The synthesis of carbohydrate-derived acylsilanes and their intramolecular free radical cyclizations with the formation of polyoxygenated cyclopentanes

Article abstract of DOI:10.1016/j.tetlet.2009.04.032

A convenient way for the synthesis of acylsilanes from arabinose, lyxose, and ribose is developed. All the chiral centers of the carbohydrate templates are conserved, and only the reducing end is transformed into the acylsilane functional group. The non-r

Full text of DOI:10.1016/j.tetlet.2009.04.032

Tetrahedron Letters 50 (2009) 3805–3808
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
The synthesis of carbohydrate-derived acylsilanes and their intramolecular
free radical cyclizations with the formation of polyoxygenated cyclopentanes
*
Che-Chien Chang, Yu-Hsien Kuo, Yeun-Min Tsai
Department of Chemistry, National Taiwan University, Taipei, Taiwan 106, ROC
a r t i c l e i n f o
a b s t r a c t
Article history:
A convenient way for the synthesis of acylsilanes from arabinose, lyxose, and ribose is developed. All the
chiral centers of the carbohydrate templates are conserved, and only the reducing end is transformed into
the acylsilane functional group. The non-reducing end of the templates can be converted into a bromide.
These bromo acylsilanes undergo efficient intramolecular radical cyclizations to give polyoxygenated
cyclopentanes.
Received 6 January 2009
Revised 14 March 2009
Accepted 6 April 2009
Available online 14 April 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Acylsilane
Radical cyclization
Polyoxygenated cyclopentanes
Acylsilane is a useful functionality which has attracted the
attentions of synthetic chemists.1 The presence of silyl group on
the carbonyl not only enhances the reactivity of a carbonyl but also
becomes a useful handle for additional transformations.
OSiR3
SiR3
O
O
radical-Brook
rearrangement
R3Si
n
n
n
1
2
Several years ago we have initiated a project to study the intra-
molecular radical cyclizations involving acylsilanes.2 As shown in
Scheme 1, intramolecular radical cyclization of acylsilane 1 gives
a cyclized alkoxy radical intermediate 2 with a silyl group attached
at the b-position of the radical. A facile radical-Brook rearrange-
ment3,4 successfully drives the reaction toward the formation of
a silyloxy-substituted radical 3. In contrast, similar radical cycliza-
tions involving formyl group are known to be reversible processes
in favor of the acyclic radicals.5 Therefore, the special feature of the
radical cyclizations of acylsilanes made this type of reaction a use-
ful method in the preparation of cyclic alcohols.6
3
n = 1, 2
Scheme 1.
Carbohydrates are unique sources of chirality in nature and
have been extensively used as building blocks for the synthesis
of enantiomerically pure and highly oxygenated derivatives.7 In
particular, the conversion of carbohydrates to carbocycles is an
area that has been extensively studied.8 Fraiser-Reid and co-work-
ers demonstrated several successful carbohydrate-based free radi-
cal cyclizations in which aldehyde groups served as the radical
Ph
O
Ph
O
acceptor (Eq. 1).9 In these cyclizations, the
x-formylalkyl radicals
O
O
O
Bu3SnH, AIBN
O
O
are immobilized by fusing with a pyranoside ring. However, alde-
hydes that carry carbohydrate skeleton, such as bromide 4, are not
good substrates for radical cyclizations.9f Thus there is a need to
examine whether acylsilane 5 can be suitable substrate for this
purpose. Herein we wish to report our synthesis of some pen-
tose-derived acylsilanes and the study of their intramolecular rad-
ical cyclizations.
HO
75−85%
R
OMe
OMe
H
R
H
R = vinyl or OMe
I
ð1Þ
OMe OBn
O
H
Br
Br
SiR3
n
Preparation of carbohydrate-derived acylsilanes has been
reported by Plantier-Royon and Portella10 using the 2-silylated-
1,3-dithiane approach developed by Brook11 and Corey.12 In this
approach, all the chiral centers of the carbohydrate substrates are
retained and the carbon skeleton is extended by one. However, the
OBn OMe O
RO
5
4
n = 3 or 4
a
-position of the resulting acylsilanes cannot have a hydroxyl or
* Corresponding author. Tel.: +886 2 33661651; fax: +886 2 23636359.
