Interplay of Conformational and Chemical Transformations of Ortho-Substituted Aromatic Nitroso Oxides: Experimental and Theoretical Study

Article abstract of DOI:10.1021/acs.joc.7b00537

The mechanism of the photooxidation of aromatic azides containing a substituent at one of the ortho positions (2,4-dimethoxyphenyl azide (1a) and 2-methyl-4-[(2E)-1-methylbut-2-en-1-yl]phenyl azide (1b)) was studied in acetonitrile. The electronic spectra and the kinetic regularities of the consumption of corresponding nitroso oxides, which are the reaction intermediates, were investigated by flash photolysis. Owing to the one-and-a-half order of the C-N and N-O bonds and asymmetric molecule structure these nitroso oxides exist as four conformers (cis/syn, cis/anti, trans/syn, and trans/anti). The conformers differ in the spectral properties and in the reactivity in various irreversible transformations. The only product, (2Z,4E)-4-methoxy-6-oxohepta-2,4-dienenitrile oxide (7a), was observed during photooxidation of 1a, whereas transformations of the nitroso oxide isomers derived from 1b led to a set of stable products: the cis/anti isomer was transformed into (3,4,7-trimethyl-3a,4-dihydro-2,1-benzisoxazol-5(3H)-ylidene)ethanal (10), the trans isomers recombined forming the corresponding nitro and nitroso compounds, and the most reactive cis/syn isomer was transformed into ortho-nitrosobenzyl alcohol 11. The last was oxidized slowly to the corresponding benzaldehyde 12. Interaction of 11 and 12 led to the formation of (Z)-1,2-bis(2-formyl-4-((2E)-1-methylbut-2-en-1-yl)phenyl)diazene-1-oxide (13). The DFT simulation and kinetic modeling of the nitroso oxide transformations as well as the product analysis allowed revealing the fine details of the mechanism of decay for these species.

Full text of DOI:10.1021/acs.joc.7b00537

Subscriber access provided by UNIV OF DURHAM
Article
Interplay of Conformational and Chemical Transformations of ortho-
Substituted Aromatic Nitroso Oxides: Experimental and Theoretical Study
Ekaterina Mikhailovna Chainikova, Alfia R. Yusupova, Sergey L. Khursan,
Aigul N. Teregulova, Alexander N. Lobov, Marat Faritovich Abdullin, Leniza
V. Enikeeva, Irek Marsovich Gubaydullin, and Rustam Lutfullovich Safiullin
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.7b00537 • Publication Date (Web): 11 Jul 2017
Downloaded from http://pubs.acs.org on July 14, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Organic Chemistry is published by the American Chemical Society.
1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Interplay of Conformational and Chemical Transformations of ortho-
Substituted Aromatic Nitroso Oxides: Experimental and Theoretical Study
,
†
†
†
†
Ekaterina M. Chainikova,* Alfia R. Yusupova, Sergey L. Khursan, Aigul N. Teregulova,
†
†
‡
‡
Alexander N. Lobov, Marat F. Abdullin, Leniza V. Enikeeva, Irek M. Gubaydullin, Rustam L.
†
Safiullin
†
Ufa Institute of Chemistry of the Russian Academy of Sciences, 71 pr. Oktyabrya, Ufa 450054, Russian Federation
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
‡Ufa State Petroleum Technological University, 1 Kosmonavtov St., Ufa 450062, Russian Federation
*Eꢀmail: kinetic@anrb.ru
ABSTRACT: The mechanism of the photooxidation of aromatic azides containing a substituent at one of the ortho
positions (2,4ꢀdimethoxyphenyl azide (1a) and 2ꢀmethylꢀ4ꢀ[(2E)ꢀ1ꢀmethylbutꢀ2ꢀenꢀ1ꢀyl]phenyl azide (1b)) was
studied in acetonitrile. The electronic spectra and the kinetic regularities of the consumption of corresponding nitroso
oxides, which are the reaction intermediates, were investigated by flash photolysis. Owing to the oneꢀandꢀaꢀhalf order
of the CꢀN and NꢀO bonds and asymmetric molecule structure these nitroso oxides exist as four conformers (cis/syn,
cis/anti, trans/syn, and trans/anti). The conformers differ in the spectral properties and in the reactivity in various
irreversible transformations. The only product – (2Z,4E)ꢀ4ꢀmethoxyꢀ6ꢀoxoheptaꢀ2,4ꢀdienenitrile oxide (7а) – was
observed during photooxidation of 1a, whereas transformations of the nitroso oxide isomers derived from 1b led to a
set of stable products: the cis/anti isomer was transformed into (3,4,7ꢀtrimethylꢀ3a,4ꢀdihydroꢀ2,1ꢀbenzisoxazolꢀ5(3H)ꢀ
ylidene)ethanal (10), the trans isomers recombined forming the corresponding nitro and nitroso compounds, and the
most reactive cis/syn isomer was transformed into orthoꢀnitrosobenzyl alcohol 11. The last was oxidized slowly to the
corresponding benzaldehyde 12. Interaction of 11 and 12 led to the formation of (Z)ꢀ1,2ꢀbis(2ꢀformylꢀ4ꢀ((2E)ꢀ1ꢀ
methylbutꢀ2ꢀenꢀ1ꢀyl)phenyl)diazeneꢀ1ꢀoxide (13). The DFT simulation and kinetic modeling of the nitroso oxides
transformations as well as the product analysis allowed revealing the fine details of the mechanism of decay for these
species.
INTRODUCTION
Aromatic nitroso oxides (ArNOO) are formed by the reaction of nitrenes in their ground triplet
1–4
state with molecular oxygen.
ArNOO have been characterized by matrix isolation
5
–9
spectroscopy. An unique feature of these labile species is their cisꢀtrans isomerism because of
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
8
,10
11
the N–O bond order of ~1.5.
Nitroso oxides have a singlet ground state and their electronic
4,12
structure may be described by superposition of main resonances A–C
(Chart 1).
Chart 1
The isomer forms of nitroso oxides differ from each another not only in spectral properties but
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
13ꢀ15
16
also in reactivity both in the decay reaction
and in reactions with oxidizable substrates.
A
1
3
consumption of most nitroso oxides studied is carried out by firstꢀorder reactions and in the
case of the trans isomer is the transformation to the cis isomer, which undergoes a nontrivial
14,15
redox isomerization via an aromatic ring opening to form diene nitrile oxide (Scheme 1).
Scheme 1. Transformations of the cis Form of Phenylnitroso Oxides
O
N O
O
O
N
O
N
O
R
R
R
Nitrile oxide, which is formed from 4ꢀmethoxyphenylnitroso oxide, was isolated and identified
14
by spectral methods. If the molecule of a parent nitroso oxide possesses a reactive center
suitable for the intramolecular addition of the CNO group then corresponding heteroꢀ and
1
5,17
carbocyclic compounds are products of a chain of transformations.
The presence of a
substituent at one of the ortho positions of aromatic nitroso oxide leads to doubling the number
of the isomeric forms of the species (Chart 2) and two isomeric products of their transformations
should be formed as a result.
Chart 2. Conformers of orthoꢀSubstituted Phenylnitroso Oxides
In this work, the mechanism of transformations of orthoꢀsubstituted nitroso oxides (2) which
are formed under photolysis of 2,4ꢀdimethoxyphenyl azide (1a) and 2ꢀmethylꢀ4ꢀ[(2E)ꢀ1ꢀ
methylbutꢀ2ꢀenꢀ1ꢀyl]phenyl azide (1b) in the presence of oxygen has been studied by
experimental and theoretical methods. By flash photolysis, the electronic spectra of the isomeric
forms of nitroso oxides 2 were obtained, and the kinetic regularities of their consumption were
2
ACS Paragon Plus Environment

