Liquid-phase dehydration of 1-octanol, 1-hexanol and 1-pentanol to linear symmetrical ethers over ion exchange resins

Article abstract of DOI:10.1016/j.apcata.2011.02.006

Dehydration of 1-octanol, 1-hexanol and 1-pentanol to di-n-octyl ether (DNOE), di-n-hexyl ether (DNHE) and di-n-pentyl ether (DNPE), respectively, has been studied in the liquid phase at 423K in a batch reactor on ion exchange resins as catalysts. Tested catalysts were the macroporous resins Amberlyst 15, Amberlyst 35, CT175, CT275 and CT276 (high crosslinking degree); Amberlyst 16, Amberlyst 36 and CT252 (medium crosslinking degree); Amberlyst 39 and Amberlyst 70 (low crosslinking degree), and the gel type ones CT224, Amberlyst 31, Amberlyst 121 and Dowex 50Wx4-50. Amberlyst 46, an ion-exchange resin sulfonated only at the polymer surface and Nafion NR50 were also tested for comparison purposes. Data show that the yield of linear symmetrical ethers (DNOE, DNHE and DNPE) highly depends on the resin structure so that best results are obtained on Amberlyst 121, Amberlyst 31 and Dowex 50Wx4-50 (the most swollen resins in reaction medium). The high ether yields are due both to the high alcohol conversions and especially to the very high selectivity to ethers (≥94%) shown by that resins. Finally, by comparing initial reaction rates of ether formation with those obtained on Amberlyst 46 an estimation of the fraction of sulfonic groups that take part in the reaction is given.

Full text of DOI:10.1016/j.apcata.2011.02.006

Applied Catalysis A: General 396 (2011) 129–139
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Liquid-phase dehydration of 1-octanol, 1-hexanol and 1-pentanol to linear
symmetrical ethers over ion exchange resins
∗
C. Casas, R. Bringué, E. Ramírez, M. Iborra, J. Tejero
Department of Chemical Engineering, Faculty of Chemistry, University of Barcelona, C/Martí i Franquès 1, E-08028 Barcelona, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Dehydration of 1-octanol, 1-hexanol and 1-pentanol to di-n-octyl ether (DNOE), di-n-hexyl ether (DNHE)
and di-n-pentyl ether (DNPE), respectively, has been studied in the liquid phase at 423K in a batch
reactor on ion exchange resins as catalysts. Tested catalysts were the macroporous resins Amberlyst
Received 27 October 2010
Received in revised form 31 January 2011
Accepted 3 February 2011
Available online 4 March 2011
1
5, Amberlyst 35, CT175, CT275 and CT276 (high crosslinking degree); Amberlyst 16, Amberlyst 36 and
CT252 (medium crosslinking degree); Amberlyst 39 and Amberlyst 70 (low crosslinking degree), and the
gel type ones CT224, Amberlyst 31, Amberlyst 121 and Dowex 50Wx4-50. Amberlyst 46, an ion-exchange
resin sulfonated only at the polymer surface and Nafion NR50 were also tested for comparison purposes.
Data show that the yield of linear symmetrical ethers (DNOE, DNHE and DNPE) highly depends on the
resin structure so that best results are obtained on Amberlyst 121, Amberlyst 31 and Dowex 50Wx4-50
(the most swollen resins in reaction medium). The high ether yields are due both to the high alcohol
conversions and especially to the very high selectivity to ethers (≥94%) shown by that resins. Finally, by
comparing initial reaction rates of ether formation with those obtained on Amberlyst 46 an estimation
of the fraction of sulfonic groups that take part in the reaction is given.
Keywords:
Alcohol dehydration
Di-n-octyl ether (DNOE)
Di-n-hexyl ether (DNHE)
Di-n-pentyl ether (DNPE)
Ion-exchange resin catalysts
©
2011 Elsevier B.V. All rights reserved.
1
. Introduction
Blends with linear ethers improve cold performance in relation to
commercial gasoil, and as boiling point of such ethers (from 453
to 528 C) coincides with the low part of distillation curve of gasoil
◦
In recent years European oil industry has made a big effort to
adapt the production facilities to maximize the gasoil production
and, at the same time, upgrade the quality of diesel fuel, as a con-
sequence of the following factors: (1) the quick rise of the whole
number of diesel fueled cars; (2) limitations in amounts of pol-
lutants in exhaust gases (particulate matter, CO, NOx, unburned
hydrocarbons, smoke) imposed by the Euro IV directive [1,2];
(443–633 K) its addition upgrades the vaporization of the blend. As
a consequence of its upgraded quality, combustion of these blends
is cleaner and pollutants amount in exhaust gases decreases sub-
stantially. Finally, it is to be noted that due to the dilution effect
by the introduction of other compounds, the sulfur and aromatic
polycyclic compounds content decreases.
(
3) the severe specifications ruled by the 2003/17/EC directive
Obtaining linear ethers is interesting for refining industry as
useful fuel components. In addition, their synthesis might be an
option to upgrade refinery streams since 1-octanol, 1-hexanol
and 1-pentanol can be produced industrially by the hydroformi-
lation of linear olefins from fluid catalytic cracking or else catalytic
polymerization of unsaturated cracker-gas [5,6]. On the other
hand, 1-octanol and 1-hexanol are the main byproducts of ethanol
oligomerization on hydroxyapatite [7]. As a consequence, DNHE
and DNOE might be alternatively considered as biodiesel for blend
with commercial gasoil
Linear symmetrical ethers are produced by the bimolecular
dehydration of primary alcohols, such as 1-pentanol, 1-hexanol
or 1-octanol over acid catalysts. As a rule, industrial processes for
obtaining linear symmetrical ethers make use of sulfuric acid as
catalyst [8]. However, acid solid catalysts have the advantages of
easier separation of the catalyst from the reaction mixture [9],
and also they yield a reaction product free of blacken compounds.
involving cetane number, distillation curve, viscosity, density, cold
flow properties and sulfur content; and (4) the introduction of
bio-compounds in the automobile fuels composition ruled by the
2
009/28/EC and 2009/30/EC directives. An option to upgrade the
quality of diesel fuel and fulfill such requirements is the reformula-
tion of gasoil by using suitable oxygenated compounds in the same
way that oxygenates were used in the 90s in gasoline reformulation.
It has been reported that the addition of linear ethers contain-
ing more than nine carbon atoms to commercial gasoil upgrades
the quality of the diesel blend [1,3,4]. Reformulated diesel fuels
show higher cetane number as a result of the addition of a long
oxygenated compound of cetane number ranging from 100 to 118.
∗ Corresponding author. Tel.: +34 93 402 1308; fax: +34 93 402 1291.
E-mail address: jtejero@ub.edu (J. Tejero).
0
926-860X/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2011.02.006

1
30
C. Casas et al. / Applied Catalysis A: General 396 (2011) 129–139
resins below 423 K, and thermostable resin Amberlyst 70 in the
Nomenclature
range 423–463 K, catalyze the dehydration reaction of 1-pentanol
to DNPE with selectivity to DNPE higher than 95% and high reac-
tion rates [15,16]. Amberlyst 70 is also efficient and selective to
ether in 1-hexanol dehydration to di-n-hexyl ether (DNHE) [17].
By-products found in both dehydration reactions are olefins and
branched ethers.
DNHE
DNOE
DNPE
dp
dpore
DVB
di-n-hexyl ether
di-n-octyl ether
di-n-pentyl ether
particle diameter (mm)
pore diameter (nm)
divinylbenzene
To the best of our knowledge, literature on 1-octanol dehy-
dration to di-n-octyl ether (DNOE) is particularly scarce. Alumina
based catalysts have been used in the gas phase dehydration of
alcohols [18,19]. In a very comprehensive study about the dehy-
dration of C5-C12 linear alcohols to fuel ethers over ␩-alumina [18],
an 1-octanol conversion of 50% with selectivity to DNOE of 65%
GHSV
gas hourly space velocity (h−1)
acid capacity (meq/g)
mol of compound j (mol)
+
[
H ]
nj
PS-DVB polystyrene divinylbenzene
rj initial reaction rate of formation of compound j
mol/(h kg cat))
−1
is reported at 523 K (fixed bed reactor, GHSV = 4 h ), whereas at
(
6
23 K selectivity to olefin is almost of 100%. Zeolite beta and Nafion-
k
Sj
selectivity to product k with respect to reactant j (%)
H have been tested in liquid phase by using batch reactors equipped
with a Dean & Stark device for water removal. In the presence of
zeolite beta, 1-octanol dehydrates selectively to C8 olefins [20]. On
the contrary, 97% of 1-octanol is converted overnight on perfluoro-
alkanesulfonic acid Nafion-H with near 100% selectivity to DNOE
at 418-423 K [21]. Despite Nafion-H is an interesting catalyst for its
thermal and chemical stability, its industrial use is economically
limited. Finally, etherification of C –C alcohols to respective di-
2
Sarea
SBET
t
surface area determined from ISEC data (m /g)
BET surface area (m /g)
time (h)
2
TOFj
turnover frequency of formation of compound j
(
mol/(h eq))
3
Vpore
Vsp
W
pore volume (cm /g)
volume of the swollen polymer (cm /g)
catalyst mass (dried) (mass units)
3
4
10
n-alkyl ether in presence Cationite KU-2 in the acidic form give rise
to linear ether yields ranging from 86% (di-n-butyl ether) to 59%
(di-n-decyl ether) [22].
