CHEMCATCHEM
FULL PAPERS
catalyst at 463 K under an H pressure of 0.7–0.9 MPa (absolute
the reactor was cooled down to RT by blowing air. The reaction
mixture was separated to liquid and solid phase by centrifugation.
The cellulose conversions obtained with the mix-milled samples or
soluble acids were also examined by the same procedure.
2
pressure at room temperature). The catalyst was durable at
least 3 times in the reuse experiments. The time course of cel-
lulose conversion indicates that the reaction takes place
through 4 steps: 1) hydrolysis of cellulose to glucose, 2) hydro-
genation of glucose to sugar alcohols, 3) glucose decomposi-
tion, and 4) degradation of sugar alcohols. The kinetic study re-
vealed that the competition between steps 2 and 3 limits the
yield of sugar alcohols at 70%, and that step 4 further decreas-
es the yield to less than 40%. Step 4 is promoted by the Ru
Water-soluble products in the liquid phase were analyzed by HPLC
(Shimadzu LC10ATVP, refractive-index detector) with a Rezex RPM–
Monosaccharide Pb+ + (8%) column (diameter 7.8ꢂ300 mm,
ꢀ1
eluent: water 0.6 mLmin , 343 K) and a Shodex Sugar SH-1011
ꢀ1
column (diameter 8ꢂ300 mm, eluent: water 0.5 mLmin , 323 K).
Products in the gas phase were quantified by gas chromatography
(GC, Shimadzu GC-8A, thermal conductivity detector) with an
catalyst under H pressure, and sugar alcohols are gradually
2
active carbon column (diameter=3ꢂ2 m, mobile phase: He
ꢀ
1
degraded over a long reaction time. The mix-milling method
increases the hydrolysis rate of cellulose, and the reaction time
can be shortened from 24 h to 3 h to reach 90% conversion.
As a result, the influence of step 4 was significantly reduced to
increase the yield of sugar alcohols to 68%, which is the high-
est ever achieved by using only solid catalysts in water under
30 mLmin , 403 K). Carbon-based yields of the liquid- and gas-
phase products were calculated with Equation (6). The conversion
of cellulose was determined by the weight difference of dried cel-
lulose before and after the reaction as shown in Equation (7). The
selectivity of product was calculated based on carbon using Equa-
tion (8).
low-pressure H . Several physicochemical analyses of the cata-
2
molescarbon in product
Yield ½% Cꢁ ¼ moles
ꢅ 100
ð6Þ
ð7Þ
ð8Þ
lyst have indicated that the small RuO ·2H O particles (1–2 nm)
2
2
carbon in charged cellulose
are quickly reduced to metal under the reaction conditions,
and this species works as the actual catalyst. These results
demonstrate that the combination of the highly dispersed Ru
catalyst and the mix-milling method enables the high-yielding
and selective production of sugar alcohols from cellulose
under low-pressure H2.
massunreacted cellulose inc: oligomers
Conversion ½%ꢁ ¼ ð1ꢀ
Þ ꢅ 100
masscharged cellulose
yieldproduct
Selectivity ½% Cꢁ ¼
ꢅ 100
conversioncellulose
Characterization of the catalyst
Experimental Section
N2 adsorption–desorption measurements were operated at 77 K
Preparation of catalyst
(
0.050ꢆp/p ꢆ0.995) by using a Belsorp mini II (BEL Japan). The
0
Supported Ru catalysts were prepared by a conventional impreg-
nation method. Typically, amounts of 1 g of AC(N) (activated
carbon Norit SX Ultra, purchased from Aldrich) was dispersed into
catalysts were pretreated under vacuum (<10 Pa) at 393 K for 4 h
to remove adsorbed water and gasses.
TEM was performed with a JEM-2100F (JEOL) at an acceleration
voltage of 200 kV. Samples were dispersed on a copper grid with
ethanol after ultrasonic pretreatment.
aqueous RuCl solution (25 mL, 8.1 mM) with stirring at RT for 16 h.
3
The water solvent was removed at 333 K under 85 hPa and then at
RT under <1 Pa for 18 h. The dried powder was reduced under H
2
ꢀ1
(
30 mLmin ) at 673 K for 2 h and passivated by diluted O (10%)
2
Ru K-edge XAFS was measured at NW10 A beam line on KEK-PF
in He at RT. The black powder was exposed to air at RT, giving
wt% Ru/AC(N).
(
Proposal No. 2010G591). The in situ measurements were per-
2
formed as follows: cellobiose (205 mg), catalyst (150 mg), and 2-
propanol/water (0.6:1.8 mL) were charged into a high-pressure
PEEK cell (internal volume of 3.5 cm ) covered with a metal frame
3
Hydrolytic hydrogenation of cellulose
[34]
(SUS304) having small windows. The reactor was purged with N2
and then heated to 413 K by two cartridge heaters, and then the
quick XAFS spectra were recorded every 0.5 min with a synchrotron
radiation (ring energy 6.5 GeV, 50 mA) through a Si(331) double-
crystal monochromator in the transmission mode. The EXAFS spec-
tra were analyzed by using REX2000 software (Rigaku) with
a spline smoothing method in the range of the wave vector k=
Cellulose was ball-milled with ZrO balls (1 cm, 1 kg) in a ceramic
bottle (900 mL) at 60 rpm for 4 d to decrease the crystallinity. The
2
degree of crystallinity (CrI) was reduced from 81% to 14%, as de-
[17]
termined by XRD. The crystalline cellulose (10.0 g) and 2 wt%
Ru/AC(N) (1.54 g) were also milled together in a ceramic pot (3.6 L)
with alumina balls (1.5 cm, 2 kg) at 60 rpm for 4 d to improve the
ꢀ
1
3
[31]
3
–15 ꢁ (k -weighted). The backscattering amplitude and the
contact. The amount of adsorbed water in the cellulose was esti-
mated from a weight loss after drying at 393 K under vacuum
phase shift of Ru–Ru coordination were extracted from the EXAFS
spectrum of Ru metal, which were applied to the curve fittings to
determine the local structure of the catalyst.
(<10 Pa) and/or by a total organic carbon analyzer (TOC, Shimadzu
SSM-5000A) to determine the carbon content of cellulose.
The hydrolytic hydrogenation of cellulose was performed as fol-
lows: ball-milled cellulose (324 mg), 2 wt% Ru/AC(N) (50 mg), and
H O (40 mL) were transferred into a high-pressure reactor (Hastel-
2
Acknowledgements
loy, OM Lab-tech MMJ-100, 100 mL). The reactor was purged with
H for 3 times and pressurized with H (0.7–0.9 MPa, absolute pres-
sure at RT), then heated to 463 K and kept for a certain time with
stirring at 600 rpm. The time at which 463 K was reached was de-
fined as the start (t=0 h) of the time course. After the reaction,
This work was supported by a Grant-in-Aid for Scientific Research
2
2
(
KAKENHI, 20226016) and for JSPS Fellow (KAKENHI, 11J03322)
from the Japan Society for the Promotion of Science (JSPS). T.K.
was a JSPS Research Fellow (DC2).
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 230 – 236 235