Reaction of N-alkyl azetidines with triphosgene

Article abstract of DOI:10.1016/j.tetlet.2015.10.036

N-Alkyl azetidines react with triphosgene (BTC) following two possible pathways: N-alkyl ring scission or ring cleavage, to give cyclic or acyclic N-carbamoyl chlorides. Predominance of one pathway over the other is governed by the nature of the substituents on the azetidine ring and on the nitrogen atom as well as by the relative stereochemistry of the ring substituents, and is examined in detail. Some azetidines were identified for their privileged reaction pathway, leading to new functionalized building blocks that were further elaborated into five- or six-membered urea or into azetidinic urea.

Full text of DOI:10.1016/j.tetlet.2015.10.036

Tetrahedron Letters 56 (2015) 6625–6628
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Reaction of N-alkyl azetidines with triphosgene
⇑
Laurence Menguy, Bruno Drouillat, François Couty
Institut Lavoisier, Université de Versailles St Quentin-en-Yvelines, UMR 8180, 45 avenue des Etats-Unis, 78035 Versailles, France
a r t i c l e i n f o
a b s t r a c t
Article history:
N-Alkyl azetidines react with triphosgene (BTC) following two possible pathways: N-alkyl ring scission or
ring cleavage, to give cyclic or acyclic N-carbamoyl chlorides. Predominance of one pathway over the
other is governed by the nature of the substituents on the azetidine ring and on the nitrogen atom as well
as by the relative stereochemistry of the ring substituents, and is examined in detail. Some azetidines
were identified for their privileged reaction pathway, leading to new functionalized building blocks that
were further elaborated into five- or six-membered urea or into azetidinic urea.
Received 9 September 2015
Revised 7 October 2015
Accepted 9 October 2015
Available online 19 October 2015
Keywords:
Azetidines
Urea
Ó 2015 Elsevier Ltd. All rights reserved.
Ring cleavage
Triphosgene
Azetidines are clearly gaining attention within the past decade
as privileged scaffolds for the development of new pharmaceuti-
cals.1 This can be ascribed both from the growing knowledge
allowing to prepare these heterocycles, especially in enantiomeri-
cally pure form,2 but also from patent strategy. In these strained
heterocycles, the nature of the protecting group on the nitrogen
atom is of utmost importance, and influences to a large extent their
reactivity which is governed by their strain: N-alkyl azetidines
being generally less prone to ring opening than their N-acyl or
N-sulfonyl analogues under acidic conditions.3 Apart from pure
synthetic considerations, the inter-conversion between N-alkyl
and N-acyl group in azetidines is an important issue which lacks
generality4 and is rendered difficult because of the ring strain. In
this context, alkyl chloroformates, which are popular reagents to
promote such reaction in tertiary amines, react mainly with
N-alkyl azetidines through a ring opening5 (Scheme 1, path A).
We report herein the use of BTC (bis-trichloromethylcarbonate,
triphosgene), a safe substitute for phosgene,6 which reacts with
N-alkyl azetidines following two possible pathways: N-alkyl cleav-
age (Scheme 1, path B), or ring cleavage (path A) depending on the
substrate. The resulting products can then be further elaborated
into different ureas, thus illustrating the chemical diversity that
can be reached with these heterocycles.7
cyano group allows for further functionalization.2g Results are
depicted in Scheme 2.
Reaction occurs readily in dichloromethane at room tempera-
ture with mono and disubstituted 1 and 2 to give exclusively
ring-opened compounds with moderate regioselectivity favouring
C-2 cleavage. When trisubstituted azetidines 3 and 4 are involved,
no reaction occurs, even with protracted reaction time.
We next selected N-benzhydryl azetidines 9, 12 and 16 as sub-
strates, aiming at favoring cleavage of the N-substituent. Thus,
commercially available 9 readily reacted with BTC8 to give a mix-
ture of carbamoyl chloride 109 and opened product 11, in a 7:3
ratio, based on the examination of the crude reaction mixture by
NMR. On the other hand, N-benzhydryl 2-cyano azetidine 12 and
trisubstituted ephedrine-derived N-benzhydrylazetidine 16 (9:1
epimeric mixture at C-2), prepared following Scheme 3,10 were
found to be completely inert towards BTC, similarly to 3 and 4, thus
illustrating the high sensitivity of this reaction towards steric
crowding around the nitrogen atom.
