Transition metal triflate catalyzed conversion of alcohols, ethers and esters to olefins

Article abstract of DOI:10.1039/c8ra02437e

Herein, we report an efficient transition metal triflate catalyzed approach to convert biomass-based compounds, such as monoterpene alcohols, sugar alcohols, octyl acetate and tea tree oil, to their corresponding olefins in high yields. The reaction proceeds through C-O bond cleavage under solvent-free conditions, where the catalytic activity is determined by the oxophilicity and the Lewis acidity of the metal catalyst. In addition, we demonstrate how the oxygen containing functionality affects the formation of the olefins. Furthermore, the robustness of the used metal triflate catalysts, Fe(OTf)₃ and Hf(OTf)₄, is highlighted by their ability to convert an over 2400-fold excess of 2-octanol to octenes in high isolated yields.

Full text of DOI:10.1039/c8ra02437e

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Transition metal triflate catalyzed conversion of
alcohols, ethers and esters to olefins†
Cite this: RSC Adv., 2018, 8, 15111
*
J. Keskivali,
A. Parviainen,
K. Lagerblom
and T. Repo
¨
Herein, we report an efficient transition metal triflate catalyzed approach to convert biomass-based
compounds, such as monoterpene alcohols, sugar alcohols, octyl acetate and tea tree oil, to their
corresponding olefins in high yields. The reaction proceeds through C–O bond cleavage under solvent-
free conditions, where the catalytic activity is determined by the oxophilicity and the Lewis acidity of the
metal catalyst. In addition, we demonstrate how the oxygen containing functionality affects the
formation of the olefins. Furthermore, the robustness of the used metal triflate catalysts, Fe(OTf)3 and
Hf(OTf)4, is highlighted by their ability to convert an over 2400-fold excess of 2-octanol to octenes in
high isolated yields.
Received 20th March 2018
Accepted 9th April 2018
DOI: 10.1039/c8ra02437e
rsc.li/rsc-advances
C–O bond cleavage at elevated temperatures in the presence of
Pd/C and at high hydrogen pressures, producing alkanes as
products.13–18 In this respect, the use of metal triates in the
Introduction
There is a great aspiration to replace fossil feedstocks with
sustainable biomass-based resources. Biomass provides a large
supply of different organic compounds, mostly in the form of
cellulose, hemicellulose and lignin.1 As a result, the selective
defunctionalization and upgrading of these oxygen rich
biomass-based substrates has been intensely researched
recently, and various approaches have been devised to obtain
value-added biochemicals, such as alkanes, isosorbide and
furfurals.2–6 In addition to lignocellulose, terpenes are a versa-
tile class of organic compound widely available from biomass,
containing different functionalities, such as alcohols, esters
and ethers.7,8 In this respect, the defunctionalization of
terpenes offers an interesting opportunity to produce high value
biomass-based olens for biopolymer applications as well as for
other purposes.9 However, more attention should be focused on
nding abundant and sustainable catalysts to produce valuable
olens from renewable natural resources.
synthesis of olens is plausible, and ideally the more expensive
metals should be replaced with more abundant and inexpensive
transition metal-based Lewis acid catalysts in the conversion of
biomass-based oxygen containing compounds to olens.
Herein, we present the correlation between the oxophilicity
and the Lewis-acidity of metal triates in the dehydration of
alcohols, and the aptitude of the selected transition metal tri-
ates, Fe(OTf)3 and Hf(OTf)4, to produce olens from various
oxygen containing biomass-based compounds under solvent-
free conditions. The durability and robustness of Fe(OTf)3 and
Hf(OTf)4 are demonstrated with a large scale synthesis of
octenes from 2-octanol. As a proof of concept for the upgrading
of essential oils to alkenes, tea tree oil is converted to olens
using the aforementioned metal triate catalysts.
Results and discussion
The dehydration of alcohols to alkenes is usually conducted
using Brønsted acid catalysts, including H2SO4, HCl or H-
ZSM5.10–12 However, these catalysts are associated with various
problems, such as low product selectivities, corrosive nature as
well as environmental hazards.3,5,10 Accordingly, the develop-
ment of more benign and selective catalysts is preferred. The
use of Lewis acidic metal dehydration catalysts has been
sparsely reported in the synthesis of olens from alcohols,
ethers and esters.10 On the other hand, metal triates, such as
those of Eu, Hf and La, have previously been reported to catalyze
We initiated the study by investigating different catalysts in the
catalytic cleavage of the alcohol C–O bond of 2-octanol (Table 1).
