Formation of a Tunneling Product in the Photorearrangement of o-Nitrobenzaldehyde

Article abstract of DOI:10.1002/anie.201705140

The photochemical rearrangement of o-nitrobenzaldehyde to o-nitrosobenzoic acid, first reported in 1901, has been shown to proceed via a distinct ketene intermediate. In the course of matrix isolation experiments in various host materials at temperatures as low as 3 K, the ketene was re-investigated in its electronic and vibrational ground states. It was shown that hitherto unreported H-tunneling dominates its reactivity, with half-lives of a few minutes. Unexpectedly, the tunneling product is different from o-nitrosobenzoic acid formed in the photoprocess: Once prepared by irradiation, the ketene spontaneously rearranges to an isoxazolone via an intriguing mechanism initiated by H-tunneling. CCSD(T)/cc-pVTZ computations reveal that this isoxazolone is neither thermodynamically nor kinetically favored under the experimental conditions, and that formation of this unique tunneling product constitutes a remarkable and new example of tunneling control.

Full text of DOI:10.1002/anie.201705140

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Formation of a Tunneling Product in the Photo-Rearrangement of
o-Nitrobenzaldehyde
Authors: Dennis Gerbig and Peter Richard Schreiner
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201705140
Angew. Chem. 10.1002/ange.201705140
Link to VoR: http://dx.doi.org/10.1002/anie.201705140
http://dx.doi.org/10.1002/ange.201705140

10.1002/anie.201705140
Angewandte Chemie International Edition
COMMUNICATION
Formation of a Tunneling Product in the Photo-Rearrangement of
o-Nitrobenzaldehyde
Dennis Gerbig[a] and Peter R. Schreiner*[a]
Abstract: The photochemical rearrangement of o-nitrobenzaldehyde
to o-nitrosobenzoic acid, first reported in 1901, has been shown to
proceed via a distinct ketene intermediate. In the course of matrix
isolation experiments in various host materials at temperatures as
low as 3 K, we re-investigated the ketene in its electronic and
vibrational ground states and show here that hitherto unreported H-
tunneling dominates its reactivity, with half-lives t1/2 of a few minutes.
Unexpectedly, the tunneling product is different from o-
nitrosobenzoic acid formed in the photoprocess: Once prepared by
irradiation, the ketene spontaneously rearranges to an isoxazolone
via an intriguing mechanism initiated by H-tunneling. High-level
corresponding aldehyde (the thermodynamic product) by
hydrogen tunneling through a high but narrow barrier.[7a] In
principle –but not shown yet– a reaction product that is neither
thermodynamically nor kinetically favored cannot be synthesized
from the same starting material by conventional means, but may
be accessible exclusively by tunneling. We demonstrate here
that this is indeed possible and consequently term this the
tunneling product.
As it turns out, 3 still shows exceptional reactivity even when
there is insufficient energy to overcome thermal barriers and in
the absence of ultraviolet or visible radiation. In the following,
computations
reveal
that
this
isoxazolone
is
neither
we provide evidence that ground state
3 undergoes an
thermodynamically nor kinetically favored under the experimental
conditions, and that formation of this unique tunneling product
constitutes a remarkable and new example of tunneling control.
intramolecular rearrangement initiated by hydrogen tunneling,
yielding the unique tunneling product 4, which is distinctly
different from the thermodynamically and kinetically favored
photoproduct 2 (Scheme 1).
The photochemical conversion of o-nitrobenzaldehyde (1) to
o-nitrosobenzoic acid (2) is a puzzling reaction that has attracted
considerable attention since its original discovery by Ciamician
and Silber.[1] Its significance lies in its diverse applications, in
particular, in the use of nitro-compounds as photo-cleavable
protecting groups in the context of “caging”, i.e., activation upon
photoirradiation, of physiologically active reagents.[2] The latter
provide great opportunities for the continuing development of
phototherapy as
a biocompatible alternative to traditional
systemic drug applications.
