂NH, HCN, CO₂, and N₂ from N₃CH₂COOCH₂CH₃) allowed mechanisms for decomposition to be proposed. Ab initio calculations were performed for these azides, and application of Koopmans' theorem to the computed orbital energies yielded vertical ionization energies that agreed with experimental values. Two main mechanisms of decomposition of organic azides of the type considered began to emerge. 2-Azidoacetic acid and 2-azidoehtyl acetate decomposed via a concerted process through a cyclic transition state to give the products, while 2-azidoethanol and azidoacetone decomposed via a stepwise mechanism through imine intermediates, which decomposed to give the products via two possible pathways.

Full text of DOI:10.1021/jp020625e

9968
J. Phys. Chem. A 2002, 106, 9968-9975
A Study of the Thermal Decomposition of 2-Azidoethanol and 2-Azidoethyl Acetate by
Ultraviolet Photoelectron Spectroscopy and Matrix Isolation Infrared Spectroscopy†
N. Hooper, L. J. Beeching, J. M. Dyke,* A. Morris, and J. S. Ogden
Department of Chemistry, UniVersity of Southampton, Southampton SO17 1BJ, U.K.
A. A. Dias,‡ M. L. Costa,‡ M. T. Barros,§ M. H. Cabral,‡ and A. M. C. Moutinho‡
Faculdade de Ciencias e Tecnologia, UniVersidade NoVa de Lisboa, 2829-215 Caparica, Portugal
ReceiVed: March 5, 2002; In Final Form: June 14, 2002
The thermal decompositions of 2-azidoethanol and 2-azidoethyl acetate have been studied by matrix isolation
infrared spectroscopy and real-time ultraviolet photoelectron spectroscopy. The products that were detected
in a flow system at different temperatures (CH2NH, H2CO, N2, CO, and HCN from N3CH2CH2OH and C2H4,
CH2NH, HCN, CO2, and N2 from N3CH2COOCH2CH3) allowed mechanisms for decomposition to be proposed.
The experimental evidence obtained is consistent with 2-azidoethyl actetate decomposing via a concerted
mechanism, similar to that found previously for azidoacetic acid, whereas the 2-azidoethanol decomposition
is consistent with a stepwise decomposition mechanism as observed previously for azidoacetone.
Introduction
thermal decomposition of azidoacetic acid and azidoacetone
were investigated using ultraviolet photoelectron spectroscopy
and matrix isolation infrared spectroscopy. In the case of
azidoacetic acid,17 the results were shown to be consistent with
a single decomposition pathway involving the ejection of N2
and the simultaneous formation of methanimine (CH2NH) and
CO2. At higher temperatures, HCN was produced. These results
were supported by later ab initio and DFT calculations,19 which
showed that decomposition occurs via a concerted reaction that
proceeds via a five-membered transition state leading to the
products (N2, CH2NH, and CO2). The decomposition of azi-
doacetone was found to be more complicated,18 producing not
only the products expected from the analogous pathway (N2,
CH2NH, HCN, and ketene, CH2CO) but also significant yields
of CO and acetaldehyde. It was suggested that alternative radical
pathways might be responsible for the observed distribution of
products.
The intrinsic instability of organic azides makes their proper-
ties difficult to measure, and experimental work has to be carried
out with care. Nevertheless, the study of the mechanisms of
their decomposition is of considerable interest both from a
fundamental viewpoint and because azides play important roles
in heterocycle synthesis,1-3 biological and pharmaceutical
processes,4-6 the preparation of semiconductor materials,7,8 and
as high-energy density materials used in solid propellants.9,10
It is generally believed that the initial stage in the thermal
decomposition of an organic azide such as R2CHN3 is the release
of molecular nitrogen and the formation of a nitrene, R2CHN.
The thermal decompositions of a number of alkyl azides,
including CH3N3, have been studied several years earlier in the
gas phase by Bock and Dammel11-15 using UV photoelectron
spectroscopy (PES). The conclusion from this work is that, for
the azides studied, N2 evolution is accompanied by a 1,2-
hydrogen shift to form an imine R2CdNH, which can undergo
further decomposition to produce simple hydrocarbons, HCN
and H2.
More recently, Dianxun et al.16 reported the UV photoelectron
spectrum of CH3N obtained by pyrolysis of CH3N3. In Dianx-
un’s work, CH3N was produced by passing CH3N3 mixed with
NO through a heated plug of molecular sieve positioned 1-2
cm above the photon source of the spectrometer. CH3N
photoelectron spectra were observed at a molecular sieve
temperature of 145 °C. It is clear from the conditions used that
CH3N was not produced in this work from a gas-phase
decomposition but from decomposition of methyl azide on the
surface of the molecular sieve in the presence of NO.16
The work presented here is part of a series of studies17,18 on
organic azides performed to investigate the mechanisms of their
thermal decomposition. In these two previous studies,17,18 the
This present paper describes related studies on the pyrolysis
of 2-azidoethanol and 2-azidoethyl acetate using the techniques
of real-time ultraviolet photoelectron spectroscopy and matrix
isolation infrared spectroscopy, supported by ab initio molecular
orbital calculations. The main aim is to investigate the mech-
anism of thermal decomposition of these azides through the
observation of the reactants, intermediates, and products at
different pyrolysis temperatures. Also included in this study is
the measurement of other spectroscopic properties for both
parent azides.