Xj
Yjk
conversion of compound j (%)
yield of product k with respect to reactant j (%)
ꢀ
ꢁ
porosity (%)
skeletal density (g/cm )
Ion-exchange resins might be efficient and selective in the dehy-
dration reaction of 1-octanol to DNOE in the same way that in DNPE
and DNHE synthesis. As a consequence, the aim of this paper is
to study the reaction in liquid phase on a series of resin catalysts
without water removal. For comparison purposes, dehydrations of
1-pentanol and 1-hexanol are conducted also on resin catalysts
other than those reported in previous works [15–17]. The relation-
ship between morphology and catalytic behavior of tested resins is
discussed as well.
3
s
Subscripts
HeOH
OcOH
PeOH
1-hexanol
1-octanol
1-pentanol
alcohol, being OcOH, HeOH or PeOH
ROH
Superscripts
0
C
C
C
C
initial
olefins, being C , C or C
pentenes
hexenes
octenes
di-n-hexyl ether
di-n-octyl ether
di-n-pentyl ether (not di-n-hexyl ether)
=
=
=
=
2. Experimental
8
6
5
=
5
=
2.1. Materials
6
=
8
DNHE
DNOE
DNPE
1-Octanol (>95% pure; Fluka) was used without further purifi-
cation. DNOE and 1-octene (Fluka) as well as bidistilled wated
were also used for analysis purposes. Alcohols, olefins and ethers
used in DNPE and DNHE reaction systems can be found elsewhere
[16,17].
Ion exchange resin catalysts were supplied by Rohm
and Haas (Amberlyst 15, Amberlyst 16, Amberlyst 31,
Amberlyst 35, Amberlyst 36, Amberlyst 39, Amberlyst 46,
Amberlyst 70 and Amberlyst 121), Purolite (CT175, CT224,
CT252, CT275 and CT276), (Dowex 50Wx4-50, and Nafion
NR50).
ꢀ
ꢀ
HeOHe C12 branched ethers
OcOOc C16 branched ethers
PeOPe C10 branched ethers
ROR
ROR
ꢀ
linear ether, being DNOE, DNHE or DNPE
branched ethers, being OcOOc , HeOHe or PeOPe
ꢀ
ꢀ
ꢀ
ꢀ
Sulfonated polystyrene-divinylbenzene (PS-DVB) resins catalyze a
great number of reactions, i.e. alkylation, hydration, etherification,
dehydration, etc. They show high concentrations of acid sites quite
uniform in nature [10], but working temperatures are limited by
their low thermal stability below 473 K (423 K typically) [11,12].
Acid strength and thermal resistance to desulfonation of PS-DVB
resins can be enhanced by adding electron attracting groups to the
sulfonated phenyl rings [10] such as halogen atoms as in the case of
the resin Amberlyst 70, with chlorine substituting hydrogen atoms
in polymer chains. In reaction systems with water release, an addi-
tional important drawback is the inhibition of the reaction by water
2
.2. Apparatus
The experiments were carried out in a 100-mL nominal stainless
steel autoclave operated in batch mode. A magnetic drive turbine
was used for mixing, and a baffle was placed inside to improve the
mixing. Experiments were performed at 423 K; the temperature
was controlled to within ± 1 by an electrical furnace. The pressure
was set at 2.0 MPa by means of N to maintain the reaction mixture
2
in liquid phase. A reactor outlet was connected directly to a liquid
sampling valve, which injected 0.1 ␮L of liquid into a GLC apparatus.
Reaction was controlled by a PC with a designed LabView software
program.
[
13,14].
It is shown in literature that ion exchange resins are suitable cat-
alysts for the synthesis of di-n-pentyl ether (DNPE). Microporous

C. Casas et al. / Applied Catalysis A: General 396 (2011) 129–139
131
2.3. Analysis
variation of nROR (mole of ROR) and nROH (mol of alcohol), respec-
tively, vs. time:
Analyses were carried out by means of a Hewlett-Packard
GLC equipped with a TCD. A HP-Pona Methyl Siloxane (HP
ꢀ
ꢁ
ꢂ
ꢃ
1
W
dnROR
dt
0
r
=
mol/ (h kg cat)
(6)
(7)
ROR
t=0
1
90915-001) capillary column (50 m length × 200 ␮m I.D. × 0.5 ␮m
ꢀ
ꢁ
ꢂ
ꢃ
width of stationary phase) was used to determine 1-octene, 2-
octene (cis/trans), 3-octene (cis/trans), 4-octene, 1-octanol, DNOE,
water and C16 branched ethers. The column was temperature-
1
dnROH
dt
0
−r
=
mol/ (h kg cat)
ROH
W
t=0
Initial turnover frequency, TOF, for both ROR formation and ROH
−1
programmed with a 10 K min initial ramp from 323 K to 523 K and
held for 6 min. Helium (30 mL/min) was the carrier gas. Analysis
details for 1-pentanol and 1-hexanol systems can be found else-
where [16,17]. All the species were identified with a GLC equipped
with MS (Agilent GC/MS 5973) and chemical database software.
consumption were estimated by dividing r0 and r0 by the acid
ROR
ROH
capacity, respectively:
0
r
ꢂ
ꢃ
0
ROR
ROR
TOF
=
mol/ (h eq)
(8)
(9)
acid capacity
0
r
ꢂ
ꢃ
0
ROH
ROH
2
.4. Procedure
TOF
=
( )
mol/ h eq
acid capacity
Resins were dried at 383 K in an oven, firstly at 1 bar for 1 h,
and then for 15 h at 10 mbar. The reactor was loaded with 1 g of
dried resin and 70 mL of alcohol (1-octanol, 57.8 g; 1-hexanol or
From Eqs. (6) to (9) it follows that DNOE formation and 1-octanol
consumption rates, and turnover frequencies, are related by:
−
rOcOH
DNOE
1
-pentanol, 57 g). The reaction mixture was pressurized to 2.0 MPa
rDNOE = SOcOH
(10)
(11)
2
by means of N , heated to 423 K and stirred at 500 rpm. The time
2
−
TOFOcOH
the mixture reached 423 K was considered as the zero time of the
experiment. Heating time was about 20 min, and alcohol conver-
sion at zero time was always less than 0.5%. Liquid samples were
taken out hourly and analyzed online to follow the reaction until
the end of the experiment for 6 h.
DNOE
TOFDNOE = SOcOH
2
3
. Results and discussion
3
.1. Structure and textural parameters of ion exchangers
In each experiment, the alcohol conversion (XROH), selectivity to
=
ROH
ROR
ROH
), olefins (SC ), and branched ethers
linear symmetrical ether (S
ꢀ
Physical properties of resins are seen in Table 1. Macroporous
resins include polymers of low (Amberlyst 39 and 70), medium
Amberlyst 16 and 36, and CT252) and high crosslinking degree
ROR
ROR
(
S
), as well as yield of linear symmetrical ether (YROH) relative
ROH
to alcohol, were estimated by the expressions:
(
0
ROH
(Amberlyst 15 and 35, and CT175, CT275 and CT276); gel type
ones include resins containing from 2 to 8 DVB%. As for sulfonation
mole of alcohol reacted
mole of alcohol initially
n
− nROH(t)
XROH(t) =
=
0
n
ROH
degree, selected resins include conventionally sulfonated (a –SO H
3
2
· n
(t) + nC
ROR
=
(t) + 2 · n
ꢀ
(t)
group per styrene ring) and oversulfonated (a –SO3H group per
benzene ring). Amberlyst 35 and 36 are oversulfonated versions of
Amberlyst 15 and 35, respectively, and CT275 and CT276 of CT175.
