Radical cyclizations of alkenyl acylphosphonate derivatives under thermal and photochemical conditions

Article abstract of DOI:10.1246/bcsj.78.1665

The use of alkenyl acylphosphonate and acylphosphine oxide derivatives as acceptors in radical cyclizations was studied under thermal and photochemical conditions, respectively. The cyclizations of alkenyl acylphosphonates under thermal conditions occurre

Full text of DOI:10.1246/bcsj.78.1665

ꢀ 2005 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 78, 1665–1672 (2005) 1665
Radical Cyclizations of Alkenyl Acylphosphonate Derivatives
under Thermal and Photochemical Conditions
ꢀ
ꢀ;1
Chang Ho Cho, Sunggak Kim, Motoki Yamane,1 Hironori Miyauchi,1 and Koichi Narasaka
Center for Molecular Design & Synthesis and Department of Chemistry, School of Molecular Science (BK21),
Korea Advanced Institute of Science and Technology, Daejeon 305-701, Korea
1Department of Chemistry, Graduate School of Science, The University of Tokyo,
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033
Received March 16, 2005; E-mail: skim@kaist.ac.kr
The use of alkenyl acylphosphonate and acylphosphine oxide derivatives as acceptors in radical cyclizations was
studied under thermal and photochemical conditions, respectively. The cyclizations of alkenyl acylphosphonates under
thermal conditions occurred smoothly in refluxing dioxane using benzoyl peroxide as an initiator; the addition of diethyl
phosphite increased the chemical yield. Photochemically induced cyclizations of alkenyl acyldiphenylphosphine oxides
at 300 nm gave similar results, although a notable difference was observed in one case. The intramolecular cyclization of
S-but-3-enyl phosphinothiolformates occurred under thermal and photochemical conditions, providing thiolactones,
whereas S-pent-4-enyl phosphinothiolformate afforded the tetrahydrothiophene derivative under similar conditions.
Radical cyclizations proved to be synthetically useful for
O
O
the construction of 5- and 6-membered rings, and mostly uti-
lized carbon–carbon multiple bonds as radical acceptors.1 Re-
cently, carbon–nitrogen double bonds, such as imines, oximes,
and hydrazones, have been known to be very effective radical
acceptors.2 However, the carbonyl group, one of the most im-
portant functional groups, has not been widely used as an ac-
ceptor because the addition of an alkyl radical to the carbonyl
group is energetically unfavorable and reversible due to the
strong ꢀ-bond strength of the carbonyl group.3 Although an
aldehyde and a ketone are not suitable as acceptors, due to
unfavorable equilibrium at the cyclization, acyl sulfides, acyl
selenides,4 and acylgermanes5 have been effectively used as
carbonyl group equivalent radical acceptors.
O
O
P(OEt)2
P(OEt)2
P(OEt)2
O
O
O
O
O
P(OEt)2
P(OEt)2
P(OEt)2
O
P(OEt)2
O
Scheme 1. Phosphonyl radical-mediated acylphosphoryl-
ation of an alkene.
The acylphosphonates were known as acylating agents, and
react with various nucleophiles, such as alcohols, amines, and
carbanions.6 In radical reactions, acyl and bis(acyl)phosphine
oxides have been widely employed as photochemical sources
of acyl radicals for use as initiators in the polymerization reac-
tions.7 Very recently, Kim et al. have shown intramolecular
acylation reactions using an acylphosphonate group as a car-
bonyl group equivalent radical acceptor.8 The intramolecular
acylation involves the addition of an electrophilic alkyl radical
onto an alkenyl bond, followed by cyclization to the acyl-
phosphonate group along with the ꢁ-elimination of a diethyl
phosphonyl group from an alkoxyl radical (Eq. 1). In the reac-
tion, the finally eliminated phosphonyl radical reacts with an
alkyl halide to generate an alkyl radical along with diethyl
phosphoryl halide. Since the phosphonyl radical is known to
add efficiently to alkenes,9 we investigated the cyclization of
alkenyl acylphosphonates based on a plausible mechanism,
depicted in Scheme 1. If the eliminated phosphonyl radical
generated by the radical cyclization could add to the starting
alkenyl moiety of the acylphosphonate, the phosphonyl radical
would mediate the acylphosphorylation of alkenyl acylphos-
phonates to provide cyclic ketones bearing the diethyl phos-
phonyl group.
O
O
P(OEt)2
AIBN
+
+
R X
O
ð1Þ
O
R
X
P(OEt)2
Results and Discussion
To initiate the radical reaction, two reaction conditions in-
volving thermal and photochemical conditions were employed.
We first examined the combination of benzoyl peroxide/dieth-
yl phosphite in benzene.9a When butenyl benzoylphosphonate
Published on the web August 30, 2005; DOI 10.1246/bcsj.78.1665

1666 Bull. Chem. Soc. Jpn., 78, No. 9 (2005)
Radical Cyclizations of Acylphosphonates
Table 1. Solvent Effect in the Radical Reaction of Acyl-
phosphonate 1aÞ
Table 2. Effect of the Amount of Diethyl Phosphite in the
Radical Cyclization of Acylphosphonate 1aÞ
Solvent
Time
Yield of 2
HPO(OEt)2 (x)bÞ
Yield
2,2,5,5-Tetramethyltetrahydrofuran
Benzene
Toluene
6 h
6 h
6 h
1 h
1 h
trace
trace
27%
75%
82%
/mol amt.
2
3
5.0
1.0
0.4
0
88%
82%
72%
64%
3%
5%
7%
7%
Propionitrile
1,4-Dioxane
a) Ratio of the reagents was 1:(PhCO2)2:HPO(OEt)2 =
1.0:0.2:1.0.
a) The reaction was performed by using reagents in the ratio of
1:(PhCO2)2:HPO(OEt)2 = 1.0:0.2:x in 1,4-dioxane for 1 h. b)
Molar amounts of diethyl phosphite to 1.