Experimental Section
Sample Preparation. Samples of 2-azidoethanol and 2-azi-
doethyl acetate were prepared as follows: 2-Azidoethanol was
prepared from 2-bromoethanol (95%, Aldrich) and sodium azide
(99%, Aldrich), mixed in a molar ratio of 1:2.5. Solid sodium
azide (2.05 g) was added slowly to a vigorously stirred solution
of 2-bromoethanol (2.62 g) in distilled water (10 mL) at 0 °C.
The solution was then allowed to warm to room temperature
and was stirred for 4 h. More solid sodium azide was then added
(1.05 g), and stirring was continued overnight at 80 °C. The
† Part of the special issue “Jack Beauchamp Festschrift”.
* To whom correspondence should be addressed. E-mail: jmdyke@
soton.ac.uk.
‡ CeFiTec, Department of Physics.
§ CQFB, Department of Chemistry.
10.1021/jp020625e CCC: $22.00 © 2002 American Chemical Society
Published on Web 08/20/2002

Decomposition of N3CH2CH2OH and N3CH2COOCH2CH3
J. Phys. Chem. A, Vol. 106, No. 42, 2002 9969
solution was then cooled, and the azidoethanol was extracted
with diethyl ether (3 × 10 mL), after which the combined ether
layers were dried (over MgSO4) and filtered. The filtrate was
concentrated by vacuum distillation, first at room temperature
and then finally at 105 °C to yield a colorless oil. In a very
similar preparative procedure, 2-azidoethyl acetate was prepared
from the reaction of ethyl bromoacetate (98%, Aldrich) with
sodium azide (99%, Aldrich) mixed in the molar ratio of 1:2.5.
To a vigorously stirred solution of ethyl bromoacetate (1.51 g)
in acetone (10 mL) at 0 °C, was slowly added a solution of
sodium azide (1.47 g) in water (10 mL). The mixture was
allowed to warm to room temperature and then heated overnight
at 60 °C. The resulting solution was extracted with dichlo-
romethane (3 × 10 mL). The combined organic layers were
washed first with 10% aqueous sodium bicarbonate solution and
then with water and dried over magnesium sulfate and filtered.
The filtered solution was concentrated by vacuum distillation,
and the residue was distilled at 115 °C to yield a colorless oil.
Photoelectron spectra were recorded in real time as the
furnace temperature was changed. For each azide, the onset of
decomposition was marked by the appearance of photoelectron
bands associated with molecular nitrogen. Spectral calibration
was achieved by reference to known ionization energies of
methyl iodide added to the chamber and traces of nitrogen and
water22,23 present in the ionization chamber. In the PES work,
the pressure in the pyrolysis region is estimated from previous
pressure measurements as ca. 10-4 Torr and the flow time
between the pyrolysis region and the photoionization region is
estimated at ca. 5 ms.
The photoelectron spectrum was also recorded for BrCH2-
CH2OH used in the preparation of the 2-azidoethanol, and it
was in good agreement with that reported previously.24 The first
and second bands of the 2-azidoethanol, at vertical ionization
energies of 9.90 ( 0.01 and 11.01 ( 0.01 eV, respectively,
were the most intense, and photoelectron spectra recorded for
N3CH2CH2OH showed no evidence of BrCH2CH2OH. Similarly,
the photoelectron spectrum was recorded for BrCH2COOCH2-
CH3, used in the preparation of N3CH2COOCH2CH3, and the
photoelectron spectrum of the 2-azidoethyl acetate showed no
evidence of bands associated with bromoethyl acetate.
The pyrolysis system used for the azidoethanol experiments
consisted of a resistively heated furnace, which was made of
molybdenum wire wrapped around an inner glass tube connected
to a power supply via tungsten feed-throughs. The operating
temperature of the furnace in the heated section was measured
by a chromel-alumel thermocouple, and the maximum operat-
ing temperature recorded was 550 °C.
The pyrolysis experiments carried out for the 2-azidoethyl
acetate were carried out on a spectrometer of similar design.
The heating method used on this instrument was radio frequency
(rf) induction heating. This could achieve much higher tem-
peratures (>2000 °C) than the resistively heated system used
for azidoethanol. It consisted of a tantalum cylinder (the rf
susceptor) positioned inside the rf heating coils a few centimeters
above the photon source. The sample vapor was then passed
through an alumina tube, which was positioned in the center of
the tantalum susceptor. As in the 2-azidoethanol experiments,
photoelectron spectra were recorded as a function of heater
temperature.
Both products were characterized by a variety of spectro-
scopic methods as described below. Care was taken to minimize
the effect of possible explosions at all stages in the preparation
and handling of the azide samples, but no untoward occurrences
were experienced during this work.
Thermal decomposition of the two azides prepared in this
work was studied at different pyrolysis temperatures and hence
different degrees of decomposition using matrix isolation
infrared spectroscopy and gas-phase photoelectron spectroscopy.
Matrix Isolation Studies. Matrix isolation infrared studies
on these azides followed a very similar pattern to those described
in our recent studies on the pyrolysis of azidoacetic acid and
azidoacetone.17,18 The inlet and pyrolysis parts of the apparatus
were identical, as were the low-temperature Displex and infrared
(IR) spectrometers. Spectra of each parent azide and its
decomposition products were obtained in nitrogen matrices, and
supporting experiments were also carried out on H2CO/N2, CH3-
NH2/N2, C2H4/N2, C2H6/N2, and CH4/N2 mixtures to aid spectral
identification. Deposition times were typically 30-60 min at a
particular superheater temperature, and any changes occurring
during this period were monitored by spectral subtraction. The
pressure in the pyrolysis region during a typical experiment was
estimated as 10-4 Torr, and matrix ratios were estimated to be
well in excess of 1000:1.