The catalyst scope covered different lanthanide and transition
metal salts, which were mainly their triates. Triic acid (HOTf)
was used as a reference Brønsted acid. The experiments were
conducted under solvent-free reaction conditions to exclude
possible solvent-effects, as, for example, coordinating solvents
have been shown to reduce the dehydration activity of Hf(OTf)4
under hydrodeoxygenation conditions.13,15 To rationalize the
results, the octene yield and conversion were plotted as a func-
tion of the oxophilicity and Lewis acidity of the used metal
cations, respectively (Fig. 1). The oxophilicity was assessed in
terms of the dissociation energy of the M–O bond, and the Lewis
acidity of the metal cations was calculated using the formula:
Department of Chemistry, University of Helsinki, A. I. Virtasen aukio 1, P.O. Box 55,
00014, Finland. E-mail: timo.repo@helsinki.
† Electronic supplementary information (ESI) available: Materials, experimental
and analysis details. See DOI: 10.1039/c8ra02437e
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 15111–15118 | 15111

View Article Online
RSC Advances
Paper
Table 1 Selected metal triflates, their properties and catalytic activity in the dehydration of 2-octanola
Entry
Catalyst
Oxophilicity (kJ molꢁ1 21
)
Lewis acidityc (ꢀ10ꢁ6 pmꢁ3
)
Conv. (%)
Octene yieldb (%)
1
2
3
4
5
6
7
8
Fe(OTf)3
Hf(OTf)4
Eu(OTf)3
Sc(OTf)3
Yb(OTf)3
Y(OTf)3
La(OTf)3
Nd(OTf)3
Al(OTf)3
Mg(OTf)2
Pr(OTf)3
Ti(OTf)4
Cr(OTf)3
FeCl3
409
791
557
674
398
715
799
703
512
394
753
662
427
409
409
409
409
—
11.18
11.18
3.53
7.26
4.59
4.11
2.73
3.16
19.59
5.36
48
>99
9
20
8
30
93
2
6
1
1
5
2
34
4
5
11
10
75
10
10
>99
73
11
11
13
14
>99%
9
10
11
12
13
14
15
16
17
18
3.09
1
18.06
12.90
11.18
11.18
11.18
11.18
—
71
35
0
0
0
Fe(NO3)3
Fe(CO2CH3)3
Fe2(SO4)3
HOTf
0
84
a
Reaction conditions: 3 h, 150 ꢂC, 0.5 mol% of catalyst using a closed batch type reaction vessel. b Mixture of 1-octene and cis/trans isomers of 2-, 3-
and 4-octene, yields obtained with GC-FID using calibration curves. c Lewis acidity ¼ Z/r3, unit 1/pm3. The effect of anions has not been taken into
account in the calculated Lewis acidity.
Lewis acidity ¼ Z/r3 (Z ¼ charge of the metal and r ¼ ionic radius chlorides form HCl when they come into contact with water (e.g.
of the cation in pm).19–21
AlCl3, HfCl4 and TiCl4), thus possessing similar unwanted
Interestingly, the plots revealed a clear dependence of the properties to Brønsted acids, such as HOTf.
reaction outcome on the oxophilicity and Lewis acidity of the
Next, we measured the conversion rates of 2-octanol (6.3
metal cations (Table 1 & Fig. 1). In general, the metals with mmol) with 0.5 mol% loadings of Fe(OTf)3, Cr(OTf)3, Al(OTf)3,
relatively high Lewis acidity (>10 ꢀ 10ꢁ6) gave signicantly Ti(OTf)4 and Hf(OTf)4 at 150 ꢂC. The reactions were monitored
better conversions than those catalysts with lower Lewis acidity using GC-FID with sampling intervals of 5 min, 15 min, 30 min,
(<10 ꢀ 10ꢁ6, Fig. 1A). Further comparison of the octene yields as 1 h, 1.5 h, 2 h and 3 h. The conversion had a linear correlation
a function of oxophilicity with the ve most Lewis acidic metal with time, indicating that the reaction follows zero-order
cations (Fe(OTf)3, Cr(OTf)3, Al(OTf)3, Ti(OTf)4 and Hf(OTf)4) kinetics (Fig. 2). This is common in catalytic reactions where
revealed a linear correlation (Fig. 1B). While Fe(OTf)3, Cr(OTf)3 a catalyst is in large excess of substrates. The order of the
and Al(OTf)3 gave octenes in moderate yields of 30–40%, the reaction arises through saturated catalyst activity, forcing the
most oxophilic Hf(OTf)4 and Ti(OTf)4 generated quantitative catalyst to work at the maximum capacity until the substrate is
conversion of 2-octanol, and octenes were formed in high 93% nearly consumed. The observed rates of the reactions with
and 71% yields, respectively. The reference Brønsted acid different metal triates were in a similar order as the initial
catalyst, HOTf, generated quantitative conversion of 2-octanol catalyst screening. Hf(OTf)4 generated the highest conversion
with 84% yield of octenes (Table 1, entry 18). In addition to rate of 2.76 mmol hꢁ1, which was three times higher than that
olen products, unidentied by-products were formed in all of with Fe(OTf)3. Overall, the order of the conversion rates is
the experiments.
