Kinetically unstable 2–isocyanophenol isolated in cryogenic matrices: Vibrational excitation, conformational changes and spontaneous tunneling

Article abstract of DOI:10.1016/j.cplett.2019.137069

Monomers of 2-isocyanophenol were generated in low-temperature solid Ar and N₂ matrices by UV-irradiation of benzoxazole and characterized by infrared (IR) spectroscopy. Near–IR narrowband excitation of the first OH-stretching overtone of 2–isocyanophenol isolated in an N₂ matrix converted the most stable cis into the higher-energy trans conformer. Interconversions between these conformers also occurred when the sample was vibrationally excited by the full, or filtered, broadband light of the IR spectrometer source, notably, with different isomerization rate constants. A spontaneous trans → cis decay, via H–tunneling, was observed for N₂ matrix kept in dark. In an Ar matrix, only the cis conformer was observed.

Full text of DOI:10.1016/j.cplett.2019.137069

Journal Pre-proofs
Research paper
Kinetically unstable 2–isocyanophenol isolated in cryogenic matrices: Vibra-
tional excitation, conformational changes and spontaneous tunneling
A.J. Lopes Jesus, Igor Reva, Cláudio M. Nunes, José P.L. Roque, Sandra
M.V. Pinto, R. Fausto
PII:
S0009-2614(19)31050-4
DOI:
https://doi.org/10.1016/j.cplett.2019.137069
CPLETT 137069
Reference:
To appear in:
Chemical Physics Letters
Received Date:
Revised Date:
Accepted Date:
29 November 2019
27 December 2019
29 December 2019
Please cite this article as: A.J. Lopes Jesus, I. Reva, C.M. Nunes, J.P.L. Roque, S.M.V. Pinto, R. Fausto,
Kinetically unstable 2–isocyanophenol isolated in cryogenic matrices: Vibrational excitation, conformational
changes and spontaneous tunneling, Chemical Physics Letters (2019), doi: https://doi.org/10.1016/j.cplett.
2019.137069
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2019 Published by Elsevier B.V.

Kinetically unstable 2-isocyanophenol isolated in cryogenic matrices:
vibrational excitation, conformational changes and spontaneous tunneling
A. J. Lopes Jesus,a,b Igor Reva,*,a Cláudio M. Nunes,a
José P. L. Roque,a Sandra M. V. Pinto,a,† and R. Faustoa
a CQC, Department of Chemistry, University of Coimbra, 3004-535, Coimbra, Portugal.
b CQC, Faculty of Pharmacy, University of Coimbra, 3004-295, Coimbra, Portugal.
____________________________
* Corresponding author. E-mail address: reva@qui.uc.pt
† Present address: Scuola Normale Superiore, Piazza dei Cavalieri, 7, I-56124 Pisa, Italy.
1

Abstract
Monomers of 2-isocyanophenol were generated in low-temperature solid Ar and N2
matrices by UV-irradiation of benzoxazole and characterized by infrared (IR) spectroscopy.
Near-IR narrowband excitation of the first OH-stretching overtone of 2-isocyanophenol
isolated in an N2 matrix converted the most stable cis into the higher-energy trans conformer.
Interconversions between these conformers also occurred when the sample was vibrationally
excited by the full, or filtered, broadband light of the spectrometer source, notably, with
different isomerization rate constants. A spontaneous trans  cis decay, via H-tunneling, was
observed for N2 matrix kept in dark. In an Ar matrix, only the cis conformer was observed.
Keywords:
2-isocyanophenol;
Matrix-isolation IR-spectroscopy;
Conformational changes;
Infrared-induced chemistry;
Isomerization rate constant;
Hydrogen atom tunneling.
2

1. Introduction
The method of matrix isolation was first developed in 1954, by the groups of Porter [1]
and Pimentel [2] with the purpose of studying short-lived reaction intermediates. In the early
sixties, Pimentel et al. reported isomerization of nitrous acid (HONO) [3,4] as a result of an
“Infrared Chemical Reaction” induced by broadband light emitted by the infrared spectrometer
light source. After these reports, the infrared-induced chemistry of matrix-isolated molecules
was barely explored over the course of three decades. However, the seminal work of Pettersson,
Lundell, Khriachtchev and Räsänen [5] rekindled interest to this topic in 1997. Monomers of
the most stable conformer (cis with respect to the OCOH dihedral) of formic acid (HCOOH)
isolated in an Ar matrix were vibrationally excited by narrowband near-infrared (NIR) light
tuned at the frequency of the observed first OH stretching overtone (6934 cm−1) [5]. This
resulted in generation and spectroscopic characterization of the elusive trans OCOH
conformer. They also found that, in an Ar matrix at 15 K, the higher energy trans conformer,
once generated, converted back to the most stable cis form on a time scale of minutes, via
hydrogen tunneling [5]. A few years later, in 2000, Khriachtchev et al. have revisited the IR-
induced chemistry of matrix-isolated HONO [6]. They performed NIR laser irradiation of the
matrix-isolated molecules of the compound at the frequency of the first OH and N=O stretching
overtones of each conformer, which resulted in selective transcis or cistrans
conformational isomerization, depending on the conformer being excited. Noteworthy, these
processes were found to be selective not just to a specific conformer but even with respect to
the matrix sites, showing the great potential of the narrowband vibrational excitations in
promoting infrared-induced conformational changes [6].
Spontaneous tunneling of trans formic acid generated in situ by NIR excitation of the
cis conformer (as described above) was investigated in detail in different noble gas matrices
(Ar, Kr, Xe) and it was found that the medium has a significant influence on the tunneling rate
(within the half-lives of minutes) [7]. Rotational isomerization of other small carboxylic acids
isolated in noble gases matrices was studied using the same approach, with focus on
photoinduced generation of higher energy conformers and tunneling [8,9]. The state-of-art of
preparation of higher-energy conformers by selective vibrational excitation was reviewed by
Khriachtchev in 2008 [10].
The next breakthrough in the field started with two simultaneous publications [11,12],
both authored by Khriachtchev et al. It was recognized that in N2-doped Ar matrices [11], as
well as in neat N2 matrices [12], the lifetimes of the higher-energy conformers of formic and
3