Page 3 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
studied in acetonitrile. Analysis of the products of the photooxidation of azides 1, theoretical
modeling of elementary stages, as well as kinetic modeling of the process of the decay of nitroso
oxides 2 allowed revealing fine details of the reaction mechanism.
N3
NOO
R1
R1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
R2
R2
1
2
:
:
R = R = OMe
Me
Me
1
2
R = Me, R =
Flash Photolysis. Under flash photolysis the azide 1a solution in acetonitrile in the presence of
a transient absorption, lifetime of which depends on the detection wavelength, was observed
O
2
in the 350–570 nm range (Figure 1). In the shortꢀwavelength region of this range (350–440 nm),
the kinetic curves of the absorption decay were described by a firstꢀorder equation (1) (Figure 1a,
ꢀ1
Curve 1) with the rate constant of 0.14 ± 0.01 s at room temperature.
ꢄꢅꢆ
ꢀ
= ꢀ + ꢀ ꢃ
(1)
ꢁ
ꢂ
where A
0
and A
∞
are optical densities of the solution immediately after excitation and at the end
of the reaction, respectively. At wavelengths of 420ꢀ480 nm, in millisecond interval, the kinetic
curves were recorded (Figure 1a, Curve 2), which also were described by eq 1; the rate constant
3
ꢀ1
equals to (3.1 ± 0.1) × 10 s . Finally, in the range from 460 to 570 nm, the optical density
dependences on time had a biꢀexponential character and were described by eq 2 (Figure 1a,
Curve 3).
ꢇ
ꢄꢅꢈꢆ
ꢇꢇ ꢄꢅꢈꢈ
ꢆ
ꢀ = ꢀ + ꢀ ꢃ
+ ꢀ ꢃ
(2)
ꢁ
ꢂ
ꢂ
Therefore, in this spectral range, a consumption of two intermediates was observed with
I
II
somewhat different reactivity; in eq 2 A
0 0 I
, A , k , and kII are initial optical densities and rate
constants of consumption of two species, respectively, while A is an end optical density of
∞
ꢀ1
solution. The rate constants obtained using this equation equal to 1.63 ± 0.08 s and 0.77 ± 0.04
ꢀ1
s at room temperature.
Under flash photolysis of oxygenꢀcontaining solutions of azide 1b in acetonitrile, the formation
of several species was also observed. The most labile of them absorbs in the wavelength range of
3
60ꢀ460 nm and had a lifetime of two orders of magnitude smaller than other species. The
kinetic curves of the decay of this intermediate were well described by eq 1. (Figure 1b, Curve 3)
ꢀ1
with the rate constant of 75 ± 2 s at room temperature. At wavelengths from 360 to 410 nm, the
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
absorption decay obeyed a biꢀexponential eq 2 (Figure 1b, Curve 1). The rate constants of the
ꢀ1
ꢀ1
consumption of two components were 0.10 ± 0.01 s and 0.50 ± 0.01 s . In the longꢀwavelength
range (λ ˃ 410 nm) kinetic curves, treatment of which by eq 1 gives the rate constant of 0.50 ±
ꢀ1
0
.01 s , were recorded (Figure 1b, Curve 3). Therefore, under flash photolysis of azide 1b in the
presence of oxygen, the formation of intermediates of three kinds differing from each another by
lifetimes and absorption spectra were observed.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Time (s) (I)
3
Time (s) (I)
4
0
0
1
2
4
5
0
0
2
6
Time (s) (II)
10
Time (s) (II)
10
5
15
20
25
5
15
20
0
0
0
0
0
0
.25
.20
.15
.10
.05
.00
0.30
(
a)
(b)
0
0
0
0
.25
.20
.15
.10
450 nm
(
2)
(Axis II)
4
10 nm
(Axis I)
(3)
(Axis I)
(
1)
(Axis II)
380 nm
3
80 nm
(1)
0.05
4
50 nm
5
10 nm
1.0
(
3)
(2)
0.00
0
.00
0.02
0.04
0.06
0.08
0.10
0.0
0.5
1.5
Time (s)
Time (ms)
Figure 1. Kinetic traces observed following flash photolysis of azides 1a (a) and 1b (b) (2.5 ×
–
4
10 M) in oxygenꢀsaturated acetonitrile (T = 295 K). Solid lines correspond to the theoretical
description.
The absorption spectra of intermediates of the photooxidation of azides 1a and 1b were
constructed from the initial optical densities calculated by a treatment of kinetic curves using eqs
1 and 2 (See Supporting Information, Figures S1 and S2 on transient absorption complete spectra
at several time). The maxima of these spectra for the most stable species (denote it by the letter
A) correspond to ~380 nm in both cases (Figure 2). The absorption maxima of the most labile
species (B) are around 450 and 410 nm in the cases of azides 1a and 1b, respectively (Figure 2).
Two remaining intermediates of the photooxidation of azide 1a, with close reactivity, absorb a
light in the longꢀwavelength range with maxima at ~490 nm (species C) and ~510 nm (species
D) (Figure 2a). Analogous intermediates of the azide 1b photooxidation are apparently
consumed with similar rate constants, and their total absorption spectrum has a maximum at the
wavelength of ~440 nm (Figure 2b). It should be noted that the residual optical density on the
kinetic curves of the consumption of the most reactive intermediates B, lifetimes of which by 2–
4
orders of magnitude smaller than lifetimes of the remaining intermediates, equals the sum of
4
ACS Paragon Plus Environment

Page 5 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
the optical densities of these intermediates at the initial moment in time (Figure 1a, Curve 2;
Figure 1b, Curve 3).
0
0
0
0
0
.16
.12
.08
.04
.00
0.25
(
a)
(b)
C + D
A
0.20
0.15
0.10
0.05
0.00
D
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
B
C
B
A
3
50
400
450
500
550
360
380
400
420
440
460
480
500
Wavelength (nm)
Wavelength (nm)
Figure 2. Absorption spectra of species formed following flash photolysis of azides 1a (a) and
–4
1
b (b) (2.5 × 10 M) in oxygenꢀsaturated acetonitrile. T = 295 K.
Table 1. Rate Constants of Decay of Species Formed under Flash Photolysis of Azides 1a
ꢀ1
and 1b in OxygenꢀSaturated Acetonitrile [k (s )] (Т = 295 К)
ArN3
intermediate
A
B
C
D
1
a
0.14 ± 0.01 3100 ± 100
0.10 ± 0.01 75 ± 2
1.63±0.08
0.77 ± 0.04
a
1b
0.50 ± 0.01
a
The species C and D are consumed with similar rate constants.
The experimental rate constants of the decay of the species A–D are summarized in Table 1.
Previously, in a number of examples, it was shown that aromatic nitroso oxides (phenylnitroso
oxide and its paraꢀsubstituted derivatives) absorb in the spectral range from 350 to 650 nm
1
1,13–15,18
depending on the nature of a substituent and solvent.
Thus, the shortꢀlived absorption
observed under flash photolysis of solutions of azides 1 in the presence of oxygen corresponds to
nitroso oxides 2a and 2b, and the species A–D are their isomeric forms shown in Chart 2. The
absorption bands will be assigned later.
Products of Azide 1а Photooxidation. Based on the results obtained earlier in the study of the
14,15
mechanism of the decay of the isomeric forms of aromatic nitroso oxides
it was expected that
two isomeric nitrile oxides, 7a and 8a, should be the endꢀproducts of unimolecular
transformations of four isomers of nitroso oxide 2a (Scheme 2). The photolysis of the solution of
ꢀ4
azide 1a (6 × 10 М) in oxygenꢀsaturated acetonitrile was performed using the filtered light of a
xenon lamp (λ = 380–500 nm) at 293 K. The progress of the reaction was monitored by reverseꢀ
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
phase HPLC (See Experimental Section). Judging from the chromatogram of the reaction
mixture (See Supporting Information, Figure S1), a single product was formed as a result of the
reaction. This was isolated and analyzed. The obtained compound has a molecular weight equal
8 9 4
to the molecular weight of nitroso oxide 2а (C H NO ) – 183. The IR spectrum exhibits an
ꢀ1
intense band at 2297 cm corresponding to stretching vibrations of the C≡N bond. Analysis of
1
13
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the NMR spectra ( H and C) allowed the conclusion to be drawn that the obtained product is
(2Z,4E)ꢀ4ꢀmethoxyꢀ6ꢀoxoheptaꢀ2,4ꢀdienenitrile oxide (7а) (yield to azide consumed was 94%)
(See Experimental Section and Supporting Information). Thus, only the cis/syn form of nitroso
oxide 2a (structure 2aꢀ3 in Scheme 2) is transformed to the endꢀproduct of the reaction.
Scheme 2. Proposed Mechanism for Photooxidation of Azides 1
Products of Azide 1b Photooxidation. It was expected that the chain of unimolecular
transformations of nitroso oxide 2b should lead to the formation of nitrile oxides 7b and 8b
(Scheme 2). In this case, the proposed reaction mechanism is supplemented by the stage of the
intramolecular [3+2]ꢀcycloaddition of the nitrile oxide group to the double bond of the allyl
15,19,20
fragment to form benzisoxazoles 9 and 10
(Scheme 3).
The azide 1b photooxidation in acetonitrile led to a complicated mixture of products
(Figure 3a). During standing of the obtained reaction mixture in the dark at room temperature,
the signal of a product with the retention time t
R
= 3.8 min decreased and finally disappeared
(Figure 3b). It was accompanied initially by growth then decreasing and full disappearance of the
R R
signal of a product with t = 4.5 min and by growth of the signal at t = 7.1 min. This process
was continued for about 72 hours. The signals with the retention times of 3.3, 5.2, and 5.9 min
were not changed (Figure 3). Thus, obviously the mechanism of the azide 1b photooxidation is
6
ACS Paragon Plus Environment

Page 7 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
more complicated than that presented in Schemes 2 and 3. One can make a preliminary
conclusion that the reaction leads to stable products with the retention times of 3.3, 5.2 and 5.9
min as well as to a labile product with the retention time of 3.8 min which is transformed via an
intermediate with the retention time of 4.5 min into an endꢀproduct with the retention time of 7.1
min. A chromatographic separation of the reaction mixture was performed immediately after
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
stopping of the reaction to isolate intermediates (t
R
= 3.8 and 4.5 min) and after standing of the
= 3.3,
.2, 5.9, and 7.1 min). Identification of the obtained products was achieved using APCIꢀMS and
reaction mixture in the dark at room temperature for 72 hours to isolate stable products (t
R
5
1
13
H and C NMR spectroscopy (See Experimental Section and Supporting Information).
Scheme 3. Intramolecular Cyclization of Nitrile Oxides
The molecular weight of the product with the retention time of 3.3 min is 205, which equals the
molecular mass of nitroso oxide 2b (C12
trimethylꢀ3a,4ꢀdihydroꢀ2,1ꢀbenzisoxazolꢀ5(3H)ꢀylidene)ethanal (10, Scheme 3). Benzisoxasole
0 is formed from the cis/anti form of nitroso oxide 2b according to Schemes 2 and 3 as a
2
H15NO ). According to NMR data, this is (3,4,7ꢀ
1
mixture of two isomers 10a and 10b (ratio of about 2:1) (Chart 3), which differ by the aldehyde
group position with respect to the double bond.
Chart 3. Isomers of Benzisoxasole 10
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
[mV]
[%]
80
60
40
20
0
(
a)
8
6
4
0
0
0
20
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
0
0
2
4
6
8
[
min.]
Time
[mV]
[%]
30
(b)
8
0
20
60
4
2
0
0
1
0
0
0
0
2
4
6
8
10
[
min.]
Time
Figure 3. Chromatograms of the reaction mixture obtained by photolysis of an oxygenꢀsaturated
acetonitrile solution of azide 1b (5 × 10 М) at 293 K. The mobile phase was acetonitrile–water
92:8), and the detection wavelengths were 250 and 320 nm. After stopping the reaction (a) and
R
standing this reaction mixture in the dark for ~36 hours (b). t = 8.0 min (azide 1b). See text for
ꢀ4
(
details.
R
The molecular weight of a labile product with t = 3.8 min also equals the molecular mass of
nitroso oxide 2b (205). The NMR spectra of this compound confirm that the aromatic ring and
allyl substituent were not affected during the reaction and that its molecule contains a methylene
group. Thus, we came to the conclusion that this product is {5ꢀ[(2E)ꢀ1ꢀmethylbutꢀ2ꢀenꢀ1ꢀyl]ꢀ2ꢀ
nitrosophenyl}methanol (11) (Chart 4), which was evidently formed from the cis/syn form of
nitroso oxide 2b.
Chart 4. Products of Photooxidation of Azide 1b
11
Me
O
7
O
2
O
O
2
10
N
O
1
O
H
N
N
7
N
N
H
2
3
5
7
Me
O
N
H
9
8
1
1
7
12'
Me
Me
2
5'
3'
6'
1'
1
4
6
5
6
3
4
OH
O
8
'
Me
12
3
5
3
4
'
6
6
4
9'
2'
4
10
5
10
12
Me
11
Me
10
12
Me
11
Me
1
2
Me
11
10'
8
7
'
8
8
Me Me
9
9
Me
Me
O
9
11'
1
1
12
13
14
15
A product with t
R
= 4.5 min, which judging from chromatograms was formed from
nitrosobenzyl alcohol 11, was not obtained in quantities sufficient for NMR analysis. However,
8
ACS Paragon Plus Environment

