Dual Radical/Polar Pudovik Reaction: Application Field of New Activation Methods

Article abstract of DOI:10.1021/jo9622441

The Pudovik reaction (addition of organophosphorus compounds containing a labile P-H bond with alkenes and alkynes) can progess via a radical or (and) ionic mechanism. A comparative and systematic study including various reagents and different activation methods (heating, photochemical or ultrasonic irradiation, and dry medium supported reactions) is presented. Photolysis is the most efficient method for the radical processes, but in a few examples, ultrasonic irradiation can be more appropriate since the reaction time is shorter and ultrasound did not induce side-reactions (in particular Z/E isomerization). Dry medium process on basic solid support is the best anionic activation (yield, time, selectivity, purification facilities). Ultrasound, by its mechanical effects, can contribute to increase yield compared to the classical thermal method under these basic conditions. All the activation methods are efficient whatever the unsaturated substrates for the phosphine reactivity, whereas the appropriate activation method is exclusively determined by the nature of the unsaturated system for the thiophosphine (or phosphine oxide) reactivity.

Full text of DOI:10.1021/jo9622441

2414
J . Org. Chem. 1997, 62, 2414-2422
Du a l Ra d ica l/P ola r P u d ovik Rea ction : Ap p lica tion F ield of New
Activa tion Meth od s
Delphine Semenzin,† Guita Etemad-Moghadam,*,† Dominique Albouy,† Ousmane Diallo,‡ and
Max Koenig*,†
FU14, Universite´ Paul-Sabatier, Baˆt. 2R1, 118 route de Narbonne, 31062 Toulouse cedex 04, France, and
Laboratoire de Chimie de Coordination, 205 route de Narbonne, 31075 Toulouse cedex, France
Received December 3, 1996X
The Pudovik reaction (addition of organophosphorus compounds containing a labile P-H bond with
alkenes and alkynes) can progess via a radical or (and) ionic mechanism. A comparative and
systematic study including various reagents and different activation methods (heating, photochemi-
cal or ultrasonic irradiation, and dry medium supported reactions) is presented. Photolysis is the
most efficient method for the radical processes, but in a few examples, ultrasonic irradiation can
be more appropriate since the reaction time is shorter and ultrasound did not induce side-reactions
(in particular Z/ E isomerization). Dry medium process on basic solid support is the best anionic
activation (yield, time, selectivity, purification facilities). Ultrasound, by its mechanical effects,
can contribute to increase yield compared to the classical thermal method under these basic
conditions. All the activation methods are efficient whatever the unsaturated substrates for the
phosphine reactivity, whereas the appropriate activation method is exclusively determined by the
nature of the unsaturated system for the thiophosphine (or phosphine oxide) reactivity.
Sch em e 1. P u d ovik Rea ction
In tr od u ction
The Pudovik reaction is one of the most versatile
pathways for the formation of carbon-phosphorus bonds
(Scheme 1) and involves the addition of compounds
containing a labile P-H bond with unsaturated systems
(alkenes, alkynes, carbonyls, imines).1,2
The products of the reaction find significant applica-
tions in a wide range of areas (industrial, biological, and
chemical synthetic uses).3 The Pudovik reaction can be
considered as a “hybrid mechanism model” since it may
proceed via an ionic and/or a radical mechanism, depend-
ing upon the structure of the unsaturated substrates and
the phosphorus reagents and the experimental condi-
tions. Although this reaction has been widely studied, a
comparative study including various reagents and par-
ticularly different new activation methods has not been
systematically done.
The choice of the various methods of activation was
made according to the polar and/or radical character of
the Pudovik reaction.
Method A: heterogeneous alkaline medium subjected
to magnetic stirring and classical heating [∆/KOH/H2O/
CH3CN].
Method B: heterogeneous alkaline medium subjected
to ultrasonic irradiation [))))/KOH/H2O/CH3CN].
Method C: basic solid support in dry medium [Al2O3/
KOH].
We now wish to report our studies of the Pudovik
reaction with a series of alkenes and alkynes either being
activated by electron-withdrawing substituents (C(O)-
OEt, C(O)Me, CN) or having nucleophilic character (Ph,
alkyl), in the presence of various compounds containing
a labile P-H bond such as diphenyl(thio)phosphine and
diethyl phosphonate (Scheme 2).
Method D: homogeneous medium subjected to ultra-
sonic irradiation [))))].
Method E: homogeneous medium in the presence of a
free radical initiator under classical heating [∆/AIBN].
Method F: homogeneous medium in the presence of a
free radical initiator subjected to ultrasonic irradiation
[))))/AIBN].
Method G: homogeneous medium subjected to photo-
chemical irradiation at 300 nm [hν].
Resu lts a n d Discu ssion
The addition reactions of phosphorus reagents 1a -d
to ethylenic 2-4 and acetylenic 5-7 compounds were
performed in either heterogeneous (methods A-C) or
homogeneous medium (methods D-G). All the products
obtained 2′-7′a -d , 5′′-7′′a -d and 8 (Scheme 2) were
isolated, and their respective yields are reported in the
Experimental Section. However, in order to rigourously
compare the efficiency of the various activation methods,
the initial yields were obtained by 31P NMR analysis of
the crude reaction mixtures prior to purification.
I. In flu en ce of th e n a tu r e of p h osp h or u s r e-
a gen ts. (a ) Dieth yl P h osp h on a te (1a ). The addition
of diethyl phosphonate (1a ) to R,â-unsaturated carbonyl
compounds such as benzalacetone (2) in homogeneous
These reactions were carried out in homogeneous or
heterogeneous medium under various activation methods
(heating, photochemical or ultrasonic irradiation, sup-
ported reactions in dry medium).
† Universite´ Paul-Sabatier.
‡ Laboratoire de Chimie de Coordination.
X Abstract published in Advance ACS Abstracts, March 15, 1997.
(1) (a) Kirby, A. J .; Warren, S. G. The Organic Chemistry of
Phosphorus; Elsevier Publishing Co.: New York, 1967. (b) Emsley, J .;
Hall, D. The Chemistry of Phosphorus; Harper & Row Publishers: New
York, 1976. (c) Wolfsberger, V. W. Chem. Zeit. 1988, 53.
(2) Pudovik, A. N.; Konovalova, I. V. Synthesis 1979, 81.
(3) (a) Ullmann’s Encyclopedy of Industry Chemistry; VHC: New
York, 1991; Vol 17 and 19. (b) Akso, DE 2 708 790, 1977 (Kowallik, J .;
Brandner, A.).
S0022-3263(96)02244-X CCC: $14.00 © 1997 American Chemical Society

Dual Radical/Polar Pudovik Reaction
J . Org. Chem., Vol. 62, No. 8, 1997 2415
Sch em e 2. Gen er a l Sch em e
Sch em e 3. Regioselective Syn th esis of 8 (P a th w a y
a ) a n d 2′a [P a th w a ys a (Two-Steps) a n d b
(On e-Step)]
min at 20 °C, giving rise to the double addition product
5′′a in 100% yield, regardless of the stoichiometry of the
reagents.
The same method (method C) produced 3′a after
addition of diethyl phosphonate (1a ) with acrylonitrile
(3) in the yield of 81% whereas the yield did not exceed
33% by method B.
The addition of the diethyl phosphonate (1a ) to the
activated unsaturated compounds (benzalacetone (2),
acrylonitrile (3), and ethyl propiolate (5)) occured in the
presence of a basic catalyst (KOH) and particularly in
the presence of this basic catalyst on solid support
(method C). No reaction of diethyl phosphonate (1a ) was
observed with the nonactivated unsaturated substrates
(diphenylethylene (4), phenylacetylene (6), and hexyne
(7)) regardless of the method of activation (under our
standard conditions) (Figure 1).
(b ) Dip h en ylt h iop h osp h in e (1b ). The diphenyl-
thiophosphine (1b) underwent addition reactions with the
activated systems such as benzalacetone (2), acrylonitrile
(3), and ethyl propiolate (5) under anionic activation by
three methods (A-C) (Figure 2).
medium is not regioselective, since it occurs at either the
CdO or the CdC double bond, giving rise to a mixture
of R-hydroxy phosphonate 8 and â-keto phosphonate 2′a .4
A similar approach to the synthesis of R-hydroxy phos-
phonates involving the reaction of dialkyl phosphonates
with nonactivated carbonyl reagents in dry medium was
reported.5 When the addition of diethyl phosphonate (1a )
to benzalacetone (2) was performed on basic solid support
in dry medium (method C), we observed that the regi-
oselectivity and the ratio of the adducts were dependent
on the reaction time and the temperature (Scheme 3).