In the series of N-benzylic substrates, diastereoisomeric
azetidines 17 and 20, readily prepared from (S)-phenylethy-
lamine,4d and fitted with a substituted benzyl group were also
reacted with BTC, to provide interesting chiral functionalized
azetidinic carbamoyl chlorides 18 and ent-18, together with
minor amounts of opened products 19 and 21. Compounds 18
and ent-18 could be conveniently isolated by flash
chromatography in 54% and 58% yields, respectively (Scheme 4).
Finally, we also tested the reaction of BTC on encumbered
trisubstituted azetidines 22 and 23 fitted with a smaller N-methyl
substituent, compared to a N-Bn (3–4) or a N-benzhydryl (16). In
this case, reaction with BTC was effective, but we noticed an
important influence of the relative configuration at C-2/C-3 on
In a preliminary series, we selected azetidines 1–4 fitted with a
N-Bn protecting group, as substrates for reaction with BTC, since it
is known that such substituent is cleaved preferably upon reaction
with alkylchloroformates. On the other hand, the presence of the
⇑
Corresponding author. Fax: +33 (0)1 39 25 44 52.
E-mail address: couty@chimie.uvsq.fr (F. Couty).
http://dx.doi.org/10.1016/j.tetlet.2015.10.036
0040-4039/Ó 2015 Elsevier Ltd. All rights reserved.

6626
L. Menguy et al. / Tetrahedron Letters 56 (2015) 6625–6628
Path A
Ph
O
BTC, DCM
Path B
BTC
rt, 12h
ROCOCl
CN
CN
Cl
N
N
Cl
CN
or
N
N
N
N
+
+
X
O
R'
R'
R'
Cl
BTC
70%
17
18
19
21
Ph
Ph
Cl
Cl
O
O
N
R
N
R
N
N-Alkyl cleavage
Ring cleavage
(8 : 2)
Ph
X
O
BTC, DCM
rt, 12h
X = OR (Alkyl
choroformate)
X = Cl (BTC)
CN
CN
Cl
Cl
CN
80%
ent-18
20
O
Scheme 1. BTC (triphosgene) reacts with N-alkyl azetidines following two path-
ways, depending on the substituents on the azetidine ring.
(8 : 2)
Scheme 4. In 17 and 20, N-phenylethyl substituent is preferably cleaved by BTC
(triphosgene).
2
CN
BTC, DCM
0°C to rt
12h
CN
Cl
4
CN
Rationalization of all the above results is not straightforward.
Particularly striking is the difference of reactivity of 1, compared
to 17 and 20, the former leading exclusively to ring opening while
the latter favoring N-dealkylation. As matter of fact, an overview of
this reaction includes two steps that should be considered
independently: (i) N-acylation of the azetidine with phosgene
(or triphosgene initially), which is mainly governed by steric
hindrance around the nitrogen atom and explain the inertia of
encumbered substrates. In this first step, each of the four conform-
ers and invertomers of the azetidine are able to react and their
relative population is not known. Then (ii) attack of the chloride
anion at the carbon atoms adjacent to the N-acylated atom, leading
either to ring opening or dealkylation. This last step was shown
Cl
N
Cl
4
Cl
N
2
N
Bn
Bn
+
Bn
1
81%
O
O
6
5
1 : 2
CN
Ph
Ph
Ph
BTC, DCM
0°C to rt
12h
CN
4
2
CN
Cl
Bn
N
Cl
Cl
7
N
Cl
+
N
Bn
Bn
2
98%
8
O
O
4 : 1
BTC, DCM
0°C to rt
12h
Ph
Me
Me
CN
No reaction
No reaction
N
Bn
3
BTC, DCM
0°C to rt
12h
Ph
previously to follow
a
SN2 mechanism2e,g and is extremely
CN
N
Bn
4
sensitive to steric hindrance. Subtle parameters can drastically
influence the ease of this step, and the privileged cleavage of the
N-(1-phenylethyl) group in 17 and 20 compared to the N-Bn
group in 1 might reflect a better stabilization of the positive
charge partially developed in the transition state brought by the
methyl substituent.
