An environmentally benign synthesis of cis-2,6-disubstituted tetrahydropyrans via indium-mediated tandem allylation/prins cyclization reaction

Article abstract of DOI:10.1021/jo7016857

(Chemical Equation Presented) In the presence of indium metal, 3-iodo-2-[(trimethylsilyl)-methyl]propene (1) reacts with sequentially added aldehydes to provide cis-2,6-disubstituted tetrahydropyrans in good yields. Evidence suggests that InI, formed upon

Full text of DOI:10.1021/jo7016857

SCHEME 1. An Indium-Mediated [m + n] Annulation
Reaction
An Environmentally Benign Synthesis of
cis-2,6-Disubstituted Tetrahydropyrans via
Indium-Mediated Tandem Allylation/Prins
Cyclization Reaction
Minh Pham, Amir Allatabakhsh, and Thomas G. Minehan*
Department of Chemistry and Biochemistry, California State
UniVersitysNorthridge, Northridge, California 91330
SCHEME 2. Indium-Mediated Reactions of 1 with
Aldehydes in Aqueous Media
thomas.minehan@csun.edu
ReceiVed August 2, 2007
tetrahydropyrans in aqueous media employing indium metal as
the sole promoter.
We have previously reported that combination of equimolar
amounts of 3-iodo-2-[(trimethylsilyl)methyl]propene (1),4 di-
carbonyl compounds (4), and indium metal gives rise to seven-
and eight-membered oxa-bridged carbocycles (5) in moderate
to high yields (Scheme 1).5 This [m + n] annulation process
takes place in aqueous media at room temperature under an
atmosphere of air and requires no externally added Lewis acid
promoter. We have proposed that indium-mediated intermo-
lecular allylation of one carbonyl of the substrate is followed
by an intramolecular silyl-Prins cyclization reaction promoted
by the indium halide salts (or Brønsted acids derived therefrom)
formed in the allylation step.
In the presence of indium metal, 3-iodo-2-[(trimethylsilyl)-
methyl]propene (1) reacts with sequentially added aldehydes
to provide cis-2,6-disubstituted tetrahydropyrans in good
yields. Evidence suggests that InI, formed upon aldehyde
(R1CHO) allylation in aqueous media, acts as a promoter
for the silyl-Prins reaction with the second equivalent of
added aldehyde (R2CHO). The preparation of cyclohexenyl-
fused pyrans via this one-pot, three-component coupling
process is presented, as is a short formal synthesis of (()-
centrolobine.
To verify that two monocarbonyl compounds could also
participate in this process, we added an excess (3.0 equiv) of
benzaldehyde to 1 and indium metal in 1:1 H2O/i-PrOH and
stirred the reaction mixture at ambient temperature for 36 h
(Scheme 2). 2,6-Diphenyltetrahydropyran 2a was obtained in
60% yield; repetition of this reaction with 3.0 equiv of hexanal
instead gave rise to 2,6-dipentyltetrahydropyran 2b in 73% yield.
In both cases, a single stereoisomer was predominant (>13:1
stereoselectivity by GC-MS analysis of the crude reaction
The widespread occurrence of substituted tetrahydropyran
moieties in natural products has inspired the development of
numerous creative synthetic approaches to this important
structural subunit.1 Although these methods are highly efficient,
the typical requirements for strictly anhydrous conditions, strong
Lewis acid promoters, low temperatures, and/or halogenated
solvents, may be seen as limitations.2,3 Herein we report an
environmentally benign one-pot preparation of 2,6-disubstituted
1
mixture), and H and 13C NMR spectra for 2a were in accord
with literature data6 reported for the cis isomer.
Given the efficiency with which symmetrical tetrahydropyrans
could be obtained by this process, we next explored the
possibility of preparing unsymmetrical tetrahydropyrans by
employing two different aldehydes added sequentially during
the reaction. Thus, stirring equimolar amounts of 1, hexanal,
and indium metal in 1:1 H2O/i-PrOH at room temperature for
10 h, followed by addition of 2.0 equiv of benzaldehyde and
stirring for 24 h, led to the formation of unsymmetrical
tetrahydropyran 2c in 45% yield (Scheme 3), along with
symmetrical tetrahydropyran 2b (10% yield) and homoallylic
alcohol 10 (40%). Further analysis of the reaction revealed that
hexanal is not completely consumed in its reaction with 1 in
the initial allylation step, and the intermediate homoallylic
(1) For a recent review of strategies for the formation of tetrahydropyran
rings in the synthesis of natural products, see: (a) Clarke, P.A.; Santos, S.
