Diastereoselective synthesis of tetrahydrofurans via palladium(0)-catalyzed [3 + 2] cycloaddition of vinylcyclopropanes and aldehydes

Article abstract of DOI:10.1021/ol800819h

(Chemical Equation Presented) Palladium(0)-catalyzed cycloadditions of malonate-derived vinylcyclopropane 1 and aldehydes to afford 2,5-cis disubstituted tetrahydrofuran derivatives are described. Pd loadings as low as 0.5 mol % were effective in catalyzing the transformation with high yields and diastereoselectivities. Electron-poor aldehydes work best, suggesting that a mechanism involving an initial aldol reaction may be operative.

Full text of DOI:10.1021/ol800819h

ORGANIC
LETTERS
2008
Vol. 10, No. 12
2541-2544
Diastereoselective Synthesis of
Tetrahydrofurans via
Palladium(0)-Catalyzed [3 + 2]
Cycloaddition of Vinylcyclopropanes
and Aldehydes
Andrew T. Parsons, Matthew J. Campbell, and Jeffrey S. Johnson*
Department of Chemistry, UniVersity of North Carolina at Chapel Hill,
Chapel Hill, North Carolina 27599-3290
jsj@unc.edu
Received April 10, 2008
ABSTRACT
Palladium(0)-catalyzed cycloadditions of malonate-derived vinylcyclopropane 1 and aldehydes to afford 2,5-cis disubstituted tetrahydrofuran
derivatives are described. Pd loadings as low as 0.5 mol % were effective in catalyzing the transformation with high yields and
diastereoselectivities. Electron-poor aldehydes work best, suggesting that a mechanism involving an initial aldol reaction may be operative.
The importance of developing synthetic methods to prepare
tetrahydrofurans is evident due to the prevalence of these
moieties in structurally and biologically interesting mol-
ecules.1 A common method used to synthesize this scaffold
involves the generation of 1,3-dipolar equivalents and sub-
sequent trapping with a dipolarophile. Dipole formation can
be achieved by the generation of a carbonyl ylide through
thermolysis2 or photolysis3 of epoxides, or more commonly
by trapping of a metallocarbenoid with a carbonyl.4,5 In these
examples, the use of an alkene dipolarophile generates the
tetrahydrofuran (THF) product. Alternatively, trapping of a
carbonyl with a three-carbon 1,3-dipole equivalent also yields
THFs. It has been demonstrated that donor-acceptor (D-A)
cyclopropanes, particularly those with heteroatom or other
specialized, highly activating donor substituents (e.g., trim-
ethylsilyl methyl), are useful three-carbon building blocks
for the Lewis acid-catalyzed cycloaddition of aldehydes6 and
(4) Reviews: (a) Doyle, M. P. Chem. ReV. 1986, 86, 919. (b) Padwa, A.
Acc. Chem. Res. 1991, 24, 22. (c) Padwa, A.; Hornbuckle, S. F. Chem.
ReV. 1991, 91, 263. (d) Padwa, A.; Weingarten, M. D. Chem. ReV. 1996,
96, 223. (e) Doyle, M. P.; Forbes, D. C. Chem. ReV. 1998, 98, 911. (f)
Padwa, A.; Pearson, W. H. Synthetic Applications of 1,3-Dipolar Cycload-
dition Chemistry toward Heterocycles and Natural Products; John Wiley
and Sons: Hoboken, NJ, 2003. (g) Padwa, A. J. Organomet. Chem. 2005,
690, 5533.
(1) Recent examples: (a) Grese, T. A.; Hutchinson, K. D.; Overman,
L. E. J. Org. Chem. 1993, 58, 2468. (b) Kim, E.-J.; Tian, F.; Woo, M.-H.
J. Nat. Prod. 2000, 63, 1503. (c) MacMillan, D. W. C.; Overman, L. E.;
Pennington, L. D. J. Am. Chem. Soc. 2001, 123, 9033. (d) Crimmins, M. T.;
She, J. J. Am. Chem. Soc. 2004, 126, 12790. (e) Tinsley, J. M.; Roush,
W. R. J. Am. Chem. Soc. 2005, 127, 10818. (f) Mihovilovic, M. D.; Bianchi,
D. A.; Rudroff, F. Chem. Commun. 2006, 3214. (g) Volz, F.; Krause, N.
Org. Biomol. Chem. 2007, 5, 1519. (h) Morinaka, B. I.; Skepper, C. K.;
Molinski, T. F. Org. Lett. 2007, 9, 1975. For a review, see: (i) Wolfe, J. P.;
Hay, M. B. Tetrahedron 2007, 63, 261.