1 was heated with diethyl phosphite (1 equiv) in the presence
of benzoyl peroxide (20 mol %) in benzene for 6 h, the reaction
did not proceed, yielding a trace amount of the desired product
2 (Eq. 2). As shown in Table 1, a similar result was obtained
using 2,2,5,5-tetramethyltetrahydrofuran as a solvent. Howev-
er, when toluene and propionitrile were used as solvents, the
desired product 2 was obtained in 27% and 75% yields, respec-
tively. An even better result was obtained when the reaction
was carried out in refluxing 1,4-dioxane, indicating that sol-
vents with active hydrogen are essential for the success of
the reaction.
gated the effect of the amount of diethyl phosphite in the reac-
tion (Table 2). As anticipated, the use of 40 mol % of diethyl
phosphite gave 72% yield of 2, which revealed that the phos-
phonyl part of product 2 came not only from the added phos-
phite, but also from the starting acylphosphonate 1. Moreover,
even in the absence of diethyl phosphite, 2 was obtained in
64% yield along with 3 (7%). The formation of 1,4-dioxanyl
byproduct 3 in all cases indicates that the reaction seems to
have been initiated by the addition of 1,4-dioxan-2-yl radical,
which was generated by hydrogen abstraction from 1,4-di-
oxane by benzoyloxyl radical. Since the use of a large excess
of diethyl phosphite (5 equiv) gave a similar result, as com-
pared to 1 equiv of diethyl phosphite, the remaining reactions
were carried out with benzoyl peroxide (20 mol %) as an ini-
tiator and diethyl phosphite (1 equiv) in refluxing 1,4-dioxane.
O
(PhCO2)2 (0.2)
HPO(OEt)2 (1.0)
O
P(OEt)2
reflux
1
ð2Þ
O
(PhCO2)2 (0.2)
O
O
HPO(OiPr)2 (1.0)
O
P(OEt)2
P(OEt)2
1,4-dioxane
reflux, 1 h
1
2
O
O
O
O
O
O
(PhCO2)2
HPO(OEt)2
P(OEt)2
P(OiPr)2
P(OEt)2
+
1,4-dioxane
reflux, 1 h
ð4Þ
2 71%
4 10%
1
ð3Þ
O
O
O
(PhCO2)2 (0.2)
HPO(OEt)2 (1.0)
O
O
O
O
P(OEt)2
+
P(OiPr)2
1,4-dioxane
reflux, 1 h
2
3
5
When 1 was treated with diethyl phosphite (1 equiv) and ben-
zoyl peroxide (20 mol %) in 1,4-dioxane, it was consumed
smoothly and, after 1 h refluxing, acylphosphorylation product
2 was obtained in 82% yield along with the formation of 1,4-
dioxanyl compound 3 (5%) (Eq. 3). Although it is not yet
clear, the key mediator, the phosphonyl radical, may be gener-
ated via i) hydrogen abstraction from diethyl phosphite by a
solvent-derived radical and/or ii) an initial addition of the sol-
vent-derived radical to the alkenyl part of 1, followed by radi-
cal cyclization and the elimination of the phosphonyl radical.
Furthermore, if the cyclization proceeded by the radical chain
mechanism depicted in Scheme 1, the acylphosphorylation
product 2 should have been obtained with a catalytic amount
or even in the absence of diethyl phosphite. Thus, we investi-
2 9%
+
4 75%
To clearly confirm whether the phosphonyl group of the
product 2 originated from diethyl phosphite or 1, phosphonyl
group scrambling experiments were performed. As shown in
Eq. 4, when 1 was reacted with diisopropyl phosphite under
the same condition, 2 bearing the diethyl phosphonyl group
originated from the starting acylphosphonate 1 was isolated as
a major product (71%) along with 4 (10%). A similar result was
also obtained with 5. The results obtained here clearly indicate
that the eliminated phosphonyl radical not only works as a rad-
ical carrier, but also efficiently adds to the alkenyl moiety of
the acylphosphonate. This acylphosphorylation reaction exhib-

C. H. Cho et al.
Bull. Chem. Soc. Jpn., 78, No. 9 (2005) 1667
O
O
O
O
O
O
(PhCO2)2 (0.2)
HPO(OEt)2 (1.0)
300 nm
P(OEt)2
PPh2
PPh2
benzene
1,4-dioxane, reflux
Me
6
12
13 93%
O
Me
O
O
O
O
7 75%
P(OEt)2
O
PPh2
PPh2
O
O
(PhCO2)2 (0.2)
HPO(OEt)2 (1.0)
1,4-dioxane, reflux
15
0%
14
P(OEt)2
1) 300 nm, 12 h, benzene
( )n
2) PhCOP(O)Ph2(0.2), 300 nm,
24 h, benzene
8a: n = 1
8b: n = 2
20%
48%
70%
3) PhCOP(O)Ph2(1.0), 300 nm,
24 h, benzene
O
O
9a 75%
9b 92%
P(OEt)2
4) (PhCO2 2 (0.2), Ph2P(O)H (0.2),
)
reflux, 2 h, dioxane
( )n
Scheme 2. Cyclizations of alkenyl acylphosphonates under
thermal conditions.
Scheme 3. Cyclizations of 12 and 14 under photochemical
and thermal conditions.
O
O
O
*
O
O
O
300 nm
PPh2
PPh2
PPh2
O
PPh2
a
b
O
O
PPh
O
O
PPh2
2
O
O
O
PPh2
PPh2
O
PPh2
Scheme 4. Plausible mechanism of a photochemically induced acylphosphorylation.
ited a wide generality. As shown in Scheme 2, the presence of
of 14 was irradiated at 300 nm for 12 h, the reaction did
not occur, and the starting material (91%) was recovered un-
changed (Scheme 3).
a methyl group in 6 did not cause any problems, yielding 7 in
75% yield. Furthermore, the reaction of aliphatic acylphos-
phonates 8a and 8b under the same conditions provided ꢂ-
phosphonomethyl cyclopentanone 9a and cyclohexanone 9b
in 75% and 92% yield, respectively.
O
O
O
O
PPh2
300 nm
benzene
PPh2
We next studied a photochemically induced acylphospho-
rylation. As we anticipated, the irradiation of a benzene solu-
tion of 8a at 300 nm for 6 h did not afford 9a, and the starting
material was recovered. Evidently, an acyl radical was not
generated under the condition due to weak UV absorption.
