Organocatalyzed synthesis of tertiary α-hydroxyphosphonates by a highly regioselective modified Pudovik reaction

Article abstract of DOI:10.1002/ejoc.201100308

An organocatalytic modified Pudovik reaction between dialkyl or diaryl phosphites and α-haloketones was used to synthesize β-chloro-α- hydroxyphosphonates in high yields with very high regioselectivity and high stereoselectivity. Aliphatic, aromatic, and

Full text of DOI:10.1002/ejoc.201100308

FULL PAPER
DOI: 10.1002/ejoc.201100308
Organocatalyzed Synthesis of Tertiary α-Hydroxyphosphonates by a Highly
Regioselective Modified Pudovik Reaction
Maria Teresa Barros*[a] and Ana Maria Faísca Phillips*[a]
Keywords: Phosphonates / Organocatalysis / Asymmetric synthesis / Diastereoselectivity
An organocatalytic modified Pudovik reaction between di-
alkyl or diaryl phosphites and α-haloketones was used to
synthesize β-chloro-α-hydroxyphosphonates in high yields
with very high regioselectivity and high stereoselectivity. Ali-
phatic, aromatic, and cyclic ketones could be used success-
fully. Under the optimized conditions it was possible to obtain
enantiomerically enriched products with a combination of
quinine and a stoichiometric racemic base, but the ee values
only went up to 40%. The presence of a β-chloro function-
ality allows further synthetic elaboration of the hydroxyphos-
phonates obtained, which could be useful for target-oriented
synthesis, as these compounds often have important bio-
logical activity.
Introduction
α-Hydroxyphosphonic acids are mimics of natural hy-
droxycarboxylic acids with the carboxylic acid group re-
placed by a phosphonic acid group. They usually show po-
tent biological activity as inhibitors of enzymes.[1] A few
examples are shown in Figure 1. Phosphonopeptide A in-
hibits human renin, the enzyme that catalyzes the conver-
sion of angiotensinogen into angiotensin, a hormone that
causes blood vessels to constrict causing a rise in blood
pressure.[2] Hydroxyphosphonate B is an inhibitor of CD-
45 tyrosine phosphatase, which plays an important role in
the regulation of cell activation and proliferation in haema-
topoetic cells, from which all types of blood cells rise.[3]
Bisphosphonate C possesses potent anticancer activity.[1] In
addition, suitably substituted α-hydroxyphosphonates may
be precursors of epoxyphosphonates and vice versa.
Figure 1. Biologically active α-hydroxyphosphonates.
These compounds may have antibiotic activity such as
the broad spectrum antibiotic (–)-(1R,2S)-1,2-epoxyprop-
ylphosphonic acid (D),[4] commercialized as Fosfomycin,
which inhibits the first step of bacterial cell wall biosynthe-
sis with little toxicity to humans and used clinically to treat
Gram positive and Gram negative bacterial infections. The
antibiotic is produced naturally in culture media of Strepto-
myces fradiae, but the chemical synthesis of synthetic ana-
logues is an active area of research, even more so nowadays
because the global problem of antimicrobial resistance has
led to a renewed interest in its use.[5]
α-Oxygenated phosphonates may be obtained by a very
short synthetic route, that is, two steps, from an inorganic
source of phosphorus, PCl3, through a modified Pudovik
reaction between dialkyl phosphites and α-haloketones.[6]
Yet, despite the simplicity, the reaction has its limitations,
and different products or mixtures of products may be ob-
tained depending on the reaction conditions, as indicated
in Scheme 1. The phosphite is normally activated in situ by
reaction with a sodium alkoxide, or a preformed sodium
dialkylphosphonate is used. However, the reaction condi-
tions used may be too harsh for the sensitive functionalities
of α-haloketones or the reaction may not be regiospecific.
Epoxide 4 alone may form from aliphatic substrates,[7] or
vinyl phosphate 5 alone,[8] or a mixture of epoxide and vinyl
phosphate,[9] or a mixture of epoxide and β-oxophos-
phonate 6,[10] depending on where the initial attack by
phosphorus takes place or on the fate of the resulting
anion. Thus, the reaction with a dialkyl phosphite and so-
[a] Department of Chemistry, REQUIMTE, Faculdade de Ciên-
cias e Tecnologia, Universidade Nova de Lisboa,
Quinta da Torre, 2829-516 Caparica, Portugal
Fax: +351-212948550
E-mail: mtbarros@dq.fct.unl.pt
amfp@dq.fct.unl.pt
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201100308.
4028
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 4028–4036

Organocatalyzed Synthesis of Tertiary α-Hydroxyphosphonates
dium alkoxide gives good yields of epoxides 4,[11] but in this by nucleophilic attack at the carbon atom α to phos-
case only aliphatic α-haloketones can be used. If aromatic phorus.[19] They open the epoxide ring making the drug in-
α-haloketones are used, the only products that are produced effective. A quaternary carbon atom at this position would
at room temperature are 1-arylvinylphosphates 5. At most likely be too hindered for enzymatic attack, and a
–35 °C, aromatic β-haloenol phosphates 7 are obtained,[12] more effective antibiotic could be the result. Although chi-
but, when α-bromo-4-nitro-acetophenone was used, the ep- ral quaternary carbon centers are notoriously difficult to
oxide was obtained. In the reaction with preformed sodium make,[20] because there have been no asymmetric versions
diethylphosphonate, 1:1 mixtures of epoxides 4 and vinyl of this reaction described as of yet, we were also interested
phosphates 5 were obtained, and the halogenated α-carbon in investigating the possibility of developing an enantiose-
cannot be tertiary.[8a,8c,9a,9c,10d,13] There has also been one lective reaction.
report that phenacyl chloride reacts with dimethyl phos-
phite in the presence of bases to give corresponding β-
chloro-α-hydroxyphosphonate 3.[14] However, in that work,
only one α-haloketone was studied. α-Tosylketones have
also been used in these reactions instead of the chlo-
roketones, but only aliphatic ones.[15] Different bases have
also been used: with KF the reaction was not regioselec-
tive;[16] with phase-transfer catalysts (TBAB–K2CO3) there
were no examples with aromatic α-haloketones.[17]
Results and Discussion
Recently a few methods have been developed using cin-
chona alkaloids to activate dialkyl phosphites in other reac-
tions leading to the production of chiral molecules often
with good to excellent ee values.[21] The subject has been
reviewed recently.[22] Hence we decided to try these organ-
ocatalysts in the Pudovik-type reaction between dialkyl
phosphites and α-halo ketones. Cinchona alkaloids may
function as kinetically active chiral Brønsted bases, depro-
tonating the phosphite. As we were particularly interested
in developing a Pudovik-type reaction that would also al-
low the use of the “elusive” aromatic α-haloketones, we
started by choosing the reaction of dialkyl phosphite 1a
with 2-chloroacetophenone as a model reaction for our
studies, and quinine as the base (Scheme 2). When 10 mol-
% catalyst was used, after 17 h tertiary hydroxyphos-
phonate 3a had formed, and 54% of the phosphite re-
mained unreacted as determined by 31P NMR spectroscopy.
After 41 h, 34% of the phosphite still remained unreacted.
Scheme 1. The products that may form in the reaction of a dialkyl
phosphite with a α-haloketone in the presence of base.
From this literature search above it is clear that there
exists at present no general method based on the modified
Pudovik reaction that can be applied chemoselectively to
the synthesis of both aromatic and aliphatic epoxides.
