A concise and first synthesis of α-aminophosphinates with two stereogenic atoms leading to optically pure α-amino-H-phosphinic acids

Article abstract of DOI:10.1002/chem.200800690

A highly stereoselective synthesis of ω-aminophosphinates by nucleophilic attack of ethyl diethoxymethylphosphinate to Ellman's N-(tert-butanesulfinyl) ketimines by using Rb₂CO₃ as a base at room temperature was reported. In the first set of experiments, (S)-(tert-butane-sulfinyl)methyl (p-bromo)phenylketimine (1j) was used as the model compound to study its reaction with ethyl diethoxymethyl-phosphinate. Ethyl diethoxymethyiphosphinate and Rb₂CO₃ in CH ₂Cl₂ were placed in a 20 ml, Schienk flask and (5)-N-tert-butanesulfinylketimines 1 were then added at room temperature. The mixture was then stirred for 3-4 d, while being carefully monitored by TLC. The H NMR analysis showed that they have minor differences with shifts at 3.5 and 4.2 ppm, which can be assigned to the diethoxymethyl and the ethoxyl groups of the products.

Full text of DOI:10.1002/chem.200800690

COMMUNICATION
DOI: 10.1002/chem.200800690
A Concise and First Synthesis of a-Aminophosphinates with Two Stereogenic
Atoms Leading to Optically Pure a-Amino-H-phosphinic Acids
Dehui Zhang and Chengye Yuan*[a]
As phosphorus analogues of natural a-aminocarboxylic
acids, synthetic study of a-aminophosphonic acids is of great
interest to organic, bioorganic, and medicinal chemists.[1]
Very recently, our group has reported a convenient method
for the preparation of optically active a-aminophosphonic
acids based on a nucleophilic addition of dialkylphosphite to
sp2-carbon atom under mild reaction conditions using N-tert-
butanesulfinyl imines as chiral auxiliaries.[2] However, direct
comparison of a-aminophosphonic acids with a-aminocar-
boxylic acids is not reasonable, since the former belong to
dibasic acids while the latter are monobasic acids in nature.
From this structural point of view, a-amino-H-phosphinic
acids are much closer to natural a-aminocarboxylic acids
(Scheme 1). It is well documented that optically active a-
aminocarboxylic acids play an irreplaceable role in the bio-
logical metabolism.[3] Consequently, it is reasonable to
assume that optically active a-amino-H-phosphinic acids
might demonstrate greater biological activity than the well-
studied a-aminophosphonic acids as it could act as a more
effective surrogate of a-aminocarboxylic acids and block the
reproduction of DNA. Unfortunately, probably due to the
unique structure, it was more difficult to synthesize optically
pure a-amino-H-phosphinic acids[4] and most of the synthet-
ic methods available only lead to racemates.[5] For these rea-
sons, further chemical and biological studies on this kind of
compounds have run aground for decades.
Adol reaction has been always an effective way to intro-
duce hydroxyl or amino group into target molecules, and
with this strategy both a-hydroxyl phosphonates and a-hy-
droxylphosphinates have been obtained successfully.[6,7] Our
group has also realized the convenient synthesis of a-amino-
phosphonates initiated by this concept.[2] To the best of our
knowledge, however, owing to the lack of effective chiral-in-
ducing reagents, syntheses of optically active a-aminophos-
phinates have not yet been reported by a nucleophilic adol
reactions. As part of our efforts, we wish to disclose a novel
and facile and highly stereoselective synthesis of a-amino-
phosphinates by nucleophilic attack of ethyl diethoxyme-
thylphosphinate to Ellmanꢀs N-(tert-butanesulfinyl)ket-
imines[8] in CH2Cl2 using Rb2CO3 as a base at room temper-
ature; subsequently these compounds can be readily con-
verted to optically pure a-amino-H-phosphinic acids. During
this process, for the first time, a pair of diastereoisomers,
which have different configurations at the phosphorus atom,
was obtained.
In the first set of experiments, we chose (S)-(tert-butane-
sulfinyl)methyl (p-bromo)phenylketimine (1j) as the model
compound to study its reaction with ethyl diethoxymethyl-
phosphinate (2; Scheme 2).[9] The reaction conditions were
chosen based on our previous work: K2CO3 as base and
CH2Cl2 as solvent were initially an ideal conditions for the
reaction, because they gave excellent results in the reactions
of dimethyl phosphonate with various Ellmanꢀs N-(tert-buta-
nesulfinyl)ketimines. However, probably due to the subtle
differences between the chemical characteristics of ethyl di-
ethoxymethylphosphinate and dimethyl phosphonate, these
conditions proved not to be so effective, in that few prod-
ucts were detected even after five days. The change of
Scheme 1. Natural a-aminocarboxylic acids and their phosphorus ana-
logues.
[a] D. Zhang, Prof. Dr. C. Yuan
State Key Laboratory of Bio-organic and Natural Products Chemistry
Shanghai Institute of Organic Chemistry
Chinese Academy of Sciences
354 Feng-lin Rd., Shanghai, 200032 (China)
Fax : (+86)(21)5492-5379
E-mail: yuancy@mail.sioc.ac.cn
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200800690.
Chem. Eur. J. 2008, 14, 6049 – 6052
ꢁ 2008 Wiley-VCHVerlag GmbH& Co. KGaA, Weinheim
6049