Special properties of Amberlyst 70 (with hydrogen atoms of poly-
mer chain substituted by chlorine), Amberlyst 46 (sulfonated only
at the polymer surface) and NR50 (a gel-type copolymer of Teflon®
and perfluoro-alkanesulfonic monomers) have to be emphasized.
Fluorine atoms of polymer chains upgrade thermal stability of NR50
and give a higher acid strength than PS-DVB resins.
Ion exchange resin catalysts are nearly spherical beads of sul-
fonated PS-DVB copolymers. If copolymerization of styrene and
divinylbenzene takes place in the presence of a solvent like as
toluene soluble in styrene–DVB mixtures but unable to react
with both co-monomers, toluene is excluded of polymer and the
spaces filled by the solvent become permanent pores (macroporous
resins). On the contrary, in the absence of a solvent, polymeriza-
tion gives place to an ensemble of entangled styrene–DVB chains
with no permanent pores in dry estate (gel-type or microporous).
Macroporous resins consist of large agglomerates of gel micro-
spheres, and each micro-sphere shows smaller nodules that are
more or less fused together [28]. In between the nodules there
is a family of very small pores (micropores), and in between the
micro-spheres a second family of intermediate pores with diameter
ROR
ROR
×
×
100 = nROH
(t) + 2 · n
100 [%, mol/mol]
ꢀ
(t) + nC
=
(t) + 2 · nROR
ꢀ
(t)
(1)
(2)
Mole of alcohol reacted to form linear ether
(t) =
Mole of alcohol reacted
ROR
ROH
S
2
· nROR(t)
×
×
100 = 2
100 [%, mol/mol]
· nROR(t) + nC (t) + 2 · nROR
=
ꢀ
(t)
=
ROH
nC
· nROR(t) + nC
=
(t)
(t) + 2 · nROR
C
S
S
(t) = 2
× 100 [%, mol/mol]
(t)
=
ꢀ
(3)
ꢀ
2 · n
ꢀ
(t)
ROR
ROH
ROR
(t) = 2
× 100 [%, mol/mol]
ꢀ
(t)
· nROR(t) + nC
=
(t) + 2 · nROR
(4)
8
–20 nm (mesopores) is observed. A third family of large pores with
diameter 30–80 nm is located between the agglomerates (macro-
pores). Macropores are permanent and can be detected by standard
techniques of pore analysis, i.e. adsorption–desorption of N2 at
77 K. Meso- and micropores are non-permanent, and appear if
resin swells by the interaction with the reaction medium. SEM
photo-micrographs (Hitachi H-2300) of Amberlyst 15 and CT224
are shown in Figs. 1 and 2, respectively, under magnification of 50
and 6000. Commercial beads are nearly spherical (Figs. 1A and 2A).
Mole of alcohol reacted to form linear ether
ROR
ROH
Y
(t) =
0
n
ROH
ROR
ROH
S
(t) · XROH(t)
[%, mol/mol]
×
100 =
(5)
1
00
Initial reaction rates of symmetrical ether (ROR) formation and
alcohol (ROH) consumption were computed from the functions of

1
32
C. Casas et al. / Applied Catalysis A: General 396 (2011) 129–139
Table 1
Characteristics and short names of tested catalysts.
Structurea
Acid capacity (meq H /g)
+
b
Sulfonation typec
dp (mm)
DVB%
Catalyst
Short name
d
Amberlyst 15
Amberlyst 35
CT175
CT275
CT276
A-15
A-35
CT-175
CT-275
CT-276
Macro
Macro
Macro
Macro
Macro
4.81
5.32
4.98
5.2
C
O
C
O
O
0.74
0.66
0.77
0.62
20
20
High
High
High
d
e
d
e
5.2
0.76
CT252
Amberlyst 16
Amberlyst 36
CT-252
A-16
A-36
Macro
Macro
Macro
5.4
4.8
5.4
O
C
O
0.78e
Medium
12 [23]
12 [24]
e
0.6-0.8
e
0.63
Amberlyst 39
Amberlyst 70
A-39
A-70
Macro
Macro
5.0
3.01
C
C
0.71e
0.57
8 [25]
8
d
d
CT224
CT-224
A-31
A-121
Micro
Micro
Micro
Micro
5.34
4.8
4.8
O
C
C
C
0.32
0.62
0.77
0.57
4
d
Amberlyst 31
Amberlyst 121
Dowex 50Wx4-50
4 [24]
2 [26]
4
e
e
DOW5050
4.95
Amberlyst 46
Nafion NR50
A-46
NR50
Macro
Micro
0.43
0.885
S
–
High
–
2.4e
a
Macroporous structure (Macro) or microporous structure (Micro).
Titration against standard base [27].
Conventionally sulfonated (C), oversulfonated (O) and sulfonated only at the polymer surface (S).
By sieving at atmospheric humidity or from SEM microphotographs.
Manufacturer data.
b
c
d
e
Fig. 1. SEM photomicrographs of Amberlyst 15. (A) General view of beads (magnification 50). (B) Surface detail of a bead (magnification 6000).
A surface view of Amberlyst 15 shows aggregates of microspheres
of about 0.5 ␮m each (Fig. 1B) in agreement with literature [29]. On
the contrary, CT224 surface is smooth and aggregates are not seen.
Amberlyst 15 morphology is illustrative of dry macroporous resins
of high and medium crosslinking degree, and CT224 of gel-type
ones.
Table 2 shows skeletal density and textural parameters esti-
mated from N -adsorption data at 77 K (Tristar 3000). N2
2
Fig. 2. SEM photomicrographs of CT 224. (A) General view of beads (magnification 50). (B) Surface detail of a bead (magnification 6000).

C. Casas et al. / Applied Catalysis A: General 396 (2011) 129–139
133
Table 2
Textural parameters of tested resins in dry state and swollen in water.
Catalyst
ꢁas (g/cm3)
Dry state: adsortion–desorption of N2 at 77 Kb
Swollen in water (ISEC method)
“True pores”
SBET (m /g)c
2
Vpore (cm /g)
3
d
dporee (nm)
ꢀ (%)
Gel polymer
2
3
dporee (nm)
3
Sarea (m /g)
Vpore (cm /g)
Vsp (cm /g)
ꢀ (%)
A-15
A-35
1.41
1.54
1.50
1.51
1.49
1.49
1.40
1.57
1.42
1.52
1.42
1.43
1.43
1.43
1.14
2.04
42.0
28.9
28.0
21.8
23.5
22.4
1.69
21.0
0.09
0.02
0.95
0.10
0.02
0.01
57.4
0.35
0.33
0.21
0.30
0.18
0.21
0.221
0.01
31.8
23.6
42.9
32.9
40.8
44.4
29.7
27.0
17.6
31.7
24.5
31.0
21.0
23.7
24.8
1.8
18.3
0.00
0.00
0.08
0
157
194
157
183
176
132
149
147
181
176
0.63
0.63
0.82
0.91
0.66
0.49
0.38
0.33
0.36
0.36
16.1
13.0
20.9
19.9
15.1
14.9
10.3
9.10
7.90
8.10
0.82
0.74
1.00
0.81
0.77
0.98
1.25
1.00
1.45
1.40
1.81
1.93
3.26
1.92
0.16
51.5
52.3
63.5
61.3
53.0
54.5
56.2
52.1
61.0
62.5
61.2
63.7
78.5
63.5
CT-175
CT-275
CT-276
CT-252
A-16
A-36
A-39
A-70
CT-224
A-31
0.14
2.9 × 10−
4
0.01
3.00
15.3
32.9
−
4
4
3.3 × 10
A-121
DOW5050
A-46
3.5 × 10−
0.00
0.00
23.0
0.04
0.26
19.2
3.20
186
0.48
10.30
0.00
2.0 × 10−
4
NR50
a
Skeletal density. Measured by helium displacement (Accupic 1330).
Samples dried at vacuum (10 mmHg, 110 C).
BET method.
Volume of N2 adsorbed relative pressure (P/P0) = 0.99.
dpore = 4Vpore/SBET or dpore = 4Vpore/Sarea, respectively.
b
c
−4
◦
d
e
particles. Pores have quite uniform size and shape. BJH pore size
analysis from desorption curve shows a unimodal distribution in
the macroporous region with a maximum around 33 nm (Fig. 4).
For Amberlyst 15 this value agree with those reported by Umar
et al. [32] and Sundmacher and Hoffmann [33]. On the contrary,
microporous resins, for example CT224, show type I isotherm but
as seen the whole pore volume is very small.
It is well known that ion-exchange resins swell in polar media.
As a result, morphology changes and non-permanent pores appear.