Furthermore, to the best of our knowledge, there is only
one report on the synthesis of tertiary alcohols based on
this reaction and in that report only one α-haloketone was
studied.[14] Considering the usefulness and synthetic versa-
Scheme 2. The reaction used as a model in this study.
We then decided to use an additional base. If the reaction
tility of α-oxygenated phosphonates it appears that a short stops at the alcohol, there is mass balance when one base
synthetic general route to β-chloro-α-hydroxyphosphonates is used in catalytic amounts, because the proton removed
would be synthetically useful. As far as we know there have from the phosphite can be used to protonate the intermedi-
been no asymmetric versions of this reaction reported ate alkoxide anion that forms. If the epoxide is to form, an
either. As a follow-up to our interest in epoxyalkylphos- additional base has to be present in stoichiometric amounts
phonate and hydroxyalkylphosphonate chemistry,[18] we de- to accept the proton so that the chiral Brønsted base can
cided to investigate the possibility of developing a general be regenerated for another catalytic cycle. Cinchona alka-
version of this reaction, which could be applied to aliphatic, loids may also function both as phase-transfer agents and
alicyclic, and aromatic substrates. The products would have a Brønsted base if they are used in combination with a solid
a quaternary carbon center α to phosphorus. Interesting inorganic base. We decided to try conditions that would
biological activity may result. For example, although nucle- allow the formation of the epoxide to take place. As stoi-
ophilic attack generally takes place β to the phosphorus chiometric base, we started with proton sponge, a kinet-
atom in epoxyphosphonates, for electronic and steric ically restricted, hindered but thermodynamically active
reasons, it is known that the enzymes that cause resistance base, which would not react with the phosphite but would
to Fosfomycin, namely, Fos A, Fos B, and Fos X, function be able to act as a proton scavenger. When quinine (10 mol-
Eur. J. Org. Chem. 2011, 4028–4036
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4029

M. T. Barros, A. M. F. Phillips
FULL PAPER
%) was used as catalyst in combination with proton sponge, Table 1. Influence of the nature of the stoichiometric base on the
reaction.[a]
the reaction proceeded smoothly and regioselectively to give
the 1-phenyl-substituted α-hydroxyphosphonate as the only
product after 6 h. No epoxide formed. Although the prod-
uct had an optical rotation of [α]1D4 = 22.0 (c = 1.51), the
enantiomeric excess was low, only 10%, as determined by
HPLC. Preliminary experiments were then conducted to in-
vestigate the effect of the nature of the solvent on the reac-
tion. It was found that polar solvents like CH2Cl2 and iPr-
OH gave inductions similar to those obtained in toluene,
but the reaction was much slower. In THF, the ee values
were even lower, and in cyclohexane, in which the reaction
components were highly insoluble, the reaction was con-
siderably retarded with respect to toluene. So toluene was
the solvent of choice for the remaining part of the work,
with the exception of the preparation of the standards for
HPLC. Standards were prepared by using as catalyst an
equimolar mixture of (–)-quinine and its pseudoenantiomer
(+)-quinidine, and as quinidine is poorly soluble in toluene,
iPrOH was the solvent used in those reactions.
Base
t
[h]
Products [%]
α-Hydroxy-
phosphonate
ee [%]
Phos-
phate
Pyridine[b]
K2CO3
41
2
5
6
3
80
78
77
100
100
0
22
21.5
0
23
11
25
10
ND
Na2CO3
Proton sponge[c]
Proton sponge[d]
0
[a] Reaction conditions: phosphonate/ketone/stoichiometric base/
quinine = 1:1.1:1:0.1, toluene, room temperature, 100% conversion.
[b] In this case the results are for 80% conversion of the phosphite
only. [c] Phosphonate/ketone = 1:1.5. [d] Phosphonate/ketone =
1:2.
Using quinine as chiral catalyst, a few other substances
were then tried as the stoichiometric base, and the results
are presented in Table 1. The inorganic base Na2CO3 gave
the best ee, 25%, but in this case the reaction was not re- fects of temperature and catalyst load were then investi-
gioselective, and approximately 22% of enol phosphate was gated, and the results are presented in Table 2. While there
formed as a byproduct. Pyridine could also be used as a was no influence of temperature on ee, the reactions became
base, giving a product with approximately the same ee but slower at lower temperatures as expected. Increasing cata-
the reaction was considerably slower.
lyst load to 30% gave a modest increase in ee from 25 to
Using Na2CO3 as the stoichiometric base, a few other 34%.
cinchona alkaloids were then tried in an attempt to improve
Since the best enantiomeric excess was obtained when
asymmetric induction, but the results did not improve (Fig- 30 mol-% Na2CO3 was used as the stoichiometric base, we
ure 2). With quinidine as catalyst, the major enantiomer then applied these conditions to the synthesis of a few phos-
formed had the opposite configuration to that obtained phonates, and the results are presented in Table 3. Even
with quinine as catalyst, as determined from the area ratios when the reaction was run with equimolar amounts of cata-
of the peaks in the HPLC trace. However, even when cupr- lyst and phosphite, the ee did not change much. However
eine was tried as catalyst, which could allow extra stabiliza- the reaction became much slower, due to the poor solubility
tion of the transition state through H-bonding with an ad- of the catalyst in the reaction medium. Other ketones,
ditional OH group, the ee values did not improve. The ef- namely, 2,4Ј-dichloroacetophenone and 2-chloropropio-
Figure 2. Effect of the nature of the chiral catalyst on the reaction. The following conditions were used: phosphonate (SM)/ketone/
Na2CO3/chiral catalyst = 1.0:1.1:1.0:0.1, toluene, 5 h; exception: quinidinium benzyl bromide: 16 h, room temp.
4030
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 4028–4036

Organocatalyzed Synthesis of Tertiary α-Hydroxyphosphonates
Table 2. The influence of temperature and catalyst load on the reaction between phosphite 1a and 2-chloroacetophenone catalyzed by
quinine/Na2CO3.[a]
Entry
T
[°C]
Catalyst
[mol-%]
t
[h]
Hydroxyphosphonate
[%]
Phosphite
[%]
Phosphate
[%]
ee[b]
[%]
1
2
3
4
5
6
7
r.t.
–25 to –35
5–6
r.t.
r.t.
5–6
–25 to –35
10
10
10
20
30
20
30
5
17
4.5
5
5
4.5
23.5
77
95
93
97
96
94
77
1.5
0
2
0
0
2
23
21.5
25
23
20
23
34
20
29
5
5
3
4
4
0
[a] Reaction conditions: phosphonate/ketone = 1:1.1 in toluene. The yields were determined from peak area ratios by 31P NMR spec-
troscopy. [b] Determined by HPLC on a chiral column.
Table 3. The synthesis of tertiary 1-hydroxyalkylphosphonates catalyzed by quinine/Na2CO3 or cupreine/Na2CO3.[a]
Entry Phosphite
Ketone
Catalyst
T
[°C]
t
[h]
3
[%]
Phosphite
[%]
5
[%]
ee[b]
[%]
1
1a
1a
1a
1a
1a
1a
1a
1-phenacyl chloride
1-phenacyl chloride
1-phenacyl chloride
1-phenacyl chloride
1-phenacyl chloride
quinine
quinine
quinine
cupreine
cupreine
quinine
quinine
r.t.