The rearrangement of 1 has been shown to proceed via
initial hydrogen transfer from the formyl to the adjacent nitro
group, yielding a ketene (3) as the primary photoproduct.[3] The
intermediacy of 3 could unequivocally be established by means
of matrix isolation infrared spectroscopy[4] as well as by ultrafast
Scheme 1. Photorearrangement of o-nitrobenzaldehyde (1): product 2 is
favored kinetically and thermodynamically, while hitherto unobserved
isoxazolone (4) is the product only accessible via a tunneling reaction.
transient
visible,
infrared
and
stimulated
Raman
spectroscopies.[5] Inherent vibrational excitation of “hot” 3 or
continued irradiation in the visible range then effect a secondary
rearrangement to final product 2 via hydroxyl group transfer in a
process akin to the concept of shuttle catalysis.[6]
In our investigations, we employed the matrix isolation
technique to generate ground state 3 to study its reactions under
cryogenic conditions. For that purpose, 1 was evaporated under
high vacuum conditions and co-deposited with a large excess of
either neon, argon, krypton, xenon, or dinitrogen onto an
optically transparent support as cold as 3 K at the time of
measurement. After photogeneration of 3 within the so-obtained
matrix by 313 nm light, we were able to discern new signals in
the ketene region of the infrared spectrum as well as a new
absorption in the visible range of the ultraviolet-visible spectrum.
We turned our attention to this intriguing reaction within the
broader scope of our studies on tunneling control[7] that must be
considered as a way of understanding and ultimately controlling
chemical reactions as a viable alternative to the well established
principle of kinetic vs. thermodynamic control. Tunneling control
was
originally identified
for
the novel
family of
8]
hydroxycarbenes:[7a,
Despite the existence of a kinetically
Much to our surprise, we observed the rapid decay of the
ketene signals within minutes, which was not attributable to
photochemical bleaching of the matrix (Fig. 1). The first-order
decay process was accompanied by the concomitant
emergence of a new signal in the infrared, located at higher
frequencies than the carbonyl signal of 2 (Table 1). To exclude
the possibility that the observed reaction was a mere result of
infrared stimulation by globar radiation inside the spectrometer,
we covered the vacuum shroud immediately after generation of
3 and foregoing initial spectra recording, only to find its expected
signals had already vanished upon the follow-up measurement.
preferable reaction pathway to the corresponding vinyl alcohol,
many hydroxycarbenes exclusively rearrange to their
[a]
Dr. Dennis Gerbig, Prof. Dr. Peter R. Schreiner, Ph.D.
Institute of Organic Chemistry
Justus–Liebig University
Heinrich–Buff–Ring 17, 35392 Giessen, Germany
E-mail: prs@uni-giessen.de
Supporting information for this article is given via a link at the end of
the document.
1
This article is protected by copyright. All rights reserved.

10.1002/anie.201705140
Angewandte Chemie International Edition
COMMUNICATION
We performed quantum chemical computations to shed light
on the tunneling mechanism and the ensuing rearrangement
product. The newly emerging infrared signal in the carbonyl
region matches closest the spectral features of 1-hydroxy-
benzo[c]isoxazol-3(1H)-one (4), which has been suggested as a
possible trace product alongside 2 in a prior matrix isolation
study (Scheme 1).[4]
Table 2. Experimentally and theoretically obtained half-lives t1/2 in minutes of
the ketene intermediate (3) generated from o-nitrobenzaldehyde (1) in
different matrix materials by irradiation with 313 nm light.
1/2 (Ne)
1/2 (Ar)
1/2 (Kr)
1/2 (Xe)
1/2 (N2)
experimental
theoretical[c]
1.70.1[a]
2.00.2[b]
2.20.1[b]
3.50.1[b]
814[a]
Figure 1. Decay of ketene intermediate (3, downward arrows) and concurrent
emergence of tunneling product (4, upward arrows) in neon at 3 K (top trace)
and argon at 8 K (bottom trace) within a 45 min period in the dark.
