The Cyclization of Allenyl-Substituted Hydroxylamines to 1,2-Oxazines: an Experimental and Computational Study

Article abstract of DOI:10.1002/ejoc.202001147

To gain a deeper understanding of the formation of the synthetically important 3,6-dihydro-2H-1,2-oxazines, the 6-endo-trig cyclization of allenyl-substituted hydroxylamines was experimentally investigated in detail employing a model compound. The solvent effect was moderate with respect to the rate, but crucial to suppress side-product formation. Surprisingly, acids or bases had no big influence on the cyclization rate. With O-deuterated allenyl hydroxylamine a high primary isotope effect was found, indicating that the proton transfer is crucial in the rate-determining step. DFT calculations evidence that the allenyl-substituted hydroxylamine is converted into an energetically similar zwitterionic intermediate with an allyl cation subunit. It cyclizes to the 1,2-oxazine as the most stable species. Alternative pathways starting from the zwitterion were computationally investigated. Interestingly, it can also undergo a fragmentation to give a pentadiene derivative and a nitroso compound. The hetero Diels–Alder reaction of these components may also deliver the 1,2-oxazine. To evaluate an alternative mechanistic scenario, calculations of the protonated allenyl-substituted hydroxylamine were also performed.

Full text of DOI:10.1002/ejoc.202001147

Full Paper
doi.org/10.1002/ejoc.202001147
EurJOC
European Journal of Organic Chemistry
Allene cyclization
The Cyclization of Allenyl-Substituted Hydroxylamines to
1,2-Oxazines: an Experimental and Computational Study
Greta Utecht-Jarzyńska,[a] Marcin Jasiński,*[a] Ernst-Ulrich Würthwein,*[b] and
Hans-Ulrich Reissig*[c]
Dedicated to Professor Grzegorz Mlostoń on the occasion of his 70th birthday
Abstract: To gain a deeper understanding of the formation of that the allenyl-substituted hydroxylamine is converted into an
the synthetically important 3,6-dihydro-2H-1,2-oxazines, the 6- energetically similar zwitterionic intermediate with an allyl cat-
endo-trig cyclization of allenyl-substituted hydroxylamines was ion subunit. It cyclizes to the 1,2-oxazine as the most stable
experimentally investigated in detail employing a model com- species. Alternative pathways starting from the zwitterion were
pound. The solvent effect was moderate with respect to the computationally investigated. Interestingly, it can also undergo
rate, but crucial to suppress side-product formation. Surpris- a fragmentation to give a pentadiene derivative and a nitroso
ingly, acids or bases had no big influence on the cyclization compound. The hetero Diels–Alder reaction of these compo-
rate. With O-deuterated allenyl hydroxylamine a high primary nents may also deliver the 1,2-oxazine. To evaluate an alterna-
isotope effect was found, indicating that the proton transfer is tive mechanistic scenario, calculations of the protonated allenyl-
crucial in the rate-determining step. DFT calculations evidence substituted hydroxylamine were also performed.
Introduction
the alkoxyallene-nitrone chemistry for the preparation of natu-
ral products and their analogs,[1h,4] the cyclization step leading
from A to the key 1,2-oxazine intermediates B was not fully
understood and only speculative mechanistic proposals were
discussed.[1g] The rate of the cyclization depends on the substi-
tution pattern of the hydroxylamine moiety; sterically hindered
derivatives react considerably slower. The seemingly uncata-
lyzed ring closure of these hydroxylamine derivatives has to be
compared with the related cyclizations of hydroxyl- or amino-
substituted alkoxyallenes that require strongly basic conditions
for their cyclizations to dihydrofuran[5] or dihydropyrrole deriva-
tives,[6] respectively. As a favored alternative method, mild Lewis
acids, such as gold(I) complexes, could also induce these cycli-
zations to the five-membered heterocycles.[7] On the other
hand, with related allenyl-substituted thiols a spontaneous for-
mation of vinyl thiiranes was observed.[8]
The addition of lithiated alkoxyallenes[1] to nitrones[2] provides
after aqueous work-up allenyl-substituted hydroxylamines A
that undergo a spontaneous 6-endo-trig cyclization to syntheti-
cally very useful 3,6-dihydro-2H-1,2-oxazine derivatives
B
thereby constituting an overall [3+3] annulation process
(Scheme 1).[3] The use of chiral nitrones afforded the corre-
sponding 1,2-oxazine derivatives generally with very high dia-
stereoselectivity. Despite of numerous synthetic applications of
[a] Dr. G. Utecht-Jarzyńska, Dr. M. Jasiński
Faculty of Chemistry, University of Łódź,
Tamka 12, 91403 Łódź, Poland
E-mail: marcin.jasinski@chemia.uni.lodz.pl
[b] Prof. E.-U. Würthwein
Organisch-Chemisches Institut and Center for Multiscale Theory and
Computation (CMTC), Westfälische Wilhelms-Universität Münster,
Corrensstrasse 40, 48149 Münster, Germany
E-mail: wurthwe@uni-muenster.de
[c] Prof. H.-U. Reissig
Institut für Chemie und Biochemie, Freie Universität Berlin,
Takustrasse 3, 14195 Berlin, Germany
E-mail: hreissig@chemie.fu-berlin.de
http://www.bcp.fu-berlin.de/en/chemie/chemie/forschung/OrgChem/reissig/
index.html
Scheme 1. Spontaneous 6-endo-trig cyclization of allenyl-substituted
hydroxylamines A into 3,6-dihydro-2H-1,2-oxazines B.
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under https://doi.org/10.1002/ejoc.202001147.
© 2020 The Authors. European Journal of Organic Chemistry published by
Wiley-VCH GmbH · This is an open access article under the terms of the
Creative Commons Attribution-NonCommercial-NoDerivs License, which
permits use, distribution and reproduction in any medium, provided the
original work is properly cited, the use is non-commercial and no modifica-
tions or adaptations are made.
In Scheme 2 we depict two probable cyclization pathways
discussed in the past.[9] In analogy to the Lewis acid-catalyzed
processes mentioned above, a proton catalysis via intermedi-
ates C and D could provide the product B. The required proton
Eur. J. Org. Chem. 2020, 6740–6748
6740
© 2020 The Authors. European Journal of Organic Chemistry
published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejoc.202001147
EurJOC
European Journal of Organic Chemistry
may be delivered by the solvent (and its impurities, the drying
This study describes the influence of solvents and various
agent applied during work-up etc.), or by the hydroxylamine additives on the rate of the 6-endo-trig cyclization of A to B
itself, serving as a mild acid (estimated pKa value ca. 14).[10] An employing a suitable model allenyl hydroxylamine and its O-
alternative mechanism without involvement of a catalyst fol- deuterated derivative. We also examined the effect of the reac-
lows a suggestion of Dulcère et al.,[11] reported for the related tion conditions on the formation of side products such as G
cyclizations of allenyl-substituted hydroxylamines lacking alk- and I. These experimental studies are supplemented by com-
oxy groups. Two concerted steps are proposed, first a reverse prehensive DFT calculations that provide information about the
Cope elimination (also known as Cope-House cyclization) to the energy profiles of the different mechanistic cyclization scenar-
vinyl-substituted aziridinium N-oxide E that, after a rotation of ios.
the vinyl group to conformer F, undergoes a sigmatropic rear-
rangement to the 1,2-oxazine derivative B (Meisenheimer 3,2-
rearrangement). This elegant mechanism has the charm that no
Results and Discussion
catalysts are required; it requires the hydroxylamine functional
group and cannot occur with the above mentioned hydroxy- or
Experimental Investigations
amino-substituted allenes.
