Biomass conversion to diesel via the etherification of furanyl alcohols catalyzed by Amberlyst-15

Article abstract of DOI:10.1016/j.jcat.2014.02.012

The etherification of furanyl alcohols produced from biomass-derived glucose and fructose has been a growing area of research for production of alternative diesel additives. We have determined that the Bronsted acidic resin catalyst, Amberlyst-15, is highly active and selective for the etherification of furanyl alcohols by both ethanol and butanol. The mechanism and kinetics of this reaction were investigated using 5-methylfurfuryl alcohol (MFA) as a probe molecule. Etherification of MFA was found to be first order in both the concentrations of furanyl alcohol and the acid sites. The mechanism of MFA etherification also holds for the etherification of 2,5-bis(hydroxymethyl) furan (BHMF) and 5-(hydroxymethyl)furfural (HMF). In the case of HMF, we find that acetalization of HMF precedes etherification in alcohol solutions. The apparent activation energy of furanyl alcohol etherification in ethanol and butanol solutions ranged from 17.0 to 26.3 kcal/mol. Electron donation/withdrawal at the 2 or 5 position of the furan ring in addition to solvent polarity was found to have significant effects on the rate of furanyl alcohol etherification.

Full text of DOI:10.1016/j.jcat.2014.02.012

Journal of Catalysis 313 (2014) 70–79
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Biomass conversion to diesel via the etherification of furanyl alcohols
catalyzed by Amberlyst-15
⇑
Eric R. Sacia, Madhesan Balakrishnan, Alexis T. Bell
Department of Chemical and Biomolecular Engineering, University of California, Berkeley, 107 Gilman Hall, Berkeley, CA 94720, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The etherification of furanyl alcohols produced from biomass-derived glucose and fructose has been a
growing area of research for production of alternative diesel additives. We have determined that the
Brønsted acidic resin catalyst, Amberlyst-15, is highly active and selective for the etherification of furanyl
alcohols by both ethanol and butanol. The mechanism and kinetics of this reaction were investigated
using 5-methylfurfuryl alcohol (MFA) as a probe molecule. Etherification of MFA was found to be first
order in both the concentrations of furanyl alcohol and the acid sites. The mechanism of MFA etherifica-
tion also holds for the etherification of 2,5-bis(hydroxymethyl)furan (BHMF) and 5-(hydroxymethyl)fur-
fural (HMF). In the case of HMF, we find that acetalization of HMF precedes etherification in alcohol
solutions. The apparent activation energy of furanyl alcohol etherification in ethanol and butanol solu-
tions ranged from 17.0 to 26.3 kcal/mol. Electron donation/withdrawal at the 2 or 5 position of the furan
ring in addition to solvent polarity was found to have significant effects on the rate of furanyl alcohol
etherification.
Received 17 December 2013
Revised 16 February 2014
Accepted 25 February 2014
Keywords:
Furanyl alcohols
Etherification
Biofuels
Diesel
Amberlyst-15
HMF
Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction
An example of a furanyl ether is 5-(ethoxymethyl)furfural
(EMF). The energy density of EMF is 30.3 MJ/L, almost 30% higher
Both the European Union [1] and the United States [2] have
implemented targets for blending significant amounts of biofuels,
on the order of 10% and greater, into the traditional fuel supply
over the course of the next ten years. This requirement necessitates
the development of new biofuel technologies through the conver-
sion of lignocellulosic biomass [3]. Since the demand for diesel ex-
ceeded that for gasoline by 15% in 2009, and the growth in demand
for diesel is expected to outpace that of gasoline by at least a factor
of two through 2030 [4], there is a growing incentive to identify
strategies for converting biomass to diesel. While biodiesel can
be produced by transesterification of palm, rapeseed, or soybean
oil, the supply of these oils is limited and is in demand for human
nutritional use. An alternative approach that has received some
attention is the synthesis of furanyl ethers [5–8]. Such ethers re-
quire little to no hydrogenation [9] and can be blended into petro-
leum-derived diesel up to 17 wt% with no adverse effects on engine
performance and a decrease in particulate emissions [10] because
of the high cetane number of ether groups [11].
than that of ethanol and comparable to that of gasoline (31.1 MJ/
L) [12]. One approach for producing EMF with high yields involves
the formation of 5-(chloromethyl)furfural (CMF) as an intermedi-
ate by acid-catalyzed dehydration of sugars in a concentrated
aqueous HCl and LiCl solution [12]. CMF will then react with etha-
nol to form EMF and HCl with 95% yield [13]. One concern with this
approach is the separation and recycling of HCl. There are also sub-
stantial concerns about the introduction of unreacted chlorine-
containing products into the automobile fuel delivery system and
engine.
We have recently reported that EMF can be produced by acid-
catalyzed etherification of HMF and ethanol [5]. The aim of the
present study is to expand the scope of furanyl ethers that can
be produced by this approach (Fig. 1) and to understand the mech-
anism and kinetics of such reactions. In the hydrogenolysis of HMF
to 2,5-dimethylfuran (DMF), 5-(methyl)furfuryl alcohol (MFA) and
2,5-bis(hydroxymethyl)furan (BHMF) are formed as by-products
[14]. These compounds can undergo etherification to form prod-
ucts that could be even more attractive than EMF as diesel blend-
ing agents because they do not contain aldehyde groups that are
known to decrease the cetane number of diesel components [11].
A further goal of this study is to elucidate the mechanism and
kinetics of furanyl ether formation with the aim of identifying
⇑
Corresponding author.
E-mail addresses: Eric.Sacia@berkeley.edu (E.R. Sacia), Balki@berkeley.edu
(M. Balakrishnan), bell@cchem.berkeley.edu (A.T. Bell).
http://dx.doi.org/10.1016/j.jcat.2014.02.012
0021-9517/Ó 2014 Elsevier Inc. All rights reserved.

E.R. Sacia et al. / Journal of Catalysis 313 (2014) 70–79
71
Fig. 1. Transformation of the HMF derived furanyl alcohols into fuel candidate molecules.
conditions for maximizing the yield of these products and under-
standing the role of the catalyst in determining selectivity.
a refrigerated cold room that was maintained at 4 °C. Etherification
reactions for HMF occurred at temperatures above the boiling
point of the solvent and were performed in sealed headspace vials
equipped with a silicone rubber septum for online sampling. In the
cases where online sampling was required, filtration of solid cata-
lysts was performed in situ during sample removal by passage
through a packed-cotton filter (see diagram in Supplemental
information).
Solutions of reactants and inert internal standard were allowed
to reach the desired reaction temperature prior to catalyst addi-
tion. If less than 4.0 mg of Amberlyst-15 was required or in all
cases for HMF etherification, the catalyst was dispensed as a
5.00 mg/mL suspension in the alcohol in which the reaction was
carried out. Otherwise, the catalyst was added directly to the reac-
tor as a solid. Samples were removed after a specified reaction
time. When soluble acids were used as the catalyst, the reaction
mixture was neutralized with NaHCO3 at the end of the
reaction period. Heterogeneous catalysts were removed by
filtration through a column packed with cotton, silica gel, and
NaHCO3 to neutralize protons liberated by ion exchange. Ethyl
acetate was then used to wash filtration columns and dilute the
solutions to the necessary concentrations for evaluation by gas
chromatography.
