Decomposition of lignin model compounds by Lewis acid catalysts in water and ethanol

Article abstract of DOI:10.1016/j.molcata.2015.09.007

The conversion of benzyl phenyl ether, diphenyl ether, diphenyl methane and biphenyl as representative model compounds for α-O₄, 5-O₄, α ₁ (methylene bridges) and 5-5′ lignin linkages was investigated. We compared the use of metal chlorides and acetates. The reactions were studied in sub- and supercritical water and supercritical ethanol between 300 and 400 °C. At low temperature in water, Lewis acids mainly catalyzed condensation of hydrolysis products of the dimeric model compounds. At higher temperature, mono-aromatic products were formed. The yield of monomeric products was higher in ethanol than in water. The preference for ethanol is due to extensive alkylation of the mono-aromatic products, which inhibits their condensation into larger products. The highest yields of deoxygenated mono-aromatics were obtained using Lewis acid catalysts at 400 °C in supercritical ethanol. The preferred Lewis acid catalysts were Fe, Cu, Ni and Al chlorides.

Full text of DOI:10.1016/j.molcata.2015.09.007

Accepted Manuscript
Title: Decomposition of lignin model compounds by Lewis
acid catalysts in water and ethanol
¨
Author: Burcu Guvenatam Erik H.J. Heeres Evgeny A. Pidko
Emiel J.M. Hensen
PII:
S1381-1169(15)30083-2
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.molcata.2015.09.007
MOLCAA 9623
To appear in:
Journal of Molecular Catalysis A: Chemical
Received date:
Revised date:
Accepted date:
4-3-2015
31-8-2015
7-9-2015
¨
Please cite this article as: Burcu Guvenatam, Erik H.J.Heeres, Evgeny A.Pidko,
Emiel J.M.Hensen, Decomposition of lignin model compounds by Lewis acid
catalysts in water and ethanol, Journal of Molecular Catalysis A: Chemical
http://dx.doi.org/10.1016/j.molcata.2015.09.007
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Decomposition of lignin model compounds by Lewis acid catalysts in water and ethanol
BurcuGüvenatama, Erik H.J. Heeresb, Evgeny A. Pidkoa,c, Emiel J.M. Hensena,*
a
Laboratory of Inorganic Materials Chemistry, Schuit Institute of Catalysis, Eindhoven
University of Technology, P.O. Box 513, 5600 MB Eindhoven, The Netherlands
b Institute for Technology and Management, University of Groningen, Nijenborgh 4, 9747
AG Groningen, The Netherlands.
c
Institute for Complex Molecular Systems, Eindhoven University of Technology, P.O. Box
513, 5600 MB Eindhoven, The Netherlands.
Corresponding author:
Emiel Hensen
E-mail address: e.j.m.hensen@tue.nl
Phone: +31-40-2475178
Highlights
3






Catalytic effects of Lewis acidic salts such as metal acetates and chlorides were investigated.
Representative di-aromatic fragments were used as simple lignin model compounds.
Reactions were carried out in two different solvent media (water and ethanol).
High reaction temperatures (300-400 °C)were required to prevent condensation reactions.
According to the type of the solvent, different reaction mechanisms were predicted.
Abstract
The conversion of benzyl phenyl ether, diphenyl ether, diphenyl methane and biphenyl as
representative model compounds for α–O4, 5–O4, α1 (methylene bridges) and 5–5′ lignin linkages was
investigated. We compared the use of metal chlorides and acetates. The reactions were studied in sub-
and supercritical water and supercritical ethanol between 300-400 °C. At low temperature in water,
Lewis acids mainly catalyzed condensation of hydrolysis products of the dimeric model compounds.
At higher temperature, mono-aromatic products were formed. The yield of monomeric products was
higher in ethanol than in water. The preference for ethanol is due to extensive alkylation of the mono-
aromatic products, which inhibits their condensation into larger products. The highest yields of
deoxygenated mono-aromatics were obtained using Lewis acid catalysts at 400 °C in supercritical
ethanol. The preferred Lewis acid catalysts were Fe, Cu, Ni and Al chlorides.
Keywords: lignin, model compound, Lewis acid salt, ethanol, water
4

1. Introduction
Lignin is an amorphous three-dimensional biopolymer that makes up about a third of
lignocellulosic biomass [1-3]. Its efficient valorization is pivotal in scenarios in which biomass serves
as a renewable feedstock to fuels and chemicals. However, the structure of lignin is complex, as it is a
polymer of three primary components, namely, p-coumaryl, coniferyl, and sinapyl alcohols, connected
via different linkages [2]. For instance, the structure of hardwood and softwood lignin is dominated by
α– and β–aryl ether bonds [1], among which β–O4 constitutes nearly half of all linkages in lignin [2].
Due to its structural complexity, the use of model compounds that contain similar linkages as those
found in lignin has become a common approach in identifying approaches to upgrade lignin [4-6]. A
representative compound that contains the α–O4 bond is benzyl phenyl ether (BPE) [3,7]. The C-O
bond energy in BPE is 234 kJ/mol [3,7]. Diphenyl ether (DPE) bond is used as a model to represent
the less common 5–O4 linkage, in which the C-O bond is much stronger (330 kJ/mol) [3,7].
Experimental studies have confirmed that the cleavage of aryl-O-aryl bonds in lignin is much more
difficult than the cleavage of alkyl-O-aryl bonds [8].
Among the many approaches that have already been explored for lignin depolymerization [2, 8],
Lewis acid salts have been only scarcely investigated. Hepditch and Thring [9] demonstrated that
NiCl2 and FeCl3 facilitated the degradation of Alcell lignin in water at 305 °C into a mixture of low-
molecular-weight products such as phenolics (syringols, guaiacols, catechols), aldehydes
(syringaldehyde and vanillin) and phenolic ketones (acetoguaiacone and acetosyringone). This
reaction in water yielded large amounts of insoluble reactor residue (~70 wt. %) in the presence of
5