As a model compound for our experimental studies, we se-
lected the easily available allenyl hydroxylamine 3.[3a,3c] Apply-
ing the standard method, generation of lithiated methoxy-
allene from 1 and n-butyllithium in tetrahydrofuran and sub-
sequent addition to enantiopure
D-glyceraldehyde-derived
nitrone 2 at low temperature furnished after aqueous work-up
the corresponding adduct 3 (Scheme 3). The THF/Et2O solution
of crude allenyl hydroxylamine 3 could be stored for weeks in
a container cooled by dry ice, and aliquots of this solution were
used for the subsequent kinetic studies. Typical 1H-NMR signals
of 3 refer to the allenic protons at ca 5.1–5.6 ppm (AB system
in C6D6 δ 5.15, 5.24, JAB = 8 Hz). The depicted syn-configuration
of product 3 is deduced from its cyclization to 1,2-oxazine 4
(see below) whose 3S,4′S-configuration has unequivocally been
proven earlier; subsequent reactions of 4 allowed X-ray analyses
of products or comparison with known enantiopure com-
pounds. The O-deuterated derivative 3-d1 was analogously pre-
pared when D2O was used for quenching the reaction. It was
important that the washing of the organic extract of this mix-
ture was also performed with deuterated water to avoid the
fast H/D exchange at the hydroxylamine oxygen. Under these
prerequisites the degree of O-deuteration is > 95 % as shown
by the cyclization of 3-d1 to 4-d1 (see below), where this high
level of D-incorporation could be proven by the 13C-NMR signal
of C-5 (92.5 ppm, JC-D = 24.1 Hz).
Scheme 2. Two pathways for the 6-endo-trig cyclization of A → B: proton-
catalyzed route via intermediate C (left hand side) and uncatalyzed route
with two concerted steps via E and F (right hand side); side reactions: retro
ene reaction of A leading to G and H and its 5-endo-trig cyclization to pyrro-
line N-oxide I.
The 6-endo-trig cyclization of A to B was clearly the major
reaction pathway in all studied cases, however, in many exam-
ples we also found fragmentation products of type G with a 3-
alkoxy-1,3-butadiene moiety (ca. 90:10 E/Z mixture, only one
isomer shown).[3c] These compounds probably arise from a retro
ene reaction and in a few cases products derived from the sec-
ond fragment, the nitroso compound H, could be identified (the
corresponding oximes if R3 bears an α-CH unit). A second side-
Scheme 3. Lithiation of methoxyallene 1 and subsequent reaction with
D
-glyceraldehyde-derived nitrone 2 to allenyl hydroxylamines 3 and 3-d1.
Aliquots of the THF solution of 3 were rapidly evaporated at
reaction occurred essentially only when cyclic nitrones were room temperature and the resulting crude allenyl hydroxyl-
employed as precursor. In their reaction with lithiated alkoxyal- amine 3 was then dissolved in the corresponding deuterated
lenes not only the expected bicyclic 1,2-oxazines B were solvent (Scheme 4). Its cyclization at 21 °C was immediately
formed, but also the isomeric pyrroline N-oxides I, formed by a followed by 1H-NMR spectroscopy (200 MHz) observing the de-
5-endo-trig cyclization.[3b,12]
creasing signals of 3 at 5.1–5.6 ppm (=CH2) and the increasing
Eur. J. Org. Chem. 2020, 6740–6748
www.eurjoc.org
6741
© 2020 The Authors. European Journal of Organic Chemistry
published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejoc.202001147
EurJOC
European Journal of Organic Chemistry
Table 1. Conversion of hydroxylamine derivative 3 into 1,2-oxazine 4 and 1,3-diene 5 according to Scheme 4 (cyclization of 3-d1 into 4-d1 in bold, entry 7) at
21 °C as determined by 1H-NMR spectroscopy at 200 MHz.
Entry
Solvent
ET(30)[a] (kcal/mol)
Ratio of 4/5
k
exp × 10–5 (s–1 [b,c]
)
t
1/2(exp) [min][b]
k
1(exp) × 10–5 (s–1
)
k )
2(exp) × 10–5 (s–1
1
2
3
4
5
6
[D12]cyclohexane
[D8]toluene
[D6]benzene
[D10]Et2O
[D8]THF
31.2
33.9
34.5
34.6
37.4
39.1
39.1
40.2
41.1
42.2
45.0
46.0
82:18
91:9
95:5
76:24
79:21
100:0
100:0[e]
91:9
40
83
102
14
16
344
51
24
348
81
29
14
11
84
71
3
23
48
3
33
75
97
11
7
8
5
3
3
–
–
2
–
3
1
2
13
CDCl3
CDCl3
344
51
22
348
78
7[d]
8
9
10
11
12
[D5]pyridine
CD2Cl2
[D6]acetone
[D6]DMSO
CD3CN
100:0
97:3
96:4
14
52
6
22
199
21
197
99:1
[a] ET(30) values for non-deuterated solvents (Ref.[13]). [b] The k values should be considered as a consumption rate of 3 or 3-d1. [c] Coefficient of determination
R2 > 0.998. [d] 3-d1 was employed as starting material. [e] Under these conditions the by-product 5 was not observed.
signals of 1,2-oxazine 4 at 4.8–5.0 ppm (5-H) and/or 3.1– two chlorinated solvents CDCl3 and CD2Cl2. This moderate influ-
3.6 ppm (3-H) and of 1,3-diene 5 at 5.05–5.25, 5.6–5.9, and 6.4– ence of the solvent polarity is compatible with a mechanism of
6.8 ppm. The ratio of the cyclization product 4 and the frag-
mentation product 5 considerably depends on the solvent em-
ployed (Table 1).
the retro ene reaction with low charge separation in the transi-
tion state.
The solvent dependence of the cyclization of 3 to the major
product 4 does not correlate with the ET(30) values and there-
fore its interpretation is not simple. The slowest rates were de-
termined for diethyl ether and tetrahydrofuran and as a conse-
quence the fragmentation to 5 occurs to considerable extent in
these solvents. But relatively slow cyclizations were also ob-
served in cyclohexane, pyridine and DMSO (all solvents are per-
deuterated). On the other hand, acetonitrile and the two chlor-
inated solvents CDCl3 and CD2Cl2 provided the highest rate
constants k1. For the cyclization in CDCl3 a high primary kinetic
isotope effect of ca. 6.7 was determined (entries 6 and 7),
demonstrating that the migration of the proton/deuteron is in-
volved in the rate-determining step. We also attempted to per-
form a similar experiment of 3-d1 in [D5]pyridine but observed
fast partial D/H exchange of the hydroxylamine moiety (proba-
bly due to traces of water and catalyzed by the basic solvent).
