N-protected 1,2-oxazetidines as a source of electrophilic oxygen: Straightforward access to benzomorpholines and related heterocycles by using a reactive tether

Article abstract of DOI:10.1002/chem.201500731

A hitherto unknown reactivity of a strained four-membered heterocycle, 1,2-oxazetidine, is reported. When reacted with organometallic compounds, this reagent provides electrophilic oxygen with a nitrogen-terminated two-carbon-atom tether. The synthetic versatility of the products obtained was demonstrated in various transformations, leading to efficient synthesis of six-, seven-, and eight-membered heterocyclic systems of pharmaceutical importance. Open and then close: The O-selective ring-opening reaction of 1,2-oxazetidines with organometallic compounds provides tethered phenol derivatives that can be used to access six-, seven-, and eight-membered heterocyclic systems in only a few steps (see scheme; Ts=tosyl).

Full text of DOI:10.1002/chem.201500731

DOI: 10.1002/chem.201500731
Full Paper
&
Synthesis Design |Hot Paper|
N-Protected 1,2-Oxazetidines as a Source of Electrophilic Oxygen:
Straightforward Access to Benzomorpholines and Related
Heterocycles by Using a Reactive Tether
[a]
[a]
[b]
[a]
˙
ˇ
¯
Tomas Javorskis, Simona Sriubaite, Gintautas Bagdziunas, and Edvinas Orentas*
Abstract: A hitherto unknown reactivity of a strained
four-membered heterocycle, 1,2-oxazetidine, is reported.
When reacted with organometallic compounds, this reagent
provides electrophilic oxygen with a nitrogen-terminated
two-carbon-atom tether. The synthetic versatility of the
products obtained was demonstrated in various transfor-
mations, leading to efficient synthesis of six-, seven-, and
eight-membered heterocyclic systems of pharmaceutical
importance.
Introduction
The introduction of oxygen atoms into organic molecules
represents one of the fundamental transformations in organic
synthesis. The oxygenated derivatives serve as starting materi-
als for a wide variety of compounds of practical importance.[1]
Formally, two types of reactions that make use of either
nucleo- or electrophilic oxygen can be recognized. Due to the
ease of generating diverse organic anions, reagents that pro-
vide electrophilic oxygen are highly desired. Commonly
occurring reactions that involve such species include alkene or
imine epoxidation with peracids, Bayer–Villiger rearrangement,
hydroxylation of carbonyl compounds, amine oxidation to
N-oxides or related derivatives, phenol synthesis through the
oxidation of the corresponding boron and silicon compounds,
and ozonolysis. Of special mention is Davis N-sulfonyloxaziri-
dine 1, which finds wide applications in oxygen-transfer
reactions (Scheme 1).[2]
Scheme 1. Possible mimicking of oxaziridine 1 reactivity toward aryl
organometallics with 1,2-oxazetidine 2a and further synthetic application of
as-obtained tethered phenol derivatives (X=H, OMe, or alkyl). Ts=tosyl.
synthetic methods to access more sophisticated derivatives of
this heterocycle.[5] 1,2-Oxazetidine has also attracted attention
from a theoretical point of view as a strained system for the in-
vestigation of the nitrogen inversion process.[6] We were
especially intrigued about the possibility of accessing phenol
derivatives directly from the corresponding aryl metals
prepared by halogen–lithium exchange, Grignard synthesis, or
ortho-metalation. Rare precedents of related transformations
exist and include the oxidation of organometallics by peroxy-
molybdenium,[7] organic peroxides,[8] molecular oxygen,[9] or
oxaziridine;[10] however, these methodologies are rarely used
and the substrate scope seems to be limited. In the case of
Davis oxaziridine 1, the main amine side product forms when
an organolithium reagent reacts with the imine generated
during the reaction. In contrast to oxaziridine, four-membered
congeners would produce an oxygenated derivative with a co-
valently attached two carbon/one nitrogen tether (Scheme 1).
Although at first glance it appears to be redundant, this tether
can be further exploited to transfer a terminal nitrogen atom
into the molecule in several ways. As a plausible application,
we envisioned a radical cyclization pathway for aromatic
compounds, leading to pharmacologically important benzo-
morpholines I.[11] Another viable option is the construction of
medium-sized heterocyclic systems II and III in only a few
In our studies on four-membered heterocyclic systems, we
became interested in whether the same reactivity profile of
oxaziridine could be achieved by using its higher, but still
strained, homologue 1,2-oxazetidine 2a (Scheme 1). The chem-
ical properties of 1,2-oxazetidines have not been investigated
in detail and few syntheses have been described to date.[3] The
erroneous proposal of the involvement of the 1,2-oxazetidine
moiety in the structure of biologically active depsipeptide hali-
peptins[4] A and B has sparked interest in the development of
˙
[a] T. Javorskis, S. Sriubaite, Dr. E. Orentas
Department of Organic Chemistry, Vilnius University
Naugarduko 24, 03225, Vilnius (Lithuania)
E-mail: edvinas.orentas@chf.vu.lt
ˇ
[b] Dr. G. Bagdziu¯nas
Department of Polymer Chemistry and Technology
Kaunas University of Technology
Radvile˙nu˛ pl. 19, 50254 Kaunas (Lithuania)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201500731.