(5) For select examples in synthesis, see: (a) Wood, J. L.; Thompson,
B. D.; Yusuff, N.; Pflum, D. A.; Mattha¨us, M. S. P. J. Am. Chem. Soc.
2001, 123, 2097. (b) Nakamura, S.; Hirata, Y.; Kurosaki, T.; Anada, M.;
Kataoka, O.; Kitagaki, S.; Hashimoto, S. Angew. Chem., Int. Ed. 2003, 42,
5351. (c) Nair, V.; Mathai, S.; Varma, R. L. J. Org. Chem. 2004, 69, 1413.
(d) Lu, C.-D.; Chen, Z.-Y.; Liu, H.; Hu, W.-H.; Mi, A.-Q.; Doyle, M. P. J.
Org. Chem. 2004, 69, 4856.
(2) (a) Huisgen, R. Angew. Chem. Int. Ed. 1977, 16, 572. (b) Huisgen,
R.; de March, P. J. Am. Chem. Soc. 1982, 104, 4953.
(3) Das, P. K.; Griffin, G. W. J. Photochem. 1984, 27, 317.
10.1021/ol800819h CCC: $40.75
Published on Web 05/27/2008
2008 American Chemical Society

ketones6a,7 to afford THF derivatives. Our group has demon-
strated that THF derivatives are accessible from malonate-
derived cyclopropanes with a carbon-donor substituent via
diastereoselective and enantiospecific Sn(OTf)2- and SnCl4-
catalyzed cycloadditions with aldehydes.8 In this Letter, we
describe a new electronically complementary method to
synthesize THF derivatives from D-A vinylcyclopropane
1 through a Pd(0)-catalyzed [3 + 2] cycloaddition with
aldehydes.
The use of Pd(0) to generate π-allylic species is a well-
studied area and is commonly utilized for the allylation of
nucleophiles.9 π-Allyl chemistry was expanded to cycload-
ditions nearly three decades ago when Trost introduced
palladium-trimethylenemethane (Pd-TMM) as a three-carbon
dipole equivalent useful for [3 + 2] cycloadditions (Figure
1, top).10 More recently, Tsuji generated the three-carbon
π-allylpalladium dipole equivalent 2 through ring-opening
of 1 by Pd2(dba)3•CHCl3 (Figure 1, bottom). The resulting
complex was shown to undergo formal [3 + 2] cycloaddi-
tions with potent electrophiles such as aryl isocyanates and
various acrylates to yield γ-lactams11 and cyclopentanes,12
respectively. On the basis of these and other examples of
hetero- and carbocycle formation from cyclopropanes,13 we
envisioned the synthesis of tetrahydrofurans via Pd(0)-
catalyzed ring-opening of 1 and subsequent cycloaddition
with aldehydes.14
as reported in the literature.13d Due to the need for highly
electrophilic substrates in Tsuji’s systems, we first examined
the use of the electron-deficient 4-trifluoromethylbenzalde-
hyde. Preliminary results are reported in Figure 2. The
reaction was sluggish when commercially available Pd(dba)2
was used, requiring 28 h to reach completion. Recent reports
demonstrating the increased reactivity of Pd(0) with electron-
rich dba derivatives prompted us to employ bis(4-methoxy-
benzylidene)acetone (MeO-dba) ligands on our Pd(0) pre-
catalyst.15 We noticed a marked improvement in reactivity
under the same reaction conditions when Pd2(MeO-dba)3 was
used (Figure 2). 2,2′-Dipyridyl (bipy) was chosen as the
ligand based on initial studies indicating superiority over
monodentate amines and phosphines.
Figure 2.
Identity of the Pd0 precatalyst has an effect on reaction
rate.
Although 2-MeTHF was found to be a superior solvent
for the transformation in Figure 2, it was not found to be
generally applicable. Competitive oligomerization of 1 was
observed when other aldehydes were employed in 2-MeTHF.