Page 9 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
we were able to measure its molecular weight which is 203, i.e. the molecular mass of nitroso
oxide 2b minus the mass of two protons. Based on this, we concluded that this intermediate of
the photooxidation of azide 1b is 5ꢀ[(2E)ꢀ1ꢀmethylbutꢀ2ꢀenꢀ1ꢀyl]ꢀ2ꢀnitrosobenzaldehyde (12)
(Chart 4).
R
The molecular weight of a compound with t = 7.1 min, which was formed as the products 11
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
and 12 were being consumed (Figure 3), is 390, which equals the sum the molecular mass of 11
1
and 12 minus the molecular mass of H
2
O: 205 + 203 – 18 = 390. The H NMR spectrum
contains nine signal groups including the aldehyde ones. In each group, the signals are paired
13
and have the same intensity. The same was observed in the С NMR spectrum, which contains
twelve groups of paired signals (See Experimental Section and Supporting Information). This
confirms that a molecule has two identical fragments connected by a nonsymmetrical bridge.
Analysis of the obtained results allowed the conclusion to be drawn that this compound is (Z)ꢀ
1,2ꢀbis(2ꢀformylꢀ4ꢀ((2E)ꢀ1ꢀmethylbutꢀ2ꢀenꢀ1ꢀyl)phenyl)diazeneꢀ1ꢀoxide (13) (Chart 4). On the
basis of the above, we can offer a formal scheme for the formation of azoxy compound 13 from
nitroso oxide 2b (Scheme 4). According to this scheme, an initial intramolecular redox
isomerization of the cis/syn form of nitroso oxide 2b into nitrosobenzyl alcohol 11 occurs. A
more detailed analysis of the proposed scheme with the using of quantumꢀchemical modeling
methods is presented below.
Scheme 4. Proposed Scheme of the Transformation of the cis/syn Form of Nitroso Oxide 2b
into the Final product 13
O
O
O
N
N
H
N
O
Me
OH [O]
O
Me
Me
Me
Me
Me
O
Me
Me
Me
H
O
N
H
Me
Me
N
+
-
H O
2
O
The molecular weights of products with the retention times of 5.2 and 5.9 min (Figure 3) are
205 and 189. Based on the NMR data, they were identified as 2ꢀmethylꢀ4ꢀ[(2E)ꢀ1ꢀmethylbutꢀ2ꢀ
enꢀ1ꢀyl]ꢀ1ꢀnitrobenzene (14) and 2ꢀmethylꢀ4ꢀ[(2E)ꢀ1ꢀmethylbutꢀ2ꢀenꢀ1ꢀyl]ꢀ1ꢀnitrosobenzene
15) (Chart 4).
(
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Simultaneous formation of nitro and nitroso compounds upon the photooxidation of aromatic
11,17,18
azides was explained by the bimolecular reaction of nitroso oxides
(Scheme 5).
Scheme 5. Bimolecular Reaction of Nitroso Oxides to Form Nitro and Nitroso Compounds
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Thus, analysis of the products of the azide 1b photooxidation leads to the conclusion that the
mechanism of the decay reaction of the isomeric forms of nitroso oxide 2b is more complex than
might be expected based on Schemes 2 and 3. Three channels of the consumption of this species
have been revealed: (i) the isomerization of the cis/anti form into benzisoxazole 10, (ii) the
isomerization of the cis/syn form into nitrosobenzyl alcohol 11, and (iii) the bimolecular decay
reaction in which all four isomeric forms could participate.
To compare the obtained results with those of the kinetic studies, the yields of the products 10,
1
1, 14, and 15, which were formed upon flash photolysis of azide 1b in the presence of oxygen,
were measured by HPLC using the calibration plots (Table 2).
Table 2. Yields of Products of the Transformations of Nitroso Oxide 2b upon Flash
a
Photolysis of Azide 1b in OxygenꢀSaturated Acetonitrile at Different Temperatures
5
5
5
[
product] × 10 (M)
[
1b] × 10
ꢁ[1b] × 10
0
T (K)
333
(M)
(M)
1
0
11
14
0.10 {4}
(6)
15
0.08 {6}
(7)
9
.7
1.75
1.88
1.58
1.12
1.34
0.41 {23}
29)
0.23 {12}
16)
0.82 {47}
(58)
(
10.6
293
264
253
251
0.97 {52}
(69)
0.12 {4}
(6)
0.08 {6}
(9)
(
10.6
10.6
10.7
0.07 {4}
(8)
0.57 {36}
(65)
0.07 {4}
(8)
0.17 {11}
(19)
0.07 {6}
(8)
0.53 {47}
(61)
0.10 {9}
(11)
0.17 {15}
(20)
0.06 {4}
0.54 {40}
(63)
0.08 {6}
(9)
0.17 {13}
(21)
(7)
a
The solution of azide was irradiated by six light pulses. In curly brackets yields (%) on azide consumed; in
parentheses the part of the product (%) in a sum of products 10, 11, 14, and 15.
As seen from this Table, nitrosobenzyl alcohol 11 (Chart 4), which formed from the cis/syn
form of nitroso oxide 2b, was the main product of the consumption of this species. Its yield
varies little with decreasing temperature from 293 to 251 K. In this temperature interval, the
percentage of benzisoxazole 10, a product of the isomerization of the cis/antiꢀ2b (Schemes 2, 3),
decreases by about twice, and the total yield of products 14 and 15 resulting from the
10
ACS Paragon Plus Environment

Page 11 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
bimolecular reaction of nitroso oxides, conversely, increases in the same proportion. Increasing
the temperature from 293 to 333 K led to a reduction of the alcohol 11 yield by 11%, the growth
of the benzisoxazole 10 that by 13%, and a very small decrease of the total yield (2%) of the
nitro and nitroso compounds 14 and 15.
DISCUSSION
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Geometries and energies. To explain the obtained results, DFT calculations were performed for
all conformers of nitroso oxides 2a and 2b (2ꢀ1–2ꢀ4, Scheme 2) as well as for the primary
products 5 and 6 (Scheme 2) of unimolecular decay of these species. In the 2a molecule, besides
the rotation of the nitroso oxide moiety around the N–O and C–N bonds determining the existing
of four conformers, a free rotation of ortho and para methoxy substituents around the CꢀO bonds
occurs. The optimal orientations of both methoxy groups were found using cis/syn isomer of 2,4ꢀ
dimethoxyphenyl nitroso oxide as a testing molecule (Chart 5, Structure II) and were used when
studying conformational transformations of 2a around NꢀO and OꢀO bonds. The search for the
optimal orientation of the substituents in nitroso oxide 2b was performed in the same way.
Chart 5. Standard Enthalpies and Gibbs Energies of the Conformers of cis/synꢀ2,4ꢀ
Dimethoxyphenyl Nitroso Oxide.
ꢀ
ꢀ
H° (kJ/mol)
G° (kJ/mol)
2.2
7.1
0.0
0.0
14.1
20.6
11.8
19.2
The structures of all conformers of 2a are displayed in Figure 4. The deviation from the
planarity of the nitroso oxide moiety relative to the 6π aromatic system due to a steric repulsion
between the NOO group and the methoxy substituent at the ortho position is a structural feature
of the cis/syn isomer 2аꢀ3. In other cases, the planarity of the NOO fragment is preserved. It was
found that the cis and trans states of 2a have a close energies, which is consistent with the results
reported in the work of ref 21 for a number of paraꢀ and orthoꢀsubstituted phenylnitroso oxides.
11
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
аꢀ2
2аꢀ3
20.7
22. 6
ꢁ
H
°
(kJ/mol)
5.1
6.0
ꢀG° (kJ/mol)
2
аꢀ1
2аꢀ4
0.0
ꢁ
H°
(kJ/mol)
3.0
3.5
ꢀ
G
°
(kJ/mol)
0.0
Figure 4. Stable conformers of nitroso oxide 2a. The standard enthalpies and Gibbs energies of
conformers with respect to 2aꢀ4 are presented. Bond distances in angstroms and bond angles in
degree.
The cis/anti isomer 2аꢀ4 is most stable conformer of 2a and the structure 2аꢀ3 is most labile,
which undoubtedly is connected to the steric effect. The structural features and relative energies
of the conformers of nitroso oxide 2b in the main are similar to those of nitroso oxide 2a. The
only important difference is the highest thermodynamic stability of the trans/anti isomer 2bꢀ1,
12
ACS Paragon Plus Environment

