M. Sasaki et al. / Carbohydrate Research 343 (2008) 848–854
853
TOC value of liquid product
inside of the reactor. The temperature difference
Liquid product yield ¼
between the inside and the outside of the reactor was
closely related to the feed rate and the reaction temper-
ature. At a feed rate of 0.5 mL/min, the difference
ranged from 7 to 9 °C. The maximum difference was
28 °C, which was observed at 400 °C with a feed rate
of 1.5 mL/min.
TOC value of feed solution
ꢁ 100
The reaction products in the liquid product were also
analyzed using an HPLC (Gilson HPLC system) with
a Shodex KS-801 column. The column oven tempera-
ture was 80 °C. The eluent was water at a flow rate of
0.5 mL/min. A refractive index (RI) detector (Shimadzu
RID-10A) provided quantitative analysis.
A 0.05 M (0.9 wt%) glucose aqueous solution was fed
into the reactor by a high performance liquid chromato-
graph pump (PU-2080 Plus, Jasco). The feed rate of
glucose solution ranged from 0.25 to 1.5 mL/min.
Another feeding system was installed for washing away
any water-soluble residue in the reactor. The reactor
volume and the feed rates of glucose solution are
known, and therefore, the space time in the reactor,
s (s) can be calculated by the following equation,
s = (V/Q)(q/q0), where V (mL) is the reactor volume
including the preheating section, Q (mL/s) is the flow
rate of glucose aqueous solution introduced, q (g/mL)
is the density of the reaction mixture at the reaction tem-
perature, and q0 (g/mL) is the density of glucose aque-
ous solution at the ambient temperature and pressure.
Both densities were assumed to be equal to the density
of pure water under same conditions because glucose
concentration was very low.
Prior to the thermochemical transformation of
glucose, water was fed into the preheater and heated
very rapidly to a preset temperature at which liquid
water changes into high-temperature steam. At a preset
temperature, the feed solution was changed from water
to glucose solution. The reaction products with high
vapor pressure were vaporized in the preheater and/or
reactor, and transferred to the cooling section through
the high-temperature steam. The steam containing the
vaporized products was cooled rapidly and condensed
in a collection bottle. The condensed product was
defined as ‘liquid product’. When the temperatures of
the preheating section and reactor reached steady state
conditions, the liquid product was collected in a collec-
tion bottle for 40 min. After the sampling, the feed
solution was changed to water with a feed rate of
5.0 mL/min for 15 min to terminate glucose transfor-
mation by rapidly decreasing the temperature. During
this procedure, the preheating section and reactor were
flushed out with hot water until all residual water-solu-
ble materials in non-vaporized products had been
displaced.
3.4. Isolation of AGP and AGF
A 0.05 M (0.9 wt%) glucose aqueous solution was con-
tinuously fed to the reactor for 8 h at 360 °C with a feed
rate of 1.0 mL/min. The liquid product obtained (482 g)
was frozen and then lyophilized to obtain a residue
(3.648 g). The ratio of AGP/AGF in the liquid product
was 2.1, determined using a HPLC (Gilson HPLC sys-
tem) with a KS-801 (Shodex) column. The residue was
purified twice by flash column chromatography on silica
gel (80 g; eluted with 2% MeOH/EtOAc, 5% MeOH/
EtOAc, and finally 10% MeOH/EtOAc) to afford crude
anhydroglucose mixture (2.271 g, 58.4%), that is, AGP
and AGF, as a pale brown syrup. The syrup was recrys-
tallized from i-propyl alcohol/EtOAc to yield pure AGP
(1.057 g, 27.2%) as colorless crystals and the mother
liquor (1.004 g, 25.8%, AGP/AGF (0.47:1)) as a pale
brown syrup. AGF and AGP in the mother liquor were
separated by a procedure21 involving the use of a cation-
exchange column in the calcium form. AGF was
obtained as the first eluting component as a pale yellow
syrup (0.663 g, 17.1% yield), and the second eluting
component contained AGP as a solid (0.298 g, 7.7%).
The combined AGP (1.355 g, 34.9%) was identical with
1
the authentic sample in all respects. H NMR and 13C
NMR spectral data of AGF were essentially identical
with those reported.24
Acknowledgments
The author, K.T., thanks the Hokuriku Industrial
Advancement Center (Kanazawa, Japan) for partial
financial support. Part of this study was supported by
the Industrial Technology Research Grant Program in
2006 from the New Energy and Industrial Technology
Development Organization (NEDO) of Japan.
3.3. Liquid product analysis
The liquid product collected was analyzed for total
organic carbon (TOC) using a TOC 5000A (Shimadzu).
The liquid product yield (carbon wt%) was evaluated
from the results of TOC analyses by the following
equation:
References
1. Radlein, D. In Fast pyrolysis of Biomass: A Handbook;
Bridgwater, A. V., Ed.; CPL Press: Newbury, UK, 2002;
Vol. 2, pp 205–241.