.
Angewandte
Communications
Substrates G and I are difficult to hydro-
genate and require longer reaction times. The
TOF of 160 for G (Table 2, entry 11) and 298 for I
(entry 13) compare well with the best reported
values under neutral conditions with catalysts 10
[
3c]
(
TOF = 62 in 16 h at 808C) and 3 (TOF = 314
[1b]
in 14 h at 1108C). No hydrogenation of methyl
oxalate (H) was observed with 12; this substrate
has not been studied with catalysts 1–10.
[8]
Activity and selectivity of complexes 12–14
were tested with substrates containing C=C
bonds. The results show significant differences
for ruthenium and osmium catalysts (Scheme 3).
Hydrogenation of a,b-unsaturated methyl 2-non-
enoate was not selective with 13 and afforded
nonanol, which is not surprising with regard to
the poor selectivity of 7 with methyl cinnama-
[
2a]
te. Unexpectedly, reduction of methyl 2-non-
enoate with 12 quantitatively afforded methyl
nonanoate. Such unusual selectivity and the lack
of further conversion to nonanol in this case are
puzzling. Osmium dimer 12 successfully cata-
lyzed reduction of methyl 3-nonenoate to 3-
nonenol, whereas ruthenium dimer 14 proved
inactive in this reaction. Methyl oleate was
hydrogenated with 12 with retention of the C=C
Scheme 3. Hydrogenation of alkenoates at 1008C and p(H )=50 bar, S/M=molar
bond to give (Z)-octadec-9-enol. The same
reaction was very sluggish and required more
catalyst with 12; the reaction also lacked selec-
tivity and afforded a mixture of octadecanol and
2
ratio of alkenoate groups to metal. [a] With tBuOK (0.5 mol%). [b] Conversion. [c] A
mixture of triglycerides of oleic (ca. 85%), linoleic (ca. 2–3%), and palmitic acids as
the main components in our samples. [d] Total yield of isolated alcohol mixture,
containing approximately 85% of oleyl alcohol.
(E)- and (Z)-octadec-9-enols.
We further tested osmium dimer 12 for
catalyst deactivation. None of the catalysts shown in
Scheme 1 have been tested with halogenated substrates,
except 6, which gave 88% conversion for methyl 4-chlor-
obenzoate by using 0.05 mol% catalyst with 10 mol%
reduction of glyceryl trioleate and a sample of domestic
olive oil, which is a natural mixture of triglycerides of oleic
(ca. 85%), linoleic (ca. 2–3%), and palmitic acids (main
components). Hydrogenation of the trioleate quantitatively
afforded neat (Z)-octadec-9-enol (oleyl alcohol) after aque-
[2b,c]
NaOMe at 1008C and p(H ) = 50 bar.
2
1
13
Complexes 12 and 14 gave 100% and 93% reduction for
methyl hexanoate in 2 and 1.5 hours, respectively. This
compares favourably with the reduction of methyl octanoate
with catalyst 6 (by using the same reaction temperature,
pressure, and catalyst loading), for which 86% conversion was
reported after 2.5 hours of reaction in the presence of
ous washing, based on comparison of the H and C NMR
spectra of the product with those of an authentic sample of the
alcohol. A similar product containing mostly oleyl (85%) and
palmityl alcohols was obtained after hydrogenation of olive
oil followed by aqueous washing. The reduction appeared
1
close to quantitative, although we noted that the H NMR
spectra of the product were missing the characteristic C11-
[
2b,c]
1
0 mol% NaOMe.
For ethyl acetate, the reduction with
1
2 was finished in three hours (Table 2, entry 8). The same
CH resonance of linoleyl alcohol (expected at 2.77 ppm in
2
reaction under p(H ) = 5 bar required a 20 times higher
CDCl ). Hydrogenation of esters of polyunsaturated acids
2
3
catalyst loading to reach 96% conversion in two hours with
will need to be studied further.
[
4a]
[1a]
5
and 86% conversion in 12 hours with 1, both at the
We finally tested the activity of the NNHPiPr catalysts for
dehydrogenative coupling of alcohols. Two complexes were
selected for this study, the ruthenium hydrido chloride 13 and
osmium dimer 12 (Table 3). We found the catalytic activity of
12 particularly impressive in ethanol and propanol at 788 and
968C respectively, as conversions of these alcohols to ethyl
acetate and propyl propionate with high yields have never
been observed below 1008C. The previously reported best
ADC catalyst 1 was inactive for ethanol and propanol and
required heating to more than 1008C to produce esters from
butanol and higher-boiling alcohols. The exceptional activity
of 12 for ester synthesis naturally correlates with the excellent
slightly higher temperatures of 105 and 1158C, respectively.
We further investigated hydrogenation of e-caprolactone,
a substrate that is known to undergo ring opening polymer-
[7]
ization (ROP) in the presence of tBuOK and similar bases.
Indeed, polymeric material was formed in an attempted
reduction of e-caprolactone with 13 in the presence of tBuOK.
When catalyzed by 14 without base, the hydrogenation
proceeded without polymerization, but stopped at 67%
conversion as if the catalyst was deactivated at that time.
Rapid reduction of e-caprolactone was achieved with osmium
dimer 12 (Table 2, entry 9).
2
774
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 2772 –2775