Thus, the 1,2-adduct 8 was instantly obtained on Al2O3/
KOH at 20 °C in 100% yield. The quantitative formation
of the 1,4-adduct 2′a required the use of Al2O3/KOH as
the solid support at 100 °C for 5 min. The addition
product 8 was the kinetic product and rearranged to the
thermodynamic product 2′a . This observation was veri-
fied by heating the R-hydroxy phosphonate 8 (in solution
or directly adsorbed on solid support in dry medium);
under these conditions, we obtained quantitatively the
corresponding â-keto phosphonate 2′a .
The best yields of the reactions of diphenylthiophos-
phine (1b) with benzalacetone (2) and acrylonitrile (3)
were obtained using method C (supported reactions in
dry medium), affording the 1,4-adduct 2′b and the
product 3′b in yield of 90% and 100% respectively, unlike
the reaction with diethyl phosphonate (1a ) and benzal-
acetone (2), no 1,2-adduct was detected.
The hydrophosphonylation of ethyl propiolate (5) by
diethyl phosphonate (1a ) in homogeneous medium usu-
ally affords a mixture of single and double addition
products in low yields.6 In our case, this reaction proved
to be regioselective on Al2O3/KOH in dry medium for 5
However, with ethyl propiolate (5), method B gave the
best results, giving rise to a mixture of the single addition
product 5′b and double addition product 5′′b in the
overall yield of 80%. Diphenylthiophosphine (1b) also
reacted with the activated substrates such as acryloni-
trile (3) and ethyl propiolate (5) under photochemical
(4) Arbuzov, B. A.; Zoroastrova, V. M.; Trudii, G. A.; Fuzhenkova,
A. V. Bull. Acad. Sci. USSR 1974, 2541-2542.
(5) Texier-Boullet, F.; Villemin, D.; Ricard, M.; Moison, H.; Foucaud,
A. Tetrahedron 1985, 41, 1259.
(6) Pudovik, A. N.; Kusovleva, R. G. Zh. Obshch. Khim. 1965, 35
(2), 354-358.

2416 J . Org. Chem., Vol. 62, No. 8, 1997
Semenzin et al.
F igu r e 3. Reactivity of Ph2PH (1c).
F igu r e 1. Reactivity of (EtO)2P(O)H (1a ).
Sch em e 4. Equ ilibr iu m betw een (Th io)p h osp h ite
h (Th io)p h osp h on a te
reaction of 1b with hexyne (7) did not take place under
sonication in the absence of AIBN (method D) but
afforded 7′b in the yield of 85% with the Z/ E ratio of
95/5 using method F [))))/AIBN]. The same reaction only
afforded 7′b in the yield of 60% by classical heating
(method E) with the Z/ E ratio of 75/25.
In summary, the diphenylthiophosphine (1b) presents
a dual reactivity toward unsaturated systems since
addition reactions occur under either radical- or base-
catalyzed conditions, depending upon the nature of the
multiple bond. A heterogeneous ionic process takes place
with electron-deficient alkenes and alkynes (benzalac-
etone (2), acrylonitrile (3), and ethyl propiolate (5)),
whereas addition reactions with electron-rich unsatur-
ated substrates (diphenylethylene (4), phenylacetylene
(6) and hexyne (7)) proceed via a radical mechanism in
homogeneous medium.
F igu r e 2. Reactivity of Ph2P(S)H (1b).
irradiation (method G) to give the adducts 3′b and 5′b
in the same yield of 40%. In the case of the vinylic
phosphine 5′b, the Z/ E ratio changed according to the
method used: Z/ E ) 30/70 with method B and 55/45
with method G. It must also be noted that the double
addition product 5′′b was only obtained in basic medium
(vide infra).
(c) Dip h en ylp h osp h in e (1c). Diphenylphosphine
(1c) underwent addition reactions with almost all the
unsaturated systems under base- or radical-catalyzed
conditions (Figure 3).
Diphenylthiophosphine (1b) could not be reacted with
the electron-rich unsaturated compounds (diphenyleth-
ylene (4), phenylacetylene (6), and hexyne (7)) under
anionic activations but it underwent addition reactions
under radical activations instead (Figure 2). The adducts
4′b, 6′b, and 7′b were obtained under photochemical
irradiation at 300 nm (method G) for 6-7 h in good yields.
In the case of the phenylacetylene (6), the reaction led to
the formation of the vinylic thiophosphine 6′b in the yield
of 60% under ultrasound irradiation in homogeneous
medium for 1 h (method D). The yield of 6′b under
sonication was improved (>90%) by adding 3 equiv of
AIBN (method F), whereas the same reaction under
classical thermal method (method E) gave 6′b in only
20%. It must also be noted that the photochemical
reaction of diphenylthiophosphine (1b) with phenylacety-
lene (6) (method G) led to a photostationary equilibrium
of the two isomers 6′b (Z/ E ) 70/30). Interestingly, the
However, the presence of the phosphorus electron lone
pair makes diphenylphosphine (1c) particularly sensitive
to oxidation reactions, especially under ultrasonic ir-
radiation, with formation of the diphenylphosphine oxide
(1d ), which is much less reactive than diphenylphosphine
(1c).
(d ) Sca le of th e Rea ctivity. It is now well-estab-
lished that the hydrophosphoryl compounds and their
sulfur analogs exist as two tautomeric forms in equilib-
rium, but the equilibrium is almost completely shifted
to the phosphoryl form7 (Scheme 4).
However, both entities can react as nucleophiles in
their (thio)phosphite and (thio)phosphonate forms in
basic media. Their P-H acidities vary over a wide range,
(7) (a) Nylen, P.Z. Anorg. Allgem. Chem. 1938, 235, 161. (b) Doack,
G. O.; Freedman, L. D. Chem. Rev. 1961, 61, 31. (c) Luz, Z.; Silver, B.
J . Am. Chem. Soc. 1961, 83, 4518.

Dual Radical/Polar Pudovik Reaction
J . Org. Chem., Vol. 62, No. 8, 1997 2417
depending on the nature of substituents on the phospho-
rus atom. The thiophosphoryl derivatives are seen to be
stronger acids than their corresponding oxygen analogs
and their pK values are a few units lower (in DMSO):8
(EtO) P(O)H
[pK ) 20.8]
Ph2P(O)H
[pK ) 20.7] [pK ) 12.8]
Ph2P(S)H
<
<
2
Furthermore, the (thio)phosphoryl group is more elec-
tronegative than trivalent phosphorus, and R2P(X)H
compounds are stronger acids than phosphines [pKPh
2PH
) 23.1]. Their anions are ambident nucleophiles, R2PX-
u R2P(X)-, but they rarely react at the oxygen or the
sulfur atom. In the case of diphenyl(thio)phosphine, the
“aryl effect” (the electron-acceptor effect of the phenyl
groups) influences the partial negative charge on the
phosphorus atom and stabilizes the anion. From the
results described in Figures 1-3, an order of reactivity
for the labile P-H reagents 1a -d can be proposed:
F igu r e 4. Comparative studies of the reactivity of the P-H
bond upon basic solid support, the “dry medium process”
(method C).
(EtO)2P(O)H < Ph2P(O)H < Ph2P(S)H < Ph2PH
(method B), since ultrasound is known to be more
effective in the case of heterogeneous reactions12,13 and
standard procedure (method A).
The order of the reagents in the scale of the reactivity
corresponds to the one in the acidity scale except for
diphenylphosphine, which is out of line dramatically.
This probably reflects the difference in nucleophilicity
between all the corresponding anions. The same trend
of reactivity was observed with the corresponding radi-
cals. Although, the abstraction of the labile hydrogen
atom bonded to phosphorus is a similar process in both
trivalent and tetrahedral compounds, it seems that the
lower chain transfer constants for tetrahedral compounds
reflect the lower stability of the (thio)phosphinyl radicals
[R2P(X)•] vs the phosphino radicals (R2P•).1a,9 Conse-
quently, it can be assumed that secondary phosphines
(e.g. Ph2PH) are more reactive than hydro(thio)phospho-
ryl analogs. Furthermore, it has been shown that
thiophosphinyl radicals (X ) S) are more stable than
phosphinyl radicals (X ) O).1b
II. In flu en ce of th e Activa tion Meth od s. a ) An -
ion ic Activa tion . It is well-known that a wide variety
of chemical reactions can be promoted in heterogeneous
medium due to the presence of acidic and/or basic sites
located on the surface of suitable solids such as alumina,
clay, silica gel, and talc.10 The advantages of such “dry
media” reactions over their corresponding homogeneous
counterparts are significant. The chemical transforma-
tions take place with better regiospecificity11 and ste-
reospecificity, the conditions are milder, and the isolation
of products is easier. The anionic activation of the
Pudovik reaction was carried out with potassium hy-
droxide supported on alumina in the absence of solvent
(“dry medium process”, method C), under sonication
Among the methods used (methods A-C), the dry
medium process can be considered as the best method
for the anionic activation of the P-H labile bond toward
alkenes and alkynes bearing strong electron-withdrawing
substituents. Figure 4 correlates the efficiency of the
method C and the nature of the unsaturated substrate.