2. Materials and methods
2.1. Experimental materials
All chemicals were used as received. 5-(hydroxymethyl)furfural
(HMF) was purchased from Sigma–Aldrich in >99% purity. Both
Amberlyst™ 15 (A-15) and NafionÒ SAC-13 were purchased from
Sigma–Aldrich. Methanol, 1-propanol, and 2-propanol used for
etherification were obtained from Fisher Scientific in >99% purity.
Absolute ethanol was purchased from Acros Organics in >99.5%
purity, while 1-butanol was purchased from Mallinckrodt Chemi-
cals in >99.4% purity. Ethyl acetate and hexanes (HPLC Grade), used
for dilution and reaction respectively, and anhydrous Na2CO3 and
NaHCO3, used for neutralization, were obtained from Fisher Scien-
tific. N,O-bis(trimethylsilyl)trifluoroacetamide with 1% TMSCl was
purchased from Acros Organics and was used for silylation of alco-
hols to enhance separation by gas chromatography.
The reactants MFA and BHMF were synthesized via sodium
borohydride reduction of 5-(methyl)furfural and HMF, respec-
tively, and were purified for use in standard techniques (see Sup-
plemental
information).
Calibration
standards for
gas
chromatography for the ethyl ethers of MFA, BHMF, and HMF were
synthesized from experimentally determined reaction conditions
and purified through column chromatography (see Supplemental
information). The purity of the synthesized compounds was veri-
fied by GC and NMR.
2.3. Product analysis
Product analysis was carried out using a Varian CP-3800 gas
chromatograph equipped with both a flame ionization detector
(FID) and a Varian 320 triple quadrupole mass spectrometer
(MS). Compounds analyzed were separated in both cases using a
FactorFour capillary column (VF-5 ms, 30 m length, 0.25 mm diam-
eter) coated with a 0.25 mm thick stationary phase (5% phenyl and
95% dimethylpolysiloxane). All quantification was performed using
the FID detector to achieve high signal to noise, while product
identification was performed by mass spectrometry. To maximize
accuracy of quantification, dodecane was used as an internal stan-
dard to rectify errors caused by variability in sample size delivery
and detector signal strength [15]. FID response factors for ethers
other than the synthesized ethyl ethers were estimated using the
effective carbon number method, which predicts FID response to
within 1.7% [15–17]. To achieve sufficient baseline separation of
2.2. Etherification reactions
The etherification of furanyl alcohols was typically carried out
using 10–100 mM concentrations of the furanyl alcohol, 0.9–
35.5 mol% acid, and either ethanol or butanol as the solvent. Reac-
tions below the boiling point of the solvent were performed using
magnetic stirring at 600 RPM in sealed scintillation vials on an IKA
RCT Basic stir plate equipped with an IKA ETS-D5 temperature
controller. Reactions at 0 °C were carried out in an ice bath with
a
thermocouple to monitor solution temperature. Reactions
between 0 °C and 25 °C were carried out on a heated stir plate in

72
E.R. Sacia et al. / Journal of Catalysis 313 (2014) 70–79
ethyl ethers of HMF and BHMF, reaction products were silylated
with N,O-bis(trimethylsilyl)trifluoroacetamide with 1% TMSCl at
70 °C for 1 h, providing consistent, baseline resolved peaks.
rate of EMMF formation varied with the extent of acid dissociation
[19], the selectivity toward EMMF was independent of acid compo-
sition, 61.5% 5.4% (Table 1). The yield of EMMF was slightly lower
in the case of triflic acid because the high acid strength resulted in
increased decomposition of EMMF. The similarity in selectivity of
different homogenous acids suggests that the etherification of
MFA occurs via general acid catalysis in homogeneous solutions.
Several heterogeneous acids were explored in an attempt to
maximize the yield of desired products through specific acid catal-
ysis while also increasing the ease of catalyst recyclability.
Brønsted acidic zeolites were considered because of their high sur-
face areas and known shape selectivity [20,21]. As can be seen in
Table 1, H-FAU had a lower selectivity than H-MFI, presumably
due to its large supercage structure which more readily allows
the formation of larger oligomers of MFA. Since zeolites were ob-
served to deactivate due to oligomer formation and consequent
plugging of the catalyst pores, these catalysts were not investi-
gated further.
2.4. Isotopic exchange of ether group of EMMF
Deuterated ethanol (ethanol-1,1,2,2,2-d5, 99.5% atom D) was pur-
chasedfromAldrich to measureethyl ether exchange rates for 2-(eth-
oxymethyl)-5-methylfuran (EMMF). EMMF that was synthesized and
purified by column chromatography was immersed in a solution of
the deuterated ethanolata concentration of10 mMat25 °C. A disper-
sionofA-15, ground and sievedto 38–53 lm, was addedsuchthatthe
solutionaveraged catalystconcentrationwas0.09 mM. Sampleswere
taken periodically, centrifuged to remove catalyst, and run on a Var-
ian CP-3800 gas chromatograph with Varian 320 triple quadrupole
mass spectrometer to assess rates of ethyl exchange.
2.5. 13C NMR investigation of A-15
High product selectivity could be achieved with Nafion SAC-13,
a dispersion of a sulfonated tetrafluoroethylene based fluoro-
copolymer over high surface area silica (Table 1, entry 6); however,
this material also slightly deactivated with time. Near quantitative
conversion of MFA to EMMF was achieved with Amberlyst-15 (A-
15), a sulfonated polystyrene resin with a high density of acid sites
(4.7 mol/kg) and large pores (30 nm) [22–23]. Moreover, this cata-
lyst could be recycled a number of times without a loss in activity
(see Supplemental information). Due to its exceptional reactivity
and selectivity, A-15 was chosen as an appropriate catalyst with
which to perform the remainder of the studies reported here.
