Synthesis of 4-phosphono-β-lactams via phosphite addition to acyliminium salts

Article abstract of DOI:10.1016/S0040-4039(03)00005-4

4-Aryl-4-phosphono-β-lactams are prepared by acylation of iminium salts with chloroacetyl chloride followed by phosphite addition and ring closure using sodium hydride as a base. Deacylation of the iminium salt is in competition with the desired addition of phosphites to acyliminium salts, which lowers the yield of the reaction.

Full text of DOI:10.1016/S0040-4039(03)00005-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 1619–1622
Synthesis of 4-phosphono-b-lactams via phosphite addition
to acyliminium salts
Christian V. Stevens,* Wannes Vekemans, Kristof Moonen and Thomas Rammeloo
Department of Organic Chemistry, Faculty of Agricultural and Applied Biological Sciences, Ghent University,
Coupure Links 653, B-9000 Gent, Belgium
Received 21 November 2002; revised 20 December 2002; accepted 23 December 2002
Abstract—4-Aryl-4-phosphono-b-lactams are prepared by acylation of iminium salts with chloroacetyl chloride followed by
phosphite addition and ring closure using sodium hydride as a base. Deacylation of the iminium salt is in competition with the
desired addition of phosphites to acyliminium salts, which lowers the yield of the reaction. © 2003 Elsevier Science Ltd. All rights
reserved.
a-Aminophosphonic acids have received a lot of atten-
tion since they mimic transition states of peptide
hydrolysis and thus affect several physiological activi-
Several phosphonylated penicillin-like derivatives have
also been evaluated in this area, although these com-
pounds did not possess promising antibacterial proper-
1
13–15
ties of the cell. Therefore, aminophosphonic acids are
ties.
A classical method for the synthesis of
often active as enzyme inhibitors, such as the a-
4-phosphono-b-lactams is via the Arbuzov reaction of
2
2,15
aminophosphonic acid alaphosphin, which inhibits
4-acetoxy-b-lactam and a trialkylphosphite.
An
alanine racemase. Also inhibition of thrombin by syn-
asymmetric synthesis of azetidine 2-phosphonic acids,
3
16
thetic aminophosphonates has been reported. Others
starting from b-amino alcohols, was reported recently.
are active as herbicides (e.g. glyphosate) or as
This synthesis involved an N-alkylation with a methyl-
ene phosphonate moiety, chlorination of the alcohol
and stereoselective ring closure with LiHMDS. Due to
this very recent publication, the preliminary publication
of our work in this area was initiated.
4
,5
neurotransmitters
(e.g. (R)-CPP and CGS 19755).
Aminodiphosphonates on the other hand, can be used
for the treatment of disorders of calcium metabolism.
In view of the wide range of their physiological activi-
ties, efforts have been made by several research groups
to develop efficient strategies for the synthesis of
aminophosphonate derivatives with a diverse substitu-
tion pattern.
In order to develop a new strategy to synthesise 4-aryl
phosphono-b-lactams, aromatic aldehydes were con-
densed with primary amines in the presence of magne-
sium sulfate leading to the corresponding imines in
good yields (82–99%).
In our efforts to evaluate heterocyclic aminophospho-
6–8
nates for agricultural applications, the phosphono-b-
lactams seemed an interesting class of compounds.
Further, the b-lactams still hold a lot of attention due
to the importance of b-lactams in the field of elastase
In order to make the precursors 5 of the phosphono-b-
lactams, two methodologies can be applied. The first
one consists of the addition of a phosphite to the imine
followed by acylation of the corresponding aminophos-
phonate. A second option involves the acylation of the
imine with the formation of an iminium salt, followed
by the addition of a trialkyl phosphite.
