A novel and convenient preparation of hypophosphite esters

Article abstract of DOI:10.1016/S0022-328X(01)01204-9

Only a few methods have been described for the preparation of hypophosphite esters (alkyl phosphinates, ROP(O)H₂). As a result, comparatively few applications have been reported, and these intermediates have not been widely exploited in organop

Full text of DOI:10.1016/S0022-328X(01)01204-9

Journal of Organometallic Chemistry 643–644 (2002) 154–163
www.elsevier.com/locate/jorganchem
A novel and convenient preparation of hypophosphite esters
Sylvine Depre`le, Jean-Luc Montchamp *
Department of Chemistry, Texas Christian Uni6ersity, Box 298860, Fort Worth, TX 76129, USA
Received 25 June 2001; accepted 31 August 2001
Abstract
Only a few methods have been described for the preparation of hypophosphite esters (alkyl phosphinates, ROP(O)H2). As a
result, comparatively few applications have been reported, and these intermediates have not been widely exploited in organophos-
phorus chemistry despite their synthetic potential. Herein, we describe a very general, practical, and high-yielding synthesis of
hypophosphite esters, based on the reaction of hypophosphorous acid and some of its salts with alkoxysilanes. This methodology
solves a number of the problems associated with previously reported reactions. Our esterification takes place at moderate
temperature in a variety of solvents, and under these conditions the esters are remarkably thermally stable. Furthermore, the
reagents employed are inexpensive and do not have to be strictly anhydrous. The scope and limitations of the method are
discussed, and representative synthetic applications for the preparation of monosubstituted phosphinate esters are described.
© 2002 Elsevier Science B.V. All rights reserved.
Keywords: Hypophosphite ester; Alkyl hypophosphite; Alkoxysilane; Phosphinic acid; Organic synthesis
Methyl and ethyl hypophosphites (MeOP(O)H2,
EtOP(O)H2) were first prepared by Kabachnik in 1960
1. Introduction
[1], by the esterification of hypophosphorous acid
Hypophosphite esters (alkyl phosphinates, ROP-
(H3PO2) with diazoalkanes. A few years later, Fitch
(O)H2) are highly sensitive to moisture, air, or heat, and
reported the esterification of crystalline H3PO2 with
have a propensity for disproportionation and decompo-
orthoformates and related compounds [2], and Nifan-
sition. The limited number of methods available for
t’ev described the direct esterification of H3PO2 with
their preparation typically do not combine generality,
alcohols under azeotropic water removal (Dean–Stark
high yield, and experimental simplicity. Therefore,
trap) [3]. To date, the latter two methods are still the
novel methods to form hypophosphite esters in high
most commonly employed, due in part to their relative
yield, from inexpensive and easily handled reagents,
experimental simplicity, and suitability for large scale
and under conditions compatible with a wide array of
reactions. Typically, hazardous crystalline H3PO2 is
subsequent reactions, would be desirable. We now re-
employed, but Fitch’s method allows the use of par-
port one such reaction, in which hypophosphite esters
are formed from hypophosphorous acid (or one of its
tially dried material if an excess of orthoformate is
employed to scavenge the residual moisture. Nifantev’s
salts) and an alkoxysilane (R% Si(OR)4−x).
x
method has the advantage of not requiring any particu-
Not only are hypophosphite esters readily hydrolyzed
lar drying step.
or oxidized compounds, but they also disproportionate
Other preparations entail the alkylation of anhydrous
thermally, thereby preventing their isolation in the pure
sodium hypophosphite with triethyloxonium te-
state via distillation. Even in solution, the esters (partic-
trafluoroborate [4], and the reaction of H3PO2 with
ularly the lower alkyls) rapidly decompose at room
alcohols in the presence of an activating agent (pivaloyl
temperature with formation of a yellow solid of high
chloride) [5]. The transesterification of MeOP(O)H2
phosphorus content. Therefore, the preparation and
handling of hypophosphite esters require great care.
with alcohols has also been studied [6].
However, yields are rarely quantitative in any of the
above described methods. In the orthoformate-based
esterification, side products formation has been exten-
* Corresponding author.
0022-328X/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(01)01204-9

S. Depre`le, J.-L. Montchamp / Journal of Organometallic Chemistry 643–644 (2002) 154–163
155
sively demonstrated [2,7]. In the Dean–Stark protocol,
yields depend on reaction temperature and time [8]. In
our own experience, some hypophosphorous acid al-
ways remains, disproportionation side products appear,
and yields are typically in the 80–90% range. The
presence of acidic components in the reaction mixture
can be problematic in subsequent reactions. The other
methods require the use of hazardous and/or expensive
reagents which are not readily applicable to medium or
large scales.
In terms of synthetic applications, Gallagher [9],
Maier [10], Schwabacher [6b,11], Stawinski [5,12] and a
few others [13] have been major contributors. Some of
the reactions are collected in Scheme 1. Still, the syn-
thetic scope of hypophosphite esters is limited by their
lack of chemical stability, as well as the method em-
ployed in their preparation. For example, reactions
under basic conditions (alkoxide, trialkylamines) have
been reported to yield a number of side products
[9b,14]. Also, reactions requiring heating for only a few
hours at 80 °C can lead to extensive decomposition
[6b,11].
valuable to both our radical reaction, and other syn-
thetic applications.
We already reported a novel method for the esterifi-
cation of monosubstituted phosphinic acids using or-
thosilicates and related compounds (Eq. (1)) [17]. Our
reaction is highly selective for phosphinylidene
(P(O)ꢀH)-containing compounds, and monosubstituted
phosphinic acids are thus esterified while disubstituted
phosphinic acids remain unreacted. It occurred to us
that hypophosphorous acid should also undergo the
reaction. This assumption turned out to be correct, and
even some hypophosphite salts can be esterified. The
results of these investigations are presented herein (Eq.
(2)).