Lewis acids.Metal triflate salts such as indium triflate, scandium triflate, ytterbium triflate, and indium
chloride have been recently used to catalyze the hydrolytic cleavage of C−O linkages in different
lignin model compounds in water [6]. For example, the use of indium triflate as a catalyst allowed
reaching a full conversion of benzyl phenyl ether at 225 °C after 3 h reaction. However, it was
concluded that further improvement of these systems is needed to increase the yield of mono-aromatic
products [6].
Lewis acidic salts have been studied much earlier as catalysts for coal liquefaction [10-12]. They
were found to promote cleavage of C-C bonds in aliphatic groups linking the aromatic units in coal,
but not aryl-aryl linkages [11]. The activity of ZnCl2 and AlCl3 as representative Lewis acid catalysts
in the conversion of such model compounds as biphenyl, diphenyl alkanes containing 1 to 4-carbon
aliphatic linkages, as well as hydroxylated biphenyl and diphenyl methane was investigated by the
Bell group [10-12]. AlCl3 was found to be more active in hydrogenation and cracking reactions of
aryl-alkyl C-C bonds than ZnCl2 [11-12]. It has been proposed that the cleavage of C-C and C-O-C
bonds follows the carbonium ion mechanism, resulting in the formation of condensation products and
tar [10]. Addition of molecular hydrogen improved the overall conversion but also led to enhanced tar
formation [11]. This was caused by hydride-facilitated condensation of aromatic species via Scholl
coupling reactions [11].
In the context of lignin model compounds, the use of metal chlorides has been explored for the
conversion of guaiacol. The hydrolysis rate of guaiacol at 380 °C in supercritical water (sc-water) was
accelerated by metal chlorides such as NaCl, CaCl2 and FeCl3. The main products of these reactions
were catechol and phenol [13]. However, similar to coal liquefaction studies, extensive tar formation
occurred in the presence of Lewis acid catalysts [13].
The choice of solvent and its physical properties under reaction conditions are key factors that
determine the efficiency of lignin depolymerization processes [2, 8]. In the last decade, the use of
water near or above its critical point (Tc = 374 °C, Pc = 218 atm) has attracted considerable attention as
a green and renewable alternative to organic solvents as a medium in synthetic fuel production,
biomass processing, waste water treatment and material synthesis [14-16].
6

Table 1 summarizes the physical properties of water in different states. The ion product or
dissociation constant of sc-water at high temperature and pressure is higher than that of water at
ambient conditions. Accordingly, sc-water will contain more H+ and OH- ions than liquid water,
making dense high-temperature sc-water an effective medium for acid and base catalyzed organic
reactions [15]. Water at pKw ≤ 14 has been argued to be a suitable medium for heterolytic reactions in
which ionic mechanisms involving charged transition states and intermediates occur. On contrary, the
use of water with pKw ≥ 14 is preferred for homolytic reactions, in which free radical mechanisms are
usually important [17]. As temperature and pressure increase, the static dielectric constant of water at
the critical point drops to a value of about 6 from 78.5 (at 258 °C) because of the reduced number of
hydrogen bonds at relatively low densities. The dielectric constant of a medium determines the
solubility of molecules [16]. The advantage of sc-water over normal water under conventional
conditions is that it behaves almost like a non-polar solvent allowing to efficiently dissolve various
non-polar compounds such as alkanes and aromatics [15-16]. Also, gases are better miscible in sc-
water. Thus, the use of sc-water as a solvent offers opportunities to effectively homogenize multiphase
reactant systems and, accordingly, improve strongly its mass-transfer characteristics [10-11]. The
specific heat capacity increases under supercritical conditions, which can be advantageous to reduce
possible hot spot formation [16]. The density of sc-water strongly influenced the hydrolysis
mechanism that led to C-O bond dissociation in DPE at high temperatures (415 – 480 °C) [14]. At low
density water (< 0.3 g/cm3) DPE conversion resulted in formation of various polycondensation
products with low yield of mono-aromatics (< 0.1 %). At higher water densities (> 0.4 g/cm3) the rate
of undesired condensation reactions paths was much lower and stoichiometric amounts of phenol were
obtained [14].
For the upgrading of lignin, another potential solvent that is green and renewable is ethanol [18]. It
is currently already used as a bio-fuel. Ethanol can be produced from biomass via fermentation and,
therefore, its production can be readily integrated in biorefinery concepts. The critical point of ethanol
(Tc = 241 °C, Pc = 6.14 MPa, ρc = 0.276 g/cm3) is at lower temperature and pressure than that of water.
Supercritical ethanol (sc-ethanol) is also less corrosive than sc-water [15-17]. Previous studies have
7

demonstrated the beneficial effect of ethanol under near- and supercritical conditions as a solvent for
separations and chemical reactions. Similar to water, ethanol becomes nearly non-polar under
supercritical conditions and its dipolarity/polarizability ranges from gas-like to non-polar liquid with
increasing temperature and pressure [19]. The main advantage of using sc-ethanol is that it maintains
significant hydrogen-bond donating acidity under supercritical conditions [19]. On the other hand, the
hydrogen-bond accepting basicity of ethanol becomes considerably weaker at elevated temperatures
[19]. Several reactions of industrial significance such as catalytic etherification from tertiary alcohol,
alkylation of toluene to produce p-ethyltoluene and non-catalytic hydrogen-transfer reductions of
aldehydes and ketones have been successfully carried out in hot ethanol up to its critical temperature
[19].
In this work, we investigate a range of inorganic salts as Lewis acid catalysts for the conversion of
representative lignin model compounds in sub- and supercritical water and ethanol with the goal to
assess their potential for lignin depolymerization. Benzyl phenyl ether, diphenyl ether, diphenyl
methane and biphenyl were selected as representative model compounds for the α–O4 (aryl-alkyl
bond), 5–O4 (aryl-aryl bond), α1 (C(aryl)-C(alkyl) bond), 5–5′ (C(aryl)-C(aryl) bond) linkages in
lignin. Chloride salts of Fe, Co, Cu, Ni and Al, acetate salts of Fe, Co, Cu, Ni and triflate salts of Sc
and Al were evaluated as Lewis acid catalysts for conversion of these model compounds. For reasons
of comparison, we employed metal triflates (M(OTf)n , M = metal) as super Lewis acid catalysts. The
reaction conditions such as temperature, solvent (water or ethanol) and solvent loading (density) were
varied in order to understand their effect on the conversion and selectivity. On the basis of reactivity
data, we discuss possible reaction mechanisms for the Lewis acid catalyzed conversion of the model
compounds.
2. Experimental methods
2.1 Materials
Benzyl phenyl ether (Aldrich, 98 %), diphenyl ether (ReagentPlus®, ≥99 %), diphenyl methane
(Aldrich, 99 %) and biphenyl (ReagentPlus®, 99.5 %) were used as received. Deionized water and
8