To assess the extent to which intraparticle mass transfer affects
the activity of A-15, experiments were conducted with particles of
progressively smaller diameter. Table 2 shows that as the diameter
Interactions of alcohols with the active site of A-15 were inves-
tigated using inverse-gated, proton-decoupled 13C NMR. Experi-
ments were carried out on a Bruker AV-600 instrument using the
zgig30 pulse program. Data analysis was performed using MNova
7 [18]. A typical experiment involved adding 0.50 g of a 50:50
(wt%) mixture of the alcohol and CDCl3 to 0.20 g of either A-15
or Amberlite XAD-1180 N that had been dried overnight in vacuo
at 65 °C. Signals were then acquired for 1000 scans. Ethanol and
2-furanethanol were used as probe molecules to measure surface
interaction with A-15 at 25 °C. 2-furanethanol was synthesized
and purified before use via Wittig olefination of furfural followed
by hydroboration/oxidation. The full procedure is described in
the Supplemental information.
was reduced from 600–800 lm to <38 lm, the activity per proton
increased 6.6-fold. Since particles of the smallest three size ranges
exhibited activity within 6% of the mean, the smallest particles of a
3. Results and discussion
defined size range, 38–53 lm, were used for measurements of
3.1. Catalyst selection
reaction kinetics. A calculation of the Thiele parameter was carried
out in order to further confirm the absence of intraparticle mass
transfer effects. The Wilke–Chang correlation was used to estimate
the effective diffusivity at 25 °C and the method of Le Bas was used
to estimate the molecular volume of MFA at the normal boiling
The etherification of MFA with ethanol was chosen to evaluate
the effectiveness of different acid catalysts. Trifluoroacetic acid, p-
toluenesulfonic acid (p-TSA), sulfuric acid, and triflic acid were
used as soluble acids for the reaction of MFA with ethanol to form
2-(ethoxymethyl)-5-methylfuran (EMMF) (Scheme 1). While the
point [24]. From these estimates, the Thiele modulus,
u, was
Scheme 1. Acid catalyzed etherification of MFA in an ethanol solution shown with presence of other oligomers.
Table 1
Screening of solid and soluble acids for the etherification of MFA to EMMF.
Entry
Acid catalyst
Reaction time (min)
MFA conversion (%)
EMMF yield (%)
EMMF selectivity (%)
1
2
3
4
5
6
7
p-TSA
120
120
60
120
120
120
120
94.3
97.1
100.0
81.9
48.1
89.8
99.7
60.0
63.7
55.4
64.6
42.8
83.1
>98%
63.6
65.6
55.4
78.9
89.0
92.5
>98%
Sulfuric acid
Triflic acid
H-FAU
H-MFI
Nafion SAC-13
A-15
In all experiments, starting materials were mixed, added to a heating plate at 25 °C, stirred at 600 RPM, and vials were removed and the acid was neutralized at indicated time
points. Conditions: Entry 1–3 – MFA (83.3 mM), 35.5 mol acid (35.5 mol%), 1.2 mL ethanol; Entry 4–5 – MFA (83.3 mM), zeolite (Si/Al = 15) 17.7 mol acid sites (35.5 mol%
acid sites), 0.6 mL ethanol; Entry 6 – MFA (83.3 mM), 35.5 mol acid sites on SAC-13 (35.5 mol% acid sites given previously measured acid loading [31]), 1.2 mL ethanol; Entry
7 – MFA (83.3 mM), 35.5 mol acid sites on A-15 (35.5 mol% acid sites), 1.2 mL ethanol.
l
l
l
l

E.R. Sacia et al. / Journal of Catalysis 313 (2014) 70–79
73
Table 2
Effect of particle size on the rate of MFA etherification catalyzed by A-15.
A-15 particle size (diameter –
lm)
Initial TOF (hꢁ1
)
600–800
75–125
53–75
38–53
<38
0.29
1.39
1.66
1.72
1.85
Reaction conditions: 25 °C, 10 mM MFA, 35.5
l
mol acid sites on A-15, 10 mL
ethanol.
100
80
60
40
20
0
CMFA
CEMMF
CMeas. Oligomers
0
20
40
60
80
100
120
140
Scheme 2. Proposed mechanism for etherification of MFA with ethanol.
Time (min)
Fig. 2. Time course study of MFA etherification in ethanol. Conditions: 25 °C,
83.3 mM MFA, 1.2 mL EtOH, 35.5 mol active sites on A-15 (as received).
monly the case for SN1 reactions of this type [19], the following
rate expression can be written:
l
dCEMMF
dt
k1k2
ꢁ1 þ k2
determined to be 0.37 for particles in the range of sizes used (38–
53 m). A Thiele modulus of less than one confirms that all mea-
surements of reaction kinetics were made in the reaction rate-lim-
þ
þ
¼ rEMMF ¼ TOFEMMF ꢀ CH
¼
CH CMFA
k
l
þ
¼ keff CH CMFA
ð1Þ
ited regime.
Since the rate of equilibration is much larger than the rate-lim-
iting step, the equation for the turnover frequency (TOF) may be
simplified as:
3.2. Etherification of MFA with Amberlyst-15
3.2.1. Kinetics of etherification
TOFEMMF ¼ K1k2CMFA ¼ keff CMFA
In Eq. (2), K1 = k1/k
ð2Þ
The kinetics of MFA etherification with ethanol was measured
in order to determine the reaction orders with respect to reactants
and the dependence of the reaction rate on temperature. The
reaction was zero order in concentrations of ethanol and water
over the ranges tested, whereas the reaction was nearly first order
(0.92 0.1) in the concentration of acid sites and MFA (0.95 0.1).
An example of the temporal evolution of the product is given in
Fig. 2, and a plot of the initial rate of MFA etherification as a func-
tion of reactant and catalyst concentrations is presented in the
Supplemental information.
Due to the position of the hydroxyl group on the carbon adja-
cent to the aryl ring of MFA, the reaction should proceed through
an SN1 mechanism, as shown in Scheme 2. Water formed by pro-
tonation of the alcohol acts as the leaving group. Since the primary
carbocation formed is unstable, the intermediate rearranges to
form the more stable oxonium ion, as shown in step (2) of
Scheme 2. The absence of diethyl ether from the products is attrib-
uted to the instability of the primary carbocation that would be
necessary to form this product at the temperatures used in this
study. The kinetics of this reaction was modeled using the Helffe-
rich approach, a method often used to describe liquid-phase
catalysis with polymeric resin catalysts, such as A-15, in the ab-
sence of transport limitations [25]. In this approach, the liquid–so-
lid interaction is treated as a pseudo-homogeneous process since
the reaction occurs in the solvation shell of the site. Assuming that
the second step in Scheme 2 is the rate-limiting step, as is com-
.
ꢁ1
The observations reported above indicate that a pseudo-homo-
geneous model provides a satisfactory description of the intrinsic
reaction kinetics. Further, it has been shown previously that, in
the absence of diffusion limitations, the reaction rate order in the
heterogeneous system matches that of the homogeneous system
for resin catalysts [25]. For this reason, the reaction is treated as
occurring at the liquid–solid interface of the cationic resin catalysts
in which polar species are enriched in the strongly acidic phase
near the surface [23,25–26].
Values of keff were determined by linear least squares regres-
sion of the initial rate of furanyl ether formation measured at
low MFA conversion (<10%) and are listed in Table 3 for ethanol
and butanol. The activation energies for MFA etherification in eth-
anol and 1-butanol were determined from Arrhenius plots (see
Supplemental information). The apparent activation energy for
etherification of MFA with ethanol is 17.0 0.6 kcal/mol and that
for etherification with 1-butanol is 19.9 1.0 kcal/mol (Table 4).