(1)
(2)
As part of our ongoing efforts aiming at the prepara-
tion of biologically active phosphinic acids, we have
undertaken the study of hypophosphites in organic
synthesis [15]. We recently developed the room temper-
ature radical reaction of hypophosphorous derivatives,
which proceeds under conditions even compatible with
hypophosphite esters prepared by the Fitch and Nifan-
t’ev protocols [16]. However, we realized that improve-
ments in the preparation of these esters would be
2. Results and discussion
2.1. Esterification
2.1.1. Reaction of hypophosphite salts and
hypophosphorous acid with orthosilicates
In our previous work with phosphinic acids [17], we
noticed that certain salts could be esterified, provided
that the corresponding phosphinate remained acidic:
the base must have a low pKa, like pyridine or aniline,
whereas stronger bases, like NaOH or Et3N, produce
unreactive phosphinate salts. We have introduced
anilinium hypophosphite (AHP) as an inexpensive and
convenient alternative to other amine salts such as the
commercially available N-ethylpiperidinium hypophos-
phite (EPHP) [15,16]. Anilinium hypophosphite is a
relatively non-hygroscopic crystalline solid which is eas-
ily handled and stored [18], and therefore an excellent
H3PO2 equivalent. In our first experiment, AHP was
reacted with orthosilicates, because we wanted to avoid
the handling of anhydrous H3PO2.
The influence of various reaction parameters was
studied, and the results are shown in Table 1. Reactions
were monitored by 31P-NMR, and the yields deter-
mined by integrating all the resonances in the spectrum.
As expected, the esterification occurred smoothly and
much more readily than with the corresponding anilin-
ium phosphinates (RPO2H2·H2NPh) or phosphinic
acids (RPO2H2) [17]: lower temperatures were sufficient
to deliver hypophosphite esters in excellent yields (typi-
Scheme 1. Synthesis with hypophosphite esters. 1a–c: CF3CO2H (2.5
equivalents), H2N(CH2)3Si(OMe)3 (2.5 equivalents), CH3CN, reflux,
2 h. 1d: (BuO)4Si (2.5 equivalents), cyclohexane, reflux, 2 h. (a) (i)
Cl3CCHO, 60%, [13a]. (ii) Et2NCH2NEt2, 150 °C, 40–60%, [13b].
(iii) CH2ꢁCHCH2OAc, t-BuOOt-Bu, 130–140 °C, 25%, [13c]. (iv)
Diketone, Et3N or heat, 10–40%, [13d]. (v) HC(OMe)3, cat., 20–50%,
[7b,13e]. (vi) RX, i-PrONa, 40–95%, [9b]. (vii) See also (c) and (h).
(b) (i) Acetone, r.t., [2,13l]. (ii) RCHO, AlꢀLiꢀBinol, 37–81%, [13k].
(c) ArI, Pd(0), propylene oxide or N-methylmorpholine, CH3CN,
23–80%, [11,6b]. (d) (i) RX, i-PrONa, 50–90%, [9b]. (ii) Alkene,
Et3B, r.t., 25–80%, [16]. (e) (i) RNH2·HCl, CH2O, Na2CO3, r.t., 55%,
[13g]. (ii) (AcNCH2)3, 100 °C, 77%, [13h]. (f) N-(trimethylsi-
lyl)succinimide, r.t., 57–60%, [13f]. PPh3, DIAD, R%OH, [13m]. (g) S8,
95%, [5,12]. (h) CH2ꢁCH(EWG), base, r.t., 60–88%, [9a,10,13i,13j].

156
S. Depre`le, J.-L. Montchamp / Journal of Organometallic Chemistry 643–644 (2002) 154–163
Table 1
Esterification of hypophosphorous derivatives with orthosilicates a
b
Entry
Hypophosphite
(RO)4Si equivalent
R
Solvent c
31P-NMR % yield d
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
NH4·H2PO2
Et3NH·H2PO2
H3PO2
1.0
0.5
0.25
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
Bu
Bu
Bu
Et
Bu
Et
Me
Bu
Bu
Bu
Bu
Bu
Bu
Et
Me
Bu
Bu
Bu
Et
C6H6
C6H6
C6H6
C6H6
CH3CN
CH3CN
CH3CN
DMF e
C6H12
PhCH3
Dioxane
Pyridine (dry)
THF
95–100
94
41
86
95–100
98
99
75–83
95
93
84–87
61
86–89
89
THF
THF
83
CH3CN
CH3CN
CH3CN
PhCH3
PhCH3
CH3CN
CH3CN
100 f
30
100
92
H3PO2
H3PO2
H3PO2
H3PO2
Et
Ph
Allyl
84 g
88
86
a All reactions were conducted for 2 h, under N2.
b EPHP, 1-ethylpiperidine hypophosphite. H3PO2 was obtained by concentrating a 50 wt.% solution in vacuo.
c Unless otherwise noted, reagent grade solvents were employed at refluxing temperature.
d Determined by integration of all the signals. A range is indicated when multiple runs were conducted. The balance is unreacted starting
material.
e 80–85 °C.
f NH3 evolves.
g Conducted at r.t. for 5 h.
cally 85–100%, Table 1) in no more than 2–3 h. A
variety of solvents could be employed (C6H6, cyclohex-
ane, toluene, THF, dioxane, CH3CN, DMF). Acetoni-
trile and benzene are excellent solvents for this reaction
(entries 1, 4–7, 16, 18, 21, 22). DMF at 80–85 °C was
less satisfactory (entry 8), presumably because its slow
decomposition generates dimethylamine which forms
unreactive Me2NH·H3PO2. Nonetheless, even under
such conditions, the yields typically remained around
80%. Only pyridine did not turn out to be a very useful
esterification solvent (entry 12). The fact that our ester-
ification proceeds well in different solvents is necessary
to use the resulting hypophosphite esters in subsequent
reactions. Reagent grade solvents were used throughout
this study, since they gave results comparable to that of
strictly anhydrous solvents.
Ammonium hypophosphite also underwent esterifica-
tion under our conditions, with concomitant evolution
of ammonia thereby driving the reaction to completion
(Table 2, entry 16). Not surprisingly, triethylammonium
hypophosphite was a poor substrate for this reaction
(entry 17). Hypophosphorous acid itself was esterified
in high yield under the same conditions (Table 1,
entries 18–22). Noteworthy is the fact that the H3PO2
employed was not anhydrous, but simply obtained by
concentrating a commercial aqueous solution in vacuo
for 20–30 min, at room temperature. Esterification
could also take place at room temperature (entry 20).
In fact, most of the reactions already occurred at room
temperature, although heating gave better and more
consistent results.
