Phosphonyl radical addition to enol ethers. The stereoselective synthesis of cyclic ethers

Article abstract of DOI:10.1016/j.tetlet.2004.04.173

Addition of diethyl phosphite or diethyl thiophosphite to enol ethers, in the presence of a radical initiator, results in the regioselective synthesis of organophosphonate or phosphonothioate derivatives, respectively, under mild conditions. This method c

Full text of DOI:10.1016/j.tetlet.2004.04.173

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 5095–5098
Phosphonyl radical addition to enol ethers. The stereoselective
synthesis of cyclic ethers
Christopher M. Jessop,a Andrew F. Parsons,a,* Anne Routledgea and Derek J. Irvineb
aDepartment of Chemistry, University of York, Heslington, York YO10 5DD, UK
bUniqema, Wilton Centre, Wilton, Redcar TS10 4RF, UK
Received 18 March 2004; revised 16 April 2004; accepted 29 April 2004
Abstract—Addition of diethyl phosphite or diethyl thiophosphite to enol ethers, in the presence of a radical initiator, results in the
regioselective synthesis of organophosphonate or phosphonothioate derivatives, respectively, under mild conditions. This method
can be applied to the stereoselective formation of substituted tetrahydrofurans and tetrahydropyrans, on cyclisation of vinyl ethers
bearing unsaturated side chains.
ꢀ 2004 Elsevier Ltd. All rights reserved.
One useful method for preparing organophosphorus
derivatives involves the regioselective addition of phos-
phonyl (R2P(O)Å) or thiophosphonyl (R2P(S)Å) radicals
to alkenes.1 Diethyl phosphite ((EtO)2P(O)H) and par-
ticularly diethyl thiophosphite ((EtO)2P(S)H), have been
shown to add efficiently to a variety of electron-rich
alkenes and dienes2 in the presence of a radical initiator.
The requirement for an electron-rich alkene can be
explained by the electrophilicity of the phosphorus-
centred radicals, while the higher yields obtained when
using diethyl thiophosphite presumably reflects a weaker
P–H bond in this compound.
triethylborane (4 · 0.3 equiv) at room temperature
(Scheme 1).5;6 This afforded phosphite 2 in quantitative
yield, following addition of the intermediate phospho-
rus-centred radical to the least hindered end of the vinyl
ether. A similar result was obtained on reaction of 1
with 5 equiv of diethyl thiophosphite to give adduct 3.3a
In both cases, 1.5 equiv of the initiator triethylborane
was added, which reflects the short chain length in these
reactions. Also, a greater excess of the phosphite, com-
pared to thiophosphite, was required for efficient addi-
tion––the excess phosphite or thiophosphite was
removed from the crude product by distillation and
could be recycled.
In contrast to reactions with alkenes, addition of phos-
phonyl and particularly thiophosphonyl radicals, to the
electron-rich double bond of enol ethers has not been
widely studied.3 This is surprising because of the mild
conditions employed and the range of important bio-
logical properties associated with b-alkoxy phosphon-
ates.4 As a consequence, the addition of diethyl
phosphite and diethyl thiophosphite to various enol
ethers, including enol ethers with unsaturated side
chains, was investigated for the first time.
Addition of diethyl phosphite and diethyl thiophosphite
to other enol ethers is also possible. Hence reaction
of diethyl thiophosphite (4.8–5 equiv) with vinyl ace-
tate, 3,4-dihydro-2H-pyran, tri-O-acetyl-D-glucal and
1-(trimethylsilyloxy)cyclohexene, in the presence of tri-
ethylborane (1.2–1.5 equiv) at rt, afforded adducts 4–7 in
63–99% yield (Fig. 1). Whereas 7 was isolated as an
equal mixture of diastereoisomers (from the NMR
Initially, diethyl phosphite (10 equiv) was reacted with
n-butyl vinyl ether 1 (1equiv) in the presence of
Keywords: Cyclisation; Phosphorus compounds; Radicals and radical
reactions.
Scheme 1. Addition of diethyl phosphite and diethyl thiophosphite to
n-butyl vinyl ether 1.
* Corresponding author. Tel.: +44-1904-432608; fax: +44-1904-4325-
16; e-mail: afp2@york.ac.uk
0040-4039/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.04.173

5096
C. M. Jessop et al. / Tetrahedron Letters 45 (2004) 5095–5098
Scheme 3. Addition of diethyl phosphite and diethyl thiophosphite to
1-methyl-4-pentenyl vinyl ether 14 and 3-vinyloxy-6-heptenyl benzoate
16 (the relative stereochemistry of the major isomers of 15 and 17 is
shown).