The degree to which the formation of EMMF is reversible (see
steps (3) and (4) of Scheme 2) was tested using isotopically labeled
ethanol-1,1,2,2,2-d5. The steps by which this reaction is thought to
occur are shown in Scheme 3. Ethanol deuterium-labeled at all but
the alcoholic hydrogen was used to prevent kinetic isotope
effects from affecting the observed exchange rate. The rate at
which the isotopic exchange occurred at 298 K was determined

74
E.R. Sacia et al. / Journal of Catalysis 313 (2014) 70–79
Table 4
by monitoring the intensities at m/z 140 and 145 (EMMF and
EMMF-d5, respectively). These experiments were carried out using
molar concentrations of EMMF representative of those occurring in
the measurements of the etherification of MFA. A plot displaying
the initial exchange rate can be found in the Supplemental infor-
mation. The ratio of the rate of MFA etherification over the rate
of isotopic exchange at 25 °C is as follows:
Apparent activation energies for MFA, BHMF, and HMF etherification with ethanol
and butanol.
Furanyl alcohol
Eapp (kcal/mol)
Pre-exp. factor (Mꢁ1/min)
Ethanol
Butanol
Ethanol
Butanol
MFA
17.0 0.6
18.1 0.8
20.0 1.4
25.7 2.5
19.9 1.0
18.6 1.6
19.3 0.9
26.3 1.5
3.53 ꢀ 1013
1.48 ꢀ 1013
3.49 ꢀ 1013
1.24 ꢀ 1016
4.96 ꢀ 1016
7.45 ꢀ 1014
4.78 ꢀ 1013
4.80 ꢀ 1016
BHMF (1st ether)
BHMF (2nd ether)
HMF
TOFEMMF
¼ 101
ð3Þ
TOFEMMF;exch
where TOFEMMF is the initial turnover frequency of EMMF formation
and TOFEMMF,exch is the initial exchange rate of the deuterated ether
group. The much lower rate of ether exchange compared to the ini-
tial rate of MFA etherification is attributable to the lower pKa of pro-
tonated ethers compared to alcohols [19], making the ether more
difficult to protonate, and the decreased polarity of EMMF com-
pared to MFA, making partitioning into the polar surface solvation
shell more difficult.
analogous to A-15 but not containing sulfonic acid groups, is
shown in spectrum A of Fig. 4. A single NMR peak is observed at
57.4 ppm corresponding to the methylene unit in a bulk chemical
environment. The absence of a shift in the position of this peak rel-
ative to that for bulk ethanol indicates that the ethanol in the pores
of Amberlite XAD-1180N interacts weakly with the unfunctional-
ized pore environment. By contrast, the spectrum of ethanol ad-
sorbed into A-15 (spectrum B) exhibits a peak at 57.4 ppm for
bulk ethanol and a second peak at 58.4 ppm corresponding to eth-
anol in exchange with the acid sites of A-15 over the timescale of
NMR relaxation. Interaction with the acidic proton of A-15 leads
to deshielding of the alcoholic carbon and a corresponding shift
downfield for the second peak. The Bloch–McConnell equations
can be used to describe the chemical exchange between the bulk
and the first coordination sphere that is observed over the NMR
timescale [27,28]. Relative peak intensities were obtained using a
deconvolution algorithm for fitting Lorentzian peaks contained in
MNova 7 [18]. The significant size of the peak at 58.4 ppm indi-
cates that a large fraction of ethanol molecules have diffused
through the bulk and undergone an equilibrated exchange with
the surface over the course of the NMR relaxation time.
Similar to what is observed for ethanol, the alcoholic carbon of
2-furanethanol in the pores of Amberlite XAD-1180N displays a
single peak located at 60.3 ppm (Fig. 4, spectrum C). When 2-fura-
nethanol is mixed with A-15, a splitting similar to that seen for
ethanol is observed (spectrum D), with the peak for the bulk alco-
hol appearing at 60.3 ppm and the peak for surface-exchanged
alcohol appearing at 60.6 ppm. The spectrum obtained upon addi-
tion of a 25:25:50 (wt%) mixture of 2-furanethanol, ethanol, and
CDCl3 to A-15 is given in spectrum E. Disappearance of the surface
peak for 2-furanethanol while both ethanol peaks remain indicates
that ethanol outcompetes 2-furanethanol for interaction with the
acidic site as a consequence of its smaller size and significantly
higher polarity.
3.2.2. The effect of A-15 on etherification selectivity
We hypothesize that, in the absence of pore-size constraints,
the high EMMF selectivity of A-15 is due to decreased partitioning
of the less polar furanyl alcohol compared to the more polar sol-
vent alcohol into the solvation shell of A-15’s acid sites. This idea
is illustrated in Fig. 3. Once a furanyl alcohol molecule enters this
acidic surface phase, it can react to form the oxonium ion and sub-
sequently reacts with an ethanol molecule (Fig. 3c) to produce
EMMF. It is well understood for SN1 reactions that, once the cation
is formed, the intermediate is relatively unselective with respect to
the nucleophilicity of the nucleophile during step 3 in Scheme 2
[19]. By increasing the ratio of the solvent alcohol (ethanol of buta-
nol) with respect to the furanyl alcohol in the phase around the
site, reactions forming crossed ethers are favored over self-cou-
pling of furanyl alcohols. The effects of this phenomenon become
apparent when equimolar quantities of methanol, ethanol, 1-pro-
panol, 2-propanol, and 1-butanol were added to a hexane solution.
As supported by the data given in Table 5, the smaller, more polar
alcohols react at higher rates with MFA than the larger, less polar
alcohols.
To confirm the role of surface partitioning as a governing factor
controlling selectivity, the surface phase of the catalyst was probed
using inverse-gated, proton-decoupled 13C NMR. Fig. 4 displays the
position of the alcoholic carbon peaks of ethanol and 2-furanetha-
nol. 2-Furanethanol was used as a proxy for MFA due to its similar
size and polarity; however, it also exhibits limited reactivity due to
the presence of an additional carbon atom between the alcohol
group and the ring. The additional carbon prevents primary carbo-
cation rearrangement and leads to relative inactivity over the time
period of NMR measurement.
The results of our NMR studies are consistent with the picture
presented in Fig. 3 and confirm that the alkyl alcohols are adsorbed
preferentially into the pores of A-15. A direct consequence of the
preferred adsorption of the alkyl alcohols is the high selectivity
for forming EMMF, as shown in Fig. 2. A loss in EMMF selectivity
does occur as the concentration of MFA in ethanol is raised; how-
The spectrum of ethanol adsorbed in the pores of Amberlite
XAD-1180N, a cross-linked polystyrene divinylbenzene copolymer
Table 3
Rate constants for etherification of MFA, BHMF, and HMF in ethanol and butanol.