absolute ethanol (Sigma Aldrich, ≥99.8 %) were used as solvent. Iron(II) acetate (Aldrich, 95 %),
copper(II)acetate (Aldrich, powder, 98 %),cobalt(II) acetate tetrahydrate (Merck, ≥98.0 %), nickel(II)
acetate tetrahydrate (Aldrich, 98 %), iron(II) chloride tetrahydrate (Aldrich, ≥99.0 %), copper(II)
chloride dihydrate (Aldrich, ≥99.0 %), cobalt(II) chloride (Aldrich, anhydrous, ≥98.0 %), nickel(II)
chloride (Aldrich, anhydrous, 98 %), aluminum(III) chloride hexahydrate (Fluka, ≥99.0 %),
aluminum(III) triflate (Aldrich, 99.9 %), scandium(III) triflate (Aldrich, 99 %), copper(I) oxide
(Aldrich, ≥99.99 % trace metals basis, anhydrous), alumina (Ketjen, CK-300 high purity) were used
without further purification. Ethyl acetate (VWR, 99.5 %) was used as an extraction solvent. N-decane
(Aldrich, anhydrous, ≥99 %) and di-n-butyl ether (Aldrich, anhydrous, 99.3 %) were used as external
standards, respectively. Tetrahydrofuran (Aldrich, anhydrous, ≥99.9 %) was used to dilute the samples
prior to analysis.
2.2 Catalytic activity measurements
All the experiments were performed in stainless-steel batch reactors with an internal volume of 13
mL. Reactions were carried out in the temperature range of 300-400 °C with a reaction time of 3 h.
The reactors were filled with either 8 mL of water or 6.5 mL of anhydrous ethanol. The reactant and
catalyst concentrations were 0.1 mol/L and 0.025 mol/L, respectively. Reactions at 300 °C and 350 °C
were carried out by placing the autoclaves in the oven of a gas chromatograph. Reactions at 400 °C
were done in a pre-heated fluidized sand bath. After reaction, the reactors were quenched in an ice
bath. The organic compounds were extracted from the water solvent by ethyl acetate in 1:1 volume
ratio with water. In case of ethanol as solvent, this extraction step was not needed.
2.3 Product analysis
For the identification and quantification of the products, GC/MS-FID analyses were performed on a
Shimadzu GC/MS-QP2010 SE series. The GC was equipped with a Restek Rtx-1701 capillary column
(60 x 0.25 mm i.d. and 0.25 μm film thickness). The column flow was split in a 1:10 volume ratio to
the MS and FID. The injector temperature was set at 250 °C. The oven temperature was kept at 45 °C
9

for 4 min, followed by heating to 280 °C at a rate of 4 °C/min and then held at 280 °C for 5 min.
Identification of products was done using the NIST11 and NIST11s libraries.
For product quantification in ethanol-mediated reactions, GCGC analysis was performed on a
Interscience Trace GCGC equipped with a cryogenic trap system and two columns: 30 m x 0.25 mm
i.d. and 0.25 μm film of Sol-gel (SGE-1MS) capillary column connected by a meltfit to a 150 cm x 0.1
mm i.d. and 0.1 μm film RestekRtx-1701 column. An FID detector was used. A dual jet modulator
was applied using carbon dioxide to trap the samples. Helium was used as the carrier gas (continuous
flow 0.6 mL/min). The injector temperature and FID temperature were set at 250 °C. The oven
temperature was kept at 40 °C for 5 minutes then heated to 250 °C at a rate of 3 °C/min. The pressure
was set at 0.7 bar at 40 °C. The modulation time was 6 s.
The GC/FID analysis used to quantify mono- and di-aromatic products had error margins of 0.1 %
and 5 %. The error margin in the quantification of higher-molecular weight (HMW) compounds was
estimated at 10 %.
The conversion of reactant (X (%)) was calculated from the initial and final amounts (mol) of the
substrate (eqn. (1)). The selectivities (Si (mol %)) of each product was calculated according to eqn. (2).
X % = initial mol of reactant - final mol of reactant ×100
(eqn. 1)
(eqn. 2)
( )
initial mol of reactant
ꢀꢁ % = mol of product i formed (ꢂꢃꢄꢂ. ꢅꢆꢇꢈ ꢉꢊꢋꢌꢍꢎ) × number of C atoms ×100
( )
total mol of products (ꢂꢃꢄꢂ. ꢅꢆꢇꢈ ꢉꢊꢋꢌꢍꢎ) × number of C atoms
For the reactions in water, the response factors were obtained via calibration using commercially
available mono- and di-aromatic compounds. When specific di-aromatic or HMW components were
not available, a representative response factor (RF) was calculated as an average of the response
factors of BPE, DPE, DPM and biphenyl. For the reactions in ethanol, alkylated products were
summed and the corresponding response factors of the parent product were used.
10

3. Results and Discussion
3.1 Benzyl phenyl ether conversion in water
We first carried out BPE conversion reactions in water at different temperatures in the presence of
Fe, Cu, Co, Ni or Al chloride or acetate salts. A blank experiment was also performed. The results of
the catalytic reactions at 300, 350 and 400 °C are summarized in Tables 2-4.
Already at 300 °C, BPE is almost completely converted within 3 h in all cases (Table 2). In the
blank reaction, the conversion was 98 %, and benzyl alcohol, phenol and benzaldehyde were formed
in addition to di-aromatics and some higher molecular-weight (HMW) products. For metal chlorides
(MCln), the main monomer product was phenol (Fig. S1). For metal acetates (M(OAc)n), phenol,
benzylic alcohol and benzaldehyde were the monomeric products. In all catalytic experiments, di-
aromatics, mainly isomers of hydroxy-DPM (2-benzyl phenol and 4-benzylphenol), and HMW
products were formed similar with blank reaction. The use of MCln decreased the monomer yield
compared to the blank. This is mainly due to the formation of more substantial amounts of di-aromatic
products. The anionic part of the metal salt affected the product distribution more than the cationic
part. Co- and Ni-acetates produced nearly similar yields of monomers as in the blank experiment.
The low monomer yield in water at 300 °C is due to the high rate of condensation of reactive
intermediates formed during BPE hydrolysis. There are pronounced selectivity differences in the
catalytic effect of chloride and acetate salts. We hypothesize that the nature of the anion determines
the stability of the metal site against hydrolysis and, therefore, the effective Lewis acidity of the
catalyst [20-22]. To verify this hypothesis, additional measurements were carried out using water-
tolerant Sc(OTf)3 and Al(OTf)3 super Lewis acids [23].
We also used bulk Al2O3 and CuO oxides as models to mimick the possible products of hydrolysis
of the metal salts that can produce oxides. The results of these catalytic tests are also given in Table 2.
BPE was fully converted with all of these catalysts. With M(OTf)n, phenol and hydroxy-DPM were
the main products formed, and roughly speaking, the yields correspond to those seen for the MCln
catalysts. It supports the conclusion that high Lewis acidity promotes condensation of reactive
intermediates, leading to C-C linked products such as substituted DPM. When Al2O3 and CuO were
11