Eur. J. Org. Chem. 2006, 9, 2045. For examples of 2,6-disubstituted
tetrahydropyrans in natural product synthesis, see: (b) Clarke, P. A.; Martin,
W. H. C. Tetrahedron 2005, 61, 5433. (c) Smith, A. B., III; Safonov, I. G.
Org. Lett. 2002, 4, 635. (d) Smith, A. B., III; Safonov, I. G.; Corbett, R.
M. J. Am. Chem. Soc. 2001, 123, 12426.
(2) For recent efforts toward the construction of tetrahydropyran rings,
see: (a) Clarke, P. A.; Martin, W. H. C. Tetrahedron Lett. 2004, 45, 9061.
(b) Marko, I. E.; Plancher, J.-M. Tetrahedron Lett. 1999, 40, 5259. (c)
Marko, I. E.; Leroy, B. Tetrahedron Lett. 2000, 41, 7225. (d) Leroy, B.;
Marko, I. E. J. Org. Chem. 2002, 67, 8744. (e) Dubost, C.; Marko, Bryans,
I. E. J. Tetrahedron Lett. 2005, 46, 4005. (f) Yu, C.-M.; Lee, J.-Y.; So, B.;
Hong, J. Angew. Chem., Int. Ed. 2002, 41, 161. (g) Boivin, T. L. B.
Tetrahedron 1987, 43, 3309.
(3) For indium-mediated tetrahydropyran and dihydropyran syntheses,
see: (a) Viswanathan, G. S.; Yang, J.; Li, C.-J. Org. Lett. 1999, 1, 993. (b)
Yang, J.; Viswanathan, G. S.; Li, C.-J. Tetrahedron Lett. 1999, 40, 1627.
(c) Zhang, W.-C.; Li, C.-J. Tetrahedron 2000, 56, 2403. (d) Chan, K.-P.;
Loh, T.-P. Org. Lett. 2005, 7, 4491. (e) Loh, T.-P.; Yang, J.-Y.; Feng, L.-
C.; Zhou, Y. Tetrahedron Lett. 2002, 43, 7193. (f) Chan, K.-P.; Loh, T.-P
Tetrahedron Lett. 2004, 45, 8387. (g) Dobbs, A. P.; Martinovic, S.
Tetrahedron Lett. 2002, 43, 7055.
(4) Compound 1 was prepared (MsCl, Et3N, THF; NaI, acetone) from
the corresponding allylic alcohol: Trost, B. M.; Chan, D. M. T.; Nanninga,
N. Org. Synth. 1984, 62, 58.
(5) Allatabakhsh, A.; Pham, M.; Minehan, T. G. Heterocycles 2007, 72,
115.
(6) For the spectral data of 2a, see ref 3e.
10.1021/jo7016857 CCC: $40.75 © 2008 American Chemical Society
Published on Web 12/21/2007
J. Org. Chem. 2008, 73, 741-744
741

TABLE 1. Scope of Indium-Mediated cis-2,6-Tetrahydropyran
SCHEME 3. Initial Attempts at Preparation of
Unsymmetrical Tetrahydropyrans
Synthesisa
% yield
entry
R1
R2
2
of 2b
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Ph
Ph
C5H11
Ph
4-Cl-Ph
Ph
4-Cl-Ph
C5H11
4-Cl-Ph
Ph
4-OMe-Ph
(E)-PhCHCH
C5H11
4-OMe-Ph
4-OMe-Ph
a
76
83
75
68
79
65
C5H11
C5H11
C5H11
i-Pr
i-Pr
i-Pr
b
c
d
e
f
g
h
i
j
k
l
m
n
SCHEME 4. Solvent Effect in the Preparation of
Unsymmetrical Tetrahydropyrans
68
Ph
60c
55c
75
d
PhCH2OCH2
4-Cl-Ph
Ph
(E)-C5H11CHCH
(Z)-DPSOCH2CHCHf
4-DPSO-C6H4(CH2)3
65
60c
70e
50c,g
d,f
a Reaction conditions: 1 + R1CHO in 1:1 H2O/THF (1M), 5-12 h;
concentration in vacuo, addition of 2.0 equiv of R2CHO in 1:1 H2O/i-PrOH
(2 M), 16 h. b Yields refer to products isolated after silica gel chromatog-
raphy. c Lower isolated yields due to formation of symmetrical tetrahydro-
pyran byproducts and/or protodesilylation of the intermediate homoallylic
alcohol (see text). d Allylation required 24 h stirring at room temperature.