Compound (SS,RC ,RP)-4 or
(SS,RC ,SP)-4’ can be readily con-
verted to its corresponding op-
tically active a-amino-H-phos-
phinic acids by refluxing in 4n
HCl (Scheme 3).
Scheme 2. The screening of reaction conditions.
Table 1. The synthesis of (SS,RC ,RP)-4 and (SS,RC ,SP)-4’.
CH2Cl2 to benzene gave better results, but the reaction still
needed more than five days to complete. Considering that
the steric hindrance of ethyl diethoxymethylphosphinate is
much larger than that of dimethyl phosphonate, we assumed
that the basicity of K2CO3 is too weak to initiate the reac-
tion. This assumption was supported by the fact that other
weak bases such as CaCO3, BaCO3, and KF were also inef-
fective in this reaction. On the basis of this assumption, we
chose Cs2CO3, which is a much stronger base, to study this
reaction. To our delight, the reactants did disappear in much
shorter time (one day); however, we still got very few prod-
ucts. We surmised that the reactants, both 1j and 2, may be
unstable under too basic conditions. Finally, we decided to
choose Rb2CO3, which has moderate basicity compared with
K2CO3 and Cs2CO3. It is interesting that we obtained two
isomers (4j and 4j’), which can be separated completely by
silica gel chromatography. Through the 1HNMR analysis,
we found that they have minor differences with shifts at 3.5
and 4.2 ppm, respectively, which can be assigned to the di-
ethoxymethyl and the ethoxyl groups of the products.
Whatꢀs more surprising is that the 31P NMR spectra indicat-
ed that each of the two isomers had single structure—they
both gave single peak in the 31P NMR spectrum! To clarify
the enantioselectivity of this reaction, we subsequently hy-
drolyzed the two isomers to their corresponding a-amino-H-
phosphinic acids and measured their optical values and
found that they had similar values, which indicated that this
reaction may afford high enantioselectivity on a-C, and it
may, at the same time, split the two P isomers. Further in-
vestigation into the reaction was then carried out and the re-
sults were summarized in Table 1.
R1
Yield [%][a,b]
4
Yield [%][a,b] 4’
1a
1b
1c
1d
1e
1 f
1g
1h
1i
1j
1k
1l
1m
1n
1o
1p
1q
1r
Ph
46 (35.23)
37 (35.29)
48 (34.57)
37 (34.97)
29 (35.37)
49 (35.07)
37 (35.35)
43 (34.82)
46 (34.64)
41 (33.84)
35 (34.49)
45 (33.90)
50 (34.54)
43 (35.33)
49 (35.20)
41 (34.13)
32 (34.02)
45 (40.73)
50 (35.72)[b]
48 (35.78)
50 (35.00)
34 (35.72)
36 (35.87)
49 (35.37)
36 (35.98)
50 (35.14)
48 (34.95)
35 (34.13)
49 (34.93)
47 (34.06)
35 (34.56)
49 (36.03)
49 (35.66)
48 (34.26)
38 (34.07)
30 (40.27)
p-CH3C6H4
p-CH3OC6H4
p-CH3SC6H4
p-morpholino-C6H4
3,4-methylenedioxy-C6H4
o-FC6H4
p-FC6H4
p-ClC6H4
p-BrC6H4
1-furanyl
1-thiophenyl
3-pyridinyl
2-naphthyl
4-biphenyl
4-cyanoC6H4
4-nitroC6H4
hexyl
[a] Isolated yield. [b] The values in the brackets are the 31P NMR chemi-
cal shift values in ppm, each product give single peak in its 31P NMR, in-
dicating its de% is more than 95% (except for 4g and 4h, for which
91% ee were obtained).
To further determine the accurate enantiomeric excess
(ee) values of the products, we oxidized (SS,RC ,RP)-4q and
(SS,RC ,SP)-4q’ into their corresponding sulfonylamide deriv-
atives (RC ,RP)-6q and (RC ,SP)-6q’ (see Scheme 4).[11,12] The
high optical purity of (RC ,RP)-6q and (RC ,SP)-6q’ are deter-
mined by HPLC (>99% ee), indicating that this synthetic
method promises to be a general and convenient approach
for the preparation of enantiomerically pure a-amino-H-
phosphinic acids.
In conclusion, the unprecedented nucleophilic attack of
ethyl diethoxymethylphosphinate to Ellmanꢀs N-(tert-buta-
nesulfinyl)ketimines by using Rb2CO3 as base, followed by
heating under reflux with 4n HCl has shown to be a highly
stereoselective and convenient synthesis of a-amino-H-phos-
phinic acids. In view of the mild conditions of this novel re-
action, the optically pure a-amino-H-phosphinic acids are
now more accessible and their potential application in bio-
logical systems is therefore feasible and encouraging. The
mechanism of this subtle reaction and the application of the
present synthetic methodology in the synthesis of optically
We are fortunately enough to obtain the single crystals of
4q and 4n’,[10] and we confirmed their structures and abso-
lute configurations by single-crystal X-ray analysis.
(Figure 1)
As shown in ORTEP drawings, it can be easily found that
the configuration of a-C of 4q and 4n’ are both R, which in-
dicated that the high enantioselectivity of a-C was achieved
in combination with the 31P NMR analysis (>95% diaste-
reomeric excess (de)). What is more interesting is that, ac-
cording to ORTEP drawings, the two isomers of phosphorus
atom were also obtained! (for 4q, the configuration of phos-
phorus atom is R; while for 4n’, the configuration of phos-
phorus atom is S). That is to say, in this novel reaction, high
enantioselectivity of both a-C and P were realized simulta-
neously, and we can call this “one stone, two birds”. Actual-
ly this is the first direct experimental evidence supporting
the stereogenic nature of phosphorus atom of a-aminophos-
phinates.
6050
www.chemeurj.org
ꢁ 2008 Wiley-VCHVerlag GmbH& Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 6049 – 6052