E-mail address: ymtsai@ntu.edu.tw (Y.-M. Tsai).
alkoxy substituent. Although several reports about the construction
0040-4039/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2009.04.032

3806
C.-C. Chang et al. / Tetrahedron Letters 50 (2009) 3805–3808
of
a
-oxy acylsilanes were known, none of these methods employed
sponding a-silyloxy cyclopentyl radical preferred to abstract a
carbohydrate starting material directly.13 We decided to explore
the possibility of adding a silyl anion to the carbonyl end of a car-
hydrogen from the face opposite to the C(2) methoxy group.
As mentioned above, we attributed the lack of reactivity of alco-
hol 6a toward silyl anion to the steric environment around the car-
bonyl group. Trying to understand the origin of this steric effect,
we analyzed the conformation of aldehyde 6a. As shown in Scheme
3, among the three staggered conformations of 6a, the carbonyl
group in conformers AI and AII is gauche to the large group RL. Con-
former AIII has an all gauche relationship and presumably also
made the carbonyl quite sterically encumbered. Based on this
analysis, we felt that the lyxose-derived aldehyde 1115 having a
C(2)–C(3) anti stereo-relationship would contain a low energy con-
former LI.18 In which, the two benzyloxy dipoles are anti to each
other, and the carbonyl group is anti to the large group RL but
gauche to the smaller benzyloxy group. We therefore hypothesized
that the carbonyl group in aldehyde 11 might be less sterically hin-
dered in comparison with aldehyde 6a.
Indeed, aldehyde 11 (Scheme 4) successfully coupled with the
silyl cuprate and afforded diol 12 in 79% yield after removal of
the trityl group. Diol 12 was converted to mesylate 13 followed
by Swern oxidation17 to give acylsilane 14 in 51% over two steps.
Replacing the mesylate group in 14 with bromide using lithium
bromide in DMF met with failure. However, this task was accom-
plished by the reaction of acylsilane 14 with tetrabutylammonium
bromide in refluxing benzene19 and yielded bromo acylsilane 15
(77%).
bohydrate to generate an
alcohol can then be oxidized to give an acylsilane with the desired
carbohydrate skeleton.
a a-silyl
-silyl alcohol.14 The resulting
As shown in Scheme 2, aldehyde 6a15 derived from arabinose
with all the secondary hydroxyl groups protected by benzyl groups
did not react with methyldiphenylsilyl lithium or the correspond-
ing silyl cuprate. With the hypothesis that this aldehyde is quite
sterically hindered due to the presence of adjacent bulky substitu-
ents, we switched to the methylated analog 6b.16 Indeed, aldehyde
6b reacted with methyldiphenylsilyl lithium and gave
hol 7 in 46% yield. By using the less basic lithium bis(methyldi-
phenylsilyl)cuprate, the yield of
to 62%.
a-silyl alco-
a-silyl alcohol 7 was improved
The trityl group in alcohol 7 was removed in a mixture of THF
and methanol in the presence of p-toluenesulfonic acid. The pri-
mary hydroxyl group of the resulting diol 8a was converted to a
methanesulfonate 8b. This material was then oxidized using the
Swern method17 to give acylsilane 8c in 45% yield over these three
steps. Treatment of 8c with lithium bromide in DMF afforded bro-
moacylsilane 9 in a 91% yield.
Free radical cyclization of acylsilane 9 using the standard tribu-
tyltin hydride method in refluxing benzene2 gave successfully the
cyclized silyl ether 10 as a mixture of two epimers (10a/10b = 2.4/
1) in a combined yield of 77%. For the sake of separation, this epi-
meric mixture was desilylated in a mixture of THF and methanol
with catalytic amount of p-toluenesulfonic acid. The resulting alco-
hols were converted to the p-bromobenzoates and were separated
to afford benzoates 10c (62%) and 10d (37%). Difference NOE
experiments showed that irradiation of H(1) at d 5.14 (CDCl3) in
the major isomer 10c resulted in a 4% enhancement of H(2) at d
OBn
OBn
OBn
BnO OBn
OBn H
BnO
H
H
H
H
RL
RL
OBn
TrO
O
CHO
CHO
H
CHO
RL
RL
OBn
H
Arabinose-type
4.03. In contrast, irradiation of H(1) in benzoate 10d at
5.43ꢀ5.48 did not show any enhancement of H(2) at
d
d
AI
AII
AIII
6a
OBn
OBn
OBn
3.83ꢀ3.89. We therefore assigned the structure of 10c as having
a 1,2-cis relationship of the two substituents. This stereochemical
outcome indicated that in the radical cyclization step the corre-
BnO OBn
H
H
OBn
CHO
BnO
H
RL
RL
H
H
TrO
O
CHO
CHO
OBn H
Lyxose-type
11
RL
H
OBn
LIII
LI
LII
OMe OMe
OR OR
(a) MePh2SiLi/THF/−78 o
C
Scheme 3.
