An efficient asymmetric synthesis of azetidine 2-phosphonic acids

Article abstract of DOI:10.1016/S0040-4039(02)00868-7

Substituted azetidinic 2-phosphonates were prepared in diastereoisomerically and enantiomerically pure form, starting from readily available β-amino alcohols. This synthesis involved a three-step sequence: (i) N-alkylation of the starting amino alcohol wi

Full text of DOI:10.1016/S0040-4039(02)00868-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 4633–4636
An efficient asymmetric synthesis of azetidine 2-phosphonic acids
Claude Agami,b Franc¸ois Coutya,* and Nicolas Rabassob
aSIRCOB, UPRESA CNRS 8086, Universite´ de Versailles, 45, avenue des Etats-Unis, 78035 Versailles Ce´dex, France
bLaboratoire de Synthe`se Asyme´trique, UMR 7611, Universite´ Pierre et Marie Curie, 4 place Jussieu, 75005 Paris, France
Received 11 April 2002; revised 2 May 2002; accepted 7 May 2002
Abstract—Substituted azetidinic 2-phosphonates were prepared in diastereoisomerically and enantiomerically pure form, starting
from readily available b-amino alcohols. This synthesis involved a three-step sequence: (i) N-alkylation of the starting amino
alcohol with a methylene phosphonate moiety, (ii) chlorination of the alcohol, and (iii) stereoselective 4-exo-tet ring closure
through an intramolecular alkylation of the lithiated aminophosphonate. © 2002 Elsevier Science Ltd. All rights reserved.
a-Aminophosphonic acids are considered as mimics of
the corresponding a-aminocarboxylic acid.1 This resem-
blance explains the large range of biological activities
displayed by the members of this class of compounds
and the applications they have found in agriculture and
medicine.2 Consequently, development of efficient
methodologies for the asymmetric synthesis of a-
aminophosphonic acids is an active area of research
and many methods are now available that provide
acyclic a-aminophosphonic acids,3 as well as phospho-
nic surrogates of proline4 and pipecolic acid derivatives5
in enantiomerically pure form. However, to our knowl-
edge, the asymmetric synthesis of 4-membered azeti-
dinic a-phosphonic acids has not been reported so far.
Nevertheless, it is interesting to note that a synthesis of
racemic 1, the phosphonic analogue of azetidine 2-car-
boxylic acid 2 was published very recently.6
The required starting aminophosphonates 4 were pre-
pared via a two-step sequence involving: (i) oxazo-
lidine-formation in the presence of formaldehyde, and
(ii) acid-catalysed ring opening of the oxazolidine fol-
lowed by nucleophilic addition of triethyl phosphite.8
Using this sequence, aminophosphonates 9, 12, 15 and
18 were prepared in good overall yields from, (S)-N-
benzyl phenylglycinol 7, (S)-N-benzyl phenylalaninol
10, (1R,2S)-ephedrine 13 and (1R,2R)-pseudoephedrine
16, respectively (Scheme 2).
Chlorination of these amino phosphonates alcohols was
effected by treatment with thionyl chloride in refluxing
dichloromethane. As previously reported in the case of
related amino nitriles,7 compound 9 having a benzylic
amine gave rearranged chloride 19 in good yield. On
the other hand, ephedrin family-derived phosphonates
15 and 18 gave chlorides 20 and 21. These compounds
were produced with total retention of configuration at
the hydroxylated stereocenter (Scheme 3). This was
proved indirectly through determination of the relative
configuration of the azetidines resulting from the cycli-
sation of these chlorides (vide infra).
We describe in this Letter a straightforward prepara-
tion of enantiomerically pure azetidinic 2-phosphonic
acids starting from readily available b-amino alcohols.
