Hydrogenation of levoglucosenone to renewable chemicals

Article abstract of DOI:10.1039/c6gc03028a

We have studied the hydrogenation of levoglucosenone (LGO) to dihydrolevoglucosenone (Cyrene), levoglucosanol (Lgol), and tetrahydrofurandimethanol (THFDM) and elucidated the reaction network over supported palladium catalysts. At low temperature (40 °C) over a Pd/Al₂O₃ catalyst, LGO is selectively hydrogenated to Cyrene. At intermediate temperatures (100 °C) over a Pd/Al₂O₃ catalyst, Cyrene is selectively hydrogenated to Lgol, with an excess of the exo-Lgol isomer produced over the endo-Lgol isomer. At higher temperatures (150 °C) over a bifunctional Pd/SiO₂-Al₂O₃ catalyst, Lgol is converted to THFDM in 58% selectivity, with 78% overall selectivity to 1,6-hexanediol precursors. The ratio of cis-THFDM relative to trans-THFDM is approximately 2.5, and this ratio is independent of the Lgol feed stereoisomer ratio. Tetrahydropyran-2-methanol-5-ketone (THP2M5one) and tetrahydropyran-2-methanol-5-hydroxyl (THP2M5H) are side-products of Lgol hydrogenolysis, but neither of these species are precursors to THFDM.

Full text of DOI:10.1039/c6gc03028a

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: S. Krishna, D.
McClelland, Q. Rashke, J. A. Dumesic and G. Huber, Green Chem., 2016, DOI: 10.1039/C6GC03028A.
This is an Accepted Manuscript, which has been through the
Volume 18 Number
7 7 April 2016 Pages 1821–2242
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Accepted Manuscripts are published online shortly after
Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1463-9262
CRITICAL REVIEW
G. Chatel et al.
Heterogeneous catalytic oxidation for lignin valorization into valuable
chemicals: what results? What limitations? What trends?
rsc.li/green-chem

Page 1 of 17
Green Chemistry
View Article Online
DOI: 10.1039/C6GC03028A
Hydrogenation of Levoglucosenone to Renewable Chemicals*
Siddarth H. Krishna, Daniel J. McClelland, Quinn A. Rashke, James A. Dumesic, George W.
Huber†
Abstract
We have studied the hydrogenation of levoglucosenone (LGO) to dihydrolevoglucosenone
(Cyrene), levoglucosanol (Lgol), and tetrahydrofurandimethanol (THFDM) and elucidated the
reaction network over supported palladium catalysts. At low temperature (40°C) over a Pd/Al2O3
catalyst, LGO can be selectively hydrogenated to Cyrene. At intermediate temperatures (100°C)
over a Pd/Al2O3 catalyst, Cyrene can be selectively hydrogenated to Lgol, with an excess of the
exo-Lgol isomer produced over the endo-Lgol isomer. At higher temperatures (150°C) over a
bifunctional Pd/SiO2-Al2O3 catalyst, Lgol is converted to THFDM in 58% selectivity, with 78%
overall selectivity to 1,6-hexanediol precursors. The ratio of cis-THFDM relative to trans-THFDM
is approximately 2.5, and this ratio is independent of the Lgol feed stereoisomer ratio.
Tetrahydropyran-2-methanol-5-ketone (THP2M5one) and tetrahydropyran-2-methanol-5-
hydroxyl (THP2M5H) are side-products of Lgol hydrogenolysis, but neither of these species are
precursors to THFDM.
1. Introduction
In recent years, lignocellulosic biomass has gained interest as a renewable feedstock for the
production of high-value oxygenated chemicals.1,2 We have recently shown that levoglucosenone
(LGO) can be produced from cellulose in 50% yield using a polar aprotic solvent and sulfuric acid
catalyst.3 This strategy uses relatively mild conditions (210°C, 20 mM H2SO4) and avoids complex
separation processes such as the vacuum pyrolysis approach used by Kawamoto et al.4 Due to
these improvements over the conventional pyrolysis approach,4, 5 LGO has the potential to become
a biomass-derived platform molecule, thereby broadening the opportunities for conversion of
cellulose to fuels and chemicals.
As shown in Scheme 1, LGO can be hydrogenated into a wide variety of chemicals. At low
temperatures (25-60oC) over a Pd/C catalyst, LGO is hydrogenated into Cyrene
(dihydrolevoglucosenone), a non-toxic solvent with similar properties to environmentally harmful
solvents such as N-methylpyrrolidone.6-8 DuPont has reported that LGO can be hydrogenated to
levoglucosanol (Lgol) followed by hydrogenolysis to tetrahydrofurandimethanol (THFDM) in
84% yield over a Pt/C catalyst.9 THFDM can be further upgraded to 1,6-hexanediol in >80% yield
using a bifunctional catalyst that contains metal and acid sites.10-13 Lgol contains three chiral
centers and is a potential platform molecule for chiral products relevant to the fine chemicals and
pharmaceuticals industry.14 THFDM is an α,ω-diol with the potential to be a polymer precursor
similar to existing α,ω-diols in the market.2,15 1,6-hexanediol is a commodity chemical used in
polyurethanes, coatings, and adhesives.12,16-18 While the biomass-derived intermediate 5-
* Electronic supplementary information (ESI) available: Gaussian DFT calculations; 13C NMR of reaction products;
CO chemisorption data; synthesis of reaction intermediates; product deconvolution methods; solvent degradation
data; and catalyst recycling data.
† Department of Chemical and Biological Engineering, University of Wisconsin-Madison, Madison, WI 53706, USA.
Email: gwhuber@wisc.edu.

