CHEMSUSCHEM
FULL PAPERS
Pd /Hb catalyst loading. Throughout these variations in reac-
This result is comparable to that obtained with the 20 wt%
catalyst loading after 8 h, which demonstrates the possibility
to decrease the catalyst loading. An interesting result of this
experiment is that we were able to observe oxygenates that
may be intermediates on the way to alkane formation. In the
8 h, 10 wt% loading experiment, we observed dodecane as
the main component of the reaction but also saw large
amounts of an oxygenate in the GC trace. Analysis by GC–MS
5
tion conditions, the selectivity to n-dodecane does not change.
NMR spectroscopy of the product gave further confirmation
that the product is dodecane. Additionally, we see no evidence
for fractionation of the carbon backbone of furoin. C , C , or
10
11
any other shorter chain length alkanes are not present in the
product mixture in any appreciable amount (Figure 1). If other
catalyst variations using different metals or supports were ex-
amined, in each case the results were poorer than that ach-
ꢀ
1
gave a molecular weight of 184 gmol for this component
with a base peak of m/z=85 characteristic of the MeTHF frag-
ment. Therefore, the only feasible structure for this is 2-heptyl-
5-methyltetrahydrofuran. This is somewhat surprising if it is
truly an intermediate if we consider our proposed reaction
pathway in which a THF-type ring is unlikely to undergo
a ring-opening reaction.
ieved with the Pd /Hb catalyst. Finally, we demonstrated that
5
2
308C is the necessary temperature to yield alkane product
successfully, as very low alkane yields were observed at 2208C
Table 2, entry 8). Interestingly, we found that no alkane was
(
produced if the HDO batch reaction is run in water or if Pd /C
5
is used in conjunction with zeolite-b.
The catalyst material, which is a very fine powder, may be
recovered from the reaction mixture by centrifugation. Direct
recycling of the catalyst in this manner is not successful in the
generation of alkane products. However, some catalytic activity
Table 2. Selected results for the conversion of Me
2
-furoin to dodecane.
[
a]
Entry
Catalyst
Catalyst loading Alkane yield n-Dodecane
ꢀ
1
is retained, and the predominant product is the 198 gmol
[wt%]
[%]
selectivity [%]
product, which we identified previously as 1,2-bis(5-methylte-
trahydrofuran-2-yl)ethane. The catalyst retains this reactivity for
more recycles. It is a little peculiar that the catalyst retains
some activity, yet does not yield alkane. We believe that this
partial loss in reactivity is because of the observed charring,
which results in coke formation on the catalyst and thus
blocks facile access to acidic sites. As mentioned earlier, the
acid-catalyzed ring-opening should be the first step in our cat-
alytic process (Path A, Scheme 3); we have accomplished this
with the fresh catalyst. However, in the case of the recycled
catalyst, it appears as though hydrogenation of the furan rings
has become the favored first step (Path B, Scheme 3), thus we
become stuck with the THF rings and are unable to complete
the HDO.
1
Pd
Pd
Pd
Pd
Pd
Pd
Pd
Pd
Pd
Ru
5
5
5
5
5
5
5
5
5
5
/Hb
/Hb
/Hb
/Hb
/Hb
/Hb
/C+Hb
/Hb
/HZSM-5
/Hb
10
10
20
20
20
27
66
70
76
60
76
0
6
8
27
37
0
87
93
94
94
83
95
–
95
95
67
98
–
[
b]
2
3
4
5
6
7
8
9
1
[
[
c]
d]
100
20
20
20
20
20
20
[
e]
[
f]
[
g]
0
1
1
Ni
5
/Hb
/Al
1
2
Pd
5
2 3
O
2
[a] All entries have an initial H pressure of 60 bar and are run at 2308C
for 8 h in ethanol solvent with 200 mg of substrate unless otherwise indi-
cated. In all cases, GC analysis shows 100% conversion of starting materi-
al. [b] The reaction time is 16 h. [c] Large-scale reaction with 500 mg sub-
strate to facilitate easier reactivation of catalyst. [d] Catalyst recovered
and reactivated from entry 4. [e] The loading for each catalyst was
The issue now becomes whether or not the catalyst can be
reactivated. After the catalyst was recovered by centrifugation,
it was loaded into a U-shaped drying tube, dried under flowing
2
1
0 wt%. [f] The reaction temperature was 2208C. [g] Dodecane yield:
9%, undecane yield: 4%, decane yield: 4%.
air, calcined at 4508C, and reduced under flowing H at 4508C.
2
This ‘reactivated’ catalyst was successful to transform Me2-
furoin to alkane, albeit in a lower yield and selectivity com-
pared to the fresh catalyst (Table 2, entries 4 and 5) Unfortu-
nately, after a second recovery followed by the same reactiva-
tion process, the catalyst material did not have activity toward
alkane production but it did produce various oxygenates, simi-
lar to that observed for the recovered catalyst without the re-
activation step. Work is ongoing in this area to improve the re-
usability of the catalyst material.
Our results here are unique and in contrast to similar reac-
tions in which the yield of diesel-range alkanes might be high
[15]
but the selectivity toward the theoretical product is low. Fur-
thermore, our system is one of a handful of dual-function cata-
lysts for the HDO of furfural derivatives and only the second
[
15]
one applied to furoins. Other studies on the HDO of furoin
have used either Pt/CsH PW O , which gave low yields of
2
12 40
alkane, or a hydrogenation catalyst with an added acid cocata-
[
10,15b]
lyst.
As a result of the high catalyst loading used in this HDO re-
action, we wanted to investigate both lower loadings and recy-
Conclusions
cling of the recovered catalyst using our Pd /Hb catalyst. The
We have presented herein new approaches for the synthesis of
furoins from biomass-derived furfurals using N-heterocyclic car-
bene organocatalysis. Our methods are green, use renewable
solvents, and allow the direct utilization of furanic aldehydes
synthesized from biomass with yields that approach 90% for
furfural as a substrate. We have explored a number of routes
for the conversion of furoins to hydrocarbon fuels through
5
use of 5 wt% catalyst gave significantly reduced activity, which
resulted in only 3% dodecane yield after 8 h at 2308C and
6
0 bar H . Going up to 10 wt% loading gave a moderate 27%
2
yield of dodecane under the same conditions. If we extended
the reaction time to 16 h with the 10 wt% loading, we ob-
tained 66% total C12 yield with 93% selectivity to n-dodecane.
ꢁ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2742 – 2747 2746