Hence the results of this experiment are questionable, however,
the consumption of 3-d1 was slower by a factor of ca. 6 and
only 4-d1 was formed as deuterated compound whereas no
deuterium was incorporated into 1,3-diene 5 formed in ca.
10 %.
Scheme 4. Cyclization of allenyl hydroxylamine 3 in different solvents to 1,2-
oxazine 4 and 1,3-diene derivative 5 (for details also see Table 1 and Table 2).
In the ether solvents [D8]THF and [D10]diethyl ether (entries
4 and 5) the results of preparative experiments were confirmed,
where 5–20 % of the side product 5 were isolated.[3c] Relatively
high amounts of this compound were also observed in other
solvents of low polarity (entries 1–3) according to the ET(30)
scale of Reichardt.[13] On the other hand, 1,3-diene 5 was
formed in a much lower extent when chlorinated solvents (en-
tries 6, 7 and 9) or aprotic polar solvents (entries 10–12), for
instance [D3]acetonitrile, were employed. Only, [D5]pyridine was
exceptional in this respect, since it also promoted the formation
of ca. 9 % of 5 (entry 8). In none of the experiments the forma-
tion of the pyrroline N-oxide 6 could be detected within the
limit of NMR accuracy.
The 1H-NMR experiments also allowed a determination of
the experimental rate constant kexp and half-life time t1/2 in the
different solvents (columns 5 and 6 of Table 1). The consump-
tion of allenyl hydroxylamine 3 is consist with a first-order rate
law (for details see Supporting Information). Within the limits
of NMR accuracy (2 %, only [D5]pyridine as solvent caused
broadened signals resulting in lower accuracy), the ratio of the
two resulting products 4 and 5 was constant during the reac-
tions and therefore the overall rate constant kexp could be di-
vided into the two rate constants, k1 for formation of 1,2-
oxazine 4 and k2 for that of 1,3-diene 5 (columns 7 and 8). It is
remarkable that the fragmentation of 3 to 5 is slightly deceler-
ated in more polar solvents and completely suppressed in the
The k1 values of Table 1 could suggest that the cyclization
of allenyl hydroxylamine 3 is faster in solvents were traces of
acids might be present (DCl in CD2Cl2 and CDCl3 or DCN in
[D3]acetonitrile) and slower in basic solvents such as ethers, pyr-
idine or DMSO. We therefore examined the influence of addi-
tives on the selectivity and rate of 3 (Table 2). The experiments
were performed in CD2Cl2 as solvent since only the 1,2-oxazine
derivative 4 is formed and no fragmentation was observed
without additive (entry 1). The addition of an excess of acid,
even the strong toluenesulfonic acid, (entries 2 and 3) or bases
(entries 4–7) has essentially no influence on the rate of the
cyclization. With the two acids even a slight retardation of the
cyclization was observed; interestingly, the strong p-toluene-
sulfonic acid did not react with the potentially acid sensitive
enol ether moieties of 3 and 4 in the relatively short reaction
Eur. J. Org. Chem. 2020, 6740–6748
www.eurjoc.org
6742
© 2020 The Authors. European Journal of Organic Chemistry
published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejoc.202001147
EurJOC
European Journal of Organic Chemistry
times. The fragmentation product 5 was not identified in these quantum chemical DFT calculations for model compound 12
experiments, however, it cannot be excluded that small and its various subsequent products were performed. On the
amounts are decomposed by the acidic additives.
TPSSTPSS/def2tzvp[14,15] + GD3BJ[16] level of theory first optimi-
zations for the gas phase were performed.[17,18] PCM calcula-
tions[19] using a solvent sphere of diethyl ether, one of the sol-
vents used in the experiments, were done in order to estimate
the influence of this polar environment on the intermediates
and products. In the following, we discuss Gibbs free energies
[kcal/mol], for the PCM-diethyl ether model the data are given
in italics in the text and with blue color in the Schemes [kcal/
mol] (see also Supporting Information for details). Results of
PCM calculations for the more polar solvent acetonitrile did not
deviate significantly from the diethyl ether results and are
therefore not presented in the Schemes.
Table 2. Cyclization of hydroxylamine derivative 3 into 1,2-oxazine 4 in CD2Cl2
at room temperature in the presence of different additives according to
Scheme 4.
t1/2(exp)
[min]
Entry
Additive
Equivalents
k )
exp × 10–5 (s–1
1
2
3
4
5
6
7
none
–
348
240
170
300
320
340
240
3.3
4.9
6.9
3.9
3.7
3.4
4.8
pTosOH
AcOH
DABCO
DMAP
Et3N
2.7
10.3
8.4
8.6
14.5
15.7
DBU
We first calculated the stability of the species involved em-
ploying nitrone 11 as precursor of model compound 12. Com-
bination of methoxyallene 1 with 11 to furnish allenyl hydroxyl-
amine 12 is a moderately endothermic process, however, the
subsequent cyclization to the 1,2-oxazine 13 clearly demon-
strates that the overall process is thermodynamically very favor-
able (Scheme 6, equation a). The alkoxyallene subunits of 1 and
12 certainly contribute to their relatively high energy level. It
should be noted in this context that the Huisgen reaction (1,3-
dipolar cycloaddition)[20] of methoxyallene 1 with nitrone 2
proceeds very slowly at room temperature (even in the pres-
ence of Lewis acids) and affords mixtures of isoxazolidine deriv-
atives with low regio- and stereoselectivity.[21] To realize the
formation of allenyl hydroxylamine 12 requires the deprotona-
tion of 1 and therefore the thermochemistry of the involved
anionic species was also calculated (Scheme 6, equation b).
Now, the addition step of 1– to nitrone 11 is exothermic (in
agreement with the experiments), and the barrier leading to
12– is very low. Despite of the better stabilization of the nega-
tive charge in 12– this species is still not particularly stable due
to the allenyl moiety. As a consequence, the 6-endo-trig cycliza-
tion (via a low barrier) to the 1,2-oxazine anion 13– is again an
exothermic step. Experimentally this very last step is not ob-
served since the lithium counterion plays a crucial role, which
has been neglected in the calculations summarized in this
equation. Due to the worse solvation of the lithium counterion
For a proton-catalyzed cyclization a stronger influence of the
acidic and basic additives can be expected. Overall, the kinetic
results, in particular the high primary kinetic isotope effect, are
in agreement with a mechanism of the cyclization 3 → 4 (and
generally of A → B) with transfer of the hydroxylamine
hydrogen in the rate-determining step.
Finally, we checked whether the above-mentioned pyrroline
N-oxides I are formed from 1,2-oxazine derivatives B by a ring
contraction process (reverse Meisenheimer 1,2-rearrangement).