5–6
r.t.
5–6
r.t.
r.t
5
4,3
2
3
2
2
18
3a, 96
3a, 94[c]
3a, 28
3a, 61
3a, 71
3b, 83[e]
3c, 59
0
2
7
39
20
0
5a, 4
5a, 4
5a, 65
5a, 0
5a, 9
5b, 17
5c, 21
34
20
36
23
28
0
2
3[d]
4
5
6
7
2,4Ј-dichloroacetophenone
2-chloropropiophenone
r.t.
13
0
[a] Reaction conditions: phosphite/ketone/catalyst/base = 1:1.1:0.3:1 in toluene. The yields were determined from peak area ratios by 31P
NMR spectroscopy. [b] Determined by HPLC on a chiral column. [c] The yield of purified 3a isolated by column chromatography was
61%. [d] The ratio of phosphite/catalyst = 1:1. [e] The yield of purified 3b isolated by column chromatography was 57%.
phenone, also reacted well, but the reactions were less re- in reaction times, presumably due to greater steric effects,
gioselective and a larger amount of phosphate was formed and whereas 2-phenyl-substituted phosphonate 4a could be
as a byproduct in each case, and the products were racemic. obtained in 6 h in the first case, it took almost 2 months for
Taking into account the fact that the reactions with the complete conversion to 4f in the second. Diphenyl phos-
quinine–sponge combination were the most regioselective, phite, on the contrary, was so reactive that the entire reac-
we decided to develop this modified Pudovik reaction as a tion had to be performed at 0 °C to be regioselective, other-
high yielding general method for the production of tertiary wise a large mixture of products was rapidly obtained. Un-
α-hydroxy-β-chlorophosphonates. Although the reaction der these conditions, the conversion into the desired hy-
proceeded well with a 1:1 ratio of phosphite/α-chloro- droxyphosphonate was 82%, with 10% phosphate, as well
ketone, it was considerably faster when a 1:1.5 ratio was as 8% of an unidentified product with δ = 128.5 ppm, prob-
used. It was found that if the reactions were not quenched ably a product of addition of phenol to the phosphite, as
as soon as the phosphite had been consumed, some vinyl the starting material contained 15% phenol. Aliphatic and
phosphate formed, presumably through the well-known 1- cyclic substrates also worked well, and a single product
hydroxyphosphonate to phosphate rearrangement,[23] as formed in these cases. When 2-chlorocyclopentanone was
base is present in the reaction mixture. The rearrangement used as substrate, there was no evidence by NMR spec-
was accompanied by subsequent elimination of chloride ion troscopy (1H, 13C, 31P) for the formation of more than one
in this case. Thus, to avoid long reaction times, a higher diastereoisomer. Thus, it appears that the phosphite nucleo-
ratio of α-chloroketone/phosphite was used. We also found phile approaches the carbonyl group from the same side in
that the rate of the reaction increased by increasing the both enantiomers of the α-haloketone, presumably on the
amount of proton sponge used, but in the method devel- face opposite to the chloro substituent, due to steric ap-
oped equimolar amounts of sponge and phosphite were proach control. Although the cyclopentanone ring system
used. With these conditions, we investigated the scope of has greater conformational flexibility than the cyclohex-
the method towards a variation in the phosphoryl group anone system and competing transition states are more
radicals, the effect of substitution in the aromatic ring, the equivalent,[24] when reagents with very large bulk are used
effect of the α-haloketone being aliphatic or cyclic, and also it is possible to obtain exceptional diastereofacial selectivity.
the effect of a greater degree of substitution at the haloge- For instance, in the reaction of 2-methylcyclopentanone
nated α-carbon atom. The results are compiled in Table 4. with lithium trisiamylborohydride, the reagent approaches
The regioselectivity of the reactions remained high, but the the cyclopentanone face opposite to the methyl group to
products were all nearly racemic (ee Յ 10%). Going from give the cis alcohol with Ն99.5% selectivity.[25] The quinine-
the cyclic 2-oxodioxaphosphorinane derivatives to aliphatic catalyzed addition of dialkyl phosphites to α-haloketones is
ethoxy phosphorus radicals, there was a marked retardation very sensitive to the bulk of the reagent, as observed in this
Eur. J. Org. Chem. 2011, 4028–4036
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4031

M. T. Barros, A. M. F. Phillips
FULL PAPER
study, when the reaction of diethyl phosphite is compared Table 4. Synthesis of tertiary 1-hydroxyalkylphosphonates.[a]
with that of the cyclic phosphite (Table 4, Entry 1 vs. 6).
Hence, presumably there was addition anti to the chlorine
atom in both starting enantiomers in this case, to give one
pair of enantiomers only. The reaction of 2-chloropropio-
phenone was also highly regioselective. Once again there
was no evidence in the spectra obtained from quinine–pro-
ton sponge reactions or quinine–pyridine reactions for the
formation of two diastereoisomers. All spectra contained a
single set of signals. Possibly with this conformationally
mobile substrate, the negatively charged nucleophile ap-
proaches from a side opposite to the electron-rich α-chlor-
ide atom and away from the medium-sized methyl group,
according to a Felkin–Anh model.[26] The large bulk of the
phosphite anion causes the selectivity observed. More than
one chromatography was normally necessary to separate
the excess amount of starting material, which tended to tail,
from the product, and hence the yields of the purified com-
pounds were lower than expected from the conversions ob-
tained.
Tertiary β-chloro-α-hydroxyalkylphosphonates 3a–h
were all prepared from α-chloroketones. When the more re-
active α-bromoketones were used as substrates, a large mix-
ture of products was obtained, the major ones being phos-
phates according to the 31P NMR spectroscopic data, and
the desired hydroxyphosphonate was present in the mixture
as a minor component only. This happened with o-, m-, or
p-nitro-2-bromoacetophenones and with p-methoxy-2-bro-
moacetophenone. The reaction with 4-nitro-2-bromo-
acetophenone was tried at 0 °C. After 2.5 h the reaction was
nearly complete, and the following product distribution was
obtained: β-bromo-α-hydroxyalkylphosphonate/unknown
phosphate 1/enol phosphate 5i/unknown phosphate 2 =
21:47:24:8. Thus, 2-bromoketones are not suitable reagents
for this reaction.
The reaction was then tried with a combination of quin-
ine/pyridine as catalysts, and the results are presented in
Table 5. These reactions were considerably slower than the
previous two, and generally it was found that the speed of
the reactions increased with increasing basicity of the
stoichiometric base. However, the products were obtained
in much higher yields than when proton sponge was used
as co-catalyst, but with similar enantiomeric excess values.
The highest ee observed with 2,4Ј-dichloroacetophenone
was obtained with this catalytic system.