MFA
BHMF
Ethanol
T
HMF
Ethanol
Butanol
T
Butanol
T
Ethanol
T k
eff
Butanol
T
(a)
(a)
(a)
(b)
(c)
(c)
(d)
(d)
T
(K)
k
eff
k
eff
keff 1
,
T
keff,2
keff,1
T
keff,2
keff
(Mꢁ1/min)
(Mꢁ1/min) (K)
(Mꢁ1/min) (K)
5.92 313.2
280.6 13.0 323.2
(Mꢁ1/min) (K)
(Mꢁ1/min) (K)
(Mꢁ1/min) (K)
(Mꢁ1/min) (K)
(Mꢁ1/min) (K)
273.2
0.86
273.2
3.20
9.1
314.2 0.41
300.2
305.2
313.2
22.6
34.5
66.4
300.2 0.42
353.2 1.50
333.2 0.27
298.2 14.0
305.7 24.6
313.2 45.4
323.2 1.18
333.2 3.1
343.2 6.0
305.2 0.69
313.2 1.76
321.2 3.31
358.2 2.24
363.2 4.84
368.2 7.58
373.2 9.45
339.8 0.55
346.5 1.09
353.2 2.68
286.7 32.3
293.5 72.3
333.2 19.8
343.2 41.5
321.2 179
Reaction conditions: (a) 10 mM MFA/BHMF, 10.0 mL EtOH/BuOH solvent, 0.90
lmol acid sites on A-15 (38–53 lm), (b) 10 mM BHMF, 10.0 mL EtOH solvent, 18 lmol acid
sites on A-15 (38–53
lm), (c) 10 mM BHMF, 7.0 mL BuOH solvent, 7.0
lmol acid sites on A-15 (38–53 lm) and (d) 100 mM HMF, 50.0 mL EtOH/BuOH solvent, 0.11 mmol acid
sites on A-15 (38–53
lm).

E.R. Sacia et al. / Journal of Catalysis 313 (2014) 70–79
75
Scheme 3. Reaction performed during isotopic exchange of EMMF to measure reversibility of ether formation.
(a)
(b)
(c)
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
O
OH
HO
OH
H2O
O+
O
OH
OH
HO3S
H
O
A-15 Pore
H
HO
O-
S
HO
O
S
OH
OH
OH
OH
OH
O
S
O
O
O
O
OH
OH
OH
OH
OH
OH
OH
Fig. 3. Schematic of active site occupation during the course of reaction: (a) before furanyl alcohol enters acid solvation shell, (b) upon furanyl alcohol reaching surface acid
sites and (c) upon formation of the reactive intermediate.
Table 5
Competitive reaction experiment forming various crossed ethers.
Alcohol incorporated into crossed ether
Initial TOF (hꢁ1
)
Methanol
Ethanol
1-Propanol
n-Butanol
2-Propanol
1.12
0.93
0.27
0.12
0.11
Reaction conditions: 25 °C, 10 mM MFA, 20 mM each alkyl alcohol, 0.90
sites on A-15 (38–53 m), 10.0 mL hexane as solvent.
l
mol acid
l
ever, even a nearly eightfold increase in the concentration of MFA
from 1.2 to 8.6 wt% only decreased the selectivity to EMMF from
98% to 84%.
3.3. Etherification of BHMF
Fig. 4. 13C NMR spectra of the alcoholic carbon of ethanol and 2-furanethanol in
Amberlite XAD-1180N and A-15.
The etherification of furanyl alcohols was extended to other C6
furanyl alcohols, including HMF and BHMF. The products formed
during the etherification of BHMF with ethanol are shown in
Scheme 4. Fig. 5 shows that the formation of 5-(ethoxymethyl)furfu-
ryl alcohol (EMFA) occurs much more rapidly than the subsequent
etherification of EMFA to form 2,5-bis(ethoxymethyl)furan (BEMF).
The second step, in which EMFA reacts to form BEMF, occurs nearly
quantitatively, while the first step occurs with lower selectivity
(ꢂ80%). The observed pattern in selectivity is likely attributable to
the higher polarity of BHMF with respect to the single ether, EMFA,
allowing BHMF to partition better into the surface phase of the cat-
alyst and form increased amounts of undesired oligomers and side
products. This hypothesis is further supported by the low relative
rates of the second reaction of EMFA to BEMF despite similar
amounts of electron donation to the ring for BHMF and EMFA.
The kinetics of the sequential etherification of BHMF is first or-
der in acid and furanyl alcohol and zero order in ethanol and water.
By analogy with the kinetics for the etherification of MFA, we as-
sume that the protonation of BHMF and EMFA is close to being
equilibrated and that the SN1 formation of the oxonium cation is
rate limiting. The rate of consumption of BHMF and the rates of for-
mation of unidentifiable oligomer products, EMFA, and BEMF are
then given by:
dCBHMF
dt
þ
þ
þ
¼ rBHMF ¼ TOFBHMF ꢀCH ¼ ꢁkeff;1CH CBHMF ꢁkUCH CBHMF ð4Þ
dCU
dt
þ
þ
¼ rU ¼ TOFU ꢀCH ¼ kUCH CBHMF
ð5Þ
dCEMFA
dt
þ
þ
þ
¼ rEMFA ¼ TOFEMFA ꢀCH ¼ keff;1CH CBHMF ꢁkeff;2CH CEMFA ð6Þ
dCBEMF
dt
þ
þ
¼ rBEMF ¼ TOFBEMF ꢀCH ¼ keff;2CH CEMFA
ð7Þ
Solving Eqs. (4)–(7) results in the following expression for the
temporal evolution of products:
½ꢁðkeff;1þkUÞCHþ ꢃtꢄ
CBHMF ¼ CBHMF;0 ꢃe
ð8Þ
Â
Ã
keff;1CBHMF;0
keff;2 ꢁkeff;1 ꢁkU
keff;1CBHMF;0
keff;2 ꢁkeff;1 ꢁkU
eff;1þkUꢄCHþ
eff;2CHþ ꢃtÞ
CEMFA
¼
ꢃ eðꢁ½k
ꢃtÞ ꢁeðꢁk
ð9Þ
ꢀ
ꢁ
À
Á
keff;2
eff;2CHþ
eff;1þkUꢄCHþ ꢃtÞ
CBEMF
¼
ꢃ eðꢁk
ꢃtÞ þ
1ꢁeðꢁ½k
ꢁ1
keff;1 þkU
ð10Þ
Â
Ã
kUCBHMF;0
keff;1 þkU
eff;1þkUꢄCHþ ꢃtÞ
CU
¼
ꢃ 1ꢁeðꢁ½k
ð11Þ

76
E.R. Sacia et al. / Journal of Catalysis 313 (2014) 70–79
In Eqs. (8)–(11), CBHMF,0 is the initial concentration of BHMF.
Values for keff,1 and keff,2 were determined by a least-squares fit
of Eqs. (8)–(11) to the data presented in Fig. 5 and are given in Ta-
ble 3. The solid curves appearing in Fig. 5 were generated by plot-
ting Eqs. (8)–(11) using the values of keff,1 and keff,2 listed in Table 3.