Nucleophilic Addition Reactions
COMMUNICATION
Scheme 3. Synthesis of optically active a-amino-H-phosphinic acids.
Scheme 4. Transformation of (SS,RC ,RP)-4q and (SS,RC ,SP)-4q’ into their
corresponding sulfonyl-amide derivatives (RC ,RP)-6q and (RC ,SP)-6q’.
added to quench it. Then the organic layer was separated and the aque-
ous layer was washed with Et2O (3”10 mL). The organic layers were
again washed with brine and dried over anhydrous Na2SO4 overnight.
The solvent was removed under reduced pressure and the residue was
subjected to silica gel chromatography (300–400 mesh, EtOAc/petroleum
3:1 to 6:1 for compound 4 and acetone/EtOAc 1:1 for compound 4’) to
afford pure 4 and 4’.
General procedure for the synthesis of (R)-(+)-a-methyl-a-amino-H-
phosphinic acid (5): A 20 mL Schlenk flask was flushed with Ar. Then 4
or 4’ (0.2 mmol) was added under Ar atmosphere followed by infusing
with 4n HCl (4 mL). The mixture was heated to reflux with stirring for
10 h before it was cooled to room temperature. The aqueous solution was
then washed with CH2Cl2 (5”5 mL), after which EtOH(30 mL) was
added to the aqueous layer and the resulting mixture was treated to re-
duced pressure to remove the solvent (water as well as EtOH) to near
dryness. The residue was again dissolved in a minimum amount of EtOH
followed by the addition of excess propylene oxide, in the process of
which a large amount of a white precipitate formed. The mixture was
stirred for 24 h and then filtered. The solid was collected while the fil-
trate was then evaporated to near dryness (a little amount of EtOHwas
usually required) and treated with EtOAc. The resulting white precipitate
was again filtrated and the solid collected in combination with that previ-
ously obtained was then washed with propylene oxide or petroleum to
afford pure 5.
Figure 1. ORTEP drawing for (SS,RC ,RP)-4q (top) and (SS,RC ,SP) 4n’
(bottom).
pure b-amino-H-phosphinic acids and g-amino-H-phosphinic
acids are currently under investigation in our laboratory.
Experimental Section
General procedure for the stereoselective synthesis of ethyl 1’,1’-diethoxy-
ethyl-(SS,RC ,Rp)-(+)-1-(tert-butanesulfinylamino)-1-aryl-1-methylphos-
phinate (4) and ethyl 1’,1’-diethoxyethyl-(SS,RC ,SP)-(+)-1-(tert-butanesul-
finylamino)-1-arylmethylphosphinate (4’): Ethyl diethoxymethylphosphi-
nate (2; 392 mg, 2 mmol) and Rb2CO3 (577 mg, 2.5 mmol) in CH2Cl2
(5 mL) were placed in a 20 mL Schlenk flask and (S)-N-tert-butanesulfi-
nylketimines 1 (0.5 mmol) were then added at room temperature. The
mixture was then stirred for 3–4 d, while being carefully monitored by
TLC. After the reaction came to the completion, water (5 mL) was
The identity and purity of the known products were confirmed by suffi-
cient spectroscopic analysis, and the new products were fully character-
ized (see the Supporting Information).
Chem. Eur. J. 2008, 14, 6049 – 6052
ꢁ 2008 Wiley-VCHVerlag GmbH& Co. KGaA, Weinheim
www.chemeurj.org
6051