Examples of reactions that lead to the formation of terti-
ary α-hydroxy alkylphosphonates are still limited.[27] Never-
theless, in recent years, organocatalyzed methods have been
[a] The reactions were performed with phosphite/ketone/quinine/
proton sponge = 1:1.5:0.1:1. [b] Conversion (conv.) indicates the
yield of product that formed as determined by 31P NMR spec-
troscopy, and within brackets is the yield of pure product after
isolation by column chromatography. [c] 25 mol-% catalyst was
used in this experiment. [d] There remained 25% unreacted phos-
phite at the end; the yield refers only to the amount of phosphite
that was converted into products. [e] The entire reaction was con-
ducted at 0 °C; the molar ratio of phosphite/ketone = 1:1.2; there
remained 23% unreacted phosphite at the end; the yield in this case
refers only to the amount of phosphite that was converted into
products.
developed for the production of such compounds having an
additional γ-oxo functionality,[28] an α-ester function-
ality,[29] and a β-nitro functionality[30] with good yields and
high enantioselectivities. The present method of synthesis
gives products with very high regioselectivity and in high
yields, but low enantioselectivities. It has the advantages
that highly complex molecules can be prepared in two steps
only from a simple inorganic phosphorus source, PCl3, and
the co-reactant β-chloroketones are small molecules often
readily available commercially. The presence of a β-chloro
4032
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 4028–4036

Organocatalyzed Synthesis of Tertiary α-Hydroxyphosphonates
Table 5. Synthesis of tertiary 1-hydroxyalkylphosphonates[a] with quinine/pyridine as catalysts.
[a] The reactions were performed with phosphite/ketone/quinine/pyridine = 1:1.5:0.1:1. [b] SM = Starting phosphite. [c] Conv. indicates
the total amount of product formed as determined by 31P NMR spectroscopy. [d] The yield refers to the amount of product isolated by
column chromatography in relation to the amount of phosphite that was converted into products. [e] Determined by HPLC on a chiral
column. [f] A ratio of phosphite/ketone = 1:1.1 was used in this reaction.
acid, used as external standard. The multiplicity of signals in 13C
substituent in the products allows further elaboration into
NMR spectra was determined with DEPT experiments. Two-di-
mensional spectra (COSY 45, HMQC, HMBC) or decoupling ex-
periments were used to help with structural determinations when-
ever necessary. IR spectra were obtained with a Perkin–Elmer Spec-
trum BX FT-IR spectrophotometer. Mass spectra were recorded
other useful phosphonates, that is, epoxyalkylphosphonates,
β-aminoalkylphosphonates, an so on.
Conclusions
with a Micromass GCT spectrometer, operating in the electron im-
pact mode or the field ionization mode, and were supplied by the
Thus we succeeded in developing a general method for
Mass Spectrometry Services of the Chemistry Department/RE-
the production of β-α-hydroxyalkylphosphonates by using
QUIMTE, FCT, UNL. Elemental analysis were performed with a
a modified Pudovik reaction that allows the introduction of
both aliphatic and aromatic substituents α to phosphorus in
Thermo Finnigan Elemental Analyzer 1112 series, by the Labora-
tory for External Services of CQFB-Lab Associado/REQUIMTE,
high yields and regioselectively. Although the enantiomeric
excess values obtained were low, going only up to 38%, this
is to the best of our knowledge the first report of an enan-
tioselective version of the modified Pudovik reaction be-
tween dialkyl phosphites and α-haloketones.
of the Department of Chemistry, FCT, UNL. For HPLC analysis,
a Merck Hitachi instrument equipped with a Chiralpak AD-H col-
umn from Daicel, and a Merck-Hitachi-4250 UV/Vis detector were
used.
5,5-Dimethyl-2-oxo-1,3,2-dioxaphosphorinane (1a): Prepared ac-
cording to the method of McConnell and Coover.[31]
General Procedure for the Synthesis of 2-Chloro-1-hydroxyalk-
ylphosphonates by using Quinine and Proton Sponge as Bases: The
reaction vessel was charged with the phosphite (1.0 mmol), fol-
lowed by dry toluene (3 mL). Once the phosphite had dissolved,
(–)-quinine (0.10 mmol) was added, and the mixture was stirred
under an atmosphere of argon until the catalyst had dissolved. Pro-
ton sponge (1.0 mmol) was added, and the mixture was again
stirred until a clear solution was obtained. The reaction mixture
was then cooled in an ice bath, and the α-chloroketone (1.5 mmol)
was added. The mixture was then warmed up, and it was stirred at
room temperature for the periods of time indicated in Table 3. The
reaction times were monitored by 31P NMR spectroscopy. When
Experimental Section
General Information: All reactions were carried out under an atmo-
sphere of argon. The reagents were obtained from commercial sup-
pliers and used without further purification. The solvents were
purified by standard methods and distilled before use. For column
chromatography, Merck silica gel 60 (230–400 mesh) was used, and
for thin layer chromatography, silica gel plates Merck 60 F254 were
used. Optical rotations were measured with an AA-1000 Polarime-
ter from Optical Activity Ltd. with 1-mL, 0.5-dm cells. NMR spec-
tra were obtained with a Bruker AR X400 NMR spectrometer or
with a Bruker Avance 400 MHz spectrometer (1H: 400 MHz, 13C:
100 MHz, 31P: 162 MHz). Chemical shifts are reported relative to all the phosphite had reacted, 1 m HCl was added, and the product
TMS. 31P NMR chemical shifts are reported relative to phosphoric
was extracted with chloroform (3ϫ). The combined organic ex-
Eur. J. Org. Chem. 2011, 4028–4036
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4033

M. T. Barros, A. M. F. Phillips
FULL PAPER
tracts were washed with water (1ϫ), and the solution was filtered
through anhydrous sodium sulfate. The solvent was removed on a
rotary evaporator to give the crude product as a solid that was
purified by column chromatography on silica gel as indicated be-
low.
as colorless crystals (26 mg, 27 %). M.p. 164–165 °C. 1H NMR
(CDCl3): δ = 0.87 (s, 3 H, CH3), 1.25 (s, 3 H, CH3), 1.89 (d, J =
6.6 Hz, 3 H, PCCCH3), 3.72 (ddd, J = 10.3, 16.4 Hz, 1 H,
POCHH), 4.01 (ddd, J = 10.4, 15.2 Hz, 1 H, POCHH), 4.34 (d, J
= 9.8 Hz, 1 H, POCHH), 4.64 (d, J = 10.8 Hz, 1 H, POCHH), 5.02
(q, J = 6.5 Hz, 1 H, PCCH), 7.34 (t, J = 7.6 Hz, 2 H, 2 Ph-H),
7.41 (t, J = 7.2 Hz, 2 H, 2 Ph-H), 7.63 (d, J = 7.6 Hz, 1 H, Ph-H)
ppm. 13C NMR (CDCl3): δ = 20.43 (s, CH3), 21.30 (s, CHClCH3),
22.10 (s, CH3), 32.30 (d, J = 7.0 Hz, POC-Cq), 65.93 (d, J =
16.1 Hz, PCCHCl), 78.09 (d, J = 5.0 Hz, POCH2), 79.61 (d, J =
6.0 Hz, POCH2), 82.36 (d, J = 165.7 Hz, P-C), 125.1 (s, 2 oC, Ph),
128.0 (s, pC, Ph), 128.4 (s, 2 mC, Ph), 139.9 (s, iC, Ph) ppm. 31P
NMR (CDCl3): δ = 10.00 ppm. MS (FI): m/z (%) = 322 (2), 321
(10), 320 (27), 319 (29), 318 (74) [M]+, 284 (2), 283 (10), 282 (37),
281 (13), 171 (5), 170 (28), 169 (19), 168 (83), 151 (17), 149 (12),
5,5-Dimethyl-2-(2Ј-chloro-1Ј-hydroxy-1Ј-phenylethyl)-1,3,2-dioxa-
phosphorinane 2-Oxide (3a): Prepared from dialkyl phosphite 1a
and 2-chloroacetophenone according to the general procedure. The
crude product was purified by preparative chromatography on sil-
ica gel (CHCl3/acetone, 2:1), to give product 3a as white crystals
(62 mg, 63%). M.p. 166–167 °C. [α]1D4 = –22.0 (c = 1.51, CHCl3).