A solvent screen revealed that toluene reduced oligomer-
ization substantially and thus was used in subsequent method
optimization experiments.
Table 1 summarizes the effect of the supporting ligand
on reaction efficiency and diastereoselection. For a uniform
comparison, reactions were arbitrarily stopped and analyzed
after 48 h. The discrepancy between cyclopropane consump-
tion (% conversion) and tetrahydrofuran yield provides an
approximate measure of competitive, nonproductive oligimer-
ization of 1 for a given Pd/ligand complex. By changing the
bidentate amine ligand from bipyridine to phenanthroline,
we noted an improvement in diastereoselectivity (entry 3).
Further improvement in diastereocontrol and reaction ef-
ficiency was realized when bathophenanthroline (bphen) was
used, although the rate was somewhat slower (entry 5). Both
monodentate and bidentate phosphine ligands performed
poorly in promoting THF formation (entries 6 and 7).
With both the diastereoselectivity and the product/oligimer
Figure 1. Trost’s Pd-TMM reagent (top) and catalytic ring opening/
cycloaddition of a malonate-derived vinylcyclopropane via π-al-
lylpalladium formation (bottom).
Preparation of 1 was achieved in one step through a double
alkylation of dimethyl malonate with 1,4-dibromo-2-butene
(6) (a) Sugita, Y.; Kawai, K.; Yokoe, I. Heterocycles 2000, 53, 657.
(b) Gupta, A.; Yadav, V. K. Tetrahedron Lett. 2006, 47, 8043
.
(7) Sugita, Y.; Kawai, K.; Yokoe, I. Heterocycles 2001, 55, 135
.
(8) (a) Pohlhaus, P. D.; Johnson, J. S. J. Org. Chem. 2005, 70, 1057.
(b) Pohlhaus, P. D.; Johnson, J. S. J. Am. Chem. Soc. 2005, 127, 16014.
(9) Reviews: (a) Tsuji, J. Acc. Chem. Res. 1969, 2, 144. (b) Trost, B. M.
Acc. Chem. Res. 1980, 13, 385. (c) Trost, B. M.; Van Vranken, D. L. Chem.
ReV. 1996, 96, 395. (d) Trost, B. M.; Crawley, M. L. Chem. ReV. 2003,
103, 2921. (e) Tsuji, J. Palladium reagents and catalysts: new perspectiVes
for the 21st century, 2nd ed.; John Wiley and Sons: Hoboken, NJ, 2004.
(10) (a) Trost, B. M.; Chan, D. M. T. J. Am. Chem. Soc. 1979, 101,
6429. (b) Trost, B. M.; Chan, D. M. T. J. Am. Chem. Soc. 1983, 105, 2315.
(c) Trost, B. M.; Chan, D. M. T. J. Am. Chem. Soc. 1983, 105, 2326. (d)
Trost, B. M. Angew. Chem., Int. Ed. 1986, 25, 1. (e) Trost, B. M.; Sharma,
S.; Schmidt, T. J. Am. Chem. Soc. 1992, 114, 7903. (f) Trost, B. M.;
Silverman, S. M.; Stambuli, J. P. J. Am. Chem. Soc. 2007, 129, 12398. (g)
Shintani, R.; Park, S.; Duan, W.-L.; Hayashi, T. Angew. Chem., Int. Ed.
2007, 46, 5901.
(13) For selected examples, see: (a) Hiroi, K.; Yamada, A. Tetrahedron:
Asymmetry 2000, 11, 1835. (b) Young, I. S.; Kerr, M. A. Angew. Chem.,
Int. Ed. 2003, 42, 3023–3026. (c) Zuo, G.; Louie, J. Angew. Chem., Int.
Ed. 2004, 43, 2277. (d) Bowman, R. K.; Johnson, J. S. Org. Lett. 2006, 8,
573. (e) Liu, L.; Montgomery, J. Org. Lett. 2007, 9, 3885. (f) Perreault, C.;
Goudreau, S. R.; Zimmer, L. E.; Charette, A. B. Org. Lett. 2008, 10, 689.
(14) For Pd-catalyzed tetrahydrofuran formation from vinyl epoxides
and alkenes, see: Shim, J.-G.; Yamamoto, Y. J. Org. Chem. 1998, 63, 3067.