TrO
Σ
TrO
H
or
(b) (MePh2Si)2CuLi/THF/−78 o
C
MeO
OH
OR
O
7
46% via (a)
62% via (b)
6a R = Bn
6b R = Me
1) (MePh2Si)2CuLi
1) MsCl, Et3N
CH2Cl2 (74%)
OBn OBn
Σ = SiMePh2
THF, −78 oC
HO
Σ
11
Tr = trityl
2) TsOH (cat. amt.)
THF/MeOH
2) Swern oxid.
(70%)
1) TsOH, THF/MeOH (77%)
BnO
OH
2) MsCl, Et3N, CH2Cl2, 0 oC (92%)
12 (79%)
3) Swern oxidation (64%)
Σ = SiMePh2
OMe OMe
OMe OMe
LiBr, DMF, 90 oC
91%
OBn OBn
OBn OBn
X
Σ
Br
Σ
Bu4NBr
Br
Σ
MsO
Σ
Z
Y
MeO
MeO
O
PhH
BnO
O
Y
X
80 o
C
BnO
8a X = OH, Y = H, Z = OH
9
15
X = OH, Y = H
13
14 X = Y = O
8b
8c
X = OMs, Y = H, Z = OH
X = OMs, Y = Z = O
(77%)
77%
Bu3SnH, AIBN (cat. amt.)
PhH, 80 oC, 2 h
10a/10b
= 2.4/1
X
Bu3SnH, AIBN (cat. amt.), PhH, 80 o
C
BnO
Y
X
1) TsOH (cat. amt.)
MeOH/THF
(76%)
10c
10d
X = OCOPhBr, Y = H
(62%)
MeO
OBn
X = OSiMePh2, Y = H
BnO
4
Y
2
16a
2) 4-BrPhCOCl
DMAP (cat. amt.)
Et3N, CH2Cl2
3
TsOH
OMe
MeO
16b X = H, Y = OSiMePh2
17a X = OH, Y = H
17b X = H, Y = OH
X = H, Y = OCOPhBr
(37%)
(cat. amt.)
MeOH/THF
X = OSiMePh2, Y = H
10a
10b X = H, Y = OSiMePh2
(77%)
Scheme 2.
Scheme 4.

C.-C. Chang et al. / Tetrahedron Letters 50 (2009) 3805–3808
3807
Therefore, the incorporation of 1H at C(1) in 19c-D is most likely
occurred through an intramolecular 1,5-hydrogen transfer process
from the benzylic position of the C(3)-benzyloxy group. We sug-
gest that the C(3)-substituent is located at a pseudo-axial position
(A) and is more accessible to deliver the benzylic hydrogen to the
radical at C(1). Whereas the C(4)-benzyloxy group adopts a pseu-
do-equatorial position, it is difficult to reach the corresponding
benzylic hydrogen for the radical.
In conclusion, we have developed a convenient way for the syn-
thesis of acylsilanes from carbohydrate templates. All the chiral
centers of the templates were conserved, and only the reducing
end was transformed into acylsilane. In the meantime, the non-
reducing end of the templates can be converted into a bromide.
We demonstrated that these bromo acylsilanes can undergo effi-
cient intramolecular radical cyclizations to give polyoxygenated
carbocycles with different stereochemical structures. One can ex-
pect the silyloxy group to serve as a handle for the conversion of
these cyclopentitols to potentially useful compounds such as
aminopentitols22 and carbocyclic nucleotides.23,24
TsOH
(cat. amt.)
X
1
4
OBn OBn
BnO
Bu3SnH
Y
Br
Σ
MeOH/THF
(71%)
AIBN
(cat. amt.)