A useful description of nature and characteristics of these pores
can be obtained by Inverse Steric Exclusion Chromatography (ISEC)
[
34]. In macroporous structures a part of new open spaces (meso-
pores, between aggregates) can be characterized by the cylindrical
pore model. However, this model is not applicable to describe
spaces between polymer chains formed by aggregates and nod-
Fig. 3. Nitrogen adsorption isotherms at 77 K of CT-276 ( ), A-15 ( ) and CT-224
).
(
adsorption isotherms of macroporous resins are of type II, typi-
cal of macroporous solids [30], as Fig. 3 shows for Amberlyst 15
and CT276 (high croslinking degree). Despite N2 adsorption data
on ion-exchange resins are scarce, those of Amberlyst 15 agree
well with literature (Table 3). Adsorption–desorption curves show
H1 hysteresis, typical of solids made by aggregates of spherical
Table 3
Volume of N2 adsorbed at 77 K for Amberlyst 15.
P/P0
N2 adsorbed (cm3 STP/g)
This work
Kunin et al. [31]
0
0
0
.82
.550
.15
27
16.5
10
25
16.7
10.8
Fig. 4. Pore size distribution from desorption N2 curve ( A-15,
CT-276), and
from ISEC data ( A-15, CT-276).

1
34
C. Casas et al. / Applied Catalysis A: General 396 (2011) 129–139
Fig. 5. ISEC pattern displayed by resins in water. (A) Polymer density distribution of microporous and low DVB% macroporous resins. (B) Id. For medium and high DVB%
macroporous resins.
ules swelling. A good view of the three-dimensional network of
swollen polymer is given by the geometrical model developed by
Ogston [35], in which micropores are described by spaces between
randomly oriented rigid rods. The characteristic parameter of this
model is the specific volume of the swollen polymer (volume of
the free space plus that occupied by the skeleton), Vsp. The Ogston
model also allows to distinguish zones of swollen gel phase of dif-
ferent density or polymer chain concentration (total rod length per
and alcohol are higher than in air as a consequence of swelling. It
is noteworthy that swelling in alcohol and water are comparable.
Swelling by water is due to interaction with sulfonic groups, but
in the case of higher alcohols an additional interaction of aliphatic
part of the alcohol molecule with polymer chains is possible. Break-
age of some resins in the swelling process from dry particles was
observed hindering to obtain reliable measurements.
−2
volume unit of swollen polymer, nm ). According to this model,
density of polymer chains is described as the total rod length per
unit of volume.
3.2. Conversion of 1-octanol, selectivity to DNOE and DNOE yield
at 423 K
As shown in Table 2, macroporous resins with high and medium
crosslinking degree have permanent macropores in dry state. It
is to be noted that Amberlyst 16, 39 and 70 resins with DVB%
close to 8 show a very low BET surface area. However, compar-
ing with ISEC data, it is clear that those new pores are open in
It is reported that 1-pentanol dehydration to DNPE and 1-
hexanol dehydration to DNHE takes place free of the influence of
internal and external mass transfer at 423 K, 500 rpm and 1 g of dry
resin with the commercial distribution of particle sizes [16,17]. As
a result, the experiments of DNOE synthesis were carried out under
the same working conditions. In all the experiments mass balance
was accomplished within 5%.
2
aqueous media: surface area ranges between 132 and 194 m /g,
3
and pore volume between 0.36 and 0.91 cm /g. Both parameters
are far higher than dry state ones showing that new intermediate
pores (mesopores) are accessible. Correspondingly pore diameter
in aqueous medium is lower than in dry state. Fig. 4 shows the pore
distribution originated by swelling in CT276 and Amberlyst 15. In
microporous resins no mesopores originate from the swelling by
water or alcohol.
Fig. 5A and B shows the pore distribution in the swollen gel
phase, quite representative of the morphology of swollen acidic
resins in aqueous alcohol solution. Vsp decreases as DVB% increases
both in gel-type and macroporous resins. It is seen that swollen
gel phase of microporous and low-crosslinked macroporous resins
−2
show chains concentrations of 0.4–0.8 nm , typical of a little dense
polymer mass; whereas medium and high crosslinked macrop-
−2
orous resins show densities of 1.5 and 2 nm , respectively, typical
of high dense polymer mass. In polymer zones with chain density
−
2
of 0.4 nm spaces in the swollen gel phase are equivalent to pores
−2
of 2.9 nm; 1.5 nm in density zones of 0.8 nm , and ≤1 nm for chain
−2
densities higher than 1.5 nm [36]. It is to be noted that spaces in
swollen gel phase of macroporous resins of DVB% are comparable
to those of a large pore zeolite.
In Table 4, resins particle size in air, water, 1-pentanol, 1-
hexanol and 1-octanol measured by a laser technique (Beckman
Coulter LS Particle size Analyzer) are shown. Particle sizes in air are
similar to those measured experimentally (Table 1). Size in water
Fig. 6. Variation of reaction medium composition with time. 1-octanol dehydration
to DNOE over 1 g dry A-39 at T = 423 K, P = 2.0 MPa, 70 cm 1-octanol ( 1-octanol,
3
ꢀ
C16 branched ethers, DNOE,
C8 olefins, water).

C. Casas et al. / Applied Catalysis A: General 396 (2011) 129–139
135
Table 4
Particle diameter of some resins swollen in different liquid media.
Resin
dp (mm)
Air
Water
1-Pentanol
1-Hexanol
0.703
1-Octanol
A-15
A-35
CT-275
CT-276
CT-252
A-16
A-36
A-39
A-70
CT-224
A-121
DOW5050
A-46
0.65
N.A.
0.61
0.729
0.60
0.563
0.571
0.54
0.551
0.342
0.441
0.599
0.776
0.741a
N.A.
0.820
0.794
0.759
0.733
0.633
a
0.750
0.625a
a
0.652
0.684a
0.690
0.729
0.768
0.783
0.515
0.824
0.0,83
0.814
0.678
0.532
0.544
0.758
0.586
0.495
0.774
a
0.678
0.634
0.733
a
a
0.585
0.756
0.565
0.503
0.738
0.743
0.655
0.573
0.477
0.745
0.607
0.802
0.808
a
Particle breakage is observed.
Fig. 6 shows a typical plot of DNOE and byproducts mole evo-
lution on Amberlyst 39. Reaction proceeds smoothly from the
beginning, DNOE being the main product. Byproducts appeared as
soon as the reaction begins, and their amount increased continu-
ously through the experiment. Detected byproducts were C olefins
8
(
1-octene, cis/trans-2-octene, cis/trans-3-octene, and 4-octene),
and C16 branched ethers; octenes being by far the main byproducts,
particularly trans-2-octene and trans-3-octene. It is to be noted that
the amountof water in the liquidphase was lower than the stoichio-
metric one formed from dehydration to DNOE and olefins. A part
of the water remains in the resin and contributes to swell the cata-
lyst according to data of Table 4. The evolution of reaction medium
was similar over all tested resins. 1-Octanol conversion increased
constantly with time, as expected, whereas selectivity to DNOE
decreased slightly from the beginning and then stabilized (Fig. 7);
selectivity to C8 olefins and C16 branched ethers (i.e. 1,2-oxybis
octane) shows the opposite behavior. Since C8 alcohols other than
1
-octanol were not detected, it is assumed that branched ethers
come from C8 olefins and 1-octanol. The scheme reaction is shown
in Fig. 8.
DNOE
1
-octanol conversion (X
), selectivity to DNOE (S
), C8
OcOH
Fig. 7. Variation of 1-octanol conversion and selectivity to DNOE with time over
A-39 (W = 1 g, T = 423 K, P = 2.0 MPa, 70 cc 1-octanol) [ octanol conversion, selec-
tivity to DNOE with respect to 1-octanol].
OcOH
=
C
ꢀ
8
) and C16 branched ethers (S OO cc OO HO c ), and yield of DNOE
olefins (S
OcOH
DNOE
OcOH
(
Y
) at t = 6 h are shown in Table 5. In general, XOcOH increases
with acid capacity of resins as it is seen by comparing Amberlyst
+
+
4
6 and Nafion NR50 (<1 meq H /g), Amberlyst 70 (≈3 meq H /g)
macroporous resins other than Amberlyst 46 the reaction takes
place essentially in the gel phase. In microporous resins (which
+
and the others (4.8–5.4 meq H /g). X
on Amberlyst 15 is about
OcOH
four times that of Amberlyst 46. Amberlyst 15 has only about a 5%
have a very small fraction of –SO H groups at the resin sur-
3
+
of –SO H groups at the polymer surface (≈0.25 meq H /g) [37,38].
face) reaction proceeds almost exclusively in the swollen polymer.
