Organic Letters
Letter
phenylpropanol (1k) and n-hexanol (1l) gave the desired
products in excellent yields. However, benzyl alcohol (1m)
suffered from the required compromise between reactivity and
selectivity, and the desired product 3m was obtained in 75%
yield. In contrast, alcohol 1n, bearing a primary alkyl chloride,
gave the corresponding product 3n in good yield while
retaining the potentially electrophilic chloride group. Polyether
1o and alkene 1p were also tolerated under the reaction
conditions, and 3o and 3p, respectively, were obtained in high
yields.
an initial investigation with a tertiary alcohol also suffered from
low reactivity, which led us to explore the use of more active
Zn catalysts for more demanding substrates. After extensive
studies, we finally identified sterically hindered bis(2,2,6,6-
tetramethyl-3,5-heptanedionato)zinc(II) (Zn(TMHD)2) as
the most active catalyst. With the new active catalyst in
hand, we investigated the scope of the reaction with
challenging alcohols (Scheme 3). To our delight, a sterically
Scheme 3. Substrate Scope of the Reaction with Demanding
Alcohols
We then explored the potential of our methodology by using
carbohydrates as substrates (Scheme 2). First, 1-O-methyl-2,3-
Scheme 2. Substrate Scope of the Reaction with
Carbohydrate Derivatives
hindered tertiary alcohol gave 3x in quantitative yield. Acyclic
tertiary alcohol 1y also gave the desired product 3y in good
yield. Unexpectedly, this catalyst possessed extraordinary
functional group tolerance. Using steroid 1z, which features a
free phenol group, the secondary alcohol reacted selectively to
smoothly afford 3z in excellent yield. Substrates with
carbamate and ester groups were also tolerated, with CbZ-
protected serine ester 1aa giving the desired H-phosphonate
diester 3aa in high yield. Notably, in this case, the carbamate
moiety was tolerated in this reaction system.
As described previously, the product 3g could be easily
oxidized to biologically important trialkyl phosphate 4g in high
yield by a slightly modified, reported procedure (Scheme 4).38
This result suggests that combining the currently described
method with the oxidation protocol would allow access to a
wide variety of trialkyl phosphates with high efficiency.39
a
b
Reaction was performed at 0 °C. Dimethyl phosphite (3 equiv) was
used.
acetonide-protected ribose was employed under the standard
reaction conditions. To our delight, the desired product 3q was
obtained in high yield. The protecting group at the 2,3-
positions could be changed to Bn groups without significant
loss of reactivity, and 3r was obtained in good yield.
Encouraged by the successful results of the ribose substrates,
we next investigated 2-deoxyribose substrates. 5-DMTr-
protected 2-deoxyriboses 1s and 1t worked well, and the
target H-phosphonate diesters 3s and 3t, respectively, were
obtained in good yields. The stereocenter at the 1-position (α-
and β-) did not affect the reactivity or selectivity. Similarly, 3-
BnO-protected 2-deoxyriboses 1u and 1v afforded 5-
phosphonylated compounds 3u and 3v in good yields. In
addition to pentose derivatives, protected glucose 1w also gave
the corresponding H-phosphonate diester 3w in high yield.
Apart from primary and secondary alcohols, to our
knowledge, there has been no successful example of the use
of tertiary alcohols in reported catalytic systems. In our study,
Scheme 4. Oxidation of H-Phosphonate
In conclusion, we have developed a highly active and
selective Zn-catalyzed phosphonylation of alcohols with
dimethyl phosphite. The reaction proceeded smoothly even
at 0 °C to give the target H-phosphonate diesters in high to
excellent yields. Various alcohols including primary, secondary,
and even tertiary alcohols of acyclic and cyclic structures,
carbohydrates, steroids, and amino acids reacted smoothly with
excellent functional group tolerance, and interestingly, a
sterically hindered Zn complex demonstrated the best activity
C
Org. Lett. XXXX, XXX, XXX−XXX