OBn
BnO
BnO
18
Σ = SiMePh2
O
PhH, 80 oC 19a
X = OΣ, Y = H
X = H, Y = OΣ
19b
(76%)
19c X = OH, Y = H
19d X = H, Y = OH
TsOH
(cat. amt.)
Bu3SnD
19c-D (with D at C(3)-benzylic position)
+ 19e (X = OH, Y = D)
+ 19f (X = D, Y = OH)
AIBN (cat. amt.)
MeOH/THF
PhH, 80 o
C
BnO
BnO
Bn
OΣ
1,5-H
19c-D
OΣ
OBn
O
transfer
O
H
O
H
Ph
Ph
A
Scheme 5.
Acknowledgments
The tin-mediated cyclization of 15 was carried out successfully
to give the cyclization products 16 in a 76% yield with a stereoiso-
meric ratio of 2.8/1 (16a/16b). The inseparable mixture of 16 was
desilylated to afford partially separable alcohols 17 (77%). The ste-
reochemistry of 17 was determined by difference NOE experi-
ments. Specifically, irradiation of H(2) in cyclopentanol 17a at d
3.94ꢀ3.99 (CDCl3) resulted in a 16% enhancement of H(1) at d
4.26ꢀ4.37. On the contrary, irradiation of H(2) in 17b at d 3.75 only
resulted in a 4% enhancement of H(1) at d 4.16ꢀ4.23. These results
indicated that the major isomer 17a exhibited a C(1)–C(2) cis-rela-
tionship. Similar selectivity was also observed for the cyclization of
acylsilane 9 as mentioned above.
Financial support of the National Science Council of the Repub-
lic of China is gratefully acknowledged. We are also grateful to the
National Center for High-performance Computing for the computer
time and facilities.
Supplementary data
Supplementary data (compound characterization data of 9, 10c,
d, 15, 17a, b, 18, 19c, and d) associated with this article can be
found, in the online version, at doi:10.1016/j.tetlet.2009.04.032.
Ribose is another pentose that exhibits a C(2)–C(3) anti stereo-
relationship. Using similar methods we synthesized the D-ribose-
References and notes
derived acylsilane 1820 (Scheme 5). Radical cyclization of 18 with
tributyltin hydride gave a 76% yield of an inseparable mixture of
cyclized products 19a and 19b (19a/19b = 1.6/1). Removal of the
silyl group then afforded alcohols 19c (43%) and 19d (28%). Differ-
ence NOE experiments of 19d showed that irradiation of H(1) at d
4.04–4.11 (CDCl3) resulted in a 6% enhancement of H(2) at d 3.43.
Conversely, irradiation of H(2) resulted in a 14% enhancement of
H(1). We therefore assigned 19d as having a C(1)–C(2) cis-
relationship.
1. For recent reviews about acylsilanes, see: (a) Ricci, A.; Degl’Innocenti, A.
Synthesis 1989, 647–660; (b) Page, P. C. B.; Klair, S. S.; Rosenthal, S. Chem. Soc.
Rev. 1990, 19, 147–195; (c) Cirillo, P. F.; Panek, J. S. Org. Prep. Proced. Int. 1992,
24, 553–582; (d) Page, P. C. B.; McKenzie, M. J.; Klair, S. S.; Rosenthal, S. In The
Chemistry of Organic Silicon Compounds; Rappoport, Z., Apeloig, Y., Eds.; John
Wiley & Sons: New York, 1998; Vol. 2, pp 1599–1665. Chapter 27; (e) Bonini, B.
F.; Comes-Franchini, M.; Fochi, M.; Mazzanti, G.; Ricci, A. J. Organomet. Chem.
1998, 567, 181–189.
2. Huang, C.-H.; Chang, S.-Y.; Wang, N.-S.; Tsai, Y.-M. J. Org. Chem. 2001, 66, 8983–
8991. and references cited therein.
3. (a) Dalton, J. C.; Bourque, R. A. J. Am. Chem. Soc. 1981, 103, 699–700; (b) Harris, J.