9
10
inhibitors and monobactam antibiotics. Since the
discovery of penicillin, numerous penicillin derivatives
have been prepared and examined for antibacterial
activity. A variety of new b-lactam containing ring
systems have been reported to show antibacterial prop-
erties,
including
the
penams,
carbapenems,
Evaluating the first approach, the addition of dialkyl
phosphites, activated by silylation (trimethylsilyl chlo-
ride) or by deprotonation (sodium hydride), to imines,
proved to be a very slow reaction. In the cases of quite
stable imines, the reaction could be performed either at
1
1,12
cephalosporins, clavulanic acids and oxapenams.
*
0
Corresponding author. Tel.: +32-9-264 59 57; fax: +32-9-264 62 43;
e-mail: chris.stevens@rug.ac.be
1
reflux (Ar=Ph; R =Bn, reflux 4–12 h; 69%) or at room
040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(03)00005-4

1
620
C. V. Ste6ens et al. / Tetrahedron Letters 44 (2003) 1619–1622
temperature over 2 weeks (88% yield). However, using
sensitive imines (e.g. Ar=pyridyl; R =Bn; 2 weeks; rt;
phosphoryl phosphonate 11 formed from chloroacetyl
1
chloride. The mechanism for this reaction was
18
99% yield), the reaction had to be performed at room
described by Britelli and involves a Perkow reaction
temperature leading to very long reaction times, mak-
ing this approach less practical. Furthermore, synthesis-
ing the aminophosphonate first also resulted in
problems in the acylation afterwards. The low reactivity
results from the low nucleophilicity of the nitrogen
followed by a Michaelis–Arbuzov reaction (Scheme 2).
In order to prove this hypothesis, an experiment was
done with chloroacetyl chloride and trimethyl phos-
31
phite. Comparison of the P NMR spectrum obtained
with those from the side product confirmed our pre-
sumption. Mechanistically, two possibilities for the for-
mation of the phosphonate 11 had to be considered.
The acylation reaction was not completed and the
phosphite reacted with the excess of acid chloride, or
the phosphite is able to abstract the acyl group from
the iminium salt. To investigate this last case, a very
small amount of the iminium salt was isolated free of
chloroacetyl chloride under an N2 atmosphere and
reacted with trimethyl phosphite. Again, phosphonate
11 was found in the reaction mixture. This indicates
that the formation of the side product is not due to the
presence of an excess of acid chloride in the reaction
mixture. Therefore, it will be necessary to alter the
reaction conditions in order to minimize the abstraction
of the acyl group and to maximize the addition to the
iminium salt in further investigations.
1
atom (e.g. Ar=pyridyl; R =naphthyl) because of the
strong electron withdrawing effect of the phosphonate
group. Therefore, we became interested in the evalua-
tion of a new approach trying to use the addition of a
phosphorus nucleophile onto an acylated iminium salt
(
Scheme 1).
First, attempts were made to isolate the iminium salts
after treatment of the aromatic imines with
2
chloroacetyl chloride. Unfortunately, due to the very
hygroscopic character of these iminium salts 3, they
could hardly be isolated and could not be characterised
properly. Therefore, after formation of the salts in
toluene at −40°C, which can be followed visually by
precipitation of the salts, the reaction mixture was
stirred for one additional hour, followed by addition of
the trialkyl phosphite. The best results were obtained
using trimethyl phosphite. Triethyl phosphite can be
used as well, however resulting in a lower yield. Triiso-
propyl phosphite proves to be too sterically hindered
for the addition. The aim of the reaction is the addition
of the phosphite to the activated iminium salt, generat-
ing a phosphonium salt which is then dealkylated by
The subsequent ring closure reactions were performed
with NaH in THF and the b-lactams were obtained in
good yields (71–99%). The formation of the b-lactam
ring via a C ꢀC bond is not exploited in detail, since
3
4
1
9,20
most lactams are prepared by a [2+2] cycloaddition.
17
the liberated chloride (Arbuzov type dealkylation).
However, the cycloaddition approach is not applicable
for the phosphono-b-lactams. These promising results
in the ring closure stimulate us to spend more time on
the optimisation of the addition reaction of phosphite
to the iminium salts.