Scheme 2. BTC (triphosgene) reacts with N-benzyl azetidines exclusively through
ring-opening.
O
Cl
BTC, 0°C, DCM
40 mn
Next, in order to implement possible uses of the produced com-
pounds, we investigated their transformation into tetrahydropy-
rimidin-2-ones and imidazolidin-2-ones, which are skeletons
existing in many synthetic biologically active compounds.11 To this
end, a mixture of carbamoyl chlorides 5 and 6 was reacted with
benzylamine to afford the corresponding mixture of regioisomeric
ureas 28 and 29, which were subjected to ring closure by
intramolecular N-alkylation of the urea nitrogen upon treatment
with NaH in THF. Pyrimidin-2-one 30 was thus obtained in 54%
overall yield. When this two-step operation was conducted with
isolated carbamoyl chloride 7, an unexpected issue was observed,
because in this case, partial b-elimination of the chloride occurred
in 31, yielding imidazolidin-2-one 33 as the major compound,
resulting from conjugate addition on the intermediate cyano
alkene 32, together with small amounts of tetrahydropyrimidin-
2-one 34, produced by nucleophilic substitution in 31. Formation
of the five-membered ring was however suppressed when an addi-
tional methyl group was introduced on the carbon backbone. Thus,
when carbamoyl chloride 24 was used as the substrate,
tetrahydropyrimidinones 36 and 37 were produced in a good
overall yields as a (6:4) epimeric mixture (Scheme 6). Relative con-
figuration in these epimers was unambiguously determined by 1H
NMR: assuming a fixed configuration for the methyl substituent at
C-6, isomers were minimized by AM1 calculations, which showed
that 4,5-cis isomer displayed a dihedral angle of 45° for H-4/H-5,
+
Cl
N
Ph
N
N
quant.
Ph
Ph
Ph
Cl
O
(7:3)
9
10
11
BTC, 0°C, DCM
40 mn
CN
No reaction
N
Ph
Ph
12
1) CH2O, Toluene,
reflux 12h
2) KCN, citric acid,
Ph
1)
, DCM,
NH
48h reflux
Ph
Ph
Ph
OH
Ph
OH
2) NaBH4, EtOH
EtOH, reflux, 3h
OH
N
NH
Me
NH2
46%
Ph
CN
Ph
81%
Me
Me
15
14
13
Ph
Ph
Ph
1) SOCl2, DCM
2) LiHMDS, THF, -78°C
BTC, DCM
rt, 12h
Me
Ph
Scheme 3. Reaction of N-benzhydryl azetidines with BTC (triphosgene).
CN
N
No reaction
16
(9:1 epimeric mixture at C-2)
Ph
the yield and selectivity: while 2,3-cis compound 22 reacted
rapidly to give good yield of isolated ring-opened regioisomer 24,
reaction with 2,3-trans isomer 23 was more sluggish and gave mix-
tures from which, compounds 25–27 were isolated (Scheme 5). It
should be mentioned that in this case, BTC behaves differently
from simple chloroformates, since the reaction of 23 with
MeOCOCl occurs uneventfully to give high yield of one regioisomer
resulting from C-2 cleavage.5b
while 4,5-trans isomer showed
a value of 180°. The large
(10.6 Hz) observed 3J H4–H5 value for H-5 in major isomer 36
and the small one (3 Hz) in minor 37 reflects an epimerization at
C-4 in basic medium.
Next was studied the functionalization of chlorocarbamoyl aze-
tidines 10, 18 and ent-18. To this end, crude mixture of 10 and 11
resulting from the reaction of 9 with BTC (Scheme 3) was directly

L. Menguy et al. / Tetrahedron Letters 56 (2015) 6625–6628
1) BTC (0.4 equiv.)
6627
Ph
Me Ph
N
BTC, DCM
rt, 2h
DCM, 0°C to rt, 1h
2) Et3N (2 equiv.), DMAP (cat.)