For this purpose, we examined the long-term stability of several
typically substituted 1,2-oxazine derivatives in CDCl3 under ex-
clusion of light (Scheme 5). Compounds 4 and 7–10 were stable
1
for at least four weeks, no changes could be seen in their H-
NMR spectra. These experiments were repeated in the presence
of an excess of triethylamine, but again no changes were ob-
served. We conclude that the 1,2-oxazines are stable even if
small amounts of DCl are present in a solvent like CDCl3.
Scheme 5. Examination of stability of 1,2-oxazine derivatives 4 and 7–10 in
CDCl3 at room temperature.
Scheme 6. Energetics of the formation of allenyl-substituted hydroxylamines
12 and 1,2-oxazine 13: a) starting from neutral components 1 and 11; b)
starting from allenyl anion 1– and 11 (in this and all subsequent schemes
only relative configurations are defined by the drawn formulas).
Computational Study
In order to mechanistically evaluate the many conceivable path-
ways of the highly reactive allenyl-substituted hydroxylamines,
Eur. J. Org. Chem. 2020, 6740–6748
www.eurjoc.org
6743
© 2020 The Authors. European Journal of Organic Chemistry
published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejoc.202001147
EurJOC
European Journal of Organic Chemistry
Scheme 7. Formation of 1,2-oxazine 13 via zwitterionic intermediates Z-14 and E-14 and other possible pathways/products.
of the intermediate anion 13– the cyclization of the ion pair is
calculated to be endothermic (17.4 kcal/mol) and the barrier is
higher (23.0 kcal/mol).[22]
In compounds like 12 a number of highly reactive functional
groups are combined: an allene subunit, an enol ether and a
hydroxylamine functionality with Lewis and Brønsted basic and
acidic centers as well as with nucleophilic and electrophilic
sites. With regard to the Brønsted acidity, the hydroxylamine
subunit is of most interest. Its pKa value (ca. 14)[10] is close to
that of water or to that of the dimethylammonium cation,
Figure 1. Transition state of the intramolecular proton transfer of allenyl
whereas the methoxyallene moiety may well act as a Brønsted
base. Thus, as shown in Scheme 7, the zwitterionic tautomers
14 are almost equal in relative energy compared to 12. Inter-
or intramolecular proton transfer may establish an equilibrium
of 12 and 14. The calculated values for the intramolecular H-
shifts are given in Scheme 7 and the transition state of this
process is depicted in Figure 1. Unfortunately, intermolecular H-
shifts cannot be calculated reliably. We assume that the barriers
are substantially lower than intramolecular barriers due to the
influence of the diethyl ether solvent for proton transfer reac-
tions.
hydroxylamine 12 leading to zwitterion Z-14.
under the reaction conditions, but there is no chance for experi-
mental identification. The four- and five-membered ring sys-
tems 16 and 17 might be formed exothermically via the respec-
tive 4-exo-trig and 5-endo-trig cyclizations, the barriers for their
formation (19–24 kcal/mol) are compatible with the reaction
conditions. By far lowest in energy (thermodynamic control) [ca.
–23 kcal/mol] among all considered possible products is the
experimentally observed 1,2-oxazine derivative 13. It may result
Thus, as central reactive intermediate we assume the zwitter- from 14 by a 6-endo-trig cyclization (barrier ca. 17–18 kcal/mol)
ions Z-14 and E-14 which may interconvert via a relatively low (Figure 2), but also by an alternative pathway involving a het-
barrier or via 12 by proton exchange. From these key intermedi- ero-Diels–Alder reaction (see below). It is also interesting to
ates, various possible products were calculationally generated: note, that a ring contraction 13 → 17 (reverse Meisenheimer
the three-membered aziridinium N-oxide 15, which is higher 1,2-rearrangement)[23] involves a high barrier (ca. 35 kcal/mol)
in energy by 9.3/6.9 kcal/mol, is situated on an extremely flat and is thermodynamically unfavorable. This computational re-
hyperface, with a transition state energy very close to the local sult is in full agreement with the experimental observations (see
minimum for 15. Thus, this species may play a role in equilibria Scheme 5).
Eur. J. Org. Chem. 2020, 6740–6748
www.eurjoc.org
6744
© 2020 The Authors. European Journal of Organic Chemistry
published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejoc.202001147
EurJOC
European Journal of Organic Chemistry
the system, and/or its isomer 20 (experimentally not observed),
which is accessible via a slightly higher barrier and also less
stable. This pathway to 13 was not considered so far and it
does not easily explain the observed retention of configuration
at C-3 of the 1,2-oxazine 13. Calculationally, the van der Waals
complex 18/19 was localized for the fragmentation products;
however, proton transfer reactions or subsequent cycloaddi-
tions could not be derived directly from this complex but only
from its free components 18 and 19.
In summary, starting from neutral 12 the formation of four-,
five- and six-membered heterocyclic rings is possible with not
very different barriers via zwitterions 14. However, from the
thermodynamic point of view, the formation of the 1,2-oxazine
derivative is clearly favored. Although the lowest barrier was
found for the formation of the (unstable) aziridinium N-oxide
15 no direct pathway could computationally be identified lead-
ing from this species to 1,2-oxazine 13.
Figure 2. Transition state of the 6-endo-trig cyclization of zwitterion E-14 to
1,2-oxazine derivative 13.
An unexpected feature of the zwitterions 14 is their low sta-
bility towards N-C-bond rapture leading to a van der Waals
complex of the neutral fragments 3-methoxy-1,3-pentadiene
(18) and nitrosomethane (19) and – quite likely – after tauto-
merism to its more stable isomer formaldehyde oxime (21).
Compound 5 (see above) was isolated from the reaction mix-
ture as a representative example of these C-N-rapture products.
The very low barrier and the higher stability of 18/19 compared
to 14 may allow an important contribution of these species in
the equilibrium. So far, we interpreted the fragmentation of
allenyl-substituted hydroxylamines as a concerted retro ene
reaction,[24] however, the calculations indicate that a two-step
process with a proton transfer preceding the N-C bond
cleavage is very likely.
As second scenario the involvement of protonated species
(see structure C of Scheme 2) should be considered, which are
formed either by self-protonation by the acidic hydroxylamine
group or by addition of external acids. We therefore investi-
gated the protonated compounds by DFT calculations in order
to get insight into pathways of 12 in the presence of acids. The
results are summarized in Scheme 8 with the allylic cation Z-22
as reference of energy. Quite similar as in the zwitterionic case
(see 14 in Scheme 7) the formation of a three-membered N-
hydroxy aziridinium ion 23 from Z-22 seems to be feasible in
an equilibrium situation; again experimental identification is
A recombination of 18/19 by hetero-Diels–Alder reaction unlikely due to its high relative energy and the low barrier to-
may also afford 1,2-oxazine 13,[25] the most stable species of wards the reverse reaction. Formation of the five-membered
Scheme 8. Formation of 1,2-oxazine 13 via allylic cations Z-22 and E-22 as key intermediates and alternative pathways.