Chem. Eur. J. 2015, 21, 1 – 9
1
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
steps by deploying Lewis acid catalyzed cyclization, hydro-
amination, or Mitsunobu chemistry (Scheme 1).
of pyramidilization of the latter, however, is significantly small-
er than that in a reported N-alkylated derivative of 1,2-oxazeti-
dine (Figure 1b).[13] This observation is in agreement with neg-
ative hyperconjugation of a nitrogen lone pair with the sulfo
group, resulting in partial planarization of the sulfonamide ni-
trogen. This effect is also known to account for the lower nitro-
gen inversion barrier in sulfonylated oxaziridines compared
with that of alkylated derivatives.[14] In the crystal structure, the
lone pair of nitrogen is placed in the bisecting plane of the
SO2 group, which corresponds to the most commonly found
conformer of sulfonamides in the solid state.[15] The length of
the NÀO bond is 1.495 ꢁ; this
Results and Discussion
To test the above-proposed ideas, we first developed a simple,
large-scale synthesis of N-tosyl-1,2-oxazetidine (2a), as well as
tert-butyloxycarbonyl (Boc)- and carbobenzyloxy (Cbz)-protect-
ed derivatives 2b and 2c, respectively, to assess the effect of
the protecting group on ring strain, and thus, on the reactivity
of 1,2-oxazetidine (Figure 1). The synthesis comprised of two
value is between those of oxazir-
idine and isoxazolidine. Interest-
ingly, the CÀC bond is unusually
short: d=1.510 ꢁ (Figure 1b).[16]
Regarding the packing of 2a in
the crystal, the two enantiomers
that result from the stereogenic
nitrogen form chains that are
held together by hydrogen
bonds between the oxazetidine
oxygen atom and sp2-CÀH-bond
hydrogen atom that is ortho to
the methyl group. The as-
formed homochiral chains are
connected through two hydro-
gen-bonding arrays comprised
of the oxygen acceptor in the
Figure 1. a) Synthesis of N-protected 1,2-oxazetidines. i) For 2a: TsCl (2.4 equiv), pyridine (Py); for 2b: Boc2O
(1.1 equiv), Et3N (1.4 equiv), CH2Cl2; for 2c: CbzCl (1.2 equiv), Et3N (1.4 equiv), CH2Cl2; ii) NaH (1.3–2.3 equiv), THF.
b) The geometry of the 1,2-oxazetidine ring; bond lengths are indicated (left: front view; right: side view). The
aromatic ring is omitted for clarity. c) Packing of two enantiomeric chains of 2a in a crystal; the short contacts are
indicated (green dashed lines).
SO2 group and the acidic ortho
sp2-CÀH hydrogen donor. Fur-
thermore, there are also short
contacts observed that result
from sp3-CÀHÀp interactions be-
tween the methyl groups and
trivial steps from hydroxylamine salt 3, which was easily avail-
able in large amounts from inexpensive starting materials.[12]
The protection of hydroxylamine and subsequent cyclization
by using sodium hydride as a base afforded gram quantities of
desired oxazetidine 2a as a colorless, bench-stable, crystalline
solid. The derivatives 2b and 2c were synthesized by following
the same protocol. Alternatively, the carbamate-protected
1,2-oxazetidines 2b and 2c could be accessed in two steps
starting from Boc- or Cbz-protected commercially available hy-
droxylamines (see the Supporting Information). The high ring
strain in compounds 2a–c was evident from unusually down-
field-shifted resonances of methylene protons (e.g., ÀCH2ÀOÀ
at d=4.90 ppm and ÀCH2ÀNÀ at d=4.63 ppm for 2c in CDCl3)
and also from readily occurring ring opening under strongly
acidic or hydrogenolysis conditions.
the aromatic ring with a hydrogen atom located on top of the
sp2-carbon atom connected to the sulfo group (Figure 1c).
To compare the reactivity of 1,2-oxazetidines 2a,b and oxa-
ziridine 1, and also to evaluate the influence of the protecting
group on ring strain, we used two test reactions. To probe the
electrophilic oxygen character, compounds 2a and 2b were
treated with stoichiometric amount of PPh3. As expected, the
tosyl derivative 2a afforded the corresponding aziridine 4a in
quantitative yield after 16 h at ambient temperature in THF
(Scheme 2a).