Using trimethyl phosphite, methyl chloride is formed in
the reaction and bubbles out of the reaction mixture
(
CAUTION: the reaction should be performed in a
properly working hood), so that theoretically only the
desired aminophosphonate 5 is present in the mixture.
A further remarkable observation that still needs some
computational background, is the ring closure of
dimethyl 1-[benzyl(chloroacetyl)amino]-3-phenyl-2-pro-
penyl phosphonate 5a. Using sodium hydride as a base,
the highly stabilised anion 12 formed, ring closes exclu-
sively to the four membered ring 6a instead of to the
expected less strained six-membered ring 13 (Scheme 3).
However, a competing reaction complicated the addi-
tion and made the yields drop to 24–56% due to the
formation of some side products, which could not be
3
1
isolated in pure form. Examination of the P NMR
spectra of several compounds, revealed that two dou-
blets are always present. These were attributed to the
Scheme 1.

C. V. Ste6ens et al. / Tetrahedron Letters 44 (2003) 1619–1622
1621
Scheme 2.
Scheme 3.
21
This is in accordance with some earlier observations
and will be looked at in more detail.
6. Stevens, C. V.; Verbeke, A.; De Kimpe, N. Synlett 1998,
180–182.
7. Stevens, C. V.; Gallant, M.; De Kimpe, N. Tetrahedron
In conclusion, a straightforward synthesis of 4-phos-
phono-b-lactams has been developed which leads to
interesting building blocks for the synthesis of functiona-
lised amino phosphonates. The optimization of the
addition reaction using other types of imines and the
biological testing of the new compounds are currently
under investigation.
Lett. 1999, 40, 3457–3460.
8. Bellesia, F.; De Buyck, L.; Colucci, M. V.; Ghelfi, F.;
Laureyn, I.; Libertini, E.; Mucci, A.; Pagnoni, U. M.;
Pinetti, A.; Rogge, T. M.; Stevens, C. V. Tetrahedron
Lett. 2001, 42, 4573–4575.
9. G e´ rard, S.; Dive, G.; Clamot, B.; Touillaux, R.;
Marchand-Brynaert, J. Tetrahedron 2002, 58, 2423–2433.
0. Ogilvie, W. W.; Yoakim, C.; D oˆ , F.; Hach e´ , B.; Lagac e´ ,
L.; Naud, J.; O’Meara, J. A.; D e´ ziel, R. Bioorg. Med.
Chem. 1999, 7, 1521–1531.
1. Ren, X.-F.; Konaklieva, M. I.; Shi, H.; Dickey, S.; Lim,
D. V.; Gonzalez, J.; Turos, E. J. Org. Chem. 1998, 63,
8898–8917.
12. D u¨ rckheimer, W.; Blumbach, J.; Lattrell, R.; Scheune-
mann, K. H. Angew. Chem. 1985, 97, 183–205.
13. Koster, W. H.; Zahler, R.; Chang, H. W.; Cimarusti, C.
M.; Jacobs, G. A.; Perri, M. J. Am. Chem. Soc. 1983,
105, 3743–3745.
1
1
References
1
2
3
4
5
. Kafarski, P.; Lejczak, B. Phosphorous Sulfur Silicon 1991,
6
3, 193–215.
. Campbell, M. M.; Carruthers, N. J. Chem. Soc., Chem.
Commun. 1980, 15, 730–731.
. Wang, C.-L. J.; Taylor, T. L.; Mical, A. J.; Spitz, S.;
Reilly, T. M. Tetrahedron Lett. 1992, 33, 7667–7670.
. Falorni, M.; Porcheddu, A.; Giacomello, G. Tetrahedron:
Asymmetry 1997, 8, 1633–1639.
14. Shiozaki, M.; Masuko, H. Heterocycles 1984, 22, 1727–
1728.
. Hamilton, G. S.; Huang, Z.; Yang, X.-J.; Patch, R. J.;
Narayanan, B. A.; Ferkany, J. W. J. Org. Chem. 1993,
15. Campbell, M. M.; Carruthers, N. I.; Mickel, S. J. Tetra-
hedron 1982, 38, 2513–2524.
