Cleavage and hydrodeoxygenation (HDO) of C-O bonds relevant to lignin conversion using Pd/Zn synergistic catalysis

Article abstract of DOI:10.1039/c2sc21657d

The development of chemical methods for the direct catalytic conversion of biomass to high value organic molecules is an area of increasing interest. The plant matter component known as lignin is a polymer consisting of aromatic rings that could provide a means of obtaining aromatic materials currently derived solely from petroleum. This report describes a bimetallic Pd/C and Zn catalytic system that can perform selective hydrodeoxygenation (HDO) of monomeric lignin surrogates as well as successfully cleave the β-O-4 linkages found in dimeric lignin model complexes and synthetic lignin polymers with near quantitative conversions and yields between 80-90%. The reaction with lignin polymer was highly selective affording methoxy substituted propylphenol as the major product. These reactions were performed in a Parr reactor operating at relatively mild temperature (150 °C) and pressure (20 bar H₂) using methanol as a solvent. Reaction products were characterized using high-pressure liquid chromatography coupled to a linear quadrupole ion trap mass spectrometer equipped with an electrospray ionization source using negative ion mode. Hydroxide ions were doped into the analyte solutions to encourage negative ion formation. This method ionizes all the mixture components to yield a single ion/analyte with no fragmentation. The catalyst is fully recyclable without the need for additional zinc. X-ray absorption spectroscopy (EXAFS) is consistent with Pd nanoparticles (4-5 nm) and no evidence of Pd-Zn alloy formation. A mechanistic hypothesis on the synergy between Pd and Zn is presented. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c2sc21657d

Chemical Science
View Article Online
View Journal
EDGE ARTICLE
Cleavage and hydrodeoxygenation (HDO) of C–O bonds
relevant to lignin conversion using Pd/Zn synergistic
catalysis†
Cite this: DOI: 10.1039/c2sc21657d
Trenton H. Parsell,ab Benjamin C. Owen,ab Ian Klein,ab Tiffany M. Jarrell,ab
Christopher L. Marcum,ab Laura J. Haupert,ab Lucas M. Amundson,ab
ab
Hilkka I. Kenttamaa,ab Fabio Ribeiro,bc Jeffrey T. Millercd and Mahdi M. Abu-Omar
*
¨
The development of chemical methods for the direct catalytic conversion of biomass to high value organic
molecules is an area of increasing interest. The plant matter component known as lignin is a polymer
consisting of aromatic rings that could provide a means of obtaining aromatic materials currently
derived solely from petroleum. This report describes a bimetallic Pd/C and Zn catalytic system that can
perform selective hydrodeoxygenation (HDO) of monomeric lignin surrogates as well as successfully
cleave the b-O-4 linkages found in dimeric lignin model complexes and synthetic lignin polymers with
near quantitative conversions and yields between 80–90%. The reaction with lignin polymer was highly
selective affording methoxy substituted propylphenol as the major product. These reactions were
performed in a Parr reactor operating at relatively mild temperature (150 ꢀC) and pressure (20 bar H2)
using methanol as
a solvent. Reaction products were characterized using high-pressure liquid
chromatography coupled to a linear quadrupole ion trap mass spectrometer equipped with an
electrospray ionization source using negative ion mode. Hydroxide ions were doped into the analyte
solutions to encourage negative ion formation. This method ionizes all the mixture components to yield
a single ion/analyte with no fragmentation. The catalyst is fully recyclable without the need for
additional zinc. X-ray absorption spectroscopy (EXAFS) is consistent with Pd nanoparticles (4–5 nm) and
no evidence of Pd–Zn alloy formation. A mechanistic hypothesis on the synergy between Pd and Zn is
presented.
Received 3rd October 2012
Accepted 8th November 2012
DOI: 10.1039/c2sc21657d
www.rsc.org/chemicalscience
fuel in the US alone.2 The most developed transformations of
biomass are based on converting sugars into alcohol but more
recently the focus has shied onto making hydrocarbons from
cellulose3 as well as utilizing the other major component of
plant biomass, lignin.