Therefore, swelling is a key factor to explain the catalytic activ-
ity of resins. However, polymer structure plays an important role
3
Since this quantity is a half of the acid capacity of Amberlyst 46
(
sulfonated only at the polymer surface) it can be inferred that in
Fig. 8. Reaction scheme for dehydration of 1-octanol to DNOE.

1
36
C. Casas et al. / Applied Catalysis A: General 396 (2011) 129–139
Table 5
Alcohol conversion (%), selectivity to ether (%), olefin (%) and branched ethers (%) with respect of alcohol, and yield of symmetrical ether (%) for 1-octanol, 1-hexanol and
-pentanol dehydration to ether at 6 h of reaction.
1
=
=
=
C
C
C
ꢀ
ꢀ
ꢀ
DNOE
OcOH
DNHE
HeOH
DNPE
PeOH
8
6
5
OcOOc
OcOH
HeOHe
HeOH
PeOPe
PeOH
DNOE
OcOH
DNHE
HeOH
DNPE
Y
PeOH
Resin
XOcOH
XHeOH
XPeOH
S
S
S
S
S
S
S
S
S
Y
Y
OcOH
HeOH
PeOH
A-15
A-35
19.1
22.5
22.2
20.8
23.0
23.0
25.0
25.8
29.5
16.1
26.6
28.2
27.7
27.8
5.6
15.7
20.2
17.1
20.8
58.3
57.0
61.9
59.0
61.1
69.6
76.6
71.8
90.6
98.1
85.7
93.8
97.3
96.1
95.4
69.8
64.7
81.2
73.3
28.3
29.1
25.4
27.7
25.8
19.0
13.5
17.0
5.0
17.9
24.0
11.7
16.6
13.4
14.0
12.7
13.3
13.1
11.4
9.9
11.2
4.4
0
6.7
3.5
1.8
2.1
0
8.6
11.3
7.1
10.0
11.2
12.8
13.7
12.2
14.1
16.0
19.1
18.5
26.7
15.8
22.8
26.5
27.0
26.7
5.3
10.9
13.1
12.7
13.9
CT-175
CT-275
CT-276
CT-252
A-16
A-36
A-39
A-70
CT-224
A-31
21.4
25.5
21.8
16.5
23.9
22.5
22.1
21.6
4.3
20.4
25.1
20.0
13.7
20.9
20.6
19.9
19.4
4.0
85.5
79.0
94.0
97.7
96.5
97.0
98.0
96.6
94.8
97.9
91.5
85.5
96.8
97.8
97.3
98.0
99.2
97.3
98.4
99.0
9.6
13.5
3.9
5.2
7.6
1.9
4.9
7.5
2.1
3.3
6.8
1.3
1.3
1.6
18.3
20.1
20.5
16.1
23.1
21.8
21.7
20.9
4.1
17.0
19.9
17.7
12.0
18.5
18.4
18.1
17.1
3.6
1.91
0.17
0.94
0.77
7.6
2.7
0.89
1.8
4.6
0
1.9
1.7
0.67
2.0
5.2
1.5
1.2
1.3
0.68
1.2
1.6
0
1.7
1.3
0.89
1.2
0
0.67
A-121
DOW5050
A-46
0.07
1.5
0
NR50
12.1
10.3
9.7
100.0
0
0.64
1.0
12.1
10.0
8.7
Reaction conditions: W = 1 g, T = 423 K, P = 2.0 MPa, 70 cm3 alcohol.
as well. By comparing resins of similar structure but different acid
capacity, e.g. Amberlyst 15 and 35 or CT175 and 276 (macroporous,
high DVB %), and Amberlyst 16 and 36 (macroporous, medium DVB
sulfonated resins Amberlyst 35 and 36 (high and medium DVB %).
In addition, by comparing Amberlyst 15 and 35; Amberlyst 16 and
36; or CT175 with CT275 and CT276 it is seen that oversulfonation
DNOE
%), it is seen that XOcOH on oversulfonated resins are higher than
does not upgrade SOcOH .
on the conventionally sulfonated homologues. However, conver-
sion on CT 275 is lower than on CT175, probably because swollen
CT275, unlike CT276, does not have spaces in the density zone
Higher DNOE yields were obtained on gel-type resins Amberlyst
121, Dowex 50Wx4-50 and Amberlyst 31, and low-crosslinked
macroporous resin Amberlyst 39. It is seen in Table 5 that on these
−
2
DNOE
of 1.5 nm
lar acid capacity but different DVB% e.g. Amberlyst15, 16 and 39
high, medium and low DVB%, respectively); Amberlyst 35 and 36,
(Fig. 5B). Moreover, by comparing resins with simi-
resins XOcOH is high and SOcOH ranges between 90 and 98%.
A look to morphology of swollen gel polymer might elucidate the
(
relationship between resin structure and catalytic activity. Tested
−
2
or CT275 and CT276 (high DVB%), it is seen that XOcOH is higher on
resins with less crosslinking degree. As seen in Table 5, the highest
XOcOH was achieved on the Amberlyst 39 (macroporous, low DVB
resins have gel phase densities between 0.4 and 2.0 nm (Fig. 5A
and B), and spaces among polymer chains range from 2.6 to less
−
2
than 1 nm [36]. Polymer zones of density 0.4 nm are accessible
to 1-octanol and the bulky DNOE, and there is no restrictions for the
reaction intermediate. On the contrary, alcohol permeation, the exit
of DNOE or the formation of the reaction intermediate are hindered
%); high XOcOH values were also found on gel-type resins Amberlyst
3
1 and 121, and Dowex 50Wx4-50.
DNOE
As a rule, S
increases with the volume of swollen gel phase.
OcOH
−
2
NR50 shows the highest selectivity to DNOE, followed by Amberlyst
in polymer zones of density 2.0 nm . Resins have to swell by 1-
octanol and water released for the reaction takes place with high
DNOE yield. In this way, microporous resins showing at the same
7
0 (macroporous, low DVB %) and gel-type resins Amberlyst 121
DNOE
and Dowex 50Wx4-50. Macroporous resins clearly show that SOcOH
decreases on increasing the DVB % (and therefore the resin stiffness)
as seen by comparing conventionally sulfonated resins Amberlyst
−
2
time high Vsp values and swollen polymer densities of 0.4–0.8 nm
(spaces of 2.9–1.5 nm) give the best DNOE yields. Gel-type resins
greatly swell in octanol-water mixtures so that a high portion of
1
5, 16 and 39 (high, medium and low DVB %, respectively), and over-
–
SO H groups are able to take part in the reaction. As a result, X
3
O
c
O
H
values are high. As for selectivity, the wide spaces between polymer
chains favor the formation and diffusion of DNOE
DNOE
Since the reaction mainly occurs in the gel phase, SOcOH is closely
related to the gel phase density of swollen resins. Amberlyst 121,
Dowex 50Wx4-50 and Amberlyst 70, with gel phase density of
−
2
0
3
9
.4 nm show selectivities to ether from 98 to 96%. In Amberlyst
−
2
DNOE
OcOH
1 and 39, with polymer density of 0.8 nm , S
4 and 90.6%. Amberlyst 16 and 36, and CT252 (macroporous,
is between
−
2
medium DVB %) have polymer density of 1.5 nm and show selec-
tivities ranging from 60 to 77. Finally, Amberlyst 15 and 35, CT175,
CT275 and CT276 (macroporous, high DVB %), with a density of
−
2
2
.0 nm , show selectivities to DNOE lower than 61%. The low
selectivity of CT224, as compared with the other gel-type resins,
of about 86% can be explained by the significant fraction of poly-
−
2
mer densities 1.5–2.0 nm . Less dense swollen resins, with wide
spaces between polymer chains, allow permeation of 1-octanol,
outside diffusion of DNOE, and formation of reaction intermedi-
ate is not restricted. Therefore, dehydration to DNOE is favored on
gel-type and macroporous resins with low DVB %, whereas in the
less swollen macroporous resins CT276, CT275, CT175, Amberlyst
15 or 35, selectivity to olefin and branched ethers increases sig-
nificantly. Finally, it seems that supplementary -SO3H groups in
Fig. 9. Selectivity to DNOE ( ), C8 olefins ( ), and C16 branched ethers ( ) with
+
3
respect 1-octanol vs. [H ]/Vsp ratio at t = 6 h (W = 1 g, T = 423 K, P = 2.0 MPa, 70 cm
-octanol).