Figure 1. Organophosphorus adducts prepared by addition of diethyl
thiophosphite or diethyl phosphite to various enol ethers.
spectra), selective anti-addition of the thiophosphonyl
radical to the double bond of tri-O-acetyl-D-glucal was
and 13 can be explained by chemo- and regioselective
addition of the phosphorus-centred radical to the enol
ether double bond followed by 5-exo trig radical cycli-
sation, which is expected to proceed via a chair-like
(Beckwith–Houk) transition state 11.8
observed to give 6 as a single diastereoisomer. The rel-
ative stereochemistry of 6 was assigned as all-trans on
1
the basis of a J4–5 value of 11 Hz in the H NMR spec-
trum together with a positive correlation between H-3
and H-5 in the NOESY NMR spectrum. Reaction of 5
and 20 equiv of diethyl phosphite with 3,4-dihydro-2H-
The stereoselective formation of tetrahydropyrans is
also possible as illustrated by the reaction of vinyl ethers
14 and 16 with diethyl phosphite (10 equiv) (Scheme 3).
Even though a terminal alkene is present in both 14 and
16, the phosphonyl radical prefers to add to the vinyl
ether double bond. This may be explained by polarity
and the fact that the electrophilic phosphorus radical
prefers to add to the more electron-rich double bond.
Tetrahydropyrans 15 and 17 were isolated as mixtures of
inseparable diastereoisomers. The assignment of ste-
reochemistry of the major isomers, shown in Scheme 3,
was made on the basis of the NMR spectra and NOESY
experiments. This stereochemistry is consistent with a
chair-like transition state for the 6-exo trig radical cyc-
lisation, while the minor diastereoisomers of 15 and 17
were assigned the all-cis stereochemistry.9
pyran and tri-O-acetyl-D-glucal was found to afford
reasonable yields of phosphonates 8 and 9, respectively.
However, the yields were lower than those observed
when using diethyl thiophosphite even when using a
large excess of the phosphite.
Following these preliminary studies, our attention
turned to the reaction of diethyl phosphite and diethyl
thiophosphite with enol ethers bearing alkene side
chains. For these substrates, the intermediate phospho-
rus-centred radical could add to the double bond of the
alkene and/or the enol ether and it was of interest to
explore the chemoselectivity of these novel reactions.
Our studies focused on the reaction of phosphorus hy-
drides with vinyl ether 10 as shown in Scheme 2. This
precursor was prepared in 75% yield by reaction of Z-
hexen-3-ol with excess ethyl vinyl ether in the presence
of the catalyst mercury(II) trifluoroacetate.7 Reaction of
10 with diethyl thiophosphite (5 equiv) or diethyl phos-
phite (10 equiv) afforded disubstituted tetrahydrofurans
12 or 13, respectively, as predominately the cis-diaste-
reoisomer (from the NMR spectra). The formation of 12
Surprisingly, when vinyl ether 14 was reacted with di-
ethyl thiophosphite (3 equiv), only secondary alcohol 18
was isolated after column chromatography (Scheme 3).
The thiophosphonyl radical prefers to undergo simple
addition to the alkene double bond of 14 and this is
presumably followed by vinyl ether hydrolysis (on col-
umn chromatography) to give the secondary alcohol.
This change in chemoselectivity could reflect the greater
electrophilicity of the phosphonyl radical, (EtO)2P(O)Å,
compared to the thiophosphonyl radical, (EtO)2P(S)Å.
There are a number of natural and biologically active
five- and six-ring ethers that could be prepared using this
novel cyclisation methodology. This includes ionophore
antibiotics such as tetronomycin 19, which is active
against several Gram-positive bacteria (Fig. 2).10 Yoshii
and co-workers have reported the only total synthesis of
19.11 This approach involves the elaboration of trisub-
stituted tetrahydropyran 20, which was prepared in 17
steps from
L
-ascorbic acid (in ꢀ10% overall yield). As
shown in Scheme 4, our initial work has concentrated on
the development of a shorter racemic synthesis of the
related silyl ether 21, using a phosphonyl radical cycli-
sation in the key step.
Scheme 2. Addition of diethyl phosphite and diethyl thiophosphite to
(Z)-1-(vinyloxy)-3-hexene 10 (the relative stereochemistry of the major
isomers of 12 and 13 is shown).

C. M. Jessop et al. / Tetrahedron Letters 45 (2004) 5095–5098
5097
Figure 2. Tetronomycin 19 and tetrahydropyran 20.
