Unravelling the Ru-Catalyzed Hydrogenolysis of Biomass-Based Polyols under Neutral and Acidic Conditions

Article abstract of DOI:10.1002/cssc.201500493

The aqueous Ru/C-catalyzed hydrogenolysis of biomass-based polyols such as erythritol, xylitol, sorbitol, and cellobitol is studied under neutral and acidic conditions. For the first time, the complete product spectrum of C₂-C₆ polyols is identified and, based on a thorough analysis of the reaction mixtures, a comprehensive reaction mechanism is proposed, which consists of (de)hydrogenation, epimerization, decarbonylation, and deoxygenation reactions. The data reveal that the Ru-catalyzed deoxygenation reaction is highly selective for the cleavage of terminal hydroxyl groups. Changing from neutral to acidic conditions suppresses decarbonylation, consequently increasing the selectivity towards deoxygenation.

Full text of DOI:10.1002/cssc.201500493

DOI: 10.1002/cssc.201500493
Full Papers
Unravelling the Ru-Catalyzed Hydrogenolysis of Biomass-
Based Polyols under Neutral and Acidic Conditions
[
a]
Peter J. C. Hausoul, Leila Negahdar, Kai Schute, and Regina Palkovits*
The aqueous Ru/C-catalyzed hydrogenolysis of biomass-based
polyols such as erythritol, xylitol, sorbitol, and cellobitol is stud-
ied under neutral and acidic conditions. For the first time, the
complete product spectrum of C –C polyols is identified and,
(de)hydrogenation, epimerization, decarbonylation, and deoxy-
genation reactions. The data reveal that the Ru-catalyzed deox-
ygenation reaction is highly selective for the cleavage of termi-
nal hydroxyl groups. Changing from neutral to acidic condi-
tions suppresses decarbonylation, consequently increasing the
selectivity towards deoxygenation.
2
6
based on a thorough analysis of the reaction mixtures, a com-
prehensive reaction mechanism is proposed, which consists of
Introduction
Cellulosic biomass is an abundant renewable resource, which
holds great promise for the sustainable production of fuels
demonstrates the high versatility of Ru-catalyzed hydrolytic hy-
drogenation and hydrogenolysis of cellulose. Yet improving
the selectivity towards target products remains a major chal-
lenge. Hence, it is imperative to gain a better understanding of
the role of the metal catalyst as well as the acid co-catalyst to
enable a rational catalyst design. Thus far, a limited number of
studies have provided insight into the reaction mechanism by
studying the resulting product mixture. Drawing on earlier
work concerning Cu, Montassier attributed the product mix-
ture obtained by the aqueous Ru-catalyzed hydrogenolysis of
polyols to the occurrence of (de)hydrogenation, dehydroxyla-
[
1–6]
and chemicals.
By employing selective depolymerization
methods, several defined platform chemicals can be obtained
from cellulose (Scheme 1). For example, acid-catalyzed hydroly-
[
24,25]
tion, retro-aldolization, and retro-Michael reactions.
Clear
evidence in support of the retro-aldol condensation reaction
was provided by Furney et al., who studied the Cu- and Ni-cat-
alyzed hydrogenolysis of 1,3-diol model compounds in the
Scheme 1. Hydrolysis and hydrolytic hydrogenation of cellulose.
[
26]
presence of 1m NaOH. Recently, Shanks et al. studied the
product mixtures of nine different polyols in the presence of
[
27]
sis yields glucose, which serves as precursor for platform mole-
Ru/C and CaO. In addition to the retro-aldol reaction, decar-
bonylation was proposed to account for the observed product
mixture. This suggests that two fundamentally different CÀC
cleavage reactions may occur under hydrogenolysis conditions.
It should, however, be noted that these studies were conduct-
ed under basic conditions. To the best of our knowledge, the
effect of acid on the product spectrum of the hydrogenolysis
of higher polyols has not been studied in detail. Here, we pres-
ent the results of the Ru/C-catalyzed hydrogenolysis of bio-
mass-based polyols under neutral and acidic conditions. The
product spectrum and reaction network under neutral condi-
tions is studied using erythritol, xylitol, and sorbitol. Subse-
quently, the influence of the addition of catalytic amounts of
silicotungstic acid (H [W SiO ]) on the product spectrum is
[7,8]
cules such as HMF and levulinic acid.
Alternatively, when
acid-catalyzed hydrolysis is combined with metal-catalyzed hy-
drogenation, sorbitol or hexitols are obtained directly as the
[
9–11]
main products.
Particularly Ru-based catalysts exhibit high
selectivities towards hexitols under relatively mild conditions.