Several solid mineral supports (neutral or activated
basic alumina and lamellar structures such as talc) were
tested. Commercial basic alumina was not efficient
compared to the more basic potassium hydroxide impreg-
nated upon alumina. The use of KOH supported on
alumina rendered the Pudovik reaction fast and re-
giospecifically. It also made the workup easier: the
reagents were adsorbed on the solid support, the reaction
occurred almost instantly at room temperature and a
simple extraction using the appropriate solvent followed
by filtration directly afforded the addition products. It
is important to note that the amount of water adsorbed
is critical since the completion and the selectivity of the
reaction heavily depended on the degree of dryness of the
basic supports.14 We finalized the cyanoethylation reac-
tion of the phosphine PH3swell-known in homogeneous
medium15 or in heterogeneous medium (liquid-liquid)13sin
heterogeneous “gas/solid” medium. We studied the influ-
ence of the varying degrees of dryness of solid support
(Al2O3/KOH in the ratio of 2.5/1), upon the progress of
the reaction (Scheme 5).
The thermogravimetric analysis of the solid support
obtained as a fine powder after evaporation for 1.5 h
under reduced pressure (15-20 mm) indicated the pres-
ence of 4% (Figure 5a) of physically adsorbed water
whereas the same analysis of another support obtained
as a granular powder (30 min under 15-20 mm) revealed
21% of water (Figure 5b). Our experimental results
(8) (a) Petrov, E. S.; Terekhova, M. I.; Malakhova, I. G.; Tsvetkov,
E. N.; Shatenshtein, A. I.; Kabachnick, M. I. J . Gen. Chem. USSR 1979,
49, 11, 2410. (b) Matveeva, A. G.; Terekhova, M. I.; Aladzheva, I. M.;
Odinets, I. L.; Petrov, E. S.; Mastryukova, T. A.; Kabachnick, M. I.
Russ. J . Gen. Chem. 1993, 63, No. 3, Part. 2.
(9) Smith, U.; Geiger, F.; Mu¨ller, A.; Markau, K. Angew. Chem., Int.
Ed. Engl. 1963, 2, 400.
(12) (a) Moon, S.; Duchin, L.; Cooney, J . V. Tetrahedron Lett. 1979,
3917. (b) Luche, J . L.; Damaino, J . C. J . Am. Chem. Soc. 1980, 102,
7927. (c) Boudjouk, K.; Han, B. H. J . Org. Chem. 1982, 47, 5030.
(13) (a) Etemad-Moghadam, G.; Rifqui, M.; Layrolle, P.; Berlan, J .;
Koenig, M. Tetrahedron Lett. 1991, 32, 5965. (b) Semenzin, D.; Etemad-
Moghadam, G.; Albouy, D.; Koenig, M. Tetrahedron Lett. 1994, 35,
3297. (c) Mirza-Aghayan, M.; Etemad-Moghadam, G.; Zaparucha, A.;
Berlan, J .; Loupy, A.; Koenig,M. Tetrahedron Assym. 1995, 6, 2643.
(14) Texier-Boullet, F.; Foucaud, A. Tetrahedron Lett. 1982, 4927.
(15) Rauhut, M. M.; Hechenbleikner, I.; Currier, H. A.; Shaeffer, F.
C.; Wystrach, V. P. J . Am. Chem. Soc. 1959, 81, 1103.
(10) (a) Lazlo, P. Preparative Chemistry Using Supported Reagents;
Academic: San Diego, 1987. (b) Clark, J . H.; Kybett, A. P.; Macquarrie,
D. J . Supported Reagents: Preparation, Analysis and Applications;
VHC: New York, 1992. (c) Smith, K. Solid Supports And Catalysts In
Organic Synthesis; Ellis Horwood-PTR Prentice Hall: New York, 1992.
(11) (a) Scheidecker, S.; Semenzin, D.; Etemad-Moghadam, G.;
Sournies, F.; Koenig, M.; Labarre, J . F. Phosphorus Sulfur 1993, 80,
85. (b) Sournies, F.; Graffeuil, M.; Segues, B.; Crasnier, F.; Semenzin,
D.; Etemad-Moghadam, G.; Koenig, M.; Labarre, J . F. Phosphorus
Sulfur 1994, 86, 1.

2418 J . Org. Chem., Vol. 62, No. 8, 1997
Semenzin et al.
F igu r e 7. Comparative studies of the reactivity of the P-H
F igu r e 5. Thermogravimetric analysis of Al2O3/KOH (2.5/1)
solid support after evaporation [(a) 1.5 h and (b) 30 min] under
reduced pressure (15-20 mmHg).
bond under photochemical irradiation at 300 nm (method G).
of unsaturated compounds was not strongly polarized
(Figure 7).
It is known that ultrasound can promote a single
electron transfer and favor the radical mechanism.16
Method D [))))] and method F [))))/AIBN] allowed us to
test this observation with respect to classical methods
for the generation of radicals such as methods E [∆/AIBN]
and G [hν].
From these results (Figures 1-3), the photochemical
activation (method G) was very efficient for the radical
addition of diphenyl(thio)phosphine (1b ,c) to electron-
rich alkenes and alkynes (Figure 7). However, the
reaction required a long irradiation time, bringing about
the isomerization of the corresponding vinylic (thio)-
phosphine adducts. No reaction took place with diethyl
phosphonate (1a ), regardless of the unsaturated sub-
strates 2-7 and the radical activation methods (meth-
ods D-G). This observation may appear paradoxical in
the light of the known examples of free radical reactivity
of diethyl phosphonate (1a )2,17 described by Pudovik, but
our standard chosen methods were less drastic. Furth-
ermore, the reaction mechanism depends on the nature
of the unsaturated substrates since, in our recent stud-
ies,18 we have shown that the addition of diethyl phos-
phonate (1a ) with imines was improved by ultrasonic
activation. The radical mechanism of this reaction was
confirm by EPR19 while the addition of the same com-
pound 1a with aldehydes20 was anionic.
F igu r e 6. Comparative studies of the reactivity of the P-H
bond upon sonication in basic heterogeneous medium (method
B).
Sch em e 5. Cya n oeth yla tion Rea ction of
P h osp h in e P H3
The cavitational effects of ultrasonic irradiation depend
on many physical parameters (viscosity, boiling point,
vapor pressure, macroscopic temperature, etc.).21,22 It
seems likely that the sonochemical effects for the studied
reaction mixtures (1a with 2-7) are not sufficiently
effective to initiate the addition reactions. It was de-
showed that the reaction only occurred with the solid
support containing a lesser amount of water (4%).
The sonication of a heterogeneous basic medium
(method B) enhanced the addition rate and improved the
yield with respect to the standard procedure (method A).
Figure 6 shows the scale of the reactivity of phosphorus
compounds using method B.
Furthermore, method B constituted an efficient method
for the preparation of the double addition product in the
reaction of diphenylphosphine (1c) with phenylacetylene
(6): the oxidized double addition adduct 6′′d was ob-
tained in a yield of 40% along side with the corresponding
single addition product 6′c (16%) and its oxidized analog
6′d (12%). For the same reaction, the dry medium
method afforded 6′′d in only 16% yield.
(16) Reitz, P. Ultrasonics 1987, 31.
(17) (a) Pudovik, A. N.; Konovalova, I. V. Zh. Obshch. Khim. 1959,
29, 3342. (b) Pudovik, A. N.; Konovalova, I. V. Zh. Obshch. Khim. 1960,
30, 2348.
(18) Hubert, C.; Oussaid, B.; Etemad-Mogadham, G.; Koenig, M.;
Garrigues, B. Synthesis 1994, 51.
(19) Hubert, C.; Munoz, A.; Garrigues, B.; Luche, J . L. J . Org. Chem.
1995, 60, 1488.
(20) Oussaid, B.; Diallo, O.; Soufiaoui, M.; Garrigues, B. Bull. Soc.
Chim. Fr. 1995, 132, 461.
(21) Margulis, M. A. Russ. J . Phys. Chem. 1985, 59, 6.
(22) (a) Suslick, K. S. Adv. Organomet. Chem. 1986, 25, 33. (b) De
Souza Barboza, J . C.; Petrier, C.; Luche, J . L. J . Org. Chem. 1988, 53,
1212. (c) Ley, S. V. C. R. M. Low Ultrasound in Synthesis; Springer
Velag: Berlin, 1989.