M.; MacInnes, I.; Walton, J. C.; Maillard, B. J. Organomet. Chem. 1991, 403, C25–
C28; (c) Tsai, Y.-M.; Ke, B.-W. J. Chin. Chem. Soc. (Taipei) 1993, 40, 641–642; (d)
Robertson, J.; Burrows, J. N. Tetrahedron Lett. 1994, 35, 3777–3780; (e)
Schiesser, C. H.; Styles, M. L. J. Chem. Soc., Perkin Trans. 2 1997, 2335–2340;
(f) Paredes, M. D.; Alonso, R. Tetrahedron Lett. 1999, 40, 3973–3976; (g) Paredes,
M. D.; Alonso, R. J. Org. Chem. 2000, 65, 2292–2304.
Surprisingly, the preference of stereochemical outcome in this
system showed that the corresponding a-silyloxy cyclopentyl rad-
ical abstracted a hydrogen atom from the same face relative to the
benzyloxy group at C(2) on the cyclopentane ring. We suspected
that intramolecular hydrogen abstraction might be the cause of
this preference. To confirm this suspicion, the cyclization of acylsi-
lane 18 was carried out using tributyltin deuteride followed by
desilylation, and we were able to isolate alcohols 19f (24%) and a
mixture (31%) of 19c-D and 19e (19c-D/19e = 45/55).
The complete deuteration at C(1) in the all-cis isomer 19f is
obvious in the 1H NMR spectrum for the disappearance of the pro-
ton signal at C(1) (d 4.10 in CD3OD). The 13C signal of C(1) also
diminished into a very small triplet of equal intensity centered at
d 69.5 (in CDCl3; JC–D = 23 Hz). In contrast, in the 1H NMR spectrum
of the trans–cis–cis isomer mixture 19c-D and 19e, the proton sig-
nal at C(1) (d 4.36 in CD3OD) only partially disappeared. The 13C
signal of C(1) of 19e appeared as a small triplet centered at d
74.1 (in CDCl3; JC–D = 23 Hz) coexisted with a signal at d 74.4
belonging to C(1) of 19c-D. The 13C signal of the benzylic carbon
(d 73.0 in CDCl3 on the C(3)-side chain also appeared as a singlet
overlapped with a triplet indicating the incorporation of a deute-
rium atom at this carbon.21
4. For a recent review about Brook rearrangement, see: Moser, W. H. Tetrahedron
2001, 57, 2065–2084.
5. (a) Beckwith, A. L. J.; Hay, B. P. J. Am. Chem. Soc. 1989, 111, 230–234; (b)
Beckwith, A. L. J.; Hay, B. P. J. Am. Chem. Soc. 1989, 111, 2674–2681; (c)
Beckwith, A. L. J.; Raner, K. D. J. Org. Chem. 1992, 57, 4954–4962.
6. Chen, M.-J.; Tsai, Y.-M. Tetrahedron Lett. 2007, 48, 6271–6274.
7. (a) Hanessian, S. Total Synthesis of natural products: The Chiron Approach;
Pergamon: New York, 1983; (b) Bols, M. Carbohydrate Building Blocks; Wiley:
New York, 1996.
8. (a) Madsen, R. Eur. J. Org. Chem. 2007, 399–415; (b) Zhou, J.; Wang, G.; Zhang,
L.-H.; Ye, X.-S. Curr. Org. Chem. 2006, 10, 625–642; (c) Ferrier, R. J.; Hoberg, J.
O. Adv. Carbohydr. Chem. Biochem. 2003, 58, 55–119; (d) Ferrier, R. J. Top. Curr.
Chem. 2001, 215, 277–291; (e) Dalko, P. I.; Sinaÿ, P. Angew. Chem., Int. Ed.
1999, 38, 773–777; (f) Marco-Contelles, J.; Alhambra, C.; Martínez-Grau, A.
Synlett 1998, 693–699; (g) Martínez-Grau, A.; Marco-Contelles, J. Chem. Soc.
Rev. 1998, 27, 155–162; (h) Ferrier, R. J.; Middleton, S. Chem. Rev. 1993, 93,
2779–2831.
9. (a) Tsang, R.; Fraser-Reid, B. J. Am. Chem. Soc. 1986, 108, 2116–2117; (b)
Tsang, R.; Fraser-Reid, B. J. Am. Chem. Soc. 1986, 108, 8102–8104; (c)
Tsang, R.; Dickson, J. K., Jr.; Pak, H.; Walton, R.; Fraser-Reid, B. J. Am.