Green Chemistry
Page 2 of 17
View Article Online
DOI: 10.1039/C6GC03028A
hydroxymethylfurfural (HMF) can also be hydrogenated into THFDM and 1,6-hexanediol,12,19-21
LGO upgrading could provide a new route to these chemicals as well as Cyrene and Lgol.
Furthermore, the fundamental chemistry of LGO hydrogenation and hydrogenolysis has not been
explored in detail in the literature.
Scheme 1: Levoglucosenone hydrogenation. Enthalpies and entropies of reaction were estimated
using Gaussian DFT calculations (see Supplemental Information).
Herein, we study the hydrogenation of LGO over supported palladium catalysts. We elucidate the
reaction network and demonstrate catalysts and conditions suitable to produce Cyrene, Lgol, or
THFDM depending on which products are desired. We also gain mechanistic insights from
investigations of the effect of metal acid sites and selectivity as a function of conversion. Our
results reveal the chemistry underlying the hydrogenation of LGO, and provide directions for
designing catalytic systems to optimize the production of several renewable chemicals, including
green solvents and polymer precursors, from lignocellulosic biomass.
2. Experimental
LGO (98%, Apollo Scientific), Cyrene (99%, Apollo Scientific), THFDM (98%, Alfa Chemicals),
tetrahydrofuran (THF) stabilized with butylated hydroxytoluene (BHT) (99.9%, Acros Organics),
and Pd(NO3) solution (10% in 10% HNO3, Sigma Aldrich), were used as received. 5% Pd/C and
low-soda γ-Al2O3 were purchased from Strem Chemicals. 1% Pd/C, Davisil SiO2, and Grade 135
amorphous SiO2-Al2O3 (Si-Al) were purchased from Sigma Aldrich.
2.1. Catalyst synthesis methods
Palladium catalysts supported on Al2O3 and Si-Al were synthesized by incipient wetness
impregnation (IWI) of a 10% Pd(NO)3 precursor onto the appropriate support. After impregnation,
catalysts were dried overnight in an oven at 110°C. The catalysts were then calcined in flowing air
at 400°C followed by reduction in flowing H2 at 260°C and passivation in 1% O2/He at room
temperature.
2.2. Catalyst characterization methods
Carbon monoxide (CO) chemisorption was carried out in a Micromeritics ASAP2020C
instrument. 100 mg of the catalyst was loaded into a glass reactor. Catalysts were degassed
followed by in situ reduction in flowing H2 at 400°C and sample evacuation. The CO
chemisorption measurement was conducted at 35°C, and repeated after subsequent evacuation to
acquire the reversible CO uptake. The irreversible CO uptake was calculated by subtracting the
total uptake in the first measurement from the reversible uptake in the second measurement.
2.3 Reaction Studies