Introduction
The push to replace petroleum-derived materials as both fuels
and chemical feedstock is motivated by many factors including
climate concerns, constantly growing energy demand, and
energy security. One material that exhibits much promise as a
source of renewable carbon feedstock is lignocellulosic biomass
and thus it has become the target of many processes designed to
turn biomass into fuels and/or useful synthetic building
blocks.1 Non-food biomass is estimated as having the potential
to produce the annual equivalent of 3 billion barrels of liquid
Of the three main biopolymers that make up the cell wall,
lignin is the most recalcitrant. It comprises up to 25% of the
plant's mass and is needed for support and water transport.4
While lignin recalcitrance has been looked upon as a major
obstacle in the production of bioethanol and other alcohols, it
presents an opportunity because it accounts for 40% of the
energy content of biomass.5 The building blocks (monomers)
that make up lignin are based on three types of 4-propenyl-
phenols, making it the only biorenewable source of aromatics.
Furthermore, lignin's high aromatic content is responsible for
its high energy content. A number of different linkages occur
naturally in lignin's structure with the most abundant being the
b-O-4 ether linkage (Fig. 1).6 Therefore, catalytic methods to
break down lignin while preserving its aromatic nature have the
potential to provide a valuable stream of material that is
currently obtained from petroleum sources exclusively.
aBrown Laboratory, Department of Chemistry, Purdue University, 560 Oval Drive, West
Lafayette, In 47907, USA. E-mail: mabuomar@purdue.edu
bThe Center for direct Catalytic Conversion of Biomass of Biofuels (C3Bio), Purdue
University, Discovery Park, 1203 West State Street, West Lafayette, IN 47907, USA
cSchool of Chemical Engineering, Purdue University, 480 Stadium Mall Drive, West
Lafayette, IN 47907, USA
dChemical Sciences and Engineering Division, Argonne National Laboratory, 9700
South Cass Avenue, Argonne, IL 60439, USA
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c2sc21657d
This journal is ª The Royal Society of Chemistry 2012
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Experimental section
Materials
Vanillyl alcohol 98%, veratryl alcohol 96%, vanillin 99%, ZnCl2,
and 5 wt% Pd/C were purchased from Aldrich (Milwaukee, WI)
and used as received. Guaiacylglycerol-b-guaiacyl ether 97% was
purchased from TCI America and used as received. The b-O-4
type articial lignin polymer was synthesized via the methods
previously reported by Ubukata et al.16
Fig. 1 b-O-4 the most prevalent linkage in lignin.
General in situ generated catalyst reactions
In a typical experiment: substrate, 5 wt% Pd/C (5–10 wt%),
ZnCl2 (1–2 wt%), and methanol (15 mL) were added to a dry
glass sleeve. The sleeve was placed into a stainless steel Parr
reactor, which was subsequently sealed. While stirring, the
mixture was purged with UHP grade H2 for ꢁ1 min., then
pressurized with H2 (30–300 psi, 2–20.4 bar). The mixture was
Currently lignin is underutilized with the bulk of produced
lignin being employed as combustible material for its high heat
value.7 One of the reasons for this limited use is the current
methods of extracting lignin from biomass are quite harsh and
lead to modication making its conversion extremely difficult.
It is worth noting that recent developments in lignin treatment
have been shown to produce lignin that is relatively unmodi-
ed, providing a better starting point for further chemical
transformations.8
Reductive disassembly/“unzipping” of lignin followed by
hydrodeoxygenation (HDO) has the advantage of increasing the
fuel value of the product by adding H2. Hydrogenation of lignin
under high pressure and temperature conditions was described
many decades ago and continues to be investigated.9,10 More
recently reductive methods under pyrolysis conditions with
formic acid at temperatures above 350 ꢀC give bio-oils with
lower O/C ratio than the starting lignin.11 Lercher and co-
workers have demonstrated hydrodeoxygenation of phenolic
bio-oil fractions to alkanes using a combination of a hydroge-
nation catalyst and a strong acid.12 Ford et al. studied hydrogen
transfer from supercritical methanol for catalytic conversion of
organosolv lignin over Cu PMO (porous metal oxides) catalyst at
300 ꢀC.13 On the homogeneous catalysis front, recent investi-
gations on model lignin compounds have demonstrated the
utility of ruthenium hydrogenation catalysts14 and nickel car-
bene hydrogenation catalysts.15 In the latter case, high nickel
loading of 25 mol% and strong base were required to affect the
difficult aryl ether bond cleavage.