(b) Ra d ica l Activa tion . Free radical addition reac-
tions were very efficient when the double or triple bond

Dual Radical/Polar Pudovik Reaction
J . Org. Chem., Vol. 62, No. 8, 1997 2419
scribed that the use of high frequency of ultrasound for
radical reactions23 or the simultaneous action of two
acoustic vibration frequencies21 is much more effective
than the action of each frequency separately. Therefore,
we used an ultrasonic bath at 536 kHz or the coupling of
the two frequencies (20 and 536 kHz) for radical genera-
tion. However, both methods offered no improvement of
yields. This nonreactivity of diethyl phosphonate (1a )
with alkenes and alkynes under radical activation could
be due to the nature of the couple “P-H reagent/
unsaturated substrate” and consequently to the addition
reaction rate of the phosphinyl radical [(EtO)2P(O)•] with
unsaturated substrate. Actually, the reaction rate is
higher for the imines18 than for our unsaturated sub-
strates 2-7.
philic character with diphenylphosphine or diphenylth-
iophosphine. When the phosphorus atom bears an
oxygen, the radical addition reaction in our particular
experimental conditions (less drastic than the literature
conditions) does not occur with ethylenic or acetylenic
compounds. Photoactivation is the most efficient method,
but the reaction time is long and entails often an
isomerization of the vinyl(thio)phosphines obtained. Ul-
trasound-induced radical reactions in homogeneous me-
dium is mostly unpredictable because several parameters
have to be taken into account such as the nature of the
reagents, the solvent’s viscosity and boiling point, and
the temperature, but it can be superior to photochemistry
since the reaction time is shorter and only the cis
configuration is obtained by this method.
The addition reaction of diphenylthiophosphine (1b)
with hexyne did not take place under sonication and the
presence of a radical initiator is required. In contrast,
the reaction of diphenylthiophosphine (1b) with pheny-
lacetylene (6) can occur under sonication in toluene
without AIBN. The influence of the nature of solvent
upon the latter reaction is observed: although the
dichloroethane is known to generate radicals under
ultrasonic irradiation,24 the presence of this solvent
rendered the sonication ineffective in our case and the
use of toluene became necessary. Consequently, the
chemical effects of ultrasound activation in homogeneous
medium remain uncertain and unpredictable as a func-
tion of the chemical and physical properties of the
reaction mixture. However, the sonication in homoge-
neous medium and in the presence of a radical initiator
can become an appropriated method versus the photo-
chemical activation for the synthesis of vinylic phospho-
rus derivatives since the reaction time is reduced to 1 h
and the stereoselectivity of the reaction reaches much to
95% in favor of the Z isomers.
Under the conditions described in this work, the P-H
reagents regioselectively add to unsaturated compounds
by nucleophilic or free-radical mechanism. In the pres-
ence of free-radical initiators or under photochemical or
ultrasound irradiation in homogeneous medium, the
reagents 1a -d add to double and triple bonds by a free-
radical mechanism and the orientation is anti-Markovni-
kov. In basic medium, the addition occurs by nucleophilic
attack of phosphorus anions to olefins or alkynes by a
Michael type mechanism and the orientation is also anti-
Markovnikov, the nucleophile going to the least-substi-
tuted carbon and the electrophile going to the most-
substituted position. In fact, the orientation cannot be
used as a diagnostic tool to indicate which mechanism is
operating.
When an anionic activation is usedsparticularly when
the unsaturated system is activated by an electron-
withdrawing substituentsthe dry media process on basic
solid support is the best method. The contribution of
ultrasound to ionic reactions in heterogeneous medium
(liquid/liquid) is mainly due to the mechanical effects
(homogenization, emulsification). Under these condi-
tions, the reaction’s rate, selectivity, and yield are
increased in relation to classical process.
In conclusion, this comparative study of the different
activation methods allowed us to define the scope of
application depending upon the nature of the unsatur-
ated substrate (activated or not), the nature of the
organophosphorus compound (phosphorus atom oxidized
or not), and the experimental conditions:
When the phosphorus atom bears an oxygen or sulfur
atom, the domains of the reactivity are well-separated.
When the substrate used is an unsaturated system
having a nucleophilic character, the mechanism is pref-
erentially radical. When the unsaturated compound
contains an electron-withdrawing substituent, the main
mechanism is ionic. Then, we notice that the mechanism
may be either radical or ionic, depending upon the nature
of the substrate, whatever the experimental method used.
When the phosphorus atom is not oxidized (in the case
of the diphenylphosphine), the fields of the radical/ionic
mechanisms are not distinguishable; there is a duality
of the mechanisms and the coexistence of these two
processes brings about a competition which depends on
the experimental conditions.
Exp er im en ta l Section
Gen er a l Com m en ts. The spectra were recorded using the
following instruments: IR spectra, Beckman IR10 or Perkin-
Elmer 225; 1H, 13C, and 31P spectra, Brucker AC80 or 250WM;
and mass spectra in the electron impact mode, Varian Mat
311A. Column chromatographic separations were executed at
normal pressure using Merck silica 60F (70-230 mesh).
Melting points were measured in open glass capillaries with
a Bu¨chi-Tottoli apparatus. Elemental analyses were per-
formed by the Microanalytical Service Laboratory of the “Ecole
Nationale Superieure de Chimie” of Toulouse, France. UV
irradiation (300 nm) was performed on Rayonet RPR100
apparatus. The ultrasound generator was a 13 mm probe
fitted to a Bioblock Vibracell 600W (20 kHz) generator dipped
into the solution.
Con clu sion
It appears in the literature that the differences of the
Pudovik reaction mechanisms depend upon the nature
of the unsaturated substrates. Although the comparison
between the different activation methods used must be
made wisely because each method is not optimized for
each reaction in order to keep the standard conditions,
our results show that the phosphorus atom environment
allows us to define more precisely the reactional domains.
Radical activation is mainly favorable to the addition
reaction between unsaturated systems having a nucleo-
Diethyl phosphonate (1a ), benzalacetone (2), acrylonitrile
(3), 1-hexyne (7) (Aldrich), diphenylphosphine (1c), 1,1-diphe-
nylethylene (4), ethyl propiolate (5) (Fluka) and phenylacety-
lene (6) (Lancaster) were used as received without prior
purification. Solid supports were purchased from J anssen
(Al2O3, 50-200 µm).
(23) Frenzel, H.; Schultes, H. J . Phys. Chem. B. 1935, 27, 421.
(24) Rifqui, M.; Etemad-Moghadam, G.; Berlan, J .; Koenig, M. Main
Group Chem. 1996, in press.
When reactions were performed via methods A-D, materi-
als were obtained from commercial suppliers and used without

2420 J . Org. Chem., Vol. 62, No. 8, 1997
Semenzin et al.
MHz, CDCl3) δ 25.8 (m); H NMR (80.13 MHz, CDCl3) δ 4.13
1
further purification. For methods E-G, reactions were carried
out under an argon atmosphere with dry distilled solvents
under anhydrous conditions.
Gen er a l P r oced u r es. Meth od s A [∆/KOH] a n d B [))))/
KOH]. To a solution of phosphorus compounds 1a -d (0.73
mmol) in acetonitrile (3 mL) was added potassium hydroxide
(1.24 mmol) in distilled H2O (2.42 mmol). An unsaturated
system 2-7 (0.73 mmol, except 2.9 mmol for acrylonitrile (3))
was added to the solution. The mixture was stirred at 70 °C
(or sonicated at 70 °C) for 1 h.
Meth od C [Al2O3/KOH]. A 1:5 ratio of phosphorus reactant/
support and a 1:1.7 molar ratio of phosphorus reactant/base
was used. The following general procedure was used: potas-
sium hydroxide supported on alumina was added to a solution
of phosphorus compounds 1a -d (0.73 mmol) and an unsatur-
ated substrate 2-7 (0.73 mmol, except 2.9 mmol for acryloni-
trile (3)) dissolved in a minimum of CH2Cl2 (3-4 mL) at room
temperature. The solvent was evaporated under reduced
pressure during 1.5 at 15-20 mmHg and the adduct was
extracted by an appropriate solvent (specified for each product
in the description of the purification).
Meth od s D [∆/AIBN], E [))))] a n d F [))))/AIBN]. To a
solution of an unsaturated compound 2-7 (0.73 mmol, except
2.9 mmol for acrylonitrile (3)) in distilled solvent (4 mL) placed
in air-dried material was added phosphorus reagents 1a -d
(0.73 mmol). For the methods D and F, AIBN (2.9 mmol) was
added to this mixture. The resulting solution was stirred and
heated or sonicated at 70 °C.
