Improved 1-butene/isobutane alkylation with acidic ionic liquids and tunable acid/ionic liquid mixtures

Article abstract of DOI:10.1016/j.jcat.2009.09.022

A new process for 1-butene/isobutane alkylation to yield C ₈-alkylates is described using binary mixtures of certain acidic imidazolium ionic liquids (ILs) and strong acids such as sulfuric or trifluoromethanesulfonic (triflic) acid. Equivalent or better conversion (>95%), C₈-alkylates selectivity (>70%) and trimethylpentane/ dimethylhexane selectivity (TMP/DMH > 7) were achieved with the IL/acid mixtures over the pure acids themselves. Six types of substituted 3-methyl-imidazolium ionic liquids were investigated, wherein acidity is imparted via either the cation with sulfonic acid groups or the anion (hydrogen sulfate) or both. Long-term studies up to 30+ recycles indicate that the catalyst stability was increased by sometimes greater than 30+% with the IL/acid mixtures over the pure acid. The ionic liquid is believed to tune the acidity, solubility, and interfacial properties, resulting in these enhanced results. In addition, this concept could also be applicable to Friedel-Crafts alkylation, acylation chemistries, or other acid-catalyzed reactions.

Full text of DOI:10.1016/j.jcat.2009.09.022

Journal of Catalysis 268 (2009) 243–250
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Improved 1-butene/isobutane alkylation with acidic ionic liquids and tunable
acid/ionic liquid mixtures
1
*
Shengwei Tang , Aaron M. Scurto, Bala Subramaniam
Center for Environmentally Beneficial Catalysis, Department of Chemical and Petroleum Engineering, University of Kansas, Lawrence, KS 66045, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 19 June 2009
Revised 11 September 2009
Accepted 26 September 2009
Available online 3 November 2009
A new process for 1-butene/isobutane alkylation to yield C8-alkylates is described using binary mixtures
of certain acidic imidazolium ionic liquids (ILs) and strong acids such as sulfuric or trifluoromethanesul-
fonic (triflic) acid. Equivalent or better conversion (>95%), C8-alkylates selectivity (>70%) and trimethyl-
pentane/dimethylhexane selectivity (TMP/DMH > 7) were achieved with the IL/acid mixtures over the
pure acids themselves. Six types of substituted 3-methyl-imidazolium ionic liquids were investigated,
wherein acidity is imparted via either the cation with sulfonic acid groups or the anion (hydrogen sulfate)
or both. Long-term studies up to 30+ recycles indicate that the catalyst stability was increased by some-
times greater than 30+% with the IL/acid mixtures over the pure acid. The ionic liquid is believed to tune
the acidity, solubility, and interfacial properties, resulting in these enhanced results. In addition, this con-
cept could also be applicable to Friedel–Crafts alkylation, acylation chemistries, or other acid-catalyzed
reactions.
Keywords:
Alkylation
1-Butene
Ionic liquid
Isobutane
Sulfuric acid
Trifluoromethanesulfonic acid
Ó 2009 Elsevier Inc. All rights reserved.
1. Introduction
Solid acids have shown promise as environmentally-benign
alternatives to the strong or mineral acids. However, solid acid cat-
Alkylation of isobutane with various olefins is a major process
in the petroleum industry for the production of gasoline. The alkyl-
ation of C3–C5 olefins with isobutane produces a mixture of highly
branched alkanes. This alkylate offers a high octane number, a low
Reid vapor pressure (RVP), and low octane sensitivity [difference
between research octane number (RON) and motor octane number
(MON)]. It contains no aromatics and nearly no sulfur. This alkylate
is usually blended with various cuts of refined crude oil. The world-
wide installed alkylation production capacity is approx. 102 mil-
lion tons/year [1], accounting for approximately 10% of the
world-wide gasoline usage.
alysts often deactivate rapidly due to fouling of the pores, resulting
in low product yield and loss of reaction selectivity [2]. The use of
supercritical reaction media was shown to provide an optimum
combination of solvent power and transport to mitigate foulant
accumulation in the catalyst pores resulting in extended catalyst
activity [3]. However, as shown by detailed economic analysis,
such efforts are not yet economically competitive with conven-
tional processes. The main challenge is to develop tailored solid
acid catalysts that are porous enough to accommodate a high acid
site density but with pore sizes that do not impose diffusion limi-
tations on the C8-alkylate products [4].
Current commercial alkylation processes are catalyzed primar-
ily by concentrated sulfuric acid or hydrofluoric acid (HF). The sul-
furic acid process produces large amounts of spent acid and acid
soluble oils. The spent acid is costly to regenerate. Anhydrous HF
is highly toxic and its leakage results in dangerously stable aerosols
at the ground level. In addition, equipment corrosion, transport,
handling of hazards, and environmental liability associated with
the disposal of spent acid are disadvantages for both processes.
Hence, efforts continue to develop alternative processes that are
relatively safe.
Recently, room temperature ionic liquids (RTILs) have been rec-
ognized as potentially green solvents for a wide variety of processes.
Ionic liquids are organic salts composed of relatively large organic
cations and inorganic or organic anions that are liquid near room
temperature (Tm < 100 °C). They have virtually no vapor pressure
and are molecularly tunable for desired properties. The combina-
tion of the cation and anion determines solubility, density, and vis-
cosity of the liquids. There are several major classes of ionic liquid
cations, and each offers the possibility to add any number of func-
tional groups. A change in the structure, functional group, or size
of the ions leads to different chemical and/or physical properties.
We have utilized ionic liquids as efficient medium for homoge-
neously catalyzed reactions using organometallic complexes [5–7].
Ionic liquids with functional groups providing Lewis or Brønsted
acidity have been synthesized for a variety of applications. The use
* Corresponding author. Fax: +1 785 864 6051.
E-mail address: bsubramaniam@ku.edu (B. Subramaniam).
Present address: College of Chemical Engineering, Sichuan University, Chengdu
1
610065, PR China.