e Product isolated as a 3:1 mixture of Z:E olefin isomers. f DPS ) tert-
butyldiphenylsilyl. g Prins reaction performed in the absence of solvent.
alcohol 6 begins to react with residual hexanal to form 2b even
before benzaldehyde is added. Furthermore, since the acidity
of the medium increases as allylation proceeds (to pH ∼2),7
protodesilylation of the intermediate homoallylic alcohol 6
becomes a major side reaction over extended reaction times.8
The resulting alcohol 7 is unreactive toward benzaldehyde under
these conditions.
To address these problems, the reaction solvent composition
was adjusted. Equimolar amounts of 1, hexanal, and indium
metal were combined in 1:1 H2O/THF and stirred at room
temperature. Complete consumption of the starting materials
by TLC analysis was observed within 5 h, and no evidence of
2b formation could be detected by GC analysis. However, when
2.0 equiv of benzaldehyde was added and the mixture was stirred
at room temperature for 16 h, unsymmetrical tetrahydropyran
2c3e was obtained in only 8% yield, with the vast majority of
the reaction mixture consisting of homoallylic alcohol 6, along
with some minor amounts (<10%) of 7. Under the assumption
that the Lewis basic solvent THF inhibits the acid-promoted
Prins reaction, we subsequently altered the procedure by briefly
concentrating the allylation reaction in vacuo (to remove THF)
just prior to the addition of benzaldehyde (dissolved in 1:1 H2O/
i-PrOH); after stirring the resulting 2 M solution at room
temperature for 16 h, a dramatically improved yield of 2c (75%)
was obtained, and only minimal amounts of 7 (<5%) were
recovered from the reaction mixture (Scheme 4). It was
subsequently discovered that similarly high yields of 2c (70-
80%) could be obtained when the Prins step was performed in
the absence of solvent by simply stirring the concentrated
allylation reaction mixture with 2 equiv of benzaldehyde
overnight. Interestingly, utilizing the same procedures but
employing stoichiometric benzaldehyde for the allylation step
and 2.0 equiv of hexanal for the Prins reaction resulted in only
a 52% yield of 2c, along with significant amounts (∼15-20%)
of 2b (vide infra). GC-MS analysis of all crude reaction
mixtures indicated product formation with >15:1 stereoselec-
tivity, and the cis stereochemistry of major product 2c was
confirmed by a two-dimensional NMR (NOESY) experiment,
in which strong cross-peaks between the axial protons at C2
and C6 were observed (see Supporting Information).
We evaluated the scope of this process by employing a variety
of alkyl, aryl, and R,â-unsaturated aldehydes in both the
allylation and Prins cyclization steps. As shown in Table 1,
diverse cis-2,6-disubstituted tetrahydropyrans can be prepared
by this method. The two-step, one-pot procedure provides
superior yields (compared to using 1:1 H2O/i-PrOH as solvent
alone, vide supra) even for the preparation of the symmetrical
tetrahydropyrans 2a and 2b (entries 1 and 2). It is also
noteworthy that acid-labile silyl-protecting groups survive the
coupling process (entries 13 and 14,16 DPS ) tert-butyldiphe-
(9) For a discussion on erosion of enantioselectivity and side-chain
exchange via the oxonium-Cope rearrangement in Prins reactions, see: (a)
Cheng-Hsia, A. L.; Loh, T.-P. Tetrahedron Lett. 2006, 47, 1641. (b) Crosby,
S. R.; Harding, J. R.; King, C. D.; Parker, G. D.; Willis, C. L. Org. Lett.
2002, 4, 577. (c) Marumoto, S.; Jaber, J. J.; Vitale, J. P.; Rychnovsky, S.