11
[a] Measurement at 3 K. [b] Measurement at 8 K. [c] computed for gas-
phase at CCSD(T)/cc-pVTZ//MP2/aug-cc-pVDZ at 0 K.
As there is insufficient thermal energy under our cryogenic
conditions, 3 cannot overcome any reaction barrier in a classic
fashion. Thus, only a quantum mechanical tunneling process
provides a satisfactory explanation for the observed reactivity.
The tunneling rate constants show an expected dependence on
The structure of 3 as initially generated prohibits the direct
formation of isoxazolone (4): Since the torsional movement of
the NOOH moiety along the C=N bond is not feasible due to the
quinoid nature of 3, an intramolecular [1,3]-hydrogen transfer
initiated by tunneling followed by instantaneous ring formation
appears as a viable pathway. This mechanistic proposal is
the matrix host material,[9] with the tunneling half-lives (t1/2
,
Table 2) increasing in the order of neon < argon < krypton <
xenon: Guest molecules are likely to be increasingly stabilized
by London dispersion interactions in the solid host material.
Molecular nitrogen, on the other hand, strongly stabilizes 3 likely
by coordination,[10] thereby increasing t1/2 by more than an order
of magnitude. Rates measured in argon at 8 K and 20 K are
also identical within the experimental margin of error, which
further supports a tunneling mechanism.
further substantiated by repeating the experiment with
1
deuterated at the formyl group (d1-1): intermediate d1-3 is
persistent in the absence of ambient light for at least several
days. A hydrogen/deuterium kinetic isotope effect this large
(computed as >106, vide infra) is highly indicative of a tunneling
process due to the pronounced influence of the mass of the
tunneling particle on the rate constant.[11] We could indeed
reproduce the kinetic isotope effect computationally by
Table 1. Infrared signals  in wavenumbers (cm–1) of the ketene intermediate
(3) and 1-hydroxy-benzo[c]isoxazol-3(1H)-one (4) in neon and argon (values
for other host materials are given in the SI); theoretical values theor and Irel (in
) computed at B3LYP/6-311++G(d,p) (unscaled; results from
anharmonic frequency computations in curly brackets).
assessing the tunneling rate constants of
3
with the
approxi-
11]
semiclassical Wentzel-Kramers-Brillouin (WKB)[7a,
mation (Table 2): While we obtain a value of several minutes for
3 at the CCSD(T)/cc-pVTZ//MP2/aug-cc-pVDZ level of theory,
the half-life of d1-3 amounts to several years.
km mol–1
approx.
description
[d]
(Ne)[a]
(Ar)[b]
ref.[c]
theor
Itheor
2122.8
2109.2
2118.3
2108.1
2118
2107
2188.1
{2145.2}
{2140.0}
1156
{794}
{108}
(C=C=O)
3
4
[f]
1256.1
{1230.8}
110
{137}
δ(NOH),
δ(CCC) arom.
[e]
1232.6
3606.6
1230.5
3581.5
3577.8
3580
3577
3783.9
1847.6
108
577
(OH)
1821.3
1814.8
1810.8
1805.7
1799.7
1804
1799
(C=O)
Figure 2. Schematic reaction path from ketene (3) to isoxazolone (4).
Quantum mechanical tunneling (QMT) only occurs through the narrowest
barrier.
1063.3
1061.1
1059.0
1057.7
(C–O), ring
deform.
[e]
1057.2
1069.6
131
In the course of these tunneling computations, we mapped
the intrinsic reaction path[12] of the transformation (Figure 2): It is
evident that the traversed enthalpic barrier only applies to the
[a] Measurement at 3 K. [b] Measurement at 8 K. [c] values in Ar from [4]
[d] gas phase values [e] not observed [f] combination δ(NOH), δ(CCC) arom. +
(C–O).