1H NMR (400 MHz, CDCl3): δ = 0.87 (s, 3 H, CH3), 1.23 (s, 3 H,
CH3), 3.86 (dd, J = 12.0, 16.1 Hz, 1 H, POCHH), 4.03 (dd, J =
11.0, 13.9 Hz, POCHH), 4.28–4.40 (m, 3 H, 2 PCCH2, POCHH),
4.45 (d, J = 10.3 Hz, 1 H, POCHH), 7.32–7.43 (m, 3 H, 3 Ph-H),
7.62 (d, J = 7.4 Hz, 2 H, 2 Ph-H) ppm. 13C NMR (CDCl3): δ =
20.48 (s, CH3), 21.95 (s, CH3), 32.46 (d, J = 8.1 Hz, POC-Cq),
50.87 (d, J = 13.1 Hz, CH2O, C-3Ј), 76.68–77.32 (overlapp. d, P-
C), 78.55 (s, POCH2), 79.33 (s, POCH2), 126.1 (s, 2 CH, Ph-C),
128.2 (s, CH, Ph-C), 128.4 (s, 2 CH, Ph-C), 136.8 (s, Cq, Ph-C)
ppm. 31P NMR (CDCl3): δ = 10.47 ppm. MS (EI): m/z (%) = 268
133 (4), 105 (4), 69 (35). IR (KBr): ν = 478, 496, 541, 618, 670,
˜
699, 754, 776, 826, 896, 922, 946, 989, 1018, 1068, 1077, 1170, 1220,
1234, 1372, 1448, 1466, 1490, 1541, 1560, 1636, 1654, 2970,
3274 cm–1. C14H20ClO4P (318.734): calcd. C 52.76, H 6.33; found
C 52.78, H 6.30.
5,5-Dimethyl-2-(2Ј-chloro-1Ј-hydroxy-1Ј-methylethyl)-1,3,2-dioxa-
(1), 154 (3), 120 (2), 106 (9), 105 (100), 102 (8), 91 (7), 83 (5), 78 phosphorinane 2-Oxide (3d): Prepared from dialkyl phosphite 1a
(4), 77 (48), 74 (5), 51 (14). IR (KBr): ν= 3236, 3059, 2969, 2905, and 1-chloro-2-propanone according to the general procedure. The
˜
2876, 1494, 1466, 1499, 1373, 1290, 1210, 1120, 1074, 1019, 998,
crude product was purified by preparative chromatography on sil-
ica gel (CHCl3/MeOH, 35:2) to give product 3d as white crystals
985, 947, 921, 868, 832, 779, 758, 726, 704, 616, 539, 478 cm–1.
C13H18ClO4P (304.707): calcd. C 51.24, H 5.96; found C 49.87, H (43 mg, 66%). M.p. 155–156 °C. [α]1D4 = 0 (c = 1.40, CHCl3). 1H
6.07. The enantiomeric excess values were determined by HPLC
analysis relative to a racemic sample prepared with a quinine/quini-
dine (1:1) as base in the presence of proton sponge. HPLC (Chi-
NMR (400 MHz, CDCl3): δ = 0.96 (s, 3 H, CH3), 1.24 (s, 3 H,
CH3), 1.59 (d, J = 15.0 Hz, PC-CH3), 3.86 (dd, J = 6.0, 11.4 Hz, 1
H, PCCHH), 3.99–4.09 (m, 3 H, PCCHH, 2 POCHH), 4.39–4.43
(m, 2 H, 2 POCHH) ppm. 13C NMR (CDCl3): δ = 20.76 (s, CH3),
21.10 (s, CH3), 21.92 (s, PC-CH3), 32.54 (d, J = 7.6 Hz, POC-Cq),
50.98 (d, J = 15.6 Hz, PCCH2), 73.65 (d, J = 160.1 Hz, P-C), 78.23
(d, J = 6.3 Hz, POCH2), 78.45 (d, J = 6.5 Hz, POCH2) ppm. 31P
NMR (CDCl3): δ = 15.26 ppm. MS (EI): m/z (%) = 150 (26), 149
(3), 105 (2), 92 (11), 83 (28), 69 (23), 68 (100), 67 (11), 57 (5), 56
ralpak AD-H, 10% iPrOH in hexane, 1.0 mLmin–1, 220 nm): tR
=
17.1 (minor), 23.3 (major) min.
5,5-Dimethyl-2-[2Ј-chloro-1Ј-hydroxy-1Ј-(4ЈЈ-chlorophenyl)ethyl]-
1,3,2-dioxaphosphorinane 2-Oxide (3b): Prepared from 5,5-di-
methyl-2-oxo-1,3,2-dioxaphosphorinane and 2,4Ј-dichloroaceto-
phenone according to the general procedure. The crude product
was purified by preparative chromatography on silica gel (CHCl3/
acetone, 9:1), followed by recrystallization (ethyl acetate/hexane) to
give product 3b as colorless flaky crystals (44 mg, 57 %). M.p.
(64), 55 (10). IR (KBr): ν = 517, 668, 691, 736, 790, 831, 869, 894,
˜
917, 948, 960, 993, 1021, 1074, 1117, 1165, 1245, 1374, 1408, 1460,
1474, 2905, 2937, 2967, 3178 cm–1. C8H16ClO4P·1/2H2O (251.644):
calcd. C 38.18, H 6.81; found C 41.33, H 7.15.
1
175 °C. H NMR (400 MHz, CDCl3): δ = 0.85 (CH3), 1.25 (CH3),
3.11 (d, J = 24.9 Hz, OH), 3.89 (ddd, J = 2.0, 10.8, 16.3 Hz, 1 H,
POCHH), 4.07 (ddd, J = 2.0, 10.3, 15.6 Hz, 1 H, POCHH), 4.27–
4.38 (m, 3 H, POCHH, 2 PCCH2), 4.50 (dd, J = 3.4, 10.5 Hz, 1 H,
POCHH), 7.38 (d, J = 8.5 Hz, 1 H, 2 Ar-H), 7.56 (dd, J = 2.1,
8.7 Hz, 1 H, 2 Ar-H) ppm. 13C NMR (CDCl3): δ = 20.53 (s, CH3),
21.94 (s, CH3), 32.50 (d, J = 8.1 Hz, POCCq), 50.67 (d, J =
5,5-Dimethyl-2-(2Ј-chloro-1Ј-hydroxycyclopentyl)-1,3,2-dioxaphos-
phorinane 2-Oxide (3e): Prepared from 5,5-dimethyl-2-oxo-1,3,2-di-
oxaphosphorinane and 2-chlorocyclopentanone according to the
general procedure. The crude product was purified by preparative
chromatography on silica gel (CHCl3/acetone, 3:1) to give product
3e as white crystals (60 mg, 84%). M.p. 191–192 °C. [α]1D7 = 0 (c =
1
13.2 Hz, PCCH2), 78.61 (d, J = 6.6 Hz, POCH2), 79.11–77.68 0.56, CHCl3). H NMR (CDCl3): δ = 1.05 (s, CH3), 1.14 (s, CH3),
(overlapp. d, P-C), 79.44 (d, J = 7.0 Hz, POCH2), 127.6 (s, 2 Ar-
C), 128.7 (s, 2 Ar-C), 134.4 (s, Cq, Ar-C), 135.5 (s, Cq, Ar-C) ppm.