Excellent hexitol yields are obtained for cellulose, which has
[
12–17]
been converted to cello-oligomers by acidic ball milling.
At elevated temperatures, hydrogenolytic CÀC and CÀO bond
breaking reactions increase and short-chain polyols such as
[18–21]
propylene and ethylene glycol are formed.
In addition, it
has been shown that the product spectrum can also be
[22,23]
steered towards light alkanes and hexane.
This clearly
4
12
40
[
a] Dr. P. J. C. Hausoul, Dr. L. Negahdar, K. Schute, Prof. Dr. R. Palkovits
Institut für Technische und Makromolekulare Chemie
RWTH Aachen University
Worringerweg 2, 52074 Aachen (Germany)
E-mail: palkovits@itmc.rwth-aachen.de
studied using sorbitol and cellobiitol (as model compound for
cello-oligomers). Based on these results and H/D exchange ex-
periments, a comprehensive mechanism is proposed for the
Ru/C-catalyzed hydrogenolysis reaction under acidic condi-
tions.
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201500493.
ChemSusChem 2015, 8, 3323 – 3330
3323
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 1. GC chromatograms (DB-23 column) of the peracetylated product mixtures of ERY (a), XYL (b), SOR (c), and RAM (d); 1-deoxygenated products are
boxed. Conditions: 2.0 g substrate, 400 mg Ru/C (5 wt%), 20 mL H O, 3 h, 6 MPa H , 423 K.
2 2
Results and Discussion
the chromatogram shown in Figure 1d. Interestingly, pentitols,
tetritols, glycerol (GLY), and ethyleneglycol (EG) are not ob-
served in the product spectrum of RAM. Apparently, a reductive
CÀO cleavage or deoxygenation reaction occurs selectively at
the terminal hydroxyl groups. Analogously, in the reaction of
XYL (Figure 1b), a set of four isomers of 1,2,3,4-pentanetetraols
(boxed) is observed next to ERY and THR. In the reaction of
ERY (Figure 1a), two isomers of 1,2,3-butanetriol (boxed) are
observed next to GLY. It is also noteworthy that all compounds
observed in the mixture of ERY are also observed in that of
XYL. The same holds for the product mixtures of XYL and SOR.
Thus, any additional peaks in a particular region can be attrib-
uted to compounds with an extended carbon chain. Based on
this and HPLC–electrospray ionization (ESI)–MS data, all major
groups of compounds could be identified. These data confirm
that Ru/C catalyzes stereoisomerization and CÀC and CÀO
bond-cleavage reactions. Below, these reactions pathways are
examined in more detail.
Product spectrum under neutral conditions
Ru/C-catalyzed hydrogenolysis of erythritol (ERY), xylitol (XYL),
and sorbitol (SOR) was performed in water at 423 K under
6
MPa H . Samples were taken periodically and peracetylated
2
prior to analysis by GC(–MS) and LC–MS. Figure 1 shows the
mixtures obtained after 3 h of reaction. In all cases the starting
polyol is accompanied by all of its possible stereoisomers [i.e.,
for ERY: threitol (THR); for XYL: ribitol (RIB) and arabitol (ARA);
and for the SOR: allitol (ALL), talitol (TAL), mannitol (MAN), ga-
lactitol (GAL), and iditol (IDI)]. The identity of these isomers
was confirmed by comparison with authentic samples. As
shown, the chromatogram can be divided into five regions
containing diols, triols, tetraols, pentaols, and hexitols. In the
pentaols region of the SOR reaction (Figure 1c), RIB, ARA, and
XYL are observed, providing evidence for CÀC cleavage reac-
tions. In the same region, a set of eight hexanepentaols
(
boxed) is observed. The structure of these compounds was
determined using the 6-deoxy sugar l-rhamnose (RAM,
Scheme 2) After a mild hydrogenation (16 h, 423 K) eight iso-
mers of 1,2,3,4,5-hexanepentaols are obtained, which corre-
spond exactly to those observed in the product spectrum of
SOR. This mixture was further reacted for 3 h at 423 K, yielding
Stereoisomerization
The formation of stereoisomers of SOR such as MAN and GAL
is commonly observed in reductive transformations of cellulose
and glucose. As the stereochemistry is of little relevance for
bulk chemical applications, reaction yields are typically ex-
pressed as summed hexitols. However, as discussed by Shanks
[
27]
et al., the formation stereoisomers can yield valuable insight
into the reaction mechanism. Stereoisomerization likely occurs
through metal-catalyzed epimerization. To gain more insight
into the reaction mechanism, H/D exchange was studied for
SOR and a-methylglucoside (AMG). AMG was selected as
a model for the cyclic units of cello-oligomers. Catalytic H/D
Scheme 2. Ru/C-catalyzed hydrogenation of RAM.