acetic acids increase by one or two orders of magnitude, approaching several hours [11,12].
This finding opened up the door for application of steady-state spectroscopic methods for
characterization of elusive high-energy conformers of many chemical systems otherwise not
amenable to experiment, after their generation in N2 matrices.
In the examples reported above, the higher-energy elusive conformers were generated
using commercially available precursor compounds. Recently, we have been using another
approach, whereby the high-energy kinetically unstable compounds are generated directly in
the cryogenic matrix, being subsequently used as targets for conformational manipulations. In
such a way, we have UV-generated (λ = 224 nm) in situ the elusive hydroxy- or amino-
derivatives of 2-formyl-2H-azirines from the respective isoxazole precursors in Ar matrices at
15 K (see Fig. 1 for the hydroxy-derivative). Once formed, the conformational isomerization
of the formyl group in these azirines was successfully promoted by vibrational excitations of
the first OH or NH2 stretching overtones [13]. In another study, we started with the
commercially available thioacetamide, which exists exclusively in its conformationally rigid
amino-thione tautomer S=C(CH3)−NH2. Using UV-excitations (λ = 265 nm), four high-energy
(38 kJ mol−1 or more with respect to the thione isomer) imino-thiol isomers
H−S−C(CH3)=N−H, differing by mutual orientations of the SH and NH groups, were generated
[14,15]. This allowed for subsequent studies of spontaneous [14] and vibrationally-induced (by
excitation of the first NH stretching overtones) [15] conformational isomerizations.
In the present work, we report on the UV-induced generation of kinetically unstable
2-isocyanophenol (2) from commercially available benzoxazole (1) isolated in cryogenic Ar
and N2 matrices. In N2 matrices, it was possible to trap two conformers of 2 and study their
mutual transformations, both light-induced and spontaneous. These experiments allowed for
the structural and vibrational characterization of the two conformers of 2-isocyanophenol,
which are reported here for the first time.
2. Experimental and computational methods
Commercial benzoxazole (1) (TCI Europe, > 98%) was used in the present study as the
starting material. The compound was placed in a glass ampoule, connected to the cryostat via
SS-4BMRG (NUPRO) needle valve with a shut-off and metering possibility. Prior to the
experiment, the cell with benzoxazole was connected to the vacuum system of a closed-cycle
helium refrigerator (APD Cryogenics, with a DE-202A expander) and the compound was
additionally purified from volatile impurities by pumping through the cryostat. Matrices were
4

prepared by simultaneous condensation of benzoxazole with a large excess of inert gas (argon
N60 and nitrogen N50, both supplied by Air Liquide) onto a cryogenic CsI window (~15 K)
which was used as optical substrate. The sublimation rate of benzoxazole was chosen to be low
enough to ensure that the species trapped in the matrices were mainly monomers. For this
purpose, during sublimation, crystals of benzoxazole were cooled to 0 °C by immersing the
ampoule with the compound into a bath with melting water-ice mixture. This allowed to reduce
the saturated vapor pressure over the compound and to improve the metering function of the
needle valve. A glass vacuum system and standard manometric procedures were applied to
deposit the matrix gas (Ar or N2), from a separate inlet port.
The compound of interest, 2-isocyanophenol (2), was photogenerated in situ by
irradiating matrix-isolated 1 with UV-light at λ = 233 nm (absorption maximum of 1 in ethanol
[16]). After photogeneration of 2, irradiations were also performed in the near-infrared (NIR)
region. Monochromatic (full width at half-maximum ~0.2 cm−1) UV or NIR light was provided
by a frequency-doubled signal or idler beam, respectively, of a Quanta-Ray MOPO-SL pulsed
optical parametric oscillator (pulse duration 10 ns, repetition rate 10 Hz). The pulse energy
used was 0.5 mJ (UV) or 8 mJ (NIR).
FTIR spectroscopy was used for vibrational characterization of the trapped species and
also for following their light-induced and spontaneous transformations. In some experiments,
a standard Edmund Optics longpass filter was used (transmission cut-off value of ∼4.50 µm)
to protect the sample from infrared light with wavenumbers above ~2200 cm−1. Other details
of the experimental setup and methodology used in the matrix deposition, UV and NIR
irradiations and acquisition of IR spectra, have been described elsewhere [17].
All theoretical calculations were performed with the Gaussian 09 software package
(Revision D.01) [18]. The geometries of two conformers of 2, and the transition state between
them, were optimized using the B3LYP [19-21] and MP2 [22] methods, combined with the
6-311++G(d,p) and 6-311++G(3df,3pd) basis sets [23-25], respectively (Cartesian coordinates
of the optimized geometries are listed in Table S1). Harmonic vibrational calculations at both
levels were carried out for the respective optimized geometries. The nature of stationary points
as true minima, or first-order transition states, was confirmed by analysis of the corresponding
Hessian matrices. The wavenumbers computed for 2 at the B3LYP/6-311++G(d,p) level were
subjected to least squares linear fit against the experimental wavenumbers observed for the cis
conformer 2c in an Ar matrix (see Section 3.2), resulting in scaling factors of 0.983 or 0.948,
below and above 3000 cm−1, respectively. The scaled wavenumbers, and absolute infrared
intensities extracted from the B3LYP vibrational calculations, were used to simulate IR spectra.
5

For this purpose, the peaks centered on the scaled frequencies, were convoluted with
Lorentzian functions having a full-width at half-maximum (fwhm) of 1 or 2 cm–1 and peak
heights equal to the calculated IR intensities.
The theoretical normal modes were analyzed by carrying out potential energy distribution
(PED) calculations. Transformation of the force constants with respect to the Cartesian
coordinates into the force constants with respect to the molecule-fixed internal coordinates
allowed the PED analysis to be carried out as described by Schachtschneider and Mortimer
[26]. The set of internal coordinates used in the PED analysis is given in Table S2 and is defined
as recommended by Pulay et al. [27]. The atom numbering is presented in Fig. S1. The resulting
vibrational assignments of the two conformers, including the elements of PED matrices greater
than 10%, are presented in Tables S3 and S4. B3LYP anharmonic vibrational computations
were also carried out using the fully automated second-order vibrational perturbative theory
(VPT2) approach of Barone and co-workers [28-30]. Calculated anharmonic wavenumbers
were not scaled.
3. Results and discussion
3.1. Relative stability of the 2-isocyanophenol conformers and energy barriers for their
interconversion
2-Isocyanophenol 2 has two planar conformers (Cs symmetry), here denoted as 2c or
2t, depending if the hydrogen of the OH group points toward (cis, c) or away (trans, t) from
the isocyano group (−N≡C), respectively (Table 1). B3LYP/6-311++G(d,p) and MP2/6-
311++G(3df,3pd) relative electronic (ΔEel) and zero-point corrected (ΔE0) energies for the two
conformers are given in Table 1. Both methods predict an energy difference (zero-point
corrected) between the two forms of 11 kJ mol−1. The lower energy of 2c can be rationalized
by considering the existence in this conformer of a stabilizing O−H···N hydrogen bond-like
interaction, as revealed by the geometric parameters given in Table 1, and also by the
experimentally observed OH stretching wavenumber redshift of 39 cm−1 in 2c relative to 2t
(Tables S3 and S4). The conformational behaviour of 2 follows the pattern exhibited by other
2-substituted phenols where the substituent acts as a hydrogen bond acceptor [31].
The computed energy barriers for the conformational interconversion (ΔΔE‡) are
included in Table 1 (the potential energy scan for the internal rotation of the OH group around
the C−O bond is shown in Fig. S2). Values of 22.1 kJ mol−1 (1850 cm−1) and 11.5 kJ mol−1
6

(960 cm−1) were obtained for the isomerization in the forward 2c  2t and backward 2c 
2t directions, respectively. A nearly identical OH torsional energy profile has been previously
reported for the isomeric 2-cyanophenol molecule (Fig. S2) with forward and backward
barriers of 22.0 and 12.4 kJ mol−1, respectively) [32].
3.2. Spectral identification of 2-isocyanophenol and NIR laser induced 2c  2t conversion
Monomers of benzoxazole 1 were isolated in Ar and N2 matrices and characterized by
IR spectroscopy. Some of the strongest bands in the infrared spectrum of 1 isolated in an Ar
matrix appear at 1526.5, 1452.6, 1239.3, 1106.0, 1069.8 cm−1 (see Fig. S3b). They agree very
well with the positions of the strongest bands in the gas phase IR and Raman spectra of
benzoxazole, which appear at 1528, 1453, 1238, 1107, 1072 cm−1 (reported by Klots and
Collier [33]) or at 1523, 1451, 1237, 1105, 1068 cm−1 (reported by Mille et al. [34]). The
experimental values also match well the strongest IR bands computed at the
B3LYP/6-311++G(d,p) level: 1535.5, 1450.6, 1233.8, 1105.3, 1061.5 cm−1 (Fig. S3c).
Matrix-isolated 1 was subjected to a series of narrowband UV-irradiations at  = 233
nm. In Ar matrix, these irradiations led to the consumption of 1 and formation of a photoproduct
with strong infrared absorptions at 3584 and 2118 cm−1 (Fig. 2). Absorptions at such
frequencies are typical of OH and −N≡C stretching vibrations. Interestingly, when Ferris et al.
studied the photochemistry of 1,2-benzisoxazole 3 at −77 and −196 °C with a  = 254-nm light
source in a KBr matrix or neat film [35], they observed a photoproduct “X” with absorption at
2130 cm−1 due to the −N≡C stretching. According to those authors, this absorption disappeared
on warming to 0 °C in dark, giving rise to the intense IR spectrum of benzoxazole 1 [36]. On
this basis, the kinetically unstable product “X” was assigned to 2 [35,36]. Besides the −N≡C
stretching and the OH stretching (at 3350 cm−1), no other absorptions of 2 had been ever
reported. Interestingly, when we studied the UV-photochemistry of matrix-isolated 3 ( = 280
nm), no signs of formation of 2 from 3 were detected [37]. However, in the present work, 2 is
generated from 1 ( = 233 nm) almost exclusively. The reliable assignment of the photoproduct
to 2 is based on the experimental observation, for the first time, of all the 18 vibrations that
have their predicted counterparts in the 2200–400 cm–1 range with IR intensities above 2 km
mol–1 (see Fig. 2 and Table S3). This also allowed for deriving a scaling factor by least-squares
linear fitting of calculated to experimental wavenumbers (Fig. S4), which was used throughout
this work.
7

Comparison of the results obtained in the two matrices (Ar vs. N2) revealed a striking
difference. In Ar matrix, the IR spectrum of the photoproduct is fully reproduced by the
theoretical spectrum predicted for 2c, while no signs of 2t were detected (Fig. 2). On the other
hand, in N2 matrix, the most intense bands of the photoproduct are nicely reproduced by the
theoretical spectrum of 2c, but a set of new weaker bands also appears at the positions where
2t is expected to absorb (compare Fig. 2 and Fig 3). For example, two bands due to the OH
stretching mode appear in N2 at 3578 (2c) and 3617 (2t) cm–1. The latter band nicely agrees
with the OH stretching frequency of phenol (OH group free of intramolecular interactions) in
N2 matrix reported at 3622 cm–1 by Khriachtchev et al. [38]. Characteristic is also the doublet
near 2120 (2c) and 2135 (2t) cm–1 (Fig. 3), ascribed to the −N≡C stretching vibration, with a
predicted wavenumber difference of 17.3 cm–1 (see Tables S3 and S4).
In the NIR spectra recorded after UV-induced generation of 2, absorption bands were
identified at 6992 cm−1 in Ar matrix, and at 6986 cm−1 in N2 matrix (see Fig. S5). These bands
are ascribed to the first OH stretching overtone [2ν(OH)] of 2c, and their experimental positions
are in good agreement with the theoretical anharmonic frequency calculated for this transition
(7001 cm−1, Fig. S5).
To investigate the possibility of conformational isomerization between 2c and 2t by
vibrational excitation, we carried out a series of monochromatic excitations at the frequency of
the 2ν(OH) absorption identified for 2c. In the Ar matrix, irradiations at 6992 cm−1 did not
produce any observable transformation. However, in the N2 matrix, irradiations at 6986 cm−1
for about 100 min led to significant changes in the IR spectrum (Fig. 4). The resulting
experimental difference IR spectrum compared with the spectra calculated for 2c and 2t clearly
indicates that excitation of the 2ν(OH) mode of 2c promotes a decrease of the bands assigned
to this conformer and a concomitant increase of the bands assigned to 2t (Fig. 4). Thanks to
these spectral changes, a detailed assignment of the infrared spectra of matrix-isolated 2, and
particularly of the minor 2t conformer, was carried out (see Table S4).
The lack of observation of 2t conformer in the Ar matrix, upon the UV or NIR irradiation
experiments, is most probably due to occurrence of fast hydrogen tunneling to the most stable
2c conformer. This behaviour has been observed for a large number of carboxylic acids with
conformers differing by 180° rotation of OH group, such as formic [5,7,8], acetic [8,9],
propionic [8], and squaric [39] acids, and in a few phenol derivatives [40-42]. On the other
hand, when N2 is used as host medium, it is sometimes possible to extend the lifetime of the
high energy rotamers, thus permitting their spectroscopic characterization [12,32,39,43,44].
This is also demonstrated here for the case of 2t conformer. According to Khriachtchev et al.,
8

the increased stability of the high energy conformers in an N2 matrix is attributed to the
formation of specific OH···N2 interactions [11,12,38,45], due to the fact that N2 molecules
have a permanent quadrupole moment (while noble-gas atoms do not have) [38].
3.3. Conformational changes induced by the broadband IR-radiation emitted from the
spectrometer
Conformational changes in matrix-isolated molecules can occur also when they are
exposed to the IR broadband radiation emitted by the spectrometer source [4,32,46]. To study
such changes for 2 in the N2 matrix, we followed the time evolution of the population of the
two conformers during the exposition of the sample to the spectrometer beam (see caption of
Fig.5). The kinetic measurements started from two different 2c : 2t population ratios in the
matrix, near 75 : 25 and 95 : 5. The first ratio was obtained by performing selective NIR
excitations at 6986 cm−1 as described above, while the second was achieved by letting the
sample in dark for 23 hours at a temperature of 11 K. For the matrix with an initial ratio of 75 :
25, the abundance of 2t was found to decrease with time, while that of 2c increased (Fig. 5a,
closed symbols). After approximately 40 minutes of exposing the matrix to the spectrometer
light source, no further changes in the populations of the two conformers were observed,
meaning that a stationary state was reached. At this state, the 2c : 2t ratio was estimated to be
near 88 : 12. For the matrix with an initial 95 : 5 conformational ratio, the abundance of the
two conformers evolved in the opposite direction, i.e. 2t was found to increase with time (see
Fig. 5a, empty symbols). When the system has approached the stationary state, the
conformational ratio (about 88 : 12) was found to be identical to that observed in the experiment
where the initial conformational ratio was 75 : 25.
The above results show that, independent of the initial conformational composition, the
system always tends to the same stationary state upon the same broadband irradiation
conditions. At this state, the rate of the 2c  2t conversion (promoted mostly by excitation of
vibration modes lying above the isomerization barrier, computed ΔΔE‡ ~ 1850 cm−1) matches
0
that corresponding to the 2t  2c conversion (induced by excitation of vibrations lying above
the isomerization barrier, computed ΔΔE‡ ~ 960 cm−1, plus the contribution of hydrogen
0
tunneling). By fitting the experimental data shown in Fig. 5a with an exponential equation,
+k
c-t t-c
y = a + be−(k
)t, the following rate constants were estimated for the 2c  2t (kc-t) and 2t 
2c (kt-c) processes: kc-t = 2.3 × 10−4 s−1 and kt-c = 1.6 × 10−3 s−1.
9

The kinetic measurements were repeated after enriching the N2 matrix with the less
stable 2t form, but in this case a longpass filter transmitting only IR-light below 2200 cm−1 was
placed between the sample and the spectrometer in order to prevent excitation of the higher
energetic vibrations, particularly the strongly absorbing ν(OH) modes. The obtained kinetic
data, which are represented in Fig. 5b, show that the populations of the two conformers also
tend towards a stationary state, with a conformational ratio of about 90 : 10. The following
isomerization rate constants were estimated from the exponential fitting: kc-t = 9.3 × 10−5 s−1
and kt-c = 8.5 × 10−4 s−1. These rate constants are, respectively, 60 and 47 % lower than those
estimated from the kinetic measurements performed without the presence of the filter. It
demonstrates that, besides the excitation of the ν(OH) and other high-frequency modes,
excitation of the low energy vibrations (below 2200 cm−1) also plays a significant role in the
conformational changes occurring in the presence of broadband IR-radiation. The fact that the
reduction of the 2t  2c conversion rate, originated by the presence of the filter, is not so
pronounced as that observed for the opposite 2c  2t conversion, seems to indicate that the
first transformation is not solely determined by a photoinduced process, an observation that is
in agreement with the 2t  2c decay by tunneling discussed in the next section.
Identical kinetic studies were previously performed for 2-cyanophenol isolated in an N2
matrix and the rate constants corresponding to the cis ↔ trans isomerizations were also
estimated for the sample exposed to unfiltered and filtered (below 2200 cm−1) broadband IR-
radiation emitted from the spectrometer [32]. In this case, the cis  trans and trans  cis
isomerization rate constants, resulting from blocking the absorptions above 2200 cm−1, were
found to be, respectively, 80 and 68 % lower than those observed for the corresponding
unfiltered processes. This represents a much more pronounced reduction of the rate constants
than that here reported for 2-isocyanophenol. Such difference can be interpreted in terms of the
position of the intense ν(NC) or ν(CN) absorptions relative to the filter cut-off frequency (
2200 cm−1). As shown in Fig. S7, for 2-isocyanophenol these absorptions are located at lower
wavenumbers (2133 and 2122 cm−1), while for the isomeric 2-cyanophenol molecule they are
located at higher wavenumbers (2239 and 2230 cm−1). This means that the presence of the filter
allows excitation of the ν(−N≡C) modes in 2-isocyanophenol but prevents excitation of the
ν(−C≡N) modes in 2-cyanophenol. Then, the higher reduction of the isomerization rate
constants for 2-cyanophenol as compared to 2-isocyanophenol can be partially ascribed to
blocking excitation of the energetic N≡C stretching mode in the former molecule, and allowing
it in the case of 2.
10

3.4. Conformational relaxation in dark
As mentioned above, keeping the N2 matrix at 11 K in dark resulted in a decrease of
the population of the less stable 2t conformer, and a concomitant increase of the population of
the most stable 2c conformer (see Fig. 6 and S6). At such a low temperature, the observed
conformational relaxation cannot take place over the 2t  2c barrier (ΔΔE‡ above 11 kJ mol−1,
0
see Table 1). This transformation should, therefore, be essentially governed by tunnelling of
the hydrogen atom [47,48]. This isomerization mechanism was found to play an important role
in the conformational relaxation of other matrix-isolated molecules with OH group linked
directly to an aromatic ring such as derivatives of phenol [32,40-42] or of cytosine [49,50].
After increasing the amount of 2t in the N2 matrix, using the procedure described above,
the 2t  2c decay process was followed as a function of time, by recording a few IR spectra
in the presence of the filter cutting the IR-radiation above 2200 cm−1. The population decrease
of 2t (and increase of 2c) is represented in Fig. 7. Fitting the experimental data with a single-
exponential function, a rate constant of k  3.2 × 10−4 s−1 was estimated for the observed dark
process, which corresponds to a half-life time of t1/2  36 min. These values are practically
coincident with those recently reported by us for the corresponding dark transformation in
2-cyanophenol isolated in a N2 matrix at 15 K (kt-c  3.6 × 10−4 s−1 ; t1/2  32 min) [32], where
a barrier with nearly identical shape separates the higher from the lower energy rotamer (see
Fig. S2), which explains the observation of similar tunneling rates.
Finally, it is of interest to compare the rate constant estimated for the 2t  2c
isomerization in dark with the rate constants estimated above for the same isomerization when
the matrix is exposed to the spectrometer source. The obtained results show that the tunneling
rate represents  38 % or  20% of the overall 2t  2c rate process in the presence of filtered
or unfiltered broadband IR-light, respectively. Hence, for the matrix exposed to the
spectrometer beam, the tunneling rate is a non-negligible contribution to the overall 2t  2c
isomerization rate, in agreement also with the results previously reported for 2-cyanophenol
[32].
4. Conclusions
Isocyanophenol (2) was photochemically produced in low-temperature Ar and N2
matrices by irradiating benzoxazole (1) with UV-light at λ = 233 nm. In the Ar matrix, only the
most stable 2c rotamer was captured and spectroscopically characterized. In contrast, both 2c
11

and 2t rotamers were identified in an N2 matrix due to the high stabilizing effect of this matrix
medium as compared to the noble gas matrix. For 2-isocyanophenol isolated in solid N2,
selective excitation of the 2ν(OH) mode of the 2c conformer with NIR laser light promoted the
2c  2t transformation. The spectral changes resulting from this conversion allowed a detailed
assignment of the IR spectra of the two conformers. When the N2 matrix is exposed to the
broadband IR-radiation (unfiltered or filtered) emitted by the spectrometer light source, the
conformational distribution always converges to a stationary state. Excitation of low energy
vibrational modes (below 2200 cm−1) was found to play an important role in the IR-induced
conformational interconversion. Not only photoinduced processes, but also tunneling
contributes to the conformational switching in the presence of the spectrometer beam. For a
sample maintained in dark, the higher-energy conformer undergoes a spontaneous tunneling
conversion into the lower energy conformer, with a half-life time  32 min. Very interestingly,
over-the-barrier (IR-induced) and through-the-barrier (hydrogen tunneling) OH
rotamerizations were found to occur on the same time scale (half-lives on the order of tens of
minutes), as it follows from the values of the obtained rate constants.
Acknowledgments
This work was supported by Project POCI-01-0145-FEDER-028973, funded by FEDER, via
Portugal 2020 - POCI, and by National Funds via the Portuguese Foundation for Science and
Technology (FCT) and by a bilateral project for scientific cooperation between Portugal (FCT)
and France (Program PESSOA) focused on studies of kinetically unstable compounds. The
Coimbra Chemistry Centre is supported by the FCT through the project UID/QUI/0313/2019,
cofunded by COMPETE. C.M.N. and I.R. acknowledge the FCT for an Auxiliary Researcher
grant and an Investigador FCT grant, respectively.
12

References
[1] I. Norman, G. Porter, Nature 174 (1954) 508-509.
[2] E. Whittle, D.A. Dows, G.C. Pimentel, J. Chem. Phys. 22 (1954) 1943-1943.
[3] J.D. Baldeschwieler, G.C. Pimentel, J. Chem. Phys. 33 (1960) 1008-1015.
[4] R.T. Hall, G.C. Pimentel, J. Chem. Phys. 38 (1963) 1889-1897.
[5] M. Pettersson, J. Lundell, L. Khriachtchev, M. Räsänen, J. Am. Chem. Soc. 119 (1997)
11715-11716.
[6] L. Khriachtchev, J. Lundell, E. Isoniemi, M. Räsänen, J. Chem. Phys. 113 (2000) 4265-
4273.
[7] M. Pettersson, E.M.S. Maçôas, L. Khriachtchev, J. Lundell, R. Fausto, M. Räsänen, J.
Chem. Phys. 117 (2002) 9095-9098.
[8] E.M.S. Maçôas, L. Khriachtchev, M. Pettersson, R. Fausto, M. Räsänen, Phys. Chem.
Chem. Phys. 7 (2005) 743-749.
[9] E.M.S. Maçôas, L. Khriachtchev, M. Pettersson, R. Fausto, M. Räsänen, J. Am. Chem.
Soc. 125 (2003) 16188-16189.
[10] L. Khriachtchev, J. Mol. Struct. 880 (2008) 14-22.
[11] K. Marushkevich, M. Räsänen, L. Khriachtchev, J. Phys. Chem. A 114 (2010) 10584-
10589.
[12] S. Lopes, A.V. Domanskaya, R. Fausto, M. Räsänen, L. Khriachtchev, J. Chem. Phys.
133 (2010) 144507 (1-7).
[13] A.J. Lopes Jesus, C.M. Nunes, R. Fausto, I. Reva, Chem. Commun. 54 (2018) 4778-
4781.
[14] S. Góbi, C.M. Nunes, I. Reva, G. Tarczay, R. Fausto, Phys. Chem. Chem. Phys. 21
(2019) 17063-17071.
[15] S. Góbi, I. Reva, I.P. Csonka, C.M. Nunes, G. Tarczay, R. Fausto, Phys. Chem. Chem.
Phys. 21 (2019) 24935-24949.
[16] R. Passerini, Journal of the Chemical Society (1954) 2256-2261.
[17] C. Araujo-Andrade, I. Reva, R. Fausto, J. Chem. Phys. 140 (2014) 064306 (1-14).
[18] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato, X.
Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M. Hada, M.
Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J.A. Montgomery, Jr., J.E. Peralta, F. Ogliaro, M.
Bearpark, J.J. Heyd, E. Brothers, K.N. Kudin, V.N. Staroverov, T. Keith, R. Kobayashi,
J. Normand, K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M.
Cossi, N. Rega, J.M. Millam, M. Klene, J.E. Knox, J.B. Cross, V. Bakken, C. Adamo,
J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C.
Pomelli, J.W. Ochterski, R.L. Martin, K. Morokuma, V.G. Zakrzewski, G.A. Voth, P.
Salvador, J.J. Dannenberg, S. Dapprich, A.D. Daniels, Ö. Farkas, J.B. Foresman, J.V.
Ortiz, J. Cioslowski, D.J. Fox, Gaussian 09. Gaussian, Inc., Wallingford, CT, 2013.
[19] A.D. Becke, J. Chem. Phys. 98 (1993) 5648-5652.
[20] S.H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 58 (1980) 1200-1211.
[21] C.T. Lee, W.T. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785-789.
[22] C. Møller, M.S. Plesset, Physical Review 46 (1934) 618-622.
[23] R. Krishnan, J.S. Binkley, R. Seeger, J.A. Pople, J. Chem. Phys. 72 (1980) 650-654.
[24] M.M. Francl, W.J. Pietro, W.J. Hehre, J.S. Binkley, M.S. Gordon, D.J. DeFrees, J.A.
Pople, J. Chem. Phys. 77 (1982) 3654-3665.
[25] M.J. Frisch, J.A. Pople, J.S. Binkley, J. Chem. Phys. 80 (1984) 3265-3269.
13

[26] J.H. Schachtschneider, F.S. Mortimer, Vibrational Analysis of Polyatomic Molecules.
VI. FORTRAN IV Programs for Solving the Vibrational Secular Equation and for the
Least-Squares Refinement of Force Constants. Project No. 31450. Structural
Interpretation of Spectra. Shell Development Co., Emeryville, CA, 1969.
[27] P. Pulay, G. Fogarasi, F. Pang, J.E. Boggs, J. Am. Chem. Soc. 101 (1979) 2550-2560.
[28] V. Barone, J. Bloino, C.A. Guido, F. Lipparini, Chemical Physics Letters 496 (2010)
157-161.
[29] J. Bloino, V. Barone, J. Chem. Phys. 136 (2012) 124108 (1-15).
[30] V. Barone, M. Biczysko, J. Bloino, Phys. Chem. Chem. Phys. 16 (2014) 1759-1787.
[31] H.G. Korth, M.I. de Heer, P. Mulder, J. Phys. Chem. A 106 (2002) 8779-8789.
[32] A.J. Lopes Jesus, C.M. Nunes, I. Reva, S.M.V. Pinto, R. Fausto, J. Phys. Chem. A 123
(2019) 4396-4405.
[33] T.D. Klots, W.B. Collier, Spectroc. Acta Pt. A-Molec. Biomolec. Spectr. 51 (1995)
1291-1316.
[34] G. Mille, G. Davidovics, J. Chouteau, Comptes Rendus Hebdomadaires Des Seances
De L Academie Des Sciences Serie B 274 (1972) 532-535.
[35] J.P. Ferris, F.R. Antonucci, R.W. Trimmer, J. Am. Chem. Soc. 95 (1973) 919-920.
[36] J.P. Ferris, F.R. Antonucci, J. Am. Chem. Soc. 96 (1974) 2014-2019.
[37] C.M. Nunes, S.M.V. Pinto, I. Reva, R. Fausto, Eur. J. Org. Chem. (2016) 4152-4158.
[38] Q. Cao, N. Andrijchenko, A.E. Ahola, A. Domanskaya, M. Räsänen, A. Ermilov, A.
Nemukhin, L. Khriachtchev, J. Chem. Phys. 137 (2012) 134305 (1-11).
[39] L. Lapinski, I. Reva, H. Rostkowska, A. Halasa, R. Fausto, M.J. Nowak, J. Phys. Chem.
A 117 (2013) 5251-5259.
[40] N. Akai, S. Kudoh, M. Nakata, J. Phys. Chem. A 107 (2003) 3655-3659.
[41] N. Akai, S. Kudoh, M. Takayanagi, M. Nakata, Chem. Phys. Lett. 356 (2002) 133-139.
[42] S. Nanbu, M. Sekine, M. Nakata, J. Phys. Chem. A 115 (2011) 9911-9918.
[43] C.M. Nunes, L. Lapinski, R. Fausto, I. Reva, J. Chem. Phys. 138 (2013) 125101 (1-12).
[44] A. Halasa, L. Lapinski, I. Reva, H. Rostkowska, R. Fausto, M.J. Nowak, J. Phys. Chem.
A 118 (2014) 5626-5635.
[45] M. Tsuge, L. Khriachtchev, J. Phys. Chem. A 119 (2015) 2628-2635.
[46] A.J. Lopes Jesus, I. Reva, C. Araujo-Andrade, R. Fausto, J. Chem. Phys. 144 (2016)
124306 (1-9).
[47] R.P. Bell, The Tunnel Effect in Chemistry, Chapman and Hall, London, 1980.
[48] D. Gerbig, P.R. Schreiner, J. Phys. Chem. B 119 (2015) 693-703.
[49] I. Reva, M.J. Nowak, L. Lapinski, R. Fausto, J. Chem. Phys. 136 (2012) 064511 (1-8).
[50] L. Lapinski, I. Reva, H. Rostkowska, R. Fausto, M.J. Nowak, J. Phys. Chem. B 118
(2014) 2831-2841.
14

cis-OH
trans-OH
2ν(OH)
H-Tunneling
Ar/N2
N2

Highlights
UV-irradiation (λ=233 nm) of benzoxazole isolated in Ar or N2 matrices yields 2-isocyanophenol.
2-isocyanophenol adopts in Ar matrices only the cis conformation, while in N2 matrices both cis and
trans conformers are captured.
Narrowband irradiation at the frequency of the first OH stretching overtone (6986 cm−1) promotes
vibrational excitation of the cis conformer, resulting in cis->trans isomerization in N2 matrices.
The trans conformer decays to the ground state cis form in N2 matrices in dark, via tunneling.
Structural and vibrational characterization of the two conformers of 2-isocyanophenol is reported.
15

Table 1. Relative electronic (ΔEel) and zero-point corrected (ΔE0) energies of 2c and 2t, energy
barriers for conformational interconversion (ΔΔE‡), and selected structural parameters,
computed at two levels of theory.a
H
H
O
O
C
C
N
N
2c
2t
B3LYP/6-311++G(d,p)
ΔEel
0.0
0.0
11.2
10.6
13.6
11.5
3.672
5.6
ΔE0
ΔΔE‡
ΔΔE‡
24.8
22.1
2.311
el
0
··· Å
109.8
−··· degree
MP2/6-311++G(3df,3pd)
ΔEel
0.0
0.0
12.1
11.4
13.3
11.0
3.640
6.0
ΔE0
ΔΔE‡
ΔΔE‡
25.4
22.4
2.266
el
0
··· Å
110.9
−··· degree
a All energies are in kJ mol−1. Relative energies ΔEel and ΔE0 are expressed with respect to the most
stable 2c conformer. E0 = Eel + ZPVE (zero-point vibrational energy). The absolute computed values of
2c are for B3LYP: Eel = −399.791881 Eh; E0 = −399.689588 Eh; and for MP2: Eel = −398.908034 Eh; E0 =
−398.805651 Eh. The energy barriers ΔΔE‡ and ΔΔE‡ are without and with the ZPVE contribution,
el
0
respectively. They are measured from the respective local minimum (either 2c or 2t).
6920 cm1
O
H
H
H
O
H
Ar, 15 K
N
N
N
2(OH)
6915 cm1
O
H
 = 224 nm
HO
HO
HO
H
3-OH-isoxazole
azirine A
azirine B
Fig. 1. UV-induced generation of elusive 3-hydroxy-2-formyl-2H-azirine (briefly, “azirine”) from
3-hydroxy-isoxazole (briefly, “3-OH-isoxazole”) in an Ar matrix at 15 K, and subsequent conformational
control of the formyl group in the azirine by NIR-induced narrowband vibrational excitation of the first
OH stretching overtones, 2ν(OH), of the respective azirine conformers, A and B.
16

O
 = 233 nm
O
N
H
C
1
2c
H
N+
Ar, 12 K
-
Calc.
(b)
100
50
0
100
50
0
2c
1
3700
3500
2100
1800
1500
1200
900
600
Ar @ 12 K,  = 233 nm
(a)
0.2
0.1
0.0
0.2
0.1
0.0
w
3700
3500
2100
1800
1500
1200
900
600
Wavenumber / cm1
Fig. 2. UV-generation of 2-isocyanophenol 2 in an Ar matrix at 12 K: (a) Experimental IR spectrum
showing changes after irradiation at  = 233 nm (18 min, ~10 mW) of matrix-isolated benzoxazole 1,
obtained as the spectrum of the irradiated sample minus the spectrum of the sample before
irradiation. The growing bands are due to photoproducts. The decreasing bands are due to 1 and are
truncated. The band marked with “w” is due to matrix-isolated water monomers; (b) Theoretical
difference IR spectrum assuming quantitative transformation of 1 into 2c, obtained as the spectrum
of 2c minus the spectrum of 1 (negative bands, truncated). In simulations, an fwhm of 1 cm–1 was used
(see the computational section for the frequency scaling details).
17

H
O
O
 = 233 nm
O
N
H
C
H
+
N+
N+
N2, 15 K
-
-
C
1
2c
2t
2c
Calc.
(b)
80
40
0
80
2t
40
0
1
1400
3700
3500
2200
2100 1500
1300
1200
1100
1000
(a)
N2 @ 15 K,  = 233 nm
0.2
0.1
0.0
0.2
0.1
0.0
3700
3500
2200
2100 1500
1400
1300
1200
1100
1000
Wavenumber / cm1
Fig. 3. UV-generation of two forms of 2-isocyanophenol 2 in an N2 matrix at 15 K: (a) Experimental IR
spectrum showing changes after irradiation at  = 233 nm (~1 h, ~10 mW) of matrix-isolated
benzoxazole 1, obtained as spectrum of the irradiated sample minus spectrum of the sample before
irradiation. The growing bands are due to photoproducts. The decreasing bands are due to 1 and are
truncated; (b) Simulated theoretical IR spectra of 2c (blue line, blue circles), 2t (red line, red squares)
and of 1 (black, negative, truncated). In simulations, the computed infrared intensities of 2c, 2t, and 1
were scaled by 0.8, 0.2, and (−1) respectively, and an fwhm of 2 cm–1 was used (see the computational
section for the frequency scaling details).
18

H
H
O
O
C
C
6986 cm1
N2, 15 K
N
N
2c
2t
Calc. (b)
80
0
160
0
100
0
2t
2c
-80
-160
-100
3700
3500
2300
2100
1600
1400
1200
1000
(a)
N2 @ 15 K, 6986 cm1
0.03
0.00
0.03
0.00
0.03
0.00
-0.03
-0.03
3500 2300
-0.03
3700
2100
1600
1400
1200
1000
Wavenumber / cm1
Fig. 4. Conformational 2c  2t isomerization in an N2 matrix at 15 K induced by vibrational excitation
of 2c. (a) Experimental IR spectrum showing changes after NIR-irradiation at 6986 cm−1 (~1.5 h, ~70
mW) of the sample that was earlier UV-irradiated at  = 233 nm, obtained as the spectrum of the NIR-
irradiated sample minus the spectrum of the sample before the NIR irradiation. The growing bands
are positive and the decreasing are negative; (b) Simulated theoretical IR spectra of 2c (blue line,
negative) and 2t (red line, positive). In simulations, the computed infrared intensities of 2c and 2t were
multiplied by (−1) and (+1), respectively, and an fwhm of 2 cm–1 was used (see the computational
section for the frequency scaling details).
19

100
90
(a)
(b)
80
2c, ~ 88 %
2c, ~ 90 %
70
2t, ~ 12 %
2t, ~ 10 %
20
10
0
0
20
40
60
80
100
0
20
40
60
80
100
Time / min
Time / min
Fig. 5. Time evolution of the population of 2c (circles) and 2t (squares) forms during the exposure of
matrix-isolated 2 (N2, 15 K) to (a) unfiltered and (b) filtered (blocked above  2200 cm−1) broadband
IR light emitted by the spectrometer source. The 2c : 2t conformational ratio at the start of the kinetic
measurements was 75 : 25 (solid symbols) or 95 : 5 (open symbols). The populations were estimated
from the integrated absorbance of the bands centered at 1341 cm−1 (2c) and 1329 cm−1 (2t), divided
by their respective calculated IR intensities (see Tables S3 and S4). Dotted traces represent the best
exponential fits to the experimental data. For all fits R2 = 0.98−0.99.
20

(b)
70
0
2c
2t
-70
1400
1200
1000
800
(a)
0.02
0.00
-0.02
1400
1200
1000
800
Wavenumber / cm1
Fig. 6. Spontaneous 2t  2c isomerization in dark. (a) Experimental IR spectrum showing changes
after keeping matrix-isolated 2 (N2, 11 K) in dark for 23 h, obtained as the spectrum at the end of
monitoring (after 23 h) minus that recorded immediately after NIR irradiations at 6986 cm−1, which
corresponds to the start of monitoring. The growing bands are positive and the decreasing are
negative. During the spectra acquisition, a longpass filter, only transmitting below 2200 cm−1, was
placed between the sample and the spectrometer source. Between measurements, a metal plate was
used to completely block the spectrometer beam. For the 2200-1300 cm−1 range, see Fig. S6. The
dashed rectangle designates a fragment of spectrum shown in Fig.7; (b) Theoretical difference IR
spectrum assuming quantitative transformation of 2t into 2c, obtained as the spectrum of 2c minus
the spectrum of 2t (negative bands). In simulations, an fwhm of 2 cm–1 was used (see the
computational section for the frequency scaling details).
21

0 min
30 min
120 min
0.04
0.02
0.00
2c
2t
1350
1340
1330
1320
1
Wavenumber / cm
Fig. 7. Fragments of infrared spectra of 2 isolated in N2 matrix at 11 K, showing the decay of
2t and the simultaneous growth of 2c in the dark. The populations of 2c and 2t were estimated
from the integrated absorbance of the bands centered at 1341 cm−1 and 1329 cm−1, divided by
their respective calculated IR intensities (see Tables S3 and S4). The solid (black) trace
represents the sample enriched with 2t by preceding NIR irradiations at 6986 cm−1, and
corresponds to the start of monitoring (t= 0 min). The spectra shown as green (dashed) and
pink (dotted) were collected after 30 and 120 min of monitoring, respectively. The sample was
kept in the dark during the decay, except for the recording IR spectra, when a longpass cut-off
filter transmitting only below 2200 cm−1 was used.
22

Author Contribution Statement
A. J. Lopes Jesus: Writing- Original draft preparation, Formal analysis;
Igor Reva: Writing - Original draft preparation, Writing - Review & Editing, Supervision;
Cláudio M. Nunes: Investigation, Conceptualization;
José P. L. Roque: Visualization, Formal analysis;
Sandra M. V. Pinto: Investigation;
R. Fausto: Funding acquisition, Resources, Supervision.
23

Declaration of interests
☒ The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.
☐The authors declare the following financial interests/personal relationships which may be
considered as potential competing interests:
24