Chem. Soc. 1987, 109, 3484–3486; (d) Fraser-Reid, B.; Vite, G. D.; Yeung,
B.-W. A.; Tsang, R. Tetrahedron Lett. 1988, 29, 1645–1648; (e) Dickson, J.
K., Jr.; Tsang, R.; Llera, J. M.; Fraser-Reid, B. J. Org. Chem. 1989, 54, 5350–

3808
C.-C. Chang et al. / Tetrahedron Letters 50 (2009) 3805–3808
5356; (f) Yeung, B. A.; Alonso, R.; Vite, G. D.; Fraser-Reid, B. J. Carbohydr.
Chem. 1989, 8, 413–427; (g) Walton, R.; Fraser-Reid, B. J. Am. Chem. Soc.
1991, 113, 5791–5799.
18. For the conformational analysis of 2,3-dihalobutanes and related compounds,
see: Eliel, E. L.; Wilen, S. H. Stereochemistry of Organic Compounds; John Wiley &
Sons: New York, 1994.
10. (a) Plantier-Royon, R.; Portella, C. Tetrahedron Lett. 1996, 37, 6113ꢀ6114; (b)
Plantier-Royon, R.; Portella, C. Synlett 1994, 527–529.
11. Brook, A. G.; Duff, J. M.; Jones, P. F.; Davis, N. R. J. Am. Chem. Soc. 1967, 89, 431–
434.
12. Corey, E. J.; Seebach, D.; Freedman, R. J. Am. Chem. Soc. 1967, 89, 434–436.
13. (a) Raobo, P.; Wicha, J. J. Org. Chem. 1994, 59, 4355–4356; (b) Cirilla, P. F.;
Panek, J. S. J. Org. Chem. 1994, 59, 3055–3063; (c) Cirillo, P. F.; Panek, J. S. J. Org.
Chem. 1990, 55, 6071–6073.
19. Brooks, G.; Edwards, P. D.; Hatto, J. D. I.; Smale, T. C.; Sonthgate, R. Tetrahedron
1995, 51, 7999–8014.
20. This acylsilane was synthesized from the corresponding aldehyde: Garegg, P. J.;
Samuelsson, B. Carbohydr. Res. 1978, 67, 267–270.
21. The 13C signals of the benzylic carbons were identified through the techniques
of HMQC and HMBC.
22. (a) Delgado, A. Eur. J. Org. Chem. 2008, 3893–3906; (b) Trost, B. M. J. Org. Chem.
2004, 69, 5813–5837.
14. Reich, H. J.; Eisenhart, E. K.; Olson, R. E.; Kelly, M. J. J. Am. Chem. Soc. 1986, 108,
7791–7800.
23. (a) Ferrero, M.; Gotor, V. Chem. Rev. 2000, 100, 4319–4347; (b) Crimmins, M. T.
Tetrahedron 1998, 54, 9229–9272.
15. Dubost, E.; Le Nouën, D.; Streith, J.; Tarnus, C.; Tschamber, T. Eur. J. Org. Chem.
2006, 610–626.
16. Miljkovic´, M.; Dropkin, D.; Hagel, P.; Habash-Marino, M. Carbohydr. Res. 1984,
24. (a) Ando, T.; Iwata, M.; Zulfiqar, F.; Miyamoto, T.; Nakanishi, M.; Kitade, Y.
Bioorg. Med. Chem. 2008, 3809–3815; (b) Ando, T.; Kojima, K.; Chahota, P.;
Kozaki, A.; Milind, N. D.; Kitade, Y. Bioorg. Med. Chem. 2008, 2615–2618; (c)
Siddiqi, S. M.; Jacobson, K. A.; Esker, J. L.; Olah, M. E.; Ji, X.; Melman, N.; Tiwari,
K. N.; Secrist, J. A., III; Schneller, S. W.; Cristalli, G.; Stiles, G. L.; Johnson, C. R.;
IJzerman, A. P. J. Med. Chem. 1995, 38, 1174–1188.
128, 11–20.
17. (a) Mancuso, A. J.; Huang, S. L.; Swern, D. J. Org. Chem. 1978, 43, 2480–2482; (b)
Mancuso, A. J.; Swern, D. Synthesis 1981, 165–185.