used, the product distributions were similar to those for the blank and M(OAc)n catalyzed experiments.
Note, however, that in the presence of Al2O3 the highest yield of monomer products was obtained (64
%) with benzyl alcohol as the dominant product formed at a selectivity of 47 %. The yield of di-
aromatics was the lowest (~30 %) for Al2O3. The reaction selectivities with CuO were identical to
those obtained with the Cu(OAc)2 catalyst, suggesting that Cu-acetate decomposed during reaction.
High-temperature hydrolysis has been earlier reported for Pd(OAc)2-catalyzed Heck coupling
reactions in sc-water at 400 °C [24]. Therefore, we propose that the higher Lewis acidity of MCln
catalysts compared with that of M(OAc)n is most likely due to the decomposition of the latter into less
reactive oxides/hydroxide compounds under the conditions of the catalytic reaction. At low
temperature and in the presence of the strong Lewis acidity of MCln and M(OTf)n, condensation
products of BPE hydrolysis are strongly favored and only very low monomer yields can be obtained.
As condensation reactions may be suppressed at higher temperature, we evaluated BPE conversion
at 350 °C and 400 °C (Table 3-4). Expectedly, BPE conversion was complete in all cases. In addition
to oxygen-containing aromatic monomers (mono-aromatic oxygenates) and oligomeric products, non-
oxygenated aromatic monomers (mono-aromatics) were observed in the product mixture. In the blank
reaction, toluene was the only mono-aromatic product. In the presence of MCln, M(OAc)n and Al2O3,
small amounts of benzene was also formed. The highest selectivity to mono-aromatics was obtained
with Cu, Ni and Al chlorides.
Except for the formation of toluene, the product compositions of the blank experiments were
similar at 300 and 350 °C. The main monomer products were benzyl alcohol (33 %), phenol (19 %)
and benzaldehyde (4 %). For all catalysts, the main product at 350 °C was phenol (selectivity 54 ± 2
%). Only small amounts of benzyl alcohol were observed for Cu(OAc)2 and Co(OAc)2, very different
from the results obtained at 300 °C, where benzyl alcohol was the main monomer for Al2O3 and
M(OAc)n. In most catalytic experiments, small amounts of benzaldehyde were observed. Especially,
Cu(OAc)2 promoted benzaldehyde formation. Previously, the aqueous-phase benzyl alcohol oxidation
to benzaldehyde has been shown to occur via hydride abstraction from benzyl alcohol [25-27].
12

The selectivity patterns for MCln, M(OAc)n and bulk Al2O3 were similar. This means that the
hydrolyzed metal species either did not form or that they were dissociated again. As expected, less
condensation products formed at higher temperature. The most pronounced decrease was seen for the
MCln catalysts, where the condensates yield decreased by almost 50 %. Cu(OAc)2 produced the least
condensation products among the catalysts evaluated at 350 °C.
At 400 °C, phenol is the main reaction product in all of the cases (Table 4). Note that the yield of
non-oxygenated mono-aromatics formed at 400 °C with MCln is higher than that obtained at a lower
temperature of 350 °C. The highest non-oxygenated mono-aromatic yield was reached by using
AlCl3. The phenol selectivity ranged between 40-53 % in the presence of Lewis acidic catalysts. For
Al2O3, the phenol selectivity dropped to 17 %. Benzyl alcohol is totally degraded at 400 °C, which
resulted in high yields of condensation (di-aromatics and HMW) products. Especially, Al2O3 promoted
this condensation path. At 400 °C and high water loading (8 mL), the carbon balance was between 70-
80 % depending on the catalyst type. When the solvent loading was 4 mL, only 40-60 % of the carbon
was found back in the products. At low density water, the selectivity of di-aromatics was below 5 %
while the selectivity to HMW products was below 10 % during the blank reaction and below 20 %
when catalysts were used. This may suggest that unidentified products with much higher molecular
weights were formed. The same trend was previously observed in base-catalyzed conversion of BPE
when the water density decreased [26]. Thus, phenol was the main product with high selectivity for the
blank reaction (74 %) and also for the M(OAc)n catalysts. In this case, the highest degree of
deoxygenation was obtained with the MCln catalysts with combined benzene and toluene selectivities
as high as 36 % for CuCl2. In brief, when the temperature is increased above 300 °C, mono-aromatics
can be obtained by BPE hydrolysis. Both MCln and M(OAc)n catalysts are active in the deoxygenation
at 400 °C. However, even at these elevated temperatures, condensation reactions cannot be fully
prevented in water.
We propose that BPE is converted according to the mechanism shown in Fig. 1. The acidity of
water itself strongly increases with temperature [27], explaining the high BPE conversion in sc-water.
Reactions between Lewis acidic metal ions with water result in the generation of acidic protons. BPE
13

will undergo the acid-catalyzed hydrolysis of the ether bond to form phenol and benzyl alcohol, which
are the main monomer products at 300 °CThese reactive products can further undergo various
transformations including self- and cross-condensation [5]. Benzyl alcohol can undergo
dehydrogenation at elevated temperatures in sc-water, resulting in the formation of benzaldehyde and
molecular H2 [28]. All these mono-aromatic oxygenated products are prone to condensation reactions
and form di-aromatic and HMW components. Benzyl alcohol readily reacts with phenol to form di-
aromatic components such as 2-benzylphenol and 4-benzylphenol (hydroxy-DPM) [26]. The reaction
of these di-aromatic species with the mono-aromatic oxygenates may proceed towards tri-nuclear
aromatics and further HMW products. The small amount of toluene detected at 350 °C can be formed
via slow hydride abstraction from the media by the benzyl cation [11], while benzene is likely the
product of the hydrodeoxygenation of phenol or decarbonylation of the benzaldehyde intermediate
[28]. We speculate that cleavage of ether cross-links was favored above 350 °C via radical pathways.
It was previously shown that there is always competition between radical and ionic reaction pathways
during thermal hydrolysis of ethers in water [25]. The ionic mechanism is the more relevant one at
lower temperatures [8.
To summarize, full conversion of BPE in water can be achieved at temperatures in the range of 300
– 400 °C. Both the catalytic and non-catalytic reactions were always accompanied by the formation of
dimeric and higher molecular weight products. It was observed that higher reaction temperatures
promoted deoxygenation resulting in an increase of the yields of benzene and toluene. At the highest
temperature of 400 °C and high water density, the use of AlCl3 afforded more non-oxygenated mono-
aromatics than the blank reaction.
3.2 Conversion of benzyl phenyl ether in ethanol
We investigated in a similar manner BPE conversion in ethanol. The results for sc-ethanol
reactions at 300, 350 and 400 °C are summarized in Tables 5 and 6. Unlike for the sc-water-mediated
reaction, the conversion of BPE in ethanol at 300 °C is not complete after 3 h. Without catalyst, the
conversion was 72 %. The product composition was also very different from the experiments done in
sc-water. Without catalyst, the monomer products were toluene and phenol with selectivities of 27 and
14

