excellent results (entry 10), whereas chloroacetic acid (CA)
was an ineffective promoter under the same conditions (entry
12). Smaller excesses of BH3‚pyridine and DCA gave
comparable results in the case of substrate 2a (entry 11),
but larger excesses often proved helpful on some of the more
demanding substrates described below. On the basis of the
results in Table 1, BH3‚pyridine in the presence of DCA was
adopted as the reducing system of choice, as described in
more detail below. The results with BH3‚pyridine and various
carboxylic acids suggest that a minimum acid strength is
necessary to promote oxime reduction by BH3‚pyridine. This
effect is consistent with findings reported by Kikugawa,9
whereby aldehydes were transformed to symmetrical ethers
with BH3‚pyridine in trifluoroacetic acid but not in acetic
acid. Interestingly, dichloroacetic acid also was found to be
a better promoter than acetic acid of sodium cyanoboro-
hydride mediated reduction (Table 1, entry 1 vs entry 14),
although conditions for application of this reducing system
were not optimized.
The scope of the BH3‚pyridine/DCA system with respect
to aldehydes and ketones was examined. First, the conditions
for oxime formation using a variety of ketones and aldehydes
were reexamined, and it was observed that for several
ketones, use of HOAc in TMOF/1,2-dichloroethane (DCE)
gave somewhat better results compared to the conditions
previously used for benzaldehyde (TMOF/THF without
HOAc). Thus, double treatment of 1a with 10 equiv of
aldehyde or ketone (0.5 M) and 3.5 equiv of HOAc (0.2 M)
in 1.3:1 DCE/TMOF for 16-18 h at room temperature was
adopted as a standard protocol for oxime formation. Table
2 summarizes the results from application of this procedure
benzophenones and R-branched acetophenones will likely
require more forcing conditions to induce oxime formation.
The oximes that formed successfully were then subjected
to one or two cycles of reduction with BH3‚pyridine (15
equiv)/DCA (22 equiv) in CH2Cl2 at room temperature for
16-18 h, with the product mixture subjected to benzoylation
followed by cleavage as described above. The results (Table
3) indicate that the majority of cases produce the desired
Table 3. Results for Conversion of 1a to Hydroxamates 5
According to Scheme 1 Using the Standard Protocols (see text)
and Aldehydes or Ketones R1(CdO)R2 a
% unreacted % hydroxamic
aldehyde or ketone
benzaldehyde
p-methoxybenzaldehyde
p-trifluoromethylbenzaldehyde
3-phenylpropionaldehydeb
acetophenone
p-nitroacetophenoneb
N-benzyl-4-piperidinone
benzophenone imine
3-pentanoneb
oxime (2)
acid (5)
<5
<5
21
<5
<5
55
<5
40
ND
>95
>95
79
>95
80
24
90
15
83 (compd 6)
a Data estimated from relative HPLC peak areas monitored at 215 nm.
b Subjected to double reduction (see text).
N-substituted benzohydroxamic acid in good to excellent
purities. However, electron-withdrawing substituents on the
aromatic ring of a starting benzaldehyde or acetophenone
appear to inhibit oxime reduction, and reduction also is quite
sluggish for the benzophenone-derived oxime.
Table 2. Results for Conversion of 1a to Oximes Using the
Standard Protocol (see text) with Aldehydes and Ketones
R1(CdO)R2
To demonstrate the sequence on a preparative scale, 1a
(1.0 g, 1.4 mmol) was reductively alkylated with 3-pentanone
(two cycles of oxime reduction), followed by acylation with
benzoyl chloride and acidolytic cleavage under the conditions
described above. After evaporation of the volatile compo-
nents, the residue (83% pure by HPLC at 215 nm; unacylated
extent of
oxime formationa
aldehyde or ketone (R1/R2)
Ph/H; p-MeO-Ph/H; p-CF3-Ph/H;
Ph-(CH2)2/H; p-NO2-Ph/Me; Et/Et;
ketone ) N-benzyl-4-piperidinone
Ph/Me; Ph/Ph (as imine)
>90%
1
precursor not detected by H NMR) was chromatographed
over silica gel to afford 6 (Scheme 1) in 60% yield and
>95% purity by HPLC monitored at 215 nm. 1H NMR (500
MHz) and high resolution mass spectral data were fully
consistent with the structure.10
70-90%
<20%
Ph/i-Pr; Ph/t-Bu; Ph/c-Pr; Ph/Ph; i-Pr/i-Pr
a Data estimated from HPLC peak areas monitored at 215 nm.
Acknowledgment. We thank Dr. R. M. Scarborough
(COR Therapeutics, S. San Francisco, CA) for valuable input
into this work.
to various aldehydes and ketones, including a number of
sterically or electronically demanding examples. Following
attempted oxime formation, resin samples were washed and
then subjected to benzoylation and cleavage conditions, as
described above, permitting detection of unreacted 1a as
benzohydroxamic acid (in some cases, accompanying hy-
drolysis to benzoic acid, presumably during cleavage, also
was observed). The results show that aliphatic and aromatic
aldehydes, unbranched acetophenones, and unbranched cyclic
and acyclic ketones are good or excellent substrates, whereas
Supporting Information Available: Experimental pro-
cedure for preparation of 6. This material is available free
OL006152G
(10) 1H NMR (500 MHz, CDCl3): δ 0.89 (t, 6H), 1.53 (m, 2H), 1.87
(m, 2H), 3.65 (m, 1H), 7.43-7.51 (m, 5H). High-resolution MS (FAB+):
calcd for C12H18NO2 (M + H) 208.1338; found 208.1337.
(9) Kikugawa, Y. Chem. Lett. 1979, 415-418.
Org. Lett., Vol. 2, No. 18, 2000
2779