Continuous, ionic liquid-catalysed propylation of toluene in a liquid-liquid biphasic reaction mode using a loop reactor concept

Article abstract of DOI:10.1002/adsc.200600414

Acidic ionic liquids of the type [cation] [bis(trifluoromethanesulfonyl) imide anion]/aluminium trichloride {[cation] [(CF₃SO ₂)₂N]/AlCl₃} have been applied to realise for the first time a continuous Friedel-Crafts alkylation reaction in liquid-liquid biphasic reaction mode using a loop reactor concept. Exemplified for the continuous propylation of toluene we have proved that the acidic catalyst layer is successfully immobilised over at least 18 h in the loop reactor. After 8 h time-on-stream, a steady state has been reached followed by at least 10 h time-on-stream without any loss of activity or change in selectivity. Our results reveal the high potential of this class of acidic ionic liquids in continuous arene functionalisation reactions.

Full text of DOI:10.1002/adsc.200600414

FULL PAPERS
DOI: 10.1002/adsc.200600414
Continuous, Ionic Liquid-Catalysed Propylation ofToluene in a
Liquid-Liquid Biphasic Reaction Mode using a Loop Reactor
Concept
Viktor Ladnak,a Norbert Hofmann,a Nicole Brausch,a and Peter Wasserscheida,*
a
Lehrstuhl für Chemische Reaktionstechnik, Universität Erlangen-Nürnberg, Egerlandstr. 3, 91058 Erlangen, Germany
Fax : (+49)-9131-852-7421; e-mail: wasserscheid@crt.cbi.uni-erlangen.de
Received: August 17, 2006
Abstract: Acidic ionic liquids of the type [cation] loop reactor. After 8 h time-on-stream, a steady state
[bis(trifluoromethanesulfonyl)imide anion]/alumini- has been reached followed by at least 10 h time-on-
um trichloride {[cation][(CF3SO2)2N]/AlCl3} have stream without any loss of activity or change in se-
G
been applied to realise for the first time a continuous lectivity. Our results reveal the high potential of this
Friedel–Crafts alkylation reaction in liquid-liquid bi- class of acidic ionic liquids in continuous arene func-
phasic reaction mode using a loop reactor concept. tionalisation reactions.
Exemplified for the continuous propylation of tolu-
ene we have proved that the acidic catalyst layer is Keywords: alkylation; aromatic substitution; contin-
successfully immobilised over at least 18 h in the uous operation; immobilisation; ionic liquids
Introduction
gated ones. Since their discovery, these “first genera-
tion” ionic liquids have been widely used in form of
Electrophilic aromatic substitution reactions – espe- their acidic compositions (ratio [cation]Cl/AlCl3 <1)
cially the group of Friedel–Crafts reactions – are as catalysts for liquid-liquid biphasic Friedel–Crafts
among the most important reactions for the prepara- reactions and similar chemistry making use of the
tion of alkylated arenes in industry. Catalysts for Frie- pronounced miscibility gap between the ionic liquid
del–Crafts alkylation reactions industrially in use are and the aromatic feedstock/product mixture.[6] As a
Lewis acids like AlCl3, FeCl3, TiCl4, ZnCl2, SbF5 and general conclusion from all acid-catalysed reactions
BF3 as well as Brønsted acids like HCl, H2SO4, H3PO4 with chloroaluminate ionic liquids it can be stated
or HF.
that the acidity of the ionic liquid depends on the
One important way in which Friedel–Crafts alkyla- amount of AlCl3 in the mixed [cation]Cl/AlCl3
tion reactions are carried out technically is the liquid system. The higher the amount of AlCl3, the higher is
phase reaction with a homogeneously dissolved AlCl3 the acidity and hence the catalytic activity of the ionic
catalyst (red-oil). In this process, the catalyst is re- liquid. However, the amount of AlCl3 in the chloro-
moved from the formed products by a washing proce- aluminate system is effectively limited by the liquid
dure with water and thus a large amount of waste range of the ionic liquid to a maximum molar ratio of
water is produced. Moreover, the washing process 2 mol AlCl3 per mol of [cation]Cl.
leads to the hydrolysis of the AlCl3 catalyst which
cannot be recycled and the formed alumina is either type [cation]
disposed off or finds applications in, e.g., street build- larger amounts of Lewis acids {e.g., up to 5 equiva-
ing. lents of AlCl3 per mol of “matrix ionic liquid”
A very promising method for the immobilisation of [cation]
[(CF3SO2)2N]}.[7] In addition they form bipha-
Recently, we have reported that ionic liquids of the
AHCTREUNG
AHCTREUNG
acidic catalysts is being offered by ionic liquids. Ionic sic systems with many organic product mixtures so
liquids represent a new class of solvents which consist that the same possibility of a multiphasic reaction
only of ions and are liquid at temperatures below mode can be applied as known from the chloroalumi-
1008C.[1–3] In recent years, ionic liquids have attracted nate systems. In this way easy product isolation and
much interest as catalysts and alternative solvents for catalyst recovery can be realised. Most interestingly,
chemical transformations.[4] Chloroaluminate ionic liq- in the system [cation]
[(CF3SO2)2N]/AlCl3 the ionic
G
uids were the first Lewis acidic ionic liquids to be liquid can be seen as a true “matrix” for the AlCl3.
studied in detail[5] and are still among the best investi- The valuable matrix ionic liquid can be recycled even
Adv. Synth. Catal. 2007, 349, 719 – 726
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
719