The excellent fit of the model to the data is evident. Apparent acti-
vation energies for each etherification step were obtained from an
Arrhenius plot of the natural log of keff,1 and keff,2 versus inverse
temperature and are listed in Table 4. The results presented in Ta-
ble 3 show that the etherification of BHMF proceeds at higher tem-
peratures than the etherification of MFA. The latter compound is
more reactive than the former because the electron-donating
methyl group on MFA helps to stabilize the transition state for
the rate-limiting step.
Scheme 4. Proposed pathway for BHMF etherification with ethanol.
As mentioned previously, BHMF displays lower combined selec-
tivity to EMFA and BEMF due to the high partitioning of BHMF into
the solvation environment of the acid sites in A-15. Since SN1 reac-
tions tend to not be selective with regard to nucleophilicity [19],
once the active oxonium ion is formed, a higher local concentration
of the furanyl alcohol around the site will cause a decrease in the
selectivity for crossed furanyl ethers. Therefore, there is a signifi-
cant decrease in selectivity to desired products as the concentra-
tion of BHMF is increased. An increase in the concentration of
BHMF from 1.3 to 9.7 wt% decreased the selectivity to BEMF from
79% to 53%.
CBHMF
CEMFA
CBEMF
10
8
6
4
2
3.4. Etherification of HMF
0
0
50
100
150
200
The temporal evolution of ether products formed during the
etherification of HMF by ethanol at 80 °C is presented in Fig. 6.
The etherified products can also react further to form ethyl levuli-
nate and ethyl formate in stoichiometric ratios at long reaction
times and high temperatures and acid loadings [5]. Due to the lim-
ited formation of these products over the timescales studied, the
kinetics of their formation was not investigated.
While HMF can undergo direct etherification, acetalization of
HMF prior to etherification is preferred in alcohol solutions (see
Scheme 5). As discussed previously, the rate-limiting step in the
etherification of furanyl alcohols involves the formation of an oxo-
nium ion, and the reaction is facilitated at milder conditions by
electron donation to the furan ring. Since HMF has an aldehyde
group that is highly electron withdrawing due to the mesomeric
effect [19], acetalization is preferred prior to etherification.
To test the validity of the reaction sequence shown in Scheme 5,
the equilibrium in the first step was shifted from 5-(hydroxy-
methyl)furfural diethylacetal (HMFDEA) to HMF by the addition
of water. The observed decrease in HMFDEA concentration led to
a corresponding decrease in the rate of formation of etherified
products (5-(ethoxymethyl)furfural diethylacetal (EMFDEA) and
5-(ethoxymethyl)furfural (EMF)), showing that HMFDEA is the ac-
tive intermediate for etherification (see Supplemental
information).
Time (min)
Fig. 5. Temporal product distribution during BHMF etherification with ethanol.
Reaction conditions: 60 °C, 10 mM BHMF, 10.0 mL ethanol as solvent, 18 mol acid
sites on A-15 (38–53 m).
l
l
concentrations of the furanyl alcohols and ethers were determined
by GC and the amount of water was determined initially by Karl
Fischer titration and by a material balance at subsequent data points.
The values of K and Kb at each temperature were determined by
a
averaging the measured values of the equilibrium constants found
throughout the course of reaction. By this means, it was determined
that, similar to the acetalization of benzaldehyde [29], formation of
the acetal is exothermic, resulting in a higher concentration of alde-
hyde products with increasing temperature. The enthalpy of reac-
tion for the equilibration of HMF with its dibutyl acetal,
ꢁ16.0 2.3 kJ/mol, and for the second equilibration step,
DH , is
a
DHb,
100
CHMF
CHMFDEA
CEMF
CEMFDEA
80
As shown in Fig. 6, equilibrium between the aldehyde and the
acetal is established rapidly due to the significantly higher rate of
acetalization compared to etherification. Therefore, the acetaliza-
tion of HMF and EMF can be treated as quasi-equilibrated. The cor-
60
40
20
0
responding equilibrium constants K and Kb are defined by Eqs.
a
(12) and (13):
CHMFDðRÞA
C
H2
ka
O
Ka
Kb
¼
¼
¼
¼
ð12Þ
ð13Þ
CHMFC2
k
ꢁa
ROH
0
50
100
150
200
C
ðRÞMFC2ROH
CðRÞMFDðRÞAC
kb
Time (min)
k
ꢁb
H2O
Fig. 6. Temporal product distribution during HMF etherification with ethanol.
Reaction conditions: 80.0 °C, 100 mM HMF, 50.0 mL ethanol solvent, 0.11 mmol
Values of K and Kb were determined at each data point by
a
measuring the concentrations of furanyl alcohols and ethers. The
acid sites on A-15 (38–53 lm).

E.R. Sacia et al. / Journal of Catalysis 313 (2014) 70–79
77
Scheme 5. Proposed pathway for HMF etherification with ethanol.
was determined to be 24.7 6 kJ/mol during the etherification of
3.5. Effects of solvent environment
HMF with butanol. Values for the equilibrium constants can be
found in the Supplemental information.
As previously discussed, the etherification step was assumed to
be first order in the concentration of the acid sites and the furanyl
alcohol (HMFDEA in Scheme 5) and zero order with respect to eth-
anol and water concentration, giving the following rate expression:
To this point, we have explored the mechanism of furanyl alco-
hol etherification and the measured kinetic parameters in ethanol
and butanol catalyzed by A-15 for several substituents on the furan
ring. In order to further elucidate the role of the alcohol solvent on
MFA, BHMF, and HMF etherification, the effect of the solvent on the
rate of etherification was also investigated. Fig. 7 shows the rate
ꢁðTOFHMF þ TOFHMFDEAÞ ¼ ðTOFEMF þ TOFEMFDEA
¼ keff CHMFDEA
Þ
ð14Þ
80
(a)
The temporal dependence of the concentration of each compo-
nent is then given by Eqs. (15)–(19) for any alcohol solvent:
MFA
25ºC
60
dCHMF
dt
¼ ꢁkaCHMFC2 þ k aCHMFDðRÞACH
ð15Þ
ꢁ
2O
ROH
40
20
0
dCHMFDðRÞA
dt
¼ kaCHMFC2 ꢁ k aCHMFDðRÞACH
ꢁ
2O
ROH
þ
ꢁ keff CH CHMFDðRÞA
ð16Þ
dC
ðRÞMFDðRÞA
¼ ꢁkbCðRÞMFDðRÞACH O þ kꢁbCðRÞMFCR2OH
2
dt
160
(b)
þ
þ keff CH CHMFDðRÞA
ð17Þ
ð18Þ
BHMF
60ºC
dC
dt
ðRÞMF
¼ kbCðRÞMFDðRÞACH O ꢁ kꢁbCðRÞMFCR2OH
120
2
80
40
dC
dt
H2
O
¼ kaCHMFC2 ꢁ k aCHMFDðRÞA
C
O þ keff CH CHMFDðRÞA
þ
ꢁ
H2
ROH
ꢁ kbCðRÞMFDðRÞACH O þ kꢁbCðRÞMFCR2OH
ð19Þ
2
There was an insufficient number of data points at short reac-
tion times to accurately obtain values for k , k , kb, and k in
a
ꢁa
ꢁb
the equilibration of acetals and aldehydes due to the high rate of
acetalization on the timescale of etherification. Therefore, values
0
8
(c)
HMFD(R)A
of k and kb were chosen such that the acetalization rate would
a
be significantly faster than the rate of etherification, and values
80ºC
for k and k were determined using Eqs. (12) and (13) and the
ꢁa
ꢁb
6
measured equilibrium constants.