ꢀ
heated to 150 C. This temperature was maintained between 2
to 8 hours depending on the substrate and catalyst loading. The
reaction was terminated by removing the heat and cooling the
reactor to room temperature (WARNING! Use caution when
handling and venting the reactor; Pd/C is pyrophoric). The glass
sleeve was removed from the Parr reactor and the mixture was
ltered to remove Pd/C. The Pd/C was washed with additional
MeOH and the ltrate was collected and diluted in a volumetric
ask. A known amount of 9,10-dihydroanthracene or 1,4-
dimethoxybenzene was added as an internal standard, and the
solution was analyzed by GC-FID and HPLC/MS as described
below.
Catalyst recycling experiments
Aer an experiment as the one described above, the ltered
solid was dried under vacuum, weighed, and reused under the
same conditions described above.
Control experiments
The above reactions were performed under varying conditions.
Having just ZnCl2 or Pd/C in the reaction vessel (or no catalyst at
all) caused the conversion of vanillyl alcohol to vanillyl methoxy
ether. When the above reactions were performed under an N2
atmosphere no reaction was observed.
Performing HDO to remove covalently bonded oxygen from
lignin offers opportunities to not only break apart the lignin's
complex structure but also to increase the overall energy density
and value of the resulting products. The difficulty with this type
of reaction is that the C–O and C]C bond strengths are
Catalyst preparation
comparable, making selective HDO versus aromatic hydroge- In a typical preparation 5 wt% Pd/C (0.250 g, 0.12 mmole Pd)
nation challenging. Hydrogenation of the aromatic ring negates was placed into a dry glass sleeve with a glass stirbar and 15 mL
some of the innate value lignin possesses and also uses up of MeOH. To this mixture Zn(OAc)2$6H2O (or ZnCl2) (0.0510 g,
additional equivalents of valuable H2. In this study, we describe 0.232 mmol) was added. The sleeve was then placed into a
a bimetallic catalyst system based on metallic Pd nanoparticles stainless steel Parr reactor. The reactor was purged with H2 for
(NP) on carbon and Zn(II) Lewis acid that can selectively cleave approximately 2 min, sealed, and pressurized to 300 psi with H2.
C–O bonds in a variety of monomeric, dimeric, and polymeric The reactor was then heated to 150 ꢀC and allowed to stand for 8
ꢀ
lignin molecules under relatively mild conditions, 150 C and hours. The reactor was cooled, vented, and the solid was ltered
30–300 psi of H2. A combination of control experiments and and washed with additional MeOH and then dried under
spectroscopic characterization of the catalyst (EXAFS) under vacuum (use caution when handling Pd/C as it is pyrophoric). A
relevant reaction conditions points to a synergistic effect portion of the ltered solid was sent to Galbraith Labs for
between Zn(II) and Pd without formation of a Zn–Pd alloy.
elemental analysis to determine the amount of Pd and Zn in the
Chem. Sci.
This journal is ª The Royal Society of Chemistry 2012

View Article Online
Edge Article
Chemical Science
catalyst, which were found to be 4.7% and 2.5%, respectively. A (CAD) to be performed for the structural determination of the
portion of the prepared catalyst sample was also sent to HPLC-separated product mixtures.
Argonne National Laboratory for additional analysis by X-ray
Absorption Spectroscopy.
