4082
E. Janus et al. / Tetrahedron Letters 47 (2006) 4079–4083
uncatalyzed endo reaction. The energy difference
between the exo and endo TSs was only À0.4 kcal/mol,
and poorly reflected the preferential formation of the
endo isomer, in accordance with experiment.
tetrafluoroboric acid (8.78 g, 100 mmol, 48% solution
in water) was added in portions with stirring. The
mixture was stirred at room temperature for 3 h and
then concentrated on a rotary evaporator. The product
was washed with ethyl acetate (2 · 20 mL) and dried
The sum of the Mulliken atomic charges in the TSs of
the atoms of cyclopentadiene and the maleate fragment
showed that the electrons were shifted from the diene to
the dienophile (0.148 and 0.156 for exo and endo, respec-
tively, but 0.201 and 0.200 for the protic IL catalyzed
reaction). Selected bond orders (BOs) were used to ap-
praise the extent of reaction at the TS. We observed that
the BOs for the formation of bonds and for the diene
and dienophile fragments of the TS spanned a reliable
range (Fig. 2). The distances and BO trends suggested
that the catalyst augmented dienophile acceptor proper-
ties only slightly. For non-catalyzed and catalyzed TSs,
the distance sums of C3–C6 and C5–C7 bond formation
were essentially the same but the sum of the BOs was
greater for the non-catalyzed TS for both endo and
exo structures. This was ascribed to the more product-
like character of the TS in the non-catalyzed reaction.
Examination of the geometries of the TSs and their
BOs evidently indicated a concerted reaction, although
involving an asynchronous mechanism. The imidazo-
lium-catalyzed process was much more synchronous:
the BO differences for the formation of carbon–carbon
bonds were 0.001 and 0.008 for exo and endo, respec-
tively, while the differences amounted to 0.036 and
under a vacuum (2–3 Torr) for 8 h at 70 ꢁC (90% yield).
1
H NMR (CD CN) 3.79 (s, 3H), 7.27 (s, 2H), 8.18
3
+
13
(s, 1H), 12.24 (s, N –H). C NMR (CD CN) 35.4,
3
122.9, 123.7 and 137.2. Elemental analysis: calcd for
C H BF N ; C 28.28, H 4.15, N 16.49. Found: C
4
7
4
2
28.56, H 4.48, N 15.25.
1-Methylimidazolium bis(trifluoromethylsulfonyl)imide
5: 1-Methylimidazolium chloride (3.56 g, 30 mmol) was
dissolved in 30 mL of distilled water and LiNTf2
(8.61 g, 30 mmol) was added. The reaction mixture
was stirred at room temperature for 10 min. After sepa-
ration of the phases, the organic phase was washed with
2 · 20 mL distilled water until chloride ions were no
longer detected (AgNO ). The liquid obtained was dried
3
1
at 70 ꢁC in a vacuum (98% yield). H NMR (DMSO-d )
6
3.89 (s, 3H), 7.66 (s, 2H), 9.03 (s, 1H), 12.17 (s, wide,
+
13
N –H).
C NMR (DMSO-d ) 35.4, 119.7, 123.0,
6
135.8 and anion 125.9 (q, –CF , J = 321 Hz). Elemental
3
analysis: calcd for C H F N O S ; C 19.84, H 1.94, N
6
7
6
3
4 2
11.57. Found: C 20.06, H 2.28, N 11.29.
1-Methylimidazolium chloride (mp = 75 ꢁC) was pre-
pared by saturating a chloroform solution of 1-methyl-
imidazole at 15 ꢁC with HCl.
0
.049 for the non-catalyzed process. Moreover, the sum
of the active orders for the non-catalyzed reaction was
close to 6.45 but was only 6.41 for the imidazolium-
catalyzed reaction and it did not depend on the exo/endo
stereoselectivity.
Typical procedure for the Diels–Alder reaction in protic
IL: 0.25 mL of IL, cyclohexanone (10 lL) as an internal
chromatographic standard, dimethyl maleate or methyl
acrylate (1 mmol) and freshly cracked cold cyclopenta-
diene (1.5 mmol) were added into a 4 mL vial containing
a small stirring bar. The reaction was conducted at
25 ꢁC. The progress of the reaction was monitored by
GC analysis over 48 h. The yield of products and
endo/exo ratios were calculated based on the GC
analysis.
In conclusion, we have reported on the use of protic,
imidazolium-based ILs as reaction media and as
Brønsted catalysts for Diels–Alder reactions. In com-
parison to conventional organic solvents, fast conver-
sion with good endo/exo selectivities was observed. On
the grounds of the semi-empirical computational model,
we have found that hydrogen bonding of protic imid-
azolium ILs to the dienophile provides a rationale for
the catalysis observed.
Acknowledgement
We are grateful for the financial support received from
the Polish Committee for Scientific Research, Grant
No. 3 T09B 053 29.
2
. Experimental and computational methods
Preparation of 1-alkyl- and 1-alkoxymethylimidazolium
salts followed the published method.27 The densities of
the obtained RTILs ranged from 1.1231 to 0.9861 g/
mL for lactates 1–2 and from 1.1396 to 1.0578 g/mL
for salicylates 3, and they were thermally stable up to
85–244 ꢁC. Gas phase PM3 semi-empirical calcula-
tions employing HyperChem (Hypercube) software were
used to model structures of reagents, transition states
References and notes
2
8
1
1. Diels, O.; Alder, K. Liebigs Ann. 1928, 460, 95–122.
2
3
4
5
. Oikawa, H.; Tokiwano, T. Nat. Prod. Rep. 2004, 21, 321–
52.
. Robertson, A.; Philip, D.; Spencer, N. Tetrahedron 1999,
5, 11365–11384.
. Manka, J. T.; Douglass, A. G.; Kaszynski, P.; Friedli, A.
C. J. Org. Chem. 2000, 65, 5202–5206.
3
(
TS) and products. The bond order (BO) calculations
were performed with MOPAC2002 software (Fujitsu
Ltd).
5
. Rulisek, L.; Sebek, P.; Havlas, Z.; Hrabal, R.; Capek, P.;
Svatos, A. J. Org. Chem. 2005, 70, 6295–6302.
1
-Methylimidazolium tetrafluoroborate 4: 1-Methylimid-
azole (8.21 g, 100 mmol) was placed in a round-
bottomed flask equipped with a stirring bar and
6. Hirao, H.; Ohwada, T. J. Phys. Chem. A 2005, 109, 816–
824.