Generation of phosphorus-centered radicals via homolytic substitution at sulfur

Article abstract of DOI:10.1021/ol0631096

(Chemical Equation Presented) A novel radical domino process relying on the homolytic cleavage of P-S bonds allows the preparation of phosphorus-containing molecules through addition of P-centered radicals onto olefins. The key step of this reaction is a homolytic substitution on a sulfur atom. The scope of the reaction is broad. Diaminophosphonyl radicals whose reactivity was unknown react smoothly with olefins. Use of tin hydride can be avoided. A radical thiophosphinoylation of triple bonds has been uncovered.

Full text of DOI:10.1021/ol0631096

ORGANIC
LETTERS
2007
Vol. 9, No. 6
1061-1063
Generation of Phosphorus-Centered
Radicals via Homolytic Substitution at
Sulfur
Paola Carta, Nicolas Puljic, Carine Robert, Anne-Lise Dhimane,
Louis Fensterbank,* Emmanuel Lacoˆte,* and Max Malacria*
UniVersite´ Pierre et Marie Curie-Paris 6, Laboratoire de chimie organique
(UMR CNRS 7611), Institut de chimie mole´culaire (FR 2769), 4 place Jussieu, C. 229,
75005 Paris, France
louis.fensterbank@upmc.fr; lacote@ccr.jussieu.fr; max.malacria@upmc.fr
Received December 22, 2006
ABSTRACT
A novel radical domino process relying on the homolytic cleavage of P−S bonds allows the preparation of phosphorus-containing molecules
through addition of P-centered radicals onto olefins. The key step of this reaction is a homolytic substitution on a sulfur atom. The scope of
the reaction is broad. Diaminophosphonyl radicals whose reactivity was unknown react smoothly with olefins. Use of tin hydride can be
avoided. A radical thiophosphinoylation of triple bonds has been uncovered.
The formation of C-P bonds has become an extremely
important tool in modern synthesis. For example, the need
for new phosphine-based ligands or phosphorus-containing
biomolecules has created a strong demand for such meth-
odologies. Because they are very mild, radical reactions are
good candidates for such a purpose and have been shown to
be especially suitable to phosphorus. As a consequence,
various methods have been devised for radical-based forma-
tion of C-P bonds. The two main options available are
addition of P-centered radicals to unsaturated compounds1
or addition of carbon radicals to phosphorus moieties.2 The
former is a very old reaction,3 and it has been reinvestigated
recently.4 The P-centered radicals have been prepared mainly
by abstraction of a hydrogen atom from the corresponding
P-H bond and by deselenylation. Both those methods have
limitations: selenium-containing molecules are expensive
and toxic. Some P-H bonds may be difficult to cleave. For
example, in our hands, no radicals were formed via P-H
bond cleavage of diaminophosphonates.
Our attention was attracted by the highly original work
of Oshima and Yorimitsu, who introduced diphosphanes as
a source of both P-radicals and phosphinylation agents.5 The
key feature of the reaction is a homolytic substitution on
(4) (a) Herpin, T. F.; Motherwell, W. B.; Roberts, B. P.; Roland, S.;
Weibel, J.-M. Tetrahedron 1997, 53, 15085. (b) Depre`le, S.; Montchamp,
J.-L. J. Org. Chem. 2001, 66, 6745. (c) Jessop, C. M.; Parsons, A. F.;
Routledge, A.; Irvine, D. Tetrahedron Lett. 2003, 44, 479. (d) Jessop, C.
M.; Parsons, A. F.; Routledge, A.; Irvine, D. J. Tetrahedron: Asymmetry
2003, 14, 2849. (e) Lopin, C.; Gouhier, G.; Gautier, A.; Piettre, S. R. J.