The X-ray structure of crystalline 2a allowed, for the first
time, the characterization of the geometric features of the N-
sulfonylated 1,2-oxazetidine ring in the solid state. The crystals
of 2a were grown from a solution in dichloromethane by slow
evaporation at room temperature. The 1,2-oxazetidine ring in
2a is almost planar, showing only a small distortion from the
plane due to the pyramidilized sp3-nitrogen atom. The degree
Scheme 2. a) The reaction of compounds 2a,b with PPh3. b) The reaction of
compounds 2a,b with SmI2.
&
&
Chem. Eur. J. 2015, 21, 1 – 9
www.chemeurj.org
2
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Surprisingly, after the same reaction time, carbamate 2b
gave solely amino alcohol 5b. This observation can be ex-
plained by a two-step mechanism that involves fast oxidative
insertion of PPh3 into the oxazetidine ring to produce inter-
mediate A and subsequent transformation of the latter into
The control experiments with acyclic tosylhydroxylamine deriv-
ative 64 (Figure 2b) produced no traces of anisoles, and there-
fore, confirmed that the ring-opening reactions with 2a were
strain-release driven. The reaction was successful with aryl and
heteroaryl compounds, such as derivatives of pyridines or thio-
phene (Figure 2a, 46–56). The notable exceptions were highly
electron-deficient organometallic derivatives that gave a very
low yield of the expected products; this was most likely to be
due to their reduced nucleophilicity (Figure 2c). Likewise, al-
kynes such as phenylacetylene or 4-methoxyphenylacetylene
were also inert. Oxazetidine 2a reacts readily with alkyl lithium
derivatives to afford the expected ring-opening products. We
did not elaborate the scope of this reaction any further be-
cause the above compounds containing an alkyl substituent
could be synthesized from the corresponding readily available
alcohols by alternative methods. Nevertheless, this type of re-
activity could be very useful in more complicated systems,
such as Trçger bases, in which the benzylic positions can be
lithiated in a straightforward fashion with sBuLi (Figure 2a, 58).
The exo-selective installation of an amino-terminated chain on
a Trçger base scaffold in high yield provided a very versatile
handle for its further functionalization.[20] In addition to carbon
nucleophiles, the lithium salt of diphenylamine also reacted
smoothly to afford the hydroxylamine derivative (see the Sup-
porting Information). The ring-opening reactions with alkenyl-
lithium reagents generated by lithium–halogen exchange or
direct lithiation proceeded smoothly to give vinyl ethers in
good yields (Figure 2a, 43–45).
In addition to halogen–lithium exchange, the organolithium
compounds obtained by directed ortho-lithiation were also
successfully applied in this reaction, as exemplified with sulfo-
namide-, amide-, and methoxy-directing groups (Figure 2a,
59–63). In this regard, the utilization of the rarely used hydrox-
ymethyl ortho-directing group in benzyl alcohols seemed to be
especially attractive because the corresponding products could
be potentially utilized in ring-closure reactions to obtain 1,4-
oxazepines III (Scheme 1, X=H). Thus, treating benzyl alcohol
71 with two equivalents of nBuLi in boiling hexane with the
subsequent addition of 2a afforded the ortho-functionalized
product in 58% yield (Scheme 3a). Despite the moderate yield,
this approach offers the advantage of using a very cheap start-
ing material and the possibility of simultaneously installing the
reactive tether for the ring-closing step. The corresponding
1,4-oxazepines were then obtained by an intramolecular Mitsu-
nobu reaction.[21] Moreover, the attachment of the substituent
on the benzylic position and ortho-lithiation can be performed
in a one-pot procedure, as demonstrated in the synthesis of
secondary and tertiary benzylic alcohols 75 and 77, starting
from aldehyde 74 and methyl benzoate 76, respectively
(Scheme 3a). It should be noted that, when lithiation is per-
formed on 2-bromobenzylic alcohols, no improvements in
yield are achieved relative to the directed ortho-metalation of
an alcohol. We found that, with the brominated starting mate-
rials, bromine/lithium exchange competed with the deprotona-
tion step, sometimes to as much as 40% for compound 78
(Scheme 3b). In case of the compound 80 (Scheme 3b), the
directed ortho-lithiation of a non-brominated substrate is not
aziridine and triphenylphosphine oxide via zwitterion
B
(Scheme 2a). Indeed, when the reaction with compound 2a
was quenched after 2 h, none of the starting material or amino
alcohol 5a were detected (see the Supporting Information),
which indicated full conversion to A (Scheme 2a). In the case
of compound 2b, the five-membered insertion product is too
unstable toward an aqueous workup and instead is directly hy-
drolyzed into the corresponding amino alcohol 5b.[17] On the
other hand, intermediate A formed from 2b does not react fur-
ther to give the corresponding aziridine; this is most likely to
be due to the unfavorable formation of a more basic nitrogen
anion in intermediate B. The reactivity profile of 1,2-oxazeti-
dine towards phosphorous nucleophiles is thus different from
that of Davis oxaziridine 1, for which the carbon atom is at-
tacked preferentially over oxygen to produce a four-membered
Wittig-type intermediate.[18] In another test reaction with SmI2,
the electron-acceptor properties of 1,2-oxazetidines were
probed. When one equivalent of the one-electron donor SmI2
was added to oxazetidine 2b, almost full conversion was ach-
ieved in a few minutes. The same reaction with sulfonyl deriva-
tive 2a required a longer time and larger amount of SmI2 (ca.