3
3
(qd, J HH ) 7.1 Hz, J HP ) 8.3 Hz, 4H), 2.79-2.44 (m, 2H),
2.26-1.85 (m, 2H), 1.33 (t, 3J HH ) 7.1 Hz, 6H); 13C NMR (62.90
3
2
MHz, C6D6) δ 118.77 (d, J CP ) 15.1 Hz), 62.51 (d, J CP ) 6.1
Hz), 21.84 (d, 1J CP ) 144.9 Hz), 16.37 (d, 3J CP ) 5.8 Hz), 11.27
(d, J CP ) 4.1 Hz); IR (CDCl3, cm-1) 2244 (CN), 1245 (PdO);
2
MS (EI) m/ z 192 (M + 1)+•, 136 [(M - CH2CH2CN - 1)]+.
O,O′-Dieth yl (1-Hyd r oxy-1-m eth yl-3-p h en yl-2-p r op e-
n yl)p h osp h on a te (8). Method C [Al2O3/KOH] was used.
Products were extracted with acetone and 8 was purified by
chromatography on silica (eluent: n-hexane/AcOEt, 1:9) to
yield 63%: yellow oil;4 Rf 0.25 (n-hexane/AcOEt, 1:9); 31P NMR
(32.44 MHz, C6D6) δ 23.8 (m); 1H NMR (250.13 MHz, C6D6) δ
7.37-7.35 (m, 2H), 7.20 (d, 3J HH ) 4.5 Hz, 1H), 7.16-7.01 (m,
3H), 6.66 (dd, 3J HH ) 4.5 Hz,3J HP ) 15.9 Hz, 1 H), 6.37 (s, 1H),
3
3
4.18-4.04 (m, 4H), 1.84 (d, J HP ) 15.8 Hz, 3H), 1.08 (t, J HH
) 7.3 Hz, 3H), 1.05 (t, J HH ) 7.3 Hz, 3H); 13C NMR (62.90
3
4
MHz, C6D6) δ 137.65 (d, J CP ) 3.0 Hz), 131.08 (s), 130.10 (d,
2J CP ) 10.5 Hz), 128.88-127.01 (m), 73.07 (d, J CP ) 161.3
1
2
2
Hz), 63.56 (d, J CP ) 7.4 Hz), 62.92 (d, J CP ) 7.8 Hz), 24.45
(d, J CP ) 2.3 Hz), 16.62 (s), 16.53 (s); IR (C6D6, cm-1) 3257
2
(OH), 1603 (CdC), 1240 (PdO); MS (EI) m/ z 284 (M)+•, 103
(CHdCHPh)+, 29 (CH3CH2)+. Anal. Calcd for C14H21O4P: C,
59.15; H, 7.45. Found: C, 59.01; H, 7.38.
3,3-Bis(d ieth oxyp h osp h in yl)p r op -2-a n oic Acid Eth yl
Ester (5′′a ). Method C [Al2O3/KOH] was used to obtain 5′′a .
The solid support was extracted with acetone and 5′′a was
purified by chromatography (SiO2; eluent:CH2Cl2/n-hexane,
1:9) to give 60% of 5′′a as a yellow oil:6 Rf 0.26 (CH2Cl2/n-
hexane, 1:9); 31P NMR (32.44 MHz, C6D6) δ 21.1 (m); 1H NMR
Meth od G [h ν/300 n m ]. A phosphorus compound 1a -d
was dissolved in 0.3 mL of deuteriated solvent and was added
to an unsaturated system 2-7 (0.73 mmol, except 2.9 mmol
for acrylonitrile (3)). This solution was irradiated at 300 nm
(the reaction time and the nature of the solvent are specified
for each product in the description of the purification).
Dip h en ylp h osp h in e Su lfid e (1b). To a solution of diphe-
nylphosphine (1c) (1 g, 5.4 mmol) in distilled toluene (10 mL)
was added S8 (173 mg, 5.4 mmol) in air-dried material. This
solution was stirred and heated at 80 °C for 2.5 h. The solvent
was removed and the crude product 1b was recrystallized
from CH3CN to yield 1.05 g of white solid (90%): mp 98-99
°C (from CH3CN);25 31P NMR (32.44 MHz, C6D6) δ 20.9 (d, 1J PH
3
3
(200.13 MHz, C6D6) δ 4.11 (qd, J HH ) 8.0 Hz, J HP ) 9.0 Hz,
3
3
3
4H), 4.07 (qd, J HH ) 8.0 Hz, J HP ) 9.0 Hz, 4H), 3.94 (d, J HH
3
2
) 8.0 Hz, 2H), 3.44 (tt, J HH ) 6.0 Hz, J HP ) 24.0 Hz, 1H),
3
2
3
3.07 (dt, J HH ) 6.0 Hz, J HP ) 16.0 Hz, 2H), 1.07 (t, J HH
)
8.0 Hz, 6H), 1.06 (t, 3J HH ) 8.0 Hz, 6H), 0.93 (t, 3J HH ) 8.0 Hz,
3H); 13C NMR (62.90 MHz, C6D6) δ 171.01 (s), 62.88 (d, 2J CP
)
)
2
1
6.3 Hz), 62.61 (d, J CP ) 6.3 Hz), 61.03 (s), 33.91 (t, J CP
2
3
135.2 Hz), 31.40 (t, J CP ) 4.4 Hz), 16.04 (d, J CP ) 3.8 Hz),
16.03 (d, 3J CP ) 3.8 Hz), 14.11 (s); IR (C6D6, cm-1) 1738 (CdO),
1252 (PdO); MS (EI) m/ z 375 (M + 1)+•, 301 [(M - CO2Et)]+•
.
(3-Oxo-1-p h en ylbu tyl)d ip h en ylp h osp h in e su lfid e (2′b)
was obtained by methods A [∆/KOH], B [))))/KOH], and C
[Al2O3/KOH].
P u r ifica tion a fter Meth od s A a n d B. The organic phase
was separated from the aqueous phase and acetonitrile was
removed under vacuum. To this mixture was added H2O. The
crude expected product was precipitated. The white solid was
separated from the supernatent and recrystallized at low
temperature from EtOH to yield 47% (method A) and 40%
(method B) of 2′b as a white solid.
P u r ifica tion a fter Meth od C. The solid support was
extracted with EtOH. This solution was concentrated and
placed at low temperature. The product 2′b was obtained
(64%) as a white solid: mp 149-151 °C (from EtOH);27 31P
NMR (32.44 MHz, C6D6) δ 50.1 (m); 1H NMR (250.13 MHz,
1
1
) 465 Hz); H NMR (80.13 MHz, C6D6) δ 7.72 (d, J HP ) 465
Hz, 1H), 7.70 (m, 10H); IR (KBr, cm-1) 3060 (Ph), 695 (PdS);
MS (EI) m/ z 219 (M + 1)+•
.
O,O′-Dieth yl (3-Oxo-1-p h en ylbu tyl)p h osp h on a te (2′a ).
This compound was obtained by method C [Al2O3/KOH]. The
supported reagents were heated at 100 °C for 5 min. The
crude product was obtained after extraction with acetone and
was chromatographed on silica (eluent: n-hexane/AcOEt, 1:9),
affording the expected product 2′a (65%) as a yellow oil:4 Rf
0.28 (n-hexane/AcOEt, 1:9); 31P NMR (32.44 MHz, C6D6) δ 27.6
(m); 1H NMR (80.13 MHz, C6D6) δ 7.53-7.04 (m, 5H), 3.92 (q,
3J HH ) 7.3 Hz, 2H), 3.84 (q, 3J HH ) 6.4 Hz, 2H), 3.80-3.58 (m,
3
1H), 3.18-2.81 (m, 2H), 1.55 (s, 3H), 1.01 (t, J HH ) 7.3 Hz,
3
3H), 0.82 (t, J HH ) 6.4 Hz, 3H); 13C NMR (62.90 MHz, C6D6)
3
2
3
2
C6D6) δ 8.28-6.76 (m, 15 H), 4.83 (td, J HH ) 2.7 Hz, J HP
∼
δ 203.52 (d, J CP ) 14.0 Hz), 137.20 (d, J CP ) 6.7 Hz), 129.77
3J HH ) 10.0 Hz, 1H), 3.30 (ddd, 3J HH ) 10.0 Hz, 3J HP ) 5.4 Hz,
2J HH ) 18.2 Hz, 1H), 2.80 (ddd, 3J HH ) 2.7 Hz, 3J HP ) 13.0 Hz,
2J HH ) 18.2 Hz, 1H), 1.29 (s, 3H); 13C NMR (62.90 MHz, C6D6)
δ 204.09 (d, 3J CP ) 14 Hz), 136.21 (s), 133.10-127.55 (m), 44.67
(d, J CP ) 4 Hz), 41.44 (d, J CP ) 53 Hz), 29.99 (s); IR (KBr,
cm-1) 3052 (Ph), 1710 (CdO), 694 (PdS); MS (EI) m/ z 364
(M)+•, 217 (Ph2PdS)+, 43 (CH3CdO)+.