Eur. J. Org. Chem. 2020, 6740–6748
www.eurjoc.org
6745
© 2020 The Authors. European Journal of Organic Chemistry
published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejoc.202001147
EurJOC
European Journal of Organic Chemistry
ring system 24 cannot be excluded which is accessible over an oxide 15 has an even lower barrier, but the product is of similar
only very small barrier (6–7 kcal/mol); its calculated energy energy as the transition state and no pathway could be identi-
(–14.1/–16.1 kcal/mol), however, allows return to the allyl cat- fied leading from 15 to 1,2-oxazines 13. Based on these calcula-
ions 22. Cyclization to form the protonated 1,2-oxazine deriva- tions the two pathways proposed in Scheme 2 are not con-
tive is possible considering the calculated barrier (ca. 20 kcal/ firmed and the mechanisms as shown in Scheme 9 with zwitter-
mol). Due to the protonated oxygen atom the primary cycliza- ion ZW as key intermediate are more likely. The cyclization to
tion product 25 is energetically disfavored (16–18 kcal/mol) but B has a moderate barrier, but the fragmentation to G and H is
a subsequent intra- or (preferably) intermolecular proton shift also very favorable. These products of a two-step retro ene reac-
may give access to the thermodynamically favored N-proto- tion can recombine to give 1,2-oxazine B by a hetero-Diels–
nated isomer 26 (–7.4/–10.2). A deprotonation of 25 or 26 to Alder reaction, a mechanism we did not consider so far. Experi-
the neutral product 13 will provide strong thermodynamic driv- ments should be designed to prove or disprove this alternative
ing force for the overall cyclization process.
pathway to 1,2-oxazine B.
As for zwitterion 14, we also studied the C-N-bond cleavage
reaction of allyl cations 22. Due to the primary formation of
the very unfavorable O-protonated nitrosomethane fragment
besides the 3-methoxy-1,3-pentadiene (18), no transition state
could be localized. After proton shift from Me-N=O-H+ the
much better tautomer 27 is formed.
To summarize the computational results of the protonated
species shown in Scheme 8, the cyclization to the observed 1,2-
oxazine derivative 13 has to pass an intermediate 25 of rela-
tively high energy though its deprotonation is energetically
very favorable. The best direct cyclization pathway leads to the
protonated pyrrolidine N-oxide 24 which is not compatible with
the experimental results. Furthermore, the observed fragmenta-
tion products should not be as easily accessible as via the zwit-
terionic intermediates 14. With exception of the five-membered
ring system 17, the relative energies (with respect to 14 or 22)
of the neutral compounds and the barriers are similar or slightly
lower than the protonated forms. Thus, zwitterions 14 and C-
protonated allenes 22 are expected to exert similar reactivity in
most cases.
Scheme 9. Formation of 1,2-oxazines B from A via zwitterion ZW.
Experimental Section
General information: reagents and solvents were purchased (Sigma-
Aldrich, Acros, Fluorochem) and used as received without further
purification. Tetrahydrofuran was dried with sodium metal in the
presence of benzophenone and distilled just before usage. Reac-
tions were carried out under argon in a flame-dried flask with addi-
tion of the components by using syringes; subsequent manipula-
tions were conducted in air. Products were purified by flash chroma-
tography on silica gel LC60A (70–200 micron, Fluorochem). Unless
stated otherwise, reported yields refer to analytically pure samples.
NMR spectra were measured with a Bruker AVIII 600 or with a Varian
Gemini 2000 BB 200 MHz instrument. Chemical shifts are reported
relative to solvent residual peaks.[26] All 13C-NMR spectra are proton-
decoupled; substitution patterns of the carbon atoms were deter-
mined by 2D NMR spectroscopy (COSY, HMQC, HMBC) and are indi-
cated as 13C NMR peak multiplicity; coupling constants J are given
in Hz. IR spectra were measured with a FTIR NEXUS spectrometer
(as KBr pellets or thin films). MS spectra were performed with a
Varian 500-MS LC Ion Trap or with a Waters SYNAPT G2-S HDMS
instrument. Melting points were measured in capillaries with a Mel-
Temp II apparatus (Aldrich) and are uncorrected. Elemental analyses
were obtained with a Vario EL III instrument.
Conclusions
The experiments performed with allenyl-substituted hydroxyl-
amine 3 and its O-deuterated analog 3-d1 show that the rate of
the formation of 1,2-oxazine derivative 4 and 4-d1, respectively,
cannot be correlated with the solvent polarity. The fragmenta-
tion product 5 is formed only in ethereal solvents in higher
extent (up to 24 %). The high primary isotope effect of ca. 6.7
clearly reveals that a proton/deuteron transfer is involved in the
rate-determining step. The cyclization rate is also not depend-
ent on the presence of acidic or basic additives, therefore, the
neutral allenyl-substituted hydroxylamines seem to be the reac-
tive key species.
The computational results confirm that for protonated spe-
cies no favorable direct pathway is accessible for the 1,2-
oxazine formation. In contrast, the calculations starting with
neutral model compound 12 offer an attractive – so far undis-
cussed – pathway to the six-membered ring system. By intra-
or (more likely) intermolecular proton transfer the zwitterionic
intermediates 14 are easily formed and show similar stability as
their precursor 12. The zwitterions 14 have various options for
further reactions, but the cyclization to 1,2-oxazine 13 via a
feasible barrier leads directly to the most stable compound of
General procedure for the synthesis of allenyl hydroxylamines
and 1,2-oxazines: The lithiated alkoxyallene was generated in situ
under inert atmosphere of dry argon by treatment of a solution of
the corresponding alkoxyallene (3.50 mmol) in freshly dried THF
(10 mL) with n-butyllithium (2.5 M in hexanes, 1.2 mL, 3.0 mmol) at
–40 °C. After 5 min, the resulting solution was cooled to –78 °C,
and a solution of the corresponding nitrone (1.0 mmol) in THF
(2.0 mL) was added dropwise. The mixture was stirred for 2–4 h,
H2O (15 mL) was added at –78 °C, and the mixture was warmed to
the system. The formation of a vinyl-substituted aziridinium N- room temperature followed by extraction with diethyl ether (3 ×
Eur. J. Org. Chem. 2020, 6740–6748
www.eurjoc.org
6746
© 2020 The Authors. European Journal of Organic Chemistry
published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejoc.202001147
EurJOC
European Journal of Organic Chemistry
10 mL). The combined organic layers were dried (Na2SO4) to pro-
was purified by column chromatography (SiO2, petroleum ether/
vide the intermediate hydroxylamines. Their cyclization was moni- EtOAc, 5:1) to afford 9 (106 mg, 44 %) as a colorless solid.
tored by TLC and after completion, the solvents were removed un-
1
M.p. 100–103 °C; H NMR (CDCl3, 600 MHz): δ = 1.23–1.39 (m, 2 H,
der reduced pressure and the crude 1,2-oxazines were purified by
5-H, 7-H), 1.67–1.80 (m, 3 H, 6-H2, 7-H), 2.25 (dbr, J ≈ 13.0 Hz, 1 H,
column chromatography.