D. Org. Lett. 2002, 4, 3919.
(10) Chan et al. have presented evidence that the intermediate in aqueous
indium-mediated allylation reactions is allylindium(I): Chan, T. H.; Yang,
T. J. Am. Chem. Soc. 1999, 121, 3228.
(11) Prepared from commercially available 3-(trimethylsilyl)allyl alcohol
(Aldrich) by mesylation (MsCl, Et3N, CH2Cl2, 0 °C) followed by iodide
displacement (NaI, acetone, rt, 24 h).
(12) Spectroscopic data (1H, 13C NMR) for 9 matched those reported by
Hinkle for the same compound with syn-2,6-stereochemistry: (a) Lian, Y.;
Hinkle, R. J. J. Org. Chem. 2006, 71, 7071. Although Li reports the use of
E-vinylsilanes for the synthesis of dihydropyrans (ref 3a), Brimble reports
that the Z-configuration of the vinylsilane is crucial for the formation of
dihydropyrans: (b) Meilert, K.; Brimble, M. A. Org. Biolom. Chem. 2006,
4, 2184.
(13) For the preparation of 10, see: (a) Majetich, G.; Song, J.-S.; Ringold,
C.; Nemeth, G. A.; Newton, M. G. J. Org. Chem. 1991, 56, 3973. (b)
Carlson, R. M. Tetrahedron Lett. 1978, 19, 111.
(7) Assessed by a simple litmus paper test.
(8) As much as 50% of desilylated homoallylic alcohol was isolated from
the reaction mixtures in some cases.
(14) For an alternative preparation of cyclohexenyl-fused pyrans, see:
Loh, T.-P.; Feng, L.-C.; Yang, J.-Y. Synthesis 2002, 7, 937.
742 J. Org. Chem., Vol. 73, No. 2, 2008

SCHEME 5. Oxonium-Cope Rearrangement Gives Rise to the Symmetrical Tetrahydropyran Byproduct
SCHEME 6. Allylation/Cyclization Reactions of Allyl
Iodides 8, 10, and 11
nylsilyl). Reactions requiring longer times for the initial
allylation step (entries 9 and 14) resulted in lower overall yields
due to partial homoallylic alcohol desilylation (of the trimeth-
ylsilyl group) and/or formation of the symmetrical tetrahydro-
pyran byproduct.
Moderate yields were also encountered in the preparation of
substrates 2h and 2l (entries 8 and 12, respectively), as was the
case for preparation of 2c when benzaldehyde was used for
allylation and hexanal for Prins cyclization. These results may
be rationalized by noting that the second added aldehyde in each
case gives rise to a less stable oxocarbenium ion intermediate
(resonance structures B and D, Scheme 5) than would be
obtained from the first added aldehyde; as a result, an alternative
reaction pathway involving a [3,3]-sigmatropic oxonium-Cope
rearrangement may take place instead of Prins cyclization,
leading to a more stable oxocarbenium ion (resonance structures
C and E, Scheme 5).9 Hydrolysis of ion C/E to the correspond-
ing aldehyde and homoallylic alcohol (F) could then give rise
to symmetrical tetrahydropyran byproducts due to reaction of
F with the excess of second aldehyde present in the reaction
medium. As described above in the preparation of 2c (Scheme
4), this problem can largely be avoided by judicious choice in
the order of aldehyde addition (note that the reverse process,
ions C/E f B/D, should be less favorable).
Several further observations were made in an attempt to
elucidate the identity of the promoter of the Prins cyclization
process. In the synthesis of 2c, upon completion of allylindium
addition to hexanal, dark burgundy salts are evident in the
reaction mixture, perhaps indicating the presence of indium(I)
iodide.10 Concentration of the reaction mixture and addition of
benzaldehyde in 1:1 H2O/i-PrOH to make a 2 M solution
resulted in a milky-white aqueous suspension (containing dark
granules) at pH ∼2. Addition of 0.5 equiv of sodium acetate as
a buffer (f pH 4) at this point completely inhibits the Prins
cyclization reaction. Addition of 1 equiv of iodine instead (to
presumably oxidize InI to InI3) and stirring at room temperature
for 24 h produced 2c in low yields (30%). To further probe the
identity of the acid (Brønsted or Lewis) promoting the Prins
process, isolated homoallylic alcohol 6 (obtained after aqueous
extraction, drying, and evaporation of the allylation reaction
mixture) was subjected to Prins cyclization with 2 equiv of
benzaldehyde in 1:1 0.1 N HCl/i-PrOH or in 1:1 H2O/i-PrOH
in the presence of either 1 equiv of InCl3 or 1 equiv of InI.