ChemSusChem 2015, 8, 3323 – 3330
www.chemsuschem.org
3324
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
exchange is well known and provides a simple means for the
[28–31]
perdeuteration of carbohydrates.
The exchange mecha-
nism is considered to involve adsorbed alkoxide species that
2
are reversibly dehydrogenated to h (C,O)-bonded carbonyls.
Surprisingly, only little is known of the relative reactivities of
[
32]
the different positions of sugar alcohols and carbohydrates.
Reactions were performed at 373 K in D O under 6 MPa H
2
2
1
3
using Ru/C and studied by quantitative C NMR spectroscopy
see the Supporting Information). Apart from slightly upfield-
shifted signals due to a- (À0.05 ppm) and b-deuteration
(
(
À0.005 ppm), no additional signals appeared. This confirms
that under these conditions the exchange proceeds with reten-
tion of the stereochemistry. In all cases, an initial induction
period of 0.5 h was observed, which is most likely due to the
reduction of a thin oxide layer on the catalyst. After that, the
non-deuterated signals decay in a first-order fashion. The ex-
trapolated rate constants are shown in Scheme 3. In line with
Figure 2. Hexitols composition. Conditions: 2.0 g substrate, 400 mg Ru/C
2 2
(5 wt%), 20 mL H O, 6 MPa H , 423 K).
À1
Scheme 3. Numbering scheme and rate constants (h ) of H/D exchange of
SOR and AMG. Conditions: 2.0 g substrate, 0.4 g Ru/C (5 wt%), 20 mL D
MPa H2, 373 K.
2
O,
6
literature reports, H/D exchange only occurred on hydroxylated
positions. The rate constants show that aldose formation
Scheme 4. Epimerization of SOR and proposed epimerization mechanism on
Ru/C.
À1
À1
(
0.47 h ) is preferred over 2-ketose formation (0.40 h ). The
À1
rate constant for 3-ketose formation (0.16 h ) is significantly
smaller and suggests that for the innermost positions steric
hindrance or strong adsorption may play a role. Compared to
2
h (C,O)-bonded carbonyl is the rate-determining step of epime-
rization.
À1
SOR, the rate constants of AMG are all smaller (0.02–0.13 h ),
confirming that the ring structure complicates dehydrogena-
tion. This indicates that the cyclic units of hydrogenated cello-
oligomers are less prone to react at the catalyst surface than
the terminal SOR moiety. Above 373 K, SOR is mostly convert-
ed to other hexitols. Figure 2 shows the evolution of the rela-
tive composition of hexitol isomers at 423 K. Initially, SOR is
rapidly converted to MAN and IDI. Over time ALL, GAL and TAL
are also formed in significant quantities and the composition
changes towards equilibrium. The more rapid formation of
MAN and IDI demonstrates that the 2- and 5-positions are con-
siderably more reactive than the 3- and 4-positions (Scheme 4).
The results are in line with the H/D exchange data discussed
above. Similar observations have also been reported for the
CÀC bond cleavage
As discussed above, the occurrence of CÀC bond-breaking re-
actions is well known in the field of metal-catalyzed hydroge-
nolysis. Particularly retro-aldol condensation and decarbonyla-
tion are often cited to account for the formation of short-chain
[
25–27]
compounds.
Both reactions are considered to involve an
initial dehydrogenation step to yield aldose or ketose inter-
mediates before cleavage.
The retro-aldol reaction can, in principle, cleave all CÀC
bonds of the polyol depending on the site of dehydrogenation
(i.e., leading to C /C , C /C , and C /C fragments) whereas de-
1
5
2
4
3
3
carbonylation selectively cleaves terminal CÀC bonds. Accord-
ingly, different reaction profiles are to be expected for these
mechanisms. For example, the methanol synthesis catalyst
CuO/ZnO/Al O shows high selectivity for short-chain polyols
[33]
epimerization of hexitols catalyzed by Ni/kieselgur. Epimeri-
zation is proposed to proceed via the conversion of the
2
1
h (C,O)-bonded carbonyl to a prochiral h (O)-mode. This spe-
2
3
[
34]
cies can either dissociate from the metal or revert back to
such as EG, 1,2-PrD, and GLY at low conversions. Figure 3a
shows the evolution of hexitols, hexanepentaols, pentitols, te-
tritols, and GLY in the reaction of SOR at 423 K (other products
are not shown for clarity). Clearly, pentitols are the main prod-
2
h (C,O)-bonding with interconversion of the enantiotopic face.