Page 13 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
although the relative energy of the cis and trans forms, as for other nitroso oxides, do not differ
very much. The presence of the ortho methyl group in nitroso oxide 2b leads to even greater
deviation from the planarity of the NOO group in the cis/syn isomer 2bꢀ3 (45.4 degrees) than in
2aꢀ3 (Figure 4). Therefore, the cis/syn isomer 2bꢀ3 is also most labile (ꢀG°= 22.4 kJ/mol).
Geometric parameters, absolute enthalpies, and Gibbs free energies of all compounds and
transition states studied are presented in Supporting Information.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Formation of Products. The theoretical modeling of the transformations of two cis forms of
nitroso oxide 2a into corresponding nitrile oxides (Scheme 2) was performed. The Gibbs
activation energy of the stage of the formation of the bicycle 5a from the cis/syn form 2aꢀ3 was
43.1 kJ/mol, whereas the activation barrier of the analogous transformation of cis/antiꢀ2a was
almost twice as high – 84.7 kJ/mol. Such a large difference in the activation energies of the two
possible pathways for the consumption of nitroso oxide 2a explains why nitrile oxide 7a was the
only product of the azide 1a photooxidation.
Analogous to 1a, we expected that benzisoxazole 9, which is formed from the cis/syn form of
nitroso oxide 2b (Schemes 2, 3), would be the main product of the azide 1b photooxidation.
However, nitrosobenzyl alcohol 11, as the main product, and benzisoxazol 10, the precursor of
which is nitrile oxide 8b formed in turn from the cis/anti isomer 2bꢀ4, were found in the reaction
mixture (Table 2). The M06ꢀL/6ꢀ311+G(d,p) calculations give the activation barrier of the orthoꢀ
≠
cyclization of the cis/anti isomer 2bꢀ4
ꢀ
G = 77.1 kJ/mol. At the same time the analogous
≠
transformation of the cis/syn isomer 2bꢀ3 with
ꢀ
G = 61.2 kJ/mol was not observed. This result
is not associated with an error of the theoretical estimation of the activation barriers using M06ꢀL
≠
functional. Indeed, the difference in the activation energies of the orthoꢀcyclization ꢀꢀG =
≠
≠
ꢀ
G (cis/anti) – ꢀG (cis/syn) were reꢀcalculated using mPWPW91 and HCTH density
functionals, the applicability of which when describing properties of nitroso oxides was
≠
demonstrated in the work of ref 21. In all cases, the ꢀꢀG values were in the interval from 16 to
19 kJ/mol. This seeming contradiction is explained by the presence of the energetically more
favorable in comparison with the orthoꢀcyclization channel of the consumption of the cis/syn
form of nitroso oxide 2b, namely the direct oxidation of the ortho methyl substituent by the
nitroso oxide group to form nitrosobenzyl alcohol 11. The Gibbs activation energy of this
channel calculated using three functionals mentioned above on average by 10 kJ/mol smaller
than the activation energy of the orthoꢀcyclization of the cis/syn isomer 2bꢀ3, i.e. the competition
of two channels of the 2bꢀ3 decay to form 7b
→ 9 and 11 → 12 → 13 is in favor of the last: the
ratio of the reaction rates w /w is about 60 (295 K). It should be noted that such intramolecular
11
7b
oxidation of an ortho substituent by the NOO group is new type of rearrangement of aromatic
13
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
nitroso oxides which was found for the first time in this work. The theoretical modeling of the
process of the formation of the compound 11 was performed in the M06ꢀL/6ꢀ311+G(d,p)
approximation and is illustrated in Figure 5. The first stage of this mechanism is the abstraction
of the H atom from the ortho methyl substituent by the terminal oxygen atom of the nitroso oxide
group of the cis/syn isomer of 2b. Favorable conditions for the occurrence of this transformation
are the spatial proximity of the reaction centers, a high chemical potential of nitroso oxide, and
reduced strength of the C–H bond owing to the possibility of stabilization of the intermediate
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
formed due to the restructuring of the
π system of the molecule. It was found that the activation
enthalpy of the H atom abstraction was 48.2 kJ/mol.
Figure 5. Energetic diagram of the compound 11 formation. Calculation by the M06ꢀL/6ꢀ
311+G(d,p) + IEFPCM approximation.
The product of the H atom abstraction is the syn isomer of hydroperoxynitrene (HPN), which
possesses a chain of three electronegative atoms bonded to each other by an ordinary covalent
bond, =NꢀOꢀO. Heterocyclic compounds containing such structural fragment were discussed by
authors of ref 22. The conformational rotation around the O–O bond and the formation of the
thermodynamically more favorable anti form of HPN is accompanied by overcoming a small
≠
activation barrier (TS2 in Figure 6,
ꢀ
H = 1.1 kJ/mol, calculation for gas phase) and by
substantial elongation of the peroxide bond (1.573 Å). A test of a stability of the anti isomer
wave function indicates a significant spin polarization in the molecule. However, the reꢀ
optimization of the structure of the anti form of HPN with account of the nonꢀequivalence of
α
and electron systems led immediately to nitrosobenzyl alcohol 11. Thus, the activation barrier
β
of the conformational transformation of hydroperoxynitrene TS2 is essentially the one of the
isomerization of HPN into 11. The account of solvation effects decreases the value of this barrier
14
ACS Paragon Plus Environment

Page 15 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
up to zero (Figure 5). Generally, the results of the theoretical modeling of the new isomerization
reaction of aromatic nitroso oxides are consistent with the preferable formation of nitrosobenzyl
alcohol 11 and testify that the limiting stage of this reaction is the intramolecular abstraction of
the H atom in 2bꢀ3, and the subsequent rapid transformation of HPN into the stable product 11
occur by the intramolecular hydroxylation of the methylene group. It is worth to note the found
new reaction of ArNOO is consistent with known chemical behavior of 1,3ꢀdipolar species
family which includes ozone, carbonyl oxides, and nitroso oxides. The H atom abstraction with
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ozone is known to be the initiation step of freeꢀradical oxidation of organic substances by O
3
–O
The direct analogy with found transformation of 2b was observed
earlier for aromatic carbonyl oxides as well. Dimesitylketone Oꢀoxide was obtained in the
CFCl – (CF Br) glassy matrix at 77 K by irradiation of corresponding diazocompound in the
2
23,24
3 2
or O –N gas mixtures.
3
2
2
presence of oxygen. Warming to room temperature resulted in the disappearance of carbonyl
oxide and the formation of 2ꢀ(hydroxymethyl)ꢀ4,6ꢀdimethylphenyl mesityl ketone as the main
25
product:
It is obvious, the observed similarity in chemical properties of 1,3ꢀdipoles is caused to its
partially biradical character as it is shown in Chart 1 (resonance C).
The formation of azoxycompound 13 occurs more likely from nitrosobenzyl alcohol 11 via
reactions being typical for nitroso compounds. We can assume that during prolonged standing of
the reaction mixture, alcohol 11 was oxidized with oxygen to aldehyde 12. Indeed, using a highꢀ
precision volumometry setup, we have showed that the reaction mixture obtained by
photooxidation of azide 1b, when standing, is absorbing oxygen in a stoichiometric ratio of 1:1,
based on alcohol 11 consumed (See Supporting Information for details). It is well known that the
26
characteristic feature of nitroso groups bonded to a carbon atom is their ability to dimerize. An
interaction of alcohol 11 with aldehyde 12 gives the mixed dimer (Scheme 6), which loses the
water molecule to form the final product 13.
15
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 6. Proposed Mechanism for the Formation of Azoxy Compound 13
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Spectral properties and identification of isomers. Obviously, the aboveꢀdescribed spectral
detection of several species formed upon photolysis of the solutions of azides 1 in the presence
of oxygen is due to the existence of conformers of nitroso oxides 2 (Scheme 2). To assign an
observed absorbance to the particular structure, the electronic spectra of the conformers were
calculated using the TDꢀDFT method. The spectral characteristics of the longꢀwavelength S–S
transitions with maximal oscillator forces (fosc) are shown in Table 3.
The results of theoretical calculations of the electronic spectra are in reasonable agreement with
the observed spectral picture (Figure 2); therefore, establishing of the correspondence between
the experimental spectra and conformers of nitroso oxides does not cause difficulties. Consider
the assignments of the observed spectral bands to the particular isomers on the example of
nitroso oxide 2a. Previously, it was found that maxima of the absorption spectra of the cis
isomers of paraꢀsubstituted phenylnitroso oxides are blueꢀshifted in compare to those of the trans
1
3
isomers. Therefore, it can be argued that the species A, the absorption maximum of which is
about 380 nm (Figure 2a), is the cis/anti isomer 2aꢀ4. In addition, this isomer is
thermodynamically most stable, which explains the low rate constant of the consumption of the
intermediate A (Table 1, Figure 1a). On the contrary, species B, the absorption spectrum of
which is also in the shortꢀwave range with
λ
max
∼450 nm (Figure 2a), was consumed faster than
the other and corresponds to the activated by a steric interaction cis/syn isomer 2aꢀ3. It is
important to note that the TDꢀDFT calculations of the spectral properties of the cis isomers are in
good agreement with experimental data both in the
λmax position and in the observed
bathochromic shift of the absorption maximum of 2aꢀ3 (Table 3).
16
ACS Paragon Plus Environment