0021-9517/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2009.09.022

244
S. Tang et al. / Journal of Catalysis 268 (2009) 243–250
of chloroaluminate-based ionic liquids has been reported in patents
[8–17] and the open literature [18–25]. Chauvin et al. [18] reported
the alkylation of isobutane with 2-butene using 1-butyl-3-methyl-
imidazolium chloride-aluminum chloride molten salts. The chloroa-
luminate anion serves as a Lewis acid catalyst. Chloroaluminate io-
nic liquids have also been investigated as catalysts for isobutane/
butene alkylation by Yoo et al. [19,20], Xu and coworkers [21–23],
Zhang et al. [24] and Bui et al. [1,25]. Chloroaluminate ionic liquids
were immobilized on silica (FK 700) or siliceous MCM-41 to form
‘‘novel Lewis acid catalysts” (NLACs) and their activities were stud-
ied by Kumar et al. [26]. While chloroaluminate ionic liquids have
been shown plausible, the selectivity and alkylate quality were sel-
dom greater than those obtained in commercial alkylation pro-
cesses. Additionally, chloroaluminate ionic liquids are extremely
oxophilic, forming adducts with C–O functionalities and reacting
irreversibly with traces of water to form HCl.
Few studies exist for the use of non-chloroaluminate ionic liq-
uids for alkylation reactions. Olah and coworkers [27] immobilized
HF on an amine-functionalized polymer that forms a poly ammo-
nium catalyst for isobutane/olefin alkylation. Harmer and cowork-
ers [28] have used ionic liquids to sequester or separate novel
superacids catalysts from reaction mixtures. The use of pure acidic
ionic liquids for alkylation reactions has been shown in the Chinese
patent literature [29,30].
For the alkylation of isobutane with butene, the desired attri-
butes of a good homogeneous catalyst and reaction media include
acid strength and tunable solubilities of isobutane and 1-butene in
the reaction mixture. The ability to protonate the olefinic mole-
cules in the feed depends on the acidity of the catalyst while the
concentration of the carbenium ions formed is strongly dependent
on the solubility of isobutane [31]. The isobutane and the alkylates
are largely insoluble in the strong acids. Hence, the organic phase
containing the olefin and excess isobutane is usually dispersed as
droplets in the continuous acid phase to obtain substantial rates
of mass transfer and high yields of the desired alkylate product.
Recently, ionic liquids with functional groups containing
Brønsted acidity have been reported. However, the application of
such ILs in alkylation reactions has not been systematically evalu-
ated. Another possibility that has as yet not been systematically
explored is to use mixtures of strong acids with ionic liquids. Here,
the ionic liquid may be used to beneficially tune not only the acid-
ity of the mixture but also various thermodynamic and interfacial
properties of the acid phase such as better miscibility of reactants
in the acid phase, separation of products from the acid phase and
reduced volatility of the acid. The solubility of strong acids in ionic
liquids and the resulting properties of the mixture are largely un-
known at present. In addition, ionic liquid with acidic functional
groups may also benefit by mixing with other strong acids. The
present work reports on investigations of isobutane/1-butene
alkylation with ILs under the following scenarios: (1) acidic ILs;
(2) mixtures of acidic ILs with strong acids such as H2SO4 or
CF3SO3H; and (3) neutral ILs mixed with strong acids. The 1-butene
conversion and C8 selectivity obtained in each of these cases is
compared with those obtained with pure strong acids at similar
pressure, temperature, and feed composition.
2. Experimental
2.1. Ionic liquid synthesis
Six different types of ionic liquids were investigated represent-
ing 5 different cations and three different anions as shown in Fig. 1.
[BMIm][HSO4] (1-butyl-3-methyl-imidazolium hydrogen sulfate)
was purchased commercially from Sigma–Aldrich Co. The others
were synthesized in house.
2.1.1. General method
2.1.1.1. 1-n-Alkyl-3-methyl-imidazolium halide [R-MIm][X]. The
ionic liquids used in this study were prepared by anion exchange
from the corresponding bromide or chloride salts of the imidazo-
lium cations ([R-MIm][X]) with different n-alkyl substituents as
described in the literature [32]. The bromide or chloride salt of
the imidazolium cation was prepared from a quaternization
Fig. 1. Structures of the cations and anions of the ionic liquids used in this investigation.

S. Tang et al. / Journal of Catalysis 268 (2009) 243–250
245
reaction of 1-methylimidazole with a slight excess of the corre-
sponding alkyl halide in acetonitrile at 313.15 K under argon
atmosphere with stirring for 3 days. Caution: These reactions can
be highly exothermic and adequate solvent volumes and/or cooling
must be provided during the reaction. The solvent was removed in
a rotary evaporator under reduced pressure at 40 °C which was
further evacuated by connection to a high vacuum (<10ꢀ4 Torr)
at 50 °C for at least 48 h.
2.1.1.7. 3-Methyl-1-(3-sulfopropyl)-imidazolium trifluoromethylsulfo-
nate ([MPSIm][OTf]). [MPSIm][OTf] was prepared using a method
described in the literature [35]. The zwitterionic intermediate as
described above was added in a stoichiometric ratio to trifluorom-
ethylsulfonic acid (triflic acid) and stirred at 40 °C until the solid
liquefies to form the ionic liquid. The IL phase was then washed
repeatedly with toluene and ether to remove non-ionic residues,
and dried in vacuo. 1H NMR chemical shifts (relative to TMS inter-
nal standard) and coupling constants J/Hz (400 MHz, CNCD3):
d = 9.33 (s, 1H), 8.53 (s, 1H), 7.45 (t, J = 1.84, 1H), 7.38 (t, J = 1.76,
1H), 4.32 (t, J = 7.12, 2H), 3.88 (s, 3H), 3.16 (q, J = 7.48, 2H), 2.37
(m, J = 7.32, 2H).
2.1.1.2.
1-Hexyl-3-methyl-imidazolium
bis(trifluoromethylsulfo-
nyl)amide ([HMIm][Tf2N]). [HMIm][Tf2N] was prepared from the
anion exchange of [HMIm][Cl] with Li[Tf2N] in deionized water
as described in the literature [32]. The denser hydrophobic IL phase
is decanted and washed six to eight times with approximately
twice the volume of water compared to the IL. The IL is then dried
under vacuum. 1H NMR chemical shifts (relative to TMS internal
standard) and coupling constants J/Hz (400 MHz, CDCl3, TMS):
d = 8.64 (s, 1H), 7.32–7.36 (m, 2H), 4.13 (t, 2H, J = 7.5), 3.89 (s,
3H), 1.83 (t, 2H, J = 7.18), 1.28–1.33 (m, 6H), 0.84 (t, 3H, J = 6.96).