16 h) and gave the expected amino alcohol 5a as the main
product. Surprisingly, with carbamate 2b, no amino alcohol
was obtained, but instead the iodinated hydroxylamine deriva-
tive 6 was isolated. The formation of this product can be ra-
tionalized by assuming the activation of the oxazetidine ring
with Lewis acid SmI2 followed by intramolecular iodide attack
on the carbon atom (Scheme 2b).[19] In this case, only one
equivalent of SmI2 is sufficient for full conversion, as opposed
to two equivalents otherwise required to reduce the NÀO
bond. These findings provide a possibility for the future use of
carbamate-protected 1,2-oxazetidines in C-selective ring-open-
ing reactions with nucleophiles by using Lewis acid activators
with non-nucleophilic counterions (e.g., triflates). In contrast to
1, compounds 2a–c were not reactive toward sulfur nucleo-
philes, such as diphenyl sulfide.
Having confirmed the high reactivity of 1,2-oxazetidines
toward nucleophilic agents, we next examined the reaction of
2a with different aryl and alkenyl organometallic compounds
(Figure 2a). The N-sulfonyl-protected 1,2-oxazetidine 2a was
chosen as a model compound based on the very good com-
patibility of this protecting group with almost all types of orga-
nometallics. Gratifyingly, we observed very fast ring opening
through nucleophilic attack on the oxygen atom. The reac-
tions, in most cases, proceeded instantaneously at À788C. To
enhance the reactivity of the organometallic reagents, two
equivalents of N,N,N’,N’-tetramethylethylenediamine (TMEDA)
were used. High yields were obtained with a range of sub-
strates with no traces of product resulting from nucleophilic
attack on nitrogen. Both organolithium and magnesium deriva-
tives were active nucleophiles; however, if possible, Br/Li ex-
change by using tBuLi or nBuLi was applied for convenience.
Chem. Eur. J. 2015, 21, 1 – 9
www.chemeurj.org
3
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 2. a) The scope of the reaction of compound 2a toward organometallic reagents. b) Control reaction with an acyclic analogue of 2a. c) Non-reactive
substrates that give a yield of <40%.
regioselective and the use of bromobenzylic alcohol was
necessary to secure the regioselectivity of the reaction.
plication of an intramolecular iron-catalyzed hydroamination
strategy by using vinyl ethers as starting materials. The results
are shown in Scheme 4a. The full conversion of vinyl ethers 43
and 44 was obtained in a few minutes in DCE or chloroform at
room temperature, yielding protected aldehydes 91 and 92 in
78 and 80% yield, respectively. In an analogous fashion, the 8-
membered ring 95 was easily assembled from precursor 94 in
75% yield. The vinyl ethers thus proved to be exceptionally
active substrates in the hydroamination reaction.[24] The reac-
tion of the alkene derivative obtained from previously de-
scribed alcohol 75 required somewhat harsher conditions
(808C) and provided alkylated 1,4-oxazepine in 40% overall
yield without purification of the intermediate alkene
(Scheme 4a). Taken together, the procedures described herein
demonstrate the potential of 1,2-oxazetidines in achieving
a short and simple synthesis of medium-ring heterocycles.
For the synthesis of the 1,4-oxazepines with substituents at
the benzylic position, we envision the use of benzaldehyde de-
rivatives for the cyclization step. This strategy would allow the
generation of cyclic N,O-acetals that can be further utilized in
Lewis acid catalyzed reactions with various nucleophiles.[22]
Thus, in situ protection of an aldehyde group with a TMEDA-
like amine generated a temporal directing group that was
used for ortho-lithiation.[23] The reaction of the as-obtained aryl
lithium with 2a afforded aldehyde derivatives 84, 87, and 88
in good to moderate yields (Scheme 3c). The cyclization step
was performed by using trimethyl orthoformate as a solvent
and FeCl3 as a catalyst; these conditions have not been
previously described in the literature (Scheme 3c, 89 and 90).
To further demonstrate the synthetic utility of the tethered
phenol derivatives, we turned our attention to the possible ap-
&
&
Chem. Eur. J. 2015, 21, 1 – 9
www.chemeurj.org
4
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
ic acid derivative 97 reacted smoothly with 2a to
give 98, which was then converted into morpholine
99 in two simple steps (Scheme 4b).
Although the synthesis of simple phenols was not
our primary goal, we recognized the potential of an
ethanolamine tether to act as a directing group for
the regioselective deprotection of substrates that
also contained methoxy groups (Scheme 4c). It was
postulated that the coordination of boron tribromide
through chelation might produce intermediate A
with the boron-activating oxygen atom of the tether.