Page 17 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
a
Table 3. TDꢀDFT Modeling of Electron Spectra of Conformers of Nitroso Oxides 2
M06ꢀL
mPWPW91
CAMꢀB3LYP
exp
conformer
λ
max
f
osc
λ
max(nm)
f
osc
λ
max(nm)
f
osc
λ
max (nm)
(nm)
2
а
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2ꢀ1 (trans/syn) 386
2ꢀ2 (trans/anti) 407
0.368
0.302
0.214
0.385
424
453
438
393
0.568
0.500
0.263
0.441
407
431
398
380
0.561
0.574
0.240
0.469
490
510
450
380
2ꢀ3 (cis/syn)
ꢀ4 (cis/anti)
403
361
2
2
b
2
ꢀ1 (trans/syn) 384
ꢀ2 (trans/anti) 403
0.590
0.549
0.228
0.465
409
429
432
381
0.585
0.531
0.222
0.462
386
403
392
363
0.626
0.623
0.196
0.433
440
440
410
380
2
2
ꢀ3 (cis/syn)
406
359
2
ꢀ4 (cis/anti)
aCalculation by the DFT/6ꢀ311+G(d, p) + IEFPCM approximation, f is oscillator force.
osc
The intermediates C and D, which slightly differed from each another in the absorption spectra
Figure 2a) and had a moderate reactivity (Table 1), were assigned to the trans/syn (C) and
(
trans/anti (D) isomers of nitroso oxide 2a. The estimations of the thermodynamic stabilities of
the trans isomers of 2a (Figure 4) do not contradict such assignment. The TDꢀDFT calculations
for 2аꢀ1 and 2аꢀ2 overestimate an electron transition energy and, accordingly, underestimate the
value of λmax. A similar pattern in the calculation of the electronic spectra of nitroso oxides was
10
also noted previously. The reason for the observed deviations is currently unclear. We note
only that the use of other density functionals and complication of the basis set, an explicit
account of solvation effect by the supermolecule method, did not improve the situation in a
meaningful way. Yet another result that deserves attention is the significant decrease in the
oscillator force of the SꢀS transition in the cis/syn isomer (Table 3). The reason for this effect is a
deviation from the planarity in the isomer 2aꢀ3 (Figure 4) and a reduction of the conjugation of
the nitroso oxide group with the aromatic ring.
A similar assignment of the absorption bands was performed for all isomers of nitroso oxide
2b; the results are shown in Table 3.
17
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Kinetic modeling. Thus, upon flash photolysis of azides 1 in the presence of oxygen, four
isomeric forms of corresponding nitroso oxides 2 were generated. Their consumptions occurred
with different rates and led to various products. Nitrile oxide 7a was the only product of the
transformations of nitroso oxide 2a (Scheme 2), whereas the compounds 10–15 were formed as a
result of the consumption of nitroso oxide 2b. Obviously, the isomers of nitroso oxides undergo
mutual conformational transformations; therefore, the measured rate constant of the isomer
decay is effective, i.e. it describes a complex of reversible conformational transitions and
irreversible transformations into the reaction products. To confirm our ideas, the kinetic
modeling of the processes of the decay of nitroso oxides was performed in the framework of the
mechanism shown in Scheme 2, which was revised taking into account particularities of the
reactivity of each nitroso oxide. All model calculations were carried out for room temperature.
The initial concentrations of the conformers were calculated from the experimental values of
their initial optical densities using the corresponding oscillator force (Table 3) or were taken to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ6
be equal to 2.5 × 10 M (the total initial concentration of the conformers was always equal to
ꢀ5
1
0 M). To calculate initial approximations for rate constants of elementary stages, stable states
of all isomers and transition states of all their transformations were localized (See Supporting
Information), the absolute values of G° of all states were found, and the Gibbs free activation
≠
energies (∆G ) were determined by the M06ꢀL/6ꢀ311+G(d,p) method. Initial values of rate
constants of elementary stages were calculated by the Eyring equation (3):
ꢑ
ꢅ ꢌ
ꢋ
ꢄꢏꢐ
ꢉ = ꢊ
exp ꢎ( ꢒꢌ
)
(3)
ꢍ
where κ is Wigner tunneling correction factor that was taken to be 1. After solving the system of
the ordinary differential equations (ODE) the kinetic curves of the consumption of the
conformers were recovered. From these curves, the effective rate constants keff were calculated
and compared with the experimentally determined those. Then, the set of elementary rate
i
constants k was changed by varying the G° values for all conformers and transition states
involved in the process of the decay of nitroso oxide, and the ODE system was resolved. The
kinetic modeling was continued as long as the value of keff did not coincide with the
experimental rate constant in the range of 1–3%. The solution of the ODE system was performed
in the framework of the oneꢀstage secondꢀorder Rosenbrock scheme using the Matlab R2014a
computing environment. The genetic algorithm, which has worked well in solving the chemical
27
kinetics problems, was chosen as optimization method.
The case of 2a. The ODE system corresponds to Scheme 7, at that the intramolecular
≠
cyclization of the cis/syn isomer 2aꢀ3 to form the intermediate 5a (
ꢀ
G = 43.1 kJ/mol), which is
quickly transformed into the endꢀproduct 7a, is the only irreversible channel of the 2а
18
ACS Paragon Plus Environment

Page 19 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
consumption. Preliminary calculations have shown that the orthoꢀcyclization of the cis/anti
≠
isomer 2аꢀ4 requires much more energy: ∆G = 84.7 kJ/mol. Therefore, the probability of the
formation of intermediate 6a, and consequently nitrile oxide 8a, is extremely small (Scheme 2).
These results explain the fact that in the experiment, only nitrile oxide 7a was found to be a
product of the azide 1a photooxidation. Therefore, during the kinetic modeling of Scheme 7, the
alternative channel of the nitroso oxide 2a consumption was ignored, believing k46 = 0. We note
that such a pronounced differentiation of the direction of the intramolecular cyclization of nitroso
oxides is in full agreement with the results in ref 28, where it was established that the presence of
the methoxy substituent at the attacked ortho position noticeably reduces the activation barrier of
the orthoꢀcyclization of nitroso oxide. This paradoxical at first glance an effect is explained by
the steric repulsion of the substituents which distorts the planar structure of nitroso oxide,
making its geometry closer to the transition state of the orthoꢀcyclization (Figure 4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The solution of the system of five differential equations with nine kinetic constants (four pairs
of conformational transformations plus the irreversible consumption of 2aꢀ3, Scheme 7) was
ꢀ5
conducted with different initial conditions. The total initial concentration of 2a was 10 M, and
the initial distribution of the isomers 2аꢀ1–2аꢀ4 (0.30:0.18:0.34:0.17, respectively) was estimated
from the initial optical densities (Figure 2a) and values of fosc (Table 3), which are proportional
to the extinction coefficients of the isomers. Furthermore, the calculation was conducted with an
equimolar distribution of the isomers because the rate constants of the reaction of molecular
oxygen with triplet nitrenes to form the cis and trans nitroso oxides are apparently close. A
closeness of the rate constants of the formation of cisꢀ and transꢀHNOO was marked previously
29
on the example of the imidogene oxidation. The results of kinetic modeling showed a
negligible effect of the initial conditions to the main kinetic regularities of the process under
≠
study. The optimal values of the Gibbs free energies and corresponding values of ∆G
°
and ∆G ,
as well as the rate constants of the elementary stages of the reaction, were found. Calculated time
dependencies of concentrations of all participants of the process (See Supporting Information)
correspond to the effective rate constants of the consumptions of all isomers of 2a (Table 4).
19
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 7. Transformations of the Conformers of Nitroso Oxide 2a
O
O
N
O
O
N
O
N O
N
O
OMe
OMek35
MeO
OMe k23
O
k32
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
OMe
OMe
OMe
OMe
a-2
2
2a-3
5a
k21
k12
k34 k43
O
O
O
O
N
N
O
O
N
OMe
OMe
OMe k14
46
k41
OMe
OMe
OMe
a-1
2
2a-4
On the basis of the kinetic analysis it can conclude that there are three stages of the reaction
with different time ranges. In the first stage (millisecond range) (Figure 1a, Curve 2), the cis/syn
isomer was consumed via two parallel directions with keff(cis/sin)
≈ k35 + k34: the irreversible
conversion into 7a (77%) and 2аꢀ4 (23%). Its concentration rapidly decreased to values which
were not detectable spectrophotometrically, and further this isomer was not detected. The
concentrations of the trans isomers were practically not changed in this time range. At the second
stage (t ≤ 1.6 sec), the parallel consumption of the trans isomers 2aꢀ1 and 2aꢀ2, the accumulation
with approximately constant rate of nitrile oxide 7a, and the reaching of the cis/anti isomer 2aꢀ4
concentration of maximum were observed. The dominant process at this step of the reaction was
the conformational transformation of the trans forms at a depth of about 80% into the
thermodynamically most stable cis/anti isomer 2aꢀ4. Finally, at the third stage (t > 1.6 sec), the
transformation of 2aꢀ4 into 2aꢀ3 and the accumulation of 7a takes place; the concentration of the
trans isomers of 2a for about 5 sec was reduced to values that were not available for the
spectrophotometric detection (Figure 1а, Curve 3). The single experimentally observed channel
of the decay of nitroso oxide 2a became the transformation of 2aꢀ4 into 2aꢀ3 (Figure 1a,
Curve 2). However, due to the reversibility of this conversion, the effective rate constant of the
consumption of 2aꢀ4 (keff(cis/anti)) equals the product of the probability of the irreversible
consumption of 2aꢀ3 by the reaction of orthoꢀcyclization (
α
= k35/(k34 + k35)) and the rate
43. At this stage, the
constant of the transformation of 2aꢀ4 into 2aꢀ3 (k43): keff(cis/anti) ≈ α⋅
k
accumulation of the only stable reaction product 7a was about 20% of its total quantity.
20
ACS Paragon Plus Environment