However, when acyldiphenylphosphine oxide 10a was used
to generate the acyl radical under the photochemically initiated
condition, the reaction proceeded smoothly, yielding cyclopen-
tanone 11a in 73% yield under the same condition. A similar
result was obtained with 10b, yielding diphenylphosphinoyl
cyclohexanone 11b in 95% yield (Eq. 5). Furthermore, the ir-
radiation of a benzene solution of 12 at 300 nm for 2 h afford-
ed the desired product 13 in 93% yield. However, contrary to
the result obtained with 6 in Scheme 2, the irradiation of 14 in
benzene at 300 nm did not give 15. When a benzene solution
ð5Þ
( )n
( )n
10a: n = 1
10b: n = 2
11a 73%
11b 95%
There are two possible reaction pathways in the photochem-
ically induced acylphosphorylation of 12. Pathway a involves
the intramolecular addition of a photochemically generated
radical pair of the acyl radical and the phosphinoyl radical to
the pendant alkenyl moiety, as shown in Scheme 4,10 whereas
pathway b involves the addition of the photochemically gener-
ated diphenyl phosphinoyl radical onto the alkenyl group, fol-
lowed by cyclization onto the acylphosphonate group. Thus,
pathway b would follow a similar mechanism to that shown
in the thermally induced cyclization of alkenyl acylphospho-

1668 Bull. Chem. Soc. Jpn., 78, No. 9 (2005)
Radical Cyclizations of Acylphosphonates
nates (Scheme 1). To gain mechanistic insights of the photo-
chemically induced cyclization, we performed a cross-over
experiment using an equimolar mixture of two different acyl-
phosphine oxides, 16 and 17. When an equimolar mixture of
16 and 17 in benzene was irradiated at 300 nm for 2 h, the
product consisted of a mixture of four products, from which
a mixture of 18 and 19 was separated from a mixture of 20
and 21 by column chromatography. The ratio of each product
XH
22a: X = O
22b: X = S
diphosgene
CH2Cl2
O
O
O
1
was determined by H NMR analysis (Eq. 6). The result ob-
P(OEt)3
CH2Cl2
X
P(OEt)2
X
Cl
tained in this study indicates that the photochemically induced
reaction does not proceed via the addition of the radical pair,
and supports pathway b.
23a: X = O
23b: X = S
24: X = O 83%
25: X = S 75%
O
O
Ph2P-OEt
CH2Cl2
O
O
P(p-Tol)2
PPh2
+
hv
EtO2C
EtO2C
MeO2C
MeO2C
O
benzene
O
X
PPh2
16
17
26: X = O 71%
27: X = S 88%
O
O
O
O
+
P(p-Tol)2
PPh2
MeO2C
MeO2C
Scheme 5. Preparation of phosphonoformate 24, phos-
phonothiolformate 25, phosphinoformate 26, and phos-
phinothiolformate 27.
ð6Þ
MeO2C
MeO2C
18 20%
19 20%
O
O
O
O
O
O
+
+
O
P(OEt)2
PPh2
P(p-Tol)2
EtO2C
EtO2C
EtO2C
EtO2C
20 21%
21 19%
O
(PhCO2)2 (0.2)
HPO(OEt)2 (1.0)
O
In order to explain why 14 acts differently under photo-
chemical conditions, several experiments were performed, as
shown in Scheme 3. To initiate the reaction, when the reaction
was carried out in the presence of 0.2 equiv of benzoyl diphen-
ylphosphine oxide, 15 was obtained in 20% yield, whereas the
use of 1 equiv of benzoyl diphenylphosphine oxide gave 15 in
48% yield. Furthermore, a thermal reaction of 14 with benzoyl
peroxide (20 mol %) and diphenylphosphine oxide (0.2 equiv)
in refluxing dioxane gave 15 in 70% yield. The results suggest
that cleavage of the carbon–phosphorous bond in 14 did not
occur under photochemical conditions, thereby failing to gen-
erate the diphenylphosphinoyl radical. However, we do not
know why 14 does not undergo photolysis to initiate the
reaction.
We next briefly studied a feasibility of using a phosphono-
formate and a phosphonothiolformate as a radical acceptor to
explore whether the present acylation would be further extend-
ed to the synthesis of lactone 2811 and thiolactone 30,12 respec-
tively (Scheme 6). Diethylphosphonoformate 24 was prepared
by the treatment of chloroformate 23a with triethyl phosphite
in dichloromethane at room temperature for 2 h (Scheme 5).13
Similarly, S-but-3-enyl diethylphosphonothiolformate (25)
was prepared from 22b by a two-step procedure.14 The treat-
ment of 22b with diphosgene in dichloromethane afforded
23b, which was further treated with triethyl phosphite to give
phosphonothiolformate 25 in 75% yield. Furthermore, the
treatment of 23a with ethyl diphenylphosphinite in dichloro-
O
O
P(OEt)2
1,4-dioxane, reflux
24
O
O
O
O
O
P(OEt)2
O
P(OEt)2
P(OEt)2
+
+
24
O
28 0%
29 11%
64%
O
(PhCO2)2 (0.2)
HPO(OEt)2 (1.0)
S
S
P(OEt)2
1,4-dioxane, reflux
25
O
O
O
O
P(OEt)2
+
+
25
S
P(OEt)2
P(OEt)2
O
30 65%
31 10%
17%
Scheme 6. Cyclizations of 24 and 25 under thermal conditions.
methane at room temperature gave S-but-3-enyl diphenylphos-
phinoformate (26) in 71% yield. Similarly, diphenylphosphi-
nothiolformate 27 was prepared in 88% yield by the treatment
of 23b with ethyl diphenylphosphinite under the same condi-
tion.

C. H. Cho et al.
Bull. Chem. Soc. Jpn., 78, No. 9 (2005) 1669
The reaction of 24 with diethyl phosphite (1 equiv) and ben-
zoyl peroxide (20 mol %) in refluxing 1,4-dioxane for 6 h did
not afford lactone 28, yielding 29 in 11% yield along with re-
covery of the starting material (64%). The failure of the cycli-
zation may be explained as an unfavorable E-conformation in
carboxylic esters, although the low reactivity of the alkoxycar-
bonyl group as a radical acceptor can not be excluded.15 How-
ever, it is known that the differences in energy are generally
smaller for the E- and Z-conformations in thiol esters than
for the corresponding carboxylic esters.16 When the radical
cyclization of 25 was carried out under similar conditions,
the desired thiolactone 30 was obtained in 65% yield along
with 31 (10%) and the starting material (17%).