2.2. Materials and reagents
Solvents and chemicals used in the synthesis of the ILs and in
the reaction experiments were commercially available and used
without further purification unless otherwise stated. N-methylim-
idazole, 1,4-butane sultone, trifluoromethanesulfonic acid, sulfuric
acid, anhydrous dichloromethane, diethyl ether, 1-chlorooctane
and bis(trifluoromethylsulfonyl)amide lithium salt were pur-
chased from Sigma–Aldrich Co.
2.1.1.3. 1-Octyl-3-methyl-imidazolium hydrogen sulfate ([OMIm][H-
SO4]). [OMIm][HSO4] was prepared from the anion exchange of
[OMIm][Cl] with H2SO4 in cooled anhydrous methylene chloride
as proposed by Fraga-Dubreuil and coworkers [33]. The IL is then
dried under vacuum. 1H NMR chemical shifts (relative to TMS
internal standard) and coupling constants J/Hz (400 MHz, CNCD3):
d = 10.75 (s, 1H), 8.76 (s, 1H), 7.40–7.44 (m, 2H), 4.18 (t, J = 7.3, 2H),
3.9 (s, 3H), 1.84 (t, J = 1.98, 2H), 1.28–1.32 (m, 10H), 0.90 (t, J = 6.6,
3H).
2.3. Alkylation apparatus and procedure
Isobutane/1-butene alkylations were performed in a 50 mL stir-
red autoclave reactor as shown in Fig. 2 and explained in detail
elsewhere [36]. Temperature and pressure were controlled and
monitored with a Camile 2500 data acquisition system. Cooling
was provided by a recirculating chiller using ethylene glycol. Due
to the exothermicity of the reaction, there was a 2–4 °C increase
in temperature upon reaction for the various catalyst mixtures.
The isobutane and 1-butene feeds (Matheson, 99+%) were pre-
mixed in a storage vessel to a specified I/O ratio, analyzed offline,
and then transferred to a syringe pump and cooled to ꢀ5 °C. The
reactor effluent was collected in a trap chilled by a dry ice and ace-
tone bath. The products were analyzed offline by gas chromatogra-
phy (GC) with a Varian CP-3380 GC, equipped with a flame
ionization detector, and a DB-Petro 100 m column (J&W Scientific).
Helium is used as the GC carrier gas and as the flame ionization
detector (FID) makeup gas. The analysis conditions were: split ra-
tio = 50:1, injector temperature = 250 °C, detector tempera-
ture = 300 °C, carrier gas flow rate = 0.8 sccm. The temperature
program for GC analysis is as follows: initial column temperature
30 °C/hold for 15 min, 0.5 °C/min to 100 °C, then 5 °C/min to
300 °C/hold for 15 min. An alkylate reference standard (Supelco)
allowed identification of the trimethylpentanes (TMP) and dimeth-
ylhexanes (DMH). The GC area percent was equated to weight per-
cent since all hydrocarbons in the reactor effluent have response
2.1.1.4. 3/4-(3-Methyl-imidazoliumyl)propyl (or butyl)-1-sulfo-
nate. Brønsted acidic imidazolium ionic liquids were prepared by
attaching alkyl-sulfonic acid to the imidazolium ring through a
zwitterionic intermediate. 1,3-Propane sultone or 1,4-butane sul-
tone was combined with 1-methyl imidazole in equimolar quanti-
ties in toluene forming the zwitterionic imidazolium sulfonate
which precipitates upon formation. After the reaction, the solid
was washed three times with diethyl ether and toluene to remove
any unreacted starting materials, and the solid was dried in vacuo.
2.1.1.5. 3-Methyl-1-(3-sulfobutyl)-imidazolium hydrogen sulfate
([MBSIm][HSO4]). [MBSIm][HSO4] was prepared by a method de-
scribed in the literature [34]. A stoichiometric amount of H2SO4
was added drop wise to the zwitterionic intermediate as described
above and stirred for several hours at 70 °C until the solid liquefies
to form the ionic liquid. The IL phase was then washed repeatedly
with toluene and ethyl ether to remove non-ionic residues, and
dried in vacuum (393 K, 0.01 Torr). 1H NMR chemical shifts and
coupling constants (relative to TMS internal standard) from a Bru-
ker 400 NMR Spectrometer: J/Hz (400 MHz, CNCD3): d = 10.89 (bs,
2H), 8.55 (s, 1H), 7.43 (s, 1H), 7.36 (s, 1H), 4.21 (t, J = 7.0, 2H), 3.88
(s, 3H), 3.09 (t, J = 7.6, 2H), 1.97 (m, 2H), 1.83 (m, 2H).
2.1.1.6. 3-Methyl-1-(3-sulfobutyl)-imidazolium trifluoro methylsulfo-
nate ([MBSIm][OTf]). [MBSIm][OTf] was prepared following
a
method described in the literature [35]. The zwitterionic interme-
diate as described above was added in a stoichiometric ratio to tri-
fluoromethylsulfonic acid (triflic acid) and stirred at 40 °C until the
solid liquefies to form the ionic liquid. The IL phase was then
washed repeatedly with toluene and ether to remove non-ionic
residues, and dried in vacuo. 1H NMR chemical shifts (relative to
TMS internal standard) and coupling constants J/Hz (400 MHz,
CNCD3): d = 9.10 (s, 1H), 8.51 (s, 1H), 7.42 (t, J = 1.80, 1H), 7.37 (t,
J = 1.74, 1H), 4.20 (t, J = 7.10, 2H), 3.88 (s, 3H), 3.18 (t, J = 7.60,
2H), 2.01 (m, J = 7.46, 2H), 1.84 (m, J = 7.56, 2H).
Fig. 2. Schematic of batch alkylation reactor unit.

246
S. Tang et al. / Journal of Catalysis 268 (2009) 243–250
factors close to unity. The combined mass of TMP and DMH is re-
ferred to as the ‘‘alkylate product”. As discussed elsewhere [37],
propane, an impurity in the isobutane feed, was used as an internal
standard for butene conversion calculations. The butene isomers
were lumped together when calculating butene conversion.
All experiments were performed in batch mode. A typical
experiment begins with the addition of the acid, ionic liquid, or
acid/IL mixture. The reactor was sealed and cooled to a desired
temperature. Approximately 20 mL of the premixed hydrocarbon
feed at a desired isobutane/1-butene ratio was pumped into the
reactor at a flowrate of 5 mL/min with stirring of the liquid phase.