Page 21 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
a
Table 4. Results of Kinetic Modeling of Decay of Nitroso Oxides 2
≠
b
+
b
ꢀ
c,d
eff
e
transformation
∆G
°
∆G
k
k
k
exp
2
a
2
аꢀ1
2аꢀ2
аꢀ3
аꢀ3
аꢀ4
2аꢀ2
2аꢀ3
2аꢀ4
5a
ꢀ0.4
9.5
70.8
88.6
56.9
47.9
80.8
2.4
0.0018
660
2.1
0.086
0.18
–
0.77
1.63
0.77
1.63
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
ꢀ20.3
–
3
100
3100
0.14
2
2400
2
2аꢀ1
11.2
0.043
3.9
0.14
f
2
b
2bꢀ1
2bꢀ2
2bꢀ3
2bꢀ4
11
6.7
ꢀ8.7
ꢀ13.0
–
63.1
86.0
63.2
65.0
91.6
82.2
54
0.0052
51
800
0.00016
0.27
–
0
.50 (15)
0.50 (15)
75 (69)
2
bꢀ2
bꢀ3
2bꢀ3
bꢀ4
2
75 (69)
25
2
2bꢀ1
6b
15.0
–
0.00055
0.024
0.23
0
.10 (16)
0.10 (16)
2bꢀ4
–
a
≠
ꢀ1
Relative stability of the isomers (∆G°), Gibbs activation energy (∆G ) in kJ/mol, rate constants in s (295 K).
Calculation by M06ꢀL/6ꢀ311+G(d, p) + IEFPCM approximation. Indices “+” and “ꢀ” correspond to direct and
reverse reaction, respectively. Calculated effective rate constant for each isomer; in the case of 2bꢀ1 and 2bꢀ2 k
corresponds to their joint consumption. Yields of corresponding products in % for nitroso oxide 2b are given in
brackets. Experimental values of keff. Rate constants kbi = 1.0 × 10 M s , krev = 2.4 s , kir = 36 s (Scheme 8).
In general, the kinetic modeling of Scheme 7 qualitatively and quantitatively describes the
observed experimental regularities. It should be noted that the Gibbs free energies of the
b
c
eff
d
e
f
6
ꢀ1 ꢀ1
ꢀ1
ꢀ1
conformers and transition states optimized during this modeling differed slightly from the G
°
values obtained by the DFTꢀcalculations. The maximal discrepancy was observed for the cis
isomers of 2a, the optimized Gibbs energies of which were decreased by 7.6 (2аꢀ3) and 5.3
(
2аꢀ4) kJ/mol. Apparently, the results of the DFT calculations somewhat underestimate the
stability of the cis conformers, perhaps because a specific solvation of the polar molecule of
nitroso oxide 2a by acetonitrile is ignored.
The case of 2b. Nitroso oxide 2b exhibited a greater variety of chemical transformations shown
in Scheme 8. Therefore, the system of eight ODEs contained thirteen elementary stages
including eight stages of conformational transitions between the isomers of 2b, two stages of
irreversible transformation of the cis isomers 2bꢀ3 and 2bꢀ4, and three elementary stages which
describe the bimolecular decay of the trans forms of 2b according to Scheme 5. To simplify the
mathematical model, the bimolecular reaction was described by the crossꢀreaction 2bꢀ1 + 2bꢀ2,
and the accumulation of products 14 and 15 was considered in the sum as a result of one
21
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
transformation. As the initial conditions for the ODE system solution, the equimolar ratio of the
isomers was used.
Scheme 8. Transformations of the Conformers of Nitroso Oxide 2b
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The optimization was performed by varying the G° values of all kinetically independent species
which were determined by the DFT calculations. The rate constants k (the bimolecular decay),
bi
k
rev (the reversible consumption of the nitroso oxide dimer), and kir (the irreversible
transformation of the dimer) were three additional optimization parameters which modeled the
recombination of the trans isomers of 2b. The condition for the convergence of a mathematical
model with experiment was the coincidence of the effective rate constants of the isomers
consumption (See above) as well as the equality of the calculated ratio of the reaction products
yields (10:11:(14 + 15) = 16:69:15%) and the experimental that (Table 2). The results of the
mathematic modeling of the decay of nitroso oxide 2b, namely the optimized G° values and the
elementary rate constants, are shown in Table 4. It can be seen that the mathematical model
describes the experiment well, i.e. correctly predicts the k values and the products yield.
eff
Despite the differences in the mechanism of the decay of nitroso oxides 2a and 2b (Schemes 7
and 8), the results of the kinetic modeling of two reactions had much in common. As was the
case for 2a, quickly, in about 20 milliseconds, the cis/syn isomer 2bꢀ3 was consumed (Figure 1,
Curve 3). The effective rate constant of the 2bꢀ3 consumption is the sum of two elementary
constants (k (cis/syn)
≈ k311 + k ), which correspond to the accumulation of product 11 (33%)
34
eff
and conformational transformation into 2bꢀ4 (67%). The accumulation of the cis/anti isomer
2bꢀ4 via the transformation of the trans isomers occurred within about 1.3 sec. The trans forms
of 2b were consumed in the parallel process, namely the conformational transformation into the
22
ACS Paragon Plus Environment

Page 23 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
cis/anti isomer and the recombination which led to the effective first order with respect to the
concentration of the trans isomers (Figure 1b, Curve 2). After 7–8 sec only the cis/anti isomer
2
bꢀ4 remained in the system (Figure 1b, Curve 1), the concentrations of the other isomers were
too small for spectrophotometric detection. The effective rate constant of the 2bꢀ4 consumption
can be represented as k (cis/anti) k46 + k43 k311/(k311 + k ), where the first and second terms
≈
⋅
eff
34
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
describing the accumulation of products 10 and 11, respectively. The percentage ratio of the
rates of two parallel processes is about 20% (10) and 80% (11), which explains the preferential
formation of nitrosobenzyl alcohol 11 (Table 2). Interesting to note that at the initial stage of the
reaction only 9% of the product 11 was formed directly from the cis/syn isomer 2bꢀ3. The
remaining 60% was accumulated during all process of 2b decay according to the scheme:
2
bꢀ4
The temperature dependence of the ratio of the yields of the products of flash photolysis of
azide 1b in the presence of oxygen testifies that when increasing the temperature from 293 to
33 K, the share of benzisoxazole 10, which was product of the irreversible transformation of the
→ 2bꢀ3 → 11.
3
cis/anti isomer of nitroso oxide 2b, significantly increased (Table 2). This is explained by the
higher activation energy of this channel in comparison with the formation of nitrosobenzyl
alcohol 11 from the cis/syn isomer (Table 4). At that, the total yield of the products of the nitroso
oxides recombination was only slightly reduced (Table 2), since this reaction does not require
large energy expenditures. Therefore, with decreasing temperature to 251 K, its share grew,
which was reflected in the percentage ratio of the products of the uniꢀ and bimolecular channels
of nitroso oxide 2b consumption (Table 4).
A complication of the kinetic scheme in the case of 2b in comparison with 2a led to an increase
in the deviations of the G
The maximal reduction of the Gibbs free energy was again observed for the cis isomers: 17.5
2bꢀ3) and 15.3 (2bꢀ4) kJ/mol (on the supposed reasons for these deviations – See above). The
maximal positive deviation of the G from the initial value (15.1 kJ/mol) was found for the
° optimal values from the initial ones obtained by the DFT calculations.
(
°
transition state of the conformational transition between the cis isomers. The detailed description
of the mathematic modeling of the nitroso oxides decay process is given in Supporting
Information.
The results of the mathematic modeling testify that the conformational transformations of
nitroso oxides 2a and 2b make a significant contribution to the kinetic regularities of their
consumption with the formation of the final products. Therefore, the rate constant of the
transformation of the cis/syn form of nitroso oxide 2a to form the bicycle 5a (Scheme 7) is
3
ꢀ1
2.4
⋅
10 s that is 1.4 times smaller than the effective rate constant of this intermediate
23
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
consumption (Table 4). An even more illustrative example is the formation of the products 10
and 11 from the cis form of nitroso oxide 2b (Scheme 8): the rate constants of these
ꢀ1
transformations are 0.024 and 25 s , respectively, which are 4 and 3 times less, respectively,
ꢀ1
ꢀ1
than the rate constants of the decay of 2bꢀ4 (0.10 s ) and 2bꢀ3 (75 s ) which were found
experimentally (Table 4). In addition, the bimolecular decay of the trans isomers of nitroso oxide
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2b complicated their conformational transformations was described by the effective first order
ꢀ1
with the rate constant of 0.50 s .
CONCLUSION
Thus, in this work, the comprehensive study of the mechanism of the photooxidation of
aromatic azides 1a and 1b containing the methoxy or methyl substituent at the ortho position was
performed. The kinetic regularities of the decay of four isomeric forms of corresponding nitroso
oxides 2a and 2b have been investigated and the quantumꢀchemical and kinetic modeling of this
reaction was carried out. Analysis of the obtained results together with data on the composition
of the products of the azides photooxidation allowed a holistic picture to be obtained, from
which it follows that the conformational transitions in orthoꢀsubstituted aromatic nitroso oxides
significantly affect the reactivity and chemical transformations of these species.
EXPERIMENTAL SECTION
HPLC grade acetonitrile (Cryochrom) were used without further purification. 2,4ꢀ
Dimethoxyphenyl azide (1a) was synthesized from 2,4ꢀdimethoxyaniline using the method
30
described for phenyl azide.
1D and 2D NMR spectra were recorded at 298 K on a Bruker Avance III 500 MHz instrument
1
with PABBO X{ H} direct detection probe in deuterated acetonitrile (Cambridge Isotope
Laboratories, Inc., degree of deuteration 99.8%). Chemical shifts for carbon and proton are
reported in parts per million (ppm) referenced to 0.0 ppm of TMS as the internal standard.
1
The H NMR spectra were acquired with a spectral width of 5.6 kHz and 32k data points and 8
1
13
1
scans, providing a digital resolution of ca. 0.5 Hz ( H 90° pulse width = 11.5 ꢂs). For C{ H}
NMR spectra (WALTZꢀ16), a spectral width of 29.7 kHz was used with 64k data points and
1
3
1
13
required quantity of scans ( C 90° pulse width = 9.7 ꢂs) . Gradient selected { H, C} HSQC
spectra were recorded using the Bruker pulse sequence library. These data were collected with
4
096
Gradient selected { H, C} HMBC spectra were collected with 4096
scans for each increment. The delay d6 was set to 71.4 ms. Spectral widths of 6.0 and 29.7 kHz
× 512 data points with 2 scans for each increment. The delay d4 was set to 1.72 ms.
1
13
×
512 data points with 4
1
13
1
13
were used in the F2 ( H) and F1 ( C) domains, respectively. { H, C} HSQC and HMBC data
1
1
were processed using a sine window in the F2 and F1 dimensions. { H, H} gsꢀCOSY data were
24
ACS Paragon Plus Environment