This assumption was borne out in practice with the
successful selective deprotection of compounds 63
and 61 by boron tribromide. When the methoxy
group is located in the ortho position to the tether,
the simultaneous activation of two groups is
achieved through intermediate B, as demonstrated
with compound 23 (Scheme 4c).
Next, we wanted to explore the synthetic applica-
tion of carbamate derivatives 2b,c in more detail.
The advantage of using a carbamate protecting
group in 2b,c is twofold: it is easily deprotected
under mild conditions to give free amine and it
would also provide a more nucleophilic nitrogen
atom after ring opening with organometallics, which
can be further exploited for various cyclization reac-
tions. The synthetic transformations with these sub-
strates are summarized in Scheme 5. Unfortunately,
the carbamate protecting group was not compatible
with unhindered organolithium or magnesium re-
agents and resulted in alkoxycarbonyl group transfer from 2b
(Scheme 5a, top right). On the other hand, by using sterically
crowded organolithium reagents, the desired products were
Scheme 3. a,) b) The ortho-lithiation cyclization strategy toward 1,4-oxazepines with
benzyl alcohols. DEAD=diethyl azodicarboxylate. c) The ortho-lithiation cyclization
strategy toward functionalized 1,4-oxazepines.
Scheme 4. a) Intramolecular iron-catalyzed hydroamination of vinyl ethers
and alkene. DCE=dichloroethane, TFA=trifluoroacetic acid. b) Preparation
of a 3,3-disubstituted morpholine derivative. c) Synthesis of phenols by
chelation-assisted regioselective deprotection.
Scheme 5. a) Selected transformations with carbamate-protected 1,2-oxaze-
tidine 2b. b) One-step synthesis of highly functionalized morpholines fused
with a pyrimidine heterocycle. c) Attempted use of organozinc reagents for
the ring-opening reaction with 2b,c.
In addition to simple aryl organometallics, the ester enolate
was also a suitable nucleophile. The lithium enolate of mandel-
Chem. Eur. J. 2015, 21, 1 – 9
www.chemeurj.org
5
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
obtained in good yields (Scheme 5a, bottom left and middle).
It is important to note that compounds 2b,c are much more
reactive towards organometallics than 2a and high yields of
ring-opening products 106 and 107 were obtained, even with
weak halogenated aryllithium nucleophiles (Scheme 5a, top
left). The higher nucleophilicity of nitrogen in the carbamate
anion allowed us to synthesize morpholine derivative 108 in
just one step from the O-methyl mandelic acid ester. The labile
Boc group was lost during workup to afford a crystalline prod-
uct that did not require chromatographic purification
(Scheme 5a, bottom right).
To take advantage of high reactivity of carbamate 2b
toward nucleophilic attack, we performed the reaction with
a relatively weakly nucleophilic, and therefore, non-reactive
toward 2a lithiated pyrimidines derived from compounds 109
and 110. After ring opening, intramolecular nucleophilic substi-
tution took place to provide morpholines 111 and 112 in good
yields (Scheme 5b). The presence of chlorine and SMe groups
in the products obtained can be used for further functionaliza-
tion of these potentially biologically active compounds. Our at-
tempts to decrease the reactivity of organolithium and magne-
sium reagents by transmetalation to the corresponding zinc
derivatives did not meet with success. Unexpectedly, the hy-
droxylamine derivatives 113 a,b were obtained in high yields in
the reaction of 2b,c with an organozinc reagent (Scheme 5c).
These results can be easily explained by assuming the same
type of Lewis acid activation of 1,2-oxazetidine as that previ-
ously described with SmI2 (Scheme 2b). The control experi-
ment revealed that MgBr2, which formed after the transmetala-
tion step, was the actual Lewis acid and source of the bromide
anion (see the Supporting Information). When the organozinc
reagent was generated by using a magnesium-free protocol,
no reaction was observed, even at room temperature, and
2b,c were fully recovered. The much higher activity of the car-
bamate derivatives 2b,c toward nucleophiles, relative to that
of the tosyl derivative 2a, seemed surprising, especially when
taking into account the fact that the electron-withdrawing
properties of the sulfonyl group should increase the electro-
philicity of the oxygen atom in the oxazetidine ring. Most
likely, conjugation of the nitrogen atom lone pair with the car-
bamate carbonyl in 2b and 2c reduces the electron density on
nitrogen, and thus, subsequently on the neighboring oxygen
atom to render it more electrophilic. In contrast to carbamate-
protected oxaziridine congeners,[18b] which are known to
provide electrophilic nitrogen, oxazetidines 2b,c react only
through the oxygen-selective ring-opening pathway.