The key step of this synthesis is based on a 4-exo-tet
anionic alkylation of a lithiated a-aminophosphonate
(5: X=PO(OEt)2, Scheme 1); this strategy was success-
fully developed in our group for the preparation of
2-cyano azetidines (6, X=CN in Scheme 1).7
As a matter of fact, this high stereoselectivity can be
explained by the participation of the amine in the
* Corresponding author. Tel.: +33(1)39254455; fax: +33(1)39254452;
e-mail: couty@chimie.uvsq.fr
Scheme 1.
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)00868-7

4634
C. Agami et al. / Tetrahedron Letters 43 (2002) 4633–4636
R3
R2
R3
R2
R3
R2
OH
O
N
R1
b
OH
a
O
P
OEt
NH
R1
N
R1
OEt
overall yields (%)
R1= Bn; R2=Ph
R1= Bn; R2=Bn
R1= Me; R2=Me
R1= Me; R2=Me
;R3=H
; R3=H
;R3=Ph
;R3=Ph
7
8
9
68
68
74
78
10
13
16
11
14
17
12
15
18
Scheme 2. Reagents and conditions: (a) formaldehyde, toluene, reflux; (b) diethylphosphite, citric acid, EtOH, reflux.
1) SOCl2
DCM
2) NaHCO3
nes were assigned on the basis of NMR studies. For
Ph
example, H2, in compound 23 appears as a doublet of
OH
Cl
O
doublet at 3.23 ppm, with 2JH2–H3=8.6 Hz and 2JH2–P
=
P
O
OEt
6.2 Hz. The former coupling constant is in agreement
with a trans relationship between the subtituents at C2
and C3, as checked after AM1 minimisation of the
conformation of 23. Particularly relevant is the long
N
N
P
92 %
Ph
OEt
OEt
Bn
Bn
OEt
19
9
4
range JC–P coupling constant of 5.9 Hz between the
methyl substituent at C4 and phosphorous atom in
compound 23. This coupling does not appear in dia-
steroisomer 24 and this observation set up unambigu-
ously a cis relationship between the C2 and C4 sub-
stituents (W pattern between carbon and phosphorous)
in 23.9 The yield of this reaction seems to be influenced
by the steric demand in the transition state: phospho-
nate anions are known to have a sp2 structure with a
tetrahedral phosphorous,10 the important size of the
tetrahedral lithiated phosphorous moiety induces a
severe steric interaction with the cis substituent of the
azetidine ring in formation. This can explain the low
yield obtained with substrate 20, in which an interac-
tion between the lithiated phosphoenolate and the C4
methyl substituent is operating during the ring closure
(Scheme 5).
Ph
Ph
OH
Cl
O
Id
P
O
OEt
OEt
N
N
P
71 %
Me
Me
OEt
OEt
Me
Me
21
18
Ph
Ph
OH
Cl
O
Id
P
O
OEt
OEt
N
N
P
70 %
Me
Me
OEt
OEt
Me
Me
20
15
Scheme 3.
The bulkiness of the lithiated phosphoenolate com-
pared to an ester enolate and hence its beneficial effect
with regard to 2,3-trans stereoselectivity during the ring
closure is also put in light in the experiment described
in Scheme 6. Indeed, when chloro ester 26, prepared
through alkylation of (S)-N-benzyl phenylglycinol 7
with ethyl bromoacetate followed by treatment with
thionyl chloride, is deprotonated with LiHMDS in the
presence of HMPA, the intramolecular alkylation now
gives a mixture of cis and trans 2-carboethoxy azetidi-
nes 27 and 28 (de: 36%). Clearly, the steric interaction
between the sp2 ester enolate and the a phenyl sub-
stituent is less severe in this case.
substitution process through an aziridinium ion and the
configuration in compound 19 (although not proven
unequivocally) was attributed on the basis of this
behaviour
On the other hand, chlorination of amino phosphonate
12 derived from phenylalaninol was disappointing since
in this case, an unseparable mixture of two compounds
that could not be separated and characterized was
obtained. This result is in contrast with the chlorination
of the related amino nitrile7 that cleanly gave the
corresponding primary chloride.