Cl
Cl
38: Nu : PhCH2NH : 48%
Me
Me
CN
N
Me
NuH
(1.5 equiv.), rt, 1h.
N
O
Me CN
N
71%
22
39: Nu :
: 48%
24
Me Ph
N
N
Ph
Ph
Nu
O
9
Ph
Cl
Cl
40: Nu :
: 35%
BTC, DCM
rt, 5h
1) BTC (0.4 equiv.)
DCM, 0°C to rt, 1h
2) Et3N (2 equiv.), DMAP (cat.)
Me Ph
N
Ph
NH
CN
O
Me CN
25 (34%)
Me Ph
Cl
N
Me
23
O
CN
PhCH OH
(1.5 equiv.), reflux 48h.
2
26 (10%)
Cl
or
NHCbz
Cl
N
Cl
56%
41
O
CN Me
27 (5%)
PhCH2NH2, Et3N,
DMAP (cat.)
DCM, rt, 12h.
CN
CN
18
N
N
Scheme 5. Effect of relative configuration in the azetidine ring for reaction with
BTC.
or
Bn
Bn
N
H
O
N
H
O
18
ent-
ent-42: 72%.
[α]D20 = - 114 (c 1.9, CH2Cl2)
42
: 80%.
[α]D20 = + 101 (c 2.0, CH2Cl2)
reacted with amines in order to prepare the corresponding
azetidinic urea. Indeed, 10 was difficult to purify on a small scale
and it was more convenient to delay the purification step. Reaction
occurred without ring cleavage and gave azetidinic urea 38–40
that could be conveniently purified at this stage by flash chro-
matography and were isolated in fair yields, considering the 7:3
selectivity of the first step. However, aniline was unreactive fol-
lowing these conditions, and benzyl alcohol required protracted
reaction time and heating, leading essentially to ring cleavage to
give 41. Alternatively, 18 and ent-18 gave 42 and ent-42 in good
yields (Scheme 7).
Scheme 7. Reactions of chlorocarbamoyl azetidines.
Acknowledgments
University of Versailles St-Quentin-en-Yvelines and CNRS are
acknowledged for financial support.
Supplementary data
In conclusion, we have demonstrated herein that triphosgene
can induce selective cleavages in N-alkyl azetidines, providing use-
ful building blocks for chemical diversity targeting cyclic ureas or
azetidinic ureas.
Supplementary data (typical experimental procedures and
characterization data for cyclic ureas 30, 33, 36, 37 and azetidines
18, 38, 42. Copies of 1H and 13C NMR data for these compounds)
associated with this article can be found, in the online version, at
http://dx.doi.org/10.1016/j.tetlet.2015.10.036.
References and notes
PhCH2NH2, Et3N,
DMAP (cat.)
DCM, rt, 12h.
1. For some recent examples, see: (a) Lowe, J. T.; Lee, M. D., IV; Akella, L. B.;
Davoine, E.; Donckele, E. J.; Durac, L.; Duvall, J. R.; Gerard, B.; Holson, E. B.;
Joliton, A.; Kesavan, S.; Lemercier, B. C.; Liu, H.; Marié, J-C.; Mulrooney, C. A.;
Muncipinto, G.; Welzel-O’Shea, M.; Panko, L. M.; Rowley, A.; Suh, B.-C.;
Thomas, M.; Wagner, F. F.; Wei, J.; Fowley, M. A.; Marcaurelle, L. A. J. Org. Chem.