The photochemically induced radical cyclization of phos-
phonothiolformates occurred smoothly and was cleaner than
the thermally induced reaction, as shown in Scheme 7. As
we observed previously, the irradiation of 26 in benzene at
300 nm for 24 h did not afford lactone 32. However, the pho-
tochemically induced reaction of 27 in benzene at 300 nm for 1
h afforded thiolactone 34 in 95% yield without the formation
of addition product 35. However, this approach turned out to
be effective only for the synthesis of 5-membered thiolactones.
The irradiation of a benzene solution of S-pent-4-enyl di-
phenylphosphinothiolformate (36a) at 300 nm for 1 h did not
give thiolactone 38a, yielding tetrahydrothiophene 39a in
97% yield. A similar result was also obtained with 36b. Since
6-exo and 7-exo ring closures in 37 are slower than 5-exo ring
closure in the cyclization of 27, intramolecular homolytic sub-
stitution at sulfur occurred, yielding 39 along with the libera-
tion of a phosphonocarbonyl radical. The loss of carbon mon-
oxide from the phosphonocarbonyl radical would generate the
diphenyl phosphinoyl radical for radical chain propagation.
The generation of acyl radicals from thiol esters under tin-free
conditions was noted previously.17
Conclusion
The results obtained in this study suggest that alkyl radicals
add to acylphosphonates, and ꢁ-fragmentation of the resulting
alkoxyl radicals affords 5- and 6-membered cyclic ketones
along with the liberation of phosphonyl radicals under thermal
and photochemical conditions. The eliminated phosphonyl
radicals mainly mediate the acylphosphorylation of alkenyl
acylphosphonates by their facile additions to alkenyl bonds.
A small portion of phosphonyl radicals is generated from the
added dialkyl phosphite, and the use of solvents with active
hydrogen, like 1,4-dioxane, is essential for the success of the
thermal acylphosphorylation. The photochemically induced
cyclizations of alkenyl acyldiphenylphosphine oxides are initi-
ated by the generation of acyl radicals and diphenylphosphi-
noyl radicals; the reactions then proceed by similar pathway
as that involved under thermal conditions.
The cyclizations of alkenyl phosphonoformates do not occur
under thermal and photochemical conditions, apparently due to
an unfavorable E-conformation in carboxylic esters. However,
the cyclization of S-but-3-enyl phosphinothiolformates gives
thiolactones under thermal and photochemical conditions,
whereas S-pent-4-enyl phosphinothiolformate affords the tetra-
hydrothiophene derivative under similar conditions, apparently
due to an intramolecular homolytic substitution at sulfur.
O
O
O
O
O
O
PPh2
300 nm
O
PPh2
O
PPh2
+
+
26
O
C6H6, 24 h
PPh2
O
26
32 0%
33 0%
84%
O
O
O
O
O
O
300 nm
+
S
PPh2
S
PPh2
S
PPh2
C6H6, 1 h
PPh2
O
27
34 95%
35 0%
O
O
O
O
O
O
300nm
O
S
PPh2
S
PPh2
S
PPh2
( )n
38a: n = 1
38b: n = 2
( )n
( )n
PPh2
PPh2
O
36a: n = 1
36b: n = 2
37
O O
C PPh2
CO
O
S
PPh2
( )n
39a: n = 1 97%
39b: n = 2 91%
Scheme 7. Cyclizations of phosphonoformates and phosphonothiolformates under photochemical conditions.

1670 Bull. Chem. Soc. Jpn., 78, No. 9 (2005)
Radical Cyclizations of Acylphosphonates
mg, 0.30 mmol) in 1,4-dioxane (2 mL), diethyl phosphite (38.6
mL, 41.3 mg, 0.30 mmol), and 1,4-dioxane (1 mL), successively.
The mixture was refluxed for 1 h and evaporated to dryness. The
residue was subjected to silica-gel column chromatography or
preparative TLC (hexane/acetone 3:2 v/v) to afford ꢄ-keto-
phosphonate 2 (72.9 mg, 0.25 mmol, 82%).
Experimental
General. 1H, 13C, and 31P NMR spectra were determined for
solutions in CDCl3 or benzene-d6 on Bruker DRX500, Bruker
Avance-400, or JEOL AL-400 NMR spectrometers, and recorded
with tetramethylsilane (TMS) (ꢃ ¼ 0), CDCl3 (ꢃ ¼ 77:0), and
85% H3PO4 aq (ꢃ ¼ 0) as internal or external references, unless
otherwise noted. Infrared spectra (IR) were recorded on HORIBA
FT-300S or VECTOR-33 spectrometers. High-resolution mass
spectra (HRMS) were obtained on JEOL JMS-SX 102A using
FAB ionization with m-nitrobenzylalcohol (NBA) as a matrix or
a VG AUTOSPEC Ultma GC/MS system using a direct insertion
probe (DIP) and an electron impact (EI) (70 eV) method. Elemen-
tal analyses were carried out at the Elemental Analysis Laborato-
ry, Department of Chemistry, Faculty of Science, The University
of Tokyo.
1,4-Dioxane was distilled over LiAlH4 under argon before use.
Dichloromethane (CH2Cl2) was distilled twice from P4O10, then
from CaH2, and stored over molecular sieves 4A. Triethyl phos-
phite and diethyl phosphite were purchased from Tokyo Kasei Ko-
gyo Co., Ltd. and distilled before use. Benzoyl peroxide (Kishida
Kagaku) was used as purchased. Wako Gel B-5F (Wako Pure
Chemical Industries, Ltd.) was used for preparative TLC. Column
chromatography was performed with PSQ100B (spherical, neu-
tral, Fuji Silicia Kagaku).