Pd K-edge extended X-ray absorption ne structure (EXAFS)
measurements
Reactions with Zn/Pd/C catalyst
X-ray absorption measurements were made on the insertion-
In a typical experiment: substrate, 5 wt% Zn/Pd/C, and meth-
anol (15 mL) were added to a dry glass sleeve. The sleeve was
placed into a stainless steel Parr reactor, which was subse-
quently sealed. While stirring the mixture was purged with UHP
grade H2 for ca. 1–2 min., then pressurized with H2 (30–300 psi,
device beamline of the Materials Research Collaborative Access
Team (MRCAT) at the Advanced Photon Source, Argonne
National Laboratory. Measurements were made in transmission
mode with the ionization chambers optimized for the
maximum current with linear response (ꢁ1010 photons detec-
ted per s) using a mixture of N2 and He in the incident X-ray
detector and a mixture of ca. 20% Ar in N2 in the transmission X-
ray detector. A third detector in the series collected a Pd foil
spectrum simultaneously with each measurement for energy
calibration. Catalyst samples were pressed into a cylindrical
holder with a thickness chosen to give a total absorbance (mx) at
the Pd K (24.350 keV) edge of about 2.0 and a Pd edge step (Dmx)
of ca. 0.5. The Pd/C catalyst before reaction was measured in air
at room temperature. The Pd/C + ZnCl2 catalyst aer reaction
was unloaded in the absence of air, placed in an air-tight
container and loaded in the EXAFS cell in a glove box at the
beamline. The used catalyst (without air exposure) was also re-
reduced at 150 and 250 ꢀC for 30 min in 4% H2/He followed by
He purge, and cooled to room temperature (RT) in a contin-
uous-ow EXAFS reactor cell. Trace oxidants in He were
removed by passing through a Matheson PUR-Gas Triple Puri-
er Cartridge. The spectra were obtained at RT under static He
atmosphere. Phase shis, backscattering amplitudes and
ꢀ
2–20.4 bar). The mixture was heated 150 C. This temperature
was maintained for 2 h, aer which heat was removed and the
reactor cooled to room temperature (WARNING! Use caution
when venting the reactor and handling Pd/C as it is pyrophoric).
The glass sleeve was removed from the Parr reactor and the
mixture was ltered to remove Zn/Pd/C. The Zn/Pd/C was
washed with additional MeOH and the ltrate was collected and
diluted in a volumetric ask. A known amount of 9,10 dihy-
droanthracene was added as an internal standard, and the
solution was analyzed by GC-FID and HPLC/MS as described
below.
GC-FID analysis
Gas chromatography was performed with an Agilent Technol-
ogies 6890N Network GC System with a J&W Scientic DB-5
capillary column. Oven temperature was ramped from 50 to
250 ꢀC at 15 ꢀC minꢂ1, and from 250 to 300 ꢀC at 25 ꢀC minꢂ1 in
constant ow mode at 1.0 mL minꢂ1. Products in the GC trace
were identied by retention time comparison to standards and
quantied by use of a standard calibration curve.
˚
XANES references were obtained from PdO for Pd–O (4 at 2.05 A)
˚
Pd foil for Pd–Pd (12 at 2.75 A). For Pd–Zn, phase and amplitude
les were calculated from FEFF. Values for So and Ds2 were
determined by tting the experimental data from Pd foil with
calculated Pd–Pd phase and amplitude FEFF functions. Stan-
dard procedures based on WINXAS97 soware were used to
extract the EXAFS data. The rst shell coordination parameters
were obtained by a least square t in r-space of the k2-weighted
Fourier transform data.
High-resolution HPLC/MSn analysis
All samples were introduced into the HPLC/MS (a linear quad-
rupole ion trap) via an autosampler as a full-loop injection
volume (25 mL) for high reproducibility. A 0.001% (m/v) ammo-
nium formate in water solution (A) and 0.001% (m/v) ammonium
formate in acetonitrile solution (B) were used as the mobile phase
solvents. Ammonium formate was used to encourage negative
ion production upon ESI. A Zorbax SB-Phenyl reversed-phase
HPLC column was used to separate the reaction mixtures. A non-
Results and discussion
Simple phenolic substrates
linear gradient elution was used for this column. The column was Table 1 shows the results of reactions performed on 4-(hydrox-
equilibrated with the start eluent for 10 min before each run and ymethyl)-2-methoxyphenol, 4-hydroxy-3-methoxybenzaldehyde,
placed in a thermostatted column compartment that maintained and 4-(methoxymethyl)-2-methoxyphenol as monomeric lignin
a temperature of 30 ꢀC to increase the reproducibility of the surrogates. The results show that when both Pd/C and Zn2+ are
retention times and peak widths.