As H/D exchange occurs readily without racemization at lower
temperatures, it can be inferred that the dissociation of
ChemSusChem 2015, 8, 3323 – 3330
www.chemsuschem.org
3325
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Scheme 5. Decarbonylation of SOR and proposed reaction mechanism on
Ru/C.
XYL are formed in roughly equal amounts. ARA and XYL can
be obtained from SOR by decarbonylation at the 1- and 6-posi-
tions, respectively (Scheme 5). Various studies confirmed that
the heterogeneous decarbonylation mechanism involves an in-
itial dehydrogenation reaction to yield an adsorbed alde-
[
35]
hyde. Activation of the aldehydic H atom, in turn, leads to
a surface-bound acyl species, which can undergo CÀC cleavage
to yield adsorbed CO and an alkyl species.
Addition of hydrogen atoms to these species results in the
formation of the corresponding alcohols and methane. This
was confirmed by GC analysis of the gas phase, which showed
the presence of only H and CH . The data presented herein
2
4
clearly identify decarbonylation as the primary CÀC bond-
breaking reaction under neutral conditions. Of course, this
does not exclude the possibility of a retro-aldol reaction occur-
ring at higher temperatures or in the presence of basic co-cat-
alysts. Further studies are required to study the differences in
product distribution arising from the participation of the retro-
aldol reaction.
CÀO bond cleavage
The results obtained for RAM (Figure 1d) confirm that the first
deoxygenation step occurs selectively at the terminal positions
of the polyol chain. This is rather counterintuitive as dehydra-
tion reactions generally proceed via the acid-catalyzed E1
mechanism, which would favor the internal positions. This,
however, is not observed. Possible mechanisms accounting for
the observed dehydroxylation include a direct metal-catalyzed
CÀO activation or a dehydration reaction.
An important hint regarding the deoxygenation mechanism
is provided by the relative composition of hexanepentaols (Fig-
ure 3c). Initially, four of the eight possible isomers are formed
from SOR. As time progresses, the other four isomers are also
formed and the composition equilibrates. This suggests that
the deoxygenation mechanism results in the racemization of
one stereocenter (i.e., either the 2- or the 5-position,
Scheme 6). This excludes a direct CÀO activation and suggests
Figure 3. Ru/C-catalyzed hydrogenolysis of SOR (a), pentitols composition
b), and hexanepentaols composition (c). Conditions: 2.0 g substrate,
00 mg Ru/C (5 wt%), 20 mL H O, 6 MPa H , 423 K.
(
4
2
2
2
the intermediacy of a sp hybridization at the 2- and 5-posi-
tions, thus supporting a dehydration reaction. The dehydration
of terminal hydroxyl groups is proposed to proceed via a con-
certed E2 mechanism involving Lewis acid and basic sites
(Scheme 5). First, a terminal hydroxyl group is adsorbed onto
the catalyst. A metal hydroxyl species can then abstract
uct of the reaction, suggesting a decarbonylation reaction.
After exhibiting a maximum at 1 h, the amount decreases and
the amounts of tetritols and GLY increase. The relative compo-
sition of pentitols (Figure 3b) evidences that initially ARA and
ChemSusChem 2015, 8, 3323 – 3330
www.chemsuschem.org
3326
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
work. The substitution patterns
after the first deoxygenation
step could be assigned based on
the high preference for terminal
hydroxyl groups. Comparison of
the diols section of the GC chro-
matograms of ERY, XYL, SOR,
and RAM provides some impor-
tant clues regarding the course
of subsequent deoxygenation
steps. In the chromatogram of
ERY (Figure 4a), EG, 1,2-PrD, 2,3-
butanediol (2,3-BuD), and 1,2-bu-
Scheme 6. Deoxygenation of SOR and proposed reaction mechanism on Ru/C.
a proton on the 2-position, leading to elimination of the pri-
mary hydroxyl group and formation of the enol. As such, the
metal hydroxyl is regenerated. The enol, in turn, converts to
the keto state and is ultimately hydrogenated to the alcohol.
Numerous studies on the hydrogenolysis of GLY have repeat-
edly shown that 1,2-PrD is the main deoxygenation product ir-
[36]
respective of the metal catalyst used. This raises the ques-
tion whether the reaction is metal catalyzed or acid catalyzed.
Reactions without Ru/C were performed at 433 K in the ab-
sence and presence of catalytic amounts of silicotungstic acid
(
STA). In the absence of STA, SOR remained completely stable
for 3 h, whereas in the presence of STA 10% 1,4-sorbitan was
[37]
formed as the sole product. This indicates that under acidic
conditions nucleophilic cyclodehydration is the preferred path-
way. The absence of other products excludes possible rear-
Figure 4. GC chromatograms (CP-SIL PONA CB column) of ERY (a), XYL (b),
SOR (c), and RAM (d) reactions. Conditions: see Figure 1.