Product 2c was obtained in all three cases, but superior
conversions (∼65%) were noted for the reaction promoted by
finely ground InI versus that promoted by either InCl3 (20%)
or 0.1 N HCl (30%). Surprisingly, simply stirring the isolated
homoallylic alcohol 6 together with 2 equiv of benzaldehyde
in the absence of either solvent or promoter led to the production
of 2c in approximately 25% yield. Taken together, these data
suggest an important role for InI in the Prins cyclization process,
along with the unexpected finding that InI is a superior promoter
in comparison to InCl3.
With these data in hand, we explored the use of alternative
allyl iodides in our one-pot allylation/Prins cyclization proce-
dure. Not surprisingly, combining 1-iodo-3-trimethylsilyl-2-
propene 811 with hexanal and benzaldehyde according to our
established protocol (with 48 h stirring for the Prins step)
furnished dihydropyran 912 in only 35% yield; the intermediate
homoallylic alcohol was also recovered in approximately 60%
yield (Scheme 6). This result can be rationalized by the
requirement for formation of a less stable secondary â-silyl
cation during the Prins cyclization step for this substrate.
Unexpectedly, however, reaction of iodide 1013 with 1 equiv
each of indium metal and hexanal in 1:1 H2O/THF, followed
by concentration and addition of excess benzaldehyde, led to a
1:1 mixture of cyclohexenyl-fused pyran 12a14 and an unidenti-
fied structural isomer in a disappointing 50% overall yield. All
attempts to improve the yield of the desired product failed, and
by analyzing the structure of the byproduct of this reaction by
NMR spectroscopy, we hypothesized that migration of silicon
was occurring during the process. As a result, we instead
exposed 1-(iodomethyl)cyclohex-1-ene 1115 to our allylation/
Prins cyclization protocol with a variety of aldehydes. Gratify-
ingly, when the Prins cyclization step was performed in the
absence of solvent, pyrans 12a-c were obtained in moderate
yields with a single stereoisomer in excess (>15:1 for 12a, 10:1
for 12b, and 13:1 for 12c). The 11.6 Hz coupling constant of
(15) Compound 11 was prepared from commercially available methyl
cyclohexenecarboxylate (Aldrich) by reduction (2.5 equiv of DIBAL-H,
THF, -78 °C), mesylation (MsCl, Et3N, THF), and iodide displacement
(NaI, acetone, rt, 24 h).
(16) Aldehyde prepared from commercially available 4-hydroxybenzal-
dehyde by silylation (TBDPS-Cl, imidazole, DMF, 24 h, rt), Horner-
Emmons reaction (triethylphosphonoacetate, NaH, THF), reduction (DIBAL-
H, THF, -78 °C), olefin hydrogenation (H2, 10% Pd-C, rt, 24 h), and
oxidation (oxalyl chloride, DMSO, Et3N, CH2Cl2).
J. Org. Chem, Vol. 73, No. 2, 2008 743

SCHEME 7. A Short Formal Synthesis of (()-Centrolobine
showed complete disappearance of the starting material, R2CHO
(0.1 equiv) was added and the reaction mixture was concentrated
in vacuo on a rotary evaporator. Further R2CHO (2 equiv) was
added, either with 1:1 H2O/i-PrOH (to make a 2 M solution) as
solvent or in the complete absence of solvent, and the mixture was
stirred at room temperature for 16 h with protection from light.
The reaction was then diluted with ether (20 mL) and washed with
1 N HCl (2 × 20 mL). The combined aqueous phases were back-
extracted with ether (1 × 50 mL). The combined organic extracts
were then dried over Na2SO4, filtered, and concentrated in vacuo.
The residue was purified by flash chromatography (silica gel,
hexane/Et2O ) 99:1 to 95:5).
(()-4-Methylene-2-pentyl-6-phenyltetrahydro-2H-pyran (2c).
