CHEMSUSCHEM
COMMUNICATIONS
sidering that HMF remained unaltered in the CO2/H2O system
(without Pd/C and hydrogen), one may suspect that the first
step is the hydrogenation of HMF to DHMF. Then, inspired by
a previous mechanism reported for the production of levulinic
acid from HMF,[9] the reaction mechanism may proceed here
through (1) the acid-catalyzed rehydration of DHMF, followed
by (2) a ring opening, and (3) a subsequent hydrogenation of
the in situ-produced C=C bond yielding HHD.
acid-catalyzed rehydration of DHMF, and finally (5) hydrogena-
tion to HHD. Considering that five consecutive steps are neces-
sary to convert inulin to HHD with 15% yield, one may deduce
that the average yield corresponding to each elementary step
is about 70% showing the rather selective efficiency of the
process.
We also focused on the use of glucose; a key raw material.
Using an identical CO2/H2O system for converting fructose as
aforementioned, no formation of HHD was observed. This
result is in agreement with previous works, as production of
furanic derivatives from glucose requires first an isomerization
of glucose to fructose which cannot be catalyzed by carbonic
acid. In such case, it was possible to produce HHD from glu-
cose in a two-step process involving (1) conversion of glucose
to HMF using our previously reported route (conversion of glu-
cose to HMF with 70% yield in an aqueous phase containing
a catalytic amount of AlCl3),[11] followed by (2) the conversion
of HMF to HHD. Using this two-step process, an overall yield of
HHD of 50% was obtained from glucose. Note that all at-
tempts to produce HHD from glucose by combining Pd/C, hy-
drogen, and AlCl3 in a single reactor failed, mostly because pH
required for the conversion of glucose to HMF and HMF to
HHD are not compatible.
In a last set of experiments, we evaluated the direct conver-
sion of fructose to HHD in a single reactor. To the best of our
knowledge, production of HHD from fructose has not been re-
ported before. Under our optimized conditions (7.5 wt% of
Pd/C and a pressure of hydrogen and CO2 of 10 bar and
30 bar, respectively), HHD was produced in a low yield (<13%)
although the conversion of fructose was complete. Hence, we
investigated the direct conversion of fructose to HHD in
a single reactor but in a sequential manner. Buxing Han et al.
and our group have previously reported that the pKa of car-
bonic acid is sufficient to catalyze the dehydration of fructose
to HMF.[10] Keeping these results in mind, fructose was first
heated in water at 1508C under 40 bar of CO2, resulting in the
formation of HMF with a maximum yield of 50% (conv. fruc-
tose=70%) after 10 h of reaction at 1508C. Then, Pd/C
(7.5 wt%), and H2 (10 bar) were introduced into the reactor
(the pressure of CO2 and temperature were maintained at
40 bar and 1508C, respectively) without intermediate purifica-
tion of the aqueous phase containing HMF. In agreement with
the results described above, HMF was converted into HHD in
72% yield (i.e., 36% yield from fructose; Scheme 6). In addi-
tion, the reaction also proceeded well starting from inulin,
a natural biopolymer of fructose. In particular, after heating
inulin in the CO2/H2O system at 1508C and 40 bar, HMF was
produced with 30% yield. Introduction of Pd/C (7.5 wt%) CO2
(40 bar) and hydrogen (10 bar) into the reactor led to the pro-
duction of HHD with 50% yield from HMF (i.e. 15% yield from
inulin) after heating the reactor at 1208C for 10 h (Scheme 6).
In such case, the catalytic system was capable of promoting in
a single reactor (1) the acid-catalyzed hydrolysis of inulin to
fructose, followed by (2) the acid catalyzed dehydration of fruc-
tose to HMF, (3) the hydrogenation of HMF to DHMF, (4) the
In summary, we show that Pd/C mediated hydrogenation of
inulin, fructose, and HMF in water and under pressure of CO2
opens a route to valuable diketone derivatives in a one-pot
process. Under optimized conditions, five reaction steps con-
secutively and selectively take place in a single reactor, with an
average yield of 70% for each step in the case of inulin. The
reactions show a carbon economy of 100%, employ only
water as a solvent, and require rather low hydrogen pressures
(10 bar). On the other hand, the ability of CO2 to produce car-
bonic acid in the presence of water not only facilitates the
work-up procedure but also offers a solution to the deactiva-
tion of acid solid catalyst, often encountered in the presence
of water. In addition, by changing the nature of the solvent, it
is possible to switch the selectivity of the reaction either to
DMF or HHD, which represents another noticeable advantage.
We do believe that the concept of using CO2 to generate acidi-
ty during hydrogenation reactions is of wide interest because
it allows complexity and diversity to be created from carbohy-
drates, with a high eco-efficiency.
Experimental Section
Analytical methods: Gas chromatography analyses were per-
formed on a Bruker GC-456 instrument equipped with an on-
column injector (2508C), an FID detector (3258C) and an HP-5 ms
column (30 mꢃ0.25 mmꢃ0.25 mm). NMR spectra were recorded on
1
Bruker ADVANCE DPX 400 spectrometers at 400.13 MHz for H and
100.6 MHz for 13C. Data for HHD, DMF, HDX, HMF, LA, DHMF, and
DHMTHF are already described in the existing literature, and were
used to ascertain the formation of targeted products.
1
Selected data for HHD: H NMR (400 MHz, CDCl3): d=2.16 (s, 3H,
ÀCH3), 2.59 (t, J=6.4 Hz, 2H, ÀCH2), 2.80 (t, J=6.4 Hz, 2H, ÀCH2),
3.12 (bs, 1H, ÀOH), 4.29 ppm (s, 2H, ÀCH2); 13C NMR (100 MHz,
CDCl3): d=29.7 (ÀCH3), 31.7 (ÀCH2), 36.8 (ÀCH2), 68.2 (ÀCH2OH);
LC/MS (ESI), m/z=132.1 uma.
Scheme 6. Catalytic conversion of fructose and inulin to HHD.
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 0000, 00, 1 – 5
&
4
&
ÞÞ
These are not the final page numbers!