[
38]
rangement reactions (e.g., pinacol-type rearrangement) and
further confirms the role of the metal catalyst.
tanediol (1,2-BuD) are observed. 1,2-BuD can be obtained from
the deoxygenation of 1,2,3-butanetriol at the 3-position where-
as 2,3-BuD can be obtained from the deoxygenation at the 1-
position. This confirms again that deoxygenation occurs on the
outermost hydroxyl groups of the polyol chain independent of
whether they are attached to a primary or a secondary carbon
atom. Analogously, in the chromatogram of XYL (Figure 4b),
three additional peaks are observed, which are assigned to 2,3-
pentanediol (2,3-PeD) and 1,2-pentanediol (1,2-PeD). In the
chromatogram of SOR (Figure 4c), another four peaks are ob-
served, which are assigned to 2,3- and 3,4-hexanediol (2,3-/3,4-
HeD). The same signals are also observed in the chromatogram
of RAM (Figure 4d), with the exception that EG is absent.
Scheme 8 shows the proposed course of the multiple deoxyge-
nation steps of polyols. After the initial cleavage of a primary
Reaction network
Based on the results discussed above, we propose that the
product spectrum observed for SOR is the result of a reaction
network of decarbonylation and deoxygenation steps as out-
lined in Scheme 7. The observation that the product spectra of
XYL, ERY, and RAM are all included in the product spectrum of
SOR (Figure 1) is well in line with the proposed reaction net-
Scheme 7. Reaction network of the Ru/C-catalyzed hydrogenolysis of hexi-
tols.
Scheme 8. Proposed deoxygenation of polyols.
ChemSusChem 2015, 8, 3323 – 3330
www.chemsuschem.org
3327
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
hydroxyl group, a polyol with both terminal primary and termi-
nal secondary hydroxyl groups is obtained. The results clearly
show that these positions exhibit comparable reactivities. How-
ever, due to the simultaneous occurrence of decarbonylation
and deoxygenation reactions, it is at present not possible to
determine the selectivity difference between these positions.
Kinetic analysis of the reaction network is currently performed
to assess the rate constants and activation energies of the indi-
vidual steps of the reaction mechanism.
Product spectrum under acidic conditions
The influence of acid on the hydrogenolysis reaction was stud-
ied by the addition of a catalytic amount of STA (0.24 mmol
+
H ) to the reactions of SOR and cellobiitol (CEL). Interestingly,
the product spectrum of the reaction of SOR with acid is iden-
tical to that without (i.e., the same compounds are observed).
Dehydration products such as sorbitan or isosorbide were not
observed. Nevertheless, the evolution of the reaction mixture
shows considerable differences. In the presence of STA, only
5
4% hexitol conversion is observed after 3 h (Figure 5a). This
is significantly less than for the reaction without STA (Fig-
ure 3a), which amounted to 91%. The maximum amount of
pentitols observed in the reaction decreased from 17% with-
out to 7% with STA. Thus, the addition of acid inhibits decar-
bonylation. In contrast, the maximum amount of hexanepen-
taols increased from 6% without to 18% with acid. This dem-
onstrates that the acidic co-catalyst has a strong influence on
the selectivity of the hydrogenolysis reaction. This is particular-
[23]
ly advantageous for the synthesis of liquid alkanes as the
degradation towards light alkanes and hydrogen consumption
are decreased.
Figure 5. Ru/C-catalyzed hydrogenolysis of SOR (a) and CEL (b) with STA.
Conditions: 2.0 g substrate, 400 mg Ru/C (5 wt%), 0.06 mmol STA, 20 mL
Cellohexitols
H
2
O, 6 MPa H
2
,423 K; cellohexitol conversion was determined by performing
A kinetic study by Fukuoka et al. demonstrated that the slow
depolymerization of cellulose and the simultaneous degrada-
LC–ESI–MS).
[39]
tion by Ru severely limits hexitol yields to 40%. The use of
ball milling as pretreatment leads to significantly improved
hexitol yields due to the formation of smaller (water-soluble)
that, the yield decreases and the amount of degradation prod-
ucts such as pentitols and hexanepentaols steadily increases.
LC–ESI–MS data also confirm a direct degradation of the cello-
[
12–17,39]
cello-oligomers.
Previous studies by us already showed
that the hydrolytic hydrogenation of cello-oligomers
such as cellobiose and cellotriose proceeds smoothly
[40,41]
in the presence of Ru/C and STA.