Org. Chem. 2003, 68, 9916. (f) Gautier, A.; Garipova, G.; Salcedo, C.;
Balieu, S.; Piettre, S. R. Angew. Chem., Int. Ed. 2004, 43, 5963. (g) Beaufils,
F.; De´ne`s, F.; Renaud, P. Angew. Chem., Int. Ed. 2005, 44, 5273. (h)
Mimeau, D.; Delacroix, O.; Gaumont, A.-C. Chem. Commun. 2003, 2928.
(i) Mu, X.-J.; Zou, J.-P.; Qian, Q.-F.; Zhang, W. Org. Lett. 2006, 8, 5291.
(j) Hirai, T.; Han, L.-B. Org. Lett. 2007, 9, 53.
(1) Leca, D.; Fensterbank, L.; Lacoˆte, E.; Malacria, M. Chem. Soc. ReV.
2005, 34, 858.
(2) (a) Bentrude, W. G. Acc. Chem. Res. 1982, 15, 117. (b) Bentrude,
W. G. In The Chemistry of Organophosphorus Compounds; Hartley, F. R.,
Ed.; Wiley: Chichester, 1990; Vol. 1, p 531.
(3) (a) Stacey, F. W.; Harris, J. F. Org. React. 1963, 13, 150. (b) Walling,
C.; Pearson, M. S. Top. Phosphorus Chem. 1966, 3, 1.
(5) Sato, A.; Yorimitsu, H.; Oshima, K. Angew. Chem., Int. Ed. 2005,
44, 1694.
10.1021/ol0631096 CCC: $37.00
© 2007 American Chemical Society
Published on Web 02/23/2007

one phosphorus atom, which releases a phosphinyl radical
that can add to terminal alkynes or be reduced to afford
phosphinated products.6 Because of our previous interest in
the elimination of oxidized P• radicals,7 we wondered whether
homolytic substitution could enable us to generate those
radicals in a convenient way.
Table 1. Formation of C-P Bonds through a Homolytic
Substitution/P-Centered Radical Addition Tandem
We turned our attention to thiophosphonates 1a-c. It has
been shown early on that sulfur is a very good target for
rapid homolytic substitution.8 Crich9 and Spagnolo/Benati
(for a tin-free version)10 used that particular reaction to
prepare acyl radicals. We felt that a related reaction could
take place with phosphorus analogues, thus affording the
corresponding P-centered radicals that could further react
with olefins. This approach would add a new method to the
existing tools available to generate and use P-centered
radicals (the key radical fragmentation would now be a P-S
bond cleavage). We also wished to gain insight on radicals
which have so far not been used in synthesis, such as the
diaminophosphonyl ones. To achieve this goal, a rapid
substitution on phosphorus is needed to avoid unproductive
early reductions. An intramolecular reaction meets this
criterion. We present herein our results.
Substrates 1a-c were chosen for this study, as they
represent a good sample of phosphorus(v) moieties. The
reactions were carried out under standard radical conditions,
i.e., slow addition of tributyltin hydride (TBTH) in the
presence of AIBN and an olefin in refluxing benzene (Table
1, method A). We selected an array of olefins with different
electronic properties (electron-poor, -rich, and neutral) to get
additional data on the philicity of the phosphorus-centered
radicals.
A typical example is shown in Table 1, entry 1. In the
presence of TBTH, 1-octene, and AIBN in refluxing benzene,
substrate 1a yielded 84% of the expected diphenyloctyl
phosphine oxide. The reaction proved quite general. Both
phosphinoyl and phosphonyl radicals led to the corresponding
P-C bond formation, and the dihydrobenzothiophene byprod-
uct could be easily separated from the desired products.
a
Method A: Bu3SnH, AIBN cat., slow addition, PhH reflux. Method
B: Bu3SnCl (10 mol %), NaBH4, t-BuOH, reflux. b 10 equiv of olefin was
used. 2 equiv of olefin was used. Product was contaminated by
c
d
(6) Sato, A.; Yorimitsu, H.; Oshima, K. J. Am. Chem. Soc. 2006, 128,
4240.
e
approximately 17% of byproduct. Product was contaminated by ap-
proximately 9% of byproduct. f Some polymerized product was also isolated.
g 10-30% of reduced starting material was also observed. h Product is not
stable.