After a certain reaction time (10–30 min), stirring was stopped and
the mixture allowed to settle for 5 min to let the catalyst separate
from the hydrocarbons. The hydrocarbon layer was then led to an
external trap cooled with a dry ice/acetone solution. Care was ta-
ken to avoid outflow of the catalyst layer. For recycling studies,
the vessel was filled with fresh hydrocarbon feed and the proce-
dure repeated.
ditional catalysts: sulfuric acid and triflic acid. The results are
shown in Entries 1 and 2 of Table 1. At 8 °C and a molar isobutane
(I) to 1-butene (O) ratio of 10, nearly quantitative conversion is
achieved in just over 10 min for both catalysts. However, the selec-
tivity for each is markedly different. Sulfuric acid provided a selec-
tivity of nearly 60% to octanes with a trimethylpentanes (TMP)/
dimethylhexanes (DMH) ratio of approximately 6. These results
are consistent with the alkylate quality of a typical commercial
plant [31] (60–65% octanes and TMP:DMH ꢁ 4–6). However, the
C8 selectivity and the TMP/DMH ratio obtained with triflic acid
were approximately 41% and 1.4, respectively. These values are
similar to those reported by Olah and coworkers [38] (20 °C, isobu-
tane/1-butene = 14, 20 min, 22.7% C8 selectivity (TMP + DMH +
methyl heptanes) and TMP:DMH = 0.17). Selectivity and alkylate
quality are strong functions of acid strength. Olah and coworkers
[38] reported that the highest C8-alkylates selectivity occurred
at an acid strength, given by the Hammett acidity scale, of
H0 = ꢀ10.7. At higher acidity however, cracking reactions are
favored and hence lighter components are also observed in the
product stream. Triflic acid is a strong acid with H0 = ꢀ14.1. Conse-
quently, the C8-alkylates selectivity obtained with pure triflic acid
is lower.
3. Results and discussion
The isobutane/1-butene alkylation was investigated using three
different acidic catalytic systems including acidic ionic liquids,
mixtures of acidic ionic liquids with strong acids such as H2SO4
and triflic acid (CF3SO3H), and neutral ionic liquids mixed with
strong acids.
3.2. Isobutane/1-butene alkylation using acidic ionic liquids
Davis and coworkers [39] covalently tethered sulfonic acid
groups to ionic liquid cations to produce Brønsted acidic ionic liq-
uids. Such ionic liquids have shown high activities in esterification
[35], alkylation of phenol with tert-butyl alcohol [34] and Friedel–
Crafts alkylation of aromatic compounds with alkene [40]. We have
synthesized three different types of acidic ionic liquids, where the
acidic group is found in either the cation or anion or both: [MPSI-
3.1. Isobutane/1-butene alkylation benchmark: sulfuric acid and triflic
acid
To provide benchmarks for the ionic liquid-based acid systems,
the alkylation of isobutane with 1-butene was performed with tra-
Table 1
Isobutane /1-butene alkylation results.
Entry Total catalyst Ionic liquid (IL) IL (wt%) Acid
weight (g)
Acid (wt%) Batch time (min) 1-Butene conversion (%) Cs selectivity (%) TMP/DMH selectivity
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
27.01
24.87
25.84
27.61
45.43
30.75
39.01
31.05
31.84
36.43
28.86
30.22
32.86
37.39
24.47
27.61
30.93
33.29
33.20
27.10
25.49
30.26
36.14
26.85
26.45
29.11
31.79
29.32
20.58
24.51
25.07
None
None
H2SO4
CF3SO3H 100.0
None
100.0
10
10
30
10
30
20
20
20
20
20
20
20
20
20
20
20
20
20
10
20
20
20
20
20
10
10
10
10
10
10
10
97.3
98.7
0.3
61.7
41.1
15.4
15.9
19.2
31.1
60.2
11.0
43.6
63.9
6.6
6.1
1.4
0.5
0.6
5.8
4.8
5.1
0.8
3.6
5.1
0.0
2.7
3.8
4.8
0.1
5.2
4.9
7.9
6.6
7.2
8.1
2.6
6.4
8.8
6.2
5.1
5.7
8.1
6.2
6.8
6.8
[HMIm][Tf2N]
[HMIm][Tf2N]
[HMIm][Tf2N]
[HMIm][Tf2N]
[HMIm][Tf2N]
[MPSIm][OTf]
[MPSIm][OTf]
[MPSIm][OTf]
[MBSIm][OTf]
[MBSIm][OTf]
[MBSIm]|OTf]
[MBSIm][OTf]
[BMIm][HSO4]
[BMIm][HSO4]
[BMIm][HSO4]
[BMIm][HSO4]
[OMIm][HSO4]
[MBSIm][HSO4]
[MBSIm][HSO4]
[MBSIm][HSO4]
[MBSIm][HSO4]
[MBSIm][OTf]
[MBSIm][OTf]
[MBSIm][OTf]
[MBSIm][OTf]
[BMIm][HSO4]
[OMIm][HSO4]
[OMIm][HSO4]
[OMIm][HSO4]
100.0
93.6
60.8
42.0
8.7
0.0
6.4
39.2
58.0
91.3
0.0
56.4
73.2
0.0
29.6
53.3
71.8
0.0
32.4
58.3
79.1
84.0
0.0
36.1
59.2
74.0
3.4
18.7
51.7
66.7
76.8
32.4
40.4
76.3
H2SO4
H2SO4
H2SO4
H2SO4
None
H2SO4
H2SO4
None
H2SO4
H2SO4
H2SO4
None
H2SO4
H2SO4
H2SO4
H2SO4
None
H2SO4
H2SO4
H2SO4
CF3SO3H
CF3SO3H
CF3SO3H
CF3SO3H
CF3SO3H
CF3SO3H
CF3SO3H
CF3SO3H
2.2
59.0
89.4
89.7
6.3
93.7
92.2
2.2
100.0
43.6
26.8
100.0
70,4
46.7
28.1
100.0
67.6
41.7
20.9
16.0
100.0
63.9
40.9
26.0
96.6
81.3
48.3
33.3
23.2
67.6
59.6
23.7
8.2
31.2
41.0
57.0
8.7
91.3
91.1
11.4
51.8
91.0
90.4
98.0
5.4
12.1
89.5
95.6
1.1
68.3
95.2
93.3
98.1
85.1
91.2
96.3
22.4
60.3
41.0
51.3
62.0
27.2
25.9
62.3
59.5
57.9
64.7
72.1
72.4
61.9
65.6
75.8
Experimental conditions: initial volume of hydrocarbon feed = 20 mL; molar I/O in feed = 10; P = 2 bar; initial T = 8 °C. For entry 19, I/O in feed = 7.5, second recycle.