5-H), 2.54–2.60 (m, 1 H, 8-H), 3.12 (dbr, J ≈ 11.3 Hz, 1 H, 4a-H), 3.34–
The syntheses and analytical data for (3S,4′S)-2-benzyl-3-(2′,2′-di-
methyl-1′,3′-dioxolan-4′-yl)-4-methoxy-3,6-dihydro-2H-1,2-oxazine
(4),[3c] 2-benzyl-3-[4′-(tert-butyldiphenylsiloxy)-butyl]-4-methoxy-
3,6-dihydro-2H-1,2-oxazine (8),[27] and (4aS,5R,6S)-5,6-isopropylid-
enedioxy-4-methoxy-4a,5,6,7-tetrahydro-2H-pyrrolo[1,2-b][1,2]-ox-
azine (10)[28] were fully described in the literature. The 1H- and 13C-
NMR spectra of the obtained 1,2-oxazines 4, 8 and 10 are in full
accordance to those reported.
3.38 (m, 1 H, 8-H), 4.19 (dd, J = 2.7, 14.1, 1 H, 2-H), 4.64 (dbr, J ≈
14.1 Hz, 1 H, 2-H), 4.74 (mc, 1 H, 3-H), 4.76 (s, 2 H, CH2Ph), 7.28–7.37
(m, 5 H, Ph) ppm; 13C NMR (CDCl3, 151 MHz): δ = 23.7, 25.3, 27.0 (3
t, C-7, C-6, C-5), 55.9 (t, C-8) 65.6 (d, C-4a), 66.4 (t, C-2), 69.0 (t,
CH2Ph), 92.4 (d, C-3), 127.1, 127.7, 128.4, 137.1 (3 d, s, Ph), 153.4 (s,
C-4) ppm; IR (KBr): ν = 3060–2815 (=C-H, C-H), 1665 (C=C), 1225,
˜
1205, 1175, 1095 (C-O) cm–1; EI-HRMS: calcd. for C15H19NO2 ([M]+)
245.1416; found 245.1420. Anal. calcd. for C15H19NO2 (245.1): C
73.44, H 7.81, N 5.71; found C 73.64, H 7.80, N 5.69.
Synthesis of 3-d1 and 4-d1: Following the general procedure, lithi-
ated methoxyallene (4.00 mmol) was treated with
D-glyceralde-
Preparation of allenyl hydroxylamine samples for kinetic meas-
urements: Following the general protocols for the preparation of
hydroxylamines 3 or 3-d1, the mixture was quenched with water or
D2O at –78 °C and allowed to reach 0 °C. Cold diethyl ether was
added in the case of 3 to achieve separation of the layers. The
organic layer was separated and very quickly washed with either
H2O or D2O and dried with solid Na2SO4 (1.5 g) at 4 °C. After ca.
15 min. of drying the first sample (0.6–1.0 mL for 3, 0.2–0.3 mL for
3-d1) was transferred via syringe into a flame-dried flask, and the
solvent was removed in vacuo (at room temperature) to give a thick
oil. The residue was dissolved in 0.8 mL of the corresponding deu-
hyde-derived nitrone (235 mg, 1.00 mmol) in freshly dried THF
(12 mL) at –78 °C for 2 h. The resulting mixture was quenched
with D2O (3.0 mL) and the mixture was allowed to reach room
temperature. The layers were separated and the organic layer was
washed with D2O (2 × 2 mL). Solid Na2SO4 (1.5 g) was added and
the cyclization of the hydroxylamine 3-d1 was monitored by TLC
(SiO2, petroleum ether/EtOAc, 7:1, visualization with p-anisaldehyde
stain). After 24 h the mixture was filtered, the solvent was removed
in vacuo, and crude product was purified by column chromatogra-
phy (SiO2, petroleum ether/EtOAc, 6:1) to give 1,2-oxazine 4-d1
(191 mg, 63 %, degree of deuteration at C-5 >95 %) as a colorless
oil.
1
terated solvent (optionally with an additive) and the H-NMR spec-
tra were measured. The remaining solution of the allenylhydroxyl-
amine was stored in the presence of the drying agent in a Dewar
container with dry ice until the next sample for the kinetics was
required.
1
[α]2D0 = +33.1 (c = 0.26, CHCl3). H NMR (CDCl3, 600 MHz): δ = 1.35,
1.40 (2 s, 3 H each, 2 Me), 3.31 (dbr, J ≈ 7.0 Hz, 1 H, 3-H), 3.56 (s, 3
H, OMe), 3.91 (d, J = 7.0 Hz, 2 H, 5′-H), 4.14 (s, 2 H, CH2Ph), 4.15 (d,
J = 14.6 Hz, 1 H, 6-H), 4.42 (dbr, J ≈ 14.6 Hz, 1 H, 6-H), 4.55 (q, J =
7.0 H, 1 H, 4′-H), 7.23–7.26, 7.29–7.32, 7.41–7.43 (3 m, 1 H, 2 H, 2 H,
Ph) ppm; 13C NMR (CDCl3, 151 MHz): δ = 26.1, 26.6 (2 q, 2 Me), 54.1
(q, OMe), 58.3 (t, CH2Ph), 63.4 (d, C-3), 64.3 (t, C-6), 66.8 (t, C-5′),
75.0 (d, C-4′), 92.5 (t, JC-D = 24.1 Hz, C-5), 108.6 (s, C-2′), 127.0, 128.2,
1H NMR of 3: ([D8]THF, 200 MHz): δ = 1.27, 1.31 (2 s, 3 H each, 2
Me), 3.34 (dbr, J = 5.7 Hz, 1 H), 3.37 (s, 3 H, OMe), 3.75 (dd, J = 7.0,
8.3 Hz, 1 H), 3.88, 3.91 (AB system, J = 8.6 Hz, CH2Ph), 3.96 (dd, J =
6.4, 8.3 Hz, 1 H), 4.46 (dt, J ≈ 6.7, 8.3 Hz, 1 H), 5.47 (sbr, 2 H, =CH2),
7.13–7.26, 7.30–7.37 (2m, 3 H, 2 H, Ph) ppm; 1H NMR of 3-d1: (CDCl3,
200 MHz): δ = 1.37 (sbr, 6 H, 2 Me), 3.41 (dbr, J = 6.8 Hz, 1 H), 3.45
(s, 3 H, OMe), 3.75–3.81 (m, 1 H), 3.92–4.15 (m, 3 H), 4.48–4.60 (m,
1H), 5.53 (sbr, 2 H, =CH2), 7.23–7.42 (m, 5 H, Ph) ppm.
128.7, 137.9 (3 d, s, Ph), 151.4 (s, C-4) ppm; IR (film): ν = 3055–2855
˜
(=C-H, C-H), 1625 (C=C), 1470, 1255, 1145, 1100, 1035, 835 cm–1
;
ESI-MS (m/z): 307.2 (100, [M + H]+). Anal. calcd. for C17H22DNO4
(306.2): C 66.64, H/D 7.90, N 4.57; found C 66.67, H/D 7.72, N 4.59.