1

C. Casas et al. / Applied Catalysis A: General 396 (2011) 129–139
137
oversulfonated resins contribute to hold closer the polymer chains
by the presence of additional hydrogen bonds, which would explain
selectivities to DNOE lower than those of conventionally sulfonated
homologues.
As Fig. 9 shows, S OD cN OO HE correlates well with [H ]/Vsp. Acid sites
per volume unit of swollen gel phase would be an excellent param-
eter to explain DNOE formation. Selective resins to DNOE have
+
+
low [H ]/Vsp values, whereas less selective CT276, CT275, CT175,
+
Amberlyst 15 and 35, show high [H ]/Vsp values. Accordingly, selec-
tivity to olefins and branched ethers is high, what suggest that
DNOE diffusion is hindered, and ether might decompose to 1-
DNOE
octanol and octenes. In this way, they show low XOcOH and SOcOH
values.
0
As seen in Table 6, initial DNOE formation (r
) and 1-octanol
DNOE
0
consumption (−r
) rates fit quite well Eq. (10), which is a proof
values are higher on gel-type resins Dowex
OcOH
0
of data reliability. r
DNOE
5
0Wx4-50, Amberlyst 31, and Amberlyst 121, in line with the high
yield of DNOE observed. As for 1-octanol consumption, Amberlyst
0
3
6 and 39 show higher −r
values mainly due to significant
OcOH
16
Fig. 10. Comparison of 1-pentanol (
versions on tested resins at t = 6 h (W = 1 g, T = 423 K, P = 2.0 MPa, 70 cm alcohol).
), 1-hexanol (
) and 1-octanol (
) con-
formation of C olefins and C branched ethers.
3
8
As for TOF, NR50 shows the highest initial value. It is explained
by the higher acid strength of –SO H groups, due to the presence
of fluorine in Nafion chains, as compared to those of PS-DVB resins
3
dehydrations. 1-pentanol and 1-hexanol dehydrations show a sim-
ilar pattern to that of 1-octanol, and the variation of 1-pentanol
[
16]. Amberlyst 46, sulfonated only at the polymer surface, should
(
(
X
PeOH) and 1-hexanol (XHeOH) conversion, and selectivity to DNPE
be outlined. Its initial TOFDNOE is clearly higher than those of the
other PS-DVB resins. This fact disclose that in swollen PS-DVB resins
DNPE DNHE
S
) and DNHE (SHeOH )with time is analogue to that found in
PeOH
DNOE synthesis.
there is a fraction of –SO H groups, located in very dense polymer
3
In this way, XPeOH and XHeOH were higher on increasing resins
acid capacity (Table 5), and alcohol conversions were higher on
oversulfonated resins than on conventionally sulfonated homo-
logues in both DNPE and DNHE synthesis (see Amberlyst 15 and
zones, unavailable for reaction. Seeing that acid sites in PS-DVB
resins have rather similar acid strength [39], it is inferred that TOF
values depend on the number of acid sites taking part in DNOE
formation. So, in the more active and selective to DNOE Dowex
3
5; or Amberlyst 16 and 36). Resin morphology is also an important
5
0Wx4-50, Amberlyst 31, 121 and 70 between 64 and 61% of sul-
factor to achieve high conversions. The same as in DNOE synthesis,
the highest XPeOH and XHeOH values were found on gel-type, macro-
porous resins of low crosslinking degree as well as Amberlyst-36.
fonic groups take part in the reaction, but in the less active and
selective Amberlyst 15 and 35, CT175, CT275 and CT276 only a
fraction of 32–27% participates.
In general, it is seen that XOcOH ≥ X
≥ X
for each resin.
As for initial –TOFOcOH, Nafion NR50 shows the highest value,
a little higher than that of Amberlyst 46, in agreement with their
different acid strength. PS-DVB resins show smaller values because
a fraction of active sites is inaccessible to 1-octanol. By comparing
with Amberlyst 46 TOF it is seen that accessible sites to alcohol
HeOH
PeOH
However, this fact could be misleading as for comparing the resin
activity in the three reaction systems. All experiments were per-
formed by using the same alcohol volume and catalysts mass.
Since the three alcohols have different molecular weight and den-
sity, the initial ratio alcohol moles (n R0 OH) to catalyst mass (W)
0
range from 73% (Amberlyst 39) to 54% (Amberlyst 35). −TOF
OcOH
values are greater than twice TOF0
greatly changes from 1-pentanol to 1-octanol. So, to compare cat-
ones, suggesting again that
DNOE
0
alysts activity suitably, the factor XROH · n
/W is used. As seen
in resins with very dense swollen gel phase is partially accessi-
ble to octanol but the formation and diffusion of DNOE is actually
hindered.
Unlike fluorine atoms in NR50, chlorine atoms in Amberlyst 70
chains do not upgrade acid strength. Chlorine atoms give Amberlyst
ROH
in Fig. 10 this factor decreases as alcohol is longer on macrop-
orous Amberlyst 15, Amberlyst 35, Amberlyst 16 and Amberlyst 36
(high and medium DVB %). On the contrary, on gel-type Amberlyst
3
1, 121, CT224 and Dowex 50Wx4-50 and on low-DVB% macrop-
orous Amberlyst 39 and 70 conversions are similar for the three
alcohols. In swollen state the first group has spaces in gel phase
less than 1 nm, where 1-octanol motion is more restricted than
that of 1-pentanol and 1-hexanol. The small differences observed
in gel-type resins highlight that spaces originated on swelling are
wide enough for the three alcohols to access to active sites. Fig. 11
compares alcohols conversion on CT224, Amberlyst 31 and Dowex
50Wx4-50. Despite the three resins have different acid capacity
the small differences in conversion factor can be explained by the
different morphology rather than by the different acid capacity.
Amberlyst 31 and Dowex 50Wx4-50 have gel phase densities of
7
0 additional thermal stability allowing at the same time the
swelling phenomenon in aqueous media. As a result, selectiv-
ity to DNOE is slightly higher than that of Amberlyst 39 (similar
crosslinking degree but conventionally sulfonated). Due to its ther-
mal stability it could be a suitable catalyst for industrial use.
3.3. Comparison of DNHE, DNPE and DNOE synthesis data
Synthesis of DNPE and DNHE from 1-pentanol and 1-hexanol,
respectively, was studied on resins of Table 1, but CT175, CT275,
CT276 and CT276. Despite DNOE yield was low on macroporous
resins with high or medium DVB%, Amberlyst 15, 35, 16 and 36
were used for comparison purposes. Data of DNPE synthesis on
Dowex50Wx4-50, CT224, Amberlyst 70 and 36, and NR50 are in
agreement with reported previously [16]. Main byproducts of DNPE
and DNHE syntheses were C5 and C olefins, respectively; however,
unlike 1-octanol dehydration to DNOE, traces of secondary, tertiary
and branched alcohols were detected in 1-pentanol and 1-hexanol
−
2
0.4 and 0.8 nm . On swelling, alcohol and ether motion is hin-
dered similarly in the three reaction systems, but in CT224 where
−
2
gel phase zones ofdensity1.5 nm predominates, XOcOH dropssug-
gesting that 1-octanol have more restricted motion in this resin
than 1-pentanol, 1-hexanol, and DNOE than DNPE and DNHE. In
oversulfonated resin CT224 additional acid centers (with regard
to Amberlyst 31 and Dowex 50Wx4-50) mainly contributes to
form new hydrogen bonds between chains; as a result accessibil-
6

1
38
C. Casas et al. / Applied Catalysis A: General 396 (2011) 129–139
Table 6
Initial rates (mol/h kgcat) and TOF (mol/h eqH ) values of 1-pentanol, 1-hexanol and 1-octanol consumption and DNPE, DNHE and DNOE formation.