Scheme 5. Reaction of phosphorus hydrides with phenylacetylene and
2-bromoacetophenone.
Wittig reaction to afford the E-alkene 26. Finally,
transfer hydrogenation resulted in efficient deprotection
of the benzyl carbonate15 to afford the corresponding
secondary alcohol. Michael-type cyclisation with KOH
t
followed by equilibration using BuOK11 afforded the
desired tetrahydropyran 21 (as a 4:1mixture of diaste-
reoisomers16). The nine-step synthesis from glycidol
22 resulted in the formation of 21 in an overall yield of
8%.
As both R- and S-enantiomers of glycidol 22 are com-
mercially available, the preparation of either enantio-
meric series of 21 is possible. The synthesis of
tetronomycin 19 will require the S-enantiomer, while the
R-enantiomer could be used as a starting material for
the synthesis of the alternative ionophore, tetronasin
(M139603).17
This work has shown that inexpensive and nontoxic
phosphorus hydrides can add efficiently to a variety of
enol ethers. Slow (syringe-pump) addition of the phos-
phorus hydride is not required for good product yields
due to the relatively slow rate of hydrogen atom
abstraction (in comparison to tin hydrides), although
this means that an excess of the phosphorus hydride is
desirable. Of particular synthetic interest is the regio-
and stereoselective formation of tetrahydrofuran and
tetrahydropyran ring systems. These cyclisation studies
clearly illustrate the different reactivity of phosphonyl
and thiophosphonyl radicals towards double bonds,
which can be explained by radical polarity. Further
evidence to support the fact that diethyl thiophosphite is
a more effective hydrogen-atom donor than diethyl
phosphite is also provided through this work. This is
consistent with other phosphorus hydride radical reac-
tions we have recently investigated, including addition
to phenylacetylene and reduction of 2-bromoacetophe-
none (Scheme 5). In both cases, reduced products were
isolated in higher yields when using diethyl thio-
phosphite.
Scheme 4. Synthesis of trisubstituted tetrahydropyran 21 (the relative
stereochemistry of the major isomer is shown). Reagents and condi-
tions: (a) TBDPSCl, imidazole, CH2Cl2, rt (90%); (b) CuI,
H2C@CHCH2MgBr, THF, 0 ꢁC to rt (77%); (c) EtOCH@CH2,
Hg(O2CCF3)2, rt (75%); (d) (EtO)2P(O)H, Et3B/O2, cyclohexane, rt
(57%); (e) sBuLi, then ClCO2Bn, )78 ꢁC to rt, THF (78%); (f) O3,
CH2Cl2, )78 ꢁC then PPh3; (g) Ph3P@CHCO2Me, CH2Cl2, rt (65%
over two steps); (h) Pd/C, 1,4–cyclohexadiene, MeOH, rt (90%);
t
(i) KOH, MeOH, rt then BuOK, THF, rt (60%).
The synthesis of 21 started by O-silylation of glycidol 22
followed by regioselective ring opening of the epoxide12
and vinyl ether formation (Scheme 4). Reaction of 23
with diethyl phosphite, using Et3B as initiator, resulted
in the formation of tetrahydropyran 24 in 57% yield as a
4:1mixture of inseparable diastereoisomers. A similar
yield of 24 (50%) was obtained when the cyclisation was
carried out at 80 ꢁC using AIBN as the initiator, al-
though the diastereoselectivity was reduced to 2:1. De-
protonation of phosphonate 24, at low temperature,
followed by addition of benzyl chloroformate and
warming to rt, resulted in an efficient b-elimination
reaction13 and the formation of carbonate 25. Only the
E-alkene isomer was isolated (after chromatography) as
Acknowledgements
1
indicated by the H NMR spectrum. Deprotonation of
24 in the absence of benzyl chloroformate resulted in the
quantitative formation of the corresponding vinyl
phosphonate bearing a secondary alcohol.14 Oxidative
cleavage of vinyl phosphonate 25 was then followed by a
ꢀ ꢁ
We thank Uniqema and the BBSRC for funding, Helene
Lepaumier for the preparation of phosphonates 8–9 and
Michael Puljic for the reduction of 2-bromoacetophen-
one.