Hf(OTf)4 > Ti(OTf)4 > Al(OTf)3 > Cr(OTf)3 > Fe(OTf)3. For further
Additional experiments were conducted with Fe(OTf)3, studies we selected Fe(OTf)3, due to its benign and non-toxic
Cr(OTf)3 and Al(OTf)3 at an elevated temperature of 165 ꢂC nature, and the most efficient Hf(OTF)4 as catalyst. The cata-
using 2-octanol as a substrate. This resulted in quantitative lytic activity and efficiency of Ti(OTf)4, Al(OTf)3 and Cr(OTf)3 in
conversions and enhanced yields of 80–85%, matching those of the following experiments could be expected to be in between
ꢂ
HOTf, Hf(OTf)4 and Ti(OTf)4 at 150 C (see Table S1 in ESI†). that of Hf(OTf)4 and Fe(OTf)3.
The replacement of the triate anion with chloride resulted in
The effect of the position of the hydroxyl group on the
a signicantly reduced yield and conversion (Table 1, entries Hf(OTf)4 and Fe(OTf)3 catalyzed dehydration reactions was
14–17). The better reactivity with triates can be explained by studied using primary, secondary and tertiary alcohols (Table
the weak electron donor and strong electron acceptor nature of 2). The reactions were conducted in a solvent-free manner in
this anion, which increases the Lewis acidity of the central a ask connected to micro distillation apparatus, allowing the
metal atom.22 In contrast, chloride, acetate, nitrate and sulphate distillation and isolation of the alkenes upon formation. To
are p-donating middle and low eld ligands, according to the ensure efficient conversion, the reactions catalyzed by Fe(OTf)3
spectrochemical series, and because of this they are unlikely to were conducted at higher temperatures than those with
impart greater Lewis acidity.23 Additionally, some of the metal Hf(OTf)4. The tertiary alcohol, 3-ethylpentan-3-ol as a substrate,
15112 | RSC Adv., 2018, 8, 15111–15118
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
Fig. 2 The conversion of 2-octanol with respect to time-on-stream.
The conversion rates were calculated based on the slopes of the fitted
curves.
decrease in the order of tert- > sec- > prim-alcohols, which is
analogous to the order of the stability of the carbocations, and is
in line with previous publications.10 Furthermore, the dehy-
dration of diethylene glycol produced a good 84% yield of 1,4-
dioxane with Hf(OTf)4 (Table 2, entry 7). The dehydration of 1-
phenylpropan-1-ol and 3-phenylpropan-1-ol resulted in forma-
tion of non-volatile compounds, and little of the desired dehy-
drated products were obtained (Table 2, entries 5 & 6).
Fig. 1 (A) The conversion of 2-octanol as a function of the Lewis
acidity of the metal triflate. The Lewis acidity is calculated for the metal
cations, and the effect of the triflate anion is not included. (B) The yield
of octenes as a function of the oxophilicity of the metal cation in the
dehydration of 2-octanol. The oxophilicity is the measured M–O bond
dissociation energy.21
The efficiency and robustness of Fe(OTf)3 and Hf(OTf)4 in the
dehydration reaction was studied by conducting experiments
using 0.06 mmol of the metal triates and 144.9 mmol of 2-
octanol, which is equal to 0.04 mol% catalyst loading. During
the reaction, 2-octanol was added dropwise using a dropping
funnel to keep the amount of substrate constant, while the
octenes formed were distilled. The catalysts converted an over
2400-fold excess of 2-octanol during the reaction, and with the
use of Hf(OTf)4 octenes were generated in 84% isolated yield in
10 hours, while Fe(OTf)3 produced octenes in 80% yield in 24
hours at 165 ꢂC. These experiments underscore the robust
performance of the metal triate catalysts in the dehydration.