S. Tang et al. / Journal of Catalysis 268 (2009) 243–250
247
m][OTf], [MBSIm][OTf], [MBSIm][HSO4], [BMIm][HSO4] and
[OMIm][HSO4] (see Fig. 1). [MPSIm][OTf] and [MBSIm][OTf] have
an acidic sulfonic acid group tethered by either a propyl or butyl
group to the imidazolium cation. [BMIm][HSO4] and [OMIm][H-
SO4] have acidity from the hydrogen sulfate group, which have
shown good activities in acetalization [41] and other acid-cata-
lyzed reactions [42–44]. The alkylation using [BMIm][HSO4] as
the catalyst (Entry 15) resulted in a low conversion, ꢁ11%, even
after 20 min (compared with 10 min for the pure H2SO4) with a
C8 selectivity of only 8.7% and almost no trimethylpentanes pro-
duced. H2SO4 is obviously a much stronger acid than [HSO4] as
the aqueous pKa of H2SO4 is ꢀ3 and for [HSO4] is 1.92. [MBSI-
m][OTf], with a butylsulfonic acid group attached to the cation
and the triflate anion (Entry 11), also yielded low conversions
and C8 selectivity while [MPSIm][OTf] with a propylsulfonic acid
group (Entry 8), yielded only slightly better conversion and selec-
tivities. Thus, the conversions and selectivities with these acidic
ILs alone are much lower than that those obtained using just the
strong acids. It should be noted that this is based upon an equal
mass basis. When converted to a unit of concentration based upon
acidic protons, the disparity in the results becomes much smaller.
For example, the acidic protons for entries 15, 11, and 1 are 0.1,
0.08, and 0.55 moles, respectively. Thus for the pure sulfuric acid
case, the molar quantity of catalyst is more than five times that
of the acidic ionic liquids.
70 wt% of H2SO4 in the IL mixture, the conversion and selectivities
are similar to those obtained with the pure acid. This suggests the
possibility of a significant reduction in acid usage for commercial
alkylations. When using the slightly acidic [HSO4] anion in
[BMIm][HSO4] with H2SO4 (Entries 15–18), the olefin conversion
increases up to approximately 50 wt% acid. However, a maximum
in the C8 selectivity is seen at approximately 58 wt% H2SO4 with
moderate TMP/DMH ratios. At ꢁ79 wt% H2SO4, the C8 selectivity
significantly drops, but yields higher TMP/DMH ratios. Increasing
the alkyl-chain length of the ionic liquids from butyl- to octyl-
([OMIm][HSO4] Entry 19) seems to increase the conversion and
selectivity slightly. The [OMIm][HSO4]/sulfuric acid system repre-
sents some of the best performance and was hence investigated
for its reuse and recyclability.
The IL, [MBSIm][HSO4] (Entries 20–23) that has acidic groups in
both the cation and anion, intrinsically yields selectivities higher
than pure sulfuric acid, but with markedly lower olefin conversion.
Higher proportions of H2SO4 in the IL increase the conversion, but
initially degrades the selectivities. However, at approximately
60 wt% H2SO4 in the IL, the conversion seems to plateau, but the
selectivities are at a minimum. The selectivities increase with fur-
ther amounts of acid, and at ꢁ74 wt% acid, the conversion and
selectivities match those of pure H2SO4. The foregoing results
clearly demonstrate that the catalytic IL/acid mixtures can be
highly tunable for conversion, C8 selectivity, and TMP/DMH ratio
by changing the structure and acidity of the cation and anion
and by varying the proportion of acid.
When acidity is manifested in both the cation and anion, as in
the case of [MBSIm][HSO4] (Entry 20), the C8 selectivity and
TMP/DMH ratio were actually slightly better than that for pure sul-
furic acid (Entry 1). However, the olefin conversion was still signif-
icantly lower. It thus appears that the acidity of these ILs is not
enough to catalyze the alkylation of isobutane/1-butene to match
the olefin conversion rate and selectivities of the strong acids.
3.3.2. Triflic acid
Mixtures of the acidic ionic liquids with triflic acid were inves-
tigated for the isobutane/1-butene alkylation using the cation,
[MBSIm]. [MBSIm][OTf], when mixed with triflic acid (Entries
24–27), increased the conversion while maintaining consistent C8
selectivity and only a slight drop in the TMP:DMH ratio. The selec-
tivities obtained with roughly 67 wt% triflic acid are significantly
better than those obtained with either the neat triflic acid (Entry
2) or neat IL (Entry 11). The long-chain 1-octyl-3-methyl-imidazo-
lium hydrogen sulfate IL ([OMIm][HSO4]), which has a slightly
acidic anion, when mixed with triflic acid at approximately 36
and 76 wt% acid again yielded conversion and selectivities that sur-
pass those obtained with pure triflic acid (Entries 29–31). The
acidic ionic liquids appear to mitigate the poorer selectivity found
with triflic acid, by possibly adjusting (i.e., lowering) the high acid
value of the pure triflic acid.