(7) (a) Bogen, S.; Gulea, M.; Fensterbank, L.; Malacria, M. J. Org. Chem.
1999, 64, 4920. (b) Leca, D.; Fensterbank, L.; Lacoˆte, E.; Malacria, M.
Angew. Chem., Int. Ed. 2004, 43, 4220. (c) Leca, D.; Song, K.; Albert, M.;
Grangeio Gonc¸alves, M.; Fensterbank, L.; Lacoˆte, E.; Malacria, M. Synthesis
2005, 1405. For an additional application of the process, see: Ouvry, G.;
Quiclet-Sire, B.; Zard, S. Z. Angew. Chem., Int. Ed. 2006, 45, 5002.
(8) For reviews, see: (a) Kampmeier, J. A.; Jordan, R. B.; Liu, M. S.;
Yamanaka, H.; Bishop, D. J. ACS Symp. Ser. 1978, 69, 275. (b) Beckwith,
A. L. J. Chem. Soc. ReV. 1993, 143. (c) Schiesser, C. H.; Wild, L. M.
Tetrahedron 1996, 52, 13265. (d) Walton, J. C. Acc. Chem. Res. 1998, 31,
99. (e) Crich, D. HelV. Chim. Acta 2006, 89, 2167. (f) Schiesser, C. H.
Chem. Commun. 2006, 4055.
(9) (a) Crich, D.; Hao, X. J. Org. Chem. 1997, 62, 5982. (b) Crich, D.;
Yao, Q. Tetrahedron 1998, 54, 305. (c) Crich, D.; Yao, Q. Org. Lett. 2003,
5, 2189. (d) Crich, D.; Yao, Q. J. Am. Chem. Soc. 2004, 126, 8232. For
other applications of homolytic substitutions of thioesters, see: (e) Benati,
L.; Leardini, R.; Minozzi, M.; Nanni, D.; Spagnolo, P.; Strazzari, S.; Zanardi,
G. Org. Lett. 2002, 4, 3079. (f) Hayes, C. J.; Herbert, N. M. A.; Harrington-
Frost, N. M.; Pattenden, G. Org. Biomol. Chem. 2005, 3, 316. (g) De Boeck,
B.; Harrington-Frost, N. M.; Pattenden, G. Org. Biomol. Chem. 2005, 3,
340.
Surprisingly, the yield of the phosphinoyl radical addition
to enol ether (entry 5) was better than the one obtained with
acrylonitrile (entry 3). Indeed, phosphinoyl radicals are
thought to be moderately nucleophilic11 and thus should react
best with electron-poor olefins. This may be due to polym-
erization.11 Nevertheless, the yield remained low even when
the amount of olefins was reduced to two equivalents to avoid
excessive polymerization of acrylonitrile. We will investigate
this aspect more thoroughly, but initial competition experi-
ments (acrylonitrile vs enol ether) tend to confirm the
(10) (a) Benati, L.; Calestani, G.; Leardini, R.; Minozzi, M.; Nanni, D.;
Spagnolo, P.; Strazzari, S. Org. Lett. 2003, 5, 1313. (b) Benati, L.;
Bencivenni, G.; Leardini, R.; Minozzi, M.; Nanni, D.; Scialpi, R.; Spagnolo,
P.; Zanardi, G. J. Org. Chem. 2006, 71, 3192.
(11) (a) Sumiyoshi, T.; Schnabel, W.; Henne, A.; Lechtken, P. Polymer
1985, 26, 141. (b) Sumiyoshi, T.; Schnabel, W. Makromol. Chem. 1985,
186, 1811. (c) Kajiwara, A.; Konishi, Y.; Morishima, Y.; Schnabel, W.;
Kuwata, K.; Kamachi, M. Macromolecules 1993, 26, 1656.