Diethyl
1-Oxo-1,2,3,4-tetrahydronaphthalen-2-ylmethyl-
1
phosphonate (2): Pale yellow oil; H NMR (500 MHz, CDCl3)
ꢃ 1.34 (distorted t, J ¼ 7:1 Hz, 6H), 1.68–1.76 (m, 1H), 1.90–1.99
(m, 1H), 2.61–2.66 (m, 1H), 2.80–2.91 (m, 2H overlapped), 2.96–
3.01 (m, 1H), 3.09–3.15 (m, 1H), 4.08–4.20 (m, 4H), 7.25 (distort-
ed d, J ¼ 7:5 Hz, 1H), 7.31 (distorted dd, J ¼ 7:6, 7.8 Hz, 1H),
7.48 (distorted ddd, J ¼ 1:1, 7.5, 7.6 Hz, 1H), 8.03 (distorted d,
J ¼ 7:8 Hz, 1H); 13C NMR (125 MHz, CDCl3) ꢃ 16.3 (d, JCP
6:0 Hz), 25.0 (d, JCP ¼ 143:3 Hz), 29.0, 30.0, 43.0, 61.5 (d, JCP
¼
¼
6:4 Hz), 61.7 (d, JCP ¼ 6:5 Hz), 126.5, 127.5, 128.7, 131.8, 133.4,
143.9, 197.7 (d, JCP ¼ 15:6 Hz); 31P NMR (202 MHz, CDCl3) ꢃ
32; IR (ZnSe) 517, 744, 960, 1020, 1053, 1234, 1456, 1599,
1682, 2981 cmꢂ1; Found: C, 60.64; H, 7.26%. Calcd for
C15H21O4P: C, 60.80; H, 7.14%.
Typical Procedure for the Intramolecular Acylphosphoryl-
ation (Photochemical Conditions). After a dry benzene solution
(2.0 mL, 0.2 M in the acylphosphine oxide) of acylphosphine ox-
ide 10a (119.3 mg, 0.4 mmol) in a quartz tube was degassed with
nitrogen for 10 min, the solution was irradiated at 300 nm in a
Rayonet photochemical reactor for 2 h. Any volatile matter was
evaporated. The residue was subjected to silica-gel column chro-
matography (hexane/acetone 3:2 v/v) to afford ꢄ-ketophosphine
oxide 11a (87.2 mg, 0.29 mmol, 73%).
Typical Procedure for the Preparation of Acylphospho-
nates. All acylphosphonates (1, 5, 6, 8a, and 8b) and acylphos-
phine oxides (10a, 10b, 12, 14, and 16) were synthesized by the
Arbuzov reaction from the corresponding acid chloride and trieth-
yl phosphite or ethyl diphenylphosphinite, respectively. The acid
chloride was in situ prepared from the corresponding carboxylic
acid according to literature.13
Diphenyl 2-Oxocyclopentylmethylphosphine Oxide (11a):
1
Colorless oil; H NMR (400 MHz, CDCl3) ꢃ 1.27–1.50 (m, 1H),
1.52–1.70 (m, 1H), 1.82–2.05 (m, 3H), 2.10–2.32 (m, 3H), 2.95
To a solution of 2-but-3-enylbenzoic acid (0.524 g, 2.97 mmol)
in CH2Cl2 (3 mL) was slowly added oxalyl chloride (2 M in
CH2Cl2, 3.0 mL, 6.00 mmol, Aldrich Chemical Co.,ꢁInc.) at 0
ꢁC. The reaction mixture was stirred for 30 min at 0 C and al-
lowed to warm up to room temperature. The solvent and volatiles
were removed in vacuo after gas generation ended. To the crude
acid chloride was added 3 mL of CH2Cl2 and cooled to 0 ꢁC; then,
triethyl phosphite (0.51 mL, 0.494 g, 2.97 mmol) was added drop-
(ddd, J ¼ 1:9, 9.9, 15.3 Hz, 1H), 7.26–7.50 (m, 6H), 7.55–7.80
(m, 4H); 13C NMR (100 MHz, CDCl3) ꢃ 20.3, 29.5 (d, JCP
73:5 Hz), 30.8, 36.4, 43.6 (d, JCP ¼ 3:6 Hz), 128.5 (d, JCP
¼
¼
11:8 Hz), 128.6 (d, JCP ¼ 11:8 Hz), 130.4 (d, JCP ¼ 9:4 Hz),
130.6 (d, JCP ¼ 9:2 Hz), 131.6 (d, JCP ¼ 2:7 Hz), 131.7 (d, JCP
¼
2:6 Hz), 132.1 (d, JCP ¼ 98:7 Hz), 133.2 (d, JCP ¼ 98:7 Hz),
218.9 (d, JCP ¼ 13:7 Hz); 31P NMR (121.5 MHz, CDCl3) ꢃ 32;
IR (ZnSe) 538, 698, 749, 1120, 1183 (P=O), 1438 (P–Ph), 1739
ꢁ
wise. The reaction mixture was stirred at 0 C for 1 h and over-
(C=O) cmꢂ1 HRMS (Mþ) Found: 298.1122, Calcd for
;
night at room temperature. The solvent was removed under re-
duced pressure to afford almost pure acylphosphonate 1. The
crude product was quickly subjected to short silica-gel column
chromatography (hexane/ethyl acetate 1:1 v/v) to give pure
acylphosphonate 1 (0.744 g, 2.51 mmol, 85%).
C18H19O2P: 298.1123.