The addition of IL to strong acids helps tune not only the acidity,
but also other characteristics such as reactant/product solubility,
surface activity, density and viscosity. Wasserscheid et al. [45] ob-
served interplay of acidity and solubility effects caused by the ionic
liquid additive in the Friedel–Crafts-alkylation of benzene with 1-
decene. Low amounts of ionic liquid result in a dramatic improve-
ment of product yield. The acidity of the catalyst can be fine-tuned
by the choice of the amount of ionic liquid additive and by the
choice of its anion. In a similar manner, it should be possible to
tune the solubilities of the alkylation reactants and products by a
suitable choice of the ions of the ionic liquid and by the amount
of added ionic liquid. Among the ILs used for creating the binary
mixtures with conventional acids, the [OMIm][HSO4] (Entry 29–
31) showed the best combination of activity and C8 alkylate selec-
tivity presumably because of improved isobutane solubility in the
binary mixture compared to other IL/acid mixtures. For isobu-
tene + 2-butene alkylation, Yoo et al. [19] observed with imidazo-
lium ionic liquids with the Lewis acid chloroaluminate anion,
that the ionic liquids bearing a larger alkyl group on their cation
([OMIm]) displayed relatively higher activity than a smaller one
([HMIm] or [BMIm]) with the same anionic composition. This
was presumably due to the higher solubility of reactants, especially
3.3. Isobutane/1-butene alkylation using mixtures of ionic liquids and
strong acids
3.3.1. Sulfuric acid
The addition of strong acids to ionic liquid solvents to catalyze
various reactions is reported in the literature [45,46]. However, in
all of these cases the acid represents a small fraction of the total
mass/moles. Here, we have found that large amounts of either
H2SO4 or triflic acid can be dissolved with ionic liquids. For in-
stance, mixing with even the hydrophobic IL, [HMIm][Tf2N], with
equal moles or mass of H2SO4 results in a single homogeneous
phase. This behavior allows fine tuning of the acidity and proper-
ties of the ionic liquid phase. Solutions of both neutral and acidic
ILs with the strong acids were used for the isobutane/1-butene
alkylation. When the alkylation is performed in the neutral ionic li-
quid, [HMIm][Tf2N], without the addition of any catalyst (Entry 3),
there was virtually no reaction (<0.5% olefin conversion). As a
greater fraction of H2SO4 was used, higher conversions were ob-
tained (Entries 4–7) until approximately 50 wt% of the acid, after
which little improvement is seen in the activity. Thus, less acid cat-
alyst seems to suffice to achieve the same level of activity/conver-
sion. The selectivity and TMP/DMH ratio significantly increase with
increased fraction of the acid. This trend is possibly due to the pref-
erential solubility of 1-butene over isobutane in the IL relative to
the pure acid. This neutral IL is known to have moderate solubility
of alkenes and lower solubility of alkanes. This equilibrium parti-
tioning would lead to isobutane starvation in the catalytic phase
and encourage 1-butene dimerization, adversely affecting the C8
and TMP selectivity.
For ILs with the sulfonic acid group, [MPSIm][OTf] and [MBSI-
m][OTf] (Entries 8–10 and 11–14, respectively), higher proportions
of sulfuric acid increase both the conversion and selectivities
throughout the concentration range. Note that at approximately

248
S. Tang et al. / Journal of Catalysis 268 (2009) 243–250
isobutane, in the [OMIm]-based IL. In general, the solubility of
hydrocarbons increases significantly with increasing length of the
alkyl group in the ILs. For a series of 1-alkyl-3-methyl-imidazolium
cations, increasing the alkyl-chain length from butyl to hexyl to oc-
tyl increases the hydrophobicity [32]. Olivier-Bourbigou and Mag-
na [47] also observed that changing the [BMIm] cation for the
longer-chain [HMIm] with a [BF4] anion converts a hydrophilic io-
nic liquid to a hydrophobic IL. These reports further indicate that
the IL with a cation of [OMIm] may have a better miscibility with
isobutane. It therefore appears that in the present work, the IL
addition not only tunes the acidity of strong acids, but also has a
strong effect on the isobutane solubility in the catalyst mixture.
It has been shown previously by our group and others that higher
I/O ratios favor isooctane yields [3].
3.4. Recycling studies of the acid/IL mixtures
The ability of the IL/acid mixtures to be recycled was investi-
gated as to the conversion, selectivity, and TMP/DMH ratio of the
alkylate. [OMIm][HSO4] and H2SO4 were used in initial studies
with a feed I/O ratio of 7.5. The isobutane/1-butene mixtures were
reacted and the reactant/product phase was decanted and re-
moved. Fresh reactants were then added to start the next run.
Fig. 3 illustrates the results of recycling the IL/acid mixture up to
nine times as compared with the results of just the pure sulfuric
acid. Similar conversion and selectivity were observed for both
the IL/acid mixture and the pure acid for about four recycles; how-
ever, the TMP/DMH ratio is better in the IL/acid mixture. After four
recycle runs, the pure sulfuric acid conversion and selectivities sig-
Fig. 4. Comparison of the reuse of catalysts based on trifluoromethanesulfonic acid:
(a) conversion and selectivity; (b) TMH:DMH; and (c) non-C8 products. Catalysts:
Either 26 g of a mixture of [OMIm][HSO4] (22.6 wt%) and 77.4 wt% CF3SO3H or
33.9 g of neat CF3SO3H. Reaction parameters: initial temperature: 8 °C; pressure:
ꢁ2 bar; batch reaction time = 10 min; volume of hydrocarbon fed = 20 mL with a
molar I/O in feed = 7.5. Legend: (ꢀ) 1-butene conversion with IL/acid mixture; (e) 1-
butene conversion with neat CF3SO3 H; (N) C8 selectivity with IL/acid mixture; (4)
C8 selectivity with neat CF3SO3 H; (d) TMP/DMH with IL/acid mixture; (s) TMP/
Fig. 3. Effects of reuse of pure H2SO4 and mixtures of [OMIm][HSO4]/H2SO4: (a)
conversion and selectivity; (b) TMH:DMH; and (c) non-C8 products. Catalysts: 33.2 g
of [OMIm][HSO4] (16 wt%) + H2SO4 (84 wt%) or 26.6 g of H2SO4. Reaction parame-
ters: temperature = 8 °C; pressure ꢁ2 bar; batch reaction time = 10 min; volume of
hydrocarbon fed = 20 mL with
a molar I/O in feed = 7.5. Legend: (ꢀ) 1-butene
conversion with IL/acid mixture; (e) 1-butene conversion with H2SO4; (N) C8
selectivity with IL/acid mixture; (4) C8 selectivity with H2SO4; (d) TMP/DMH
selectivity with IL/acid mixture; (s) TMP/DMH selectivity with H2SO4; (j) the
concentration of light compounds (C7 and C7ꢀ) with IL/acid mixture; (h) the
concentration of light compounds (C7 and C7ꢀ) with H2SO4; (.) the concentration of
DMH with neat CF3SO3H; (j) the concentration of light compounds (C7 and C7-
)
with IL/acid mixture; (h) the concentration of light compounds (C7 and C7-) with
neat CF3SO3H; (.) the concentration of heavy compounds (C9 and C9+) with IL/acid
heavy compounds (C9 and C9+) with IL/acid mixture; and (
heavy compounds (C9 and C9+) with H2SO4.