Page 3 of 17
Green Chemistry
View Article Online
DOI: 10.1039/C6GC03028A
Reactions were carried out in 45 or 75 mL Parr Hastelloy high-pressure batch reactors. A magnetic
stir bar and appropriate amount of catalyst and feedstock were added to the reactor. The reactor
was purged four times with 500 psi H2, followed by pressurizing to 500 psi H2. The heat-up time
was 10-15 minutes. The reaction mixture was stirred at 750 rpm. After the appropriate time, the
reactor was quenched in ice water, depressurized, and opened. The reaction products were filtered
using a 0.2 m PTFE (polytetrafluoroethylene) syringe filter prior to analysis.
For LGO hydrogenation at low temperature (40°C), catalysts were pre-reduced at 100°C in flowing
H2 and loaded into Parr reactors in a glovebox to avoid exposure to air. For reactions at higher
temperature (>40°C), passivated catalysts were loaded directly into the reactor without pre-
reduction.
Dip-tube experiments were carried out with 60 mL reaction mixture in a 75 mL Parr batch reactor
equipped with a 1/8-inch dip-tube. During sampling, 0.6 mL was purged followed by 0.6 mL
collected for analysis. Purging was done to remove dip-tube liquid collected at earlier reaction
times. A stainless steel filter (500-mesh) was used to prevent solids from entering the dip-tube.
After each sample, the reactor was re-pressurized with H2.
2.4 Product analysis methods
2.4.1 GC
Products were quantified using a Shimadzu Gas Chromatograph equipped with a Flame Ionization
Detector (FID) with liquid injection. A Restek RTX-VMS capillary column (Length: 30 m, ID:
0.25 mm, film thickness: 1.4 m) was used. The injection port and FID were maintained at 240°C.
The injection volume was 1 L and a split ratio of 50 was used. The column temperature ramp
was as follows: hold 1 min at 40°C, ramp 10°C/min to 180°C, ramp 3°C/min to 240°C, hold 5 min
at 240°C.
Because Lgol, THP2M5one (tetrahydropyran-2-methanol-5-ketone), and THP2M5H
(tetrahydropyran-2-methanol-5-hydroxyl) are not commercially available, these compounds were
synthesized and their identities were confirmed by 13C NMR. Lgol and THP2M5one were assumed
to have GC sensitivities equal to the average sensitivity of LGO and Cyrene. THP2M5H and
2MTHFA (2-methyl-tetrahydrofurfuryl alcohol) were assumed to have GC sensitivities equal to
that of THFDM. 2MTHFA was identified using a Shimadzu Gas Chromatograph equipped with a
Quadrupole Mass Spectrometer with liquid injection (GCMS-QP2010S). The ion source and
interface temperatures were set to 200°C, and a solvent cut was used to avoid introducing the
solvent into the MS.
It was observed that cis-THFDM was overlapped by THP2M5one in the GC. THP2M5one was
quantified independently using HPLC, and the concentration of cis-THFDM was calculated by
subtracting the known concentration of THP2M5one from the total concentration of these two
species in the GC. It was also observed that trans-THFDM was overlapped by one of the
stereoisomers of THP2M5H in the GC. The relative concentrations of cis-THFDM, trans-THFDM,
and the two stereoisomers of THP2M5H were acquired by quantitative 13C NMR. For experiments
with sampling over time, the concentration of trans-THFDM at earlier reaction times was
calculated from the cis-THFDM concentration, assuming that the cis/trans ratio is not a function
of conversion. This assumption was validated by running the Lgol hydrogenolysis reaction to high

Green Chemistry
Page 4 of 17
View Article Online
DOI: 10.1039/C6GC03028A
conversion as well as intermediate conversion and confirming that the THFDM cis/trans ratio was
similar in both cases. Then, the concentration of the overlapped isomer of THP2M5H at earlier
time points was calculated by subtracting the known concentration of trans-THFDM from the total
concentration of these two species in the GC. Product deconvolution using a combination of GC,
NMR, and HPLC is discussed in detail in the Supplemental Information.
2.4.2. HPLC
THP2M5one was quantified using a Shimadzu high performance liquid chromatograph (HPLC)
with a BioRad Aminex 87H column and 5 mM H2SO4 mobile phase. The column was operated at
a flow rate of 0.6 mL/min and a temperature of 30oC. A photodiode array (UV) detector was used
at a UV wavelength of 206 nm.
2.4.3 NMR
Quantitative 13C NMR spectra were collected on a Bruker Avance 500 MHz spectrometer at room
13
temperature, using an inverse-gated decoupling pulse sequence with a 30o pulse. The C NMR
spectra were absolute-referenced to the associated 1H NMR spectra. Unless otherwise indicated,
256 scans were used, with an acquisition time of 1 sec and a relaxation delay of 15 sec. Reaction
products exo-Lgol, endo-Lgol, cis-THFDM, trans-THFDM, THP2M5one, and the two
stereoisomers of THP2M5H were identified using 13C NMR. Samples were prepared for analysis
by evaporating the THF solvent from 5-10 mL of sample, then adding 1-3 mL of d6-DMSO. 13C
NMR spectra for these compounds are provided in the Supplemental Information.
3. Results and Discussion
3.1. Hydrogenation of LGO to Cyrene
LGO hydrogenation to Cyrene was carried out over a 0.4 wt% Pd/Al2O3 catalyst. THF was selected
as the reaction solvent because this solvent was effective in our previous work for cellulose
dehydration to LGO.3 As shown in Figure 1, at low temperature (40°C), LGO was hydrogenated
to Cyrene in 100% selectivity. These results agree with the findings of Clark et al., who observed
quantitative conversion of LGO to Cyrene over a Pd/C catalyst at room temperature.6
Levoglucosenol, the unsaturated alcohol resulting from selective hydrogenation of the C=O bond
of LGO, was not observed at any time. Similarly, Ronzon et al. found that complete C=C bond
hydrogenation occurred prior to C=O hydrogenation in the hydrogenation of 2-cyclohexenone (a
molecule with the same 2-enone functionality as is present in LGO) over Rh/SiO2 catalysts.22 In
general, hydrogenation of C=C bonds is thermodynamically favored compared to hydrogenation
of C=O bonds.23 Based on the data in Figure 1, the initial turnover frequency (TOF) for LGO
hydrogenation at these conditions was 345 mol Cyrene per mol surface Pd per min. Catalyst
recycling data suggest that this catalyst undergoes deactivation under these conditions
(Supplemental Information, Figure S6).