2
This article is protected by copyright. All rights reserved.

10.1002/anie.201705140
Angewandte Chemie International Edition
COMMUNICATION
initial [1,3]-hydrogen shift and that ring formation and
rehybridization of the isoxazolone nitrogen in fact occur
add yet another intriguing mechanistic feature to the rich
chemistry of the o-nitrobenzaldehyde system more than a
century after the original conception of its photorearrangement.[1]
We anticipate that the explicit consideration of tunneling and the
extent of control it exerts on reactivity is not limited to this
particular system but is of general nature, thereby affecting the
design of chemical reactions beyond the classic limits of TST.
subsequently in
a barrierless fashion (cf. Supplementary
Materials). Accordingly, additional stationary points along the
reaction path could not be located.
Upon further inspection of the ground state potential energy
surface around 3, which was also established at the
CCSD(T)/cc-pVTZ//MP2/aug-cc-pVDZ level of theory, it is
apparent that the formation of the usually observed
photoproduct 2 is thermodynamically preferred over that of 4
(Scheme 2).
Experimental Section
Matrix isolation. Experiments at 8 K were conducted on a Leybold
cryostat system consisting of a RGD 210 closed-cycle refrigerator cold
head and a RW2 compressor unit. Experiments at 3 K were carried out
on
a
Sumitomo cryostat system with
a RDK 408D2 closed-cycle
refrigerator cold head and
a
F-70 compressor unit. For infrared
measurements, the vacuum shroud was equipped with polished
potassium bromide windows. A polished cesium iodide window mounted
in the sample holder on the cold head served as the matrix substrate.
For ultraviolet-visible measurements, both the vacuum shroud and the
sample holder were outfitted with polished barium fluoride windows.
Temperatures of the matrices were determined with silicon diodes at the
base of the sample holders. Deposition temperatures were varied for the
matrix host gases, with 3 K for neon, 12 K for argon, 15 K for dinitrogen,
25 K for krypton, and 30 K for xenon. For irradiation of the matrices, a
200 W Osram HBO 200 high pressure xenon(mercury) lamp was used in
conjunction with a Bausch & Lomb monochromator. For wavelengths
≥ 405 nm, a Schott 375 nm cutoff filter was added to the beam path.
Spectroscopic measurements. Infrared spectra of matrix experiments at
8 K were recorded on a Bruker Equinox IFS55 Fourier-transform infrared
spectrometer, matrix experiments at 3 K were measured on two Bruker
Vertex 70 instruments (standard potassium bromide or wide-range
beamsplitters). Spectral data were recorded in the range from 6000 to
Scheme 2. Reaction pathways open to the ketene intermediate (3) computed
at CCSD(T)/cc-pVTZ//MP2/aug-cc-pVDZ + zero point vibrational energy (∆H0);
ground state enthalpies in black, transition state enthalpies in red. Reaction to
o-nitrosobenzoic acid (2) does not occur despite being favored kinetically and
thermodynamically.
400 cm–1 with a resolution of 0.5 cm–1
. Measurements in the ultraviolet-
visible region were conducted on Jasco V-670 and V-760
spectrophotometers in the range of 190–800 nm with a resolution of 1 nm.
Moreover, the enthalpic barrier separating 2 from precursor
3 in its ground state is by far the lowest, which should, under
consideration of classic Transition State Theory (TST),[13]
strongly favor the formation of 2 kinetically as well. In stark
contrast, however, neither reaction to 2 (nor back to 1) occurs,
but only 4 forms in the dark process: Its appearance is entirely
attributable to tunneling through a substantially higher, yet
sufficiently narrow barrier (Figure 2). This observation marks
another example of tunneling control and, to the best of our
knowledge, the first report of an exclusive tunneling product.