As shown in Table 1, various tetraalkylorthosilicates
(Si(OR)4, R=Me, Et, Bu, Allyl, Ph) could be em-
ployed. Even 0.5 equivalents of reagent delivered a high
yield of butyl hypophosphite (entry 1 vs. 2), which is
consistent with our previous observations with
RPO2H2. More importantly, the hypophosphite esters
formed appear to be exceptionally thermally stable
under our conditions, as shown in Fig. 1. Even methyl
hypophosphite is stable over at least 8 h in refluxing
acetonitrile using our method, while it is reported to
completely decompose within 1 h at 80 °C with the
orthoformate method [6b,11]. Similarly, EtOP(O)H2
suffers less than 10% decomposition over 1 day in
refluxing acetonitrile. Dialkyl phosphite ((RO)2P(O)H)
does form slowly, and its rate of formation depends on
reaction temperature and mirrors the disappearance of
the alkyl hypophosphite. While ester stability is greatest
in CH3CN, it remains significant in other solvents
(generally less than 20% decomposition in 1 day). Such

S. Depre`le, J.-L. Montchamp / Journal of Organometallic Chemistry 643–644 (2002) 154–163
157
unprecedented thermal stability opens up a number of
possibilities for synthetic applications, provided that the
organosilicon by-products do not interfere.
Reactions can be conducted without difficulty up to a
100 mmol scale. Higher scales should also be appropri-
ate, but this was not checked in this work. Concentra-
tion appears relatively unimportant. Excellent results
were obtained in the 0.2–1.0 M concentration range, so
we conducted the esterification at 0.5 M hypophosphite
concentration as a standard procedure. The reaction
also takes place in neat alkoxysilane reagent at 80 °C,
although the yield is lowered, and heterogeneity can be
problematic. Some of the reactions in Table 1 give
heterogeneous mixtures, especially when AHP was es-
terified in non-polar solvents such as cyclohexane. Al-
though it can be argued that the NMR yield might be
relatively inaccurate in some of these cases, the ob-
served precipitate appears to be related to silicon by-
products and not to phosphorus, as established using
trimethylphosphate as an NMR internal standard. Fur-
Fig. 1. Thermal stability of hypophosphite esters.
thermore, these heterogeneous runs were employed in
subsequent reactions with no significant lowering in the
overall yield.
Table 2
Esterification of hypophosphorous derivatives with alkoxysilanes a
b
Entry Hypophosphite
Alkoxysilane
Solvent 31P-NMR %
yield c
Our reaction parallels Fitch’s method [2], except that
no particular excess of orthosilicate needs to be em-
ployed with concentrated H3PO2, and thermally stable
hypophosphite esters are formed with significantly less
(or no) phosphorus-derived by-products. It is also an
improvement upon Nifant’ev’s method [3], which nor-
mally requires an excess of alcohol, and becomes a
competition between esterification and decomposition
over time. As a result, our novel synthesis of hypophos-
phite esters is a convenient alternative to those previ-
ously reported, and should prove useful in the synthesis
of organophosphorus compounds.
1
2
3
4
5
6
7
8
9
10
11
12
13
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
PhNH3·H2PO2
H3PO2
H3PO2
H3PO2
H3PO2
H3PO2
H3PO2
H3PO2
H3PO2
H3PO2
PrSi(OMe)3
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
C6H12
DMF
CH3CN
EtOH
90–100
40
95–100
85–95
95–100
67
d
PrSi(OMe)3
OctSi(OEt)3
Me2Si(OEt)2
PrSi(OMe)3
PrSi(OMe)3
PrSi(OMe)3
OctSi(OEt)3
OctSi(OEt)3
OctSi(OEt)3
Me2Si(OEt)2
e
70–80
95–100
43
i-PrOH
CH3CN
60 f
e
94–100
85
11
Me2Si(Oi-Pr)2 CH3CN
H2N(CH2)3Si- CH3CN
(OMe)3
2.1.2. Reaction of hypophosphorous acid and anilinium
hypophosphite with alkyltrialkoxysilanes and
dialkyldialkoxysilanes
14
15
16
17
H3PO2
HCl·H2N(CH2 CH3CN
)3Si(OMe)3
HCl·H2N(CH2 PhCH3
)3Si(OMe)3
61
68
H3PO2
Monosubstituted phosphinic acids can also be es-
terified with trialkoxysilanes, although the reaction is
not as general as with orthosilicates [17]. Since hy-
pophosphorous derivatives are significantly more reac-
tive, we focused on alkoxysilanes as esterification
reagents. A wide variety of such organosilicon com-
pounds is available inexpensively or can be prepared
easily. Several alkyltrialkoxysilanes, diethoxydimethyl-
silane, and diisopropoxydimethylsilane were chosen as
representative reagents. The results are collected in
Table 2.
Again, hypophosphite esters were obtained in high
yields from either AHP or concentrated H3PO2, and
acetonitrile proved to be an optimal solvent for these
reactions. The products stability profiles using alkoxysi-
lanes were similar to those obtained using orthosilicate
H3PO2
TFA·H2N(CH2 CH3CN
)3Si(OMe)3
90–95
100 g
NaH2PO2·H2O
(EtO)3SiCl
CH3CN
a All reactions were conducted for 2–3 h, under N2, in refluxing
reagent grade solvent. Unless otherwise noted, one equivalent of
alkoxysilane was employed. DMF was used at 80–85 °C.
b H3PO2 was obtained by concentrating a 50 wt.% solution in
vacuo for 20–30 min.
c Determined by integration of all the signals. A range is indicated
when multiple runs were conducted. The balance is typically unre-
acted starting material.
d 0.5 equivalent.
e Two equivalents were employed.
f i-PrOP(O)H2 is the major product. (Other components:
MeOP(O)H2: 5%; H3PO2: 30%.)
g Run in duplicate.

158
S. Depre`le, J.-L. Montchamp / Journal of Organometallic Chemistry 643–644 (2002) 154–163
reagents (data not shown). Tri- and di-alkoxysilanes
can be useful to replace orthosilicates, particularly in
the case of hazardous tetramethylorthosilicate. As little
as one equivalent R%Si(OR)3 was necessary to bring
about a nearly quantitative reaction. Diethoxydimethyl-
silane and diisopropoxydimethylsilane also worked
well, although in this case, two equivalents were neces-
sary to obtain the corresponding ester in high yield
(entries 4, 11, 12). Again, these findings are consistent
with our previous work on the esterification of mono-
substituted phosphinic acids.