53 %, respectively. In the presence of Lewis acids, the product distribution depended more on the
anionic than on the cationic part. It was found that ethanol was reactive under the chosen reaction
conditions. In addition to phenol and toluene, substantial amounts of alkylated benzyl alcohol (alkyl-
BzOHs) and alkylated phenols (alkyl-PhOHs) were observed (Fig. S2). The products of ethanol
conversion include 1,1-diethoxyethane and diethyl ether as major products as well as acetaldehyde and
butenes (Fig. S2).
The use of M(OAc)n led to much lower BPE conversion compared with the blank reaction. The
selectivities to phenol, toluene and alkyl-PhOHs were in the range of 42 %-52 %, 20 %-27 %, and 3
%-9 %, respectively, for M(OAc)n. In addition, alkyl-BzOHs were only formed in trace amounts for
Fe(OAc)2 and Cu(OAc)2. The mono-aromatic oxygenates selectivity and the ratio of phenol and alkyl-
PhOHs were only slightly influenced by the nature of the cation. The reaction with Co(OAc)2 resulted
in the highest selectivity to alkyl-PhOHs (9 %) and, accordingly, the lowest phenol selectivity (42 %).
This indicates that Co(OAc)2 is a good alkylation catalyst at 300 °C. Fe and Ni acetates showed
slightly lower alkylation activity, resulting in very similar selectivities to phenol (50 and 52 %,
respectively) and alkyl-PhOHs (5 and 7 %, respectively). In the presence of Cu(OAc)2 catalyst, the
selectivity towards alkylated products was the lowest (~3 %).
MCln catalysts strongly promoted BPE conversion in sc-ethanol. FeCl2 was able to nearly fully
convert BPE within 3 h at 300 °C. CuCl2 and AlCl3 gave slightly lower BPE conversions (90 %),
while CoCl2 and NiCl2 converted, respectively, 80 and 83 % of BPE. All MCln catalysts gave very
similar monomer selectivities. The combined selectivities to mono-aromatic oxygenate compounds
was around 70 %. The stronger Lewis acid MCln led to higher yields of the alkylation products (14 %-
23 %) compared to use of M(OAc)2. Interestingly, almost all Lewis acid catalysts considered here
decreased the toluene selectivity compared to the blank experiment.
The condensation products were mainly di-aromatics such as diphenylmethane (DPM), hydroxy-
DPM, 1,2‒diphenylethane, and ethyl-hydroxy-DPM as well as HMW products, mainly tri-nuclear
aromatics with molecular weights higher than 258 g/mol. The formation of DPM and
15

1,2‒diphenylethane might suggest radical-type condensation reactions of toluene. M(OAc)2 produced
more condensed products than MCln.
For Al2O3, the conversion of BPE was 74 %, similar to the value for the blank reaction. However,
the product composition was very different. The mono-aromatic oxygenates selectivity of 62 % was
higher than for M(OAc)2 and lower than for MCln. Alkylation of mono-aromatic oxygenates was
strongly promoted in the presence of Al2O3, resulting in the highest selectivity towards alkyl-BzOHs
(25 %). The toluene selectivity was only 17 % which may relate to the promotion of radical
condensation paths leading to di-aromatic and HMW products. CuO gave a BPE conversion of 84 %
with substantial amounts (37 %) of toluene and less mono-aromatic oxygenates (46 %). With CuO
alkylation of mono-aromatic oxygenates was not possible (<1 %). In addition, condensation reactions
led to the formation of mainly hydroxy-DPM and DPM type of di-aromatic components (~16 %).
Thus, the catalytic results of Al2O3 and CuO did not correspond to the behavior of the corresponding
Lewis acidic salts in sc-ethanol at 300 °C as they did in sc-water.
BPE conversion was complete at a reaction temperature of 350 °C and 400 °C (Table 6). The
mono-aromatics selectivity, mainly toluene, increased substantially at the expense of mono-aromatic
oxygenates. In the non-catalytic alcoholysis of BPE, toluene (39 %) and phenols (42 %) were almost
obtained in equimolar ratio at 350 °C. Alkylation of phenol led to the formation of small amounts of
alkyl-PhOHs (3 %). In addition, ethanol products such as acetaldehyde, diethyl ether, 2-butanol and
1,1-diethoxyethane were observed. In the presence of catalyst, more alkylated mono-aromatic
oxygenates were found. Alkylation of mono-aromatics was much slower than alkylation of mono-
aromatic oxygenates. The mono-aromatics yield was the highest (45 %) with Cu(OAc)2. Similar with
the results at 300 °C, the mono-aromatic oxygenates fraction consisted of phenol, alkyl-BzOHs and
alkyl-PhOHs for MCln.
At 350 °C, the use of bulk catalysts had a significant influence on the selectivity towards alkylated
products compared to the blank reaction. Catalytic activity of Al2O3 was comparable with the
corresponding chloride salts AlCl3, while CuO exhibited similar behavior with the corresponding
acetate salt Cu(OAc)2. Al2O3 catalyzed the formation of alkyl-BzOHs. These products did not form
16

with CuO. In ethanol, a better mass balance closure was observed than in water, for most experiments
above 85 %.
At 400 °C, ethanol conversion was strongly increased. As a result, the composition of the product
mixture was very different from that obtained at lower temperatures. As the product mixtures were
complex, we combined results of GC/MS-FID and GCGC. Fig. 2 shows a representative
GCGCchromatogram for the reaction of BPE with CuCl2 in ethanol at 400 °C. Ethanol products
such as butenes, acetaldehyde, ethyl ethers and carboxylic acids (acetic, propanoic and butanoic acid)
were observed. A wide range of aliphatic products (mainly alkanes and some alkenes) were formed as
well as alkylated mono-aromatics bearing alkyl side chains. Alkylation of aromatics was possible at
the meta-, para- and ortho-positions. Mono-aromatic oxygenates were mainly present as alkylated-
PhOHs (Fig. 2). Naphthalenes might be formed via oligomerization reactions of monomeric products.
They can be also likely formed via cyclization of hexyl-alkylated aromatics. Condensation further
results in formation of di-aromatic components.
The reaction data obtained at 400 °C are also collected in Table 6. At elevated temperature,
deoxygenation became more prominent, as follows from the much higher yield of mono-aromatics,
when MCln was the catalyst. Comparatively, the blank reaction and the use of M(OAc)n produced less
mono-aromatics and more mono-aromatic oxygenates. The product distribution did not vary much for
the MCln catalysts. As an example, the main products with CuCl2 were alkylated mono-aromatic, di-
aromatic, mono-aromatic oxygenate, naphthalene and di-aromatic products formed with selectivities
of 60 %, 13 %, 22 % and 5 %, respectively. With M(OAc)n, selectivities for mono-aromatics and
naphthalenes amounted to ~41 % and ~22 %, respectively. Co(OAc)2 displayed the lowest
deoxygenation activity and the highest di-aromatics selectivity. MCln produced less undesired
condensation products than M(OAc)n at 400 °C. This is in line with the results for the reactions at
lower temperatures (300 °C and 350 °C). Interestingly, in the absence of catalysts, the selectivity to di-
aromatics and naphthalenes was very low, suggesting a crucial role of Lewis acid catalysts for the
condensation reactions. Furthermore, the conversion of ethanol was much less pronounced in the
blank reaction compared to catalytic experiments.
17