We moved on to study the formation of the isomers during
the dehydration reactions. Based on the literature, two possible
pathways were recognized: (i) migration of the charge in the
carbocation intermediate and (ii) alkene hydration (Scheme
1).24,25 The isomerization mechanisms were studied by heating
1-octene in the presence of Hf(OTf)4 with and without water at
150 ꢂC. In the absence of water, 1-octene seemingly underwent
oligomerization, forming a viscous dark brown oil. The
measured 1H NMR spectrum shows that the signals corre-
sponding to the alkene protons were suppressed and shied,
and also that the saturated carbon chain C–H signals were
clearly broadened (see Fig. S23 in ESI†). This oligomerization
might explain the formation of the non-volatile compounds
was converted to 3-ethylpent-3-ene in isolated yields of 79% and
84% using Fe(OTf)3 and Hf(OTf)4, respectively (Table 2, entry 3).
With the secondary, alcohols 2-octanol and cyclohexanol,
higher reaction temperatures were required than with the
tertiary alcohol. With Fe(OTf)3, octenes from 2-octanol were
obtained in a slightly higher 91% yield than the 85% yield with
Hf(OTf)4 (Table 2, entry 1). Various isomers of octene were
observed, and the product ratio of 2-, 3- and 4-octene to 1-octene
was 10 : 1 with both the catalysts (see Fig. S5 in ESI†). Cyclo-
hexanol was converted to cyclohexene in 80% and 78% yield
with Hf(OTf)4 and Fe(OTf)3, respectively (Table 2, entry 4). Full
conversion of the primary alcohol, 1-octanol, was achieved aer
6 h at 180 ꢂC using Fe(OTf)3 as a catalyst (Table 2, entry 2).
Unexpectedly, dioctyl ether was the main product with 80%
yield, which was accompanied by octene in only 2% yield, and
the rest being unidentied side products. Due to this, no further
Fe(OTf)3 catalyzed experiments with primary alcohols were
conducted. With Hf(OTf)4, 1-octanol was converted to octenes,
generating 46% yield in 6 hours and 65% yield in 12 hours at
1
from the aromatic alcohols.26 Based on the H NMR analysis,
ꢂ
180 C (Table 2, entry 2). The ratio of 2-, 3- and 4-octene to 1-
octene isomers or oligomers do not form in the presence of
water, which indicates that the isomers are formed through the
octene was 30 : 1. The reactivity of the alcohols was found to
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 15111–15118 | 15113

View Article Online
RSC Advances
Paper
a
Table 2 The results of dehydration of different alcohols with 0.5 mol% loading of Fe(OTf)3 or Hf(OTf)4
Fe(OTf)3 catalyst
Hf(OTf)4 catalyst
Entry
Substrate (mmol)
t (h)
T (ꢂC)
Yieldb (%)
t (h)
T (ꢂC)
Yieldb (%)
1
2
3
6
165
180
91
2
1.5
12
150
180
85
65
3
4
0.5
2.5
130
165
79
80
0.5
3.5
110
150
84
78
5
6
0.5
130
3
0.5
110
3
—
—
—
—
—
—
6
180
180
N.d.
85c
7
4.5
N.d. ¼ not detected. b Isolated olen yields obtained by distillation of the products during the reaction. Analyzed using GC-MS, 1H NMR and 13
C
a
NMR (see ESI). c 1,4-Dioxane was the reaction product.
emphasis was on the secondary and tertiary monoterpene
alcohols, due to their more reactive nature. Using menthol and
dihydromyrcenol, 71–84% isolated yields of the corresponding
terpenes and dihydromyrcenes were obtained with both metal
triates (Table 3, entries 1 & 2). With Fe(OTf)3, a-terpineol was
successfully converted to limonene, a-terpinen, g-terpinen and
a-terpinolene in 76% yield. Meanwhile, Hf(OTf)4 gave only 16%
yield of the olens and mainly formed a viscous non-volatile oily
substance similar to with the aromatic alcohols (Table 3, entry
3). The formation of non-volatile oils was also observed with
linalool and citronellol (Table 3, entries 4 & 5). In fact, all of the
secondary and tertiary alcohols forming these non-volatile oils
had electron donating groups, such as methyl, connected to the
sp2-hybridized carbon. This is known to facilitate cationic
polymerization.9 It is also known that metal halides are capable
of catalyzing cationic polymerization, and thus it is likely that
the non-volatile oils herein are a result of metal triate catalyzed
cationic oligomerization.27,28 It is therefore rational that the
more Lewis acidic and oxophilic, and thus stronger dehydration
catalyst, Hf(OTf)4 also catalyzes the oligomerization more effi-
ciently than Fe(OTf)3 (Table 3, entry 3). The dehydration of
sugar alcohols was studied using sorbitol, which is readily
available from cellulose- and hemicellulose-based glucose.29 As
a result, isosorbide was generated in yields of 77% and 78%
with Fe(OTf)3 and Hf(OTf)4 as the catalysts, respectively (Table
3, entry 6). In comparison to the previously published metal
triate-catalyzed conversion of sorbitol to isosorbide, the use of
Fe(OTf)3 and Hf(OTf)4 afforded isosorbide in similar yields but
in shorter reaction times and with lower catalyst loadings.30,31
Scheme 1 Two putative pathways for octene isomerization: carbo-
cation migration and hydrolysis induced isomerization.
migration of the carbocation in the transition state. This
proposition is in agreement with the previous report on the
isomerization of alkenes.24 The addition of water decreases the
oligomerization through the formation of a biphasic system
and the solvation of Hf(OTf)4, thus limiting the interaction of
the catalyst and the alkenes.
Due to the promising results obtained using the model
alcohols, we conducted Fe(OTf)3 and Hf(OTf)4 catalyzed dehy-
dration of the biomass-based monoterpene alcohols. The reac-
tions were conducted in a solvent-free manner using micro
distillation apparatus to isolate the products as they formed
(Table 3). Some reactions were conducted under reduced pres-
sure to ensure the distillation of the products. The main
15114 | RSC Adv., 2018, 8, 15111–15118
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
Table 3 The results of the solvent-free dehydration of alcohols with 0.5 mol% loading of Fe(OTf)3 or Hf(OTf)4
Fe(OTf)3 catalyst
Hf(OTf)4 catalyst
Entry
1
Substrate (mmol)
t (h)
T (ꢂC)
Yielda (%)
71
t (h)
T (ꢂC)
Yielda (%)
82
Formed productsd
4
165
3
150
2
3
4
5
0.5
0.5
0.5
—
130
130
110
—
82
76
3
0.5
0.5
0.5
24
110
110
110
180
84
16
3
—
7
6b
3
150
77c
3
130
78c
a
b
Isolated yields obtained by distillation of the olens during the reaction. Analyzed using GC-MS, 1H NMR and 13C NMR (see ESI). Batch type
reaction was conducted in vacuo to increase the efficiency of the dehydration, and product yields were measured using HPLC-FID and
c
calibration curves, with isosorbide as the reaction product. Product yields were determined using HPLC-FID and calibration curves, with
isosorbide as the reaction product. d Product distributions are presented in the ESI.
As metal triates could potentially catalyze the conversion of highest yield of 29% was obtained aer 24 hours at 180 ꢂC
essential oils to olens, we composed an experiment as a proof (Table 4, entries 8 & 9). No alkenes were detected using Hf(OTf)4
of concept using tea tree oil as a substrate in Fe(OTf)3 and without additional water (Table 4, entry 10). Evidently, water is
Hf(OTf)4 catalyzed dehydration reactions. Tea tree oil consists an essential additive to convert the ethers to alcohols through
mainly of terpinen-4-ol, thus providing an ample source of hydrolysis, followed by the yield-reducing cumbersome dehy-
biomass-based alcohol for the synthesis of olens.32,33 With the dration of the primary alcohols to alkenes. However, Fe(OTf)3
use of Fe(OTf)3, a mixture of products, mainly a- and g-terpi- could not convert dioctyl ether to alkenes or alcohols in
nenes, was generated in 53 wt% yield from tea tree oil in 1 hour signicant amounts with either approach (see Table S2 in ESI†).
under a reduced pressure of 57 mmHg at 130 ꢂC. However, the The attempted conversion of dimethyl tetrahydrofuran to
Hf(OTf)4-catalyzed reaction produced non-volatile products at olens through hydrolysis and subsequent dehydration failed
110 ꢂC under reduced pressure, and as a result no olens were to produce any olen products with either of the used catalysts,
obtained. The reaction results resemble those of the a-terpineol Fe(OTf)3 or Hf(OTf)4.
dehydration, where Fe(OTf)3 gave a good olen yield while
Hf(OTf)4 generated mostly non-volatile products (Table 3,
entry 3).
We also investigated two metal triate-catalyzed non-
hydrogenolytic approaches to the scission of ester and ether
C–O bonds: (i) direct cleavage and (ii) hydrolysis, followed by
the dehydration of the hydroxyl groups to alkenes (Scheme 2).
Previously, ester and ether C–O bond cleavage has been ach-
ieved through hydrogenolysis catalyzed by Hf(OTf)4 and Pd/
C.18,34 Accordingly, on dioctyl ether and octyl acetate we carried
out several reactions with and without water to see the effect of
hydrolysis (Table 4). It turned out that the use of Hf(OTf)4 and
water (1–2 mol eq.) resulted in the hydrolysis of dioctylether,
forming 1-octanol. Subsequently, the as-formed 1-octanol was
Scheme
2 Two theoretical pathways for alkene formation from
dehydrated to octenes, although in relatively low yields. The
dioctyl ether and octyl acetate.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 15111–15118 | 15115

View Article Online
RSC Advances
Table 4 The results of the conversion of dioctyl ether and octyl acetate into octenes with 0.5 mol% loading of Fe(OTf)3 or Hf(OTf)4
Paper
a
Entry
Catalyst
Fe(OTf)3
Fe(OTf)3
Substrate (mmol)
Time (h)
Temp. (ꢂC)
Octene yieldb (%)
1
2
3
3
170
4
4
d
170
3
4
5
6
7
Fe(OTf)3
36
3
180
150
170
170
180
68c
13
5
Hf(OTf)4
Hf(OTf)4
Hf(OTf)4
Hf(OTf)4
d
3
3
20
80c
22
d
d
8
9
Hf(OTf)4
Hf(OTf)4
8
180
180
7
24
29
10
Hf(OTf)4
8
180
N.d.
a
N.d. ¼ not detected. b Yield determined using GC-FID and calibration curves. c Isolated yields obtained by distillation of the olens during the
reaction. Analyzed using GC-MS, 1H NMR and 13C NMR. d 1 equivalent of H2O added to induce hydrolysis of the ether/ester bond.
The hydrolysis experiments with octyl acetate generated conversion using Fe(OTf)3 and Hf(OTf)4, although the deoxy-
octenes with both approaches and catalysts (Table 4, entries 2 & genation of carboxylic acids has been reported to occur under
5). In general, the addition of water resulted in the formation of similar conditions with Pd or Pt as catalysts.36
1-octanol and acetic acid as the hydrolysis products. However,
this approach leads to low alkene yields, as dehydration of the
primary hydroxyl group is difficult even with Hf(OTf)4 (see
above). In this respect, the addition of water should be avoided.
Without water, the direct ester C–O cleavage approach resulted
in the formation of acetic acid and octene from octyl acetate
through the scission of the alkoxy C–O bond (Table 4, entries 1,
4 & 6). This direct approach generated high isolated yields of
Conclusions
In summary, highly oxophilic and Lewis acidic metal triates
can be used to catalyze the cleavage of C–O bonds of biomass-
based alcohols, esters and ethers to give the corresponding
olens. The catalytic activity of the transition metal triates
improved with increasing oxophilicity and Lewis-acidity.
68% and 80% of octenes with Fe(OTf)3 and Hf(OTf)4, respec-
Understanding of this correlation aids in the development of
tively (Table 4, entries 3 & 7). Compared to the previous Lewis
new transition metal based defunctionalization catalysts. The
reactivity order of the different oxygen containing species to
acid catalyzed synthesis of olens, our setup allows the use of
benign and abundant transition metal triate catalysts as well
alkenes is tert-alcohols > sec-alcohols > esters > prim-alcohols >
as various alcohols, esters and ethers as substrates.10,35
ethers. Alcohols and esters can be converted efficiently to
olens under solvent-free conditions using metal triate cata-
Furthermore, the solvent-free approach enables the isolation of
the products in good yields as they form.
lysts. However, the conversion of ethers to alkenes requires
The compatibility of aldehydes and carboxylic acids as
initial hydrolysis of the ether bond to an alcohol, which is then
substrates was also investigated. Under the applied conditions,
followed by the dehydration of the formed alcohol. The proof of
aldehydes underwent an immediate oligomerization to non-
concept experiments demonstrated that the studied Fe(OTf)3
volatile oils, while carboxylic acids did not undergo any
and Hf(OTf)4 are very robust catalysts, as in the solvent-free
15116 | RSC Adv., 2018, 8, 15111–15118
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
continuous distillation process they converted an over 2400-fold
excess of 2-octanol into octenes in high yields of 84% and 80%,
respectively. Also, the usability of essential oils as substrates
was established by the dehydration of tea tree oil, producing
alkenes in up to 53 wt% yield.
Notes and references
1 S. V. Vassilev, D. Baxter, L. K. Andersen, C. G. Vassileva and
T. J. Morgan, Fuel, 2012, 94, 1–33.
2 G. W. Huber, J. Chheda, C. B. Barrett and J. A. Dumesic,
Science, 2005, 308, 1446–1450.
3 S. Van de Vyver, J. Geboers, P. A. Jacobs and B. F. Sels,
ChemCatChem, 2011, 3, 82–94.
4 A. A. Rosatella, S. P. Simeonov, R. F. M. Frade and
C. A. M. Afonso, Green Chem., 2011, 13, 754–793.
Experimental
General procedure for the synthesis of olens
The desired amount of catalyst (0.5 mol%, unless otherwise
stated) was placed into a reaction vessel, to which an appro-
priate amount of substrate and any additives, such as water,
were added. The vessel was heated with an oil bath at a suitable
temperature. Batch type reactions were conducted in a closed
vial, and the distillation reactions were conducted using micro
distillation apparatus at atmospheric or reduced pressure,
depending on the substrate. The forming products and side
products, water or acetic acid, were distilled and collected in
a product ask, which was cooled with an ice bath to minimize
evaporation of the products. The distillation reactions were
continued until completion, and no more products could be
recovered. Aer the reaction, when water was the side product
the product ask was placed in a freezer for ease of separation
of the olen products from the frozen water. In the case of acetic
acid as the side product, the acetic acid–olen mixture was
washed with 1 M NaOH(aq) to neutralize the acetic acid. Pure
olen products were obtained aer drying with Na2SO4 and
ltration. The reaction products were then analysed using GC-
¨
5 R. Karinen, K. Vilonen and M. Niemela, ChemSusChem, 2011,
4, 1002–1016.
6 M. Rose and R. Palkovits, ChemSusChem, 2012, 5, 167–176.
7 A. Corma Canos, S. Iborra and A. Velty, Chem. Rev., 2007, 107,
2411–2502.
8 O. Lasekan and K. A. Abbas, Crit. Rev. Food Sci. Nutr., 2012,
52, 726–735.
9 P. A. Wilbon, F. Chu and C. Tang, Macromol. Rapid Commun.,
2013, 34, 8–37.
10 T. J. Korstanje, E. F. de Waard, J. T. B. H. Jastrzebski and
R. J. M. Klein Gebbink, ACS Catal., 2012, 2, 2173–2181.
¨
11 M. Eisenacher, S. Beschnitt and W. Holderich, Catal.
Commun., 2012, 26, 214–217.
12 M.-H. Lee, S.-W. Lee, Y.-M. Jeong, D.-Y. Park and J.-Y. Ryu,
WO 2005035468, 2005.
13 Z. Li, R. S. Assary, A. C. Atesin, L. A. Curtiss and T. J. Marks, J.
Am. Chem. Soc., 2014, 136, 104–107.
¨
14 J. Keskivali, P. Wrigstedt, K. Lagerblom and T. Repo, Appl.
1
FID, GC-MS, H NMR and 13C NMR. For a detailed description
Catal., A, 2017, 534, 40–45.
15 H. J. Song, J. Deng, M. S. Cui, X. L. Li, X. X. Liu, R. Zhu,
W. P. Wu and Y. Fu, ChemSusChem, 2015, 8, 4250–4255.
16 A. D. Sutton, F. D. Waldie, R. Wu, M. Schlaf, A. Louis III and
J. C. Gordon, Nat. Chem., 2013, 5, 428–432.
17 A. E. King, T. J. Brooks, Y.-H. Tian, E. R. Batista and
A. D. Sutton, ACS Catal., 2015, 5, 1223–1226.
18 T. L. Lohr, Z. Li and T. J. Marks, ACS Catal., 2015, 5, 7004–
7007.
of the experimental setups see the ESI.†
GC-analysis
The reaction products were analysed using an Agilent Tech-
nologies 7890B GC System equipped with a ame ionization
detector (FID) and an Agilent Technologies 5977C MSD mass
detector (MS). HP-5MS UI (30 m, 0.25 mm, 0.25 mm) was used as
the column in the GC-analysis.
19 K. Mikami, M. Terada and H. Matsuzawa, Angew. Chem., Int.
Ed., 2002, 41, 3554–3572.
NMR-analysis
20 R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr.,
Theor. Gen. Crystallogr., 1976, 32, 751–767.
21 J. A. Dean, Lange’s Handbook of chemistry, McGraw-Hill, Inc,
New York, 1985.
22 S. Voigt, H. Glaum, S. Schlecht, J. Magull, D. Fenske and
K. Dehnicke, Communications, 1997, 3, 1994–1995.
23 Y. Shimura, Bull. Chem. Soc. Jpn., 1988, 61, 693–698.
Nuclear magnetic resonance (NMR) experiments (1H and 13C)
were conducted using a Varian Mercury Plus 300 MHz (1H
frequency 300 MHz and 13C frequency 75 MHz) and deuterated
solvents.
HPLC-analysis
´
24 V. Macho, M. Kralik, E. Jurecekova, J. Hudec and L. Jurecek,
High-performance liquid chromatography (HPLC) analysis was
conducted using an Agilent Technologies 1200 Series HPLC
equipped with Phenomenex Rezex ROA – organic acid H+ (8%)
(300 ꢀ 7.80 mm) column and refractive index detector (RID). A
5 mM H2SO4 aqueous solution was used as the eluent with the
Appl. Catal., A, 2001, 214, 251–257.
25 M. B. Smith and J. March, March’s advanced organic
chemistry, John Wiley & Sons, Inc., New Jersey, 6th edn, 2007.
26 C. H. Bartholomew, Appl. Catal., A, 2001, 212, 17–60.
27 W. J. Roberts and A. R. Day, J. Am. Chem. Soc., 1950, 72, 1226–
1230.
ow rate of 0.35 ml minꢁ1
.
28 J. Lu, M. Kamigaito, M. Sawamoto, T. Higashimura and
Y.-X. Deng, J. Appl. Polym. Sci., 1996, 61, 1011–1016.
Conflicts of interest
There are no conicts to declare.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 15111–15118 | 15117

View Article Online
RSC Advances
Paper
29 W. Deng, X. Tan, W. Fang, Q. Zhang and Y. Wang, Catal. 33 G. Swords and G. L. K. Hunter, J. Agric. Food Chem., 1978,
Lett., 2009, 133, 167–174. 26(3), 734–737.
30 F. Liu, K. De Oliveira Vigier, M. Pera-Titus, Y. Pouilloux, 34 T. L. Lohr, Z. Li, R. S. Assary, L. A. Curtiss and T. J. Marks,
´ ˆ
J.-M. Clacens, F. Decampo and F. Jerome, Green Chem.,
Energy Environ. Sci., 2016, 9, 550–564.
2013, 15, 901–909.
35 Z. Zhu and J. H. Espenson, J. Org. Chem., 1996, 61, 324–328.
˚
ˇ
´
¨
¨
31 K. Stensrud, E. Hagberg, S. Howard and E. M. Rockafellow, 36 M. Snare, I. Kubickova, P. Maki-Arvela, K. Eranen and
WO 2014137619 A1, 2014.
D. Y. Murzin, Ind. Eng. Chem. Res., 2006, 45, 5708–5715.
32 J. J. Brophy, N. W. Davies, I. A. Southwell, I. A. Stiff and
L. R. Williams, J. Agric. Food Chem., 1989, 37, 1330–1335.
15118 | RSC Adv., 2018, 8, 15111–15118
This journal is © The Royal Society of Chemistry 2018