As expected, the alkyl hypophosphite ester is transes-
terified in the presence of another alcohol (Table 2,
entry 10), but the reaction is incomplete, as previously
reported in the literature [9]. Based on our previous
experience [17], we searched for an organosilicon
reagent which could be removed by extraction after a
subsequent reaction. While alkoxysilane by-products
are usually easy to remove by chromatography on silica
gel, purification via a simple extractive work-up might
be more practical. We expected and verified that 3-
aminopropyltrimethoxysilane is unreactive (entry 13) as
it forms a hypophosphite salt of insufficient acidity,
while preformed 3-aminopropyltrimethoxysilane hydro-
chloride did react in moderate yield (entries 14 and 15).
We also found in situ-prepared aminopropy-
ltrimethoxysilane trifluoroacetate to be an excellent es-
terification reagent (entry 16). Some applications of this
convenient system are described in the following
section.
example, Me2SiCl2+ROH) could also be employed.
Future improvements along these lines are underway
and would provide an inexpensive and more general
alternative to the alkylation of anhydrous NaH2PO2
with triethyloxonium tetrafluoroborate [4].
3. Synthetic applications
With a general preparation of hypophosphite esters
at hand, we turned our attention to synthetic applica-
tions. If the organosilicon by-products present in the
reaction mixture are not compatible with a subsequent
synthetic step, our esterification would not be very
useful. Fortunately, this turns out not to be the case,
and some preliminary results are presented below to
illustrate the practical utility of the reaction. In the
following section, all isolated yields are unoptimized.
More systematic studies and novel applications will be
reported in due course. The remarkable thermal stabil-
ity of the hypophosphite esters prepared under our
conditions should allow the development of novel
methodology for the synthesis of organophosphorus
compounds previously not accessible using other
esterifications.
3.1. Room temperature radical reaction of
hypophosphite esters
Our initial motivation to develop a new synthesis of
hypophosphite esters stemmed from our study of hy-
pophosphorous derivatives in radical reactions (Scheme
2) [16]. As discussed earlier, we first used the orthofor-
mate and Dean–Stark methods to study the Et3B-ini-
tiated radical reaction of hypophosphite esters, and
these results are described elsewhere (Scheme 2,
Method A). Here, we focus our attention on the
alkoxysilane-promoted esterification of AHP and
H3PO2 and demonstrate it is an excellent method for
the synthesis of monosubstituted phosphinate esters
(Scheme 2, Method B).
Finally, even commercial NaH2PO2·H2O was directly
esterified without prior drying, using (EtO)3SiCl (entry
17), thus suggesting that other reagent systems (for
In a typical procedure (Scheme 3), a solution of
hypophosphite ester (2–2.5 equivalents) is cooled to
room temperature, treated with an alkene (one equiva-
lent), and Et3B (1 M in hexane, 0.1–1 equivalent) is
then added to the open reaction vessel. After 2–3 h, the
reaction mixture is worked up and purified, affording
the phosphinate ester (see Section 5). The correspond-
ing phosphinic acid is formed in much smaller amounts
than in Method A, because little or no residual H3PO2
contaminates the starting material. Interestingly, aceto-
nitrile was found to be a very good solvent for this
methodology, although it has not been widely used in
radical reactions. Clearly, the organosilicon by-prod-
ucts present in the reaction mixture do not interfere
with the radical process, and the reaction should be
Scheme 2. Room temperature radical reaction of hypophosphite
esters.
Scheme 3. Examples of room temperature radical reactions. 1a–c:
CF3CO2H (2.5 equivalents), H2N(CH2)3Si(OMe)3 (2.5 equivalents),
CH3CN, reflux, 2 h. 1d: (BuO)4Si (2.5 equivalents), cyclohexane,
reflux, 2 h.

S. Depre`le, J.-L. Montchamp / Journal of Organometallic Chemistry 643–644 (2002) 154–163
159
ceeds satisfactorily in several solvents. The broad scope
of this esterification and its demonstrated applicability
to further synthetic manipulations should prove a use-
ful alternative to previously available methodology.
Future directions include further optimization of the
reaction conditions, the preparation of polymer sup-
ported reagents, improvements upon existing reactions
of hypophosphorous esters, and more importantly the
development of novel reactions for the synthesis of
organophosphorus compounds. The unique thermal
stability profile of the hypophosphite esters shows great
promise to achieve the latter objective, and opens up
several new research avenues.
Finally, with the ability to prepare various hypophos-
phite esters in excellent yields, we can focus our atten-
tion on harnessing the previously unrealized potential
of these compounds to conduct diastereo- and enantio-
selective reactions [19]. Desymmetrization of the PꢀH
bonds in ROP(O)H2 should prove feasible and would
provide a simple entry into P-chiral compounds. Work
along these lines is actively being pursued in our labo-
ratory, and results will be disclosed in due course.
Scheme 4. Representative reactions of hypophosphite esters.
useful for the synthesis of monosubstituted phosphinate
esters.
5. Experimental
3.2. Other reactions
Various types of reactions previously described in the
literature (Scheme 1) can also be conducted. A few
representative examples are shown in Scheme 4. Again,
the organosilicon compounds present along with the
ester do not seem to create any difficulty in the subse-
quent step.
Hypophosphite esters prepared with our method un-
dergo (Scheme 4) addition to carbonyls (products 1e,
1f), Michael addition (product 1g), and even palladium-
catalyzed cross-coupling (product 1h). The thermal sta-
bility of the hypophosphite esters suggests that the
latter reaction could be extended to previously unreac-
tive substrates such as aryl bromides. Work along these
lines is currently in progress.
Finally, phenyl hypophosphite (Table 1, entry 21)
was prepared for the first time, and its transesterifica-
tion was demonstrated with the synthesis of menthyl
hypophosphite (Scheme 4, 1i). Transesterification or in
situ preparation of alkoxysilanes from the correspond-
ing chlorides should prove useful to further expand the
scope of hypophosphite esters in organic synthesis.
5.1. General chemistry
1H-NMR spectra were recorded on a Varian XL-300
spectrometer. Chemical shifts for H-NMR spectra are
1
reported (in parts per million) relative to internal te-
tramethylsilane (Me4Si, l=0.00 ppm) with CDCl3 as
solvent. 13C-NMR spectra were recorded at 75 MHz.