Maximum hexi-
tols yields of up to 80% are typically observed. Kinet-
ic analysis of the reaction network revealed that the
hydrogenation of cello-oligomers is considerably
faster than hydrolysis, leading to ‘cellohexitols’ (i.e.,
cello-oligmers with a terminal hexitol unit). To access
the hydrogenolysis of these oligomers we used cello-
biose as model substrate (Scheme 9). Cellohexitols
have been prepared by performing a mild hydroge-
nation of cellobiose at 373 K for 16 h. Figure 5b
shows the conversion of the resulting cellohexitol
mixture in the presence of STA at 423 K. The sub-
strate is rapidly converted within 1 h, and a maximum
yield of 79% hexitols is observed after 0.5 h. After Scheme 9. Hydrolytic hydrogenation and hydrogenolysis of cellohexitols.
ChemSusChem 2015, 8, 3323 – 3330
www.chemsuschem.org
3328
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
hexitols (see the Supporting Information). In contrast to the
degradation network of SOR, degradation is much less exten-
sive. This suggests that either the ether linkage or the high hy-
drolysis rate prevent further degradation.
reaction was monitored by periodically taking sample from the au-
toclave. Samples were filtered through a 25 mm polyamide filter
prior to analysis.
NMR and GC–LC analysis
Conclusions
Sample solutions of H/D exchange reactions were measured direct-
13
1
ly by quantitative C{ H} NMR spectroscopy (Bruker Avance III
600 MHz, integrals were normalized to carbon atom; assignments
The complete product spectra of C –C polyols obtained
2
6
[42–44]
through the hydrogenolysis of erythritol, xylitol, sorbitol, and
cellobiitol have been identified. Analysis of the product spec-
trum revealed a complex reaction network consisting of ster-
eoisomerization and CÀC and CÀO bond cleavage reactions.
Based on H/D exchange data and the evolution of stereoiso-
mers it is proposed that Ru/C catalyzes (de)hydrogenation, epi-
merization, decarbonylation, and dehydration reactions. The
occurrence of retro-aldol condensation under the employed
conditions could not be confirmed. As such, the primary Ru-
catalyzed degradation pathways of higher polyols have been
identified. The addition of silicotungstic acid lead to a suppres-
sion of decarbonylation and an enhancement of deoxygena-
tion reactions. ESI–MS analysis of the reaction of cello-oligo-
mers revealed that they are also degraded by Ru/C. With the
insights gained and the methods developed for this study, it is
now possible to close the mass balances of the liquid-phase
products. It is expected that these results are also relevant for
other catalyst systems and can benefit research aimed at the
hydrogenolysis and/or hydrodeoxygenation of cellulosic bio-
mass.
were based on literature reference data).
0.5 mL aliquots were dried using a Eppendorf Speedvac system
303 K, 8 h). The resulting residue was dissolved in 1 mL acetic an-
In all other cases,
(
hydride/pyridine mixture (1:1 v/v) and left to react for 3 days at
room temperature with periodic mixing and shaking. Subsequently,
the sample solutions were measured by GC [Thermo Scientific
Trace GC system equipped with an Agilent DB-23 column (internal
diameter: 0.25 mm; length: 60 m; film thickness: 0.25 mm; isobaric:
0
.1 MPa He; temperature gradient: 353–527 K) or a CP-SIL PONA
CB column (internal diameter: 0.21 mm; length: 50 m; film thick-
ness: 0.21 mm; isobaric: 0.1 MPa He; temperature gradient: 323–
5
03 K)], GC–MS (Thermo Scientific Trace 1310 system equipped
with a single quadrupole MS, EI+, 70 eV), and/or HPLC–ESI–MS
[
(
Shimadzu LC-MS 2020 system using a LiChrospher 100 column
RP-18e; length: 25 cm; particle size: 5 mm; binary gradient 30–
5
0% B with the remainder being A (10 mm ammonium acetate
aqueous solution with 0.1% formic acid); B: acetonitrile with 0.1%
formic acid)]. All compounds were calibrated using the external
standard method. Isomeric products were treated as possessing
equal response factors. Hexitols were calibrated using sorbitol. The
pentitols were calibrated using xylitol. The tetritols were calibrated
using erythritol. The hydrogenated solutions of l-rhamnose and
cellobiose were used to quantify hexanepentaols and cellohexitols,
respectively.