2
2
(d, J CP ) 2.5 Hz), 128.61 (s), 127.32 (s), 62.67 (d, J CP ) 6.7
2
2
Hz), 61.72 (d, J CP ) 7.1 Hz), 44.07 (d, J CP ) 1.7 Hz), 39.51
1
3
3
(d, J CP ) 139.7 Hz), 16.47 (d, J CP ) 5.8 Hz), 16.28 (d, J CP
)
5.5 Hz); IR (C6D6, cm-1) 1738 (CdO), 1252 (PdO); MS (EI) m/ z
284 (M)+•, 241 (M - C(O)CH3)+.
2
1
O,O′-Dieth yl (2-Cya n oeth yl)p h osp h on a te (3′a ). This
adduct was obtained by methods B [))))/KOH] and C [Al2O3/
KOH].
(2-Cya n oeth yl)d ip h en ylp h osp h in e Su lfid e (3′b). Meth-
ods A [∆/KOH], B [))))/KOH], C [Al2O3/KOH], and G [hν/
300nm] were used for the synthesis of 3′b.
P u r ifica tion a fter Meth od s A a n d B. The organic phase
was separated from aqueous phase, removed, and added to
toluene. The brown precipitate was filtered off and taken up
in EtOH. This mixture was placed at low temperature and
3′b was obtained (47% with method A and 76% with method
B) as white needles.
P u r ifica tion a fter Meth od B. The organic solvent was
separated from the aqueous phase and n-hexane was added
to the organic phase. The crude product 3′a was obtained as
a yellow oil. This oil was separated from the solvents and was
dried under vacuum to yield 12% of the expected adduct 3′a .
P u r ifica tion a fter Meth od C. The solid support was
extracted with MeOH and the filtrate was purified on chro-
matography on silica (eluent: MeOH) to yield 55% of the
adduct 3′a as a yellow oil:26 Rf 0.60 (MeOH); 31P NMR (32.44
P u r ifica tion a fter Meth od C. Products were extracted
with acetone and the solvent was removed under vacuum.
(25) Peters, G. J . Am. Chem. Soc. 1960, 82, 4751.
(26) Ginsburg, V. A. Zh. Obshch. Khim. 1960, 30, 3987-3992.
(27) Issleib, K.; J asche, K. Chem. Ber. 1967, 100, 412-420.

Dual Radical/Polar Pudovik Reaction
J . Org. Chem., Vol. 62, No. 8, 1997 2421
Compound 3′b was obtained after recrystallization from EtOH
at low temperature (82%) as white needles.
P u r ifica tion a fter Meth od G. The solvent was evapo-
rated and n-hexane was added. After some hours at low
temperature, 61% of 3′c was obtained as a white solid: mp
P u r ifica tion a fter Meth od G. Deuteriated benzene was
removed after 6 h under irradiation and was added to
petroleum ether. The adduct 3′b was precipitated, separated
from supernatent, and recrystallized from EtOH with a yield
of 35%: mp 128-129 °C (from EtOH);28 31P NMR (32.44 MHz,
43-45 °C (from n-hexane);
31P NMR (32.44 MHz, C6D6) δ
-16.9 (m); 1H NMR (250.13 MHz, C6D6) δ 7.18-7.01 (m, 10H),
1.75-1.59 (m, 4H); 13C NMR (62.90 MHz, C6D6) δ 137.38 (d,
2
1J CP ) 13.5 Hz), 132.97 (d, J CP ) 19.2 Hz), 129.23 (s), 128.93
3
3
2
1
(d, J CP ) 6.9 Hz), 94.90 (d, J CP ) 18.1 Hz), 24.19 (d, J CP
)
CDCl3) δ 41.1 (m); H NMR (80.13 MHz, CDCl3) δ 7.96-7.67
15.7 Hz), 13.73 (d, J CP ) 24.1 Hz); IR (KBr, cm-1) 3067 (Ph),
1
(m, 4H), 7.54-7.41 (m, 6H), 2.81-2.62 (m, 4H); 13C NMR
2246 (CN), 1429 (CH2CN); MS (EI) m/ z 239 (M)+•
.
Anal.
(62.90 MHz, CDCl3) δ 132.28 (d, 4J CP ) 2.8 Hz), 131.09 (d, 2J CP
1
3
Calcd for C15H14NP: C, 75.30; H, 5.90; N, 5.85. Found: C,
74.70; H, 6.00; N, 5.62.
) 10.5 Hz), 130.91 (d, J CP ) 82.4 Hz), 129.05 (d, J CP ) 12.5
Hz), 118.60 (d, 3J CP ) 20.0 Hz), 28.89 (d, 1J CP ) 56.5 Hz), 11.29
(s); IR (KBr, cm-1) 3060 (Ph), 2239 (CN), 1437 (CH2CN), 697
(2,2-Dip h en yleth yl)d ip h en ylp h osp h in e (4′c). This com-
pound was synthetized by method G [hν/300nm] after 26.5 h.
The product 4′c was purified by chromatography (SiO2; elu-
ent: CH2Cl2/AcOEt, 7:3) to give 23% of a colorless oil:27 Rf 0.90
(CH2Cl2/AcOEt, 7:3); 31P NMR (32.44 MHz, C6D6) δ -20.9 (m);
1H NMR (80.13 MHz, C6D6) δ 7.45-7.28 (m, 4H), 7.15-6.90
(PdS); MS (EI) m/ z 271 (M)+•, 218 [(M - CH2CH2CN) + 1]+•
Anal. Calcd for C15H14NPS: C, 66.40; H, 5.20; N, 5.16.
Found: C, 66.15; H, 5.18; N, 5.07.
.
(2,2-Dip h en yleth yl)d ip h en ylp h osp h in e Su lfid e (4′b).
This synthesis was realized by method G [hν/300 nm] in
deuteriated benzene. The solvent was removed under vacuum.
Compound 4′b was recrystallized from MeOH at low temper-
ature and obtained with a yield of 55% as a white solid: mp
154-155 °C (from MeOH); 31P NMR (32.44 MHz, C6D6) δ 40.5
2
3
3
(m, 16H), 4.06 (td, J HP ∼ J HH ) 7.9 Hz, 1H), 2.81 (d, J HH
)
)
7.9 Hz, 2H); 13C NMR (62.90 MHz, C6D6) δ 145.33 (d, J CP
1
2
9.0 Hz), 128.80-122.21 (m), 48.73 (d, J CP ) 15.7 Hz), 114.36
(s), 36.38 (d, J CP ) 14.8 Hz); IR (C6D6, cm-1) 3057 (Ph); MS
1
1
(m); H NMR (80.13 MHz, C6D6) δ 7.75-7.50 (m, 4H), 7.18-
(EI) m/ z 366 (M)+•, 199 (Ph2PCH2)+, 181 [(M - Ph2P)+1]+, 77
(Ph)+.
3
3
6.79 (m, 16H), 5.23 (td, J HH ) 6.8 Hz, J HP ) 14.9 Hz, 1H),
3
2
3.05 (dd, J HH ) 6.8 Hz, J HP ) 11.3 Hz, 2H); 13C NMR (62.90
E t h yl 3-(Dip h en ylp h osp h in o)p r op -2-en oa t e (5′c).
Method B [))))/KOH] was used in this synthesis. The organic
phase was separated from the aqueous phase and was ex-
tracted with n-hexane. The acetonitrile phase contained
compound 5′′d (see description of 5′′d ) and the hexane phase
compound 5′c. The n-hexane solution was concentrated and
placed at low temperature. A precipitate formed. The super-
natent was separated and the solvent was evaporated to give
40% of 5′c as a white solid: 31P NMR (32.44 MHz, C6D6) δ
3
1
MHz, C6D6) δ 144.1 (d, J CP ) 8.3 Hz), 134.1 (d, J CP ) 80.3
1
Hz), 131.54-126.51 (m), 45.6 (s), 38.8 (d, J CP ) 55.2 Hz); IR
(KBr, cm-1) 3051 (Ph), 694 (PdS); MS (EI) m/ z 398 (M)+•, 218
[(M - CH2CHPh2) + 1]+•, 181 (Ph2CHCH2)+, 140 (PhPdS)+•
.