1062
Org. Lett., Vol. 9, No. 6, 2007

nucleophilic character of the phosphinoyl radical. In the enol
ether case, a minor unidentified phosphorus-containing
byproduct was also observed. Phosphonyl radicals also gave
fair yields of products (entries 7-10). Our methodology was
extended to diaminophosphonyl radicals (entries 11-14),
thus demonstrating their synthetic potential.
Encouraged by these results, we decided to eliminate
TBTH, as tin byproducts are difficult to remove and toxic.
As a first step toward that goal, we decided to use Stork’s
catalytic conditions (Bu3SnCl (10 mol %), NaBH4, t-BuOH,
reflux; method B).12 The additions still worked, albeit
generally in slightly reduced yields. To eliminate tin entirely,
we took advantage of the Benati/Spagnolo trick and replaced
the radical trigger with a triple bond (see 5a-c, Table 2).
with 5a and hexene, we isolated some products arising from
trapping of phosphinoyl radicals by 5a. This side reaction
attracted our attention, as it could be used to our advantage
(Scheme 1). Indeed, removal of the olefin led to a clean
Scheme 1. Thiophosphorylation of Triple Bonds
Table 2. Tin-Free Formation of C-P Bonds
formal cycloisomerization of thiophosphine oxide 5a (72%
yield), during which the C-P bond formation from phos-
phinoyl radicals formed in the initiation process triggered
the homolytic substitution that installed the sulfur moiety.
Cyclic sulfide 6 was obtained as a 70:30 mixture of isomers,
whose geometry was determined by NOE. This last reaction
opens new perspectives to our homolytic substitution: to the
best of our knowledge, this is the first example of a
thiophosphinoylation of triple bonds, during which the two
different heteroatoms are introduced on the hydrocarbon
backbone in the same synthetic step.
To conclude, we have devised a new way of generating
phosphorus-centered radicals. Homolytic substitution of
thiophosphine oxides, thiophosphonates, and thiodiamino-
phosphonates led to the corresponding radicals, which added
smoothly onto olefins. The diaminophosphonyl radical
additions to olefins have not been previously reported. We
could avoid using tin hydride by switching to derivatives of
pent-4-ynethiol and thiophenol as the mediator. The efficient
addition of P• onto unsaturated compounds led us to uncover
an unprecedented thiophosphinoylation reaction, in which
both heteroatoms are introduced at both ends of the triple
bond. We are currently assessing the full scope of this new
process, particularly when chiral phosphorus-containing
moieties are involved. Our results will be presented in due
course.
a
28% of the hydrothiolation of the triple bond was also observed. b 5%
c
of the hydrothiolation of the triple bond was also observed. 10% of the
hydrothiolation of the triple bond was also observed.
The homolytic substitution was neatly triggered by the
thiyl radical addition onto the triple bonds. Except in the
phosphonyl case (Table 2, entry 3), this new cascade
delivered much better yields of phosphorus-containing
products, especially with acrylonitrile. Some reactions de-
livered byproducts arising from reduction of the intermediate
vinyl radicals (Table 2, entries 3-5), which complicated the
purification process. Nonetheless, the elimination of tin
reagents increases the synthetic interest of our procedure.
Phosphinoyl radicals are also known to add onto triple
bonds.4g When only two equivalents of olefins was employed
Acknowledgment. We thank CNRS, IUF, and UPMC
for financial support. Technical support from ICSN (Gif sur
Yvette, France) is gratefully acknowledged (HRMS and
elemental analyses).
Supporting Information Available: Full characterization
(including NMR spectra) of all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(12) Stork, G.; Sher, P. M. J. Am. Chem. Soc. 1986, 108, 303.
OL0631096
Org. Lett., Vol. 9, No. 6, 2007
1063