+
0
0
0
0
0
0
0
0
0
TOF0
TOF0
TOF0
DNPE
Resin
−r
−r
−r
r
r
r
−TOF
−TOF
−TOF
OcOH
HeOH
PeOH
DNOE
DNHE
DNPE
OcOH
HeOH
PeOH
DNOE
DNHE
A-15
A-35
20.0 ± 1.6 21.7 ± 2.0 32.7 ± 2.1
20.0 ± 3.2 26.8 ± 6.7 49.7 ± 1.1
4.7 ± 0.1
5.1 ± 0.3
5.2 ± 0.6
4.7 ± 0.4
5.6 ± 0.5
6.1 ± 0.4
8.4 ± 0.4 12.4 ± 0.8 4.2 ± 0.3
7.9 ± 1.9 17.0 ± 0.9 3.8 ± 0.6
4.5 ± 0.4
5.0 ± 1.3
6.8 ± 0.4 1.0 ± 0.0 1.7 ± 0.1 2.6 ± 0.2
9.4 ± 0.2 1.0 ± 0.1 1.5 ± 0.4 3.2 ± 0.2
1.1 ± 0.1
CT-175
CT-275
CT-276
CT-252
A-16
A-36
A-39
A-70
CT-224
A-31
16.2 ± 0.8
3.1 ± 0.2
3.8 ± 0.1
0.9 ± 0.1
1.1 ± 0.1
20.7 ± 0.6
1.1 ± 0.1
18.5 ± 0.7 27.3 ± 3.2 36.1 ± 2.2
25.8 ± 2.0 40.1 ± 5.9 45.3 ± 6.4
25.5 ± 2.2 22.7 ± 0.4 31.2 ± 0.5
12.9 ± 0.4 19.6 ± 1.4 21.1 ± 1.7
19.5 ± 0.5 46.5 ± 0.8 32.2 ± 2.3
7.5 ± 1.0 12.7 ± 0.3 15.8 ± 0.6 3.9 ± 0.1
9.4 ± 0.7 17.8 ± 1.6 19.5 ± 3.8 4.8 ± 0.4
8.9 ± 1.1 10.8 ± 1.0 14.2 ± 0.2 5.1 ± 0.4
5.7 ± 0.7
7.4 ± 1.1
4.5 ± 0.1
6.5 ± 0.5
8.7 ± 0.1
4.7 ± 0.6
5.0 ± 0.2
4.6 ± 0.2
9.7 ± 0.7
7.5 ± 0.5 1.6 ± 0.2 2.7 ± 0.1 3.3 ± 0.1
8.4 ± 1.2 1.7 ± 0.1 3.3 ± 0.3 3.6 ± 0.7
6.2 ± 0.1 1.8 ± 0.2 2.2 ± 0.2 2.9 ± 0.0
7.0 ± 0.6 2.1 ± 0.1 3.6 ± 0.5 2.6 ± 0.2
6.0 ± 0.4 1.6 ± 0.1 2.3 ± 0.3 2.7 ± 0.2
6.2 ± 0.5 2.2 ± 0.3 2.4 ± 0.2 2.8 ± 0.2
5.4 ± 0.3 2.1 ± 0.1 2.3 ± 0.1 2.5 ± 0.2
5.4 ± 0.6 2.0 ± 0.2 2.2 ± 0.2 2.6 ± 0.3
9.9 ± 0.5 3.4 ± 0.4 4.0 ± 0.2 4.6 ± 0.2
6.3 ± 0.4 10.9 ± 1.4
7.8 ± 0.6 4.3 ± 0.1
8.5 ± 0.4 12.1 ± 1.6 14.2 ± 0.9 3.7 ± 0.1
21.3 ± 0.6 22.3 ± 3.0 29.6 ± 2.6 10.3 ± 1.2 11.3 ± 1.0 13.6 ± 0.7 4.4 ± 0.1
A-121
19.9 ± 1.9 24.0 ± 1.0 26.3 ± 1.2
9.9 ± 0.3 11.2 ± 0.7 11.9 ± 1.0 4.2 ± 0.4
DOW5050 19.7 ± 2.5 22.7 ± 0.9 26.6 ± 2.8 10.1 ± 1.1 10.8 ± 1.2 12.7 ± 1.4 4.0 ± 0.5
A-46
NR50
3.0 ± 0.4
4.2 ± 0.3
4.2 ± 0.2
1.5 ± 0.2
1.7 ± 0.1
4.7 ± 0.4
2.0 ± 0.1 7.0 ± 0.9
7.9 ± 0.7
9.7 ± 1.0 11.0 ± 0.5
4.2 ± 0.2
5.3 ± 0.8 8.9 ± 0.7 10.9 ± 1.1 12.4 ± 0.6 4.7 ± 0.3 5.3 ± 0.4 6.0 ± 1.0
Reaction conditions: W = 1 g, T = 423 K, P = 2.0 MPa, 70 cm3 alcohol.
DNPE
PeOH
DNHE
HeOH
DNOE
on each resin. Conversely, selec-
OcOH
found that S
> S
> S
tivity to olefins and branched ethers increase from DNPE synthesis
to that of DNOE (Table 5). This fact suggests that diffusion of alcohol
and linear ether, as well as the formation of reaction intermediate
are more hindered as alcohol is larger, and as a result when pro-
duced ether is bulkier. Diffusion of long-chain ethers in swollen
−
2
resins with zones of density ranging from 1.5 to 2 nm (Fig. 5B)
is hindered, and its decomposition to olefin and alcohol favored.
DNPE
DNHE
HeOH
DNOE
are small in
OcOH
As Fig. 12 shows, differences in SPeOH , S
and S
DNOE
microporous resins. The drop in SOcOH for CT224 is noticeable, and
similarly to XOcOH drop, it is explained because CT224 has zones
of higher chain density than the other gel-type resins. Swelling
depends on DVB% of resins so that crosslinking degree is indicative
of their ability for linear ether formation. Fig. 13 compares selec-
tivity to linear ether of four resins. It is observed that selectivity
to ether decreases on increasing crosslinking degree. Moreover, in
DNPE
PeOH
DNOE
OcOH
0
stiffer resins the difference between S
and S
is higher.
Initial 1-pentanol (−r0
) and 1-hexanol (−r
) consump-
Fig. 11. Influence of alcohol length and sulfonation type on alcohol conversion at
PeOH
HeOH
3
tion rates and DNPE (r0
) and DNHE (r0
t = 6 h (W = 1 g, T = 423 K, P = 2.0 MPa, 70 cm alcohol) [ 5.34 meq/g,
4.95 meq/g,
) formation rates fit
DNPE
DNHE
4
.8 meq/g].
0
well Eq. (10). As seen in Table 6 the higher r
values are found
DNPE
on Amberlyst 39 (macroporous; low DVB %), Amberlyst 16 and
3
6 (medium DVB %), and Amberlyst 35 (high crosslinking degree).
ity to gel phase of 1-pentanol or 1-hexanol is higher than that of
0
In the case of DNHE, higher r
resin CT224 and macroporous Amberlyst 16 and 36 (medium DVB
values are shown by gel-type
1
-octanol.
DNHE
As for selectivity to linear ether, Fig. 12 shows that gel-type
resins have high selectivity ranging from 95 to 99%. In general, it is
Fig. 13. Influence of crosslinking degree on the selectivity to DNPE, DNHE and DNOE
Fig. 12. Selectivity to DNPE (
t = 6 h (W = 1 g, T = 423 K, P = 2.0 MPa, 70 cm alcohol).
), DNHE (
), and DNOE (
) on tested resins at
(W = 1 g, T = 423 K, P = 2.0 MPa, 70 cc alcohol, t = 6 h) [ 20 DVB%,
DVB%, 4 DVB%].
12 DVB%,
8
3

C. Casas et al. / Applied Catalysis A: General 396 (2011) 129–139
139
). For each resin, it is clearly seen that r0
< r
0
< r
0
on
%
tion is hindered in less swollen polymers. Linear ether formation
is influenced by resin swelling and it is generally observed that
DNPE
DNHE
DNOE
macroporous resins with high and medium crosslinking degree
Amberlyst 15, Amberlyst 35, Amberlyst 16 and Amberlyst 36). On
0
0
0
(
r
> r
> rDNOE.
DNPE
DNHE
gel-type resins (Amberlyst 31, Amberlyst 121, Dowex 50Wx4-50,
CT224) and macroporous of low crosslinking degree (Amberlyst 39,
Amberlyst 70) differences are smaller. Bulkier ethers production is
more hindered by the less swollen macroporous resins in agree-
ment with initial rate data. The less dense gel phase of swollen
microporous resins even the rates of DNPE, DNHE and DNOE syn-
theses, as they are able to accommodate better bulky molecules.
Acknowledgements
Financial support was provided by the Science and Education
Ministry of Spain (project CTQ2007-60691/PPQ). The authors thank
Rohm and Haas France and Purolite for providing Amberlyst and
CT ion exchange resins, respectively. We thank also Dr. Karel Jera-
bek of Institute of Chemical Process Fundamentals (Prague, Czech
Republic) for the structural and textural analysis made by the ISEC
method. Finally, support of Porosimetry, SEM and Cytometry Ser-
vices of the University of Barcelona is kindly acknowledged
Once more, the slightly denser structure of CT224 is highlighted
0
ROR
due to higher r
differences than the other gel-type resins.
