Basso et al.
Experimental Section
Compound (+)-4. Benzyl alcohol (3.7 mL, 36.0 mmol) was
added dropwise to a stirred suspension of exo-3,6-epoxy-1,2,3,6-
tetrahydrophthalic anhydride (2.0 g, 12.0 mmol) and quinine
(4.8 g, 13.2 mmol) in a 1:1 mixture of toluene and carbon
tetrachloride (240 mL) at -55 °C under nitrogen. The reaction
mixture was stirred at this temperature for 5 days and then
extracted with 2 N HCl (2 × 200 mL). The combined water
phase was then extracted with ethyl acetate (2 × 100 mL),
and the combined organics were anhydrified and dried under
vacuo. The crude material was purified via flash chromatog-
raphy (eluent EtOAc/PE 1:1 + 1% AcOH), yielding 3.1 g (95%)
of pure product. Rf ) 0.34 (eluent: DCM/Et2O ) 1:1, 1%
AcOH). Mp: 120-121 °C dec. [R]20D: +30.0 (c 1.00, MeOH).
1H NMR (300 MHz): δ 2.88 [2H, s]; 5.06 [1H, d, J 12]; 5.17
[1H, d, J 12]; 5.27 [1H, broad s]; 5.34 [1H, broad s]; 6.46 [1H,
dd, J 9, 1]; 6.48 [1H, dd, J 9, 1]; 7.30-7.35 [5H, s]. 13C NMR
(75 MHz): δ 46.9 (CH); 47.3 (CH); 67.2 (CH2); 80.4 (CH); 80.7
(CH); 128.4 (CH); 128.5 (CH); 128.6 (CH); 135.5 (C); 136.4
(CH); 136.8 (CH); 171.1 (C); 176.8 (C). Anal. Calcd for
C15H14O5: C, 65.69; H, 5.15. Found: C, 65.82; H, 5.28.
FIGURE 2. Low-energy conformation of the cyclic intermedi-
ate.
Among the various mechanisms proposed for these
multicomponent reactions,15 we argued that a stepwise
pathway, similar to that suggested by Ugi and others4,16
for the classical Ugi reaction, involving a six member
cyclic intermediate formed after condensation of the
aldehyde with 8, followed by displacement of the car-
boxylate by the isocyanide, could account for our experi-
mental results. To explain the complete stereoselectivity
of the reaction, preliminary conformational calculations17
have indicated that the preferred conformation of the
cyclic intermediate is the one illustrated in Figure 2, with
the N-benzyl group trans with respect to the aldehyde
side chain and cis with respect to the bicycle C-3
hydrogen. The isocyanide should attack this intermediate
from the side opposite to the carboxylic oxygen, thus
generating, in the case of (+)-8, an amino acid derivative
of the L-series.
Compound (-)-4 was obtained analogously by quinidine-
mediated asymmetrization. [R]20D: -29.7 (c 3.22, MeOH).
Compound (-)-6. Compound (+)-4 (1.3 g, 4.74 mmol) was
dissolved in dry THF (25 mL) under nitrogen, together with
triethylamine (1.3 mL 9.48 mmol). The solution was cooled to
-30 °C, and ethylchloroformate (0.7 mL, 7.11 mmol) was added
dropwise. The solution was allowed to warm to -10 °C, and
then sodium azide (0.77 g, 11.8 mmol) dissolved in water (5
mL) was added and the solution was let warm to room
temperature. The solution was then diluted with ethyl acetate
(25 mL) and washed with brine (2 × 20 mL). The organic phase
was dried over magnesium sulfate, concentrated in vacuo, and
redissolved in methylene chloride (30 mL). The resulting
solution, containing a small amount of triethylammonium
chloride, was put in the dark at 0 °C, and bromine (5.14 mmol,
0.26 mL) dissolved in methylene chloride (20 mL) was added
dropwise. The reaction was allowed to warm to room temper-
ature and stirred overnight. Solvent and excess bromine were
removed under vacuo, and crude material was taken up in
diethyl ether (40 mL) and anhydrified with sodium sulfate.