The enantiomeric purity of compound 22 had to be
checked, since the rearrangement occurring during the
transformation 9“19 might have proceeded with some
extent of racemisation. To this end, ent-22 was pre-
pared from (R)-N-benzyl phenylglycinol ent-9. Exami-
nation of the 31P NMR spectra of a mixture of
enantiomers 22 in the presence of (S)-a-(tri-
fluoromethyl)benzyl alcohol11 showed an efficient sepa-
ration of the 31P signals for each enantiomers. This
method allowed an accurate determination of the enan-
The ring closure of chlorides 19–21 was next achieved
by treatment with LiHMDS in THF at −78 to 0°C. In
all cases, only the 1,3-trans isomer 22–24 could be
isolated in this reaction, but the yield was highly depen-
dent on the substitution pattern of the substrate
(Scheme 4).
An SN2 mechanism is probably operating during this
ring closure. The relative configurations in the azetidi-

C. Agami et al. / Tetrahedron Letters 43 (2002) 4633–4636
4635
purification by ion-exchange chromatography (Dowex
1X2), to give phosphonic acids 29–31 in, respectively,
86, 82 and 80% yields. Alternatively, aminophospho-
nate 22 could be N-debenzylated by hydrogenolysis to
give amine 32. When this hydrogenolysis was con-
ducted in the presence of (Boc)2O, N-Boc protected
azetidine 33 was obtained (Scheme 7).
O
Ph
OEt
OEt
Cl
Ph
P
LiHMDS, -78 to 0°C
75%
O
N
P
N
OEt
Bn
Bn
19
OEt
22
O
Ph
OEt
OEt
Cl
Ph
Ph
P
Id.
O
P
O
O
O
OH
OH
OH
N
P
Ph
P
N
Ph
P
Ph
Me
30%
OEt
OEt
OH
OH
Me
OH
Me
20
N
N
N
23
O
P
Me
Bn
Me
Ph
OEt
OEt
Cl
29
30
31
Id.
O
O
P
O
N
P
N
Me
OEt
OEt
50%
OEt
OEt
OEt
OEt
Ph
Ph
P
Me
Me
a
b
22
21
24
N
N
H
Boc
Scheme 4.
32
33
Scheme 7. Reagents and conditions: (a) EtOH, HCl, cat.
Pd(OH)2, H2, 40%. (b) (Boc)2O, AcOEt, H2, cat. Pd/C, 60%.
Cl
In conclusion, we have reported the first asymmetric
synthesis of azetidinic 2-phosphonic acids and
derivatives12 starting from readily available b-amino
alcohols. Further studies are in progress in our group in
order to delineate the scope of this cyclisation with
other phosphorous-containing moieties.
H
Me
N
Me
P
LiO
OEt
EtO
Scheme 5.
Acknowledgements
Ph
OH
Cl
a
b
O
The authors wish to thank Professor J. C. Cherton and
Dr. M. J. Pouet for their help in the determination of
enantiomeric purity of compound 22.
O
7
N
N
Ph
OEt
Bn
Bn
26
OEt
25
References
Ph
COOEt Ph
COOEt
Bn
c
+
1. Frank, A. W. Crit. Rev. Biochem. 1984, 16, 51.
2. See inter alia, and references cited therein (a) Ryglowski,
A.; Kafarsky, P. Tetrahedron 1996, 52, 10685–10692; (b)
Chung, S. K.; Kang, D. H. Tetrahedron: Asymmetry
1996, 7, 21–24; (c) Fazio, A.; Loreto, M. A.; Tardella, P.
A. Tetrahedron Lett. 2001, 42, 2185–2187; (d) Castelot-
Deliencourt, G.; Roger, E.; Pannecoucke, X.; Quirion, J.
C. Eur. J. Org. Chem. 2001, 3031–3038; (e) Smith, A. B.,
III; Yager, K.; Taylor, C. M. J. Am. Chem. Soc. 1995,
117, 10879–10888.