Chemical shifts for 13C-NMR spectra are reported (in
parts per million) relative to CDCl3 (l=77.0 ppm) or
C6D6 (l=128.5 ppm). 31P-NMR spectra were recorded
at 121 MHz, and chemical shifts reported (in parts per
million) relative to external 85% phosphoric acid (l=
0.0 ppm). Radial chromatography was carried out with
a Harrison Associates Chromatotron using 1, 2, or 4
mm layers of Silica Gel 60 PF254 containing gypsum (E.
Merck). Ethyl acetate–hexanes mixtures were used as
the eluent for chromatographic purifications. TLC
plates were visualized by immersion in anisaldehyde
stain (by volume: 93% C2H5OH, 3.5% H2SO4, 1%
CH3COOH, and 2.5% anisaldehyde) followed by heat-
ing. Organic solutions of products were dried over
MgSO4, and filtered.
4. Conclusions and future directions
5.2. Reagents and sol6ents
A novel preparation of hypophosphite esters has
been developed. The reaction of hypophosphorous
derivatives with various alkoxysilanes is inexpensive,
does not require strictly anhydrous reagents, and pro-
Organosilicon reagents were purchased from United
Chemical Technologies, Aldrich, or Lancaster, and
were used as received. Caution! Tetramethyl orthosili-
cate (MeO)4Si may cause blindness. Wear appropriate

160
S. Depre`le, J.-L. Montchamp / Journal of Organometallic Chemistry 643–644 (2002) 154–163
protection. Sodium hypophosphite hydrate, and aq.
hypophosphorous acid (50 wt.%), were obtained from
Aldrich and used as received. Hypophosphorous acid
was concentrated in vacuo on a rotary evaporator, at
room temperature (r.t.) for 20–30 min before reaction.
Triethylammonium hypophosphite was prepared ac-
cording to Stawinski et al. [5]. Ammonium hypophos-
phite was prepared as described in Ref. [21]. Unless
otherwise noted, HPLC grade or reagent grade solvents
were used throughout. When anhydrous solvents were
used, they were prepared as follows: dioxane was dried
was filtered and the off-white crystalline precipitate was
washed with cold CH3COCH3. The filtrate was concen-
trated under reduced pressure, and a second crop of
crystalline hypophosphite was obtained by adding ace-
tone. (Note: if desired, a third crop can be collected
similarly, or alternatively, the mother liquor is concen-
trated to dryness and dried in vacuo. This crop is very
slightly impure but can still be used.) The first two
crops were combined and washed with ether, then dried
in vacuo over P2O5 for 24 h. Anilinium hypophosphite
(304 g, 91%) was obtained as light yellow needles (m.p.
113–114 °C). It can be stored at r.t. for several months
,
over activated 4 A molecular sieves, and stored under
1
N2; pyridine and triethylamine were distilled from
without affecting its reactivity profile. H-NMR (D2O)
,
CaH2 and stored under N2 over activated 4 A molecu-
l 8.10–8.45 (br s, 3H), 7.12 (d, J=520 Hz, 2H),
7.0–7.4 (m, 5H); 31P-NMR (D2O) l 3.7 (t, JPꢀH=520
Hz). Anilinium hypophosphite was also reported previ-
ously [22].
lar sieves; tetrahydrofuran (THF) was distilled under
N2 from sodium benzophenone ketyl, and used immedi-
ately; benzene was distilled immediately before use,
from CaH2 under N2; diisopropylethylamine was dis-
tilled from CaH2 and stored under N2. Anhydrous
acetonitrile and DMF were obtained after drying over
5.5. Esterification: representati6e procedure (Tables 1
and 2)
,
activated 3 A molecular sieves, and were stored under
N2.
In a typical procedure, a solution (or suspension) of
the hypophosphorous compound (5 mmol), alkoxysi-
lane (5 mmol), in solvent (10 ml) is refluxed for 2 h
under N2. At that time, the yield is determined by
31P-NMR.
5.3. 31P-NMR yield measurements
NMR yields were determined by integration of all
the 31P signals. Thermal stability studies (Fig. 1) were
conducted by monitoring the reactions at timed inter-
val. Some representative NMR data is collected in
Table 3.
5.6. Synthetic applications
Isolated yields are unoptimized.
5.4. Anilinium hypophosphite
5.6.1. Room temperature radical reactions (Scheme 3)
Anilinium hypophosphite was prepared as previously
described by us [15]. Aniline (196 g, 2.1 mol) was added
over 30 min via an addition funnel, to an ice-cold aq.
solution of H3PO2 (50 wt.%, 278 g, 2.1 mol). The light
brown solution rapidly turned into a thick slurry. This
5.6.1.1. Methyl (2-pi6aloyloxy-ethyl)phosphinate (1a).
Concentrated H3PO2 (initially 50 wt.% aq. solution,
0.678 g, 5.1 mmol) was dissolved in CH3CN (HPLC
grade, 20 ml) and the reaction flask was equipped with
a reflux condenser and placed under N2. Trifluoroacetic
Table 3
Representative 31P-NMR data of hypophosphite esters in different solvents a
Hypophosphite
Chemical shifts (l) (coupling constants JPꢀH (Hz))
CH3CN
C6H6–toluene
THF–dioxane
DMF
Other
MeOP(O)H2
EtOP(O)H2
19.5 (566, 13)
16.2 (565, 10)
16.5 (563, 10)
17.6 (592)
17.2 (568, 11)
13.0 (572, 10)
15.0 (572, 11)
15.5 (558, 13)
12.5 (564, 9)
13.2 (558, 9)
–
–
17.0 (563, 13)
13.3 (561, 10)
14.5 (562, 9)
–
–
19.2 (566, 13)
16.0 (564, 10)
16.3 (564, 9)
–
–
19.9 (575, 13) b
14.1 (570) b
15.6 (570) b
–
BuOP(O)H2
PhOP(O)H2
AllylOP(O)H2
i-PrOP(O)H2
MenOP(O)H2
–
9.5 (561) c
11.3 (563, 11) d
–
–
–
a Esters obtained under a variety of conditions (hypophosphite starting material, concentration, alkoxysilane), generally had chemical shifts
within 0.5 ppm of the given value. ROP(O)H2 coupling patterns: RꢁMe: tq; R=Et, Bu, Allyl: tt; RꢁPh, t; Rꢁi-Pr, Men, td.
b According to Ref. [2]. The esters were obtained by the orthoformate method.
c According to Ref. [9b].
d Prepared with the Dean–Stark method, but using cyclohexane as the solvent, see Ref. [16].