Page 5 of 17
Green Chemistry
View Article Online
DOI: 10.1039/C6GC03028A
Figure 1. Hydrogenation of LGO over 1.75 mg 0.4 wt% Pd/Al2O3 (17.5 mg diluted 10x in SiO2) in
a batch reactor with dip-tube sampling. Conditions: 40°C, 500 psi H2, 60 mL 55 mM LGO feed in
THF. Solid lines between points are visual aids.
3.2. Hydrogenation of Cyrene to Lgol
Cyrene hydrogenation to Lgol was investigated at higher temperature (100°C) and at higher
catalyst amounts compared to LGO hydrogenation. As shown in Figure 2, under these conditions
Cyrene was hydrogenated to Lgol in 100% selectivity. Based on the data in Figure 2, the initial
turnover frequency (TOF) for Cyrene hydrogenation at these conditions was 8.5 mol Lgol per mol
surface Pd per min. Catalyst recycling data suggest that this catalyst undergoes more significant
deactivation under these conditions compared to the case of LGO hydrogenation (Supplemental
Information, Figure S7). An excess of the exo-Lgol isomer was produced over the endo-Lgol
isomer (exo/endo = 4.6), and the stereoisomer ratio was not a function of conversion. Augustine
argued that in the stereoselective hydrogenation of prochiral ketones (i.e., ketones which already
contain stereocenters), hydrogen atom addition from the less hindered direction should be
kinetically favored.24 We hypothesize that the observed excess of exo-Lgol results from
preferential hydrogen atom addition to the ketone group of Cyrene from the opposite side as the
anhydro-bridge.5 Similarly, Zanardi et al. found that the C=O bond of LGO was selectively
reduced to the exo-alcohol when sodium borohydride was used as the reductant.14 Based on our
results, we have proposed the reaction network for LGO hydrogenation to Lgol shown in Scheme
2.

Green Chemistry
Page 6 of 17
View Article Online
DOI: 10.1039/C6GC03028A
Figure 2. Hydrogenation of Cyrene over 150 mg 0.4 wt% Pd/Al2O3 in a batch reactor with dip-
tube sampling. Conditions: 100°C, 500 psi H2, 60 mL 55 mM Cyrene in THF feed. Solid lines
between points are visual aids.
Scheme 2: Proposed reaction network for LGO hydrogenation.
The ratio of exo-Lgol to endo-Lgol in Cyrene hydrogenation for different supported palladium
catalysts is shown in Table 1. While the conversion values were different for the different
experiments in Table 1, Figure 2 indicates that the stereoselectivity does not depend on conversion.
Cyrene hydrogenation over a 1 wt% Pd/C catalyst resulted in exo/endo = 2.3, while hydrogenation
over a 0.4 wt% Pd/Al2O3 catalyst resulted in exo/endo = 4.6. Since the stereoisomer ratio varies
with the catalyst used, it is governed by kinetics rather than thermodynamics. The stereoisomer
ratio is a function of the catalyst support, as indicated by the stereoisomer ratio over 1% and 5%
Pd/C catalysts (1.7-2.3) versus 0.4% and 5% Pd/Al2O3 catalysts (4.6-4.7) and the 1% Pd/Si-Al
catalyst (3.5). Comparing 0.4 wt% Pd/Al2O3 (~2 nm particles, as shown in Table S1) with 5 wt%
Pd/Al2O3 (~10 nm particles), the stereoisomer ratios were similar, suggesting that the stereoisomer
ratio is not affected by changes in the particle size. Additionally, as shown in Table 1, the

Page 7 of 17
Green Chemistry
View Article Online
DOI: 10.1039/C6GC03028A
stereoisomer ratios over the 0.4 wt% Pd/Al2O3 were comparable at temperatures of 100°C and
150°C. This result suggests that the apparent activation energies for exo-Lgol and endo-Lgol
formation are similar. While the stereoisomer ratio was observed to change with the choice of
support, the underlying reason for this dependence is not known at present.
Table 1. Cyrene Hydrogenation –Stereoisomer Ratio
Catalyst
Temperature [°C]
Conversion [%]
Lgol Isomer Ratio
(exo/endo)
1% Pd/C
5% Pd/C
0.4% Pd/Al2O3 100
0.4% Pd/Al2O3 150
5% Pd/Al2O3
1% Pd/Si-Al
100
100
71
2.3
1.7
4.6
4.4
4.7
3.5
100
0-100*
100
100
95
100
100
Conditions: 500 psi H2, 10 mL 55 mM Cyrene in THF feed, 50 mg cat, 3h reaction time.
*Taken from Figure 2.
3.3. Hydrogenolysis of Lgol to THFDM
The hydrogenolysis of Lgol to THFDM involves a ring-rearrangement to form the 5-membered
ring of THFDM. Using a Pd/Si-Al catalyst at 150°C, 58% selectivity towards THFDM was
achieved at 90% Lgol conversion, as shown in Table 2. A ~2.8:1 excess of cis-THFDM was
produced over trans-THFDM. In contrast, when THFDM is produced from HMF hydrogenation,
the product mixture has been reported to comprise predominantly (>95%) the cis-THFDM isomer,
most likely because the C=C bonds in the furan ring of HMF would be hydrogenated from the
same direction.12,25 THFDM can be converted into 1,2,6-hexanetriol followed by further upgrading
to 1,6-hexanediol.10-13 We expect that both cis-THFDM and trans-THFDM can be upgraded to
1,6-hexanediol.
The major side-products of Lgol hydrogenolysis were THP2M5H and 2MTHFA, as shown in
Scheme 3. Both of these species contain two stereoisomers, but the stereochemistry could not be
assigned due to the low yields of these products. DuPont also reported THP2M5H as a minor
product of Lgol hydrogenolysis.9 Although we do not know of any literature studies of THP2M5H
hydrogenolysis, we expect that THP2M5H, like THFDM, can be upgraded to 1,2,6-hexanetriol
through ring-opening at the C-O bond adjacent to the hydroxymethyl group. Therefore, the overall
selectivity to 1,6-hexanediol precursors THFDM and THP2M5H is 78%. We note that degradation
of the THF solvent in the presence of Lgol is negligible under these reaction conditions, as
presented in the Supplemental Information.