Cross-Over Experiment of a Photochemically Induced
Acylphosphorylation. After a dry benzene solution (2.0 mL)
of acylphosphine oxide 16 (41.4 mg, 0.1 mmol) and 17 (47.1
mg, 0.1 mmol) in a quartz tube was degassed with nitrogen for
10 min, the solution was irradiated at 300 nm in a Rayonet photo-
chemical reactor for 2 h. Any volatile matter was evaporated. The
residue was subjected to silica-gel column chromatography (hex-
ane/acetone 3:2 v/v) to afford a mixture of ꢄ-ketophosphine ox-
Diethyl 2-But-3-enylbenzoylphosphonate (1): Yellow oil;
1H NMR (500 MHz, CDCl3) ꢃ 1.37 (distorted t, J ¼ 7:0 Hz,
6H), 2.31–2.35 (m, 2H), 2.95 (t, J ¼ 7:8 Hz, 2H), 4.23–4.29 (m,
4H), 4.95–5.03 (m, 2H overlapped), 5.82–5.90 (m, 1H), 7.29 (dis-
torted d, J ¼ 7:5 Hz, 1H), 7.36 (distorted dd, J ¼ 7:6, 7.8 Hz, 1H),
7.49 (distorted dd, J ¼ 7:5, 7.6 Hz, 1H), 8.43 (distorted d, J ¼ 7:8
Hz, 1H); 13C NMR (125 MHz, CDCl3) ꢃ 16.3 (d, JCP ¼ 5:6 Hz),
1
ides, 18 and 19 (33.6 mg, H NMR ratio 0.99:1.00, 20%, 20%),
and a mixture of ꢄ-ketophosphine oxides, 20 and 21 (37.1 mg,
1H NMR ratio 1.09:1.00, 21%, 19%).
3-Oxo-4-diphenylphosphinoylmethylcyclopentane-1,1-di-
33.5, 35.4, 63.9 (d, JCP ¼ 7:4 Hz), 115.0, 126.1, 131.5 (d, JCP
¼
1
3:9 Hz), 132.5, 133.2, 134.6 (d, JCP ¼ 63:0 Hz), 137.9, 143.6
(d, JCP ¼ 9:3 Hz), 201.6 (d, JCP ¼ 170:5 Hz); 31P NMR (202
MHz, CDCl3) ꢃ ꢂ1; IR (ZnSe) 568, 744, 773, 922, 974, 1016,
1250, 1652, 2981 cmꢂ1; Found: C, 60.81; H, 7.12%. Calcd for
C15H21O4P: C, 60.80; H, 7.14%.
carboxylic Acid Dimethyl Ester (18): Colorless oil; H NMR
(400 MHz, C6D6) ꢃ 1.78–1.93 (m, 1H), 2.26 (dd, J ¼ 12:0, 13.7
Hz, 1H), 2.65–2.80 (m, 2H), 2.85–2.97 (m, 2H), 3.10 (s, 3H),
3.21 (s, 3H), 3.22–3.38 (m, 1H), 6.93–7.07 (m, 6H), 7.64–7.79
(m, 4H); 13C NMR (100 MHz, C6D6) ꢃ 30.6 (d, JCP ¼ 72:4 Hz),
37.4, 42.6 (d, JCP ¼ 3:6 Hz), 44.1, 52.5 (d, JCP ¼ 2:9 Hz), 55.2,
128.7 (d, JCP ¼ 10:8 Hz), 128.8 (d, JCP ¼ 11:1 Hz), 130.9 (d,
JCP ¼ 9:1 Hz), 131.2 (d, JCP ¼ 9:2 Hz), 131.5 (d, JCP ¼ 3:0
Typical Procedure for the Intramolecular Acylphosphoryl-
ation (Thermal Conditions). To benzoyl peroxide (15.4 mg,
0.06 mmol) were added a solution of acylphosphonate 1 (88.9

C. H. Cho et al.
Bull. Chem. Soc. Jpn., 78, No. 9 (2005) 1671
Hz), 131.6 (d, JCP ¼ 2:8 Hz), 133.6 (d, JCP ¼ 97:7 Hz), 135.1 (d,
JCP ¼ 97:7 Hz), 171.0, 171.2, 212.7 (d, JCP ¼ 13:5 Hz); 31P NMR
(121.5 MHz, C6D6) ꢃ 29; IR (ZnSe) 504, 702, 744, 1121, 1439 (P–
Ph), 1733 (C=O) cmꢂ1; HRMS (Mþ) Found: 414.1231, Calcd for
C22H23O6P: 414.1232.
But-3-enyl Diethylphosphonoformate (24): Colorless oil;
1H NMR (400 MHz, CDCl3) ꢃ 1.29 (t, J ¼ 9:5 Hz, 6H), 2.32–
2.42 (m, 2H), 4.11–4.25 (m, 6H), 4.96–5.10 (m, 2H), 5.60–5.75
(m, 1H); 13C NMR (100 MHz, CDCl3) ꢃ 16.0 (d, JCP ¼ 5:5 Hz),
32.6, 64.2 (d, JCP ¼ 11:8 Hz), 65.6, 117.5, 133.0, 166.7 (d, JCP
¼
3-Oxo-4-(bis(4-methylphenyl)phosphinoyl)methylcyclopen-
tane-1,1-dicarboxylic Acid Dimethyl Ester (19): Colorless oil;
1H NMR (400 MHz, C6D6) ꢃ 1.78–1.93 (m, 1H), 1.94 (d, J ¼ 11:9
Hz, 6H), 2.19–2.35 (m, 1H), 2.65–2.80 (m, 2H), 2.85–2.97 (m,
2H), 3.10 (s, 3H), 3.21 (s, 3H), 3.22–3.38 (m, 1H), 6.79–6.89
(m, 4H), 7.64–7.79 (m, 4H); 31P NMR (121.5 MHz, C6D6) ꢃ 29.
3-Oxo-4-diphenylphosphinoylmethylcyclopentane-1,1-dicar-
356:5 Hz); 31P NMR (121.5 MHz, CDCl3) ꢃ ꢂ4; IR (ZnSe) 559,
586, 1022, 1164 (P=O), 1212, 1274, 1719 (C=O), 2985 cmꢂ1
HRMS (Mþ) Found: 236.0840, Calcd for C9H17O5P: 236.0814.
;
Supporting Information
Additional spectral data of all new compounds. This material is
available free of charge on the web at http://www.csj.jp/journals/
bcsj/.
1
boxylic Acid Diethyl Ester (20): Colorless oil; H NMR (400
MHz, C6D6) ꢃ 0.74 (t, J ¼ 7:2 Hz, 3H), 0.83 (t, J ¼ 7:2 Hz, 3H),
1.81–1.93 (m, 1H), 2.31 (dd, J ¼ 12:0, 13.6 Hz, 1H), 2.70–2.83
(m, 2H), 2.89–3.02 (m, 2H), 3.23–3.36 (m, 1H), 3.68–3.80 (m,
2H), 3.86 (q, J ¼ 7:1 Hz, 2H), 6.93–7.07 (m, 6H), 7.64–7.80
References
1
a) B. Giese, ‘‘Radicals in Organic Synthesis: Formation of
(m, 4H); 13C NMR (100 MHz, C6D6) ꢃ 13.7, 13.8, 30.7 (d, JCP
¼
Carbon–Carbon Bonds,’’ Pergamon, New York (1986). b) D. P.