2012, 77, 7187–7211; (b) Pizzonero, M.; Dupont, S.; Babel, M.; Beaumont, S.;
Bienvenu, N.; Blanqué, R.; Cherel, L.; Christophe, T.; Crescenzi, B.; De Lemos, E.;
Delerive, P.; Deprez, P.; De Vos, S.; Djata, F.; Fletcher, S.; Kopiejewski, S.;
L’Ebraly, C.; Lefrançois, J.-M.; Lavazais, S.; Manioc, M.; Nelles, L.; Oste, L.;
Polancec, D.; Qhénéhen, V.; Soulas, F.; Triballeau, N.; van der Aar, E.;
Vandeghinste, N.; Wakselman, E.; Brys, R.; Saniere, L. J. Med. Chem. 2014, 57,
10044–10057; (c) Nicholls, D. J.; Wiley, K.; Dainty, I.; MacIntosh, F.; Philips, C.;
Gaw, A.; Mards, C. K. J. Pharmacol. Exp. Ther. 2015, 353, 340; (d) Rzasa, R. M.;
Frohn, M. J.; Andrews, K. L.; Chmait, S.; Chen, N.; Clarine, J. G.; Davis, C.;
Eastwood, H. A.; Horne, D. B.; Hu, E.; Jones, A. E.; Kaller, M. R.; Kunz, R. K.;
Miller, S.; Monenschein, H.; Nguyen, T.; Pickrell, A. J.; Porter, A.; Reichelt, A.;
Zhao, X.; Treanor, J. J. S.; Allen, J. R. Bioorg. Med. Chem. Lett. 2014, 22, 6570–
6585; (e) Hart, T.; Macias, A. T.; Benwell, K.; Brooks, T.; D’Alessandro, J.;
Dokurno, P.; Francis, G.; Gibbons, B.; Haymes, T.; Kennett, G.; Lightowler, S.;
Mansell, H.; Matassova, N.; Misra, A.; Padfield, A.; Parsons, R.; Pratt, R.;
Robertson, A.; Walls, S.; Wong, M.; Roughley, S. Bioorg. Med. Chem. Lett. 2009,
19, 4241–4244.
CN
N NHBn
CN
+
Cl
Bn
BnHN
28
N
Cl
Bn
5
6
+
29
O
O
2 : 1
CN
Bn
NaH, THF, 50°C, 48h
N
N
Bn
30
: 54% overall
O
PhCH2NH2, Et3N,
DMAP (cat.)
Bn Ph
N
DCM, rt, 12h.
BnHN
BnHN
Cl
7
O
O
CN
70%
Ph
31
NaH, THF,
50°C, 48h
Ph
O
Bn Ph
N
CN
Bn
+
CN
N
N
Bn
N
N
Bn
Bn
CN
32
O
2. For recent reviews on azetidines, including their preparations, see: (a)
Dejaegher, Y.; Kuz’menok, N. M.; Zvonok, A. M.; De Kimpe, N. Chem. Rev.
2002, 102, 29–60; (b) Couty, F.; Evano, G.; Prim, D. Mini-Rev. Org. Chem. 2004, 1,
133–148; (c) Brandi, A.; Cicchi, S.; Cordero, F. M. Chem. Rev. 2008, 108, 3988–
4035; (d) Couty, F. Synthesis of Azetidines In Science of Synthesis: Houben-Weyl
Methods of Molecular Transformations; Enders, D., Ed.; Georg Thieme: New York,
2009; Vol. 40a, pp 773–817; (e) Couty, F.; Evano, G. Synlett 2009, 3053–3064;
(f) Bott, T. M.; West, F. G. Heterocycles 2012, 84, 223–264; (g) Couty, F.;
Drouillat, B.; Evano, G.; David, O. Eur. J. Org. Chem. 2013, 2045–2056.
3. (a) Ghorai, M. K.; Shukla, D.; Das, K. J. Org. Chem. 2009, 74, 7013–7022; (b)
Dwidedi, S. K.; Gandhi, S.; Rastogi, N.; Singh, V. K. Tetrahedron Lett. 2007, 48,
5375–5377.
4. The N-Bn to N-Boc interconversion through hydrogenolysis appears to be most
reported procedure, but can lead to ring cleavage and is not compatible with
the presence of some functional groups. See for examples: (a) Brauner-
Osborne, H.; Bunch, L.; Chopin, N.; Couty, F.; Evano, G.; Jensen, A. A.; Kusk, M.;
Nielsen, B.; Rabasso, N. Org. Biomol. Chem. 2005, 3, 3926–3936; (b)
33
(60%)
34 (6%)
PhCH2NH2, Et3N,
DMAP (cat.)
DCM, rt, 12h.