S. Depre`le, J.-L. Montchamp / Journal of Organometallic Chemistry 643–644 (2002) 154–163
161
acid (0.39 ml, 5.1 mmol), and 3-aminopropy-
ltrimethoxysilane (0.90 ml, 5.1 mmol) were then added
successively at r.t. An exothermic reaction immediately
took place, and the mixture was heated to reflux. After
2 h, the reaction mixture was allowed to cool to r.t. The
flask was open to air, vinyl pivalate (0.30 ml, 2 mmol)
was added neat via syringe, followed by triethylborane
(1 M in hexanes, 2.0 ml, 2 mmol), and the heteroge-
neous mixture was stirred in air at r.t. for 2 h. 31P-
NMR analysis indicated the product at l 39 (86%). The
crude reaction mixture was treated with EtOAc and aq.
KHSO4. The organic layer was washed successively
with saturated aq. NaHCO3 (1×), and brine (1×).
Drying, concentration, and purification by radial chro-
matography (4 mm thickness, EtOAc–hexane 1:1, v/v,
EtOAc) afforded 1a (0.215 g, 55%) as a colorless oil:
1H-NMR (CDCl3) l 7.20 (d, J=548 Hz, 1H), 4.25–
4.45 (m, 2H), 3.83 (d, J=12 Hz, 3H), 2.15–2.35 (m,
2H), 1.21 (s, 9H); 13C-NMR (CDCl3) l 177.6, 57.2, 52.7
(d, JPOC=7 Hz), 38.4, 28.5 (d, JPC=94 Hz), 26.8;
31P-NMR (CDCl3) l 36.4 (dm, JPꢀH=548 Hz).
mm thickness, EtOAc–hexane 1:5, 1:1, v/v, EtOAc)
afforded 1d (0.689 g, 65%) as a colorless oil: H-NMR
1
(CDCl3) l 7.06 (d, J=530 Hz, 1H), 7.1–7.3 (m, 5H),
3.9–4.15 (m, 2H), 2.63 (t, J=15 Hz, 2H), 1.55–1.85
(m, 8H), 1.3–1.45 (2H), 0.94 (t, J=7 Hz, 3H); 13C-
NMR (CDCl3) l 141.5, 128.2, 125.7, 65.9 (d, JPOC=7
Hz), 35.2, 32.2 (d, JPOCC=6 Hz), 32.0 (d, JPCC=16
Hz), 28.5 (d, JPC=94 Hz), 20.2 (d, JPCCC=3 Hz), 18.6,
13.4; 31P-NMR (CDCl3) l 39.5 (d, JPꢀH=530 Hz).
5.6.2. Addition to carbonyls (Scheme 4)
5.6.2.1. Ethyl(phenylhydroxymethyl)phosphinate (1e). A
mixture of concentrated H3PO2 (initially 50 wt.% aq.
solution, 1.370 g, 10.4 mmol), and octyltriethoxysilane
(2.855 g, 10.3 mmol), in CH3CN (HPLC grade, 20 ml),
was refluxed for 2 h under N2 and cooled to r.t.
Benzaldehyde (0.96 ml, 9.4 mmol) and anhydrous i-
Pr2NEt (1.80 ml, 10.3 mmol) were added via syringe to
the cloudy reaction mixture. Stirring was continued at
r.t. for 22 h. 31P-NMR analysis indicated the product as
a 1:1 mixture of diastereoisomers (l 38.0 and 35.5) in
81% combined yield. The cloudy, biphasic mixture ob-
tained after concentration was partitioned between
EtOAc and aq. KHSO4. The organic layer was washed
with saturated aq. NaHCO3 (1×), and brine (1×).
Drying, concentration, and purification by radial chro-
matography (4 mm thickness, EtOAc–hexane 1:5, 1:1,
v/v, EtOAc) afforded 1e (0.992 g, 52%, 1:1 mixture of
Compounds 1b and 1c were prepared similarly.
5.6.1.2. Methyl cyclohexylphosphinate (1b). 1H-NMR
(CDCl3) l 6.80 (dd, J=521, 2 Hz, 1H), 3.78 (d, J=11
Hz, 3H), 1.6–2.0 (m, 6H), 1.15–1.5 (m, 5H); 13C-NMR
(CDCl3) l 52.8 (d, JPOC=8 Hz), 37.4, 36.1, 35.7 (d,
J
PC=89 Hz), 25.0–26.2 (multiple peaks, could not be
deconvoluted), 23.9 (d,
J
PCC=6 Hz); 31P-NMR
1
(CDCl3) l 46.7 (d, JPꢀH=521 Hz).
diastereoisomers) as a colorless oil: H-NMR (CDCl3)
l 7.7–7.8 (0.5H), 7.2–7.5 (m, 5H), 5.8–6.0 (0.5H),
5.5–5.9 (br, 1H), 4.8–5.0 (m, 1H), 3.8–4.1 (m, 2H),
1.1–1.3 (m, 3H); 13C-NMR (CDCl3) l 135.2 (2), 128.4,
128.1 (2), 127.9, 127.5 (2), 127.4, 126.8, 126.7 (2), 71.7
(d, JPC=98 Hz), 71.3 (d, JPC=110 Hz), 63.4 (d,
5.6.1.3. Methyl (3-(tert-butoxycarbonylamino)propyl)-
phosphinate (1c). H-NMR (CDCl3) l 7.09 (d, J=533
Hz, 1H), 4.93 (bs, 1H), 3.79 (d, J=12 Hz, 2H), 3.15–
3.25 (m, 2H), 1.7–1.9 (m, 4H), 1.44 (s, 9H); 13C-NMR
(CDCl3) l 155.9, 52.8 (d, JPOC=7 Hz), 40.5 (d,
1
J
POC=7 Hz), 63.1 (d, JPOC=8 Hz), 30.1 (d, JPCC=7
J
J
PCCC=6 Hz), 40.4, 28.3, 25.8 (d, JPC=94 Hz), 21.4 (d,
Hz), 29.6 (d, JPCC=6 Hz), 16.1 (d, JPOCC=6 Hz), 16.0
PCC=3 Hz); 31P-NMR (CDCl3) l 40.5 (dm, JPꢀH
=
(d, JPOCC=7 Hz); 31P-NMR (CDCl3) l 36.6 (d, JPꢀH
554 Hz), 32.6 (d, JPꢀH=549 Hz).
=
521 Hz).
5.6.1.4. Butyl (4-phenylbutyl)phosphinate (1d) [17]. Con-
centrated H3PO2 (initially 50 wt.% aq. solution, 1.385 g,
10.5 mmol) and tetrabutoxysilane (3.233 g, 10.1 mmol)
were taken up in cyclohexane (reagent grade, 20 ml).