Summarizing, at 300 °C condensation reactions in sc-ethanol are slower than in sc-water. MCln
catalysts gave the lowest yield of condensation products. However, the condensation reactions towards
stable C-C linked di- and tri- aromatic compounds could not be completely suppressed by the higher
reaction temperature. At 350 °C, in addition to complete conversion of BPE, the increased yields of
mono-aromatics (up to 45 mol %) showed that the deoxygenation paths became strongly favored.
Results indicated that BPE was converted via different reaction mechanisms in sc-ethanol and sc-
water. In ethanol, the formation of phenol and toluene at 300 and 350 °C points to alcoholysis of BPE
involving molecular hydrogen formed via ethanol dehydrogenation (Fig. 3). The formation of
hydrogen is consistent with the observation of acetaldehyde in the product mixture. An alternative
explanation is homolytic cleavage of the ether bond, resulting in phenoxy and benzyl radicals that lead
to phenol and toluene upon recombination with ethanol solvent. Phenol is more susceptible to
alkylation than benzene, which explains the predominant formation of alkylated mono-aromatic
oxygenates at low temperature. At temperatures below 400 °C, the transformation of ethanol was
limited to dehydrogenation and alkylation of mono-oxygenate species. However, at 400 °C ethanol
conversion became more extensive and resulted in a wide range of aliphatic hydrocarbons (alkanes
and alkenes). In fact, in the presence of Lewis acid catalysts, ethanol was involved in alkane coupling
reactions to form aliphatic products and to alkylate all types of alcoholysis products. Furthermore, the
radical nature of the reaction resulted in increased selectivity to condensation products at the higher
temperature. Condensation products were mainly produced by combination of phenol, benzyl alcohol
and toluene. The isomers of hydroxy-DPM were the principle di-aromatic products. Other abundant
types of condensation products obtained at 400 °C were DPM, 1,2‒diphenylethane, ethyl-hydroxy-
DPM. Formation of HMW components was limited at 400 °C in ethanol. Only trace amount of tri-
nuclear aromatics with molecular weights higher than 258 were detected. Naphthalenes were the main
condensation products at 400 °C.
The proposed reaction mechanism for naphthalene formation is shown in Fig. 4 where phenol, as a
monomer product of alcoholysis of BPE, is converted to naphthalene. Previously, naphthalene
formation was reported in pyrolysis studies of anisole [29]. It was proposed that the reaction is
18

initiated by the formation of phenoxy radicals that undergo decarbonylation reaction to yield
cyclopentadienyl (C5H5) radicals [29]. Being the main precursor of naphthalene and polycyclic
aromatic hydrocarbons, C5H5 radicals undergo self-recombination towards naphthalene [29].
Consistent with this, the catalyst that showed the highest yield of mono-aromatic oxygenates (mainly
phenols), i.e. Co(OAc)2 showed the lowest naphthalene selectivity.
3.3 Conversion of diphenyl ether
We also examined the effect of Lewis acid salts on the decomposition of DPE, the model
compound for the 5–O4 ether aryl-O-aryl lignin linkage. Previously, it was reported that cleavage of
aryl-O-aryl ether bond in sc-water is much more difficult than that of alkyl-aryl ether bond[8]. As we
hypothesized that relatively high reaction temperatures and strong Lewis acid catalysts are necessary
to cleave the 5–O4 ether bond, reactions were only carried out at 350 and 400 °C in sc-water and sc-
ethanol.
Without catalyst, DPE was not converted at 350 °C in sc-water and only trace amounts of phenol
were observed (Table S1). DPE conversion was also very low for all Lewis acids considered here. The
highest conversion (8 %) was for Co(OAc)2 with 4-hydroxy biphenyl as the main reaction product. In
all other cases, phenol was the dominant product. At 400 °C, DPE conversion was higher (Table S1).
In the blank reaction, the phenol selectivity was 54 %, and various condensation products including 4-
hydroxy biphenyl were also formed. For MCln, no condensation products were observed. The main
products were phenol (87 % ± 3 %) and benzene (13 % ± 3 %). NiCl2, however, gave more benzene,
showing its high deoxygenation activity. For M(OAc)n, the DPE conversion was higher than for MCln,
but it led to a higher yield of condensed products at the expensive of the phenol selectivity. Again,
Co(OAc)2 gave almost exclusive formation of 4-hydroxy biphenyl (95 %) with phenol as the only
other product.
Previously, it has been shown that DPE hydrolysis can be catalyzed by the strong BF3 Lewis acid at
380 °C. Four equivalents of BF3 hydrolyzed 58 % DPE to phenol in water (at 380 °C, 30 min)[30].
When the BF3 concentration was lowered four times, the phenol yields decreased to 12 %. Varga et al.
19

[30] compared the catalytic activities of BF3 (strong Lewis acid) with HBF4 (Brønsted acid) on DPE
hydrolysis and illustrated that BF3 is nearly twice as effective as HBF4. Accordingly, we surmise that
the relatively low DPE conversion in the present study is due to the low Lewis acid loading.
We propose the following reaction mechanism for the hydrolysis of DPE in water (Fig. 5). Acid-
catalyzed hydrolysis of DPE leads to the formation of two phenol molecules. The reaction proceeds
via highly unstable phenyl-cation species. The recombination of these species with water is the
predominant path at 350 °C, resulting in the exclusive formation of phenol [14]. At elevated
temperatures or in the presence of specific catalysts such as Co(OAc)2, the recombination of these
activated monomer species towards 4-hydroxy biphenyl becomes preferred. Interestingly, no
oligoaromatics formed for MCln at 400 °C. Also benzene was observed in some cases, and especially
NiCl2 was active in the deoxygenation of phenol.
DPE conversion in sc-ethanol at 350 °C and 400 °C are summarized in Table S2. DPE conversion
at 350 °C was very low for all catalysts (3 %-10 %). There was only little influence of the presence of
the type of catalyst. The only product observed was 4-hydroxy biphenyl. DPE conversion was higher
at 400 °C (Table S2). The conversion of ethanol led to hydrocarbons with up to 20 C atoms. A
representative GCGCchromatogram of the reaction mixture formed by the alcoholysis of DPE with
CuCl2 in ethanol is shown in Fig. 6. The reaction products consisted of mono-aromatics, aliphatics and
carboxylic acids. Importantly, a substantial amount of deoxygenated mono-aromatic products were
formed. These products were also alkylated by the solvent. The extensive alkylation of mono-
aromatics hampers subsequent condensation reactions. This may explain the relatively low selectivity
to HMW products [31].In the blank reaction, DPE conversion was very low and alkyl-aromatics (86
%), phenol (9 %) and alkyl-PhOHs (5 %) were obtained. The product of ethanol conversion in the
absence of Lewis acidic salts was limited to light aliphatics. Nevertheless, it is important to note that,
even without catalyst, high selectivity towards alkylated mono-aromatic products was obtained at 400
°C. In the presence of the Lewis acid catalysts, formation of mono-aromatics increased.
Deoxygenation led to the formation of water and also alkylated mono-aromatic products. As a result,
phase separation of the reaction mixture occurred. In the presence of MCln and M(OAc)n, DPE
20