N
N
Bn
28
27
Scheme 6. Reagents and conditions: (a) ethyl bromoacetate,
NaHCO3, NaI, DMF, 86%; (b) SOCl2, DCM, 58%; (c) LiH-
MDS, THF/HMPA, −78°C, 48%.
tiomeric excesses of 22 and ent-22 and these com-
pounds proved to be enantiomerically pure, within the
precision of NMR (300 MHz).
3. (a) Maury, C.; Gharbaoui, T.; Royer, J.; Husson, J. P. J.
Org. Chem. 1996, 61, 3687–3693; (b) Ref. 2e.
4. (a) Katritzky, A. R.; Cui, X. L.; Yang, B.; Steel, P. J. J.
Org. Chem. 1999, 64, 1979–1985; (b) Groth, U.; Richter,
L.; Schollkopf, U. Tetrahedron 1992, 48, 117–122; (c)
Belyaev, A.; Borloo, M.; Augustyns, K.; Lambeir, A. M.;
De Meester, I. Tetrahedron Lett. 1995, 36, 3755–3758.
Further chemical transformations of azetidinic phos-
phonate esters 22–24 were also studied. Hydrolysis of
the phosphonate moiety was effected by treatment with
bromotrimethylsilane in acetonitrile,6 followed by

4636
C. Agami et al. / Tetrahedron Letters 43 (2002) 4633–4636
5. (a) Maury, C.; Wang, Q.; Gharbaoui, T.; Chiadmi, M.;
57.42; H, 8.69; N, 4.07. Compound 19: Rf: 0.5 (diethyl
ether), [h]2D0 −33.5 (c 2.5, CHCl3). IR (film) 3063, 3032,
2980, 2919, 2843, 1674, 1598, 1230 (PꢀO), 1050 (PꢁOꢁC),
728. 1H NMR (250 MHz, CDCl3): 1.17 and 1.19 (2 t, 6H,
J=7.1 Hz), 2.86 (d, 2H, J=9.6 Hz), 3.21 (dd, 1H, J=1.2
and 13.3 Hz), 3.80 (dd, J=1.5 and 13.6 Hz), 3.85–4.04
(m, 4H), 4.92 (t, 1H, J=7.1 Hz), 7.11–7.26 (m, 10H); 13C
NMR: 16.5 (d, J=6.3 Hz), 49.4 (d, J=154.7 Hz), 60.2 (d,
J=3.9 Hz), 61.1, 61.7 (t, J=3.9 Hz), 62.8 (d, J=3.9 Hz),
127.3, 127.6, 128.2, 128.4, 129.0, 138.2, 140.1; Anal.
Calcd for C20H27ClNO3P: C, 60.68; H, 6.87; N, 3.54.
Found: C, 60.57; H, 6.98; N, 3.44. Compound 23: Rf:
0.34 (ethyl acetate/ethanol: 98/2), [h]2D0 −62.9 (c 1.3,
CHCl3). IR (film) 3060, 3028, 2978, 1603, 1497, 1453,
1242 (PꢀO), 1026 (PꢁOꢁC);. 1H NMR (250 MHz,
CDCl3): 0.74 (d, 3H, J=6.5 Hz), 1.20 (t, 3H, J=7.1 Hz),
1.25 (t, 3H, J=7.1 Hz), 2.35 (s, 3H), 3.86 (dd, 1H, J=6.2
and 5.8 Hz), 4.00–4.16 (m, 6H), 7.12–7.28 (m, 5H); 13C
NMR: 12.1, 16.7 (d, J=5.9 Hz), 37.3 (d, J=5.9 Hz), 41.3
(d, J=3.9 Hz), 61.7 (d, J=5.9), 62.2 (d, J=5.9 Hz), 62.5
(t, J=135.9 Hz), 63.6 (d, J=45.3 Hz), 126.7, 128.2, 128.3,
137.4; 31P NMR (Ref: H3PO4): 26.2; anal. calcd for
C15H24NO3P: C, 60.59; H, 8.14; N, 4.71. Found: C,
60.43; H, 8.31; N, 4.60. Compound 29: Rf: 0.22 (ethanol/
30% NH4OH/H2O: 30/10/3), Mp: 223–229°C; [h]2D0 −2.5
(c 0.6, 1 M NaOH). 1H NMR (250 MHz, CD3OD):
3.89–3.99 (m, 1H), 4.09–4.22 (m, 1H), 4.24–4.34 (m, 1H),
4.30 (d, 1H, J=12.1 Hz), 4.34–4.44 (m, 1H); 4.55 (d, 1H,
J=13.2 Hz), 7.17–7.35 (m, 5H), 7.46–7.56 (m, 2H), 7.65–
7.73 (m, 2H); 13C NMR: 39.0, 56.2 (d, J=11.8 Hz), 58.3,
71.4 (d, J=137.8 Hz), 128.0, 128.4, 129.7, 130.3, 130.7,
130.9, 140.1. HRMS: (MH+) calcd 304 1103. Obs.