Page 25 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
1
1
collected with 4K
× 512 data points with 2 scans for each increment. For the { H, H} NOESY
NMR experiments, the solution was degassed to remove any dissolved oxygen. The following
parameters and procedures are commonly employed: spectral width 6.0 kHz, 4K data matrix and
256 time increments of 2 transients each, mixing time 0.5 s. Fourier transformations were carried
out with zeroꢀfilling using the shifted sineꢀbell apodization function in both dimensions.
The following abbreviations (or combinations thereof) were used to explain multiplicities: s =
singlet, d = doublet, t = triplet, q = quartet, m = multiplet, br = broad. Coupling constants, J, were
reported in Hertz unit (Hz).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
UVꢀvis spectra were recorded on a Shimadzu UVꢀ365 spectrometer.
The atmospheric pressure chemical ionization mass spectra (APCIꢀMS) were obtained on a
HPLC massꢀspectrometer LCMSꢀ2010EV (Shimadzu) (direct syringe sample inlet, sample
solution was in acetonitrile, mobile phase was acetonitrile/water (85:15)) in positive and negative
ions mode at the corona discharge needle ionizing electrode potential of 4.5 and ꢀ3.5 kV,
respectively. The mobile phase flow rate was 0.1 mL/min. The temperature and voltage of the
о
interface capillary was 250 С and 5 – (ꢀ5) V, respectively. The nebulizer gas (nitrogen) flow
о
rate was 2.5 L/min. The temperature and voltage of the curved desolvation line was 250 С and 5
о
–
(ꢀ5) V, respectively, heater temperature was 200 С.
High resolution mass spectra were recorded on a mass spectrometer MALDIꢀTOF Autoflex
III (Bruker, Germany) with sinapic acid as a matrix. Samples of compounds were prepared by
the “dried droplet” method (1:10).
HPLC analysis was carried out on an YL9100 liquid chromatograph equipped with a twoꢀ
wave UV–Vis detector. The reverseꢀphase analytical column was a ProteCol C18 GP125 4.6
50 mm (SGE), the mobile phase was acetonitrile–water (92:8 or 50:50), the flow rate of mobile
phase was 1 mL/min, and the column was operated at room temperature. A ReprosilꢀPur C18ꢀ
AQ 10 250 mm column (Dr. Maisch GmbH) was used for preparative separation of the
reaction mixtures. The mobile phase was acetonitrile.
×
2
×
16
Kinetic Experiments. A flash photolysis system of known design was used for the kinetic
experiments. The photolytic source was an IFP 5000ꢀ2 lamp; the maximum pulse energy was
400 J at U = 5 kV, С = 32 µF; ∼90% light energy was emitted in 50 µs. The reactor was a quartz
cell with an optical path length l = 10 cm, inner diameter of ~1 cm and volume of ~8 mL. The
–4
flash photolysis of 2.5
×
10 M azides 1 solutions saturated with oxygen was performed with
filtered light (UFSꢀ2 light filter; transmittance range
λ = 270–380 nm).
Synthesis of 2ꢀmethylꢀ4ꢀ[(2E)ꢀ1ꢀmethylbutꢀ2ꢀenꢀ1ꢀyl]aniline. 26.5 g (0.25 mol) of oꢀ
toluidine, 17.6 g (0.26 mol) of piperylene, 7.5 g (0.056 mol) of AlCl
3
, and 22 mL of hexane was
25
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
placed into a rotating stainless steel autoclave. The autoclave was heated at 130°C for 4 hours,
and then cooled. The reaction mixture was neutralized with a 30
organic layer was extracted with tertꢀbuthylꢀmethylether and dried over KOH. The solvent was
evaporated; the residue was distilled under vacuum. The fraction boiling at 145–150 C (10
% NaOH solution in water. The
°
mmHg) was 2ꢀmethylꢀ4ꢀ[(2E)ꢀ1ꢀmethylbutꢀ2ꢀenꢀ1ꢀyl]aniline which was obtained as a yellow
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
liquid in a yield of 14.6 g (29
%
). H NMR (CDCl
3
, 500 MHz):
δ
H
6.95 (1H, d, J = 2.0 Hz, H3);
6.94 (1H, dd, J = 8.5, 2.0 Hz, H5); 6.67 (1H, d, J = 8.5 Hz, H6); 5.65 (1H, ddq, J = 15.2, 6.7, 1.1
Hz, H9); 5.49 (1H, dqd, J = 15.2, 6.7, 1.1 Hz, H10); 3.36 (1H, m, H8); 2.21 (3H, s, H7); 1.73
13
(
dd, 3H, J = 6.0, 1.1 Hz, H11); 1.36 (3H, d, J = 6.7 Hz, H12). C NMR (CDCl , 125 MHz): δC
3
1
42.5 (s, C4), 137.0 (d, C9), 136.7 (s, C2), 129.1 (d, C3), 125.4 (d, C5), 122.8 (d, C10), 115.1 (d,
C6), 41.5 (d, C8), 21.6 (q, C12), 17.9 (q, C11), 17.4 (q, C7).
ꢀMethylꢀ4ꢀ[(2E)ꢀ1ꢀmethylbutꢀ2ꢀenꢀ1ꢀyl]phenyl azide (1b) was synthesized from 2ꢀmethylꢀ
ꢀ[(2E)ꢀ1ꢀmethylbutꢀ2ꢀenꢀ1ꢀyl]aniline via the procedure described in the literature for transꢀ1ꢀ(oꢀ
2
4
3
1
1
azidophenyl)ꢀ1ꢀpentene. 1b was obtained as a yellow liquid. H NMR (CDCl
3
, 500 MHz):
.09 (1H, dd, J = 8.5, 2.0 Hz, H5); 7.06 (1H, d, J = 8.5 Hz, H6); 7.03 (1H, d, J = 2.0 Hz, H3),
.60 (1H, ddq, J = 15.2, 6.7, 1.1 Hz, H9); 3.38 (1H, m, H8); 2.22 (3H, s, H7); 1.71 (3H, dd, J =
δH
7
5
6
13
3
.0, 1.1 Hz, H11); 1.34 (3H, d, J = 6.7, H12). C NMR (CDCl , 125 MHz): δC 143.0 (s, C4),
136.1 (d, C9), 136.0 (s, C4), 130.0 (d, C3), 125.8 (d, C5), 123.8 (d, C10), 41.8 (d, C8), 21.5 (q,
C12), 17.9 (q, C11).
Photooxidation of azide 1a. Azide 1a (7.3 mg, 0.041 mmol) was dissolved in 60 mL of
acetonirile and placed in a thermostatically controlled (293 K) reactor. To saturate the solution
2
with oxygen, O was bubbled through it for 5 min. The resulting solution was further purged
with oxygen and irradiated by means of a xenon lamp through SSꢀ5 filter (360–500 nm) until the
starting material had disappeared. The reaction mixture was concentrated to about 0.5 mL and
separated chromatographically. (2Z,4E)ꢀ2ꢀmethoxyꢀ6ꢀoxoheptaꢀ2,4ꢀdienenitrile oxide (7а) was
1
isolated as a yellow oil in an amount of 7.0 mg (93%): H NMR (CD CN, 500 MHz):
δH 7.90
3
(1H, d, J = 11.4 Hz, H2), 5.84 (1H, dd, J = 11.4, 1.4 Hz, H3), 5.35 (1H, d, J = 1.4 Hz, H5), 3.79
1
3
(
3H, s, H8), 3.66 (3H, s, H7). C NMR (CD3CN, 125 MHz):
δC 167.6 (s, C6), 164.5 (s, C4),
136.2 (d, C2), 103.9 (d, C3), 96.7 (d, C5), 57.1 (q, C7), 52.0 (q, C8), 37.3 (s, C1). APCIꢀMS, m/z
+
ꢀ
+
+
(
relative intensity): [M+H] 184 (100), [MꢀH] 182 (100), [M+H+MeCN] 225 (23), [2M+H]
+
367 (13). HRꢀMS (MALDI) calcd for C
8
H
10NO
4
[M+H] : 184.0604. Found: 184.0631. IR
ꢀ1
4
ꢀ1
ꢀ1
spectrum (KBr), ν/ cm : 2297 (C≡N). UVꢀVis (MeCN): λmax = 295 nm (ε = 1.0 × 10 M cm ).
Photooxidation of the azide 1b (18.6 mg, 0.092 mmol in 180 mL acetonitrile) was carried out
in a manner similar to 1a except that its solution was irradiated with light of wavelength 270–
26
ACS Paragon Plus Environment