Following the general procedure, starting materials 1 (114 mg, 0.44
mmol), hexanal (44 mg, 0.44 mmol), and benzaldehyde (94 mg,
0.88 mmol) were combined to provide 2c as a colorless oil (81
mg, 0.33 mmol, 75%): 1H NMR (400 MHz, CDCl3) δ 7.41-7.37
(m, 5H), 4.81 (m, 2H), 4.35 (dd, J ) 1.6, 10.8 Hz, 1H), 3.45 (m,
1H), 2.50 (d, J ) 13.2 Hz, 1H), 2.32 (d, J ) 13.2 Hz, 1H), 2.24 (t,
J ) 12.8 Hz, 1H), 2.06 (t, J ) 11.6 Hz, 1H), 1.71-1.32 (m, 7H),
0.92 (t, J ) 6.0 Hz, 3H); 13C NMR (100 MHz, CDCl3) δ 145.1,
142.8, 128.3, 127.3, 125.8, 108.6, 80.1, 78.8, 42.9, 40.6, 36.3, 31.9,
25.1, 22.6, 14.1; GC-MS m/z 244 (M+).
the C2 proton in each product, in addition to the C2 proton-
C6 proton cross-peaks in the NOESY spectrum of 12a (see
Supporting Information), confirmed the relative stereochemistry
of the major diastereomers as 2,3-anti, 2,6-syn.
Finally, we realized that product 2n (Table 1) might serve as
an intermediate in a formal synthesis of the natural product
centrolobine.17 Oxidative cleavage of the alkene of 2n under
Jin’s conditions,18 followed by silyl ether deprotection, furnished
ketone 13,19 a compound that has been converted to racemic
centrolobine in quantitative yield by Clarke et al.20 (Scheme
7).
In summary, we have demonstrated an efficient one-pot
synthesis of cis-2,6-disubstituted tetrahydropyrans from substi-
tuted allyl iodides using indium metal as the sole promoter. This
environmentally benign protocol takes place in aqueous media
under conditions that tolerate acid-sensitive alcohol protecting
groups. Further experiments to extend the scope and delineate
the mechanism of this process are underway and will be reported
in due course.
Experimental Section
General Procedure for the Synthesis of 2a-n, 12a-c. To a 1
M solution of 3-iodo-2-[(trimethylsilyl)methyl]propene (1) in 1:1
THF/H2O were added R1CHO (1 equiv) and indium metal (1 equiv).
The reaction mixture (wrapped in aluminum foil for protection from
light) was stirred at room temperature for 5-12 h. When TLC
Acknowledgment. This paper is dedicated with great respect
to Professor Yoshito Kishi on the occasion of his 70th birthday.
We thank the National Institutes of Health (SC2 GM081064-
01), the ACS Petroleum Research Fund (No. PRF 45277-B1),
and Research Corporation (No. CC3343) for their generous
support of our research program. Mass spectra were provided
by the UC Riverside mass spectrometry department.
(17) (a) Boehrsch, V.; Blechert, S. Chem. Commun. 2006, 18, 1968. (b)
Prasad, K. R.; Anbarasan, P. Tetrahedron 2006, 63, 1089. (c) Carreno, M.
C.; Mazery, R. D.; Urbano, A.; Colobert, F.; Solladie´, G. J. Org. Chem.
2003, 68, 7779. (d) Marumoto, S.; Jaber, J. J.; Vitale, J. P.; Rychnovsky,
S. D. Org. Lett. 2002, 4, 3919. (e) Evans, P. A.; Cui, J.; Gharpure, S. J.
Org. Lett. 2003, 5, 3883. (f) Chandrasekhar, S.; Prakash, S. J.; Shyamsunder,
T. Tetrahedron Lett. 2005, 46, 6651. (g) Jennings, M. P.; Clemens, R. T.
Tetrahedron Lett. 2005, 46, 2021.
Supporting Information Available: Detailed experimental
procedures, spectroscopic data, and 1H NMR spectra for all
compounds in Table 1 and Schemes 6 and 7, as well as NOESY
spectra for compounds 2c and 12a. This material is available free
of charge via the Internet at http://pubs.acs.org.
(18) Yu, W.; Mei, Y.; Kang, Y.; Hua, Z.; Jin, Z. Org. Lett. 2004, 6,
3217.
(19) NMR spectral data (1H, 13C) for 13 were in complete accord with
those reported for this compound by Clarke et al. (ref 20).
(20) Clarke, P. A.; Santos, S. Eur. J. Org. Chem. 2006, 9, 2045.
JO7016857
744 J. Org. Chem., Vol. 73, No. 2, 2008