.
Angewandte
Communications
OBn provided a similar outcome, furnishing the desired
major product 3ab in 65% yield and 92% ee (Table 2,
entry 13). Products 3 have been assigned an S configuration
by X-ray diffraction analysis of 3ja (Supporting Information,
Figure SI-1).[13] The structure of 3 is closely related with
phosphorus-chromones.[4d]
Furthermore, the results shown in entry 14 of Table 1
encouraged us to further screen various solvents in the
reaction of 1a and 2a with C13 as the catalyst. The
examination of solvent effects revealed that CH3CN was the
best solvent for this reaction, providing 4aa as the sole
product in 90% yield and 94% ee (Supporting Information,
Table SI-2, entries 1–8). Lowering the reaction temperature
to 08C or ꢀ158C and reducing the catalyst loading to
10 mol% did not improve the reaction outcomes (Table SI-
2, entries 9–11).
Product 4ab can be easily transformed into tetrahydro-
pyran 5ab as a single stereoisomer after hydrogenation with
Pd/C in THF and condensation with aniline in 73% yield and
96% ee in two steps. The crystal structure of 5ab has been
determined by X-ray diffraction and the ORTEP drawing is
shown in Figure SI-2 in the Supporting Information,[13] which
provides information on the absolute stereochemistry of
products 4 (S configuration) as well (Scheme 1). The structure
of 5ab is also very similar to that of phosphono-glycosyl
acetates.[3a]
Under the optimized conditions to produce 4aa as the sole
reaction product in good yield and high ee value, the reaction
scope was investigated by using various phosphonates 1 with
several allenic esters 2; the results of these experiments are
summarized in Table 3. As can been seen, the reactions
produced cycloadducts 4 in high yields and high enantiomeric
excesses for a wide range of substrates 1 with various aromatic
groups, heteroaromatic rings, alkyl or naphthyl groups
(Table 3, entries 1–12). When the ester moiety of the phos-
phonates was changed from OiPr to OMe or OtBu, the
reactions also proceeded smoothly to give the desired
products 4ma and 4na in 87% yield with 93% ee and 83%
yield with 95% ee, respectively (Table 3, entries 13 and 14).
Changing the ester moiety of allenic esters 2 from OEt to
OBn provided a similar outcome, affording the desired
product 4ab in 85% yield and 96% ee (Table 3, entry 15).
Scheme 1. Transformation of Product 4ab. Bn=benzyl, DMF=dime-
thylformamide, EDCI=N’-(ethylcarbonimidoyl)-N,N-dimethyl 1,3-pro-
panediamine, HOBT=hydroxybenzotriazole.
Based on the above experimental results and our previous
mechanistic studies on the formation of racemic cycloaddition
products,[9n] we have done theoretical investigations on
transition state (TS) models A–D (Scheme 2), which have
been proposed as the key transition states for the formation of
products 3 and 4. These TS structures (see Figure 2) were
optimized at the mPW1K/6-31G(d) level of theory, and their
relative energies were calculated at the mPW1K/6-31 +
G(2d)//mPW1K/6-31G(d) level of theory.[12] When C8 is
used in the reaction, the energy of key TS A, which leads to
product 3, and involves a hydrogen bond between the CONH
moiety in C8 and the carbonyl group and phosphonate moiety
in substrate 1, is lower than that of the TS B leading to
product 4 by 18.2 kJmolꢀ1. Thus, when C8 is used as the
catalyst, product 3 is the major product. When C13, which
lacks a hydrogen bond donor, is used as the catalyst, it cannot
form hydrogen bonds with the substrates. Furthermore, we
determined that the energy of the key TS C leading to product
3 is higher than that of the TS D leading to product 4 by
6.4 kJmolꢀ1, owing to the steric repulsion between the
naphthyl moiety of C13 and the phosphonate group of 1.
In conclusion, we have developed a novel asymmetric
[4+2] cycloaddition of b,g-unsaturated a-ketophosphonates
with allenic esters catalyzed by organocatalysts derived from
cinchona alkaloids, which affords the corresponding major
adducts (phosphonate-substituted functionalized pyran or
dihydropyran derivatives) in high yields and enantioselectiv-
ities. Calculations indicate that stereocontrol is determined by
hydrogen-bonding effects or steric interactions. Cinchona-
alkaloid-derived organocatalysts which possess either a hydro-
gen bond donor or a sterically bulky group can give different
cycloadducts. These reactions differ from previous [4+2]
cycloadditions through steric interactions;[10b,c] herein, we first
report that different types of cycloadducts can be afforded
through hydrogen-bonding interactions or pure steric repul-
sion. These corresponding phosphonate-substituted function-
Table 3: Substrate scope of the C13-catalyzed asymmetric [4+2] cycli-
zation of phosphonates 1 and 2.[a]
Entry R1
R2
R3
Product
ee of 4
(% yield)[%][b] [%][c]
1
2
3
4
5
6
7
8
C6H5 (1a)
Et (2a)
Et (2a)
Et (2a)
Et (2a)
Et (2a)
Et (2a)
Et (2a)
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
iPr
Me
4aa (90)
4ba (92)
4ca (91)
4da (93)
4ea (90)
4 fa (88)
4ga (89)
4ha (91)
4ia (85)
4ja (87)
4ka (87)
4la (85)
4ma (87)
94
92
91
95
95
94
95
93
90
94
94
94
93
95
96
4-ClC6H4 (1b)
4-BrC6H4 (1c)
4-MeC6H4 (1d)
4-MeOC6H4 (1e)
3-MeOC6H4 (1 f)
3-MeC6H4 (1g)
2-MeOC6H4 (1h) Et (2a)
9
2-furyl (1i)
1-naphthyl (1j)
Me (1k)
n-propyl (1l)
C6H5 (1m)
C6H5 (1n)
C6H5 (1a)
Et (2a)
Et (2a)
Et (2a)
Et (2a)
Et (2a)
Et (2a)
10
11
12
13
14
15
tBu 4na (83)
4ab (85)
Bn (2b) iPr
[a] All reactions were carried out using 1 (0.10 mmol) and 2 (0.12 mmol)
in CH3CN (2.00 mL) for 24 h. [b] Yield of isolated product. [c] Deter-
mined by HPLC analysis on a chiral stationary phase.
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 11328 –11332