Quantum chemical calculations: Quantum chemical calculations
at the DFT level TPSSTPSS/def2-TZVP+GD3BJ dispersion correc-
tion[14–16] were performed using the Gaussian 09, Revision D.01[17]
and the Gaussian 16, Revision B.01[18], packages of programs for
the gas phase as well as for diethyl ether and acetonitrile as solvent
using the PCM solvent sphere.[19] The transition state localizations
were started with reaction path calculations by stepwise, independ-
ent elongation of relevant bonds. Then transition state searches or
QST2 calculations on the basis of the obtained 3D-hyperfaces fol-
lowed. We cannot exclude that due to the steric complexity of the
reacting systems further transition state conformations and config-
urations exist. In many cases, for transition states IRC-calculations
were subsequently performed in order to characterize them unam-
biguously, followed be complete geometry optimizations leading
to the respective minima.
4-Methoxy-2-methyl-3-(p-tolyl)-3,6-dihydro-2H-1,2-oxazine (7):
Following the general procedure, N-methyl-C-(p-tolyl)nitrone
(149 mg, 1.00 mmol) was treated with lithiated methoxyallene (2 h).
After the cyclization of intermediate allenyl hydroxylamine was
complete (16 h), the crude mixture was purified by column chroma-
tography (SiO2, hexanes/EtOAc, 4:1) to give 7 (160 mg, 73 %) as a
pale yellow solid.
1
M.p. 29–31 °C. H NMR (CDCl3, 600 MHz): δ = 2.34 (s, 3 H, Tol), 2.43
(s, 3 H, NMe), 3.48 (s, 3 H, OMe), 4.05 (s, 3-H), 4.42, 4.58 (2 sbr, 1 H
each, 6-H2), 4.88 (pseudo-t, J ≈ 3.2 Hz, 1 H, 5-H), 7.14, 7.21 (2 dbr,
J ≈ 8.2 Hz, 2 H each, Tol) ppm; 13C NMR (CDCl3, 151 MHz): δ = 21.2,
42.9, 54.7 (3 q, Tol, NMe, OMe), 66.5 (t, C-6), 70.6 (d, C-3), 92.4 (d, C-
5), 128.9, 129.5, 133.9, 137.7 (2 d, 2 s, Tol), 154.6 (s, C-4) ppm; IR
(KBr): ν = 3025–2820 (=C-H, C-H), 1670 (C=C), 1360, 1225, 1085,
˜
1045 (C-O) cm–1; ESI-MS (m/z): 220.2 (100, [M + H]+). Anal. calcd. for
C13H17NO2 (219.1): C 71.21, H 7.81, N 6.39; found C 71.49, H 7.73, N
6.16.
Acknowledgments
G. U.-J. and M. J. thank the National Science Centre for financial
support (NCN Preludium grant 2016/21/N/ST5/01254). The au-
thors also thank Dr. Paweł Tokarz (University of Łódź) for help
in kinetic measurements. H.-U. R. thanks Freie Universität Berlin
4-Benzyloxy-2,4a-dihydropiperidino[1,2-b][1,2]-oxazine (9): Fol-
lowing the general procedure, 2,3,4,5-tetrahydropiperidine N-oxide
(99.1 mg, 1.00 mmol) was treated with lithiated benzyloxyallene at
ca. –100 °C (Et2O/CO2 bath) for 4 h. The resulting allenyl hydroxyl-
amine required 2 days for complete cyclization. The crude product for support. We are very thankful to Dr. Christian Mück-Lichten-
Eur. J. Org. Chem. 2020, 6740–6748
www.eurjoc.org
6747
© 2020 The Authors. European Journal of Organic Chemistry
published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/ejoc.202001147
EurJOC
European Journal of Organic Chemistry
[14] J. M. Tao, J. P. Perdew, V. N. Staroverov, G. E. Scuseria, Phys. Rev. Lett.
2003, 91, 146401.
[15] R. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297–3305.
[16] a) S. Grimme, S. Ehrlich, L. Goerigk, J. Comput. Chem. 2011, 32, 1456–
1465; b) S. Grimme, A. Hansen, J. G. Brandenburg, C. Bannwarth, Chem.
Rev. 2016, 116, 5105–5154.
feld (Universität Münster) for his support and for helpful discus-
sions. Open access funding enabled and organized by Projekt
DEAL.
Keywords: Allenes · Cyclization · DFT calculation ·
[17] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.
Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Na-
katsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G.
Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hase-
gawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A.
Montgomery Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Broth-
ers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghava-
chari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega,
J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J.
Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski,
G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö.
Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09,
Revision D.01, Gaussian, Inc., Wallingford CT, 2009.
[18] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.
Cheeseman, G. Scalmani, V. Barone, G. A. Petersson, H. Nakatsuji, X. Li,
M. Caricato, A. V. Marenich, J. Bloino, B. G. Janesko, R. Gomperts, B. Men-
nucci, H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov, J. L. Sonnenberg, D.
Williams-Young, F. Ding, F. Lipparini, F. Egidi, J. Goings, B. Peng, A. Pet-
rone, T. Henderson, D. Ranasinghe, V. G. Zakrzewski, J. Gao, N. Rega, G.
Zheng, W. Liang, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, K. Throssell,
J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. J. Bearpark, J. J. Heyd,
E. N. Brothers, K. N. Kudin, V. N. Staroverov, T. A. Keith, R. Kobayashi, J.
Normand, K. Raghavachari, A. P. Rendell, J. C. Burant, S. S. Iyengar, J.
Tomasi, M. Cossi, J. M. Millam, M. Klene, C. Adamo, R. Cammi, J. W. Och-
terski, R. L. Martin, K. Morokuma, O. Farkas, J. B. Foresman, and D. J. Fox,
Gaussian 16, Revision B.01, Gaussian, Inc., Wallingford CT, 2016.
[19] J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 2005, 105, 2999–3093.
[20] a) R. Huisgen, Angew. Chem. Int. Ed. Engl. 1963, 2, 565–598; Angew. Chem.
1963, 75, 604–637; b) R. Huisgen, Angew. Chem. Int. Ed. Engl. 1963, 2,
633–645; Angew. Chem. 1963, 75, 742–754; c) recent review: M. Breugst,
H.-U. Reissig, Angew. Chem. Int. Ed. 2020, 59, 12293–12307; Angew. Chem.
2020, 132, 12389–124040.
1,2-Oxazine · Reaction mechanisms
[1] Selected reviews: a) R. Zimmer, Synthesis 1993, 165–178; b) R. Zimmer,
H.-U. Reissig, 1-Methoxyallenyllithium in Electronic Encyclopedia of Rea-
gents for Organic Synthesis John Wiley & Sons, Chichester, 2005; second
up-date 2018. DOI: https://doi.org/https://doi.org/10.1002/047084289X;
c) T. Lechel, H.-U. Reissig, Pure Appl. Chem. 2010, 82, 1835–1844; d) A.
Nedolya, O. Tarasova, O. G. Volostnykh, A. L. Albanov, L. V. Klyba, B. A.
Trofimov, Synthesis 2011, 2192–2204; e) R. Zimmer, H.-U. Reissig, Chem.