conversion was much higher. The products are extensively alkylated mono-aromatics. By using CuCl2,
a conversion of 48 % could be reached with a selectivity of 94 % to mono-aromatics. Very similar
results were obtained with AlCl3.
We propose for DPE conversion the reaction mechanism depicted in Fig. 7. Homolytic dissociation
of DPE produces phenyl and phenoxy radicals in ethanol. At 350 °C, radicals cannot be stabilized and
underwent recombination reactions to form the sole product 4-hydroxy biphenyl. At 400 °C, ethanol
was converted to aliphatic hydrocarbons and it was very reactive for the alkylation of the DPE
alcoholysis products. Alkylation and in-situ generation of hydrogen are expected to stabilize these
radicals. As a result, extensively alkylated mono-aromatics and alkyl phenyl ethers were formed. Aryl-
alkyl ether bonds formed in the course of reaction were easily hydrodeoxygenated to form alkyl-
aromatics and aliphatics (alkanes and alkenes). More importantly, in the presence of Lewis acidic
salts, it was possible to reach nearly complete deoxygenation which resulted in high yields of mono-
aromatic products.
We briefly discuss the influence of the solvent by using the concept of donor number for reactants
and solvents. The donor number (DN) can be determined calorimetrically as the negative value of the
standard enthalpy change for the 1:1 adducts formation between compound solvent and antimony
pentachloride (SbCl5) in 1,2-dichloroethane at 25 °C [32]. DPE and phenol have donor numbers (DN)
of 6 and 11, respectively [30]. However, the donor numbers for water and ethanol are much higher at
18 (gas) and 33 (liquid) and 32, respectively [32]. Previously, the combination of strong Lewis acid
and very low donicity solvents was considered necessary to hydrolyze DPE at lower temperatures (250
°C) [30]. The advantage of ethanol as a solvent in our system might be that its conversion to higher
hydrocarbons at 400 °C resulted in the formation of a solvent media with much lower DN numbers
(e.g., DNhexane = 0). This might be the reason for the higher conversion of DPE in ethanol at 400 °C. It
has also been reported that H2 was required for both hydrolysis and hydrogenolysis reaction of DPE
catalyzed by Ni/SiO2 in water [33], where C–O bond cleavage was reported as a rate determining step
instead of ring hydrogenation. In our studies, ethanol also acts as hydrogen donor solvent, especially at
400 °C, accelerating the conversion of DPE.
21

In summary, we showed that the cleavage of 5–O4 type of ether bond was much more difficult than
the α–O4 cleavage, even at 400 °C in the presence of Lewis acid catalysts. DPE reactions proceeded at
much lower rates in sc-water than in sc-ethanol. This is the oppositive of what was observed for BPE
conversion. At 400 °C in sc-ethanol, DPE was converted in high yield to monomer products that
additionally underwent deoxygenation reactions and extensive alkylation by ethanol. It is argued that
both the deoxygenation and alkylation protects the products from condensation reactions.
3.4 Conversion of diphenyl methane
We also investigated the reactivity of Lewis acid catalysts towards the cleavage of DPM as a model
for the cleavage of aryl-alkyl C-C α1–type lignin linkages. This reaction was only done at 350 °C in
sc-water and sc-ethanol.
In water, DPM conversion was low (max 18 % for Cu(OAc)2 and AlCl3) and the products in
decreasing order of yield were condensation products (mostly HMW products) > mono-aromatic
oxygenates (benzaldehyde and phenol) > aromatics (mostly benzene) (Table S3). With MCln, the
benzene yield was the highest. The conversion of DPM in water was dominated by condensation
reaction paths. Tri-nuclear aromatics were formed in higher yields than di-aromatics. In sc-ethanol,
DPM was not converted at all, except when CuCl2 was the catalyst. In this case, DPM conversion was
5 % with 99 % selectivity to toluene.
Fig. 8 depicts the proposed reaction mechanism for the high-temperature conversion of DPM in
water. It has been shown previously that a direct oxygen supply from sc-water is necessary for the
hydrolysis of DPM [34]. We propose that hydroxyl radicals first attack to the methylene carbon and
form diphenyl methanol (benzhydrol), which is very reactive and unstable in sc-water [34].
Subsequently, diphenyl methanol is converted to DPM and benzophenone [34]. Benzophenone can be
formed in sc-water as the first stable product from DPM [34-35]. In sc-water, decomposition of
benzophenone into benzoic acid has also been reported [34]. We speculate that benzophenone is
further converted to the phenyl radical and benzoic acid via homolytic cleavage. The radical can be
converted to benzene via H· radicals from water or form phenol by directly reacting with water. On the
22

other hand, benzoic acid can be converted to benzaldehyde via dehydroxylation. The ketone group can
easily undergo hydrogenation reactions to form benzyl alcohol, which is further hydrodeoxygenated to
toluene. However, in our studies, intermediate products such as benzhydrol, benzoic acid and benzyl
alcohol were not detected. Benzophenone was observed only in very small quantities.
Moreover, in the presence of MCln, both stabilization of phenyl radical by H· radicals and
hydrodeoxygenation of benzaldehyde were promoted. On the contrary, M(OAc)n promoted
recombination of phenyl cations resulting in di-aromatic products such as biphenyl. In addition,
condensation of mono-aromatic oxygenates formed hydroxy-DPM and tri-nuclear aromatic products.
In summary, it is only possible to cleave the α1 C-C bond in DPM in sc-water. Oxygenation of the
α1 bond in sc-water led to formation of monomers but also to extensive condensation reactions in the
presence of Lewis acid catalysts. The use of sc-ethanol as solvent did not lead to conversion of DPM.
3.5 Conversion of biphenyl
Biphenyl was not converted in sc-water or in sc-ethanol with or without catalysts in the
temperature range of 300-400 °C. The recalcitrance of biphenyl is well known [8], and it is due to the
high bond strength of the 5–5′ C-C linkage. This suggests that phenyl-phenyl linkages of lignin will be
very difficult to break down into smaller aromatic units, even in the presence of Lewis acidic salts.
4. Conclusions
The catalytic effect of Lewis acidic salts on the cleavage of different types of lignin linkages in
representative di-aromatic model compounds was investigated. Full conversion of benzyl phenyl ether
(α–O4) can be achieved in water and ethanol solvents at the temperatures between 300 and 400 °C.
The main reaction products are phenol, toluene and benzyl alcohol. Condensation reactions towards
higher molecular weight products occurred at higher rates in water than in ethanol. Higher Lewis
acidity also promoted these condensation reactions. The conversion of diphenyl ether (5–O4 model)
was low, even at 400 °C. In ethanol, the conversion was substantially higher, especially with metal
chloride catalysts. At 400 °C deoxygenation of the monomeric products was almost complete. Ethanol
was involved as reactant in the alkylation of the aromatic products of lignin model compound
23