0
ROR
As observed in DNOE synthesis, NR50 shows the highest TOF
values, and Amberlyst 46 achieves the highest values on PS-DVB
resins. Since in Amberlyst 46 (sulfonated only at the polymer sur-
face) all acid sites are accessible to alcohol, from the fact that
References
PS-DVB resins have lower TOF0 values than Amberlyst 46 it can
ROR
[
[
[
[
1] IEA Implementing Agreement Advanced Motor Fuels. Annex XXVI. Alcohols
and Ethers as Oxygenates in Diesel fuel.
2] J.I. Ciria, Technical Article: Estudio de la combustión en Motores Diesel, 2003,
www.wearcheckiberica.es.
3] G.C. Pecci, M.G. Clerici, F. Giavazzi, F. Ancillotti, M. Marchiona, R. Patrini, IX Int.
Symp. Alcohol Fuels 1 (1991) 321–335.
4] G.A. Olah, US Patent US5520710A (1996).
be assumed that only a fraction of acid centers take part in the
reaction. Taking as a reference initial TOF of DNPE and DNHE forma-
tion on Amberlyst 46, it can be drawn that in macroporous resins
Amberlyst 15, Amberlyst 35, Amberlyst 16 and Amberlyst 36 the
fraction of active centers taking part in DNPE synthesis ranges from
5
2
6 to 79%; it is between 44 and 67% in DNHE synthesis and between
9 and 52% in the DNOE synthesis. It is also seen that the fraction
[5] Fischer-Tropsch Archive, www.Fischer-Tropsch.org.
[
[
6] D.P. Krinkin, D.M. Rudkovskii, A.G. Trifel, Chem. Technol. Fuels Oils 1 (7) (1965)
11–514.
7] T. Tsichida, S. Sakuma, US Patent Appl. 20070255079 (2007).
5
of active sites taking part in the reaction decreases in the order
DNPE > DNHE > DNOE syntheses (i.e. a 56% of Amberlyst 15 active
sites take part in DNPE synthesis, a 44% in DNHE one and a 29%
in that of DNOE). In relation to high-selective resins (Amberlyst
[8] J.F. Norris, G.W. Rigby, J. Am. Chem. Soc. 54 (5) (1932) 2088–2100.
9] A. Corma, H. García, Top. Catal. 48 (2008) 8–31.
10] P.F. Siril, H.E. Gross, D.R. Brown, J. Mol. Catal. A: Chem. 279 (2008) 63–68.
11] Rohm Haas Website, www.rohmhaas.com/products/industrial/catalysis
(03/2009).
[
[
[
&
7
0, Amberlyst 39, Amberlyst 121, Amberlyst 31, and Dowex50x4)
[
[
12] Dowex Website, www.dow.com (03/2009).
13] R. Bringué, J. Tejero, M. Iborra, J.F. Izquierdo, C. Fité, F. Cunill, Top. Catal. 45
the fraction of active sites involved in ether synthesis is nearly the
same in the three reaction systems (54–61%), i.e. 54% of acid sites
of Amberlyst 121 are accessible for DNPE formation, 58% for DNHE
formation and 61% for that of DNOE. In the case of CT224, it is seen
again that the fraction of sites taking part in DNOE synthesis is lower
(
2007) 181–186.
[14] X. Yin, B. Yang, S. Goto, Int. J. Chem. Kinet. 27 (1995) 1065–1074.
[15] J. Tejero, F. Cunill, M. Iborra, J.F. Izquierdo, C. Fité., J. Mol. Catal. A: Chem.
1
82–183 (2002) 541–554.
16] R. Bringué, M. Iborra, J. Tejero, J.F. Izquierdo, F. Cunill, C. Fité, V.J. Cruz, J. Catal.
44 (2006) 33–42.
[17] E. Medina, R. Bringué, J. Tejero, M. Iborra, C. Fité, Appl. Catal. A 374 (2010) 41–47.
[
(
47%) than in DNPE and DNHE ones (58%). Similar trends are found
2
in alcohol consumption.
[
[
18] R.J.J. Nel, A. de Klerk, Ind. Eng. Chem. Res. 48 (2009) 5230–5238.
19] N.P. Makgoba, T.M. Sakuneka, J.G. Koortzen, C. Van Schalkwyk, J.M. Botha, C.P.
Nicolaides, Appl. Catal. A 297 (2006) 145–150.
4
. Conclusions
[
20] I. Hoek, T.A. Nijhuis, A.I. Stankiewicz, J.A. Moulijn, Appl. Catal. A 266 (2004)
1
09–116.
The dehydration of 1-octanol to DNOE at 423 K and 2.0 MPa takes
[
[
21] G.A. Olah, T. Shamma, G.K.S. Prakash, Catal. Lett. 46 (1997) 1–4.
22] O.N. Karpov, R.M. Bystrova, L.G. Fedosyuk, Zh. Prikl. Khim. 40 (1) (1967)
219–223.
23] P.K. Kiviranta-Pääkkönen, L.K. Struckmann (née Rihko), J.A. Linnekoski, A.O.I.
Krause, Ind. Eng. Chem. Res. 37 (1998) 18–24.
24] S.C.M. dos Reis, E.R. Lachter, R.S.V. Nascimento, J.A. Rodrigues Jr., M.G. Reid, J.
Am. Oil Chem. Soc. 82 (2005) 661–665.
25] M.L. Honkela, A. Root, M. Lindblad, A. Outi, I. Krause, Appl. Catal. A 295 (2005)
place successfully over ion exchange resins as well as 1-hexanol and
ROR
1
-pentanol dehydrations. High YROH values are found on resins with
[
[
[
[
high acid capacity and Vsp values. Best resins for DNOE synthesis
are microporous resins with 2–4 DVB%. In general oversulfonated
resins do not upgrade the behavior of their conventionally sul-
fonated homologues
216–223.
By comparing the synthesis of DNPE, DNHE and DNOE it is found
26] Kapila et al., US Patent US2005/0137428A1 (2005).
[27] S. Fischer, R. Kunin, Anal. Chem. 27 (1955) 1191.
[28] A. Guyot, D.C. Sherrington, in: P. Hodge (Ed.), Syntheses and Separations Using
Polymers Supports, Wiley, Chichester, 1988, p. 15 (Chapter 1).
[29] M. Iborra, J. Tejero, J.F. Izquierdo, F. Cunill, Br. Polym. J. 23 (1990) 117–127.
[30] G. Leofanti, M. Padovan, G. Tozzola, B. Venturelli, Catal. Today 41 (1998)
0
0
as a rule for each resin that XPeOH · n
/W > XHeOH · n
/W >
PeOH
HeOH
0
XOcOH · n
/W, but differences are small in microporous resins. It
OcOH
DNPE
DNHE
DNOE
OcOH
is also observed that S
> S
> S
on macroporous resins
PeOH
HeOH
but differences in selectivity lessens on decreasing the crosslinking
degree. High selectivity to symmetrical linear ether was found on
microporous resins for the three ethers. Swelling is a key factor
to achieve high selectivity to linear ether, and a clear relationship
207–219.
[
31] R. Kunin, E.F. Meitzner, J.A. Oline, S.A. Fischer, N. Frish, Ind. Eng. Chem. Prod.
Res. Dev. 1 (1962) 140–144.
[32] M. Umar, D. Patel, B. Saha, Chem. Eng. Sci. 64 (2009) 4424–4432.
[33] K. Sundmacher, U. Hoffmann, Chem. Eng. Sci. 49 (1994) 3077–3089.
34] K. Jerabek, ACS Symp. Ser. 635 (1996) 211.
35] A.G. Ogston, Trans. Faraday Soc. 54 (1958) 1754.
[36] K. Jerabek, L. Holub, 13th International IUPAC Conference on Polymers and
Organic Chemistry, Montreal, 2009.
37] S. Ihm, M. Chung, K. Park, Ind. Eng. Chem. Res. 27 (1988) 41–45.
38] R.B. Diemer Jr., K.M. Dooley, B.C. Gates, R.L. Albright, J. Catal. 74 (1982) 373–381.
[39] S. Koujout, B.M. Kiernan, D.R. Brown, H.G.M. Edwards, J.A. Dale, S. Plant, Catal.
Lett. 85 (2003) 33–40.
+
between the parameter H /Vsp and selectivity is observed.
[
[
0
NR50 showed the highest TOF values due to its higher acid
0
strength. TOF values of Amberlyst 46, a special resin with all sul-
fonic groups at the polymer surface, and so wholly accessible to
alcohol, were the highest between PS-DVB resins. Lower TOF0 val-
ues shown by the other PS-DVB resins reveal that only a fraction
of active centers is available for reaction, so that ether forma-
[
[