Figure 3. Synthesis of benzomorpholines by photochemical radical
cyclization. PIDA=phenyliodonium diacetate, NIS=N-iodosuccinimide.
cyclization reaction of N-centered radicals. Surprisingly, despite
a number of studies on the cyclization of tosyl amines,[26] the
application of this type of reaction for the synthesis of pharma-
cologically important benzomorpholine derivatives has not
been thoroughly investigated.[27] More importantly, there are
no examples of this radical chemistry being applied in
heterocyclic systems, such as pyridines. This motivated us to
investigate the radical cyclization strategy in more detail.[28]
The previously synthesized compounds with the N-tosyl
ethanolamine tether were first subjected to the most widely
used Suꢂrez conditions.[29] Under these conditions, PIDA reacts
with iodine to produce an electrophilic iodine source, iodine
acetate, which further reacts with tosyl amine to generate the
N-iodo derivative. This intermediate dissociates through homo-
lytic cleavage under illumination with visible light to generate
the N-centered radical. In our study, we used K2CO3 as a base
to deprotonate the tosyl amine and also to scavenge the acid
produced during the reaction. As seen from Figure 3,
a number of substrates afforded the expected benzomorpho-
lines in high to moderate yields. We also observed that the use
of K2CO3 base had the advantage of almost completely inhibit-
ing the iodination of the starting materials or products (except
119; Figure 3), as often observed in similar reactions.[26] In gen-
eral, the substrates with electron-donating groups readily react
under Suꢂrez conditions. However, we noticed that some sub-
strates, such as pyridine 49b (Figure 2a), formed an unreactive
complex with PIDA from which starting material was recovered
after reductive (aq. Na2SO3) workup. In a few other instances,
the standard conditions provided a negligible amount of prod-
uct. Having suspected that the use of PIDA was responsible for
these complications, we devised an alternative procedure to
The presence of a tosyl amino functionality in the phenol
derivatives described so far encouraged us to consider the
possible transfer of the nitrogen atom into the aromatic ring.
He and co-workers recently described the formation of qui-
none derivatives and not the expected benzomorpholines in
a rhodium-catalyzed reaction with the same tethered phenol
derivatives (Figure 3).[25] Our simple approach described herein
can provide the starting materials for this reaction in one step
from the corresponding aryllithiums. We reasoned, however,
that the sought-after benzomorpholine derivatives could be
easily obtained through the radical rather than metal-catalyzed
&
&
Chem. Eur. J. 2015, 21, 1 – 9
www.chemeurj.org
6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
generate N-iodo derivatives. Thus, by treating the problematic
compounds 19 and 51 (Figure 2) with two equivalents of NIS
in the presence of K2CO3, products 126 and 128 were obtained
(Figure 3). NIS served as a donor of electrophilic iodine, which
was transferred to the tosyl amine anion. The use of substrates
containing cyano, sulfonamide, or amide groups afforded less
than 30% yield of the product. Likewise, only activated pyri-
dines with a methoxy group were reactive. This observation is
in full agreement with the electrophilic character of the sulfo-
namide radical. Unfortunately, neither method was successful
with ferrocene- and benzofurane-derived starting materials
(Figure 2, 32 and 34) due to a high sensitivity of these com-
pounds toward oxidants. Interestingly, when starting material
31, with two isopropyl ortho substituents, was subjected to
Suꢂrez conditions, a high yield of 1,4-oxazepine 130 was ob-
tained (Scheme 6a). In this case, the N-centered radical formed
(Scheme 4a) can be used instead of radical cyclization to direct
the tether to the benzylic position. It should also be noted
that with some substrates unexpected products were ob-
tained. For example, the reaction of compound 23, with ortho-
and para-methoxy groups, afforded quinone product 133
rather than the desired benzomorpholine (Scheme 6c). This re-
action was very specific for this particular substitution pattern
and the same type of product was obtained with pyridine ana-
logue 46. The moderate yield of 134 was likely to be related
to its high chemical reactivity because it was shown that 134
was spontaneously hydrolyzed into compound 135 in solution
by atmospheric moisture (Scheme 6c). Since both substrates
gave the same products with NIS, hypervalent iodine species
are unlikely to be involved in the reaction mechanism beyond
the electrophilic iodine generation step. The plausible mecha-
nism is outlined in Scheme 6c. The N-centered radical under-
goes ipso addition to generate a new radical, which benefits
from stabilization by two methoxy groups. This radical is then
further oxidized into the corresponding cation, which affords
the observed product after hydrolysis (Scheme 6c). Com-
pounds 133 and 135 have three differentiated carbonyl
groups and can be useful synthons for various transformations
of practical relevance.
Conclusion
We developed a simple and scalable synthesis of carbamate-
and tosyl-protected 1,2-oxazetidines and demonstrated their
synthetic applications. We showed that the strategy proposed
herein represented
a very concise and straightforward
synthesis of six-, seven-, and eight-membered heterocyclic
systems from simple starting materials.