Experimental Section
Acknowledgements
d-(+)-cellobiose (>98%) and d-isosorbide (>98%) were obtained
from Alfa Aesar. Ethylene glycol, 1,2-propanediol, glycerol, 1,2-buta-
nediol, 2,3-butanediol, Ru/C (5 wt%), silicotungstic acid hydrate, al-
litol, d-talitol, l-iditol, and l-rhamnose monohydrate (>99%) were
obtained from Sigma–Aldrich. d-(À)-Sorbitol (molecular Biology
grade) was obtained from AppliChem. Galactitol, d-(+)-arabitol,
mannitol, d-ribitol, xylitol, and erythritol were obtained from Supel-
co. Methyl-a-d-glucopyranoside (>99%) was obtained from Fluka.
Deuteriumoxide (99.9% D) was obtained from Deutero GmbH. 1,4-
The authors kindly thank Heike Fickers-Boltz, Elke Biener, and
Hannelore Eschemann for the GC analyses; Noah Avraham for
HPLC analyses; and Ines Bachmann for NMR analyses. This work
was supported by the Robert Bosch Foundation within the Robert
Bosch Junior Professorship for the efficient utilization of renewa-
ble resources. This work was performed as part of the Cluster of
Excellence “Tailor-Made Fuels from Biomass” funded by the Excel-
lence Initiative by the German federal and state governments to
promote science and research at German universities.
[42]
sorbitan was prepared according to a literature procedure.
Autoclave reactions
H/D exchange and hydrogenolysis experiments were performed in
a 50 mL batch-type high-pressure autoclave reactor. Typically, sub-
strate (2.0 g), Ru/C (0.4 g), and, where applicable, silicotungstic acid
Keywords: biomass
polyols · ruthenium
· deoxygenation · hydrogenolysis ·
(
0.175 g) were added into a glass-lined stainless steel reactor
[1] C. H. Zhou, X. Xia, C. X. Lin, D. S. Tong, J. Beltramini, Chem. Soc. Rev.
equipped with a sampling valve and charged with H O or D O
2011, 40, 5588–5617.
2
2
(
20 mL). The reactor was flushed by pressurizing and depressuriz-
[2] D. M. Alonso, S. G. Wettstein, J. A. Dumesic, Chem. Soc. Rev. 2012, 41,
ing three times with 6 MPa H at room temperature. The reactor
8075–8098.
2
[
3] J. P. Ma, W. Q. Yu, M. Wang, X. Q. Jia, F. Lu, J. Xu, Chin. J. Catal. 2013, 34,
was pressurized with 6 MPa H2 and then heated to the defined
temperature (373–423 K). The time zero was set at the beginning
of the isothermal reaction stage. For cellobiose and l-rhamnose,
the reactions were performed in two stages. First, the reaction mix-
492–507.
[
[
[
[
4] Q. H. Zhang, F. Jerome, ChemSusChem 2013, 6, 2042–2044.
5] M. Besson, P. Gallezot, C. Pinel, Chem. Rev. 2014, 114, 1827–1870.
6] M. J. Climent, A. Corma, S. Iborra, Green Chem. 2014, 16, 516–547.
7] R.-J. van Putten, J. C. van der Waal, E. de Jong, C. B. Rasrendra, C. B.
Heeres, J. G. de Vries, Chem. Rev. 2013, 113, 1499–1597.
ture was heated to 373 K for 16 h under 6 MPa H . After cooling
2
the autoclave to room temperature, the gas phase was flushed
twice and pressurized with 6 MPa H . Subsequently, the autoclave
[8] F. Schüth, R. Rinaldi, N. Meine, M. Käldstrçm, J. Hilgert, M. D. Kaufman
2
was heated to the desired reaction temperature. Progress of the
Rechulski, Catal. Today 2014, 234, 24–30.
ChemSusChem 2015, 8, 3323 – 3330
www.chemsuschem.org
3329
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[
9] X. R. Liu, X. C. Wang, S. X. Yao, Y. J. Jiang, J. Guan, X. D. Mu, RSC Adv.
014, 4, 49501–49520.
10] A. Ruppert, K. Weinberg, R. Palkovits, Angew. Chem. Int. Ed. 2012, 51,
564–2601; Angew. Chem. 2012, 124, 2614–2654.
[26] K. Wang, M. C. Hawley, T. D. Furney, Ind. Eng. Chem. Res. 1995, 34,
3766–3770.
[27] K. L. Deutsch, D. G. Lahr, B. H. Shanks, Green Chem. 2012, 14, 1635–
1642.
2
[
[
[
[
[
[
[
[
[
[
[
2
11] C. Luo, S. Wang, H. Liu, Angew. Chem. Int. Ed. 2007, 46, 7636–7639;
Angew. Chem. 2007, 119, 7780–7783.
12] J. Geboers, S. van de Vyver, K. Carpentier, K. de Blochouse, P. Jacobs, B.