Anal. Calcd for C26H23PS: C, 78.36; H, 5.82. Found: C, 78.17;
H, 5.78.
Dip h en yl-cis-â-styr ylp h osp h in e Su lfid e (6′b). Adduct
6′b was obtained by methods E [))))], F [))))/AIBN], and G [hν,
300 nm, 6 h]. After the synthesis by each method, the solvent
was evaporated. Hexane was added and the crude product
6′b was precipitated. The product was purified by recrystal-
lization from EtOH after methods E (40%) and G (30%) and
by chromatography (SiO2; eluent: n-hexane/AcOEt, 99:1) after
synthesis by method F (65%). Compound 6′b was obtained
as a white solid: Rf 0.30 (n-hexane/AcOEt, 99:1); mp ) 108-
109 °C (from EtOH);29 31P NMR (32.44 MHz, C6D6) δ 29 (m);
1H NMR (250.13 MHz, C6D6) δ 7.99-7.89 (m, 3H), 7.65-7.60
(m, 2H), 7.16-7.15 (m, 2H), 7.03 (d, 3J HH ) 13.6 Hz, 1H), 7.05-
-11.2 (m) (cis isomer), -13.2 (m) (trans isomer); 1H NMR
3
(80.13 MHz, C6D6) (ratio cis/trans, 50:50) δ 7.76 (dd, J HH
)
)
2
3
2
16.8 Hz, J HP ) 4.0 Hz, 1H), 7.10 (dd, J HH ) 12.3 Hz, J HP
3
3
1.5 Hz, 1H), 6.51 (dd, J HH ) 12.3 Hz, J HP ) 15.4 Hz, 1H),
3
3
3
5.83 (dd, J HH ) 16.8 Hz, J HP ) 6.8 Hz, 1H), 4.18 (q, J HH
)
3
3
7.2 Hz, 2H), 3.56 (q, J HH ) 7.2 Hz, 2H), 1.03 (t, J HH ) 7.2
3
Hz, 3H), 0.95 (t, J HH ) 7.2 Hz, 3H).
Dip h en yl-cis-â-st yr ylp h osp h in e (6′c). Methods
D
[∆/AIBN], F [))))/AIBN], and G [hν/300nm] were used. The
purification was the same for each method: 6′c was purified
by chromatography (SiO2; eluent: n-hexane/AcOEt, 9:1) to
yield 81% (method D), 40% (method F), and 41% (method G)
of 6′c cis as a white solid: 31P NMR (32.44 MHz, C6D6) δ -24.8
2
3
6.75 (m, 8H), 6.11 (dd, J HP ) 17.6 Hz, J HH ) 13.6 Hz, 1 H);
13C NMR (62.90 MHz, C6D6) δ 146.00 (d, 2J CP ) 2.5 Hz), 136.18
1
(s), 134.49 (d, J CP ) 85.1 Hz), 131.63-127.69 (m), 124.08 (d,
1J CP ) 81.5 Hz); IR (KBr, cm-1) 3051 (Ph), 1595 (CdC), 694
1
(m); H NMR (80.13 MHz, C6D6) δ 7.65-6.90 (m, 11H), 6.40
(PdS); MS (EI) m/ z 320 (M)+•, 218 [(M - PhCHdCH) + 1]+•
,
(dd, 2J HP ) 2.4 Hz, 3J HH ) 12.6 Hz, 1H); 13C NMR (62.90 MHz,
78 (Ph + 1)+•
.
2
1
C6D6) δ 144.31 (d, J CP ) 19.1 Hz), 140.03 (d, J CP ) 10.8 Hz),
2
1
(1-Hexen yl)d ip h en ylp h osp h in e Su lfid e (7′b). The mix-
ture was irradiated for 7 h under 300 nm light (method G).
The toluene was removed and 7′b was obtained as a white
solid after recrystallization in EtOH at low temperature with
a yield of 34%: 31P NMR (32.44 MHz, CDCl3) δ -14.1 (m) (cis
137.52 (s), 133.23 (d, J CP ) 19.2 Hz), 130.14 (d, J CP ) 8.2
Hz), 128.60 (m); IR (CH2Cl2, cm-1) 3051 (Ph), 1438 (CdC); MS
(EI) m/ z 288 (M)+•
.
(1-Hexen yl)d ip h en ylp h osp h in e (7′c). This adduct was
synthetized by method G [hν, 300nm, 19 h]. The deuteriated
toluene was evaporated. The colorless oil obtained was
extracted with n-hexane. The solution was concentrated and
placed at low temperature. The precipitate was separated and
purified by chromatography (SiO2; eluent: n-hexane/AcOEt,
9:1). The product 7′c was obtained with a yield of 41% (93%
cis isomer/7% trans isomer) as a white solid: 31P NMR (32.44
MHz, C6D6) δ -14.1 (m) (cis isomer) (93%), -31.7 (m) (trans
isomer) (7%); 1H NMR (250.13 MHz, C6D6) (ratio cis/trans, 93:
7) δ 7.48-7.06 (m, 20H), 6.35-6.18 (m, 4H), 2.70-2.52 (m,
2H), 2.09-1.89 (m, 2H), 1.47-1.11 (m, 8H), 0.92-0.69 (m, 6H);
13C NMR (62.90 MHz, C6D6) (ratio cis/trans, 93:7) δ 149.01 (d,
isomer), -31.7 (m) (trans isomer); 1H NMR (250.13 MHz,
3
CDCl3) (cis isomer) δ 8.01-7.52 (m, 10 H), 6.57 (tdd, J HH
)
)
12.4 Hz, 3J HH ) 7.6 Hz, 3J HP ) 42.9 Hz, 1H), 6.32 (tdd, 3J HH
4
2
12.4 Hz, J HH ) 1.52 Hz, J HP ) 23.6 Hz, 1H), 2.27 (m, 2H),
1.23 (m, 4H), 0.78 (t, 3J HH ) 7.9 Hz, 3H); 13C NMR (62.90 MHz,
1
CDCl3) (cis isomer) δ 153.50 (s), 134.02 (d, J CP ) 83.2 Hz),
3
3
131.25 (d, J CP ) 13.1 Hz), 128.03 (d, J CP ) 12.3 Hz), 123.06
1
3
(d, J CP ) 84.4 Hz), 30.67 (d, J CP ) 9.5 Hz), 30.46 (s), 22.22
(s), 13.61 (s); IR (C6D6, cm-1) (cis isomer) 3054 (Ph), 1434
(CdC); MS (EI) m/ z 300 (M)+•, 271 (M - CH2CH3)+, 108
(PhP)+•. Anal. Calcd for C18H21PS: C, 71.97; H, 7.05. Found:
C, 71.55; H, 7.05.
2
1
2J CP ) 33.6 Hz), 147.96 (d, J CP ) 24.8 Hz), 140.29 (d, J CP
)
1
1
(2-Cya n oeth yl)d ip h en ylp h osp h in e (3′c). Methods B
[))))/KOH] and G [hν/300nm] was used for the synthesis of 3′c.
P u r ifica tion a fter Meth od B. The organic liquid phase
was separated from the aqueous phase and the solvent was
removed under vacuum. The crude product 3′c precipitated
after addition of n-hexane. This solution was concentrated and
placed at low temperature to yield 76% of 3′c.
10.5 Hz), 140.08 (d, J CP ) 10.6 Hz), 133.07 (d, J CP ) 19.0
1
Hz), 133.04 (d, J CP ) 18.5 Hz), 128.82-127.69 (m), 35.04 (d,
3J CP ) 12.8 Hz), 31.71 (s), 31.19 (d, J CP ) 11.9 Hz), 30.95 (s),
3
22.61 (s), 22.57 (s), 14.20 (s), 14.16 (s); IR (C6D6, cm-1) (ratio
cis/trans, 93:7) 3054 (Ph), 1434 (CdC); MS (EI) m/ z 268 (M)+•,
239 (M - CH2CH3)+, 108 (PhP)+•. Anal. Calcd for C18H21P:
C, 80.57; H, 7.89. Found: C, 80.63; H, 7.99.
(3-Oxo-1-p h en ylbu tyl)d ip h en ylp h osp h in e Oxid e (2′d ).
Methods A [∆/KOH], B [))))/KOH], C [Al2O3/KOH], and G [hν,
300nm, 39 h] were used to obtain 2′d .
(28) Pudovik, A. N.; Batyeva, E. S. J . Gen. Chem. USSR 1969, 314.
(29) Aguiar, A. M.; Daigle, D. J . Org. Chem. 1965, 30, 2826.

2422 J . Org. Chem., Vol. 62, No. 8, 1997
Semenzin et al.