present the benzyl alcohol and aldehyde groups can be selectively
The PDA detector for HPLC was set at 214 nm. The condi- deoxygenated in good yields without hydrogenation of the
tions for ionization of the analytes and ion injection into the phenyl ring. Experiments that used either Pd/C or Zn2+ inde-
mass spectrometer were optimized using a stock solution of pendently showed formation of 4-(methoxymethyl)-2-methox-
2-methoxy-4-propylphenol in a 0.015% NaOH 50 : 50 (m/v) yphenol, which is a result of benzyl alcohol protection with
acetonitrile–water solution. For mass spectrometric detection, MeOH, the solvent. This type of protection/activity is known17
data-dependent scans were used on a 7 Tesla Thermo Scientic and can be promoted by a hydrogenation catalyst. It is worth
LQIT/FT-ICR mass spectrometer allowing for high-resolution noting that once this product is formed it can no longer be
exact mass measurements and tandem mass spectrometry deoxygenated, suggesting that HDO is either faster than
(MSn) experiments based on collision-activated dissociation protection or the synergy between the Pd/C and Zn2+ catalyst
This journal is ª The Royal Society of Chemistry 2012
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Table 1 Catalytic hydrogenation of monomeric lignin surrogates.a
Reactant
Catalyst
%Conversionb
>99
Product
%Product yieldb
Zn/Pd/C
80
Pd/C
>99
ꢁ60c
ZnCl2
>99
>99
NR
ꢁ60c
Zn/Pd/C
Zn/Pd/C
81
N/Ad
a
b
Reaction conditions: Parr reactions @ 300 psi H2, 150 ꢀC, 8 h in MeOH. Conversion and yield based on GC analysis aer ltration to remove
c
catalyst. A standard for these reaction could not be obtained so yields in these reactions are approximated using the response factor of 3,4-
dimethoxybenzyl alcohol. This reactivity was tested by taking a MeOH solution of the product generated by reacting 4-(hydroxymethyl)-2-
d
methoxyphenol with 5% Pd/C ltered to remove any solid, then adding Pd/C and ZnCl2 to the solution and running the reaction under the
conditions listed above.
prevents the formation of the ether. Changing the counter ion type of general Lewis acid catalysis, we tested AlCl3 and SnCl2 as
introduced with zinc from chloride to acetate showed no effect additives in place of Zn2+. No effect was observed in either case.
on the catalyst activity or reaction selectivity.
Reactions with 4-(hydroxymethyl)-2-methoxyphenol produced
The robustness and ability to recycle the catalyst was tested 4-(methoxymethyl)-2-methoxyphenol as the major product.
by ltering the Zn/Pd/C and simply reusing it in a new reaction. Hence, the catalytic enhancement and HDO selectivity is Zn
Selectivity and activity were maintained for at least three specic. It is also worth noting that the solvent inuences what
consecutive cycles with no need to add additional Zn2+ to the products are formed. When the solvent is CH2Cl2 and a Lewis
reaction mixture. Elemental analysis of the ltered catalyst acid is present the aromatic ring is readily hydrogenated to give
shows that the Pd on carbon is picking up Zn2+ from the solu- cyclohexanol.17 However, in methanol hydrogenation of the
tion. Furthermore, the activity and selectivity of recycled catalyst aromatic ring is suppressed.
with Zn2+ added in situ, and a priori prepared Zn/Pd/C in the
presence of H2 but absence of substrate are identical.
b-O-4 Dimeric and polymeric substrates
Zhou and coworkers have shown recently that Lewis acids
can be used in conjunction with Pd/C to help control the The b-O-4 linkage of the lignin macromolecule is the most
selectivity of phenols hydrogenation.18 To probe for a similar abundant repeating subunit. Therefore, devising a catalyst that
Chem. Sci.
This journal is ª The Royal Society of Chemistry 2012

View Article Online
Edge Article
Chemical Science
amount of 3-(3-methoxy-4-hydroxyphenyl)-1-propanol (eqn (1)).
Fig. 2 shows the GC-FID and HPLC/UV chromatograms of the
products. The minor product 3-(3-methoxy-4-hydroxyphenyl)-1-
propanol was not detected by GC. The identity of the products
was established by high-resolution tandem mass spectrometry
(HPLC/MS).
Carrying out the reaction for longer periods of time caused
no decrease in the amount of 3-(3-methoxy-4-hydroxyphenyl)-1-
propanol (minor product) produced. When 3-(4-hydroxy-
phenyl)-1-propanol was allowed to react under the same
conditions, conversion to 4-propylphenol was not observed.