The effective rate coefficient for the etherification step, keff, was
then determined by fitting the concentration profiles predicted by
Eqs. (15)–(19) to the experimental data points for HMF etherifica-
tion in ethanol and 1-butanol using least-squares regression. The
curves through the data shown in Fig. 6 exemplify the quality of
fit for our model of the kinetics. Measured values of keff and the
activation energies for HMF etherification in ethanol and butanol
are given in Tables 3 and 4, respectively, whereas the Arrhenius
plots from which the activation energies were derived are shown
in the Supplemental information.
4
2
0
10
15
20
25
30
35
Alcohol Dielectric Constant (ε)
Finally, the extent to which HMF selectivity to the desired EMF
and EMFDEA products could be maintained was also explored. For
HMF concentrations of 1.3 wt% in ethanol, the combined selectivity
to ethyl ether products was 93% at 90% HMF conversion. By
increasing the loading of HMF to 9.6 wt%, the selectivity decreased
slightly to 83%.
Fig. 7. Effects of solvent environment on rate of furanyl alcohol etherification for
MFA, BHMF (first ether formation), and HMFD(R)A in methanol (d), ethanol (.), 1-
propanol (j), 2-propanol (ꢀ), and 1-butanol (N). Reaction conditions: (a) 25 °C,
10 mM MFA, 10.0 mL alcohol solvent, 0.90
10 mM BHMF, 10.0 mL ethanol as solvent, 0.90
80 °C, 100 mM HMF, 50.0 mL alcohol as solvent, 0.11 mmol acid sites on A-15.
l
mol acid sites on A-15, (b) 60 °C,
lmol mM acid sites on A-15 and (c)

78
E.R. Sacia et al. / Journal of Catalysis 313 (2014) 70–79
coefficient for each substrate as a function of the dielectric con-
stant of the solvent. The dielectric constant of each solvent was
determined at the corresponding reaction temperature using pre-
viously published values [30]. The solvent dielectric constant is
chosen as a descriptor because it is known that for SN1 reactions
involving a neutral leaving group, increasing solvent polarity
(viz., dielectric constant) decreases the rate of reaction [19]. This
trend is a consequence of the fact that the positive charge is orig-
inally highly localized on the protonated alcohol intermediate,
shown in step (2) of Scheme 2. The transition state for the release
of H2O involves delocalization of the charge so that it is shared be-
tween the ring and the leaving group. A more polar solvent is bet-
ter able to stabilize the reactants with respect to the transition
state; therefore, the rate decreases with increasing solvent polar-
ity. This effect is largely manifested in the intrinsic rate constant
(k2 of Eq. (2)). Since the furanyl alcohol will also more readily par-
tition into a less polar solvation shell containing butanol than one
containing methanol, the effect of solvent polarity on reaction rate
is further enhanced.
withdrawing substituents. Progressing from BHMF to HMF, the
ring substituent at the 2 position becomes more electron with-
drawing. Since the aldehyde functionality of HMF is electron with-
drawing due to resonance and inductive effects, the acetal is
formed prior to etherification in alcohol solutions with A-15. The
acetal, however, still causes the observed activation barrier for
the etherification of HMF to be higher than that for etherification
of BHMF or MFA. The mechanism of acetal formation proposed
for HMF explains why this compound forms ethers at mild condi-
tions in alcoholic solvents.
Acknowledgments
This work was funded by the Energy Biosciences Institute.This
material is also based upon work supported by the National Sci-
ence Foundation Graduate Research Fellowship under Grant No.
DGE 1106400. The authors would also like to acknowledge the con-
tributions of Taeryong Kim and Matthew Deaner for the experi-
mental portion of this manuscript.
For MFA and BHMF etherification, the effect of changes in sol-
vent polarity on rate is very similar. For HMF, acetalization pre-
cedes etherification, and the variation of the rate coefficient with
dielectric constant is no longer monotonic. During the etherifica-
tion of HMF in methanol, the active species undergoing etherifica-
tion is the dimethyl acetal of HMF. Likewise, for etherification of
HMF in butanol, it is the dibutyl acetal of HMF that undergoes
etherification. Therefore, the observed dependence on dielectric
constant can be attributed to competing effects between solvent
polarity, which increases the intrinsic rate constant for the larger
alcohols, and the ‘‘bulkiness’’ of the non-polar acetal group, which
affects the partitioning of HMFD(R)A to the acid site’s solvation
shell. The highly non-polar acetal that is formed for the larger alco-
hols, like butanol, prevents partitioning and proper orientation into
the solvation shell of the A-15 resin. Since the methoxy groups of
the acetal in the methanol solvent are less bulky and more polar
than their butoxy counterparts in butanol, the intermediate can
better reach the acid site. Therefore, the competition between sol-
vent polarity and the size of the acetal intermediate causes the
trend observed in Fig. 7c.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2014.02.012.
References
[1] Renewable Energy Road Map, Renewable Energies in the 21st Century:
Building A More Sustainable Future, COM(2006) 848 Final, Brussels, 2007.
[2] J.R. Regalbuto, Cellulosic biofuels—got gasoline?, Science 325 (2009) 822–824
[3] C. Somerville, H. Youngs, C. Taylor, S.C. Davis, S.P. Long, Feedstocks for
lignocellulosic biofuels, Science 329 (2010) 790–792.
[4] F.S. Namat Abu Al-Soof, Brahim Aklil, Mohammad Taeb, Mohammad Khesali,
Mohammad Mazraati, Benny Lubiantara, Taher Najah, Amal Alawami, Claude
Clemenz, Nadir Guerer, Garry Brennand, Jan Ban, Joerg Spitzy, Douglas Linton,
James Griffin, Martin Tallett, Petr Steiner, Ula Szalkowska, in: J. Griffin (Ed.),
World Oil Outlook 2010, Organization of the Petroleum Exporting Countries,
Vienna, Austria, 2010.
[5] M. Balakrishnan, E.R. Sacia, A.T. Bell, Etherification and reductive etherification
of
5-(hydroxymethyl)furfural:
5-(alkoxymethyl)furfurals
and
2,5-
bis(alkoxymethyl)furans as potential bio-diesel candidates, Green Chem. 14
(2012) 1626–1634.