Green Chemistry
Page 8 of 17
View Article Online
DOI: 10.1039/C6GC03028A
Scheme 3: Proposed reaction network for Lgol hydrogenolysis.
3.3.1. Effect of metal and acid sites
As shown in Table 2, Lgol was unreactive in the presence of a 1% Pd/Al2O3 catalyst at 150°C,
indicating that acid sites are necessary for Lgol hydrogenolysis. Using the pure Si-Al support
resulted in 32% Lgol conversion. Low selectivity (19%) was observed to THP2M5one, a product
of acid-catalyzed Lgol isomerization. A physical mixture of 1% Pd/Al2O3 and Si-Al resulted in
lower conversion (41%) and lower THFDM selectivity (29%) compared to the Pd/Si-Al catalyst
(90% conversion, 58% THFDM selectivity), indicating that close proximity of metal (Pd) and acid
(Si-Al) sites26,27 is favorable for this reaction. Acid catalysts are known to carry out ring
rearrangement reactions by protonating ether oxygen atoms to facilitate C-O bond cleavage.28
Downstream hydrogenolysis products of THFDM, such as 1,2,6-hexanetriol, were not observed.
The selectivities to identified products do not add to 100% when pure Si-Al support (19%) or a
physical mixture of Pd/Al2O3 and Si-Al (48%) are used, indicating the formation of unaccounted
degradation products in these cases.

Page 9 of 17
Green Chemistry
View Article Online
DOI: 10.1039/C6GC03028A
Table 2. Lgol Hydrogenolysis
Selectivity [%]
Catalyst
Conversion Cis-
Trans-
2MTHFA THP2M5one THP2M5H
[%]
1% Pd/Si-Al 89.9
THFDM THFDM
42.5
-
15.4
-
17.1
-
0
0
3.4
-
18.9
0
20.1
-
0
1% Pd/Al2O3
Si-Al
0
31.7
0
17.4
3.4*
11.7
1% Pd/Al2O3 40.9
+ Si-Al
18.6
Conditions: 150°C, 500 psi H2, 10 mL 60 mM Lgols in THF (exo/endo = 1.7), 100 mg cat, 3h
reaction time. Si-Al calcined at 400oC prior to reaction.
*The reported selectivity towards THFDM over the Si-Al catalyst is most likely an impurity, as no
metal sites are present to carry out hydrogenolysis.
3.3.2. Selectivity versus conversion
The reaction network for Lgol hydrogenolysis was further investigated by carrying out sampling
from a batch reactor over time. As calculated from Figure 3, the selectivity of Lgol conversion to
THFDM was ~55% and was not a strong function of conversion. Because the selectivities to the
side-products 2MTHFA and THP2M5H were also not strong functions of conversion, these
products are produced in parallel with THFDM rather than in series (Scheme 3). A separate
experiment with THFDM as feedstock (10 mL 60 mM THFDM feed, cis/trans = 4.8) at the same
reaction conditions (150°C, 500 psi H2, 50 mg 1% Pd/Si-Al, 3h reaction time) displayed near-zero
conversion with small amounts of 2MTHFA produced (4% yield), providing further evidence that
the dominant pathway for 2MTHFA formation is in parallel with THFDM rather than in series.
The concentration of THP2M5one is nearly zero throughout the reaction. Based on the data in
Figure 3, the initial turnover frequency (TOF) for Lgol hydrogenolysis to THFDM at these
conditions was 0.84 mol THFDM per mol surface Pd per min. Catalyst recycling data suggest that
this catalyst undergoes deactivation under these conditions similar to the case of LGO
hydrogenation (Supplemental Information, Figure S8). The carbon balance decreased to ~85%
during the reaction, indicating that minor degradation reactions are occurring at these reaction
conditions.