Results
Photoelectron Spectroscopy. All photoelectron spectra
recorded in this work were obtained using HeI (21.22 eV)
radiation. The two spectrometers used, which were very similar
and have been described elsewhere,20,21 were single detector
instruments designed for high-temperature pyrolysis studies. For
each azide, photoelectron spectra of the parent compound and
its decomposition products were obtained by passing the azide
vapor through a heated furnace positioned inside the ionization
chamber of the spectrometer, several centimeters above the
photon source.
Characterization of N3CH2CH2OH. 2-Azidoethanol was
characterized in the vapor phase by mass spectrometry and
ultraviolet photoelectron spectroscopy, and also in the liquid
1
phase by H NMR, 13C NMR, Raman, and infrared spectro-
scopy. The 70 eV electron impact mass spectrum showed a
parent peak at 87 (10%), together with two intense fragments
at 28 (40%, N2 or NCH2+) and 31 (100%, CH2OH+) amu.
+
Weaker signals were also observed at 45 (5%, CH2CH2OH+),
43 (5%, N3H+), 29 (15%, COH+), and 27 (5%, NCH+) amu.
No signals were seen that could be attributed to unreacted
BrCH2CH2OH, and the mass spectrum was in good agreement
with that reported previously.25
The vapor pressures of the parent materials (N3CH2CH2OH
and N3CH2COOCH2CH3) were sufficient at room temperature
to allow photoelectron spectra to be obtained with good signal-
to-noise ratios by direct pumping on a liquid sample held in a
small flask through a needle valve outside the spectrometer
ionization chamber, rather than having to place sample vials
on the ledge of the heated inlet system inside the ionization
chamber of the spectrometer as was done for the less-volatile
azidoacetic acid.17 The operating resolution of each spectrometer
(typically 35 meV fwhm as measured for the argon (3p)-1 signal
obtained with He(I) radiation) was unaffected when the heating
system was in operation, even at the highest temperature used.
Two signals were observed in the 300 MHz 1H NMR
spectrum of 2-azidoethanol in CDCl3 solution. These were at
3.41 and 3.77 ppm relative to TMS with integrated relative
intensities of 0.4:0.6, respectively. After expansion, it could be
seen that both peaks were triplets, but the band at 3.77 ppm
had a much more intense central band; this was due to the -OH
proton appearing in exactly the same place as one of the -CH2
groups of protons. Two signals were also observed in the 13C
NMR spectrum at 53.7 and 61.5 ppm with integrated relative

9970 J. Phys. Chem. A, Vol. 106, No. 42, 2002
Hooper et al.
TABLE 1: Significant IR Bands (cm-1) Observed in Matrix
Isolation Studies on the Pyrolysis of 2-Azidoethanol and
2-Azidoethyl Acetate
N2 matrix (letter in
Figures 1 and 3)a
previous studies
assignment
3623, 2977, 2958, 2930,
2888, 2137, 2105,
1282, 1074
N3CH2CH2OH
2116, 1753, 1373, 1351,
1293, 1199, 1034
2139(F)
N3CH2COOCH2CH3
2139
2347, 662
CO (ref 30)
CO2 (ref 31)
2347(O), 662(S)
3032(B), 2919(C), 1637, 3033, 2920, 1637, 1450, CH2NH (ref 29)
1450, 1352(H), 1127(I), 1353, 1128, 1065
1065(J)
3287(A), 747/727
2864(D), 2800(E),
1738(G),
3287, 747/737
2865, 2800, 1740,
1499, 1244, 1167
HCN (refs 29 and 34)
H2CO (ref 33)
1497, 1245, 1169
3106(M), 3077, 2988(N), 3107, 3077, 2989, 1889, C2H4 (ref 32)
1888(P), 1437(Q),
947(R)
1438, 947
1306
2967, 2885, 1263,
816(L), 970(K)
1306 (most intense)
CH4 (ref 35)
unassigned features
appearing
on pyrolysis
a Wavenumber accuracy ) (1 cm-1
.
Figure 1. Nitrogen matrix infrared bands observed during pyrolysis
studies on 2-azidoethanol. Absorptions denoted by an asterisk are
assigned to H2O. Band identification is shown in Table 1. Spectrum a
was obtained with no superheating (parent 2-azidoethanol sublimed at
room temperature). Spectrum b was obtained after passage through
superheater at 570 °C.
intensities of 1:1. H-C correlation was also carried out, and
good correlation was found between the two sets of peaks
indicating that the 3.77 ppm (1H NMR) peak and the carbon at
61.5 ppm (13C NMR) were from the same group (-CH2OH)
and the 3.41 ppm (1H NMR) and the signal at 53.7 ppm (13C
NMR) were from the other group (N3-CH2-).
The most intense absorption band in the liquid infrared
spectrum was found at 2123 cm-1, which was assigned to the
N3 group. Other prominent bands were present at 3442 cm-1
(OH stretch), 2988 and 2928 cm-1 (CH stretch), and 1296 and
1066 cm-1. The infrared bands in a nitrogen matrix were in
general shifted only slightly from the liquid-phase values, and
their positions and relative intensities are shown in Figure 1a
and in Table 1. The only significant differences between the
liquid and matrix infrared spectra were in the position of the
O-H stretching band, which was significantly higher (3623
cm-1) for the isolated molecule, and in the multiplet structure
observed for the N3 absorption at ca. 2120 cm-1. This band is
quite broad in the liquid spectrum, and the resolution of at least
two components in this band in the matrix spectrum perhaps
reveals the existence of more than one conformer.