Acknowledgements
The research was funded by the European Social Fund under
ˇ
Global Grand Measure (VP1-3.1-SMM-07-K-03-007). We thank
Lukas Taujenis (Department of Analytical and Environmental
Chemistry, Vilnius University) for assistance with mass
spectrometry. We are grateful to Dr. Vaidotas Navickas (Fine
Chemicals & Biocatalysis Research - Biocatalysis, BASF SE, Lud-
wigshafen, Germany) for the generous gift of some chemicals.
Scheme 6. a) Synthesis and mechanism of 1,4-oxazepine formation by
photochemical radical cyclization. b) The regioselective formation of
benzomorpholine. c) Untypical cyclization pathway with 2,4-dimethoxy-
substituted substrates and the proposed reaction mechanism.
Keywords: heterocycles · 1,2-oxazetidines · oxygen · radicals ·
synthesis design
isomerizes to form a C-centered benzylic radical through [1,7]-
hydrogen transfer. After oxidation of the benzylic radical, the
product is formed by an SN1-type intramolecular reaction. To
test which reaction pathway of the radical, [1,7]-hydrogen
transfer or addition to the aromatic ring, was preferred, we
performed the reaction with compound 131. The exclusive for-
mation of benzomorpholine 132 was observed in this case
(Scheme 6b). On the other hand, if the 1,4-oxazepine product
is desired, the hydroamination strategy described earlier
[1] For some recent applications, see: a) D. I. Coppi, A. Salomone, F. M.
Perna, V. Capriati, Chem. Commun. 2011, 47, 9918–9920; b) R. Mansue-
to, V. Mallardo, F. M. Perna, A. Salomone, V. Capriati, Chem. Commun.
2013, 49, 10160–10162; c) D. I. Coppi, A. Salomone, F. M. Perna, V. Cap-
riati, Angew. Chem. Int. Ed. 2012, 51, 7532–7536; Angew. Chem. 2012,
124, 7650–7654.
[2] For reviews, see: a) F. A. Davis, A. C. Sheppard, Tetrahedron 1989, 45,
5703–5742; b) F. A. Davis, B. C. Chen, Chem. Rev. 1992, 92, 919–934.
[3] a) J. E. Baldwin, A. K. Bhatnagar, S. Chun Choi, T. J. Shortridge, J. Am.
Chem. Soc. 1971, 93, 4082–4084; b) G. Pifferi, P. Consonni, J. Heterocyclic
Chem. 1972, 9, 159–160; c) R. Luisi, V. Capriati, S. Florio, E. Piccolo, J.
Chem. Eur. J. 2015, 21, 1 – 9
www.chemeurj.org
7
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Org. Chem. 2003, 68, 10187–10190; d) V. Capriati, S. Florio, R. Luisi, A.
Salomone, C. Cuocci, Org. Lett. 2006, 8, 3923–3926.
[4] A. Randazzo, G. Bifulco, C. Giannini, M. Bucci, C. Debitus, G. Cirino, L.
Gomez-Paloma, J. Am. Chem. Soc. 2001, 123, 10870–10876.
[5] B. B. Snider, J. R. Duvall, Tetrahedron Lett. 2003, 44, 3067–3070.
[6] J. D. Readio, J. Org. Chem. 1970, 35, 1607–1611.
[19] An analogous ring-opening reaction has also been observed with epox-
ides, see: M. Matsukawa, T. Tabuchi, J. Inanaga, M. Yamaguchi, Chem.
Lett. 1987, 2101–2102.
[20] J. Artacho, E. Ascic, T. Rantanen, C. J. Wallentin, S. Dawaigher, K. E. Berg-
quist, M. Harmata, V. Snieckus, K. Wꢃrnmark, Org. Lett. 2012, 14, 4706–
4709.
[7] N. J. Lewis, S. Y. Gabhe, M. R. DeLaMater, J. Org. Chem. 1977, 42, 1479–
1480.
[21] For recent examples of 1,4-oxazepine synthesis, see: a) M. Yar, E. M.
McGarrigle, V. K. Aggarwal, Org. Lett. 2009, 11, 257–260; b) J. Rujirawa-
nich, T. Gallagher, Org. Lett. 2009, 11, 5494–5496.
[22] W. N. Speckamp, M. J. Moolenaar, Tetrahedron 2000, 56, 3817–3856.
[23] D. L. Comins, Synlett 1992, 8, 615–625.
[24] One example of intermolecular FeCl3-catalyzed hydroamination of vinyl
ethers has recently been reported, see: Ch. Dal Zotto, J. Michaux, A.
Zarate-Ruiz, E. Gayon, D. Virieux, J.-M. Campagne, V. Terrasson, G. Piet-
ers, A. Gaucher, D. Prim, J. Organomet. Chem. 2011, 696, 296–304.
[25] Y. X. Deng, J. P. Xie, W. W. Zhang, P. Yin, J. Yu, L. He, Chem. Eur. J. 2012,
18, 1077–1082.