Page 27 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
3
80 nm (UFSꢀ2 filter). After the disappearance of the starting azide the reaction mixture was
divided in half. The first part was concentrated at once and separated chromatographically to
obtain the intermediate products 11 and 12 along with stable products 10, 14, and 15. The second
portion was stood in the dark for 72 hours and then concentrated. Chromatography of this
mixture yielded the products 10, and 13 – 15. Compounds 11, 13 – 15 are unstable under
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
MALDI conditions.
1
{
5ꢀ[(2E)ꢀ1ꢀMethylbutꢀ2ꢀenꢀ1ꢀyl]ꢀ2ꢀnitrosophenyl} methanol (11, 2.1 mg): H NMR (CD
3
CN,
500 MHz):
δH 7.77 (1H, d, J = 1.8 Hz, H3), 7.22 (1H, dd, J = 8.3, 1.8 Hz, H5), 6.61 (1H, d, J =
8.3 Hz, H6), 5.69 (2H, s, H7), 5.64 (1H, ddq, J = 15.2, 6.8, 1.1 Hz, H9), 5.58 (1H, dqd, J = 15.2,
6.1, 1.1 Hz, H10), 3.58 (1H, m, H8), 1.67 (3H, dd, J = 6.1, 1.1 Hz, H11), 1.35 (3H, d, J = 6.8 Hz,
13
3
H12). C NMR (CD CN, 125 MHz): δC 164.0 (s, C2), 157.5 (s, C4), 144.2 (s, C2), 135.6 (d,
C9), 128.1 (d, C3), 126.8 (d, C5), 126.0 (d, C10), 111.1 (d, C6), 60.8 (t, C7), 43.8 (d, C8), 21.4
+
(q, C12), 18.2 (q, C11). APCIꢀMS, m/z (relative intensity): [M+H] 206 (25).
5ꢀ[(2E)ꢀ1ꢀMethylbutꢀ2ꢀenꢀ1ꢀyl]ꢀ2ꢀnitrosobenzaldehyde (12) was tentatively identified by
+
molecular mass on the basis of APCI mass spectra. APCIꢀMS, m/z (relative intensity): [M+H]
ꢀ
204 (9), (MꢀH) 202 (100).
1
(3,4,7ꢀTrimethylꢀ3a,4ꢀdihydroꢀ2,1ꢀbenzisoxazolꢀ5(3H)ꢀylidene)ethanal (10, 3.8 mg): 10a – H
NMR (CD
3 H
CN, 500 MHz): δ 10.03 (1H, d, J = 7.8 Hz, H12), 6.34 (1H, s, H6), 5.87 (1H, d, J =
7.8 Hz, H11), 4.53 (1H, dq, J = 12.7, 6.2 Hz, H3), 3.91 (1H, m, J = 6.7 Hz, H4), 3.16 (1H, dd,
J = 12.7, 6.7 Hz, H3a), 2.07 (3H, s, H10), 1.45 (3H, d, J = 6.2 Hz, H8), 1.10 (3H, d, J = 6.7 Hz,
13
H9). C NMR (CD
3
CN, 125 MHz):
δ
C 191.8 (d, C12), 160.2 (s, C5), 158.0 (s, C7a), 134.9 (s,
C7), 133.4 (d, C6), 127.7 (d, C11), 80.4 (d, C3), 55.0 (d, C3a), 30.6 (d, C4), 19.0 (q, C8), 17.3
1
(q, C10), 16.7 (q, C9). 10b – H NMR (CD CN, 500 MHz): δH 10.13 (1H, d, J = 7.9 Hz, H12),
3
7.27 (1H, s, H6), 5.83 (1H, d, J = 7.9 Hz, H11), 4.49 (1H, dq, J = 12.8, 6.2 Hz, H3), 3.18 (1H,
dd, J = 12.8, 6.8 Hz, H3a), 2.90 (1H, m, H4), 2.11 (3H, s, H10), 1.44 (3H, d, J = 6.2 Hz, H8),
1
3
1.04 (3H, d, J = 6.7 Hz, H9). C NMR (CD
3 C
CN, 125 MHz): δ 191.8 (d, C12), 159.4 (s, C5),
158.0 (s, C7a), 134.5 (s, C7), 126.9 (d, C11), 126.2 (d, C6), 80.5 (d, C3), 55.1 (d, C3a), 38.2 (d,
+
C4), 18.9 (q, C8), 17.5 (q, C10), 16.4 (q, C9). APCIꢀMS, m/z (relative intensity): [M+H] 206
ꢀ
+
(
93), [MꢀH] 204 (67). HRꢀMS (MALDI) calcd for C H NO [M+H] : 206.1176. Found:
12 16 2
4
ꢀ1
ꢀ1
2
06.1050. UVꢀVis (MeCN): λmax = 315 nm (ε = 1.0 × 10 M cm ).
Z)ꢀ1,2ꢀBis(2ꢀformylꢀ4ꢀ((2E)ꢀ1ꢀmethylbutꢀ2ꢀenꢀ1ꢀyl)phenyl)diazene 1ꢀoxide (13, 2.0 mg): H
NMR (CD CN, 500 MHz): H 10.33 (1H, s, H7'), 10.32 (1H, s, H7), 8.05 (1H, d, J = 8.2 Hz,
H6), 7.98 (1H, d, J = 8.2 Hz, H6'), 7.84 (1H, d, J = 2.1 Hz, H3'), 7.75 (1H, d, J3ꢀ5 = 2.0 Hz, H3),
1
(
3
δ
7.67 (1H, dd, J = 8.2, 2.1 Hz, H5), 7.66 (1H, dd, J = 8.2, 2.1 Hz, H5'), 5.67 (2H, m, H9 and H9'),
27
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
.59 (1H, dqd, J = 15.2, 6.0, 1.1 Hz, H10), 5.56 (1H, dqd, J = 15.2, 6.0, 1.1 Hz, H10'), 3.65 (1H,
m, H8), 3.60 (1H, m, H8'), 1.68 (6H, both dd, J = 6.0, 1.1 Hz, H11 and H11'), 1.38 (3H, d, J =
13
3
6.7 Hz, H12), 1.37 (3H, d, J = 6.7 Hz, H12'). C NMR (CD CN, 125 MHz): δC 191.9 (d, C7'),
190.7 (d, C7), 152.2 (s, C4), 149.1 (s, C4'), 148.4 (s, C1), 143.6 (s, C1'), 136.2 (d, C9'), 135.8 (d,
C9), 134.3 (d, C5'), 133.6 (d, C5), 131.7 (s, C2), 131.1 (s, C2'), 129.1 (d, C3'), 128.9 (d, C3),
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
126.3 (d, C10), 125.8 (d, C10'), 125.3 (d, C6), 123.6 (d, C6'), 42.87 (d, C8'), 42.85 (d, C8), 21.59
(q, C12), 21.58 (q, C12'), 18.13 (q, C11), 18.12 (q, C11'). APCIꢀMS, m/z (relative intensity):
+
ꢀ
4
ꢀ1
ꢀ1
[
M+H] 391 (100), [MꢀH] 389 (47). UVꢀVis (MeCN): λmax = 330 nm (ε = 1.0 × 10 M cm ).
1
2
ꢀMethylꢀ4ꢀ[(2E)ꢀ1ꢀmethylbutꢀ2ꢀenꢀ1ꢀyl]ꢀ1ꢀnitrobenzene (14, 4.9 mg): H NMR (CD
3
CN, 500
MHz):
Hz, H5), 5.61 (1H, ddq, J = 15.2, 6.7, 1.1 Hz, H9), 5.54 (1H, dqd, J = 15.2, 6.0, 1.1 Hz, H10),
.50 (1H, m, H8), 2.55 (3H, s, H7), 1.66 (3H, dd, J = 6.0, 1.1 Hz, H11), 1.32 (3H, d, J = 6.7 Hz,
δH 7.91 (1H, d, J = 8.5 Hz, H6), 7.28 (1H, d, J = 2.0 Hz, H3), 7.24 (1H, dd, J = 8.5, 2.0
3
13
H12). C NMR (CD CN, 125 MHz):
δC 153.7 (s, C4), 148.4 (s, C1), 135.9 (d, C9), 134.8 (s,
3
C2), 132.5 (d, C3), 126.7 (d, C5), 125.9 (d, C6), 125.8 (d, C10), 43.0 (d, C8), 21.6 (q, C12), 20.7
+
ꢀ
(q, C7), 18.1 (q, C11). APCIꢀMS, m/z (relative intensity): [M+H] 206 (78), [MꢀH] 204 (67).
3
ꢀ1
ꢀ1
UVꢀVis (MeCN): λmax = 265 nm (ε = 9.3 × 10 M cm ).
ꢀMethylꢀ4ꢀ[(2E)ꢀ1ꢀmethylbutꢀ2ꢀenꢀ1ꢀyl]ꢀ1ꢀnitrosobenzene (15, 2.2 mg): H NMR (CD
00 MHz): 7.47 (1H, d, J = 2.0 Hz, H3), 7.07 (1H, dd, J = 8.5 Hz, H5), 6.27 (1H, d, J = 8.5
1
2
3
CN,
5
δ
H
Hz, H6), 5.62 (1H, ddq, J = 15.2, 6.7, 1.1 Hz, H9), 5.55 (1H, dqd, J = 15.2, 6.0, 1.1 Hz, H10),
3.49 (1H, m, H8), 3.28 (3H, s, H7), 1.66 (3H, dd, J = 6.0, 1.1 Hz, H11), 1.33 (3H, d, J = 6.7 Hz,
13
H12). C NMR (CD
3
CN, 125 MHz):
C
δ 166.1 (s, C1), 157.4 (s, C4), 143.9 (s, C2), 135.7 (d,
C9), 132.3 (d, C3), 126.0 (d, C10), 125.7 (d, C5), 108.5 (d, C6), 43.5 (d, C8), 21.4 (q, C12), 18.1
+
ꢀ
(q, C11), 17.5 (q, C7). APCIꢀMS, m/z (relative intensity): [M+H] 190 (100), [MꢀH] 188 (40),
+
4
ꢀ1
ꢀ1
[
M+H+MeCN] 231 (38). UVꢀVis (MeCN): λmax = 320 nm (ε = 1.2 × 10 M cm ).
Methods of Calculation. The wave function of nitroso oxides has a prominent multiꢀ
configurational character; therefore, an intensive account is required for both static and dynamic
29
electron correlation. Highꢀlevel calculations are unacceptable for aromatic nitroso oxides
because of resource constraints; therefore, the use of methods of the density functional theory
(DFT), namely M06ꢀL, mPWPW91, OLYP, and HCTH, which adequately describe the structure,
energetic and spectral properties of aromatic nitroso oxides, seemed to be a good alternative to
21
the sophisticated timeꢀconsuming methods.
Quantum chemical calculations were performed on the cluster supercomputer of Ufa Institute
32
of Chemistry of the RAS using Gaussian 09, Revision C01 program package. Visualization of
33
the results was performed in ChemCraft program. All geometric and energy values given in the
28
ACS Paragon Plus Environment

Page 29 of 31
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
text (unless otherwise specified) were obtained using the M06ꢀL and mPWPW91 functionals in
conjunction with the Pople basis set of triple valence splitting – 6ꢀ311+G(d,p). It was found that
the wave functions of ground states and products of intramolecular reactions of nitroso oxides
are stable with respect to breaking symmetry of α and β spin electronic systems, whereas
transition states show some spin polarization. The calculations were carried out in an
approximation of isolated molecules. The polarizable continuum model (IEFPCM) was used to
account a solvent effect. The full optimization of the structures of all isomeric forms, transition
states of conformational transitions and irreversible transformations of aromatic nitroso oxides
under study has been carried out using the aboveꢀmentioned methods. The type of stationary
points was determined by the number of negative elements in the diagonalized Hessian matrix.
The correspondence of the transition states to the reagents and products was confirmed by
scanning the intrinsic reaction coordinate. The relative energetic values (energies of transition
states of reactions and conformational transformations, energies of conformers and heat effects
of reactions) were calculated as the difference between the absolute enthalpies of the final (or
transition) and initial states of a transformation under study. The absolute enthalpies were
calculated as the sum of total energy, zero point vibration energy, and thermal correction for
enthalpy change from zero to 298 K. The last values were obtained from the calculations of
frequencies using the known equations of statistical thermodynamics.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
SUPPORTING INFORMATION
1
13
Copies of H NMR and C NMR, IR spectra and APCI mass spectra; cartesian coordinates of
the stationary points of nitroso oxides 2a and 2b; reaction enthalpies ꢁH°, and activation
‡
enthalpies ꢁH for the intramolecular reactions of 2a and 2b; kinetic modeling data of the 2a and
2b conformers decay. This material is available free of charge via the Internet at
http://pubs.acs.org.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
The theoretical investigations were performed and NMR and APCI mass spectra were recorded
on equipment installed in the Center for Collective use “Khimiya” (Chemistry) at Ufa Institute of
chemistry of the Russian Academy of Sciences.
29
ACS Paragon Plus Environment