Therefore, it can be concluded that 3-(3-methoxy-4-hydroxy-
phenyl)-1-propanol is not an intermediate en route to the
major product 2-methoxy-4-propylphenol. Carrying out the
reaction with guaiacylglycerol-b-guaiacyl ether under the same
conditions without the Zn2+ co-catalyst leads to signicantly
less b-O-4 cleavage. The major components of the reaction
mixture were starting material or a methoxy protected form of
the starting material with less than 5% conversion to guaiacol
and 3-(3-methoxy-4-hydroxyphenyl)-1-propanol (HPLC shown in
the ESI†).
Given the success of the Zn/Pd/C catalyst, it was tested on a
b-O-4 synthetic lignin polymer that has a Mn of 3390 and DPn of
12.1 (eqn (2)).16 This polymer was cleaved and hydro-
deoxygenated under the same conditions as the aforemen-
tioned dimer.
While the polymer was not initially very soluble in MeOH
(instead becoming sticky), we observed complete conversion
with excellent yields (80%) of two major products, 2-methoxy-4-
propylphenol (56%) and 3-(3-methoxy-4-hydroxyphenyl)-1-
propanol (44%). Fig. 3 displays the HPLC/UV chromatogram of
the reaction products alongside the mass spectrum of each
product. High-resolution exact mass measurements and MS3
experiments were used to identify the products. The ratio of the
two products is different than what was observed for the reac-
tion with the dimer (eqn (1)) above. This difference is explained
on the basis that the polymer chain ends with a propyl alcohol
and a greater amount of 3-(3-methoxy-4-hydroxyphenyl)-1-
propanol is to be expected.
Fig. 2 Caption (A) GC-FID chromatogram and (B) HPLC/UV chromatogram of
the products from the hydrogenation of guaiacylglycerol-b-guaiacyl ether over
Zn/Pd/C.
can selectively cleave this type of ether linkage and undergo
HDO would provide a means of unzipping the very complex
polymeric structure down into smaller molecules that have
higher energy value. To this end we studied the hydro-
deoxygenation of dimeric and polymeric compounds contain-
ing b-O-4 ether linkages.
The model compound guaiacylglycerol-b-guaiacyl ether was
hydrogenated and deoxygenated at 150 ꢀC under 20 bar of
hydrogen using 5 wt% of a prepared Zn/Pd/C catalyst in MeOH
giving complete conversion and yielding primarily two products
guaiacol and 2-methoxy-4-propylphenol in addition to a small
Catalyst characterization
Elemental analysis of the prepared Zn/Pd/C catalyst conrmed
that the Pd/C was indeed picking up and retaining the Zn2+
This journal is ª The Royal Society of Chemistry 2012
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Fig. 3 HPLC/UV and (-)ESI MS analysis of the products resulting from the hydrogenation of a synthetic b-O-4 lignin polymer over Zn/Pd/C catalyst.
added to the reaction mixture. On the samples that were tested XANES and EXAFS spectra of our catalyst were compared to Pd
approximately 50% of the Zn added to the solution was retained foil (Fig. 4) and an authentic Pd–Zn alloy on C.
giving a catalyst of 2.5 wt% Zn and 5 wt% Pd on C. This
composition corresponds to ca. 1 : 1 molar ratio of Pd to Zn.
The EXAFS ts of the catalysts aer reaction were consistent
with only Pd neighbors, i.e., Pd nano-particles and inclusion of
EXAFS and XANES were obtained on the Pd/C catalyst (as a Pd–Zn scattering path gave poor results. We observed no
received in air), aer reaction and aer reaction but reduced at evidence in our catalyst samples of close Pd–Zn neighbors,
higher temperature, e.g., 250 ꢀC. The XANES and EXAFS (not negating the hypothesis that a bimetallic Pd–Zn alloy is the
˚
shown) of the Pd/C catalyst exposed to air, i.e., before reaction, catalyst. The Pd–Pd bond distances of 2.75 A are fully consistent
is partially oxidized. A linear combination XANES t with PdO with palladium foil and the Pd–Pd coordination numbers of
and Pd metal indicates that approximately 30% of the metallic about 10 are consistent with an average particle size of 4–5 nm.
nano-particles are oxidized. The EXAFS also show Pd–O bonds Upon heating our catalyst to 250 ꢀC, which is much higher than
˚
with a t coordination number of 1.7 at 2.05 A consistent with the reaction conditions, changes in both the XANES and the Pd
the XANES. The XANES and EXAFS ts are given in Table 2. The K-edge suggest that a small amount of Pd–Zn alloy is formed,
Table 2 EXAFS and XANES fits of Pd-ZnCl2/C catalysts
Fraction from
XANES
Est.