Filtration of sodium sulfate and triethylammonium chloride
and evaporation of the solvents yielded almost pure 5 as an
inseparable 1:1 mixture of diastereoisomers. Crude 5 was then
dissolved in toluene (40 mL) and heated at reflux for 30 min
until the release of nitrogen ceased. 4-Methoxybenzyl alcohol
(1.2 mL, 9.36 mmol) was then added, and the solution was
allowed to cool to room temperature overnight. The crude
material was partially crystallized, and the mother waters
were purified by flash chromatography (eluent: gradient from
EtOAc/PE 2:8 to EtOAc/PE 3:7). The solid and the product of
the chromatography were combined, yielding 2.1 g (3.62 mmol,
81%) of product as two diastereoisomers. This mixture was
finally dissolved in dry THF (40 mL) and cooled to 0 °C. A
substechiometric amount of TiCl4 dissolved in methylene
chloride (approximately 0.5 mL of a 1.8 M solution) and Zn
dust (0.71 g, 10.9 mmol) were added in this sequence. The
resulting suspension was stirred overnight, while three more
aliquots of catalyst were added at regular intervals. The solid
was then filtered over Celite and the crude material extracted
with 1 M HCl (20 mL) and brine (20 mL). The organic phase
was dried over magnesium sulfate, concentrated in vacuo, and
purified by flash chromatography (eluent: gradient from
EtOAc/PE 3:7 to EtOAc/PE 1:1) yielding 1.4 g (95%) of pure
product. Rf ) 0.49 (eluent: EtOAc/PE ) 1:1). Mp: 105 °C.
[R]20D: -4.40 (c 1.00, CHCl3). 1H NMR (300 MHz): δ 2.85 [1H,
d, J 8]; 3.78 [3H, s]; 4.29 [1H, dd, J 10, 8]; 4.74 [1H, s]; 4.93
[1H, d, J 12]; 4.94 [1H, d, J 12]; 5.01 [1H, d, J 12]; 5.12 [1H,
s]; 5.14 [1H, d, J 12]; 5.43 [1H, d, J 10]; 6.44 [2H, broad s];
6.85 [2H, d, J 8]; 7.20-7.40 [7H, m]. 13C NMR (75 MHz): δ
47.1 (CH); 52.7 (CH); 55.2 (CH3); 66.7 (CH2); 66.8 (CH2); 80.1
(CH); 83.9 (CH); 113.8 (CH); 128.3 (CH); 128.4 (CH); 128.5
Although this mechanism is in accordance with the
observed stereochemistry of the final compounds, more
detailed studies18 are required for better understanding
the reasons of the strong stereocontrol and to rule out
other possible hypotheses.
Conclusions
In conclusion, we have reported the synthesis and
application of a novel chiral auxiliary for the preparation
of optically pure R-amino acid derivatives via the Ugi
reaction. The availability of a large number of aldehydes
and of both enantiomers of 8, the mild conditions used
in the Ugi reaction and in the cleavage of the auxiliary,
make this methodology extremely useful for the prepara-
tion of a wide range of differently structured N-alkylated
and not alkylated L- and D-amino acid derivatives. Other
synthetic elaborations of intermediate 9 as well as the
use of a convertible isocyanide to obtain free R-amino
acids are under study in our laboratory, and the results
will be reported in due course.
(15) (a) Ugi, I.; Kaufhold, G. Liebigs Ann. Chem. 1967, 709, 11-28.
(b) Ugi, I.; Offermann, K.; Herlinger, H.; Marquarding, D. Liebigs Ann.
Chem. 1967, 709, 1-10.
(16) (a) Flanagan, D. M.; Joullie´, M. M. Synth. Commun. 1989, 19,
1-12. (b) Bock, H.; Ugi, I. J. Prakt. Chem. 1997, 339, 385-389.
(17) The four possible isomers of the postulated cyclic intermediate
were minimized using MM2 (ChemBats3D software) and the energies
compared, furnishing, as most stable isomer, the one reported in Figure
2.
(18) Attempts to prove the involvement of the cyclic postulated
intermediate were performed, but without success. When the bicyclic
amino acid 8 was mixed with benzaldehyde using MeOH-d4 as the
solvent, neither transient peaks nor any stable intermediate was
observed by NMR, even after several days; however, after addition of
the isocyanide, the reaction proceeded smoothly to give the desired
Ugi adduct. A parallel experiment was conducted mixing compound 8
with benzyl isocyanide, but also in this case no transformation was
observed until benzaldehyde was added to the mixture. These experi-
ments do not prove but do not exclude the involvement of the
postulated intermediate as metastable species in the reaction.
578 J. Org. Chem., Vol. 70, No. 2, 2005