r) the concentration of
mixture; and (
CF3SO3H.
r) the concentration of heavy compounds (C9 and C9+) with neat

S. Tang et al. / Journal of Catalysis 268 (2009) 243–250
249
Fig. 5. Comparison of performance of [OMIm][HSO4] + CF3SO3H mixtures at different I/O ratios: catalyst used at I/O = 10:25.1 g of [OMIm][HSO4] (23.7 wt%) + CF3SO3H
(76.3 wt%); catalyst used at I/O = 7.5: 26.1 g of [OMIm][HSO4] (22.6 wt%) + CF3SO3H (77.4 wt%). Reaction parameters: temperature = 8 °C; pressure = ꢁ2 bar; batch reaction
time = 10 min; volume of hydrocarbon feed = 20 mL. Legend: (ꢀ, h) 1-butene conversion and (4, N) C8 selectivity; d, s: TMP/DMH selectivity.
nificantly decreased. After eight recycles, the conversion and C8
selectivity were about half of their values observed during the first
run (Fig. 3a). With the IL/acid mixture, the conversion only de-
creased by approximately 20% by recycle 9, with nearly unchanged
C8 selectivity (Fig. 3a), albeit with a much reduced TMP/DMH ratio
(Fig. 3b). The formation of lighter (C7 and C7ꢀ) and heavier (C9 and
C9+) compounds showed similar trends with the neat acid and the
IL/acid mixture up to seven recycles. However, after seven recycles,
the amounts of heavier compounds (C9 and C9+) decreased for the
IL/acid mixture relative to neat sulfuric acid (Fig. 3c) indicating
that the addition of [OMIm][HSO4] to sulfuric acid impedes the for-
mation of acid soluble heavies.
The [OMIm][HSO4] IL was used in the mixtures of triflic acid
with a feed I/O ratio of 7.5. Fig. 4 illustrates that even after 25 recy-
cles, the IL/acid mixture maintains high conversion and C8 selectiv-
ity (Fig. 4a) with better TMP/DMH selectivity (Fig. 4b) than the
pure triflic acid. This IL/acid mixture also does not suffer from
the initial low conversion and selectivity that hampers the pure tri-
flic acid. The high initial acidity of the pure triflic acid leads to
cracking reactions which degrades the selectivity, but the IL seems
to moderate this. From Fig. 4c, it may be inferred that after seven
alkylation reaction. The ionic liquids with inherent acidity in either
the cation or the anion resulted in low conversions and poor selec-
tivities. In contrast, the ILs that had acidity in both the cation and
the anion provided better selectivities even though the conversions
were still low. The mixture of these ILs with sulfuric acid and triflic
acid resulted in superior performance over either the ILs them-
selves, or, importantly, the pure acids. This is especially true for
mixtures with the triflic acid, whose acidity is believed to be too
high for the high selectivities sought for alkylation reactions. Ionic
liquids with longer-chain alkyl groups, such as [OMIm][HSO4],
were found to have better selectivity, which is believed to be due
to higher isobutane solubility in the IL. The recyclability of the IL/
acid mixtures were also superior to the pure acids even out to 25
different recycling runs. These results suggest that ILs offer the
ability to produce liquid acid catalysts with tuned acidity, reactant
solubilities, product separation and mass transfer properties that
may in turn be exploited to optimize the performance of acid cat-
alyzed transformations.
Acknowledgments
recycles, the concentrations of lighter compounds (C7 and C7ꢀ
)
This research was partly supported by the National Science
Foundation Engineering Research Center Grant (EEC-0310689).
S.T. gratefully acknowledges a Visiting Researcher Fellowship from
the China Scholarship Council (20063142) and the grant of a leave
of absence from Sichuan University to pursue research at the CEBC.
were about 40% with triflic acid, implying that cracking reactions
were prominent. However, the concentrations of these lighter
compounds were less than 10% with the IL/acid mixture and the
conversion decreased after 29 cycles with a slight increase of heavy
(C9 and C9+) compounds while the C8 selectivity was stable. In con-
trast, with neat triflic acid, the presence of these heavier com-
pounds increased sharply when the conversion decreased after
38 cycles. These results clearly show that the addition of IL to
the triflic acid inhibits the formation of heavies and the ensuing
deactivation. The spent [OMIm][HSO4]/triflic acid catalyst was
slightly colored after the recycle runs. In comparison, the spent sul-
furic acid was opaque. This suggests that the deactivation of the
mixed IL/triflic acid catalyst may only be partly due to the entrain-
ment of heavy compounds in the acidic solution. Further, the C8
selectivity and the TMP/DMH ratio obtained with the binary
[OMIm][HSO4]/CF3SO3H mixture were higher than those obtained
with H2SO4 and the lifetime of H2SO4 was only 3–6 cycles (Fig. 3).
Fig. 5 illustrates that the conversion and selectivities obtained
at the higher I/O ratio of 10 were nearly identical to the values ob-
tained at an I/O ratio of 7.5. The ability to use lower I/O ratios with-
out sacrificing alkylate yields and stability is desirable to lower
isobutane separation costs.
References
[1] T.L.T. Bui, Investigations on Alkylation of Isobutane with 2-Butene Using
Highly Acidic Ionic Liquids as Catalysts, Dissertation, University of Bayreuth,
2007.
[2] J. Weitkamp, Y. Traa, Catal. Today 49 (1999) 193.
[3] C.J. Lyon, V.R. Sarsani, B. Subramaniam, Ind. Eng. Chem. Res. 43 (2004) 4809.
[4] K. Gong, S. Chafin, K. Pennybaker, D. Fahey, B. Subramaniam, Ind. Eng. Chem.
Res. 47 (2008) 9072.
[5] A. Ahosseini, W. Ren, A.M. Scurto, Ind. Eng. Chem. Res. 48 (2009) 4254.
[6] A. Ahosseini, W. Ren, A.M. Scurto, in: K. Hutchenson, A.M. Scurto, B.