Kinetics. Tunneling kinetics of matrix-isolated 3 (2) were evaluated by
recording infrared spectra in regular intervals for a period of several
hours after 30–60 s irradiation of the matrix with 313 nm light. Between
repeated measurements, the optical path was blocked to avoid prolonged
irradiation of the matrix by the instrument globar. Additionally, a 3.6 m
cutoff filter was installed to effectively block infrared light > 2700 cm–1
and prevent obscuring the ground state reactivity of 3 (2) due to
vibrational overtone excitation.
Chemicals. Samples of 1 were sublimed before use. Synthesis of d1-1
was performed via reduction of methyl o-nitrobenzoate to d3-o-
nitrobenzylic alcohol with lithium aluminium deuteride and subsequent
Swern oxidation. The crude product was purified by flash column
chromatography as well as sublimation and showed an estimated
isotopic purity of > 90% for deuterium. For matrix isolation, host gases of
99.999% purity were used.
Note that 4 can in principle transform into 2 via a transition
state close in energy to the entrance channel (approx.
+2.2 kcal mol–1; cf. Supplementary Materials). While this
pathway is not operative under matrix isolation conditions, it
might well be relevant in solution at ambient temperatures,
similarly to closely related compounds.[14] Besides tunneling in
the rearrangement of 3, photoproduct 2 and its deuterated
analog O-d-2 undergo conformational tunneling between E- and
Computational. All computations employing density functional[16] and
Møller-Plesset[17] perturbation theory were performed with Gaussian09[18]
in revision E.01. Intrinsic reaction paths were mapped using the
Hratchian predictor-corrector algorithm and projected frequencies were
computed along the paths.[19] Coupled cluster[20] energies for each point
on a path were determined with CFOUR 1.0.[21] WKB computations were
carried out with Wolfram Mathematica 10.[22]
Z-conformers, thus mimicking the behavior of many other
15]
carboxylic acids.[9,
Experimental details and kinetic data
regarding these processes are provided in the Supplementary
Materials.
By unveiling the spontaneous rearrangement of ground state
3 in cryogenic matrices due to a tunneling process, we could
3
This article is protected by copyright. All rights reserved.

10.1002/anie.201705140
Angewandte Chemie International Edition
COMMUNICATION
3526-3529; e) D. Ley, D. Gerbig, J. P. Wagner, H. P. Reisenauer, P. R.
Schreiner, J. Am. Chem. Soc. 2011, 133, 13614-13621.
Keywords: matrix isolation
• nitrobenzaldehyde • photo-
chemistry • physical organic chemistry • tunneling
[9]
M. Pettersson, E. M. S. Maçôas, L. Khriachtchev, J. Lundell, R. Fausto,
M. Räsänen, J. Chem. Phys. 2002, 117, 9095-9098.
[10] K. Marushkevich, M. Räsänen, L. Khriachtchev, J. Phys. Chem. A 2010,
114, 10584-10589.
[11] W. H. Miller, N. C. Handy, J. E. Adams, J. Chem. Phys. 1980, 72, 99-
112.
[1]
[2]
G. Ciamician, P. Silber, Ber. Dtsch. Chem. Ges. 1901, 34, 2040-2046.
a) J. A. Barltrop, P. J. Plant, P. Schofield, Chem. Commun. London
1966, 822-823; b) A. Patchornik, B. Amit, R. B. Woodward, J. Am.
Chem. Soc. 1970, 92, 6333-6335; c) J. Engels, E. J. Schlaeger, J. Med.
Chem. 1977, 20, 907-911; d) J. H. Kaplan, B. Forbush, J. F. Hoffman,
Biochemistry 1978, 17, 1929-1935.
[12] a) K. Fukui, J. Phys. Chem. 1970, 74, 4161-4163; b) K. Fukui, Acc.
Chem. Res. 1981, 14, 363-368.
[13] a) H. Eyring, J. Chem. Phys. 1935, 3, 107-115; b) M. G. Evans, M.
Polanyi, Trans. Faraday. Soc. 1935, 31, 875-894.