Sels, Chem. Commun. 2010, 46, 3577–3579.
[28] H. J. Koch, R. S. Stuart, Carbohydr. Res. 1977, 59, C1–C6.
[29] H. J. Koch, R. S. Stuart, Carbohydr. Res. 1978, 64, 127–134.
[30] T. Maegawa, Y. Fujiwara, Y. Inagaki, Y. Mongushi, H. Sajiki, Adv. Synth.
Catal. 2008, 350, 2215–2218.
[31] Y. Fujiwara, H. Iwata, Y. Sawama, Y. Monguchi, H. Sajiki, Chem. Commun.
2010, 46, 4977–4979.
[32] Y. Sawama, Y. Yabe, H. Iwata, Y. Fujiwara, Y. Monguchi, H. Sajiki, Chem.
Eur. J. 2012, 18, 16436–16442.
[33] L. Wright, L. Hartmann, J. Org. Chem. 1961, 26, 1588–1596.
[34] K. Tajvidi, P. J. C. Hausoul, R. Palkovits, ChemSusChem 2014, 7, 1311–
1317.
[35] M. Mavrikakis, M. A. Barteau, J. Mol. Catal. A 1998, 131, 135–147.
[36] J. ten Dam, U. Hanefeld, ChemSusChem 2011, 4, 1017–1034.
[37] J. U. Oltmanns, S. Palkovits, R. Palkovits, Appl. Catal. A 2013, 456, 168–
173.
[38] W. B. Smith, Tetrahedron 2002, 58, 2091–2094.
[39] T. Komanoya, H. Kobayashi, K. Hara, W.-J. Chun, A. Fukuoka, ChemCatCh-
em 2014, 6, 230–236.
13] J. Geboers, S. Van de Vyver, K. Carpentier, P. Jacobs, B. Sels, Green Chem.
2011, 13, 2167–2174.
14] J. Geboers, S. Van de Vyver, K. Carpentier, P. Jacobs, B. Sels, Chem.
Commun. 2011, 47, 5590–5592.
15] J. Hilgert, N. Meine, R. Rinaldi, F. Schuth, Energy Environ. Sci. 2013, 6,
9
2–96.
16] W. W. Zhu, H. M. Yang, J. Z. Chen, C. Chen, L. Guo, H. M. Gan, X. G. Zhao,
Z. S. Hou, Green Chem. 2014, 16, 1534–1542.
17] Y. Liu, L. G. Chen, T. J. Wang, Y. Xu, Q. Zhang, L. L. Ma, Y. H. Liao, N. Shi,
RSC Adv. 2014, 4, 52402–52409.
18] R. Palkovits, K. Tajvidi, J. Procelewska, R. Rinaldi, A. Ruppert, Green
Chem. 2010, 12, 972–978.
19] R. Palkovits, K. Tajvidi, A. M. Ruppert, J. Procelewska, Chem. Commun.
2011, 47, 576–578.
20] Y. Liu, C. Luo, H. C. Liu, Angew. Chem. Int. Ed. 2012, 51, 3249–3253;
Angew. Chem. 2012, 124, 3303–3307.
[40] L. Negahdar, J. U. Oltmanns, S. Palkovits, R. Palkovits, Appl. Catal. B
2014, 147, 677–683.
[
[
21] A. Q. Wang, T. Zhang, Acc. Chem. Res. 2013, 46, 1377–1386.
22] Y. Osaka, Y. Ikeda, D. Hashizume, M. Iwamoto, Biomass Bioenergy 2013,
[41] L. Negahdar, P. J. C. Hausoul, S. Palkovits, R. Palkovits, Appl. Catal. B
2015, 166–167, 460–464.
5
6, 1–7.
[42] G. J. Stockburger, US4297290A, 1981.
[
[
[
23] Y. Liu, L. Chen, T. Wang, Q. Zhang, C. Wang, J. Yan, L. Ma, ACS Sustaina-
ble Chem. Eng. 2015, 3, 1745–1755.
24] D. K. Sohounloue, C. Montassier, J. Barbier, React. Kinet. Catal. Lett. 1983,
[43] S. J. Angyal, R. Le Fur, Carbohydr. Res. 1980, 84, 201–209.
[44] S.-C. Ho, H. J. Koch, R. S. Stuart, Carbohydr. Res. 1978, 64, 251–256.
2
2, 391–397.
25] C. Montassier, J. C. Menezo, L. C. Hoang, C. Renaud, J. Barbier, J. Mol.
Catal. 1991, 70, 99–110.
Received: April 8, 2015
Published online on September 3, 2015
ChemSusChem 2015, 8, 3323 – 3330
www.chemsuschem.org
3330
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim