Effect of Alcohol Structure on the Kinetics of Etherification and Dehydration over Tungstated Zirconia

Article abstract of DOI:10.1002/cssc.201801067

Linear and branched ether molecules have attracted recent interest as diesel additives and lubricants that can be produced from biomass-derived alcohols. In this study, tungstated zirconia was identified as a selective and green solid acid catalyst for the direct etherification of primary alcohols in the liquid phase, achieving ether selectivities of >94 % for C₆–C₁₂ linear alcohol coupling at 393 K. The length of linear primary alcohols (C₆–C₁₂) was shown to have a negligible effect on apparent activation energies for etherification and dehydration, demonstrating the possibility to produce both symmetrical and asymmetrical linear ethers. Reactions over a series of C₆ alcohols with varying methyl branch positions indicated that substituted alcohols (2°, 3°) and alcohols with branches on the β-carbon readily undergo dehydration, but alcohols with branches at least three carbons away from the -OH group are highly selective to ether. A novel model compound, 4-hexyl-1dodecanol, was synthesized and tested to further demonstrate this structure–activity relationship. Trends in the effects of alcohol structure on selectivity were consistent with previously proposed mechanisms for etherification and dehydration, and help to define possible pathways to selectively form ethers from biomass-derived alcohols.

Full text of DOI:10.1002/cssc.201801067

Accepted Article
Title: Effect of Alcohol Structure on the Kinetics of Etherification and
Dehydration over Tungstated Zirconia
Authors: Julie Rorrer, Suresh Pindi, Dean Toste, and Alexis T. Bell
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemSusChem 10.1002/cssc.201801067
Link to VoR: http://dx.doi.org/10.1002/cssc.201801067
A Journal of
www.chemsuschem.org

10.1002/cssc.201801067
ChemSusChem
Effect of Alcohol Structure on the Kinetics of Etherification and Dehydration over Tungstated Zirconia
Julie Rorrer,1,2 Suresh Pindi,1,3 F. Dean Toste,1,3 Alexis T. Bell*1,2
1Energy Biosciences Institute
University of California
Berkeley, CA 94720
2Department of Chemical and Biomolecular Engineering
University of California
Berkeley, CA 94720
3Department of Chemistry
University of California
Berkeley, CA 94720
Submitted to: ChemSusChem
To whom correspondence should be addressed: alexbell@berkeley.edu
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
Abstract
Linear and branched ether molecules have attracted recent interest as diesel additives and lubricants
that can be produced from biomass-derived alcohols. In this study, tungstated zirconia was identified as a
selective and green solid acid catalyst for the direct etherification of primary alcohols in the liquid phase,
achieving ether selectivities of >94% for C6-C12 linear alcohol coupling at 393 K. The length of linear
primary alcohols (C6-C12) was shown to have a negligible effect on apparent activation energies for
etherification and dehydration, demonstrating the possibility to produce both symmetrical and asymmetrical
linear ethers. Reactions over a series of C6 alcohols with varying methyl branch positions indicated that
substituted alcohols (2°, 3°) and alcohols with branches on the β-carbon readily undergo dehydration, but
alcohols with branches at least 3 carbons away from the –OH group are highly selective to ether. A novel
model compound, 4-hexyl-1-dodecanol, was synthesized and tested in order to further demonstrate this
structure-activity relationship. Trends in the effects of alcohol structure on selectivity were consistent with
previously proposed mechanisms for etherification and dehydration, and help to define possible pathways
to selectively form ethers from biomass-derived alcohols.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
1. Introduction
As the adverse effects of anthropogenic climate change continue to increase, the global need for
clean and renewable energy is now larger than ever before. One method of lowering net atmospheric CO2
emissions is to replace conventional fossil fuels and petroleum-derived specialty chemicals with renewable
alternatives.[1] Ethers have attracted recent interest as renewable fuels and lubricants. Symmetric and
asymmetric linear ethers in the C8-C12 range are suitable diesel additives due to their high cetane numbers.[2–
5] Branched ethers such as methyl tert-butyl ether can also be added to gasoline because of their high octane
numbers.[6] Longer chain branched ethers in the C28-C32 range have low pour points and excellent lubricant
properties,[7–9] and could replace poly-alpha-olefin lubricants, which are produced from fossil reserves.[10,11]
Biomass-derived alcohols are promising candidates for producing renewable ethers. Ethanol and
butanol can be obtained in high yields through fermentation processes such as ABE fermentation.[12–14]
Longer chain linear alcohols like 1-octanol and 1-dodecanol can be produced from the condensation of
biomass-derived furan platform molecules, and the hydrolysis of triglycerides and fatty acids,
respectively,[15] and then built up into branched alcohols with higher carbon numbers through the Guerbet
reaction.[16] Some general approaches to producing linear and branched alcohols from biomass-derived
platform molecules are summarized in Scheme 1.
Direct etherification of biomass-derived alcohols over an acid catalyst is an attractive method of
producing ethers because it does not require hydrogen gas and precious metal catalysts like reductive
etherification.[17] The competing reaction for alcohol dehydration over an acid catalyst is unimolecular
dehydration to form alkenes, which is thermodynamically favored at temperatures above approximately
350 K. Due to their high volatility, and the propensity to form gums, alkenes are not desired in fuel and
lubricant blends.[18]
In a previous study, tungstated zirconia was identified as a selective catalyst for the liquid phase
etherification of 1-dodecanol, with ether selectivities above 94% at moderate temperatures (393 K).[19] This
study suggested that the high selectivity to ether is due to a cooperative effect between Brønsted- and Lewis-
acid sites on the surface of the catalyst which promote the bi-molecular etherification reaction. This research
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
motivates the current study, which aims to evaluate the extent to which tungstated zirconia can be used to
produce a variety of symmetrical and asymmetrical ethers, and mixtures of ethers, for fuel and lubricant
applications.
Kinetic studies of alcohol dehydration reactions have traditionally been centered on gas phase
unimolecular dehydration.[20–22] Experimental and computational studies of gas phase unimolecular
dehydration of alcohols over γ-Al2O3 have supported the theory that increasing the substitution of the
alcohol increases the stability of the intermediate in the rate limiting step for dehydration, resulting in a
decrease in activation barrier for dehydration with increasing substitution.[20,23]
While studying the effect of alcohol structure in the gas phase where unimolecular dehydration
dominates has provided some insight into the relationship between unimolecular dehydration rates and
substitution, the kinetics of etherification must also be explored in order to understand the relationship
between alcohol structure and selectivity to ether. Some efforts have been made to investigate the catalytic
conversion of alcohols to both symmetrical and asymmetrical ethers at moderate temperatures where
etherification dominates. Various acidic resins and zeolites have been shown to be effective in the liquid
phase etherification of 1-pentanol, 1-hexanol, and 1-octanol to produce symmetrical ethers.[24–27] Tejero, et.
al. have demonstrated that it is also possible to produce asymmetrical ethers like ethyl-hexyl ether and
ethyl-octyl ether through cross-coupling of 1-hexanol or 1-octanol with ethanol or diethylcarbonate.[5,28]
But, to our knowledge, the literature is lacking a comprehensive study of the kinetics of liquid
phase etherification reactions over a large range of long-chain 1-alcohols (C6-C12), and a more thorough
investigation into the effect of branched alcohols on the formation of symmetrical and asymmetrical ethers
in the lubricant range (C12-C36). While acidic resins such as Amberlyst 70 and Nafion NR-50 have been
shown to be effective for etherification of linear alcohols in the liquid phase, measuring accurate rate data
is difficult due to the swelling of these resins in the solvent, which causes changes in the number of
accessible active sites over time.[29] Tungstated zirconia is thermally stable, does not deactivate or swell in
the reactant media, and has been shown to be an effective catalyst for etherification of alcohols. Thus, this
paper will comprehensively study the effects of linear alcohol length, carbon chain branches, and positions
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
of carbon chain branches on the kinetics of etherification and dehydration reactions over tungstated zirconia
in the liquid phase, with the aim of understanding the relationship between ether selectivity and alcohol
structure.
Scheme 1: General approach to producing linear and branched alcohols from biomass-derived platform
molecules.
2. Results and Discussion
2.1 Catalyst Characterization
Tungstated zirconia (WOx/ZrO2, 12.6 wt% W) was characterized with X-ray diffraction, Raman
spectroscopy, and DRIFTS FTIR with adsorbed pyridine to determine the structure and confirm the
presence of Brønsted- and Lewis-acid sites. From the XRD pattern in Figure 1a, the peaks at 2θ angles of
30, 35, 50, and 59o (yellow) indicate the presence of tetragonal zirconia,[30–32] and the peaks between 23-
25o (pink) represent the presence of bulk tungsten oxide.[33] From the Raman spectra in Figure 2b, the bands
at bands at 274, 715, and 807 cm-1 (pink) indicate the W-O stretching of bulk tungsten oxide,[34] and the
band at 1020 cm-1 (yellow) is indicative of polymeric tungsten oxide terminated by a tungsten oxygen
double bond.[35] The DRIFTS FTIR spectra of adsorbed pyridine shown in Figure 1c demonstrates that
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
both Brønsted- and Lewis- acid sites are present on the surface of the catalyst, as evidenced by bands of
adsorbed pyridine at 1540 cm-1 (blue), and 1444 cm-1 (green), respectively. Additional catalyst
characterization is summarized in Table 1.
Figure 1: Characterization of 12.6 wt% W (WOx/ZrO2), a) X-ray diffraction spectrum, b) Raman spectrum,
c) DRIFTS FTIR spectrum of adsorbed pyridine at 393 K.
Table 1: Characterization of 12.6 wt% WOX/ZrO2
Property
Value
Weight loading of W (%)
12.6
Zirconia Crystal Structure (XRD)
BET Surface Area (m2 g-1)
Brønsted Acid Sites (mol eq. H+ m-2)
Ratio of Brønsted / Lewis Acid Sites
Tetragonal
51
1.78*10-6 ± 1.4*10-7
0.90 ± 0.02 (393 K)
2.2 Effect of Alcohol Structure on Kinetics: Linear Primary Alcohols (C6-C12)
To build symmetrical and asymmetrical linear ethers in the diesel range, it is necessary to
investigate the effect of alcohol chain length on the apparent kinetics of etherification and dehydration over
tungstated zirconia. The selectivity to ether for C6-C12 alcohols (1-hexanol, 1-heptanol, 1-octanol, 1-
nonanol, 1-decanol, 1-undecanol, and 1-dodecanol) was measured for temperatures between 393 - 418 K.
The results are summarized in Table 2.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
Table 2: Selectivity of primary linear alcohols (C6-C12) to symmetrical ethers (C12
-
C24) over tungstated zirconia with varying temperature.
Selectivity to Symmetrical Ether (%) / (Conversion of Alcohol (%))
Reactant
393 K
398 K
94 /(6)
80 /(7)
92 /(6)
89 /(7)
95 /(8)
92 /(6)
94 /(7)
403 K
408 K
413 K
418 K
1-Hexanol
1-Heptanol
1-Octanol
1-Nonanol
1-Decanol
1-Undecanol
1-Dodecanol
>99 /(4)
>99 /(5)
>99 /(4)
>99 /(4)
>99 /(4)
>99 /(5)
94 /(6)
88 /(11)
80 /(11)
92 /(12)
85 /(12)
94 /(11)
89 /(9)
88 /(22)
82 /(26)
83 /(20)
86 /(20)
90 /(18)
88 /(17)
83 /(26)
89 /(27)
80 /(25)
85 /(28)
81 /(28)
88 /(25)
82 /(25)
80 /(30)
87 /(39)
80 /(38)
80 /(33)
77 /(37)
85 /(38)
78 /(42)
84 /(38)
86 /(11)
Reaction Conditions: 1 h, 100 mg WOx/ZrO2 (12.6 wt% W), 600 RPM, 564 µmol
reactant, 125 µmol n-tetradecane. Remaining selectivity is attributed to alkenes, and
trace amounts of branched ether. Error within ±2% for data at 393 – 413 K, and within
±4% for 418 K.
For all of the linear alcohols studied, the selectivity to ether decreases with increasing temperature.
However, the carbon chain length does not appear to have any significant effect on the selectivity. Nel and
Klerk observed that ether selectivity of C6-C12 linear alcohols decreased with increasing carbon chain length
for gas phase reactions over ɳ-alumina at 523 K.[2] The authors claimed that their system was limited by
external mass transfer, which could explain why they observed a decrease in selectivity for longer chain
alcohols.
The apparent activation energies for etherification and dehydration of linear alcohols in the liquid
phase were obtained by measuring the initial rates of ether and alkene formation over a range of
temperatures. A sample time-course study of 1-dodecanol etherification over WOx/ZrO2 (12.6 wt% W) at
393 K is shown in Figure S1 in the Supporting Information, which demonstrates that at conversions below
approximately 25%, the kinetics are pseudo-zero order. The rates, selectivities, Arrhenius plots, and
temperature ranges for which the initial rates were obtained are provided in Figure S2, and Tables S1-S2 in
the Supporting Information.
Figure 2 shows the apparent activation energies for etherification and dehydration of primary linear
alcohols 1-hexanol, 1-heptanol, 1-octanol, 1-nonanol, 1-decanol, 1-undecanol, and 1-dodecanol. Overall,
the activation barriers for unimolecular dehydration in Figure 2b (154-157 kJ mol-1) are greater than the
activation barriers for etherification in Figure 2a (126-127 kJ mol-1), which explains the decrease in
selectivity to ether with increasing temperature observed in Table 2. The activation energies for linear
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
alcohol etherification do not change significantly with increasing carbon chain length, implying that the
length of linear alcohols has little effect on the kinetics of etherification and dehydration.
Figure 2: a) Apparent activation energies for etherification, b) apparent activation energies for dehydration
of C6-C12 primary alcohols over WOx/ZrO2 (12.6 wt% W).
For applications in diesel fuel, mixtures of linear alcohols could be tuned in the reactant feed to
create a distribution of ethers that contains a desired molecular weight distribution. A similar tuning of
molecular weights has been demonstrated previously for the trimerization of ketones for diesel fuel
mixtures.[36] Because the etherification activation energies and reaction rates do not vary significantly
between linear alcohols in the C6-C12 range, a mixture of alcohols will couple in a nearly random statistical
distribution. An example distribution study is shown in Figure 3. Seven linear alcohols (1-hexanol, 1-
heptanol, 1-octanol, 1-nonanol, 1-decanol, 1-undecanol, and 1-dodecanol) in equimolar ratios were reacted
together over tungstated zirconia at 393 K. The theoretical selectivity distribution for random cross-
coupling of alcohols is shown in black, and the experimental distribution of ether products is shown in
green. Figure 3 shows that there is a slight preference for coupling of the shorter chain alcohols, which
could be explained by the fact that the experiment was run at high conversion (>90%), where the inhibiting
effects of water and ether are no longer negligible. Under these conditions, it is possible that the bulkier
ethers such as di-dodecyl ether introduce steric effects and thus have a greater inhibiting effect on the rate
of consumption of the longer chain alcohols. The observation that cross-coupling of linear alcohols is nearly
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
statistical suggests that the distribution of linear alcohols in the feed could be tuned to produce a variety of
desired ether blends.
Figure 3. Ether Distribution for C6-C12 alcohols. Reaction conditions: 393 K, 48 h, 100 mg WOx/ZrO2 (12.6
wt% W), 600 RPM. Equimolar ratio of 1-hexanol, 1-heptanol, 1-octanol, 1-nonanol, 1-decanol, 1-
undecanol, and 1-dodecanol, with n-tetradecane as an internal standard. Conversion >90%.
2.3 Effect of Alcohol Structure on Kinetics: Position of Carbon Chain Branches
To systematically study the effect of the position of carbon chain branches on etherification
kinetics, a series of C6 alcohols with varying degrees of substitution and positions of methyl groups were
reacted over tungstated zirconia. Table 3 shows the selectivity to ether for the C6 alcohols at 393 K. Primary
C6 alcohols with a methyl group at least 3 carbons away from the –OH group (1-hexanol, 4-methyl-1-
pentanol, and 3-methyl-1-pentanol) exhibit a very high selectivity to ether. As the methyl group approaches
the β-carbon (2-methyl-1-pentanol), the ether selectivity rapidly decreases. Secondary and tertiary C6
alcohols (2-hexanol and 2-methyl-2-pentanol) rapidly dehydrated under these conditions, forming no ether.
Table 3: Selectivity of reactions over C6 alcohols.
Reactant
Conversion Selectivity Selectivity to
to Ether
Alkenes/Other
1-Hexanol
11%
>99%
<1%
4-Methyl-1-Pentanol 10%
3-Methyl-1-Pentanol 10%
2-Methyl-1-Pentanol 14%
>99%
<1%
87%
<1%
<1%
<1%
13%
>99%
>99%
>99%
2-Hexanol
>99%
2-Methyl-2-Pentanol >99%
Reaction Conditions: 393 K, 4 h, 100 mg WOx/ZrO2 (12.6 wt% W),
600 RPM, 168 µL reactant, 100 µL n-tetradecane as internal
standard.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
Figure 4: a) Apparent activation energies for etherification, b) apparent activation energies for dehydration
of C6 alcohols (1-hexanol, 4-methyl-1-pentanol, 3-methyl-1-pentanol, 2-methyl-1-pentanol, 2-hexanol, and
2-methyl-2-pentanol) over WOx/ZrO2 (12.6 wt% W).
To further study the relationship between branching position and kinetics, the apparent activation
energies of etherification and dehydration of the C6 alcohols were measured, and are shown in Figure 4.
Temperature ranges were selected such that initial rates of both ether and alkene could be measured, with
the exception of 2-methyl-2-pentanol, for which ether could not be detected at any of the temperatures in
the range of 333-418 K. The initial rates, Arrhenius plots, and temperature ranges are provided in Figure
S3, and Tables S3 and S4 in the Supporting Information.
Figure 4a shows that the apparent activation barrier for etherification is fairly independent of the
reactant. But from Figure 4b, it is clear that as the substitution of the alcohol increases, the barrier for
unimolecular dehydration decreases. The dehydration activation barriers for 1-hexanol, 4-methyl-1-
pentanol, and 3-methyl-1-pentanol are 157 ± 13, 158 ± 10, and 155 ± 1 kJ mol-1, respectively, indicating
that the barriers are within error for alcohols with branching at least 3 carbons away from the –OH group.
But for 2-methyl-1-pentanol, 2-hexanol, and 2-methyl-2-pentanol, the barriers are 124 ± 2, 114 ± 5, and
100 ± 12 kJ mol-1, respectively, indicating that the dehydration barrier decreases with increasing substitution
of the alcohol. In our previous study, measurements of kinetic isotope effects for 1-hexanol dehydration
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
over tungstated zirconia suggested that the rate limiting step for unimolecular dehydration is the cleavage
of a β-carbon hydrogen bond.[19] This is consistent with the findings in Figure 4, since the increased
substitution of the α- and β-carbons would increase the stability of the carbocation intermediate formed
during unimolecular dehydration, making dehydration more favorable. Also found in our previous study,
the dehydration of 1-hexanol produced both 1-hexene and 2-hexene, as verified by NMR; however, no
methyl shift was observed, as might be expected for the dehydration of a substrate such as 3,3-di-methyl-
2-butanol.[19,37]
In addition, steric effects that limit α-carbon oxygen bond formation between two alcohols could
explain the preference towards unimolecular dehydration for 2° and 3° alcohols.[23] Similar relationships
between alcohol structure and dehydration activation barriers have been observed in the literature for other
solid acid catalysts. Activation energies of 145, 141, 121, and 110 kJ mol-1 have been reported for gas phase
unimolecular dehydration of ethanol, 1-propanol, 2-propanol, and 2-methyl-2-propanol, respectively, over
γ-Al2O3, indicating that increasing substitution decreases the activation barrier for unimolecular
dehydration.[20,23] Mpourmpakis, et. al. observed that dehydration activation barriers of ethanol, 1-propanol,
2-propanol, and 2-methyl-2-propanol also decreased with increasing substitution for dehydration reactions
over γ-Al2O3, ZrO2, and TiO2.[21] The relationship between alcohol substitution and reactivity for
unimolecular dehydration is in agreement with the results presented here for C6 alcohols in the liquid phase,
and the relative activation energies for etherification provide an explanation for the selectivity trends
observed in Table 3.
2.4 Effect of Alcohol Structure: Length of Carbon Chain Branches
While the kinetics of linear alcohols do not vary significantly from C6-C12, the addition of branches
changes the selectivity to ether. To further investigate the effects of branch size on ether selectivity, a series
of Guerbet alcohols with increasing carbon chain backbone length and branch length were reacted with 1-
dodecanol in 1:1 molar ratios at low conversions to form symmetrical and asymmetrical ethers in the
lubricant range (C16-C36). The pathways for these reactions are shown in Scheme 2, and the results are
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
summarized in Table 3. Symmetrical branched ethers (1E) are only formed in small amounts from the
coupling of small Guerbet alcohols (2-ethyl-1-hexanol and 2-butyl-1-octanol), and are not formed from the
coupling of bulkier alcohols (>C16). In addition, the reactivity of Guerbet alcohols decreases with increasing
bulkiness.
Scheme 2. Pathways for 1-dodecanol etherification and dehydration with varying Guerbet alcohols (C8-
C24) over tungstated zirconia.
Table 3: Effect of branch length of Guerbet alcohols on the selectivity of ethers and alkenes in reaction with 1-dodecanol.
Guerbet Alcohol
R1 R2 Conversion of 1A (%) Dodecanol Selectivity (%)
Conversion of 1B (%) Guerbet Alcohol Selectivity (%)
1C
63
1D
20
1F
16
1D
11
1E
3
1G
86
2-Ethyl-1-Hexanol
4
2
14
33
2-Butyl-1-Octanol
6
8
10
4
6
8
18
21
23
69
71
72
72
16
10
7
15
20
21
22
26
13
9
11
18
20
17
1
0
0
0
89
82
80
83
2-Hexyl-1-Decanol
2-Octyl-1-Dodecanol
2-Decyl-1-Tetradecanol
12 10 23
6
3
Reaction Conditions: 393 K, 4 h, 100 mg WOx/ZrO2 (12.6 wt% W), 600 RPM, 5.6*10-4 mol 1-dodecanol, 5.6*10-4 mol Guerbet alcohol, 3.9*10-4 mol n-
tetradecane as internal standard.
The relatively low selectivity towards branched ethers from the etherification of Guerbet alcohols
is likely due to the relatively low activation barrier for unimolecular dehydration of alcohols with
substitution on the β-carbon. From Table 3, the addition of methyl branches affected the selectivity to ether
to a lesser extent when the branch was 3 carbons away from the –OH group. Thus it was hypothesized that
etherification of a longer chain branched alcohol with the branch further down the carbon chain would
result in higher selectivity to ether than a Guerbet alcohol of the same chain length. To test this, the model
compound 4-hexyl-1-dodecanol was synthesized and reacted over tungstated zirconia at 393 K, as shown
in Scheme 3. The synthesis and characterization of 4-hexyl-1-dodecanol is detailed in the Supporting
Information.
Scheme 3: Reaction of 4-Hexyl-1-Dodecanol over tungstated zirconia.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
Table 4: Etherification and dehydration of 4-hexyl-1-dodecanol.
Alcohol
R1 R2 Temperature Reaction Time Conversion of Selectivity (%)
[K]
393
[h]
4
Alcohol (%)
23
Ether
40
Alkenes
60
4-Hexyl-1-Dodecanol
4-Hexyl-1-Dodecanol
4-Hexyl-1-Dodecanol
6
6
6
8
8
8
383
383
4
11
22
49
52
51
48
16
Reaction Conditions: 600 RPM, 100 mg WOx/ZrO2 (12.6 wt% W), 600 RPM, 150 µL 4-hexyl-1-
dodecanol, 3.9*10-4 mol n-tetradecane as internal standard. Note: 10% mass balance missing for
16 h reaction.
From Table 4, it can be observed that there is significant improvement in selectivity to ether for the
C18 alcohol when the branching position is further away from the –OH group, reaching 40% selectivity to
ether (2B) at a conversion of 23% at 393 K. Upon lowering the temperature to 383 K, the selectivity was
further increased to 49%, which is expected as etherification was shown to be more favorable at lower
temperatures. The results given in Table 3 indicate that 2-hexyl-1-decanol (C16) and 2-octyl-1-dodecanol
(C20) do not produce symmetrical ethers in reactions with 1-dodecanol over tungstated zirconia. For
additional comparison, Guerbet alcohols 2-hexyl-1-decanol and 2-octyl-1-dodecanol were also reacted over
tungstated zirconia under the same conditions as those used for the reaction of 4-hexyl-1-dodecanol in the
absence of 1-dodecanol. No ether was detected for either of these reactions (Table S6, Supporting
Information).
Thus, placing the branch further away from the –OH group of the alcohol significantly improves
the selectivity to ether. This raises the question of how branched alcohols with branches further down the
carbon chain could be produced from biomass. One potential route would be to start with the condensation
of biomass-derived furfural and 2-methyl-furan,[38] then selectively hydrogenate to produce 5-pentyl-1-
decanol, which could further be reacted over an acid catalyst to form ether, as shown in Scheme 4.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
Scheme 4: Proposed route to produce branched lubricants from biomass-derived platform molecules.
3. Conclusions
Tungstated zirconia was identified as an effective solid acid catalyst for the direct etherification of
primary linear alcohols in the liquid phase. The effect of alcohol structure on etherification and dehydration
reactions was studied by investigating the effects of alcohol length, positions of carbon chain branches, and
size of carbon chain branches. The rate of dehydration relative to etherification was shown to increase with
increasing temperature, which is consistent with the thermodynamic preference for dehydration with
increasing temperature. The positions of carbon chain branches were shown to have significant effects on
ether selectivity. Addition of carbon chain branches to the α and β carbons on the alcohol were shown to
increase the dehydration selectivity, which is consistent with the mechanism proposed in an earlier work.
It was also demonstrated that alcohols with carbon chain branching at least 3 carbons away from the –OH
group exhibit significantly higher ether selectivity than corresponding alcohols with branches on the α- or
β-carbons. Activation energy trends for etherification and dehydration provided a more complete
explanation for the ether selectivity trends observed over a range of alcohol structures. Understanding the
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
effects of alcohol structure on direct etherification reactions is crucial for utilizing biomass-derived alcohols
to make diesel additives, lubricants, and specialty chemicals.
4. Experimental Section
4.1 Materials
All chemicals obtained commercially were used without further purification. The following
chemicals were obtained from Sigma-Aldrich: 1-hexanol (>98%), 1-octanol (>99%), 1-decanol (>99%), 1-
dodecanol (>98%), 2-hexanol (>99%), 2-methyl-1-pentanol (>99%), 3-methyl-1-pentanol (>99%), 4-
methyl-1-pentanol (>97%), 2-ethyl-1-hexanol (>99%), 2-butyl-1-octanol (>95%), 2-hexyl-1-decanol
(>97%), 2-octyl-1-dodecanol (>97%), 2-decyl-1-tetradecanol (97%), dodecane (>99%), hexane (>99%),
1-hexene (>99%), and pyridine (99.8%). The following chemicals were obtained from Spectrum Chemical:
1-heptanol (>98%), 1-octene (>99%), and n-tetradecane (>99%), which was used as an internal standard
for analytical purposes. The following chemicals were obtained from TCI: 1-nonanol (>99%), 1-undecanol
(>99%), di-n-octyl ether (>95%), di-n-decyl ether (>95%), and di-dodecyl ether (>95%), 1-decene (>95%)
and 1-dodecene (>95%). Di-n-hexyl ether (>98%) was obtained from Alpha Aesar. Synthesis and
characterization of 4-hexyl-1-dodecanol is provided in the Supporting Information, Scheme S1, Figures S4-
S7.
4.2 Synthesis and Characterization of Tungstated Zirconia
Tungstated zirconia (WOx/ZrO2, 12.6 wt% W) was synthesized and characterized using previously
reported methods.[19,30,39,40] Powder X-ray diffraction (XRD) patterns for WOx/ZrO2 (0-22 wt% W) were
taken with a Bruker D8 GADDS diffractometer equipped with a Cu−Kα source (40 kV, 40 mA). Raman
spectra were obtained with a LabRAM HR Horiba Scientific Raman spectrometer equipped with a 633 nm-
1 laser. BET surface area measurements were performed with a Micrometrics TriStar BET and pretreated
with a Micrometrics FlowPrep 060. The ratio of Brønsted- to Lewis-acid sites was determined from IR
spectra of adsorbed pyridine, using a previously reported method.[19] Spectra were acquired using a Thermo
Scientific Nicolet 6700 Fourier Transform Infrared Spectrometer (FT-IR) equipped with a Diffuse
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
Reflectance Infrared Fourier Transform Spectrometer (DRIFTS) cell. ICP Elemental analysis was
performed by Galbraith Laboratories, Inc. in order to determine tungsten weight loadings.
4.3 Batch Reactions
All reactions were carried out in sealed 12 mL Q-Tube batch reaction vessels with magnetic stirring
at 600 RPM using an IKA C-MAG HS 10 digital hot plate with temperature control accurate to within +/-
1 K. A separate batch reaction was performed for each time point for kinetic studies in order to ensure
consistency of volume and concentration of each sample. All reactions over tungstated zirconia were carried
out solvent-free with 100 mg of catalyst, 100 µL of n-tetradecane as an internal standard, and between 84-
205 µL of reactant depending on the reaction conditions (unless otherwise noted). Reaction products were
dissolved in acetone for analysis with GC/MS.
4.4 Product Analysis
The analysis of products was carried out using a Varian CP-3800 Gas Chromatograph/ Mass
Spectrometer (GC/MS). The products were quantified using the flame ionization detector (FID) and were
identified using the Varian 320 triple quadrupole mass spectrometer (MS). N-tetradecane was used as an
internal standard to ensure accurate product quantification. FID response factors for pure commercial
compounds were obtained from linear calibration curves, and the effective carbon number method was used
to predict response factors for products which are not available in high purity. This method predicts FID
response to within ±1.7%.[41] Mass balances for all reactions were achieved within ± 5% unless otherwise
noted.
4.5 Kinetic Analysis
To measure the initial rates of etherification and unimolecular dehydration of 1-hexanol, 1-
heptanol, 1-octanol, 1-nonanol, 1-decanol, 1-undecanol, 1-dodecanol, 4-methyl-1-pentanol, 3-methyl-1-
pentanol, 2-methyl-1-pentanol, 2-hexanol, and 2-methyl-2-pentanol, individual batch reactions for each
time point were carried out at each temperature. Mass transfer limitations were found to be negligible at
stirring speeds of 600 RPM for catalyst particle sizes less than 250 µm in diameter. This conclusion was
determined by measuring the initial catalytic activity as a function of stirring speed and particle size.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
For each experiment, conversion (Xa), selectivity to ether (Sether), selectivity to alkenes (Salkenes
were defined as follows:
)
ꢂꢃꢄ ꢃꢅꢆ ꢇꢈꢆꢉꢊꢈꢋ
ꢌꢍꢌꢊꢌꢆꢄ ꢂꢃꢄ ꢃꢅ ꢆ
ꢀꢁ =
ꢂꢃꢄꢈꢒ ꢃꢅ ꢆ ꢇꢈꢆꢉꢊꢈꢋ ꢊꢃ ꢅꢃꢇꢂ ꢈꢊℎꢈꢇ
ꢎꢏꢐℎꢏꢑ
=
=
ꢂꢃꢄꢈꢒ ꢃꢅ ꢆ ꢇꢈꢆꢉꢊꢈꢋ
ꢂꢃꢄꢈꢒ ꢃꢅ ꢆ ꢇꢈꢆꢉꢊꢈꢋ ꢊꢃ ꢅꢃꢇꢂ ꢆꢄꢗꢈꢍꢈ
ꢂꢃꢄꢈꢒ ꢃꢅ ꢆ ꢇꢈꢆꢉꢊꢈꢋ
ꢎꢁꢓꢔꢏꢕꢏꢖ
Initial rates of ether and alkene formation (ro) were calculated from the experimentally measured
curve of moles of product formed per mass of catalyst versus time (t).
Apparent activation energies were calculated from an Arrhenius plot of the natural log of the initial
rates versus the inverse temperature. At low conversions (< 25%), the reaction is pseudo-zero order in
reactant, indicating that the surface of the catalyst is saturated with reactant. Under these conditions, the
inhibiting effects of water and ether, and the formation of additional products such as branched ether and
oligomerized alkenes are negligible.[19]
The Arrhenius relation was given by:
ꢝꢁ,ꢁꢙꢙ
ꢇ = ꢗꢁꢙꢙ = ꢚꢁꢙꢙꢈꢛꢜ �
ꢘ
�
ꢞꢟ
Where ro is the initial rate of ether or alkene formation, kapp is the apparent rate constant for the reaction,
Aapp is the apparent pre-exponential factor, and Ea,app is the apparent activation energy. For the Arrhenius
fitting, the pre-exponential factor was fixed for both the etherification and dehydration plots of the C6
alcohols and linear alcohols, based upon the assumptions that the active sites for each reaction are
equivalent, and that the entropy of adsorption of the most abundant surface intermediates and the change in
entropy of the transition states are both consistent across the various substrates. The pre-exponential factors
were fit to minimize the standard error on the least squares regression for each group of substrates. Error
bars for activation energies were taken from the standard error of the least squares regression. Arrhenius
plots, initial rates, pre-exponential factors, activation energies, and the temperature ranges for each reaction
are given in Figures S2-S3 and Tables S1-S4 in the Supporting Information.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
Acknowledgements
This work was funded by the Energy Biosciences Institute. J. Rorrer would also like to acknowledge
funding from the National Science Foundation Graduate Research Fellowship (Grant Number DGE
1106400). The authors thank Adam Grippo, Deepak Jadhav, Amit Gokhale, Christopher Ho, and
Konstantinos Goulas for useful discussion.
References:
[1]
C. Sievers, Y. Noda, L. Qi, E. M. Albuquerque, R. M. Rioux, S. L. Scott, ACS Catal. 2016, 8286–
8307.
[2]
[3]
[4]
[5]
[6]
[7]
R. J. J. Nel, A. De Klerk, Ind. Eng. Chem. Res. 2009, 48, 5230–5238.
A. Arteconi, A. Mazzarini, G. Di Nicola, Water. Air. Soil Pollut. 2011, 221, 405–423.
A. L. Maximov, A. I. Nekhaev, D. N. Ramazanov, Pet. Chem. 2015, 55, 1–21.
J. Guilera, E. Ramírez, C. Fité, J. Tejero, F. Cunill, Catal. Sci. Technol. 2015, 2238–2250.
M. J. Pilling, Low-Temperature Combustion and Autoignition, Elsevier, 1997.
D. A. L. Leslie R. Rudnick, Ross A. Kremer, Alkylated Diphenyl Ether Lubricants, 1996, Patent
US 5552071 A.
[8]
[9]
W. Beilfuss, R. Gradtke, W. Siegert, K.-H. Diehl, K. Weber, Glyceryl Ethers as Preservatives for
Cooling Lubricants, 2007, US 7268102 B2.
D. Jadhav, A. M. Grippo, S. Shylesh, A. A. Gokhale, J. Redshaw, A. T. Bell, ChemSusChem 2017,
10, 2527–2533.
[10] T. Zolper, Z. Li, C. Chen, M. Jungk, T. Marks, Y.-W. Chung, Q. Wang, Tribol. Lett. 2012, 48,
355–365.
[11] R. Benda, J. Bullen, A. Plomer, J. Synth. Lubr. 1996, 13, 41–57.
[12] N. Qureshi, H. P. Blaschek, J. Ind. Microbiol. Biotechnol. 2001, 27, 287–291.
[13] T. J. Schwartz, B. J. O’Neill, B. H. Shanks, J. A. Dumesic, ACS Catal. 2014, 4, 2060–2069.
[14] S. Sreekumar, Z. C. Baer, E. Gross, S. Padmanaban, K. Goulas, G. Gunbas, S. Alayoglu, H. W.
Blanch, D. S. Clark, F. D. Toste, ChemSusChem 2014, 94720, 1–5.
[15] P. Anbarasan, Z. C. Baer, S. Sreekumar, E. Gross, J. B. Binder, H. W. Blanch, D. S. Clark, F. D.
Toste, Nature 2012, 491, 235–239.
[16] C. Carlini, A. Macinai, A. M. Raspolli Galletti, G. Sbrana, J. Mol. Catal. A Chem. 2004, 212, 65–
70.
[17] M. Balakrishnan, E. R. Sacia, A. T. Bell, Green Chem. 2012, 14, 1626–1634.
[18] R. C. C. Pereira, V. M. D. Pasa, Fuel 2006, 85, 1860–1865.
[19] J. Rorrer, Y. He, F. D. Toste, A. T. Bell, J. Catal. 2017, 354, 13–23.
[20] S. Roy, G. Mpourmpakis, D. Hong, D. G. Vlachos, A. Bhan, R. J. Gorte, ACS Catal. 2012, 2,
1846–1853.
[21] P. Kostestkyy, J. Yu, R. J. Gorte, G. Mpourmpakis, Catal. Sci. Technol. 2014, 4, 3861.
[22] G. Larsen, E. Lotero, L. M. Petkovic, D. S. Shobe, J. Catal. 1997, 169, 67–75.
[23] M. Kang, J. F. DeWilde, A. Bhan, ACS Catal. 2015, 5, 602–612.
[24] R. Bringué, E. Ramírez, M. Iborra, J. Tejero, F. Cunill, Chem. Eng. J. 2014, 246, 71–78.
[25] C. Casas, R. Bringué, E. Ramírez, M. Iborra, J. Tejero, Appl. Catal. A Gen. 2011, 396, 129–139.
[26] E. Medina, R. Bringué, J. Tejero, M. Iborra, C. Fité, Appl. Catal. A Gen. 2010, 374, 41–47.
[27] I. Hoek, T. A. Nijhuis, A. I. Stankiewicz, J. A. Moulijn, Appl. Catal. A Gen. 2004, 266, 109–116.
[28] J. Guilera, R. Bringué, E. Ramírez, M. Iborra, J. Tejero, Ind. Eng. Chem. Res. 2012, 51, 16525–
This article is protected by copyright. All rights reserved.

10.1002/cssc.201801067
ChemSusChem
16530.
[29] J. A. Melero, J. Iglesias, G. Morales, Green Chem. 2009, 11, 1285–1308.
[30] Y. Zhao, W. Li, M. Zhang, K. Tao, Catal. Commun. 2002, 3, 239–245.
[31] S. R. Ginjupalli, S. Mugawar, P. Rajan N., P. Kumar Balla, V. R. C. Komandur, J. Chem. Technol.
Biotechnol. 2014, 89, 1890–1897.
[32] G. Larsen, E. Lotero, S. Raghavan, R. D. Parra, C. A. Querini, Appl. Catal. A Gen. 1996, 139,
201–211.
[33] R. Balzer, V. Drago, W. H. Schreiner, L. F. D. Probst, J. Braz. Chem. Soc. 2014, 25, 2026–2031.
[34] S. Kuba, P. Concepción Heydorn, R. K. Grasselli, B. C. Gates, M. Che, H. Knözinger, Phys.
Chem. Chem. Phys. 2001, 3, 146–154.
[35] D. E. López, K. Suwannakarn, D. A. Bruce, J. G. Goodwin, J. Catal. 2007, 247, 43–50.
[36] E. R. Sacia, M. Balakrishnan, M. H. Deaner, K. a. Goulas, F. D. Toste, A. T. Bell, ChemSusChem
2015, 8, 1726–1736.
[37] H. Pines, W. O. Haag, J. Am. Chem. Soc. 1961, 83, 2847–2852.
[38] M. Balakrishnan, E. R. Sacia, A. T. Bell, ChemSusChem 2014, 7, 1078–1085.
[39] C. D. Baertsch, S. L. Soled, E. Iglesia, J. Phys. Chem. B 2001, 105, 1320–1330.
[40] M. Hino, K. Arata, J. Chem. Soc. Chem. Commun. 1988, 1259–1260.
[41] R. D. Down, J. H. Lehr, Environmental Instrumentation and Analysis Handbook, John Wiley &
Sons, 2005.
Graphical Abstract:
This article is protected by copyright. All rights reserved.