P u r ifica tion a fter Meth od s A a n d B. The solution of
acetonitrile was separated from the aqueous phase, evapo-
rated, and taken up in n-hexane. The precipitate formed was
separated, the supernatent was purified by chromatography
(SiO2; eluent: AcOEt/n-hexane, 8:2), and the adduct 2′d was
obtained after recrystallization from EtOH (77% for the
method A, 59% for the method B).
phase was concentrated and was purified by chromatography
(SiO2; eluent: CH2Cl2/acetone, 1:1) to yield 20% of 5′′d as a
brown oil.
P u r ifica tion a fter Meth od C. The solid support was
extracted with MeOH. This solution was purified by chroma-
tography (see Purification after Method B). The yield of 5′′d
was 20%.
P u r ifica tion a fter Meth od C. The solid support was
extracted with acetone and the solvent was removed under
vacuum. Compound 2′d was precipitated as a white solid in
the mixture ClCH2CH2Cl/n-hexane, separated, and dried to
yield 75%.
P u r ifica tion a fter Syn th esis in Et2O. The mixture was
directly purified by chromatography (same conditions) after
20 min at room temperature to give 73% of 5′′d as a brown
oil: 31P NMR (32.44 MHz, CDCl3) δ 31.7 (m); 1H NMR (200.13
MHz, CDCl3) δ 7.88-7.74 (m, 8H), 7.32-7.25 (m, 12H), 4.27
(tt, 3J HH ) 5.7 Hz, 2J HP ) 20.1 Hz, 1H), 3.69 (q, 3J HH ) 7.2 Hz,
2H), 2.85 (td, 3J HH ) 5.7 Hz, 3J HP ) 15.0 Hz, 2H), 0.94 (t, 3J HH
) 7.2 Hz, 3H); 13C NMR (62.90 MHz, CDCl3) δ 171.36 (s),
131.90-128.27 (m), 129.93 (d, 1J CP ) 21.6 Hz), 61.44 (s), 30.46
(s), 29.28 (s), 13.81 (s); IR (CDCl3, cm-1) 3060 (Ph), 1728 (CdO),
P u r ifica tion a fter Meth od G. The compound obtained
in the mixture after 39 h under irradiation was 2′c but was
oxidized during the purification to give 2′d . After evaporation
of the deuteriated chloroform, the yellow oil obtained was
taken up in a mixture ClCH2CH2Cl/n-hexane and the com-
pound 2′d was precipitated at low temperature (76%): mp 170
°C (from EtOH); 31P NMR (32.44 MHz, CDCl3) δ 33.6 (m); 1H
NMR (250.13 MHz, CDCl3) δ 8.11-7.26 (m, 15H), 4.39 (ddd,
1438 (C-O), 1190 (PdO); MS (EI) m/ z 503 (M + 1)+•
.
1,2-Bis(d ip h en ylp h osp h in yl)-1-p h en yleth a n e (9). This
adduct was obtained after synthesis by methods A [∆/KOH],
B [))))/KOH], and C [Al2O3/KOH].
P u r ifica tion a fter Meth od s A a n d B. The crude product
9 was precipitated in the reaction mixture at room tempera-
ture. The precipitate was separated from the supernatant,
washed in H2O, and recrystallized at low temperature from a
CH2Cl2/n-hexane mixture. Yields were 14% after method A
and 40% after method B.
3
2
3J HH ) 3.0 Hz, J HH ) 10.0 Hz, J HP ) 7.2 Hz, 1H), 3.46 (ddd,
2J HH ) 17.9 Hz, 3J HH ) 10.0 Hz, 3J HP ) 5.4 Hz, 1H), 3.07 (ddd,
2J HH ) 17.9 Hz, J HH ) 3.0 Hz, J HP ) 11.2 Hz, 1H), 2.01 (s,
3
3
3H); 13C NMR (62.90 MHz, CDCl3) δ 205.30 (d, J CP ) 12.8
3
1
Hz), 135.77 (d, J CP ) 5.5 Hz), 135.72-127.14 (m), 132.07 (s),
43.49 (s), 41.08 (d, 1J CP ) 68.7 Hz), 30.59 (s); IR (CDCl3, cm-1
)
,
3061 (Ph), 1817 (CdO), 1263 (PdO); MS (EI) m/ z 348 (M)+•
201 (Ph2PdO)+, 77 (Ph)+, 43 (CH3CdO)+. Anal. Calcd for
C22H21O2P: C, 75.85; H, 6.08. Found: C, 75.17; H, 5.99.
(2,2-Dip h en ylet h yl)d ip h en ylp h osp h in e Oxid e (4′d ).
This adduct was purified after the synthesis by method B [))))/
KOH]. The organic phase was separated from the aqueous
phase and was extracted with acetone (3 × 10 mL). Compound
4′d was recrystallized at low temperature from n-hexane to
give 48% of a white solid or was purified by chromatography
(SiO2; eluent: CH2Cl2/AcOEt, 7:3): Rf 0.59 (CH2Cl2/AcOEt,
7:3); mp 213-214 °C (from n-hexane); 31P NMR (32.44 MHz,
C6D6) δ 25.4 (m); 1H NMR (80.13 MHz, C6D6) δ 7.73-7.47 (m,
P u r ifica tion a fter Meth od C. The compound 9 was
extracted with CH2Cl2. This solution was concentrated, added
to n-hexane, and placed at low temperature to yield 16% of 9
as a white solid: mp 169-172 °C (from CH2Cl2/n-hexane);30
31P NMR (101.26 MHz, CDCl3) δ 35.3 (d, 3J PP ) 46.8 Hz), 30.0
3
(d, J PP ) 46.8 Hz); 1H NMR (300.13 MHz, CDCl3) δ 8.07-
3
3
6.81 (m, 25 H), 4.26 (dddd, J HP ) 4.5 Hz, J HH ) 11.6 Hz,
3J HH ) 1.4 Hz, 2J HP ∼ J HH ) 11.6 Hz, 1H), 3.12 (dddd, 2J HH
)
3
3
2
3
15.3 Hz, J HH ) 11.6 Hz, J HP ) 2.8 Hz, J HP ) 3.9 Hz, 1H),
3
2
2
3
2.8 (dddd, J HH ) 1.4 Hz, J HH ) 15.3 Hz, J HP ) 9.2 Hz, J HP
) 13.7 Hz, 1H); 13C NMR (75.5 MHz, CDCl3) δ 131.91 (d, J CP
1
3
3
1
4H), 7.16-6.88 (m, 16H), 4.90 (td, J HH ) 7.0 Hz, J HP ) 11.9
Hz, 1H), 2.87 (dd, 3J HH ) 7.0 Hz, 2J HP ) 10.9 Hz, 2H); 13C NMR
(62.90 MHz, DMSO D6) δ 144.36 (d, 3J CP ) 8.0 Hz), 133.77 (d,
) 31.8 Hz), 131.53 (d, J CP ) 20.4 Hz), 132.16-127.01 (m),
39.38 (d, 1J PC ) 66 Hz), 30.14 (d, 1J PC ) 69 Hz); IR (KBr, cm-1
)
3070 (Ph), 1181 (PdO); MS (EI) m/ z 506 (M)+•, 305 (M - 201)+,
201 (Ph2PdO)+, 104 (M - 2 × 201)+•, 77 (Ph)+.
1J CP ) 96.8 Hz), 131.02-125.92 (m), 44.42 (s), 34.12 (d, 1J CP
)
68.0 Hz); IR (KBr, cm-1) 3054 (Ph), 1178 (PdO); MS (EI) m/ z
382 (M)+•, 202 [(M - CH2CHPh2) + 1]+•. Anal. Calcd for C26H23
OP: C, 81.66; H, 6.06. Found: C, 80.62; H, 6.02.
-
Ack n ow led gm en t. We are grateful to the Center
National de la Recherche Scientifique (CNRS) and the
Conseil Re´gional de la Re´gion Midi-Pyre´ne´es for finan-
cial support and to David Pleynet for his helpful
assistance.
Eth yl 3,3-Bis(d ip h en ylp h osp h in yl)p r op ion a te (5′′d ).
This compound was synthetized by methods B [))))/KOH] and
C [Al2O3/KOH] and in Et2O after 20 min at room temperature
(exothermic reaction).
P u r ifica tion a fter Meth od B. The acetonitrile phase was
separated from the aqueous phase and was extracted with
n-hexane (3 × 10 mL). The solution of n-hexane contained
the compound 5′c (see purification of 5′c). The acetonitrile
J O9622441
(30) Stang, P. J .; Arif, A. M.; Zhdankin, V. V. Tetrahedron 1991,
47, 26, 4539-4546.