Green Chemistry
Page 10 of 17
View Article Online
DOI: 10.1039/C6GC03028A
Figure 3. Hydrogenolysis of Lgol (exo/endo = 1.7) over 450 mg 1 wt% Pd/Si-Al in a batch reactor
with dip-tube sampling. Conditions: 150°C, 500 psi H2, 60 mL 60 mM Lgol in THF feed. Solid
lines between points are visual aids.
3.3.3. Effect of Lgol stereochemistry
We investigated the relationship between the Lgol feedstock stereoisomer ratio and the THFDM
product stereoisomer ratio. We produced a Lgol feedstock containing an exo/endo ratio of 4.7
using 0.4 wt% Pd/Al2O3 catalyst (Figure 2). When the Lgol hydrogenolysis reaction was carried
out using this feedstock (Figure 4), the THFDM product stereoisomer ratio was nearly equal to the
ratio obtained in the experiment shown in Figure 3 (cis/trans = 2.5 vs cis/trans = 2.6). This result
indicates that the mechanism of Lgol hydrogenolysis to THFDM passes through an intermediate
which erases the stereochemistry of the feed. Furthermore, when THFDM (cis/trans = 4.8) was fed
over the Pd/Si-Al catalyst at the same reaction conditions (as discussed above), the THFDM
stereoisomer ratio did not change. This result indicates that THFDM production is irreversible,
and the THFDM stereoisomer ratio is governed by kinetics rather than thermodynamics.

Page 11 of 17
Green Chemistry
View Article Online
DOI: 10.1039/C6GC03028A
Figure 4. Hydrogenolysis of Lgol (exo/endo = 4.5) over 450 mg 1 wt% Pd/Si-Al in a batch reactor
with dip-tube sampling. Conditions: 150°C, 500 psi H2, 60 mL 60 mM Lgol in THF feed. Solid
lines between points are visual aids.
3.3.4. Determining whether THP2M5one is a THFDM precursor
THP2M5one can be produced from the acid-catalyzed isomerization of Lgol in the absence of a
metal catalyst, via cleavage of the C-O bond of the anhydro-bridge. Amberlyst 70, a solid acid
catalyst, was used to generate a product mixture containing 30.8 mM Lgol and 21.7 mM
THP2M5one (see Supplemental Information), and this mixture was fed over a Pd/Si-Al catalyst
(Figure 5) to determine whether THP2M5one is an intermediate between Lgol and THFDM. An
additional experiment with pure Lgol at a similar initial concentration (29.8 mM) was conducted
to provide a direct comparison (Figure 6).
The THFDM concentrations over time are comparable in Figures 5 and 6. If it is assumed that
Lgol is the sole precursor to THFDM, the selectivity of Lgol to THFDM is nearly identical in
Figure 5 (63%) versus Figure 6 (59%). In contrast, the THP2M5H concentration is significantly
higher in Figure 5 than in Figure 6. If it is assumed that the THP2M5one present in Figure 5 is just
hydrogenated to THP2M5H, and additionally that the Lgol in Figure 5 is converted to THP2M5H
in 35% selectivity (as in Figure 6), then the selectivity of THP2M5one conversion to THP2M5H
in Figure 5 is 88%. These results indicate that THP2M5one is hydrogenated to THPM2H and is
not a THFDM precursor (Scheme 3). We note that we cannot rule out the possibility of a second
route from Lgol to THP2M5H which does not pass through THP2M5one. Because the 2MTHFA
concentrations are comparable in Figures 5 and 6, 2MTHFA is not produced through THP2M5one
or THP2M5H.
It should be noted that the product deconvolution methods are likely to be subject to greater error
when the concentrations of cis-THFDM and THP2M5one are comparable (see Supplemental
Information), as in Figure 5. One potential source of error is in the assumption that the GC

Green Chemistry
Page 12 of 17
View Article Online
DOI: 10.1039/C6GC03028A
sensitivities of THP2M5one and Lgol are equal. Any error in the measured concentration of
THP2M5one would translate to an error in the concentration of THFDM. However, small errors
in the THP2M5one and THFDM concentrations would not affect the interpretation of Figure 5.
Figure 5. Hydrogenolysis of a feedstock containing THP2M5one and Lgol, over 450 mg 1 wt%
Pd/Si-Al in a batch reactor with dip-tube sampling. Conditions: 150°C, 500 psi H2, 60 mL feed
containing 30.8 mM Lgol and 21.7 mM TH2m5one in THF. Time zero defined as time when the
reaction temperature was reached. Solid lines between points are visual aids.
Figure 6. Hydrogenolysis of Lgol (exo/endo = 1.7) over 450 mg 1 wt% Pd/Si-Al in a batch reactor
with dip-tube sampling. Conditions: 150°C, 500 psi H2, 60 mL 30 mM Lgol in THF feed. Solid
lines between points are visual aids.