31P NMR (CDCl3): δ = 10.42 ppm. MS (FI): m/z (%) = 338 (2),
305 (1), 304 (4), 303 (3), 302 (11), 190 (65), 188 (100), 151 (14), 149
1.68–1.83 (m, 1 H), 1.90–2.11 (m, 3 H), 2.20–2.44 (m, 3 H), 4.06–
4.25 (m, 4 H, 2 POCH2), 4.53–4.66 (m, 1 H, PCCH) ppm. 13C
NMR (CDCl3): δ = 20.57 (d, J = 10.9 Hz, CH2), 21.17 (s, CH3),
21.81 (s, CH3), 32.58 (s, POC-Cq), 33.85 (d, J = 12.4 Hz, 2 CH2),
(3), 69 (5). IR (KBr): ν = 3226, 2970, 1560, 1491, 1465, 1458, 1374, 64.18 (d, J = 10.3 Hz, PCCH), 76.89 (d, J = 6.8 Hz, CH2, POCH2),
˜
1229, 1207, 1119, 1097, 1073, 1021, 996, 984, 924, 834, 778, 730,
547 cm–1. C13H17Cl2O4P (339.146): calcd. C 46.04, H 5.05; found
C 46.06, H 5.24. The enantiomeric excess values were determined
by HPLC. HPLC (Chiralpak AD-H, 10 % iPrOH in hexane,
1 mLmin–1, 220 nm): tR = 15.54, 16.67 min.
77.53 (d, J = 6.44 Hz, CH2, POCH2), 79.65 (d, J = 169.9 Hz, P-C)
ppm. P NMR (CDCl3): δ = 18.50 ppm. MS (EI): m/z (%) = 150
(27), 149 (3), 120 (5), 118 (20), 96 (2), 86 (3), 83 (55), 72 (34), 69
(10), 68 (100), 67 (5), 62 (12), 59 (74), 56 (74), 55 (82). IR (KBr):
1
ν = 475, 538, 616, 655, 772, 787, 866, 922, 945, 954, 988, 1029,
˜
1081, 1181, 1239, 1293, 1308, 1371, 1388, 1448, 1459, 1460, 1470,
2757, 2874, 2973, 3255 cm–1. C10H18ClO4P·H2O (286.689): calcd.
C 41.90, H 7.03; found C 44.94, H 7.33.
5,5-Dimethyl-2-(2Ј-chloro-1Ј-hydroxy-1Ј-phenylpropyl)-1,3,2-dioxa-
phosphorinane 2-Oxide (3c): Prepared from 5,5-dimethyl-2-oxo-
1,3,2-dioxaphosphorinane and 2-chloropropiophenone according
to the general procedure. The crude product was purified by pre-
parative chromatography on silica gel (CHCl3/acetone, 10:1), fol-
lowed by recrystallization (ethyl acetate/hexane) to give product 3c
Diethyl (2-Chloro-1-hydroxy-1-phenylethyl)phosphonate (3f): Pre-
pared from diethyl phosphite and 2-chloroacetophenone according
to the general procedure. The crude product was purified by pre-
4034
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 4028–4036

Organocatalyzed Synthesis of Tertiary α-Hydroxyphosphonates
parative chromatography on silica gel (CHCl3/acetone, 6:1), fol-
lowed by recrystallization (ether/hexane). Product 3f was obtained
as white crystals (35 mg, 44%). M.p. 105–106 °C. 1H NMR
(400 MHz, CDCl3): δ = 1.07 (t, J = 6.5 Hz, 3 H, CH3), 1.22 (t, J
1.33 (s, 3 H, CH3), 4.04 (dd, J = 9.6, 21.8 Hz, 2 H, POCH2), 4.20
(d, J = 10.4 Hz, 2 H), 5.47 (d, J = 24.5 Hz, 2 H, POCCH2), 7.75
(d, J = 8.7 Hz, 2 H, Ar-H), 8.23 (d, J = 8.7 Hz, 2 H, Ar-H) ppm.
31P NMR (CDCl3): δ = –13.68 ppm. 13C NMR (CDCl3): δ = 20.26
= 6.6 Hz, 3 H, CH3), 3.40 (br. s, 1 H, OH), 3.64–3.81 (m, 1 H, (CH3), 21.77 (CH3), 32.22 (POCH2Cq), 78.25 (2 POC), 96.97
POCHH), 3.81–4.00 (m, 1 H, POCHH), 4.00–4.31 (m, 4 H, 2 (OC=CH2), 125.0 (2 CH, Ph), 128.5 (2 CH, Ph), 129.2 (Cq), 134.1
POCH2, PCCH2), 7.22–7.42 (m, 3 H, 3 Ph-H), 7.54 (d, J = 6.5 Hz, (Cq), 151.8 (Cq) ppm.
2 H, 2 Ph-H) ppm. 13C NMR (CDCl3): δ = 16.20 (s, 2 CH3), 50.57
5,5-Dimethyl-2-[1Ј-(4ЈЈ-nitrophenyl)vinyloxy]-1,3,2-dioxaphosphor-
(d, J = 14.1 Hz, PC-CH2), 63.60 (s, POCH2), 64.02 (s, POCH2),
77.32–74.67 (overlapp. d, P-C), 126.20 (s, 2 CH, Ph), 128.2 (s, 2
CH, Ph), 128.3 (s, Cq, Ph), 136.7 (s, Cq) ppm. 31P NMR (CDCl3):
δ = 19.18 ppm. MS (FI): m/z (%) = 331 (2), 330 (12), 329 (14), 328
(63), 327 (22), 326 (100), 293 (1), 292 (5), 291 (3), 290 (13), 281 (3),
inane 2-Oxide (5i):: This compound was formed as a byproduct in
the reaction between diethyl phosphite and 2-bromo-4Ј-nitroace-
tophenone. 1H NMR (CDCl3): δ = 0.87 (s, 3 H, CH3), 1.32 (s, 3
H, CH3), 3.97 (dd, J = 10.9, 22.2 Hz, 2 H, POCH2), 4.18 (d, J =
10.6 Hz, 2 H), 5.29 (OC=CH2), 7.75 (d, J = 8.7 Hz, 2 H, 2 mCH,
Ph), 8.23 (d, J = 8.7 Hz, 2 H, 2 oCH, Ph) ppm. 31P NMR (CDCl3):
δ = –13.51 ppm. MS (FI): m/z (%) = 315 (1) [M + 2], 314 (4) [M
+ 1], 313 (16) [M]+, 281 (3), 246 (4), 245 (27), 244 (4), 243 (26),
165 (10), 150 (2), 69 (11).
191 (2), 190 (3), 189 (3), 188 (4), 138 (2), 69 (1). IR (KBr): ν =
˜
3271, 2995, 1654, 1492, 1421, 1224, 1091, 1046, 1030, 1003, 982,
872, 800, 722, 576 cm–1. C12H18ClO4P·HCl (329.154): calcd. C
43.79, H 5.82; found C 44.27, H 5.90.
Diphenyl (2-Chloro-1-hydroxy-1-phenylethyl)phosphonate (3g): Pre-
pared from diphenyl phosphite and 2-chloroacetophenone accord-
ing to the general procedure, with the exception that the entire reac-
tion took place at 0 °C instead of at room temperature. The reac-
tion was also quenched at 0 °C. The crude product was purified
by preparative chromatography on silica gel (hexane/EtOAc, 3:1),
followed by recrystallization (hexane/EtOAc). Product 3g was ob-
tained as colorless needles (68 mg, 70%). M.p. 139–141 °C. 1H
NMR (CDCl3): δ = 3.83 (br. s, 1 H, OH), 4.31 (t, J = 11.8 Hz, 1
H, PCCHH), 4.45 (dd, J = 4.8, 11.8 Hz, 1 H, PCCHH), 6.79 (d, J
= 7.9 Hz, 2 H, Ph-H), 7.00–7.22 (m, 6 H, Ph-H), 7.22–7.32 (m, 2
Supporting Information (see footnote on the first page of this arti-
cle): 1H and 13C NMR spectra of compounds 3a–h and selected
HPLC traces.
Acknowledgments
A. M. F. P. is thankful for the financial support of the Fundação
para a Ciência e Tecnologia, MCTES (grant SFRH/BCC/15809/
2206).
H, Ph-H), 7.32–7.45 (m, 3 H, Ph-H), 7.71 (d, J = 7.4 Hz, 2 H, Ph- [1] O. I. Kolodiazhnyi, Tetrahedron: Asymmetry 2005, 16, 3295–
H) ppm. 13C NMR (CDCl3): δ = 50.14 (d, J = 14.1 Hz, PCCH2),
75.62–77.32 (overlapp. d, P-C), 120.3 (s, CH, Ph-C), 120.6 (s, CH,
Ph-C), 125.3 (s, CH, Ph-C), 125.4 (s, CH, Ph-C), 126.5 (s, CH, Ph-
C), 128.5 (s, CH, Ph-C), 129.5 (s, CH, Ph-C), 129.7 (s, CH, Ph-C),
135.8 (s, Cq, Ph-C), 150.2 (s, Cq, Ph-C) ppm. 31P NMR (CDCl3):
δ = 12.33 ppm. MS (EI): m/z (%) = 352 (0.14) [M]+, 234 (11), 170
3340.
[2] J. F. Dellaria Jr., R. G. Maki, H. H. Stein, J. Cohen, D. Whit-
tern, K. Marsh, D. J. Hoffman, J. J. Plattner, T. J. Perun, J.
Med. Chem. 1990, 33, 534–542.
[3] R. F. Frechette, C. Ackerman, S. Beers, R. Look, J. Moore,
Bioorg. Med. Chem. Lett. 1997, 7, 2169–2172.
[4] B. G. Christensen, W. J. Leanza, T. R. Beattie, A. A. Patchett,
B. H. Arison, R. E. Ormond, F. A. Kuehl, G. Albers-Schon-
berg, O. Jardetzky, Science 1969, 166, 123–125.
[5] M. E. Falagas, A. P. Grammatikos, A. Michalopoulos, Expert
Rev. Anti. Infect. Ther. 2008, 6, 593–600
[6] For reviews on the modified Pudovik reaction see, for example:
a) B. Iorga, F. Emery, P. Savignac, Synthesis 1999, 207–224; b)
P. Savignac, B. Iorga, Modern Phosphonate Chemistry, CRC
Press, Boca Raton, 2003, pp. 139–181; for a general review on
epoxide chemistry that also covers epoxyalkylphosphonates,
see, for example: c) E. G. Lewars, “Oxiranes and Oxirenes” in
Comprehensive Heterocyclic Chemistry (Eds.: A. R. Katritzky,
C. W. Rees), Pergamon Press, Oxford, 1984, vol. X, p. 95.
[7] a) B. A. Arbuzov, M. E. Movsesyan, Dokl. Akad. Nauk SSSR,
Otdel. Khim. Nauk 1959, 267 [Chem. Abstr. 1959, 53, 19850f];
b) G. Sturtz, Bull. Soc. Chim. Fr. 1964, 2333.
[8] a) N. Kreutzkampf, H. Kayser, Chem. Ber. 1956, 89, 1614; b)
V. S. Abramov, A. L. Shalman, Zh. V. Molodykh, J. Gen.
Chem. USSR (Engl. Transl.). 1968, 38, 529 [Chem. Abstr. 1968,
69, 27487a]; c) G. A. Russell, F. Ros, J. Am. Chem. Soc. 1985,
107, 2506–2511.
[9] a) B. A. Arbuzov, V. S. Vinogradova, N. A. Polezhaeva, Dokl.
Akad. Nauk SSSR 1958, 121, 641 [Chem. Abstr. 1959, 53,
1180i]; b) B. A. Arbuzov, V. S. Vinogradova, N. A. Polezhaeva,
A. K. Shamsutdinova, Bull. Acad. Sci. USSR, Div. Chem. Sci.
(Engl. Transl.) 1963, 12, 1257; c) B. A. Arbuzov, V. S. Vinogra-
dova, N. A. Polezhaeva, Izv. Akad. Nauk SSSR, Ser. Khim.
1960, 832 [Chem. Abstr. 1960, 54, 24454f]; d) A. Meisters, J. M.
Swan, Aust. J. Chem. 1965, 18, 168–172.
(2), 105 (94), 94 (100), 77 (36), 65 (4), 58 (8). IR (KBr): ν = 476,
˜
494, 524, 604, 617, 688, 706, 768, 899, 922, 951, 974, 1008, 1024,
1072, 1098, 1156, 1179, 1207, 1230, 1261, 1450, 1488, 1588, 1654,
2611, 3049, 3066, 3341, 3386 cm–1. C20H18ClO4P (388.784): calcd.
C 61.79, H 4.67; found C 59.27, H 4.92.
Diphenyl (2-Chloro-1-hydroxy-1-methylethyl)phosphonate (3h): Pre-
pared from diphenyl phosphite and 1-chloro-2-propanone accord-
ing to the general procedure. The crude product was purified by
preparative chromatography on silica gel (hexane/EtOAc, 2:1), fol-
lowed by recrystallization (EtOAc/hexane) to give product 3h as
colorless, sugar-like crystals (58 mg, 57%). M.p. 108–110 °C. 1H
NMR (CDCl3): δ = 1.57 (d, J = 16.4 Hz, 3 H, CH3), 3.78 (dd, J =
7.1, 11.4 Hz, 1 H, PCCHH), 3.93 (dd, J = 9.2, 11.4 Hz, 1 H,
PCCHH), 4.05 (br. s, OH), 7.00–7.14 (m, 6 H, Ph-H), 7.14–7.26
(m, 4 H, Ph-H) ppm. 13C NMR (CDCl3): δ = 20.67 (s, CH3), 49.57
(d, J = 17.1 Hz, PCCH2), 71.93 (d, J = 163.0 Hz, P-C), 120.6 (s, 2
CH, Ph-C), 125.4 (s, CH, pPh-C), 129.7 (s, 2 CH, Ph-C), 150.2 (d,
J = 8.1 Hz, Cq, oPh-C) ppm. 31P NMR (CDCl3): δ = 16.31 ppm.