Curran, ‘‘In Comprehensive Organic Synthesis,’’ ed by B. M.
Trost and I. Fleming, Pergamon Press, Oxford (1991), Vol. 4,
pp. 715–831. c) C. P. Jasperse, D. P. Curran, and T. L. Fevig,
Chem. Rev., 91, 1237 (1991). d) W. B. Motherwell and D. Crich,
‘‘Free Radical Chain Reactions in Organic Synthesis,’’ Academic,
London (1992). e) ‘‘Radicals in Organic Synthesis,’’ ed by P.
Renaud and M. P. Sibi, Wiley VCH, Weinheim (2001).
72:6 Hz), 37.4, 42.7 (d, JCP ¼ 3:6 Hz), 44.1, 55.4, 61.8, 128.7 (d,
JCP ¼ 11:0 Hz), 128.8 (d, JCP ¼ 11:0 Hz), 130.9 (d, JCP ¼ 9:2
Hz), 131.2 (d, JCP ¼ 8:7 Hz), 131.5 (d, JCP ¼ 2:6 Hz), 131.5 (d,
JCP ¼ 2:8 Hz), 133.6 (d, JCP ¼ 97:7 Hz), 135.1 (d, JCP ¼ 97:7
Hz), 170.6, 170.9, 213.0 (d, JCP ¼ 13:7 Hz); 31P NMR (121.5
MHz, C6D6) ꢃ 29; IR (ZnSe) 505, 540, 699, 748, 1120, 1181
(P=O), 1439 (P–Ph), 1731 (C=O), 2983 cmꢂ1; HRMS (Mþ)
Found: 442.1546, Calcd for C24H27O6P: 442.1545.
2
a) A. G. Fallis and I. M. Brinza, Tetrahedron, 53, 17543
3-Oxo-4-(bis(4-methylphenyl)phosphinoyl)methylcyclopen-
tane-1,1-dicarboxylic Acid Diethyl Ester (21): Colorless oil;
1H NMR (400 MHz, C6D6) ꢃ 0.82 (t, J ¼ 7:2 Hz, 3H), 0.90 (t, J ¼
7:2 Hz, 3H), 1.85–2.12 (m, 1H), 2.00 (d, J ¼ 12:3 Hz, 6H), 2.42
(dd, J ¼ 12:0, 13.7 Hz, 1H), 2.77–2.92 (m, 2H), 3.01–3.11 (m,
2H), 3.47 (ddd, J ¼ 1:8, 8.7, 13.6 Hz, 1H), 3.75–3.89 (m, 2H),
3.93 (q, J ¼ 7:0 Hz, 2H), 6.90–6.99 (m, 4H), 7.70–7.83 (m,
(1997). b) G. K. Friestad, Tetrahedron, 57, 5461 (2001). c) H.
Miyabe, M. Ueda, and T. Naito, Synlett, 2004, 1140. d) S. Kim,
Adv. Synth. Catal., 346, 19 (2004).
3
a) A. L. J. Beckwith and B. P. Hay, J. Am. Chem. Soc., 111,
230 (1989). b) A. L. J. Beckwith and B. P. Hay, J. Am. Chem.
Soc., 111, 2674 (1989). c) R. Walton and B. Fraser-Reid, J. Am.
Chem. Soc., 113, 5791 (1991).
4H); 13C NMR (100 MHz, C6D6) ꢃ 13.7, 13.8, 21.2, 31.0 (d, JCP
¼
4
5
S. Kim and S. Y. Jon, Chem. Commun., 1996, 1335.
a) S. Kiyooka, Y. Kaneko, H. Matsue, M. Hamada, and
72:6 Hz), 37.6, 42.8 (d, JCP ¼ 3:6 Hz), 44.2, 55.5, 61.7, 61.7,
129.4 (d, JCP ¼ 11:8 Hz), 129.6 (d, JCP ¼ 11:9 Hz), 130.7 (d,
JCP ¼ 98:9 Hz), 131.0 (d, JCP ¼ 9:4 Hz), 132.3 (d, JCP ¼ 9:2
Hz), 132.3 (d, JCP ¼ 100:1 Hz), 141.7 (d, JCP ¼ 2:6 Hz), 170.6,
170.9, 213.0 (d, JCP ¼ 13:5 Hz); 31P NMR (121.5 MHz, C6D6)
ꢃ 29; IR (ZnSe) 505, 528, 658, 741, 809, 1118, 1180 (P=O),
1446 (P–Ph), 1731 (C=O), 2982 cmꢂ1; HRMS (Mþ) Found:
470.1859, Calcd for C26H31O6P: 470.1858.
R. Fujiyama, J. Org. Chem., 55, 5562 (1990). b) D. P. Curran
and H. Liu, J. Org. Chem., 56, 3463 (1991). c) D. P. Curran
and M. Palovich, Synlett, 1992, 631. d) D. P. Curran, U.
Diederichsen, and M. Palovich, J. Am. Chem. Soc., 119, 4797
(1997). e) U. Diederichsen and D. P. Curran, J. Organomet.
Chem., 531, 9 (1997).
6
a) M. Sekine, M. Satoh, H. Yamagata, and T. Hata, J. Org.
Typical Procedure for the Preparation of Phosphonofor-
mates and Phosphonothiolformates. All phosphonoformates
(24 and 26) and phosphonothiolformates (25, 27, 36a, and 36b)
were synthesized by the Arbuzov reaction from the corresponding
chloroformate or chlorothiolformate and triethyl phosphite or eth-
yl diphenylphosphinite. The chloroformate and chlorothiolformate
were in situ prepared from the corresponding alcohol or thiol
using diphosgene and triethylamine.