Page 13 of 17
Green Chemistry
View Article Online
DOI: 10.1039/C6GC03028A
3.3.5. Proposed mechanism
A mechanism for the hydrogenolysis of Lgol to THFDM that is consistent with the above findings
is shown in Scheme 4. The above experiments show that Lgol hydrogenolysis does not proceed
through initial isomerization of Lgol to THP2M5one, nor through the hydrogenated side-product
THP2M5H. We suggest that the mechanism proceeds through initial cleavage of the anhydro-
bridge of Lgol to form a surface intermediate containing a single 6-membered ring. This species
could potentially be a partially hydrogenated intermediate as shown in Scheme 4.
Secondly, this species is ring-opened via protonation of the pyran ring oxygen to form an acyclic
surface intermediate, similar to the hydrogenolysis of tetrahydropyran-2-methanol (THP2M) to
1,6-hexanediol.28 The observation that the stereochemistry of the Lgol feed is erased in this
reaction supports the hypothesis that the ring rearrangement occurs through an acyclic intermediate
which contains only one chiral center, as opposed to a concerted ring rearrangement (we note that
loss of the hydrogen atom at the C5 position of Lgol is another way to erase the stereochemistry of
the feed). Ring rearrangement occurs via nucleophilic attack of the carbocation at the C5 position
by the alcohol oxygen at the C2 position. The stereochemistry of THFDM is determined by the
spatial direction of this ring closure step. Ring closure from the less hindered direction results in
the cis-THFDM configuration, consistent with the observation that a 2.5:1 ratio of cis-THFDM to
trans-THFDM is observed. The observation that close proximity of metal and acid sites promotes
this reaction suggests that the hydrogen atom addition and ring rearrangement steps may need to
occur in concert in order to carry out this reaction effectively.
Scheme 4: Proposed mechanism for Lgol hydrogenolysis to THFDM
4. Conclusions
We have shown that LGO can be selectively hydrogenated to Cyrene at low temperature (40°C),
and further hydrogenated to two stereoisomers of Lgol at higher temperatures (100°C). An excess
of the exo-Lgol isomer is favored over supported Pd catalysts across multiple particle sizes and
temperatures, and the stereoisomer ratio is kinetically controlled. Lgol can be further upgraded to
THFDM at 150°C in 58% selectivity over a bifunctional Pd/Si-Al catalyst, with an overall
selectivity of 78% to 1,6-hexanediol precursors. An excess of cis-THFDM to trans-THFDM was
produced over Pd/Si-Al, independent of the stereoisomer ratio of the Lgol feedstock, suggesting
that the initial stereochemistry at the alcohol position of Lgol is erased during the reaction.

Green Chemistry
Page 14 of 17
View Article Online
DOI: 10.1039/C6GC03028A
Although THP2M5one can be produced from the acid-catalyzed isomerization of Lgol,
THP2M5one is hydrogenated to THP2M5H, and neither of these species are THFDM precursors.
Our results demonstrate that LGO has the potential to become a platform molecule used for the
production of high-value oxygenated chemicals, such as Cyrene, Lgol, or THFDM, from
lignocellulosic biomass.
5. Acknowledgements
This material is based upon work supported by the Department of Energy, Office of Energy
Efficiency and Renewable Energy (EERE), under Award Number DE-EE0006878. S.H.K.
acknowledges that this material is based upon work supported by the National Science Foundation
under Grant No. DGE-1256259. We thank the UW-Madison Department of Chemistry for use of
Bruker Avance 500 MHz NMR Spectrometer. A generous gift from Paul J. Bender enabled this
spectrometer to be purchased [2012].
6. References
1.
2.
G. W. Huber, S. Iborra and A. Corma, Chemical Reviews, 2006, 106, 4044-4098.
J. N. Chheda, G. W. Huber and J. A. Dumesic, Angewandte Chemie International Edition,
2007, 46, 7164-7183.
3.
4.
5.
6.
F. Cao, T. J. Schwartz, D. J. McClelland, S. H. Krishna, J. A. Dumesic and G. W. Huber,
Energy & Environmental Science, 2015, 8, 1808-1815.
H. Kawamoto, S. Saito, W. Hatanaka and S. Saka, Journal of Wood Science, 2007, 53,
127-133.
F. Shafizadeh, R. H. Furneaux and T. T. Stevenson, Carbohydrate Research, 1979, 71,
169-191.
J. Sherwood, M. De bruyn, A. Constantinou, L. Moity, C. R. McElroy, T. J. Farmer, T.
Duncan, W. Raverty, A. J. Hunt and J. H. Clark, Chemical Communications, 2014, 50,
9650-9652.
7.
8.
M. De bruyn, J. Fan, V. L. Budarin, D. J. Macquarrie, L. D. Gomez, R. Simister, T. J.
Farmer, W. D. Raverty, S. J. McQueen-Mason and J. H. Clark, Energy & Environmental
Science, 2016, 9, 2571-2574.
J. Zhang, G. White, M. Ryan, A. J. Hunt and M. J. Katz, ACS Sustainable Chemistry &
Engineering, 2016.
9.
USA Pat., US 8865,940 B2, 2014.
10.
T. Buntara, I. Melián-Cabrera, Q. Tan, J. L. G. Fierro, M. Neurock, J. G. de Vries and H.
J. Heeres, Catalysis Today, 2013, 210, 106-116.
11.
12.
13.
T. Buntara, S. Noel, P. H. Phua, I. Melián-Cabrera, J. G. de Vries and H. J. Heeres, Topics
in Catalysis, 2012, 55, 612-619.
T. Buntara, S. Noel, P. H. Phua, I. Melián-Cabrera, J. G. de Vries and H. J. Heeres,
Angewandte Chemie International Edition, 2011, 50, 7083-7087.
M. R. Nolan, G. Sun and B. H. Shanks, Catalysis Science & Technology, 2014, 4, 2260-
2266.
14.
15.
16.
M. M. Zanardi and A. G. Suárez, Tetrahedron Letters, 2009, 50, 999-1002.
C. Moreau, M. N. Belgacem and A. Gandini, Topics in Catalysis, 2004, 27, 11-30.
1,6-Hexanediol Market by Application (Polyurethanes, Coatings, Acrylates, Adhesives,
Unsaturated Polyester Resins, Plasticizers, and Others) and By Geography (NA, Europe,