Figure 2. He(I) photoelectron spectrum recorded for 2-azidoethanol
at different stages of pyrolysis. Panel a shows the photoelectron
spectrum of the parent 2-azidoethanol. The numbered bands are listed
in Table 3. Panel b shows the photoelectron spectrum obtained on
pyrolysis of 2-azidoethanol at a temperature of 240 °C. Panel c shows
the photoelectron spectrum obtained for complete pyrolysis of 2-azi-
doethanol at 380 °C.
point on the potential energy surface with all real vibrational
frequencies.
The liquid Raman spectrum exhibited counterparts to all of
the above bands except for the OH stretch.
The highest occupied molecular orbital in the ground state
(orbital 23) is weakly antibonding between the terminal nitrogen
atoms of the N3 group, nonbonding between the adjacent
nitrogen atoms, and weakly antibonding in the C-N direction.
Orbital 22 is π-bonding in the terminal N-N bond, weakly
antibonding between the other N-N group, and weakly C-N
antibonding. Orbital 21 consists of a π(C-O) bonding orbital,
The He(I) photoelectron spectrum of 2-azidoethanol is shown
in Figure 2a. Assignment of this spectrum is achieved with
reference to ab initio molecular orbital calculations carried out
as part of this study. The computed minimum energy geometry
of N3CH2CH2OH was obtained at the MP2/6-31G** level using
the Gaussian 94 code. The geometrical parameters listed in Table
2 from the MP2/6-31G** calculations correspond to a stationary

Decomposition of N3CH2CH2OH and N3CH2COOCH2CH3
J. Phys. Chem. A, Vol. 106, No. 42, 2002 9971
TABLE 2: Computed Structural Parameters of
Azidoethanol at the MP2/6-31G** Level Showing the Atom
Numberinga
Pyrolysis experiments were carried out in the range ca. 100-
600 °C, and even at the lowest temperatures, there was evidence
for some decomposition, with the generation of new products.
However, traces of the parent azide could still be detected up
to 500 °C.
Figure 1b shows the spectrum obtained at a superheater
temperature of 570 °C. An intense feature is still present close
to the original N3 doublet at 2137/2105 cm-1, but closer
examination shows that this now appears as a sharp feature at
2139 cm-1 with a shoulder at 2146 cm-1. This new band
(labeled F) is assigned to CO. Apart from a weak residual N3
feature near 2100 cm-1, all other bands corresponding to parent
absorptions are clearly absent, having been replaced by numer-
ous new features. However, despite the complexity of these
matrix infrared spectra, many of these new features can be
assigned from previous studies,17,18 on the basis of a consider-
ation of expected reaction products. The results are summarized
in Table 1.
bond
length (Å)
angle
value (deg)
C1-C2
1.5134
1.4221
1.4781
0.9635
1.2449
1.1633
1.094
C1-C2-N3
C2-C1-O4
H9-O4-C1
C2-N3-H10
C1-C2-H7/8
C2-C1-H5/6
N3-N10-N11
106.94
105.73
107.81
115.39
109.58
109.32
180.00
C1-O4
C2-N3
O4-H9
N3-N10
N10-N11
C1-H5/6
C2-H7/8
1.092
a One minimum energy structure was found at the MP2/6-31G**
level.
CH2NH is unequivocally identified by at least five charac-
teristic peaks: B, C, H, I, and J, while band A indicates the
formation of HCN. A fourth prominent pyrolysis product is
identified as H2CO (bands D, E, and G) both by comparison
with previously published work and also as a result of supporting
matrix infrared experiments. The identification of these four
products accounts for the majority of the new features produced
on pyrolysis, but six additional weak features were routinely
observed at 2967, 2885, 1306, 1263, 970, and 816 cm-1, the
assignment of which was unknown. In an attempt to establish
their identity, supporting matrix infrared experiments were
carried out on CH3NH2, CH4, C2H4, and C2H6 isolated separately
in nitrogen matrixes, with the aim of obtaining a suitable match
for at least two frequencies for each compound. In practice,
only the most intense feature of CH4 gave a positive match for
TABLE 3: Comparison of Experimental and Computed
Vertical Ionization Energies (VIEs) of 2-Azidoethanol
∆SCF values
MP2/6-31G**
band
no.
exptl
KT
KT ×
AIE
(eV)
VIE
(eV)
MO
VIE (eV) VIE (eV) 0.92 eV
23
22
21
20
19
18
17
16
15
1
2
3
4
5
6
9.90
11.01
11.47
12.64
14.06
15.03
10.18
11.93
12.38
13.80
15.23
15.67
17.82
17.86
18.26
9.37
10.98
11.38
12.70
14.01
14.42
16.39
16.43
16.80
9.94
10.64
9.99
11.30
7
17.42
the frequency at 1306 cm-1
.
and orbital 20 is bonding between the two carbon atoms and
C-O antibonding.
Thermal Decomposition Experiments: Photoelectron Spec-
troscopy. The He(I) photoelectron spectrum obtained for
azidoethanol is shown in Figure 2a, and the photoelectron
spectrum obtained upon partial pyrolysis is shown in Figure
2b. The temperature of the furnace under these conditions was
approximately 240 °C as estimated from chromel-alumel
thermocouple measurements. The He(I) photoelectron spectrum
recorded for complete pyrolysis is shown in Figure 2c, recorded
at a furnace temperature of approximately 380 °C.
Signals arising from N2, CO, and HCN are clearly observed
on pyrolysis (Figure 2b,c). In addition, pyrolysis also produced
four further features: a broad band with a VIE of 10.65 eV,
which is associated with the vibrationally resolved band at 12.66
eV (VIE), and a sharp feature at 10.90 eV (VIE), which is
associated with the vibrationally resolved band at 15.84 eV
(VIE). These can be attributed to the first two bands of CH2-
NH and the first two bands of CH2O, respectively, by
comparison with known spectra of methanimine15,28 and form-
aldehyde.22
No spectral features were observed that indicated the possible
formation of an imine intermediate (HNdCHCH2OH, first VIE
computed at the ∆SCF QCISD/6-31G** level for the X1A′ state
is 9.96 eV) or a nitrene intermediate (NCH2CH2OH, first VIE
computed at the ∆SCF QCISD/6-31G** level for the X3A′′ state
is 10.59 eV). Also, it is significant that the bands assigned to
N2, H2CO, CH2NH, CO, and HCN appear together on pyrolysis
and show the same temperature profiles. No spectral features
were observed associated with CH4, indicating that if it is
produced it is a very minor product.
Table 3 compares the experimental vertical ionization energies
(VIEs) with VIEs obtained from application of Koopmans’
theorem to the SCF/6-31G** molecular orbital energies obtained
at the MP2/6-31G** computed minimum energy geometry
(Table 2). The values computed were, as expected, higher than
the experimental values, but scaling by a factor of 0.9226,27
afforded much better agreement with the experimental VIEs.
This result was similar to the azidoacetic acid17 and azidoacetone
cases18 in which a factor of 0.9 was used. Also included in Table
3 are the first two adiabatic ionization energies (AIEs) and VIEs
for 2-azidoethanol calculated by the ∆SCF method using the
optimized geometries of the cation obtained at the MP2/6-31G**
level. These show good agreement with experimental values.
Thermal Decomposition Experiments: IR Matrix Isola-
tion Spectroscopy. Figure 1a shows a typical nitrogen matrix
infrared spectrum obtained from a sample of N3CH2CH2OH
deposited from the vapor phase without superheating. The
absorptions denoted by the asterisk (/) indicate the most intense
infrared absorptions of matrix-isolated H2O and arise from traces
of water in the system. The most intense parent absorption
consists essentially of a doublet at 2137/2105 cm-1, one
component of which exhibits weaker shoulders. These absorp-
tions are assigned to the N3 unit, and the complex appearance
is attributed to different conformers trapped in the matrix. Our
previous study on azidoacetone18 also showed a multiplet
structure in this spectral region. Other prominent features were
noted at 1282 and 1074 cm-1, which are in similar positions to
those found in the neat liquid, and a relatively high OH stretch
at 3623 cm-1, which owes its position to the absence of
hydrogen bonding.
Characterization of N3CH2COOCH2CH3. 2-Azidoethyl
acetate was characterized in the vapor phase by mass spec-

9972 J. Phys. Chem. A, Vol. 106, No. 42, 2002
Hooper et al.
Figure 3. Nitrogen matrix infrared bands observed during pyrolysis
studies on 2-azidoethyl acetate. Absorptions denoted by an asterisk are
assigned to H2O. Band identification is shown in Table 1. Spectrum a
was obtained with no superheating (parent ester sublimed at room
temperature). Spectrum b was obtained after passage through super-
heater at ∼500 °C.
Figure 4. He(I) photoelectron spectrum recorded for 2-azidoethyl
acetate at different stages of pyrolysis. Panel a shows the photoelectron
spectrum of the parent 2-azidoethyl acetate. The numbered bands are
listed in Table 5. Panel b shows the photoelectron spectrum obtained
on pyrolysis of 2-azidoethyl acetate at a temperature of 500 °C. Panel
c shows the photoelectron spectrum obtained for complete pyrolysis
of 2-azidoethyl acetate at 775 °C.
trometry and ultraviolet photoelectron spectroscopy, and also
in the liquid phase by 1H NMR, 13C NMR, Raman, and infrared
spectroscopy. The 70 eV electron impact mass spectrum showed
an intense parent peak at 129 amu (100%) together with the
following fragments: 73 (20%, CO2CH2CH3+), 56 (55%, N3-
CH2+), 45 (25%, OCH2CH3+), 42 (20%, N3+), and 29 amu
(45%, CH2CH3+). No signals were seen that could be attributed
to unreacted BrCH2COOCH2CH3.
The infrared spectrum recorded in a nitrogen matrix is shown
in Figure 3a, and the main absorptions are summarized in Table
1.
The He(I) photoelectron spectrum of 2-azidoethyl acetate is
shown in Figure 4a. As in the azidoethanol case, assignment of
the spectrum is made by use of ab initio molecular orbital
calculations with the Gaussian 94 code. The minimum energy
structure at the MP2/6-31G** level is shown in Table 4. This
was the lowest energy conformer. The highest occupied mo-
lecular orbital (orbital 34) is antibonding between the terminal
nitrogens on the N3 group. Orbital 33 is delocalized over the
N3CH2CO2- grouping and is antibonding between each pair
of bonded atoms. Orbital 32 has a major contribution from the
lone pair on the oxygen on the C-O group and a minor bonding
contribution from the C-N group. Orbital 31 is localized on
the two oxygens and is mainly O2p antibonding in character.
The experimental VIEs have been compared with VIEs
obtained from application of Koopmans’ theorem to the SCF/
6-31G** molecular orbital energies obtained at the MP2/
6-31G** computed minimum energy geometry (see Tables 4
and 5). The Koopmans’ theorem values are, as expected, higher
than the experimental values, but scaling by 0.92 gives much
better agreement with experiment, as in the azidoethanol case.
Also included in Table 5 is the first AIE and VIE computed by
the ∆SCF method using optimized geometries of the cationic
states obtained at the MP2/6-31G** level. The ∆SCF first VIE
shows good agreement with the experimental value.
1
In the 400 MHz H NMR spectrum of 2-azidoethyl acetate
in CDCl3, three signals were observed at 4.25, 3.86, and 1.30
ppm with integrated relative intensities of 2:2:3, respectively.
These three signals were assigned as follows: The signal at
4.25 ppm was a quartet corresponding to the two protons on
the CH2 in the -OCH2CH3 group, the signal at 3.86 ppm was
a singlet corresponding to the N3CH2- group, and the signal at
1.30 ppm corresponds to the protons on the terminal methyl
group. In the 13C NMR of the 2-azidoethyl acetate four strong
signals were observed at 14.20, 50.44, 61.96, and 168.4 ppm.
1
Good H-13C correlation was found between the two sets of
NMR signals, indicating that the signal at 14.20 ppm corre-
sponds to the terminal methyl group, the signal at 50.44 ppm
corresponds to the N3CH2- group, and the signal at 61.96 ppm
corresponds to the CH2 in the -OCH2CH3 group. The final
signal at 168.4 ppm is due to the carbon in the carbonyl group.
In the liquid infrared spectrum, intense absorptions were
observed at 1749, 2110, 2935, and 2993 cm-1. These were
assigned to the CdO stretch, the N3 group, and the C-H
stretches (2935 and 2993 cm-1), respectively. The liquid Raman
spectra exhibited counterparts to each of these bands observed
in the infrared spectrum.
Thermal Decomposition Experiments: IR Matrix Isola-
tion Spectroscopy. Figure 3a shows a typical nitrogen matrix

Decomposition of N3CH2CH2OH and N3CH2COOCH2CH3
J. Phys. Chem. A, Vol. 106, No. 42, 2002 9973
TABLE 4: Computed Structural Parameters of 2-Azidoethyl
Acetate at the MP2/6-31G** Level Showing the Atom
Numberinga
Thermal Decomposition Experiments: Photoelectron Spec-
troscopy. The He(I) photoelectron spectrum obtained for
2-azidoethyl acetate is shown in Figure 4a. The photoelectron
spectra obtained on partial and complete pyrolysis are shown
in Figure 4b,c recorded at furnace temperatures of 500 and 775
°C, respectively. The bands associated with the parent azide
decrease, as expected, as the furnace temperature increases, and
bands associated with C2H4, CH2NH, HCN, CO2, and N2 appear.
The temperature profiles of these product bands are, within
experimental error, the same, all reaching a maximum intensity
at the same furnace temperature. No photoelectron bands
associated with an imine intermediate (HNCHCOOCH2CH3,
X1A′, first VIE computed at the ∆SCF MP2/6-31G** level
10.57 eV) or a nitrene intermediate (NCH2COOCH2CH3, X3A′′,
first VIE computed at the ∆SCF MP2/6-31G** level 10.03 eV)
were observed in these pyrolysis experiments.
Mechanisms of Gas-Phase Decomposition. To present
mechanisms for the gas-phase thermal decomposition of 2-azi-
doethanol and 2-azidoethyl acetate that are consistent with the
experimental photoelectron and matrix infrared spectroscopic
evidence, it is valuable to make a comparison with the earlier
results obtained for 2-azidoacetic acid and 2-azidoacetone. The
available evidence obtained for these organic azides showed
that decomposition occurred via a concerted mechanism leading
to formation of molecular nitrogen and the final reaction
products or occurred in a stepwise manner with loss of molecular
nitrogen and a proton transfer to form an imine, which
subsequently decomposes.
bond
length (Å)
angle
value (deg)
C1-C2
C2-O3
O3-C4
C4-C5
N6-C5
N7-N6
N8-N7
C-H
1.5127
1.4544
1.3498
1.5202
1.4609
1.2493
1.1616
∼1.09
C1-C2-O3
C2-O3-C4
O3-C4-C5
C4-C5-N6
C5-N6-N7
N6-N7-N8
O16-C4-O3
C-C-H
110.7
115.05
109.7
112.56
115.49
171.07
125.68
∼110
C4dO16
1.2173
a Six minimum energy structures were found at both the HF/6-31G**
and MP2/6-31G** levels corresponding to different rotational conform-
ers about the C5 and C1 centers. They all have very similar bond lengths
and angles. At the 6-31G**/MP2 level, their total energies were found
to be within 0.65 kcal mol-1. The ∆SCF computed first and second
VIEs and AIEs are all within 0.05 eV, and hence for simplicity, only
the results obtained for the lowest energy minimum (that shown in the
diagram above) are included in the text.
In the case of azidoacetic acid, CO2, CH2NH, and N2 were
observed at the same time on pyrolysis, with HCN being
produced from CH2NH at higher temperatures. The results are
consistent with a concerted mechanism which can be written
as
TABLE 5: Comparison of Experimental and Computed
Vertical Ionization Energies (VIEs) of 2-Azidoethyl Acetate
∆SCF values
band
no.
exptl
KT
KT ×
AIE
(eV)
VIE
(eV)
MO
VIE (eV) VIE (eV) 0.92 eV
34
33
32
31
30
29
28
27
26
25
24
1
2
3
4
5
9.74
10.82
11.33
12.60
13.20
10.27
11.87
12.68
12.82
13.77
14.57
14.71
15.60
16.09
16.71
17.91
9.45
10.93
11.66
11.80
12.66
13.41
13.53
14.35
14.80
15.38
16.47
9.76
9.94
This mechanism is consistent with results of DFT and ab initio
molecular orbital calculations performed by Cordeiro et al.19
on the stepwise and concerted dissociation processes of 2-azi-
doacetic acid.
The results obtained in this work for 2-azidoethyl acetate,
from which C2H4, CH2NH, CO2, and N2 are produced at the
same time with the same concentration profile as a function of
temperature, are also consistent with such a concerted mecha-
nism, which can be written as
infrared spectrum obtained from a sample of N3CH2COOCH2-
CH3 deposited from the vapor phase without superheating. The
absorptions arising from matrix-isolated H2O are again denot-
edby an asterisk (/), and the intense N3 mode in the parent azide
now lies at 2116 cm-1. Other prominent features are to be found
at 1753 and 1199 cm-1 and are in similar positions to those
found in the neat liquid. A list of the more significant bands is
given in Table 1.
Pyrolysis experiments were carried out in the range ca. 100-
550 °C, and Figure 3b shows the spectrum obtained at a
superheater temperature of ca. 500 °C. At this temperature, all
of the parent absorptions are effectively absent, but several ofthe
new features that have appeared can be assigned from previous
studies.
Both CH2NH (bands C, H, I, and J) and HCN (band A) are
again present, and the very intense new band labeled R is
identified as being due to C2H4 by comparison with previously
published work (Table 1). Closer inspection reveals that a further
four bands (M, N, P, and Q) may also be assigned to C2H4.
CO2 is unequivocally identified by bands O and S, leaving only
very weak features unidentified.

9974 J. Phys. Chem. A, Vol. 106, No. 42, 2002
Hooper et al.
This mechanism would explain the production of C2H4, CO2,
CH2NH, and N2 at the same temperatures (as is observed) with
HCN being produced by subsequent pyrolysis of CH2NH.
However, the temperatures used for the 2-azidoethyl acetate
pyrolyses were greater than those used for the azidoacetic acid
pyrolysis and are probably sufficient to lead to some decom-
position of CH2NH to HCN in the heated flow tube. Hence, it
is felt that the results obtained for 2-azidoethyl acetate are
consistent with a concerted mechanism with C2H4, CO2, CH2-
NH, and N2 being the major products.
This is very similar to the stepwise mechanism put forward for
the decomposition of azidoacetone, and as in the azidoacetone
case, it is also possible that this multistep process proceeds via
a radical mechanism.
Conclusions
2-Azidoethanol and 2-azidoethyl acetate have been synthe-
sized and characterized by a range of techniques, including
matrix isolation infrared and UV photoelectron spectroscopy.
The photoelectron spectra of azidoethanol and azidoethyl acetate
consist of seven and five resolved bands in the 9.0-21.0 eV
ionization energy region, respectively. Ab initio calculations
have been performed for these azides, and application of
Koopmans’ theorem to the computed orbital energies yields
vertical ionization energies that are in satisfactory agreement
with experimental values.
In the case of the azidoacetone, the simultaneous production
of CH2NH, H2CO, HCN, CO, N2, and CH3CHO meant that,
although a concerted mechanism of the type
The results of this pyrolysis study of 2-azidoethanol and
2-azidoethyl acetate, as well as the results from earlier studies
on the thermal decomposition of 2-azidoacetic acid and 2-azi-
doacetone, indicate that two main mechanisms of decomposition
of organic azides of the type considered are beginning to emerge.
2-Azidoacetic acid and 2-azidoethyl acetate decompose via a
concerted process through a cyclic transition state to give the
products, whereas 2-azidoethanol and azidoacetone decompose
via a stepwise mechanism through imine intermediates, which
decompose to give the products via two possible pathways.
which gives rise to H2CdCdO, CH2NH, N2, and HCN, is
possible, a stepwise pathway, possibly involving radicals, had
to be invoked to explain production of CH3CHO and CO
simultaneously with the other products. For example,
Acknowledgment. The research reported in this paper was
carried out as part of the Reactive Intermediates RTN EC
Network. Other financial support from the Leverhulme Trust
(to Southampton) and a POCTI grant (to Lisbon) is gratefully
acknowledged. The authors acknowledge valuable discussions
with Dr. E. P. F. Lee and Dr. M. Grossel.
References and Notes
(1) Patai, S. The Chemistry of the Azide Group; Interscience: New
York, 1971.
(2) Scriven, E. F. V.; Turnbull, K. Chem. ReV. 1988, 88, 298.
(3) Vantinh, V.; Stadlbauer, W. J. Heterocycl. Chem. 1996, 33, 1025.
(4) Maurus, R.; Bogumil, R.; Nguyen, N. T.; Mauk, A. G.; Brayer, G.
Biochem. J. 1998, 332, 67.
For 2-azidoethanol, it was expected that N2, H2CO, and CH2-
NH would be observed on pyrolysis from the following reaction:
(5) Schuster, G. B.; Platz, M. S. AdV. Photochem. 1992, 17, 69.
(6) Matsumara, Y.; Shiozawa, T.; Matsushita, H.; Terao, Y. Biol.
