Journal of the American Chemical Society
Communication
We evaluated the aldol reaction with the various dialdehyde
1
compounds in Figure 4 and monitored by H NMR in D2O
1
Figure 2. (a) Cartoons of complexes A and B. (b) Partial H NMR
spectrum of 2a@cavitand 1. The peaks labeled with red rectangles are
from the dynamic complex A, and the peaks labeled with blue
numbers are from the more static complex B.
“yo-yo like” between two J-shaped conformations. Complex B
is the complex of dialdehyde monohydrate 2a with the
cavitand 1, and the guest assumes a folded arrangement that is
more static. Earlier studies indicate that a typical methylene
group fixed at the bottom of the cavity appears at −2.7
ppm,23,29,32 and complex B features such upfield signals. The
ratio of complex A to B is about 0.6 to 0.4 based on 1H NMR
integration (Figure S2). This is unremarkable, as hydration
equilibria for simple aliphatic aldehydes in water are typically
near unity.33,34 The dihydrate is also expected to be present,
but it is soluble enough in water (D2O) to remain in bulk
solvent as shown by NMR signals between 1.5 to 2 ppm.
We used cavitand 1 as a chaperone to form a 12-membered
ring, cyclododecene-1-aldehyde 3b, from the corresponding
C13 dialdehyde 2b (Figure 3). The complex was prepared by
Figure 4. Scope of selective intramolecular aldol condensations in
cavitand 1. Yields are isolated, and conversions are based on the ratio
1
of the substrates and products from H NMR after extraction of the
reaction solution. a Reaction was conducted with 50 mg of 2c.
chain lengths proceeded smoothly (2b−2f), giving the desired
products in good yields. Dialdehydes bearing heteroatoms also
afford the desired macrocycles in moderate to good yields
(2g−2j). When the reaction was conducted with 50 mg of 2c,
the yield (75%) of 3c decreased slightly.
The chain length affects the rate of reaction. The reaction
rates forming 12-membered macrocycles were slow (2b, 2g,
2h), and dodecanedial (2a) did not react. These medium-sized
rings show large transannular strains, and perhaps the two
aldehydes are too deep in the cavity to interact with the acid
and base catalysts in bulk solution. Stoichiometric amounts of
cavitand 1 are required for the macrocyclization, but the
cavitand can be recovered and reused with no effect on the
reaction outcomes. Since direct competition experiments
showed macrocycle 3c to be a better guest than dialdehyde
Allyl deuterated product (55% D incorporation) was observed
when 2c was treated in D2O under standard conditions. After
further experimentation, higher deuteration (>95%) was
achieved by increasing the amount pyrrolidine from 1 to 3
equiv (Figure 5).35
Figure 3. Selective intramolecular aldol condensation in the cavitand.
Conversion based on the ratio of the substrates and products from 1H
NMR after extraction of the mixture. The yield of 3b was 71% while 4
was not detected.
stirring the dialdehyde and cavitand 1 for 2 h (2.0 mM),
followed by treatment with 1 equiv each of pyrrolidine,
propylamine, acetic acid, and triethylamine at 37 °C,
conditions based on precedents from Gellman’s studies.8 The
1
reaction was monitored by H NMR spectroscopy (in D2O,
In summary, intramolecular aldol condensation of dialde-
hydes for macrocyclization can be achieved in a cavitand
obtained in 71% isolated yield after 20 h. The cyclodimer 4
was not detected. In contrast, when the reaction was
performed in 4% H2O, 96% isopropanol in the absence of
cavitand 1, and under the same conditions, the desired
compound 3b was not obtained. Instead, the cyclodimer 4
(20% yield) was isolated. The same reaction was performed in
H2O in the absence of cavitand 1 and under the same
conditions, guaranteeing high dilution. The cyclic compound
3b or 4 was not detected (83% conv); instead, oligomers were
formed.
Figure 5. Formation of deuterated product.
2191
J. Am. Chem. Soc. 2021, 143, 2190−2193