[8] a) G. A. Baramki, H. S. Chang, J. T. Edward, Can. J. Chem. 1962, 40, 441–
444; b) Y. Hou, C. Y. Meyers, M. Akomeah, J. Org. Chem. 2009, 74, 6362–
6364; c) M. Taddei, A. Ricci, Synthesis 1986, 633–635; d) F. A. Davis, G. S.
Lai, J. Wei, Tetrahedron Lett. 1988, 29, 4269–4272; e) R. Willand-Charn-
ley, B. W. Puffer, P. H. Dussault, J. Am. Chem. Soc. 2014, 136, 5821–5823.
[9] K. A. Parker, K. A. Koziski, J. Org. Chem. 1987, 52, 674–676.
[10] F. A. Davis, P. A. Mancinelli, K. Balasubramanian, U. K. Nadir, J. Am. Chem.
Soc. 1979, 101, 1044–1045.
ˇ
[11] J. Ilas, P. S. Anderluh, M. S. Dolenc, D. Kikelj, Tetrahedron 2005, 61,
7325–7348.
[26] a) H. Togo, Y. Hoshina, M. Yokoyama, Tetrahedron Lett. 1996, 37, 6129–
6132; b) H. Togo, Y. Hoshina, T. Muraki, H. Nakayama, M. Yokoyama, J.
Org. Chem. 1998, 63, 5193–5200; c) Y. Li, Q. Ding, G. Qiub, J. Wu, Org.
Biomol. Chem. 2014, 12, 149–155.
[12] L. Bauerand, K. S. Suresh, J. Org. Chem. 1963, 28, 1604–1608.
[13] S. Florio, V. Capriati, R. Luisi, A. Salomone, C. Cuocci, Z. Kristallogr.- New.
Crys. Struct. 2006, 221, 398–400.
[14] W. B. Jennings, S. P. Watson, M. S. Tolley, J. Am. Chem. Soc. 1987, 109,
8099–8100.
[15] A. Parkin, A. Collins, Ch. J. Gilmore, C. C. Wilson, Acta Crystallogr. Sect. B
2008, 64, 66–71.
[16] F. H. Allen, O. Kennard, D. G. Watson, L. Brammer, A. G. Orpen, R. J.
Taylor, J. Chem. Soc. Perkin Trans. 2 1987, S1–S19.
[17] To the best of our knowledge, there is only one example of a similar
1,3,2l5-oxazaphospholidine derivative being reported in literature. It
was shown that it could be converted into the corresponding aziridine
by heating or it was easily hydrolyzed to give the amino alcohol; see:
R. Appel, M. Halstenberg, Chem. Ber. 1976, 109, 814–818.
[18] a) Y. Hata, M. Watanabe, J. Am. Chem. Soc. 1979, 101, 6671–6676; b) J.
Vidal, S. Damestoy, L. Guy, J. C. Hannachi, A. Aubry, A. Collet, Chem. Eur.
J. 1997, 3, 1691–1709.
[27] Only a single example of the radical approach to benzomorpholine
with a triflimide derivative has been reported, see ref. [26b].
[28] For recent examples of benzomorpholine synthesis, see: a) P. Thansan-
dote, E. Chong, K.-O. Feldmann, M. Lautens, J. Org. Chem. 2010, 75,
3495–3498; b) Z. Lu, S. S. Stahl, Org. Lett. 2012, 14, 1234–1237.
[29] a) R. L. Dorta, C. G. Francisco, E. J. Suꢂrez, J. Chem. Soc. Chem. Commun.
1989, 1168–1169; b) P. de Armas, C. G. Francisco, R. Hernꢂndez, J. A. Sal-
azar, E. J. Suꢂrez, J. Chem. Soc. Perkin Trans. 1 1988, 3255–3265; c) R.
Carrau, R. Hernꢂndez, E. J. Suꢂrez, C. Betancor, J. Chem. Soc. Perkin
Trans. 1 1987, 937–943.
Received: February 22, 2015
Published online on && &&, 0000
&
&
Chem. Eur. J. 2015, 21, 1 – 9
www.chemeurj.org
8
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
FULL PAPER
&
Synthesis Design
˙
ˇ
T. Javorskis, S. Sriubaite, G. Bagdziunas,
¯
E. Orentas*
&& – &&
N-Protected 1,2-Oxazetidines as
a Source of Electrophilic Oxygen:
Straightforward Access to
Benzomorpholines and Related
Heterocycles by Using a Reactive
Tether
Open and then close: The O-selective
ring-opening reaction of 1,2-oxazeti-
dines with organometallic compounds
provides tethered phenol derivatives
that can be used to access six-, seven-,
and eight-membered heterocyclic
systems in only a few steps (see
scheme; Ts=tosyl).
Chem. Eur. J. 2015, 21, 1 – 9
www.chemeurj.org
9
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