Soc. Rev. 2014, 43, 2888–2903; f) M. A. Tius, Chem. Soc. Rev. 2014, 43,
2979–3002; g) H.-U. Reissig, R. Zimmer, Synthesis 2017, 49, 3291–3302;
h) V. M. Schmiedel, H.-U. Reissig, Curr. Org. Chem. 2019, 23, 2976–3003.
[2] Reviews on additions of nucleophiles to nitrones: a) M. Lombardo, C.
Trombini, Synlett 2000, 759–774; b) R. Matute, S. Garcia-Vinuales, H.
Hayes, M. Ghirardello, A. Daru, T. Tejero, I. Delso, P. Merino, Curr. Org.
Chem. 2016, 13, 669–686. c) A. A. Tabolin, A. Y. Sukhorukov, S. L. Ioffe,
Chem. Rec. 2018, 10, 1489–1500; d) S.-I. Murahashi, Y. Imada, Chem. Rev.
2019, 119, 4683–4716.
[3] a) W. Schade, H.-U. Reissig, Synlett 1999, 632–634; b) R. Pulz, S. Cicchi, A.
Brandi, H.-U. Reissig, Eur. J. Org. Chem. 2003, 2003, 1153–1156; c) M.
Helms, W. Schade, R. Pulz, T. Watanabe, A. Al-Harrasi, L. Fišera, I. Hlobil-
ová, G. Zahn, H.-U. Reissig, Eur. J. Org. Chem. 2005, 2005, 1003–1019.
[4] a) B. Bressel, H.-U. Reissig, Org. Lett. 2009, 11, 527–539; b) C. Parmeggiani,
F. Cardona, L. Giusti, H.-U. Reissig, A. Goti, Chem. Eur. J. 2013, 19, 10595–
10604; c) T. Pecchioli, F. Cardona, H.-U. Reissig, R. Zimmer, A. Goti, J. Org.
Chem. 2017, 83, 5835–5844.
[5] a) S. Hoff, L. Brandsma, J. F. Arens, Recl. Trav. Chim. Pays-Bas 1969, 88,
609–619; b) D. Gange, P. Magnus, J. Am. Chem. Soc. 1978, 100, 7746–
7747.
[6] a) M. G. Okala Amombo, A. Hausherr, H.-U. Reissig, Synlett 1999, 1871–
1874; b) V. Breuil-Desvergnes, J. Goré, Tetrahedron 2001, 57, 1939–1950;
c) O. Flögel, H.-U. Reissig, Synlett 2004, 895–897; d) M. G. Okala Amombo,
O. Flögel, S. Kord Daoroun Kalai, S. Schoder, U. Warzok, H.-U. Reissig,
Eur. J. Org. Chem. 2017, 1965–1972; e) for calculations of the 5-endo-trig
cyclizations leading to pyrrole and furan derivatives: F. Cumine, A. Young,
H.-U. Reissig, T. Tuttle, J. A. Murphy, Eur. J. Org. Chem. 2017, 2017, 6867–
6871.
[21] B. Dugovič, L. Fišera, H.-U. Reissig, Eur. J. Org. Chem. 2008, 2008, 277–
284.
[22] For the influence of counterions on 5-endo-trig cyclizations leading to
dihydropyrrole derivatives see the calculations in ref.[6e]
.
[7] a) B. Gockel, N. Krause, Org. Lett. 2006, 8, 4485–4488; b) M. Brasholz, H.-U.
Reissig, Synlett 2007, 1294–1298; c) M. Brasholz, B. Dugovič, H.-U. Reissig,
[23] Discovery of the Meisenheimer 1,2-rearrangement: a) J. Meisenheimer,
H. Greeske, A. Willmersdorf, Ber. Dtsch. Chem. Ges. 1922, 55, 513–522;
b) Review: J. J. Li. [1,2]-Meisenheimer rearrangement in Name Reactions,
Springer Verlag, Berlin, Heidelberg, 2009.
[24] For a review on the nitroso ene reaction, see: W. Adam, O. Krebs, Chem.
Rev. 2003, 103, 4131–4146.
Synthesis 2010, 3855–3864; also see ref.[6d]
.
[8] M. Jasiński, G. Mlostoń, M. Stolarski, W. Costa, H.-U. Reissig, Chem. Asian
J. 2014, 9, 2641–2648.
[9] For a first discussion see ref.[3c,1g]
.
[10] The pKa value (in water) of unsubstituted hydroxylamine is reported to
be 13.7. A. F. Holleman, E. Wiberg, N. Wiberg, Lehrbuch der Anorganischen
Chemie. 102. Ed., Walter de Gruyter, Berlin 2007.
[11] a) E. Dumez, J.-P. Dulcère, Chem. Commun. 1998, 479–480; b) E. Dumez,
R. Faure, J.-P. Dulcère, Eur. J. Org. Chem. 2001, 2001, 2577–2588; c) For a
recent publication suggesting a Meisenheimer rearrangement of a styryl-
substituted aziridinium N-oxide into a 1,2-oxazine derivative, see: M.
Hasegawa, T. Suga, T. Soeta, Y. Ukaji, J. Org. Chem. 2020, 85, 11258–
11264.
[25] The hetero-Diels–Alder reaction of moderately electron-rich 1,3-dienes
with nitroso compounds occurs at room temperature or below. For typi-
cal examples, see: a) E. C. Taylor, K. McDaniel, J. S. Skotnicki, J. Org. Chem.
1984, 49, 2500–2501; b) G. Kresze, B. Ascherl, H. Braun, H. Felber, Org.
Prep. Proced. Int. 1987, 19, 329–426; c) P. Sancibrao, D. Gori, C. Kouklov-
sky, G. Vincent, Chem. Eur. J. 2013, 19, 5557–5560.
[26] G. R. Fulmer, A. J. M. Miller, N. H. Sherden, H. E. Gottlieb, A. Nudelman,
B. M. Stoltz, J. E. Bercaw, K. I. Goldberg, Organometallics 2010, 29, 2176–
2179.
[12] An earlier attempt to study the influence of solvents and other reaction
parameters on the formation of bicyclic 1,2-oxazines such as 10 and
their isomeric pyrroline N-oxides provided no conclusive results. C. Man-
nino, H.-U. Reissig, unpublished results 2003.
[27] M. Jasiński, G. Utecht, A. Fruziński, H.-U. Reissig, Synthesis 2016, 48, 893–
905.
[28] M. Jasiński, E. Moreno-Clavijo, H.-U. Reissig, Eur. J. Org. Chem. 2014, 2014,
442–454.
[13] a) C. Reichardt, Angew. Chem. Int. Ed. Engl. 1979, 18, 98–110; Angew.
Chem. 1979, 91, 119–131; b) V. G. Machado, R. I. Stock, C. Reichardt,
Chem. Rev. 2014, 114, 10429–10475.
Received: August 24, 2020
Eur. J. Org. Chem. 2020, 6740–6748
www.eurjoc.org
6748
© 2020 The Authors. European Journal of Organic Chemistry
published by Wiley-VCH GmbH