˚
Sample
Treatment
None
PdO
0.30
Pd
Scatter
N
R, A
Ds2(ꢃ 103)
E0, eV
size, nm
4.5
5% Pd/Ca
0.70
Pd–O
1.7
6.4
8.3
9.7
8.0
2.05
2.74
2.74
2.75
2.75
2.70
2.75
2.75
2.75
2.70
1.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.1
ꢂ1.0
ꢂ1.5
0.3
Pd–Pd
Pd–Pd
Pd–Pd
Pd–Pd
Pd–Zn
Pd–Pd
Pd–Pd
Pd–Pd
Pd–Zn
1b Zn/Pd/C
before catalysis
None
0.05
—
0.95
1.0
4.0
5.0
5.5
H2 150 ꢀC
H2 250 ꢀC
ꢂ1.0
0.3
2.5
5.1
0.3
ꢂ0.8
ꢂ1.2
ꢂ5.5
2c Zn/Pd/C
aer catalysis
None
—
—
1.0
1.0
10.0
10.0
8.4
5.0
5.0
5.5
H2 150 ꢀC
H2 250 ꢀC
1.9
Commercially available 5% Pd/C. b Prepared Zn/Pd/C catalyst generated as described in the experimental section and analyzed before catalysis.
a
c
Pd/C isolated aer catalytic conversion of vanillyl alcohol to 2-methoxy-4-methylphenol.
Chem. Sci.
This journal is ª The Royal Society of Chemistry 2012

View Article Online
Edge Article
Chemical Science
see Table 2, indicating that if PdZn bimetallic nano-particles substrate via binding to OH groups inducing reactivity with Pd–
were formed under reaction conditions, these would have been H on the surface via hydrogen spillover. An alternative mecha-
detected. Furthermore, authentic Pd–Zn alloy catalyst on C gave nistic hypothesis is that the Zn2+ ions are desorbed into solution
low conversion of the model dimer guaiacylglycerol-b-guaiacyl at the reaction temperature of 150 ꢀC. Zinc binds the substrate
ether, similar to results observed for Pd/C alone.
and activates its cleavage upon encountering Pd–H sites on the
catalyst surface. Once the reaction mixture is cooled, the Zn2+
ions are readsorbed onto the C surface. We tested the latter
hypothesis by collecting liquid samples from the reaction under
operating temperature and pressure and analyzed for Zn
content by ICP/MS. Zn was detected in amounts that are
comparable to blank controls where no zinc was added to the
Parr reactor. While this negative result cannot be used to make a
denitive claim that Zn is being mobilized off the catalyst
surface, it favors retention of zinc on carbon. Both of these
mechanistic proposals are presented pictorially in Fig. 5.
Mechanistic implications
The chemical observations in conjunction with XAS are
consistent with Zn adsorbing on carbon, most likely onto –OH
sites, and not forming a Pd–Zn alloy or direct Pd–Zn interaction.
The Zn2+ ions could participate in one of two ways. Activate the
Conclusion
A combined Zn/Pd/C catalyst is far more effective than Pd/C
alone for the cleavage of b-O-4 lignin molecules and the
subsequent hydrodeoxygenation (HDO) of the aromatic frag-
ments without loss of aromatic functional groups. This catalyst
selectively cleaves recalcitrant ether bonds of model lignin
dimers and polymers, removes alcohol oxygen atoms on the
alkyl chains and maintains the valuable aromaticity of the
starting materials, under relatively mild conditions, 150 ꢀC and
300 psi of H2. The catalyst is robust and fully recyclable aer
each catalytic reaction without the need to add additional zinc.
XAS illustrated that the active form of the catalyst is not a
bimetallic Pd–Zn alloy. A mechanistic hypothesis for the
synergy between Zn2+ and Pd is presented in Fig. 5.
2
ꢂ1
˚
Fig. 4 Pd K-edge Fourier transform of Zn/Pd/C catalyst (k : Dk ¼ 2.6 ꢂ 12.3 A
;
solid-magnitude of FT and dotted-imaginary part FT). Blue: Pd foil. Red: Zn/Pd/C
after reaction.