(15) (a) Fairlamb, I. J. S.; Kapdi, A. R.; Lee, A. F. Org. Lett. 2004, 6,
4435. (b) Firmansjah, L.; Fu, G. C. J. Am. Chem. Soc. 2007, 129, 11340.
(c) White, D. E.; Stewart, I. C.; Grubbs, R. H.; Stoltz, B. M. J. Am. Chem.
Soc. 2008, 130, 810.
(11) Yamamoto, K.; Ishida, T.; Tsuji, J. Chem. Lett. 1987, 1157.
(12) Shimizu, I.; Ohashi, Y.; Tsuji, J. Tetrahedron Lett. 1985, 26, 3825.
2542
Org. Lett., Vol. 10, No. 12, 2008

hydes, providing high yields and short reaction times (entries
8-10, 12, 15). Electron-rich aldehydes are problematic, with
only trace product obtained with p-anisaldehyde, despite
complete consumption of cyclopropane (entry 3). The linear,
unbranched aliphatic hexanal was also a successful substrate,
although the dr was poor (entry 16). An aliphatic aldehyde
containing R-branching was unreactive (entry 17). Although
the reaction rate can often be increased by boosting the
catalyst loading, nonproductive oligimerization increases as
well. Catalyst loading and reaction temperature were vari-
ables found to be in a delicate balance that required some
fine-tuning for each substrate; therefore, the reaction condi-
tions in Table 2 vary slightly as the aldehyde partner changes.
We believe that electron-rich aldehydes are not electro-
philic enough to undergo nucleophilic attack by π-allylpal-
ladium complex 2. As a result, the majority of the starting
material oligomerizes prior to cycloaddition. It is interesting
that in the absence of aldehyde, only 15% conversion of the
cyclopropane to oligomer occurs.16a To examine the effect
of the aldehyde on oligomerization, we carried out an
experiment using the standard reaction conditions with 0.50
Table 1. Ligand Effect on Reaction Efficiency and
Diastereoselectiona
drc
% conversionb % yieldb (cis/trans)
entry
ligand
1
2
2,2′-bipyridyl
4,4′-dimethyl-2,2′-
dipyridyl
83
45
48
32
79:21
87:13
3
4
5
6
7
phenanthroline
bathocuproine
bathophenanthroline
PPh3
78
61
65
12
85
47
14
48
9
87:13
77:23
89:11
64:36
57:43
rac-binap
16
a Conditions: 1.0 equiv of 1, 6.0 equiv of aldehyde, 0.0125 equiv of
Pd2(MeO-dba)3, 0.05 equiv of ligand, 0.30 mL of C7H8 ([1]0 ) 1.09 M),
b
1
48 h, rt under N2. Determined by H NMR analysis using a mesitylene
internal standard (average of at least two trials). c Determined by 1H NMR
spectroscopy.
1
equivalents of p-anisaldehyde. Analysis by H NMR spec-
troscopy revealed 91% consumption of 1 but only 22% of
the aldehyde.16b The reaction was accompanied by significant
oligimer formation, suggesting that p-anisaldehyde plays
some role in copromoting the oligomerization.
ratio maximized using the bathophenanthroline/Pd2(MeO-
dba)3-derived complex, we examined a variety of electroni-
cally diverse aromatic aldehydes with this catalyst system
(Table 2). This method works best for electron-poor alde-
Table 2. Cycloaddition of 1 with Aldehydes 3a-qa
equiv temp time
drd
[Pd]2 (°C) (h) % yieldc (cis/trans)
b
entry
R
1
2
C6H5 (3a)
0.0075 40 25
53
71
<5
80
91
64
80
93
98
96
95
99
92
98
100
89
<5
89:11
92:8
nd
92:8
89:11
87:13
83:17
83:17
91:9
80:20
89:11
82:18
98:2
4-MeC6H4 (3b)
4-MeOC6H4 (3c)
4-BrC6H4 (3d)
4-ClC6H4 (3e)
4-FC6H4 (3f)
4-MeO2CC6H4 (3g) 0.005
4-NCC6H4 (3 h)
4-F3CC6H4 (3i)
0.025
0.025
rt 96
40 24
Figure 3. Proposed catalytic cycle.