5098
C. M. Jessop et al. / Tetrahedron Letters 45 (2004) 5095–5098
J 18.6, 7.6, PCH2), 1.55 (2H, quintet, J 6.4, OCH2CH2),
1.40–1.30 (2H, m, CH2CH3), 1.30 (6H, t, 7.0,
2 · POCH2CH3), 0.91(3H, t, 7.3, CH2CH3); dC
(67.9 MHz, CDCl3) 70.6 (PCH2CH2O), 64.4 (OCH2),
61.5 (d, JCP 6, POCH2), 31.6 (OCH2CH2), 28.9 (d, JCP
138, PCH2), 19.1 (CH2CH3), 16.3 (d, JCP 7,
2 · POCH2CH3), 13.7 (CH3); m=z (CI, NH3) 239
(M+Hþ, 100%); Found 239.1405. C10H24O4P requires
for M+Hþ, 239.1412.
References and notes
J
J
ꢁ
1. See, for example: (a) Deprele, S.; Montchamp, J.-L. J.
Org. Chem. 2001, 66, 6745–6755; (b) Herpin, T. F.;
Motherwell, W. B.; Roberts, B. P.; Roland, S.; Weibel,
J.-M. Tetrahedron 1997, 53, 15085–15100; (c) Lopin, C.;
Gautier, A.; Gouhier, G.; Piettre, S. R. Tetrahedron Lett.
2000, 41, 10195–10200; (d) Dubert, O.; Gautier, A.;
Condamine, E.; Piettre, S. R. Org. Lett. 2002, 4, 359–
362; (e) Piettre, S. R. Tetrahedron Lett. 1996, 37, 2233–
2236; (f) Piettre, S. R. Tetrahedron Lett. 1996, 37, 4707–
4710; (g) Rey, P.; Taillades, J.; Rossi, J. C.; Gros, G.
Tetrahedron Lett. 2003, 44, 6169–6171.
7. (a) Gurjar, M. K.; Krishna, L. M.; Reddy, B. S.;
Chorghade, M. S. Synthesis 2000, 557–560; (b) Ghosh,
S.; Raychaudhuri, S. R.; Salomon, R. G. J. Org. Chem.
1987, 52, 83–90; (c) Nakajima, R.; Urabe, H.; Sato, F.
Chem. Lett. 2002, 4–5; (d) McMurry, J. E.; Kocovsky, P.
Tetrahedron Lett. 1985, 26, 2171–2172.
2. Jessop, C. M.; Parsons, A. F.; Routledge, A.; Irvine, D.
Tetrahedron Lett. 2003, 44, 479–483.
8. For related alkoxymethyl radical cyclisations, see: (a)
Rawal, V. H.; Singh, S. P.; Dufour, C.; Michoud, C. J.
Org. Chem. 1993, 58, 7718–7727; (b) Lolkema, L. D. M.;
Hiemstra, H.; Al Ghouch, A. A.; Speckamp, W. N.
Tetrahedron Lett. 1991, 32, 1491–1494.
9. For the major diastereoisomer of 15, a correlation between
H6 and the C-5 methyl group was observed in the NOESY
spectrum. Correlations between H2 and H6 were observed
for both isomers of 15, which suggest that the minor
diastereoisomers of 15 and 17 have the all cis-stereochem-
istry.
3. (a) Gautier, A.; Garipova, G.; Dubert, O.; Oulyadi, H.;
Piettre, S. R. Tetrahedron Lett. 2001, 42, 5673–5676; (b)
Trofimov, B. A.; Khil’ko, M. Y.; Nedolya, N. A.;
Vyalykh, E. P. J. Org. Chem. USSR (Engl. Transl.)
1981, 17, 1038–1040; (c) Brel’, A. K.; Tyurenkov, I. N.;
Strel’tsova, G. V.; Matyukhova, N. P.; Nazarenko, V. N.;
Agarkova, N. S. Pharm. Chem. J. (Engl. Transl.) 1988, 22,
118–120; (d) Nishiwaki, T. Tetrahedron 1965, 21, 3043–
3049.
4. See, for example: (a) Diana, G. D.; Zalay, E. S.; Salvador,
U. J.; Pancic, F.; Steinberg, B. J. Med. Chem. 1984, 27,
691–694; (b) Yoshino, K.; Kohno, T.; Morita, T.;
Tsukamoto, G. J. Med. Chem. 1989, 32, 1528–1532; (c)
Kelley, J. L.; Linn, J. A.; McLean, E. W.; Tuttle, J. V. J.