decomposition. It was found that dissociation of the α1 bond in diphenyl methane was only possible in
water. This reaction most likely involves the oxidation of the bridging methylene group to
benzophenone. Although benzaldehyde, benzene, phenol and toluene were formed in the course of
diphenyl methane conversion, condensation reactions led to formation of high-molecular weight
products. Biphenyl representing the most stable 5–5′ lignin linkage could not be converted into mono-
aromatic molecules under the studied reaction conditions.
Acknowledgement
This research has been performed within the framework of the CatchBio program. The authors
gratefully acknowledge the support of the Smart Mix Program of the Netherlands Ministry of
Economic Affairs and the Netherlands Ministry of Education, Culture and Science.
References
[1] D. Dimmel, in C. Heitner, D.R. Dimmel, J.A. Schmidt (Eds.), Lignin and lignans: advances in
chemistry, CRS Press, Boca Raton, FL, 2010, pp. 1–11.
[2] J. Zakzeski, P.C.A. Bruijnincx, A.L. Jongerius, B.M. Weckhuysen, Chem. Rev. 110 (2010) 3552–
3599.
[3] E. Dorrestijn, L. Laarhoven, I. Arends, P. Mulder, J. Anal. Appl. Pyrolysis 54 (2000) 153–192.
[4] C. Zhao, J. A. Lercher, Angew. Chem. 124 (2012) 6037–6042.
[5] J.He, L. Lu, C. Zhao, D. Mei, J. A. Lercher, J. Catal. 311 (2014) 41–51.
[6] L. Yang, Y. Li, P. E. Savage, Ind. Eng. Chem. Res. 53 (2014) 2633–2639.
[7] W. Scheppingen, E. Dorrestijn, I. Arends, P. Mulder J. Phys. Chem. A 101 (1997) 5404–5411.
[8] C. Zhao, J.A. Lercher, in: K. Triantafyllidis, A. Lappas, M. Stocker (Eds.),The role of catalysis
for the sustainable production of bio-fuels and bio-chemicals, Elsevier, 2013, pp.289–320.
[9] M.M. Hepditch, R. W. Thring, Can. J. of Chem. Eng. 78 (2000) 226–231.
[10] D.P. Mobley, A.T. Bell, Fuel 58 (1979) 661–666.
[11] N.D. Tayler, A.T. Bell, Fuel 59 (1980) 499–506.
[12] S.S. Salim, A.T. Bell, Fuel l61 (1982) 745–754.
24

[13] G.L. Huppert, B. C. Wu, S.H. Townsend, M.T. Klein, S. C. Paspek, Ind. Eng. Chem. Res. 28
(1989) 161–165.
[14] J.M.L. Penninger, R.J.A. Kersten, H.C.L. Baur, J. of Supercrit. Fluids 16 (1999) 119–132.
[15] P.E. Savage, Chem. Rev. 99 (1999) 603–621.
[16] D. Bröll, C. Kaul, A. Krämer, P. Krammer, T. Richter, M. Jung, H.Vogel, P. Zehner, Angew.
Chem. Int. Ed. 38 (1999) 2998–3014.
[17] X. Xu, C. De Almeida, M. J. Antal, Jr., J. Supercrit. Fluids 3 (1990) 228–232.
[18] R. Ma, W. Hao, X. Ma, Y. Tian, Y. Li, Angew. Chem. 126 (2014) 7438–7443.
[19] J. Lu, E. C. Boughner, C. L. Liotta, C. A. Eckert, Fluid Phase Equilib. 198 (2002) 37–49.
[20] T. Okuhara, Chem. Rev. 102 (2002) 3641–3666.
[21] F. Fringuelli, F. Pizzo, L. Vaccaro, Tetrahedron Lett. 42 (2001) 1131–1133.
[22] P.G. Jessop, J. of Supercrit. Fluids 38 (2006) 211–231.
[23] S. Kobayashi, M. Sugiura, H. Kitagawa, W.W.-L. Lam, Chem. Rev. 102 (2002) 2227–2302.
[24] A. Kruse, E. Dinjus, J. of Supercrit. Fluids 39 (2007) 362–380.
[25] M. Siskin, G. Brons, S. N. Vaughn, Energ. Fuels 4 (1990) 488–492.
[26] V. Roberts, S. Fendt, A. A. Lemonidou, X. Li, J. A. Lercher, Appl. Catal., B. 95 (2010) 71–77.
[27] G. Gonzalez, D. Montane, AIChE J. 51 (2005) 971–981.
[28] C.C. Tsao, Y. Zhou, X. Liu, T.J. Houser, J. of Supercrit. Fluids 5 (1992) 107–113.
[29] M. Nowakowska, O. Herbinet, A. Dufour, P.‐A. Glaude, Combust. Flame 161 (2014) 1474–1488.
[30]T.R. Varga, Z. Fazekas, Y. Ikeda, H. Tomiyasu, J. of Supercrit. Fluids 23 (2002) 163–167.
[31] X. Huang, T.I. Korányi, M.D. Boot, E.J.M. Hensen, ChemSusChem8 (2014) 2276–2288.
[32] K. Izutsu, Properties of solvents and solvent classification, in: Electrochemistry in
nonaqueoussolutions, second ed., Wiley-VCH, 2009,pp. 1–24.
[33] J. He, C. Zhao, D. Mei, J.A. Lercher, J. Catal. 309 (2014) 280–290.
[34] H. Tagaya, Y. Suzuki, T. Asou, J. Kadokawa, K. Chiba, Chem. Lett. 27(1998) 937–938.
[35] C.-W. Chen, A.-N.Ko, J. Chin. Chem. Soc. 58(2011) 1–11.
25

Figure captions:
Figure 1: Proposed reaction mechanism for BPE conversion in water.
Figure 2: GCGCchromatogram of the products of BPE alcoholysis in sc-ethanol in the presence of
CuCl2 at 400 °C after a reaction time of 3 h.
Figure 3: Proposed reaction mechanism for BPE alcoholysis in ethanol.
Figure 4: Proposed reaction mechanism of naphthalene formation from phenol in sc-ethanol (adopted
from [25]).
Figure 5: Proposed reaction mechanism for DPE hydrolysis in water.
Figure 6: GCGC chromatograms of the products of DPE alcoholysis in sc-ethanol in the presence
of CuCl2at 400 °C after a reaction time of 3 h.
Figure 7: Proposed reaction mechanism for DPE alcoholysis in ethanol.
Figure 8: Proposed reaction mechanism for DPM hydrolysis in sc-water at 350 °C ([H2] equivalent).
Graphical abstract
26

Figure 1: Proposed reaction mechanism for BPE conversion in water.
Figure 2:GCGCchromatogram of the products of BPE alcoholysis in sc-ethanol in the presence of
CuCl2 at 400 °C after a reaction time of 3 h.
27

Figure 3: Proposed reaction mechanism for BPE alcoholysis in ethanol.
28

Figure 4: Proposed reaction mechanism of naphthalene formation from phenol in sc-ethanol (adopted
from ref. [25]).
Figure 5: Proposed reaction mechanism for DPE hydrolysis in water.
29

Figure 6:GCGC chromatograms of the products of DPE alcoholysis in sc-ethanol in the presence of
CuCl2 at 400 °C after a reaction time of 3 h.
Figure 7: Proposed reaction mechanism for DPE alcoholysis in ethanol.
30

Figure 8: Proposed reaction mechanism for DPM hydrolysis in sc-water at 350 °C ([H2] equivalent).
Table captions:
31