Me Ph
N
BnHN
Cl
24
O
Me CN
35
Ph
94%
NaH, THF,
50°C, 48h
72%
Ph
Me
CN
Bn
Me
CN
4
Bn
5
+
4
5
6
6
N
N
N
N
Me
Me
O
37
O
36
Scheme 6. Transformation of ring opened products into cyclic urea.

6628
L. Menguy et al. / Tetrahedron Letters 56 (2015) 6625–6628
Sivaprakasam, M.; Hansen, K. B.; David, O.; Nielsen, B.; Traynelis, S. T.; Clausen,
8. For an example of 4-aryl-N-benzhydryl azetidine reacting with phosgene to
give a precursor of relevant pharmaceutical, see: Adams, D. R.; Bentley, J.;
Bodkin, C. D.; Cliffe, I. A.; Davidson, J. E. P.; Mansell, L.; Monck, N. J.; Shepherd,
R. G.; Shepherd, J. M.; US Pat. 6,831,078, 2004; See also: Parmar, D.; Henkel, L.;
Dib, J.; Rueping, M. Chem. Commun. 2015, 2111–2113.
9. For previous preparation of 10 from azetidine and phosgene (6% yield), see:
Alekperov, R. K.; Leshchinskaya, V. P.; Nosova, V. S.; Chervin, I. I.; Kostyanovskii,
R. G. Chem. Heterocycl. Compd. (NY New-York, United States) 1987, 23, 749–751.
10. This scheme follows our previously reported strategy for the preparation of 2-
cyano azetidines. See: Agami, C.; Couty, F.; Evano, G. Tetrahedron: Asymmetry
2002, 13, 297–302.
R. P.; Couty, F.; Bunch, L. Chem. Med. Chem. 2009, 4, 110–117; (c) Drouillat, K. B.;
Wright, K.; Marrot, J.; Couty, F. Tetrahedron: Asymmetry 2012, 23, 690–696; (d)
Couty, F.; Evano, G.; Vargas-Sanchez, M.; Bouzas, G. J. Org. Chem. 2005, 70,
9028–9031; (e) Kovács, E.; Faigl, F.; Mucsi, Z.; Nyerges, M.; Hegedüs, L. J. Mol.
Catal. A: Chem. 2014, 305, 217–224.
5. (a) Kenis, S.; D’hooghe, M.; Verniest, G.; Dang Thi, T. A.; The, C. P.; Nguyen, T. V.;
De Kimpe, N. J. Org. Chem. 2012, 77, 5982–5992; (b) Vargas-Sanchez, M.;
Lakhdar, S.; Couty, F.; Evano, G. Org. Lett. 2006, 8, 5501–5504; (c) Ma, S.-H.;
Yoon, D. H.; Ha, H.-J.; Lee, W. K. Tetrahedron Lett. 2007, 48, 269–271; (d) Barrett,
A. G. M.; Dozzo, P.; White, A. J. P.; Williams, D. Tetrahedron 2002, 58,
7303–7313.
11. For tetrahydropyrimidin-2-ones, see Zhang, L. L.; Sun, J.; Yan, C.-G. Mol. Divers.
6. BTC generates three equivalents of phosgene ClCOCl upon nucleophilic attack
of a nucleophile, through decomposition of the produced trichloromethoxide
anion. It was therefore used herein as a substoichiometric 40% molar ratio with
respect to the starting azetidines.
7. For an example of functionalized azetidines reacting with BTC through ring
expansion, see: Couty, F.; Drouillat, B.; Lemée, F. Eur. J. Org. Chem. 2011,
794–801.
2014, 18, 79–89. and references cited therein; For a recent example of
imidazodin-2-ones, see: Sun, S.; Zhang, Z.; Kodumuru, V.; Pokrovskaia, N.;
Fonarev, J.; Jia, Q.; Leung, P.-Y.; Tran, J.; Ratkai, L. G.; McLaren, D. G.; Radomski,
C.; Chowdhury, S.; Fu, J.; Hubbard, B.; Winther, M. D.; Dales, N. A. Bioorg. Med.
Chem. Lett. 2014, 24, 520–525.