Subramaniam (Eds.), Gas Expanded Liquids and Near-Critical Media: Green
Chemistry and Engineering, ACS Symposium Series 1006, Washington, DC,
2009, pp. 218–234 (Chapter 11).
[7] A. Ahosseini, W. Ren, A.M. Scurto, Chem. Today 25 (2007) 40.
[8] S. Elomari, US Patent 20070225538, 2007.
[9] T.V. Harris, M. Driver, S. Elomari, H.C. Timken, US Patent 2008142413, 2008.
[10] S. Elomari, S. Trumbull, H.K.C. Timken, R. Cleverdon, US Patent 2006135839,
2006.
[11] S. Elomari, H.C. Timken, US Patent 20080146858, 2008.
[12] S. Elomari, S. Trumbull, H.K.C. Timken, R. Cleverdon, US Patent 7432409B2,
2008.
[13] J. Chen, H. Zou, G. Chu, L. Shao, B. Chen, CN Patent 1907924A, 2007.
[14] B. Chen, C. Huang, J. Zhang, P. Ren, Y. Li, CN Patent 1836780A, 2006.
[15] Z. Liu, C. Xu, C. Huang, Y. Liu, CN Patent 1432627A, 2003.
[16] Z. Liu, C. Xu, C. Huang, US Patent 20040133056, 2004.
[17] J. Li, Z. Liu, J. He, G. Lu, X. Zeng, CN Patent 101244972A, 2008.
[18] Y. Chauvin, A. Hirschauer, H. Oliver, J. Mol. Catal. 92 (1994) 155.
4. Summary
Six types of ionic liquids and IL/acid mixtures with sulfuric acid
and triflic acid were utilized as catalysts in the isobutane/1-butene

250
S. Tang et al. / Journal of Catalysis 268 (2009) 243–250
[19] K. Yoo, V.V. Namboodiri, R.S. Varma, P.G. Smirniotis, J. Catal. 222 (2004) 511.
[20] K.S. Yoo, H. Kim, D.J. Moon, P.G. Smirniotis, in: Annual Meeting & Fall Showcase
05 AIChE Conference Proceedings, Cincinnati, OH’05, 2005, pp. 289i/1–289i/8.
[21] C. Huang, Z. Liu, C. Xu, B. Chen, Y. Liu, Appl. Catal. A 277 (2004) 41.
[22] Z. Liu, R. Zhang, C. Xu, R. Xia, Oil Gas J. 104 (2006) 52.
[23] Y. Liu, R. Hu, C. Xu, H. Su, Appl. Catal. A 346 (2008) 189.
[24] J. Zhang, C. Huang, B. Chen, P. Ren, M. Pu, J. Catal. 249 (2007) 261.
[25] T.L.T. Bui, W. Korth, A. Jess, Opportunities and Challenges at the Interface
between Petrochemistry and Refinery, DGMK/SCI-Conference, 10–12 October
2007, Hamburg, Germany, pp. 297–304.
[33] J. Fraga-Dubreuil, K. Bourahla, M. Rahmouni, J.P. Bazureau, J. Hamelin, Catal.
Commun. 3 (2002) 185.
[34] J. Gui, H. Ban, X. Cong, X. Zhang, J. Mol. Catal. A: Chem. 225 (2005) 27.
[35] Y. Gu, F. Shi, Y. Deng, J. Mol. Catal. A: Chem. 212 (2004) 71.
[36] C.J. Lyon, B. Subramaniam, C.J. Pereira, in: J.J. Spivey, G.W. Roberts, B.H. Davis
(Eds.), Catalyst Deactivation. Studies in Surface Science and Catalysis, vol. 139,
Elsevier, 2001, p. 221.
[37] M. Clark, B. Subramaniam, Ind. Eng. Chem. Res. 37 (1998) 1243.
[38] G.A. Olah, P. Batamack, D. Deffieux, B. Török, Q. Wang, Á. Molnár, G.K.S.
Prakash, Appl. Catal. A 146 (1996) 107.
[26] P. Kumar, W. Vermeiren, J. Dath, W.F. Hoelderich, Appl. Catal. A 304 (2006)
131.
[39] A.C. Cole, J.L. Jensen, I. Ntai, K.L.T. Tran, K.J. Weaver, D.C. Forbes, J.H. Davis Jr., J.
Am. Chem. Soc. 124 (2002) 5962–5963.
[27] G.A. Olah, T. Mathew, A. Goeppert, B. Török, I. Bucsi, X. Li, Q. Wang, E.R.
Marinez, P. Batamack, R. Aniszfeld, G.K.S. Prakash, J. Am. Chem. Soc. 127 (2005)
5964.
[28] M.A. Harmer, C.P. Junk, V.V. Rostovtsev, W.J. Marshall, L.M. Grieco, J. Vickery, R.
Miller, S. Work, Green Chem. 11 (2009) 517.
[29] J. Chen, H. Song, C. Xia, Z. Tang, X. Zhang, CN Patent 101177371A, 2008.
[30] J. Chen, C. Xia, X. Zhang, E. Guo, CN Patent 101210192A, 2008.
[31] A. Corma, A. Martinez, Catal. Rev. – Sci. Eng. 35 (1993) 483.
[32] J.G. Huddleston, A.E. Visser, W.M. Reichert, H.D. Willauer, G.A. Broker, R.D.
Rogers, Green Chem. 3 (2001) 156.
[40] K. Qiao, C. Yokoyama, Chem. Lett. 33 (2004) 472.
[41] N. Gupta, Sonu, G.L. Kad, J. Sing, Catal. Commun. 8 (2007) 1323.
[42] V. Singh, S. Kaur, V. Sapehiyia, J. Singh, G.L. Kad, Catal. Commun. 6 (2005)
57.
[43] J. Singh, N. Gupta, G.L. Kad, J. Kaur, Synth. Commun. 36 (2006) 2893.
[44] Z. Zhang, W. Wu, B. Han, T. Jiang, B. Wang, Z. Liu, J. Phys. Chem. B 109 (2005)
16176.
[45] P. Wasserscheid, M. Sesing, W. Korth, Green Chem. 4 (2002) 134.
[46] R.X. Ren, J.X. Wu, Org. Lett. 3 (2001) 3727.
[47] H. Olivier-Bourbigou, L. Magna, J. Mol. Catal. A: Chem. 182–183 (2002) 419–437.