C. Yuan and D. Zhang
[8] a) For reviews relating Ellmanꢀs N-(tert-butanesulfinyl)ketimines,
see: J. A. Ellman, T. D. Owens, T. P. Tang, Acc. Chem. Res. 2002, 35,
984; b) D. Mortona, R. A. Stockman, Tetrahedron 2006, 62, 8869–
8905; c) P. Zhou, B. C. Chen, F. A. Davis, Tetrahedron 2004, 60,
8003–8030; d) F. A. Davis, B. C. Chen, Chem. Soc. Rev. 1998, 27, 13;
e) F. A. Davis, B. C. Chen, Aldrichimica Acta 2005, 38, 93–103;
f) F. A. Davis, S. Lee, H. Yan, D. D. Titus, Org. Lett. 2001, 3, 1757–
1760.
[9] M. J. Gallagher, H. Honegger, Aust. J. Chem. 1980, 33, 287–294.
[10] a) For the correlation of the single-crystal structures of 4q and 4n’
with the 1HNMR data, please see the Supporting Information;
b) CCDC 675307 (4q) and 675308 (4n’) contain the supplementary
crystallographic data for this paper. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
Acknowledgements
The project was supported by the National Nature Science Foundation of
China (Grant No. 20272075 and 20072052).
Keywords: amino-H-phosphinic acids · aminophosphinates ·
enantioselectivity · phosphorus · synthetic methods
[1] V. P. Kukuhar, H. R. Hudson, Aminophosphorus Acids and Amino-
phosphinic Acids, Wiley, Chichester (UK), 2000.
[2] Q. Y. Chen, C. Y. Yuan, Synthesis 2007, 3779–3786.
[3] R. F. Weaver, Molecular Biology, McGraw-Hill, New York, 2005.
[4] T. Haruki, T. Yamagishi, T. Yokomatsu, Tetrahedron: Asymmetry
2007, 18, 2886–2893, and the references therein.
[11] G. Borg, M. Chino, J. A. Ellman, Tetrahedron Lett. 2001, 42, 1433–
1436.
[12] It should be emphasized that difficulties were encountered during
the synthesis of the racemic 6q and 6q’ with S configuration on the
sulfur atom. The racemates, which consist of four isomers prepared
from racemic N-tert-butanesulfinylketimines, have different chirali-
ties on the sulfur atom, so it is necessary for us to eliminate the chir-
ality at the sulfur atom to obtain two pairs of isomers with two
chiral atoms (C and P). Only with the above treatment, HPLC anal-
yses became possible.
[5] a) M. Drag, J. Oleksyszyn, Tetrahedron Lett. 2005, 46, 3359–3362;
b) R. Tyka, G. Hagele, R. Boetzel, Phosphorus Sulfur Silicon Relat.
Elem. 1991, 62, 75–82; c) X. Y. Jiao, C. Verbruggen, M. Borloo, W.
Bollaert, A. de Groot, R. Dommisse, A. Haemers, Synthesis 1994,
23–24.
[6] a) H.-J. Cristau, J.-L. Pirat, D. Virieux, J. Monbrun, C. Ciptadi, Y.-A.
Bekro, J. Organomet. Chem. 2005, 690, 2472–2481; b) B. Kaboudin,
Tetrahedron Lett. 2003, 44, 1051–1054; c) D. Ridder, H. Wunderlich,
M. Braun, Eur. J. Org. Chem. 1998, 1071–1076.
Received: April 11, 2008
[7] G. Tadeusz, Phosphorus Sulfur Silicon Relat. Elem. 1993, 85, 59–64.
Published online: May 19, 2008
6052
www.chemeurj.org
ꢁ 2008 Wiley-VCHVerlag GmbH& Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 6049 – 6052