Page 15 of 17
Green Chemistry
View Article Online
DOI: 10.1039/C6GC03028A
Asia-Pacific,
&
ROW)
-
Trends
and
Forecasts
to
2019,
http://www.researchandmarkets.com/research/zs4gnb/16hexanediol, (accessed January
22, 2016).
17.
18.
S. Van de Vyver and Y. Roman-Leshkov, Catalysis Science & Technology, 2013, 3, 1465-
1479.
P. Werle, M. Morawietz, S. Lundmark, K. Sörensen, E. Karvinen and J. Lehtonen, in
Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA,
2000.
19.
B. Xiao, M. Zheng, X. Li, J. Pang, R. Sun, H. Wang, X. Pang, A. Wang, X. Wang and T.
Zhang, Green Chemistry, 2016, 18, 2175-2184.
20.
21.
J. Tuteja, H. Choudhary, S. Nishimura and K. Ebitani, ChemSusChem, 2014, 7, 96-100.
R. Alamillo, M. Tucker, M. Chia, Y. Pagan-Torres and J. Dumesic, Green Chemistry,
2012, 14, 1413-1419.
22.
23.
E. Ronzón and G. Del Angel, Journal of Molecular Catalysis A: Chemical, 1999, 148, 105-
115.
P. Mäki-Arvela, J. Hájek, T. Salmi and D. Y. Murzin, Applied Catalysis A: General, 2005,
292, 1-49.
24.
25.
R. L. Augustine, Catalysis Today, 1997, 37, 419-440.
X. Hu, R. J. M. Westerhof, L. Wu, D. Dong and C.-Z. Li, Green Chemistry, 2015, 17, 219-
224.
26.
27.
28.
F. Leydier, C. Chizallet, A. Chaumonnot, M. Digne, E. Soyer, A.-A. Quoineaud, D. Costa
and P. Raybaud, Journal of Catalysis, 2011, 284, 215-229.
E. J. M. Hensen, D. G. Poduval, P. C. M. M. Magusin, A. E. Coumans and J. A. R. v. Veen,
Journal of Catalysis, 2010, 269, 201-218.
M. Chia, Y. J. Pagán-Torres, D. Hibbitts, Q. Tan, H. N. Pham, A. K. Datye, M. Neurock,
R. J. Davis and J. A. Dumesic, Journal of the American Chemical Society, 2011, 133,
12675-12689.
29.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman,
G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M.
Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. J. Bearpark, J.
Heyd, E. N. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K.
Raghavachari, A. P. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega,
N. J. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R.
Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J.
Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J.
Cioslowski and D. J. Fox, Gaussian 09, Revision C.01, Gaussian, Inc.: Wallingford, CT,
2009.
30.
31.
NIST Chemistry WebBook, http://webbook.nist.gov/, (accessed 10/1/16, 2016).
D. W. Rogers, Y. Zhao, M. Traetteberg, M. Hulce and J. Liebman, The Journal of Chemical
Thermodynamics, 1998, 30, 1393-1400.
32.
M. S. Miftakhov, I. N. Gaisina, F. A. Valeev and O. V. Shitikova, Russian Chemical
Bulletin, 1995, 44, 2350-2352.

Green Chemistry
Page 16 of 17
View Article Online
DOI: 10.1039/C6GC03028A
33.
34.
T. J. Connolly, J. L. Considine, Z. Ding, B. Forsatz, M. N. Jennings, M. F. MacEwan, K.
M. McCoy, D. W. Place, A. Sharma and K. Sutherland, Organic Process Research &
Development, 2010, 14, 459-465.
A. Tagirov, I. Biktagirov, Y. Galimova, L. Faizullina, S. Salikhov and F. Valeev, Russian
Journal of Organic Chemistry, 2015, 51, 569-575.

Page 17 of 17
Green Chemistry
View Article Online
DOI: 10.1039/C6GC03028A
Table of Contents Graphic and Text
Herein, we elucidate the reaction network for catalytic hydrogenation of the biomass-derived intermediate
levoglucosenone into several different renewable chemicals.