[14] Y. V. Il'ichev, M. A. Schwörer, J. Wirz, J. Am. Chem. Soc. 2004, 126,
4581-4595.
[3]
a) P. De Mayo, S. Reid, Q. Rev. Chem. Soc. 1961, 15, 393-417; b) M.
V. George, J. C. Scaiano, J. Phys. Chem. 1980, 84, 492-495.
S. Kuberski, J. Gȩbicki, J. Mol. Struct. 1992, 275, 105-110.
[15] a) E. M. S. Maçôas, L. Khriachtchev, M. Pettersson, R. Fausto, M.
Räsänen, Phys. Chem. Chem. Phys. 2005, 7, 743-749; b) S. Amiri, H.
P. Reisenauer, P. R. Schreiner, J. Am. Chem. Soc. 2010, 132, 15902-
15904.
[4]
[5]
a) S. Laimgruber, W. J. Schreier, T. Schrader, F. Koller, W. Zinth, P.
Gilch, Angew. Chem. Int. Ed. 2005, 44, 7901-7904; b) S. Laimgruber, T.
Schmierer, P. Gilch, K. Kiewisch, J. Neugebauer, Phys. Chem. Chem.
Phys. 2008, 10, 3872-3882.
[16] a) A. D. Becke, Phys. Rev. A 1988, 38, 3098-3100; b) A. D. Becke, J.
Chem. Phys. 1993, 98, 5648-5652.
[6]
[7]
B. N. Bhawal, B. Morandi, ACS Catalysis 2016, 6, 7528-7535.
a) P. R. Schreiner, H. P. Reisenauer, D. Ley, D. Gerbig, C.-H. Wu, W.
D. Allen, Science 2011, 332, 1300-1303; b) D. Ley, D. Gerbig, P. R.
Schreiner, Org. Biomol. Chem. 2012, 10, 3781-3790.
[17] C. Møller, M. S. Plesset, Phys. Rev. 1934, 46, 618-622.
[18] M. J. Frisch et al., Gaussian Inc., Wallingford CT, 2009.
[19] a) H. P. Hratchian, H. B. Schlegel, J. Phys. Chem. A 2002, 106, 165-
169; b) H. P. Hratchian, H. B. Schlegel, J. Chem. Phys. 2004, 120,
9918-9924.
[8]
a) P. R. Schreiner, H. P. Reisenauer, F. C. Pickard, A. C. Simmonett,
W. D. Allen, E. Mátyus, A. G. Császár, Nature 2008, 453, 906-909; b) D.
Gerbig, H. P. Reisenauer, C.-H. Wu, D. Ley, W. D. Allen, P. R.
Schreiner, J. Am. Chem. Soc. 2010, 132, 7273-7275; c) D. Gerbig, D.
Ley, H. P. Reisenauer, P. R. Schreiner, Beilstein J. Org. Chem. 2010, 6,
1061-1069; d) D. Gerbig, D. Ley, P. R. Schreiner, Org. Lett. 2011, 13,
[20] J. Čížek, J. Chem. Phys. 1966, 45, 4256-4266.
[21] CFOUR v. 1.0, 2010, http://www.cfour.de. See SI for full citation.
[22] Mathematica
10.0,
Wolfram
Research
Inc.,
2014,
http://www.wolfram.com.
4
This article is protected by copyright. All rights reserved.

10.1002/anie.201705140
Angewandte Chemie International Edition
COMMUNICATION
COMMUNICATION
Triggered by Tunneling: The ketene
Dennis Gerbig, Peter R. Schreiner*
intermediate
photoexcited
derived
o-nitrobenzaldehyde
from
Page No. – Page No.
rearranges to a hitherto unobserved
tunneling-exclusive isoxazolone
product under cryogenic conditions,
thereby constituting a new reactivity
paradigm in chemistry.
Formation of a Tunneling Product in
the Photo-Rearrangement of o-
Nitrobenzaldehyde
5
This article is protected by copyright. All rights reserved.