[6] L. Bing, Z. Zhang, K. Deng, Efficient one-pot synthesis of 5-
(ethoxymethyl)furfural from fructose catalyzed by a novel solid catalyst, Ind.
Eng. Chem. Res. 51 (2012) 15331–15336.
[7] L. Hu, G. Zhao, W. Hao, X. Tang, Y. Sun, L. Lin, S. Liu, Catalytic conversion of
biomass-derived carbohydrates into fuels and chemicals via furanic aldehydes,
Roy. Soc. Chem. Adv. 2 (2012) 11184–11206.
4. Conclusions
Amberlyst-15 has been identified to be a highly active and
selective catalyst for the etherification of MFA, BHMF, and HMF
by C1–C4 alcohols. High selectivity to crossed ethers is achieved
through the formation of a solvation shell of polar C1–C4 alcohols
formed around the active site of the catalyst, as shown in Fig. 3.
The proposed mechanism envisions that the oxonium ion formed
in the rate-limiting step quickly reacts with a nearby nucleophile,
such as ethanol. Due to the high relative abundance of the polar
alcohol solvent in the solvation shell of the acid, crossed ethers
are formed with high selectivity rather than undesired oligomeri-
zation of furanyl alcohols. The polarity of the alcohol around the
acid site can also have a significant effect on the rate of etherifica-
tion. Decreasing the solvent polarity increases the rate of product
formation by nearly an order of magnitude, as was noted when
methanol was replaced by 1-butanol as the solvent for MFA and
BHMF etherification.
Systematic investigation of the etherification of C6 furanyl alco-
hols demonstrates that the degree to which the substituents at the
2/5 position of the ring withdraw electron density from the ring
strongly affects the rate of etherification. Since the methyl group
of MFA donates electron density to the furan ring, the reactive oxo-
nium intermediate is formed more readily and, correspondingly,
the activation energy is lower than that observed for electron-
[8] B. Liu, Z. Zhang, One-pot conversion of carbohydrates into 5-
ethoxymethylfurfural and ethyl D-glucopyranoside in ethanol catalyzed by a
silica supported sulfonic acid catalyst, Roy. Soc. Chem. Adv. 3 (2013) 12313–
12319.
[9] E.-J. Ras, S. Maisuls, P. Haesakkers, G.-J. Gruter, G. Rothenberg, Selective
hydrogenation
heterogeneous catalysts, Adv. Synth. Catal. 351 (2009) 3175–3185.
[10] P. Imhof, A.S. Dias, E. de Jong, G.-J. Gruter, OA02 Furanics: Versatile
of
5-ethoxymethylfurfural
over
alumina-supported
–
Molecules for Biofuels and Bulk Chemicals Applications, NAM Abstract, 2009.
[11] M.J. Murphy, J.E. Taylor, R.L. McCormick, in: NREL (Ed.), Compendium of
Experimental Cetane Number Data, Golden, Colo., 2004.
[12] M. Mascal, E.B. Nikitin, Towards the efficient, total glycan utilization of
biomass, ChemSusChem 2 (2009) 423–426.
[13] M. Mascal, Edward B. Nikitin, Direct, high-yield conversion of cellulose into
biofuel, Angew. Chem. Int. Ed. 47 (2008) 7924–7926.
[14] M. Chidambaram, A.T. Bell, A two-step approach for the catalytic conversion of
glucose to 2,5-dimethylfuran in ionic liquids, Green Chem. 12 (2010) 1253–
1262.
[15] J.T. Scanlon, D.E. Willis, Calculation of flame ionization detector relative
response factors using the effective carbon number concept, J. Chromatogr. Sci.
23 (1985) 333–340.
[16] A.D. Jorgensen, K.C. Picel, V.C. Stamoudis, Prediction of gas chromatography
flame ionization detector response factors from molecular structures, Anal.
Chem. 62 (1990) 683–689.
[17] T. Holm, Aspects of the mechanism of the flame ionization detector, J.
Chromatogr. A 842 (1999) 221–227.
[18] MestReNova, Mestrelab Research, Santiago de Compostela, Spain, 2012.
[19] E.V. Anslyn, D.A. Dougherty, Modern Physical Organic Chemistry, University
Science Books, Sausalito, California, 2006.

E.R. Sacia et al. / Journal of Catalysis 313 (2014) 70–79
79
[20] R.Q. Snurr, A.T. Bell, D.N. Theodorou, Investigation of the dynamics of benzene
in silicalite using transition-state theory, J. Phys. Chem. 98 (1994) 11948–
11961.
[21] M.W. Anderson, J. Klinowski, Direct observation of shape selectivity in zeolite
ZSM-5 by magic-angle-spinning NMR, Nature 339 (1989) 200–203.
[22] Rohm&Haas, Amberlyst 15DRY: Industrial Grade Strongly Acidic Catalyst,
Technical Data Sheet, Dow Chemical Company, 2005.
stabilizing solvents and its consequence for catalysis, J. Am. Chem. Soc. 119
(1997) 11826–11831.
[27] H.M. Mcconnell, Reaction rates by nuclear magnetic resonance, J. Chem. Phys.
28 (1958) 430–431.
[28] T.J. Swift, R.E. Connick, Nmr-relaxation mechanisms of 017 in aqueous
solutions of paramagnetic cations and lifetime of water molecules in first
coordination sphere, J. Chem. Phys. 37 (1962) 307.
[23] C. Buttersack, Accessibility and catalytic activity of sulfonic acid ion-exchange
resins in different solvents, React. Polym. 10 (1989) 143–164.
[24] B.E. Poling, J.M. Prausnitz, J.P. O’Connell, The Properties of Gases and Liquids,
McGraw-Hill, 2001.
[29] M.R. Altıokka, H.L. Hosßgün, Kinetics of hydrolysis of benzaldehyde dimethyl
acetal over Amberlite IR-120, Ind. Eng. Chem. Res. 46 (2007) 1058–1062.
[30] C. Wohlfahrt, Static dielectric constants of pure liquids and binary liquid
mixtures, in: O. Madelung (Ed.), Landolt–Börnstein
– Group IV Physical
[25] A. Chakrabarti, M.M. Sharma, Cationic ion exchange resins as catalyst, React.
Polym. 20 (1993) 1–45.
Chemistry: Numerical Data and Functional Relationships in Science and
Technology, Springer-Verlag, Berlin, 1991.
[26] D. Fa˘rcaßsiu, A. Ghenciu, G. Marino, K.D. Rose, Strength of solid acids and acids
[31] Y. Liu, E. Lotero, J.G. Goodwin Jr, A comparison of the esterification of acetic
acid with methanol using heterogeneous versus homogeneous acid catalysis, J.
Catal. 242 (2006) 278–286.
in solution. Enhancement of acidity of centers on solid surfaces by anion