To a solution of homoallyl alcohol 22a (260 mL, 3.0 mmol) in
CH2Cl2 (10 mL), diphosgene (540 mL, 4.5 mmol), and triethyl-
amine (418 mL, 3.0 mmol) were added at 0 ꢁC. After being stirred
for 1 h at room temperature, the solvent and excess diphosgene
were removed under reduced pressure and filtered through a celite
pad using diethyl ether. After the crude chloroformate was dis-
solved in dichloromethaneꢁ(10 mL), triethyl phosphite (510 mL,
3.0 mmol) was added at 0 C. After being stirred for 1 h at room
temperature, the solvent was evaporated under reduced pressure
and the residue was separated by silica-gel column chromatogra-
phy (ethyl acetate/n-hexane 1:1 v/v) to give phosphonoformate
24 (588.1 mg, 83%).
Chem., 45, 4162 (1980). b) M. Sekine, A. Kume, and T. Hata,
Tetrahedron Lett., 22, 3617 (1981). c) M. Sekine, A. Kume, M.
Nakajima, and T. Hata, Chem. Lett., 1981, 1087.
7
a) A. Kajiwara, Y. Konishi, Y. Morishima, W. Schnabel,
K. Kuwata, and M. Kamachi, Macromolecules, 26, 1656 (1993).
b) S. Jockusch, I. V. Koptyug, P. F. McGarry, G. W. Sluggett,
N. J. Turro, and D. M. Watkins, J. Am. Chem. Soc., 119, 11495
(1997). c) M. T. L. Rees, G. T. Russell, M. D. Zammit, and
T. P. Davis, Macromolecules, 31, 1763 (1998).
8 a) S. Kim, C. H. Cho, and C. J. Lim, J. Am. Chem. Soc.,
125, 9574 (2003). b) C. H. Cho and S. Kim, Can. J. Chem.,
(2005), in press.
9
For addition of phosphorus-centered radicals to alkenes,
see for example: a) J. M. Barks, B. C. Gilbert, A. F. Parsons,
and B. Upeandran, Tetrahedron Lett., 42, 3137 (2001). b) J. M.
Barks, B. C. Gilbert, A. F. Parsons, and B. Upeandran, Synlett,
2001, 1719. c) S. R. Piettre, Tetrahedron Lett., 37, 2233 (1996).
d) G. W. Sluggett, P. F. McGarry, I. V. Koptyug, and N. J. Turro,
J. Am. Chem. Soc., 118, 7367 (1996). e) S. R. Piettre, Tetrahedron
Lett., 37, 4707 (1996). f) R. L. Kenney and G. S. Fisher, J. Org.

1672 Bull. Chem. Soc. Jpn., 78, No. 9 (2005)
Radical Cyclizations of Acylphosphonates
Chem., 39, 682 (1974). g) H. J. Callot and C. Benezra, Can. J.
Chem., 49, 500 (1971). h) W. K. Busfield, I. D. Grice, and I. D.
Jenkins, Aust. J. Chem., 48, 625 (1995). i) C. Lopin, A. Gautier,
G. Gouhier, and S. R. Piettre, Tetrahedron Lett., 41, 10195
(2000). j) O. Dubert, A. Gautier, E. Condamine, and S. R. Piettre,
Org. Lett., 4, 359 (2002). k) J. E. Brumwell, N. S. Simpkins, and
N. K. Terrett, Tetrahedron, 50, 13533 (1994). l) D. H. R. Barton,
D. O. Jang, and J. C. Jaszberenyi, J. Org. Chem., 58, 6838 (1993).
m) C. M. Jessop, A. F. Parsons, A. Routledge, and D. Irvine,
Tetrahedron Lett., 44, 479 (2003). n) C. Lopin, G. Gouhier,
A. Gautier, and S. R. Piettre, J. Org. Chem., 68, 9916 (2003). o)
C. Lopin, G. Gouhier, A. Gautier, and S. R. Piettre, J. Am. Chem.
Soc., 124, 14668 (2002).
2, 389 (2000).
12 a) M. Tada, T. Nakamura, and M. Matsumoto, J. Am.
Chem. Soc., 110, 4647 (1988). b) M. Tada, M. Matsumoto, and
T. Nakamura, Chem. Lett., 1988, 199. c) I. Ryu, T. Okuda, K.
Nagahara, N. Kambe, M. Komatsu, and N. Sonoda, J. Org. Chem.,
62, 7550 (1997).
13 M. Depature, J. Diewok, J. Grimaldi, and J. Hatem, Eur. J.
Org. Chem., 2000, 275.
14 C. J. Salomon and E. Breuer, Synlett, 2000, 815.
15 a) O. Exner, ‘‘The Chemistry of Double-Bonded Func-
tional Groups,’’ ed by S. Patai, Interscience, London (1977), p. 1.
b) E. L. Eliel and S. H. Wilen, ‘‘Stereochemistry of Organic Com-
pounds,’’ Wiley-Interscience, New York (1994), pp. 618–621.
16 D. M. Pawar, A. A. Khalil, D. R. Hooks, K. Collins, T.
Elliott, J. Stafford, L. Smith, and E. A. Noe, J. Am. Chem. Soc.,
120, 2108 (1998).
10 a) D. H. R. Barton, D. Crich, and P. Potier, Tetrahedron
Lett., 26, 5943 (1985). b) F. Gagosz and S. Z. Zard, Synlett,
1999, 1978.
11 a) E. Lee, C. H. Yoon, T. H. Lee, S. Y. Kim, T. J. Ha,
Y. Sung, S.-H. Park, and S. Lee, J. Am. Chem. Soc., 120, 7469
(1998). b) S. Tsunoi, I. Ryu, T. Okuda, M. Tanaka, M. Komatsu,
and N. Sonoda, J. Am. Chem. Soc., 120, 8692 (1998). c) S.
Kreimerman, I. Ryu, S. Minakata, and M. Komatsu, Org. Lett.,
17 a) J. H. Penn and F. Liu, J. Org. Chem., 59, 2608 (1994). b)
D. Crich and Q. Yao, J. Org. Chem., 61, 3566 (1996). c) L. Benati,
R. Leardini, M. Minozzi, D. Nanni, P. Spagnolo, S. Strazzari, and
G. Zanardi, Org. Lett., 4, 3079 (2002).