The resulting white suspension was refluxed for 2.5 h,
under N2. The reaction mixture was allowed to cool to
r.t. 4-Phenyl-1-butene (0.63 ml, 4.2 mmol) was added
neat via syringe, followed by triethylborane (1 M in
hexanes, 4.2 ml, 4.2 mmol). The rubber septum was
removed and the reaction mixture was stirred in air at
r.t. for 2 h. 31P-NMR analysis indicated the product at
l 39.3 (76%). The crude reaction mixture was treated
with EtOAc (100 ml) and aq. HCl (1 N, 20 ml). The
organic layer was washed successively with saturated
aq. NaHCO3 (1×), and brine (1×). Drying, concen-
tration, and purification by radial chromatography (4
5.6.2.2. Ethyl (1-hydroxy-cyclohexyl)phosphinate (1f).
Concentrated H3PO2 (initially 50 wt.% aq. solution,
0.678 g, 5.1 mmol) and diethoxydimethylsilane (1.372 g,
9.3 mmol) were dissolved in CH3CN (HPLC grade, 10
ml). The resulting colorless solution was refluxed for 2
h, under N2, then cooled to r.t. Cyclohexanone (0.54
ml, 5.2 mmol), and anhydrous Et3N (0.72 ml, 5.2
mmol) were added, and the reaction mixture was
refluxed for 2 h. EtOAc and aq. NaHSO4 were added.
The organic layer was washed successively with satu-
rated aq. NaHCO3 (1×), and brine (1×). Drying,
concentration, and purification by radial chromatogra-
phy (4 mm thickness, hexane, EtOAc–hexane 1:1, v/v,
EtOAc) afforded 1f (0.310 g, 31%) as a colorless liquid:
1H-NMR (CDCl3) l 6.70 (d, J=530 Hz, 1H), 4.1–4.3
(m, 2H), 1.2–1.9 (m, 10H), 1.36 (t, J=7 Hz, 3H);

162
S. Depre`le, J.-L. Montchamp / Journal of Organometallic Chemistry 643–644 (2002) 154–163
13C-NMR (CDCl3) l 70.5 (d, JPC=115 Hz), 62.8 (d,
POC=8 Hz), 30.1 (d, JPCC=7 Hz), 29.6 (d, JPCC=6
Hz), 25.33, 19.8 (d, JPCCC=3 Hz), 19.7, 16.2 (d,
POCC=5 Hz); 31P-NMR (CDCl3) l 41.3 (d, JPꢀH=530
Acknowledgements
J
Acknowledgment is made to the Robert A. Welch
Foundation (P-1435), and to the donors of The
Petroleum Research Fund, administered by the ACS
(34334-G1 and 36915-AC1), for the generous support
of this research.
J
Hz).
5.6.3. Conjugate addition (Scheme 4)
5.6.3.1. Ethyl (2-benzyloxycarbonyl-ethyl)phosphinate
(1g). A mixture of concentrated H3PO2 (initially 50
wt.% aq. solution, 0.657 g, 5.0 mmol), and octyltri-
ethoxysilane (1.6 ml, 5.0 mmol), in CH3CN (HPLC
grade, 10 ml), was refluxed for 2 h under N2, then
cooled to r.t. Benzyl acrylate (0.75 ml, 4.9 mmol) and
1,1,3,3-tetramethylguanidine (0.63 ml, 5.0 mmol) were
added, and the cloudy reaction mixture was stirred for
3 h. 31P-NMR analysis showed the product at l 38.1.
The solution was partitioned between EtOAc and aq.
KHSO4. The organic layer was washed successively
with saturated aq. NaHCO3 (1×), and brine (1×).
Drying, concentration, and purification by radial chro-
matography (4 mm thickness, EtOAc–hexane 1:5, 1.1,
v/v, EtOAc) afforded 1g (0.819 g, 64%) as a colorless
oil: 1H-NMR (CDCl3) l 5.3–5.4 (bs, 5H), 7.19 (d,
J=545 Hz, 1H), 5.13 (s, 2H), 3.95–4.2 (m, 2H), 2.6–
2.8 (m, 2H), 2.0–2.2 (m, 2H), 1.33 (t, 7 Hz, 3H);
13C-NMR (CDCl3) l 171.4 (d, JPCCC=13 Hz), 135.2,
128.3, 128.1, 128.0, 66.6, 62.3 (d, JPOC=7 Hz), 25.8
(JPCC=3 Hz), 23.2 (JPC=95 Hz), 15.9 (JPOCC=6 Hz);
31P-NMR (CDCl3) l 36.4 (dm, JPꢀH=545 Hz).
References
[1] M.I. Kabachnik, A.E. Shipov, T.A. Mastryukova, Bull. Acad.
Sci. USSR 1 (1960) 138.
[2] S.J. Fitch, J. Am. Chem. Soc. 86 (1964) 61.
[3] E.E. Nifant’ev, L.P. Levitan, J. Gen. Chem. USSR 35 (1965)
762.
[4] H.W. Pinnick, M.A. Reynolds, Synth. Commun. 9 (1979) 535.
[5] J. Stawinski, M. Thelin, E. Westman, R. Zain, J. Org. Chem. 55
(1990) 3503.
[6] (a) M.J. Gallagher, H. Honegger, J. Chem. Soc. Chem. Com-
mun. (1978) 54;
(b) A.W. Schwabacher, A.D. Stefanescu, Tetrahedron Lett. 37
(1996) 425.
[7] (a) M.J. Gallagher, H. Honegger, Tetrahedron Lett. 34 (1977)
2987;
(b) M.J. Gallagher, H. Honegger, Aust. J. Chem. 33 (1980) 287;
(c) M.J. Gallagher, R. Garbutt, L.Y. Hua, G.H. Lee, Phospho-
rus Sulfur Silicon 75 (1993) 201.