Pharm. Bull. 1995, 18, 1805.
(7) Tindall, C.; Hemminger, J. Surf. Sci. 1995, 330, 67.
(8) Bridges, A. S.; Greef, R.; Jonathan, N. B. H.; Morris, A.; Parker,
G. Surf. ReV. Lett. 1994, 1, 573.
(9) Kubota, N. J. Propul. Power 1995, 11, 677.
(10) Liu, Y. L.; Hsuie, G. H.; Chiu, Y. S. J. Appl. Polym. Sci. 1995, 58,
579.
(11) Bock, H.; Dammel, R. Angew. Chem., Int. Ed. Engl. 1987, 26, 504.
(12) Dammel, R., Ph.D. Thesis, University of Frankfurt, Frankfurt,
Germany 1985.
(13) Bock, H.; Dammel, R.; Aygen, S. J. J. Am. Chem. Soc. 1983, 105,
7681.
(14) Bock, H.; Dammel, R. J. Am. Chem. Soc. 1988, 110, 5261.
(15) Bock, H.; Dammel, R.; Horner, L. Chem. Ber. 1981, 114, 220.
(16) (a) Jing, W.; Zheng, S.; Xinjiang, Z.; Xiaojun, Y.; Maofa, G.;
Dianxun, W. Angew. Chem., Int. Ed. 2001, 40, 305. (b) Dianxun, W. Private
communication.
(17) Dyke, J. M.; Groves, A. P.; Morris, A.; Ogden, J. S.; Dias, A. A.;
Oliveira, A. M. S.; Costa, M. L.; Barros, M. T.; Cabral, M. H.; Moutinho,
A. M. C. J. Am. Chem. Soc. 1997, 119, 6883.
The observation of CO and HCN at the same time as N2, H2-
CNH, and H2CO was not anticipated if the above mechanism
was taking place. The experimental evidence provided by PES
and matrix isolation infrared experiments showed that CO, N2,
H2CO, CH2NH, and HCN are observed at the same time with
the same concentration profile as a function of temperature
within experimental error. Therefore, the following stepwise
mechanism is put forward to explain the decomposition:
(18) Dyke, J. M.; Groves, A. P.; Morris, A.; Ogden, J. S.; Catarino, M.
I.; Dias, A. A.; Oliveira, A. M. S.; Costa, M. L.; Barros, M. T.; Cabral, M.
H.; Moutinho, A. M. C. J. Phys. Chem. A 1999, 103, 8239.
(19) Cordeiro, M. N. D. S.; Dias, A. A.; Costa, M. L.; Gomes, J. A. N.
F. J. Phys. Chem. A 2001, 105, 3140.
(20) Dyke, J. M.; Morris, A.; Jonathan, N. Electron Spectroscopy:
Theory, Techniques and Application; Academic Press: London, 1979; Vol.
3.

Decomposition of N3CH2CH2OH and N3CH2COOCH2CH3
J. Phys. Chem. A, Vol. 106, No. 42, 2002 9975
(21) Dyke, J. M.; Morris, A.; Josland, G. D.; Hastings, M. P.; Francis,
P. D. High Temp. Sci. 1986, 22, 95.
(22) Turner, D. W.; Baker, C.; Baker, A. D.; Brundle, C. R. Molecular
Photoelectron Spectroscopy; Wiley-Interscience: New York, 1971.
(23) Eland, J. H. D. Photoelectron Spectroscopy; Butterworths: London,
1974.
(28) Peel, J. B.; Willett, G. D. J. Chem. Soc., Faraday Trans 2 1975,
71, 1799.
(29) Jacox, M. E.; Milligan, D. E. J. Mol. Spectrosc. 1975, 56, 333.
(30) Ogden, J. S. Unpublished observation. See also: Dubost, H.;
Abouaf-Marguin, L. Chem. Phys. Lett. 1972, 17, 269 for a discussion of
CO in matrixes.
(24) (a) Baker, A. D.; Betteridge, D.; Kemp, N. R.; Kirby, R. E. Anal.
Chem. 1971, 43, 375. (b) Kobayashi, T. Phys. Lett. 1984, 103A, 424.
(25) Olivera, A. M.; Barros, M. T.; Martins, A. M.; Cabral, M. A. R.;
Dias, A. A.; Costa, M. L.; Cabral, M. H.; Moutinho, A. M. C.; Jennings,
K. R. Rapid Commun. Mass Spectrom. 1999, 13, 559.
(26) Basch, H.; Robin, M. B.; Kuebler, N. A.; Baker, C.; Turner, D. W.
J. Chem. Phys. 1969, 51, 52.
(31) Fredin, L.; Nelander, B.; Ribbegard, G. J. J. Mol. Spectrosc. 1974,
53, 410.
(32) Frei, H.; Fredin, L.; Pimentel, G. C. J. Chem. Phys. 1981, 74,
397.
(33) (a) Moore, C. B.; Pimentel, G. C. J. Chem. Phys. 1963, 38,
2816. (b) Khoshkoo, H.; Nixon, E. R. Spectrochim. Acta 1973, 29A,
603.
(34) King, C. M.; Nixon, E. R. J. Chem. Phys. 1968, 48, 1685.
(35) Nelander, B. J. Chem. Phys. 1985, 82, 5340.
(27) Robin, M. B.; Brundle, C. R.; Kuebler, N. A.; Ellison, G. B.;
Wiberg, K. B. J. Chem. Phys. 1972, 57, 1758.