3
4e
5
0.0125 rt 24
0.005 40 12
0.0125 rt 96
40 27
0.005 rt 5.5
Data from the aldehyde screen provided mechanistic insight
into this reaction. Electron-poor aldehydes react at a much faster
rate and require a lower catalyst loading, supporting an aldol-
type mechanism in which π-allylpalladium complex 2 attacks
the electrophilic aldehyde to form alkoxide 5 (Figure 3).17 Ring
closure and displacement of the Pd(0) catalyst yields the THF
product to complete the catalytic cycle. The inability of
p-anisaldehyde to participate productively in the reaction is
consistent with this hypothesis.
6
7e
8e
9
0.0125 rt
0.0025 rt
4
5
10e 4-O2NC6H4 (3j)
11 3-ClC6H4 (3k)
12e 3-O2NC6H4 (3l)
13 2-ClC6H4 (3m)
14 2-FC6H4 (3n)
15e 2-O2NC6H4 (3o)
16 nC5H11 (3p)
0.0125 40 21
0.005
0.025
0.005
rt
40
40 20
2
3.5
95:5
0.0125 rt
3
87:13
69:31
nd
0.0125 40 22
0.0125 40 24
(16) Conditions: (a) One equivalent of 1 (0.326 mmol), 0.025 equiv of
Pd2(MeO-dba)3 (0.0082 mmol), 0.10 equiv of bphen (0.0164 mmol), 0.30
mL C7H8 ([1]0 ) 1.09 M), 40 °C, 17.5 h. (b) One equivalent of 1 (0.326
mmol), 0.50 equiv of p-anisaldehyde (0.163 mmol), 0.025 equiv of
Pd2(MeO-dba)3 (0.0082 mmol), 0.10 equiv of bphen (0.0164 mmol), 0.30
mL of C7H8 ([1]0 ) 1.09 M), 40 °C, 18 h under N2. 1H NMR analysis was
carried out using a mesitylene internal standard.
17 cC6H11 (3q)
a Conditions: 1 equiv of 1, 6 equiv of RCHO, 4 equiv of bphen relative
to equiv Pd2(MeO-dba)3, 0.30 mL of C7H8 ([1]0 ) 1.09 M), under N2. b [Pd]2
) Pd2(MeO-dba)3. c Isolated yield after column chromatography (average
of at least two trials). d Determined by 1H NMR analysis of the unpurified
product. e Three equivalents of aldehyde were used.
(17) (bphen)PdCl2 does not catalyze the cycloaddition, suggesting that
a mechanism involving a Pd(II) Lewis acid is unlikely.
Org. Lett., Vol. 10, No. 12, 2008
2543

Recent work in our laboratory showed that the rear-
rangement of enantioenriched vinylcyclopropanes via a
π-allylnickel complex affords dihydrofurans without loss
of stereochemical fidelity.13c On the basis of these results,
we were interested in examining the stereochemical
outcome of the Pd(0)-catalyzed [3 + 2] cycloaddition
when enantioenriched (S)-1 (er: 99:1) was used. Under
the standard reaction conditions (Figure 4), we observed
achiral intermediate and/or racemization of the starting
material.
In conclusion, we have developed an efficient, diastereo-
selective method to synthesize tetrahydrofuran derivatives
via a Pd(0)-catalyzed [3 + 2] cycloaddition of vinylcyclo-
propanes and aldehydes. This reaction works well for
electron-poor aldehydes, which is complementary to the
Lewis acid-catalyzed cycloaddition developed in our labora-
tory. Research to develop an asymmetric variant of this reac-
tion is currently underway.
Acknowledgment. This work was supported by the NSF
(CHE-0239363 and CHE-0749691). J.S.J. is an Alfred P.
Sloan Fellow and a Camille-Dreyfus Teacher Scholar.
Additional support from 3M, Eli Lilly, and Amgen is
gratefully acknowledged.
Supporting Information Available: Experimental pro-
cedures, spectroscopic and analytical data for all new com-
pounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
Figure 4. Use of enantioenriched starting material results in nearly
racemic product.
nearly complete degradation of enantioenrichment in the
product (er: 52.5:47.5), suggesting the generation of an
OL800819H
2544
Org. Lett., Vol. 10, No. 12, 2008