304 1107.
Tomas, A.; Royer, J.; Husson, H. P. Tetrahedron 1997,
53, 3627–3636; (b) Jacquier, R.; Ouazzani, F.; Roumes-
tant, M. L.; Viallefont, P. Phosphorous Sulfur 1988, 36,
73–78; (c) Katrizky, A. H.; Qiu, G.; Yang, B.; Peter, J. J.
Org. Chem. 1998, 63, 6699–6703.
6. Otmar, M.; Pola´kova´, L.; Masoj´ıdkova´, M.; Holy´, A.
Collect. Czech. Chem. Commun. 2001, 66, 507–516.
7. Agami, C.; Couty, F.; Evano, G. Tetrahedron: Asymme-
try 2002, 13, 297–302.
8. Nifant’ev, E. E.; Runova, L. M.; Shestakova, T. G.;
Bogatyreva, E. V.; Ronkov, V. I. J. Gen. Chem. USSR
1980, 50, 243–247.
9. Hanrahan, J. R.; Taylor, P. C.; Errington, W. J. Chem.
Soc., Perkin Trans. 1 1997, 493–502.
10. Denmark, S. E.; Dorow, R. L. J. Am. Chem. Soc. 1990,
112, 864–866.
11. Pirkle, W. H.; Hoekstra, M. S. J. Am. Chem. Soc. 1976,
98, 1832–1839.
1
12. All new compounds gave H, 13C and 31P NMR data in
accordance with their structure. Selective data: Com-
pound 15: Rf: 0.2 (diethyl ether/petroleum ether), [h]D20
+48 (c 0.5, CHCl3). IR (film) 3380, 3063, 2991, 2925,
2873, 2786, 1496, 1455, 1214 (PꢀO), 1158 (PꢁOꢁC). 1H
NMR (250 MHz, CDCl3): 0.8 (d, 3H, J=6.8 Hz), 1.19
and 1.20 (2 t, 6H, J=7.1 Hz), 2.35 (s, 3H), 2.72–2.81 (m,
1H), 2.73 (dd, 1H, J=9.8 and 15.8 Hz), 2.98 (dd, 1H,
J=11 and 15.8 Hz), 3.90–4.02 (m, 4H), 4.16 (bs, 1H),
4.77 (d, 1H, J=3.8 Hz), 7.10–7.27 (m, 5H); 13C NMR:
8.3, 16.3 (d, J=5.7 Hz), 41.8, 49.5 (d, J=167.3 Hz), 61.9
(t, J=7.4 Hz), 65.4 (d, J=12 Hz), 74.3, 125.9, 126.6,
127.7, 142.9; 31P NMR (Ref: H3PO4): 27.9; anal. calcd for
C15H26NO4P: C, 57.13; H, 8.31; N, 4.44. Found: C,