MS (EI): m/z (%) = 234 (3), 140 (2), 97 (2), 95 (6), 94 (100), 77 (4),
69 (3), 66 (22), 65 (17). IR (KBr): ν = 3307, 1589, 1489, 1430, 1400,
˜
1374, 1258, 1224, 1179, 1168, 1133, 1024, 1008, 938, 770, 689, 644,
616, 590, 513, 466 cm–1. C15H16ClO4P·H2O (344.728): calcd. C
52.26, H 5.26; found C 54.58, H 5.19.
5,5-Dimethyl-2-[1Ј-(phenyl)vinyloxy]-1,3,2-dioxaphosphorinane
2-
Oxide (5a): This compound was formed as a byproduct in reactions
[10] a) B. A. Arbuzov, B. P. Lugovkin, N. P. Bogonostseva, Zh.
Obshch. Khim. 1950, 20, 1468 [Chem. Abstr. 1951, 45, 1506a];
b) A. B. N. Pudovik, Zh. Obshch. Khim. 1955, 25, 2173 [Chem.
between diethyl phosphite and 2-chloroacetophenone, as indicated
1
in the text and tables. H NMR (CDCl3): δ = 0.92 (s, 3 H, CH3),
Eur. J. Org. Chem. 2011, 4028–4036
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4035

M. T. Barros, A. M. F. Phillips
FULL PAPER
Abstr. 1956, 50, 8486i]; c) N. P. Bogonostseva, Uch. Zap. Kazan.
Gos. Univ. 1956, 116, 71 [Chem. Abstr. 1957, 51, 6581g]; d) A.
Arbuzov, V. S. Vinogradova, N. A. Polezhaeva, Izv. Akad. Nauk
SSSR 1959, 41 [Chem. Abstr. 1959, 53, 15035e].
mura, N. Shibata, T. Toru, Adv. Synth. Catal. 2008, 350, 1209–
1212; f) F. Fini, G. Micheletti, L. Bernardi, D. Pettersen, M.
Fochi, A. Ricci, Chem. Commun. 2008, 4345–4347; for the reac-
tion between diethyl phosphite and N-Boc protected imines,
see: g) D. Pettersen, M. Marcolini, L. Bernardi, F. Fini, R. P.
Herrera, V. Sgarzani, A. Ricci, J. Org. Chem. 2006, 71, 6269–
6272; for an example of the phospha-Michael addition of di-
phenyl phosphite to nitroalkenes, see: h) J. Wang, L. D.
Heikkinen, H. Li, L. Zu, W. Jiang, H. Xie, W. Wang, Adv.
Synth. Catal. 2007, 349, 1052–1056.
For a review on organocatalytic asymmetric synthesis of or-
ganophosphorus compounds, see: a) L. Albrecht, A. Albrecht,
H. Krawczyk, K. A. Jørgensen, Chem. Eur. J. 2010, 16, 28–48;
for reviews on cinchona alkaloids in asymmetric catalysis, see:
b) S.-K. Tian, Y. Chen, J. Hang, L. Tang, P. McDaid, L. Deng,
Acc. Chem. Res. 2004, 37, 621–631; c) T. Hashimoto, K. Ma-
ruoka, Chem. Rev. 2007, 107, 5656–5682; for a review on asym-
metric organocatalysis with chiral Brønsted bases that includes
cinchona alkaloids, see: d) C. Palomo, M. Oiarbide, R. López,
Chem. Soc. Rev. 2009, 38, 632–653.
F. Hammerschmidt, S. Schmidt, Eur. J. Org. Chem. 2000, 12,
2239–2245 and references cited therein.
T. Ohwada, Chem. Rev. 1999, 99, 1337–1375.
N. T. Anh, U. E. Eisenstein, Nouv. J. Chim. 1977, 1, 61–70.
S. Krihnamurthy, H. C. Brown, J. Am. Chem. Soc. 1976, 98,
3383–3385.
D. Y. Kim, D. F. Wiemer, Tetrahedron Lett. 2003, 44, 2803–
2805.
[11] See, for example, ref.[10d]
[12] Y. Ding, X. Huang, Heteroat. Chem. 1999, 10, 55–60.
[13] B. Springs, P. Haake, J. Org. Chem. 1976, 46, 1165–1168.
[14] M. A. Ruveda, S. A. De Licastro, Tetrahedron 1972, 28, 6012–
6018.
[15] a) S. Hirai, R. G. Harvey, E. V. Jensen, Tetrahedron 1966, 22,
1625–1640; b) T. Hanaya, K. Yasuda, H. Yamamoto, Bull.
Chem. Soc. Jpn. 1993, 66, 2315–2322 and references cited
therein.
[22]
[16] F. Texier-Boullet, A. Foucaud, Tetrahedron Lett. 1980, 21,
2161–2164.
[17] a) K. Kossev, K. Troev, D. M. Roundhill, Phosphorus, Sulfur,
Silicon Relat. Elem. 1993, 83, 1–7; see also b) A. N. Pudovik,
R. D. Gareev, Zh. Obshch. Khim. 1972, 42, 1861; J. Gen. Chem.
USSR (Engl. Transl.) 1972, 42, 1845; c) R. Nikolova, A. Boji-
lova, N. A. Rodios, Tetrahedron 1998, 54, 14407–14420.
[18] a) A. M. M. M. Phillips, M. J. Mphahlele, A. M. Modro, A. T.
Modro, A. Zwierzak, Phosphorus, Sulfur Silicon Relat. Elem.
1992, 71, 165–174; b) M. T. Barros, A. M. F. Phillips, Tetrahe-
dron: Asymmetry 2010, 21, 2746–2752.
[23]
[24]
[25]
[26]
[19] R. E. Rigsby, K. L. Fillgrove, L. A. Beihoffer, R. N. Arm-
strong, Method. Enzymol. 2005, 401, 367–379.
[20] M. Bella, T. Gasperi, Synthesis 2009, 10, 1583–1614.
[21] Organocatalyzed reactions of dialkyl or diaryl phosphites cata-
lyzed by cinchona alkaloids or modifications thereof: for Pudo-
vik reactions between dialkyl phosphites and aromatic alde-
hydes, see: a) H. Wynberg, A. A. Smaardijk, Tetrahedron Lett.
1983, 24, 5899–5900; b) A. A. Smaardijk, S. Noorda, F.
van Bolhuis, H. Wynberg, Tetrahedron Lett. 1985, 26, 493–496;
c) A. O. Kolodiazhna, V. P. Kukhar, O. I. Kolodiazhnyi, Phos-
phorus, Sulfur Silicon Relat. Elem. 2008, 183, 728–729; for a
Pudovik reaction between dialkyl phosphites and α-keto esters,
see: d) F. Wang, X. Liu, X. Cui, Y. Xiong, X. Zhou, X. Feng,
Chem. Eur. J. 2009, 15, 589–592; for the hydrophosphonylation
of aldimines, see: e) S. Nakamura, H. Nakashima, A. Yama-
[27]
[28]
[29]
R. Dodda, C.-G. Zhao, Org. Lett. 2007, 9, 943–945.
F. Wang, X. Liu, X. Cui, Y. Xiong, X. Zhou, X. Feng, Chem.
Eur. J. 2009, 15, 589–592.
T. Mandal, S. Samanta, G. C.-G. Zhao, Org. Lett. 2007, 9, 943–
945.
[30]
[31]
R. L. McConnell, H. W. Coover Jr., J. Org. Chem. 1959, 24,
630–635.
Received: March 4, 2011
Published Online: May 9, 2011
4036
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 4028–4036