5
8, 7263–7270.

1
622
C. V. Ste6ens et al. / Tetrahedron Letters 44 (2003) 1619–1622
1
1
6. Agami, C.; Couty, F.; Rabasso, N. Tetrahedron Lett.
002, 43, 4633–4636.
6.31 (1H, d, JHꢀP=22.8 Hz, NCHP); 6.8–6.9 (2H, m,
2
CHarom); 7.1–7.2 (3H, m, CHarom); 7.3–7.4 (3H, m,
CHarom); 7.6–7.7 (2H, m, CHarom).
7. Representative experimental procedure: To a dry round-
bottom flask, 0.97 g (5 mmol) of N-(benzylidene)-benzyl-
amine was added to 9 ml of dry toluene. The solution
was stirred and cooled to −40°C under an N2 atmo-
sphere. Then, a solution of 0.62 g of chloroacetyl chloride
in 2 ml of dry toluene was added dropwise. A white
precipitate was formed in the reaction mixture. The mix-
ture is stirred for 10 min at −40°C and then a solution of
13
C NMR l (68 MHz, CDCl , ppm): 41.85 (CH Cl);
3
2
49.40 (NC6 H Ph); 53.19 (OMe, JCꢀP=7.3 Hz); 53.84
2
(OMe, J =7.3 Hz); 54.16 (NCH, JCꢀP=158.6 Hz);
CꢀP
125.59; 127.19; 127.31; 128.34; 128.39; 128.68; 128.98
(2×); 130.37; 130.49 (CH, Ph); 132.61 (Cquat, Ph, JCꢀP
=
2.4 Hz); 136.51 (Cquat, Ph); 167.98 (C=O, JCꢀP=3.7 Hz).
3
1
P NMR l (109 MHz, CDCl , ppm): 22.48.
3
−
1
0
.68 g of trimethyl phosphite in 2 ml of dry toluene was
IR (NaCl, cm ): 1662, 1254.
+
added. After the addition, the precipitate dissolved again
due to the dealkylation which is kept at 80°C for 3 h.
Finally, the solvent was evaporated and the dimethyl 1-
MS (m/z, %): 381 (M , 5); 274 (26); 215 (11); 202 (8); 201
(90); 197 (15); 196 (10); 95 (8); 93 (8); 91 (100).
18. Britelli, D. R. J. Org. Chem. 1985, 50, 1845–1847.
19. Ternansky, R. J.; Morin, J. M. In The Organic Chemistry
of i-Lactams; Georg, G. I., Ed. Novel Methods for
Constructing the b-Lactam Ring; VCH Publishers, 1993;
pp. 269–290.
[
benzyl (chloroacetyl)amino]benzyl phosphonate was
obtained pure using flash chromatography (80/20 EtOAc/
petroleum ether; Rf=0.28; yield=56%).
Spectral data:
1
H NMR l (270 MHz, CDCl , ppm): 3.68 (3H, d,
20. Gerona-Navarro, G.; Bonache, M. A.; Herranz, R.; Gar-
c ´ı a-L o´ pez, M. T.; Gonz a´ lez-Mu n˜ iz, R. Synlett 2000, 9,
1249–1252.
21. Bossio, R.; Marcos, C. F.; Marcaccini, S.; Pepino, R.
Tetrahedron Lett. 1997, 38, 2519–2520.
3
JHꢀP=10.9 Hz, OMe); 3,80 (3H, d, JHꢀP=10.9 Hz, OMe);
3
.84 (1H, d, JAB=12.5 Hz, CH6 H Cl); 3.93 (1H, d,
A B
JAB=12.5 Hz, CH H
NC H6 H Ph); 5.02 (1H, d, J =18.2 Hz, NCH H6 BPh);
A B AB A
6
Cl); 4.84 (1H, d, JAB=18.2 Hz,
B
A