[8] I. Devedjiev, V. Ganev, R. Stefanova, G. Borisov, Phosphorus
Sulfur 31 (1987) 7.
[9] (a) M.J. Gallagher, J. Sussman, Phosphorus 5 (1975) 91;
(b) M.J. Gallagher, M.G. Ranasinghe, I.D. Jenkins, Phosphorus
Sulfur Silicon 115 (1996) 255;
(c) M.J. Gallagher, M.G. Ranasinghe, I.D. Jenkins, J. Org.
Chem. 61 (1996) 436.
[10] L. Maier, Helv. Chim. Acta 56 (1973) 489.
[11] H. Lei, M.S. Stoakes, A.W. Schwabacher, Synthesis (1992) 1255.
[12] J. Jankowska, J. Cieslak, A. Kraszewski, J. Stawinski, Tetrahe-
dron Lett. 38 (1997) 2007.
[13] (a) B.E. Ivanov, L.A. Kudryavtseva, Bull. Acad. Sci. USSR 7
(1968) 1544 for examples on the use of hypophosphite esters;
(b) B.E. Ivanov, L.A. Kudryavtseva, Bull. Acad. Sci. USSR 7
(1967) 1447;
5.6.4. Pd-catalyzed cross-coupling (Scheme 4)
5.6.4.1. Butyl phenylphosphinate (1h) [17,20]. A solution
of anilinium hypophosphite (0.952 g, 6 mmol) and
tetrabutoxysilane (1.933 g, 6 mmol) in CH3CN (12 ml)
was refluxed for 2 h, under N2. After cooling to r.t.,
iodobenzene (0.25 ml, 2 mmol), anhydrous Et3N (0.30
ml, 2 mmol), and Cl2Pd(PPh3)2 (0.025 g, 0.04 mmol),
were added successively. The reaction mixture was then
refluxed for 5 h. At that point, the black mixture was
concentrated under reduced pressure, and the residue
partitioned between EtOAc and aq. KHSO4. The or-
ganic layer was washed successively with saturated aq.
NaHCO3 (1×), and brine (1×). Drying, concentra-
tion, and purification by radial chromatography (4 mm
thickness, hexane, EtOAc–hexane 1:1, v/v, EtOAc) af-
(c) I.A. Aleksandrova, L.I. Ufimtseva, Bull. Acad. Sci. USSR 6
(1971) 1218;
(d) V.I. Vysotskii, A.S. Skobun, M.N. Tilichenko, J. Gen. Chem.
USSR 49 (1979) 1721;
(e) M.V. Livantsov, A.A. Prishchenko, I.F. Lutsenko, J. Gen.
Chem. USSR 56 (1986) 2195;
(f) M.V. Livantsov, A.A. Prishchenko, I.F. Lutsenko, J. Gen.
Chem. USSR 56 (1986) 1976;
(g) I.A. Natchev, Phosphorus Sulfur 37 (1988) 133;
(h) I.A. Natchev, Liebigs Ann. Chem. (1988) 861;
(i) C.G. Caldwell, S.P. Sahoo, S.A. Polo, R.R. Eversole, T.J.
Lanza, S.G. Mills, L.M. Niedzwiecki, M. Izquierdo-Martin, B.C.
Chang, R.K. Harrison, D.W. Kuo, T.-Y. Lin, R.L. Stein, P.L.
Durette, W.K. Hagmann, Bioorg. Med. Chem. Lett. 6 (1996)
323;
(j) A.E. Wroblewski, J.G. Verkade, J. Am. Chem. Soc. 118
(1996) 10168;
(k) T. Yamagishi, T. Yokomatsu, K. Suemune, S. Shibuya,
Tetrahedron 55 (1999) 12125;
1
forded 1h (0.300 g, 80%) H-NMR (CDCl3) l 7.81 (d,
J=7 Hz, 1H), 7.76 (J=7 Hz, 1H), 7.58 (d, J=562 Hz,
1H), 7.55–7.6 (m, 1H), 7.45–7.55 (m, 2H), 3.95–4.15
(m, 2H), 1.6–1.8 (m, 2H), 1.35–1.5 (m, 2H), 0.92 (t,
J=7 Hz, 3H); 13C-NMR (CDCl3) l 132.4 (d, JPCCCC
=
3 Hz), 130.3 (d, JPCCC=12 Hz), 129.4 (d, JPC=132
Hz), 128.1 (d, JPCC=14 Hz), 65.1 (d, JPOC=7 Hz),
31.8 (JPOCC=6 Hz), 18.2, 12.9; 31P-NMR (CDCl3) l
25.3 (dm, JPꢀH=563 Hz).
(l) E.K. Baylis, Tetrahedron Lett. 36 (1995) 9389;
(m) I.D. Grice, P.J. Harvey, I.D. Jenkins, M.J. Gallagher, M.G.
Ranasinghe, Tetrahedron Lett. 37 (1996) 1087.

S. Depre`le, J.-L. Montchamp / Journal of Organometallic Chemistry 643–644 (2002) 154–163
163
[14] (a) C.J.R. Fookes, M.J. Gallagher, J. Chem. Soc. Perkin Trans.
1 (1975) 1876;
(b) C.J.R. Fookes, M.J. Gallagher, H. Honegger, J. Chem. Soc.
Chem. Commun. (1978) 324.
[15] J.-L. Montchamp, Y.R. Dumond, J. Am. Chem. Soc. 123 (2001)
510.
[19] To our knowledge, this has not been investigated. Shibuya
and coworkers ([13k]) have conducted the enantioselective
hydrophosphinylation of aldehydes, and obtained good
enantiomeric excesses at the a-carbinol center, but did
not observe any induction at the chiral phosphorus
center.
[16] S. Depre`le, J.-L. Montchamp, J. Org. Chem. 66 (2001) 6745.
[17] Y.R. Dumond, R.L. Baker, J.-L. Montchamp, Org. Lett. 2
(2000) 3341.
[18] Anilinium hypophosphite is handled in air, and can be stored for
several months at room temperature in a dark glass container.
[20] H. Dahn, V.V. Toan, M.N. Ung-Truong, Magn. Reson. Chem.
30 (1992) 1089.
[21] J.-L. Montchamp, F. Tian, J.W. Frost, J. Org. Chem. 60 (1995)
6076.
[22] H. Schmidt, Chem. Ber. 81 (1948) 477.