Fig. 5 Proposed pathways for HDO of b-O-4 lignin substrate over Zn/Pd/C catalyst. Pathway (a) involves substrate binding to Zn2+ sites and hydrogen spillover from Pd
sites. Pathway (b) represents desorption of Zn ions into solution, activation of substrate via binding, and it is the Zn bound substrate that reacts with Pd sites on the
catalyst's surface.
This journal is ª The Royal Society of Chemistry 2012
Chem. Sci.

View Article Online
Chemical Science
Edge Article
8 F. Abdelka, H. Ammar, B. Rousseau, M. Tessier,
R. El. Gharbi and A. Fradet, Biomacromolecules, 2011, 12,
3895.
9 E. E. Harris, J. D'Ianni and H. Adkins, J. Am. Chem. Soc., 1938,
60, 1467.
Acknowledgements
This work was supported as part of the Center for Direct Cata-
lytic Conversion of Biomass to Biofuels (C3Bio), an Energy
Frontier Research Center (EFRC) funded by the U.S. Depart-
ment of Energy, Office of Science, Office of Basic Energy
Sciences under Award Number DE-SC0000997.
10 M. Nagy, K. David, G. J. P. Britovsek and B. A. J. Ragauskas,
Holzforschung, 2009, 63, 513.
11 G. Gelletstedt, J. Li, I. Eide, M. Kleinert and T. Barth, Energy
Fuels, 2008, 22, 4240.
12 C. Zhao, Y. Kou, A. A. Lemonidou, X. Li and J. A. Lercher,
Chem. Commun., 2010, 46, 412.
13 (a) G. S. Macala, T. D. Matson, C. L. Johnson, R. S. Lewis,
A. V. Iretskii and P. C. Ford, ChemSusChem, 2009, 2, 215;
(b) K. Barta, T. D. Matson, M. L. Fettig, S. L. Scott,
A. V. Iretskii and P. C. Ford, Green Chem., 2010, 12,
1640.
References
1 J. van Haveren, E. L. Scott and J. Sanders, Biofuels, Bioprod.
Bioren., 2008, 2, 41.
2 M. Paster, J. L. Pellegrino and T. M. Carole, Industrial
Bioproducts: Today and Tomorrow, 2003, Available
electronically
products.html.
at
http://www1.eere.energy.gov/biomass/
´
14 (a) Transfer hydrogenation from alcohols: J. M. Nichols,
L. M. Bishop, R. G. Bergman and J. A. Ellman, J. Am. Chem.
Soc., 2010, 132, 12554; (b) Hydrogenolysis: A. Wu,
B. O. Patrick, E. Chung and B. R. James, Dalton Trans.,
2012, 41, 11093.
15 A. G. Sergeev and J. F. Hartwig, Science, 2011, 332,
439.
16 T. Kishimoto, Y. Uraki and M. Ubukata, Org. Biomol. Chem.,
2008, 6, 2982.
17 P. N. Rylander, Hydrogenation Methods, Engelhard
Corporation Edison, New Jersey, 1985.
18 H. Liu, T. Jiang, B. Han, S. Liang and Y. Zhou, Science, 2009,
326, 1250.
¨
¨
3 P. Azadi, R. Carrasquillo-Flores, Y. J. Pagan-Torres, E. I. Gurbuz,
R. Farnood and J. A. Dumesic, Green Chem., 2012, 14, 1573.
4 J. R. Regalbuto, Science, 2009, 325, 822.
5 J. J. Bozell, J. E. Holladay, D. Johnson and J. F. White, Top
Value-Added Chemicals from Biomass, DOE Pacic
Northwest National Laboratory Report PNNL-16983, 2007.
6 J. Zakzeski, P. C. A. Bruijnincx, A. L. Jongerius and
B. M. Weckhuysen, Chem. Rev., 2010, 110, 3552.
7 (a) J. Lora, Industrial Commercial Lignins: Sources,
Properties and Applications, in Monomers, Polymers, and
Composites from Renewable Resources, ed. M. N. Belgacem
and A. Gandini, Elsevier, Amsterdam, 2008; (b) M. Kleinert
and T. Barth, Chem. Eng. Technol., 2008, 31, 736.
Chem. Sci.
This journal is ª The Royal Society of Chemistry 2012