Med. Chem. 1993, 36, 3455–3463; (d) Hakimelahi, G. H.;
Moosavi-Movahedi, A. A.; Sadeghi, M. M.; Tsay, S.-C.;
Hwu, J. R. J. Med. Chem. 1995, 38, 4648–4659; (e) Liu,
H.; Li, W.; Kim, C. U. Bioorg. Med. Chem. Lett. 1997, 7,
1419–1420; (f) Cermak, D. M.; Wiemer, D. F.; Lewis, K.;
Hohl, R. J. Bioorg. Med. Chem. 2000, 8, 2729–2737; (g)
Williams, D. M.; Jakeman, D. L.; Vyle, J. S.; Williamson,
M. P.; Blackburn, G. M. Bioorg. Med. Chem. Lett. 1998,
8, 2603–2608.
ꢀ
10. Keller-Juslen, C.; King, H. D.; Kuhn, M.; Loosli, H.-R.;
Pache, W.; Petcher, T. J.; Weber, H. P.; Wartburg, A. J.
Antibiot. 1982, 35, 142–150.
11. Hori, K.; Hikage, N.; Inagaki, A.; Mori, S.; Nomura, K.;
Yoshii, E. J. Org. Chem. 1992, 57, 2888–2902.
12. Dixon, D. J.; Ley, S. V.; Reynolds, D. J. Chem. Eur. J.
2002, 8, 1621–1636.
13. Sulfides, sulfoxides and sulfones are known to undergo
similar eliminations: (a) Maezaki, N.; Izumi, M.; Yuyama,
S.; Sawamoto, H.; Iwata, C.; Tanaka, T. Tetrahedron
2000, 56, 7927–7945; (b) Barnes, N. J.; Davidson, A. H.;
Hughes, L. R.; Procter, G. J. Chem. Soc., Chem. Commun.
ꢀ
1985, 1292–1294; (c) Caturla, F.; Najera, C. Tetrahedron
5. All new compounds gave consistent spectral and high
resolution mass spectroscopic data.
1997, 53, 11449–11464; (d) Krief, A.; Kenda, B.; Remacle,
B. Tetrahedron 1996, 52, 7435–7463.
6. Typical experimental procedure: To a stirred solution of
diethyl phosphite (1.38 g, 10.0 mmol) in cyclohexane
14. Subsequent oxidative cleavage (OsO4/NaIO4 or O3) and
Wittig reaction of alcohol (25, R@H) was less efficient
than for benzyl carbonate (25, R@CO2Bn).
15. Shue, Y.-K.; Carrera, G. M.; Tufano, M. D.; Nadzan, A.
M. J. Org. Chem. 1991, 56, 2107–2111.
16. Reaction of (26, R@H) with 0.2 equiv of tBuOK at )65 ꢁC
afforded the C-2 epimer of tetrahydrofuran 21.
17. Tetronasin (M139603) has a similar structure to 19 but has
opposite configurations at the 10 common chiral centres.
(a) Hori, K.; Kazuno, H.; Nomura, K.; Yoshii, E.
Tetrahedron Lett. 1993, 34, 2183–2186; (b) Ley, S. V.;
Brown, D. S.; Clase, J. A.; Fairbanks, A. J.; Lennon, I. C.;
(15 cm3) was added n-butyl vinyl ether
1 (100 mg,
1.0 mmol) and a 1 M solution of triethylborane in hexanes
(0.3 cm3, 0.3 mmol) at ambient temperature. After 6 h,
further triethylborane (0.3 cm3, 0.3 mmol, 1M solution in
hexanes) was added and the mixture stirred for 48 h while
adding triethylborane (3 · 0.3 cm3, 3 · 0.3 mmol, 1M solu-
tion in hexanes) over this period. The mixture was then
concentrated in vacuo. Excess diethyl phosphite was
removed by kugelrohr distillation (75 ꢁC, 2 mmHg). Puri-
fication by column chromatography on silica (ethyl
acetate) afforded O,O-diethyl 2-butoxyethyl-phosphonate
2 (235 mg, 99%) as a colourless oil. Rf 0.4 (ethyl acetate);
mmax (CH2Cl2) 2983 (s), 2962 (s), 2873 (s), 1245 (s, P@O),
1099 (s, P–O), 962 (s) cmꢁ1; dH (400 MHz, CDCl3) 4.18–
4.05 (4H, m, 2 · POCH2), 3.66 (2H, dt, J 11.6, 7.3,
PCH2CH2O), 3.43 (2H, t, J 6.4, OCH2), 2.11 (2H, dt,
ꢀ
Osborn, H. M. I.; Stokes (nee Owen), E. S. E.; Wads-
worth, D. J. J. Chem. Soc., Perkin Trans. 1 1998, 2259–
2276; (c) Martinek, T.; Riddell, F. G.; Rutherford, T. J.;
Sareth, S.; Weller, C. T. Chem. Commun. 1998, 1893–
1894.