Viktor Ladnak et al.
FULL PAPERS
in cases where an aqueous work-up would become using ionic liquids of the general type AlCl3/[cation]
N
necessary after prolonged use of the acidic ionic [(CF3SO2)2N] (ratio>1.5) being applied as the cata-
liquid. This is due to the hydrolytic stability and the lyst phase. Our aim was to gain from such a study a
hydrophobic character of the (CF3SO2)2N anion. In more detailed insight into the technical potential of
contrast, the organic cation is usually lost into the continuous Friedel–Crafts alkylation reactions with
aqueous medium if acidic ionic liquids of the type ionic liquids.
[cation]Cl/AlCl3 are worked-up hydrolytically.
Today, a detailed literature survey reveals many
recent publications describing acidic ionic liquid cata- Results and Discussion
lysts in Friedel–Crafts alkylation reactions.[8–11] Most
of them also deal with the recovery and reuse of the Preliminary Experiments in a Pressureless, Semi-
acidic phase in form of repetitive batch mode experi- Batch Reactor
mentation. Several recycling experiments were de-
scribed to be affected by a slight loss of activity,[12] es- Influence of the AlCl3/
N
N
pecially at higher recycling rates. This may have to do the Applied Ionic Liquid
not only with intrinsic catalyst instability but also with
the specific conditions of the repetitive batch experi- First, we aimed to understand influencing factors in
ments. This methodology suffers from the necessity to the catalytic system under investigation. Therefore,
completely recover the catalyst phase after a run for the propylation of toluene was investigated in a batch
proper re-injection into a new batch. Loss of catalyst reactor system with propene bubbled through a cata-
phase and also contamination with traces of humidity lyst/organic biphasic system with vigorous stirring (for
are usually inevitable at least to some extent in this details see Experimental Section) to determine the in-
procedure and this leads to deactivation due to the fluence of the matrix ionic liquid to AlCl3 ratio on the
extreme hygroscopic nature of the Lewis acidic ionic activity and selectivity of the reaction. Figure 2 shows
liquid systems.
the results for the system AlCl /
G
N
In the present paper we report for the first time a ([BMIM]=1-butyl-3-methyl-imidazolium) in the com-
Friedel–Crafts alkylation using acidic ionic liquids as position range of 0.65 to 3.68 in this pressureless ap-
the catalyst phase that is immobilised in a loop reac- paratus at 158C.
tor under continuous reaction/extraction conditions.
Figure 1 shows the particular reaction under investi- served with a maximum yield of meta-cymene at high-
gation (here schematically represented in a stirred est acidities. For molar ratios AlCl3/[BMIM]
A clear tendency of the selectivities could be ob-
AHCTREUNG
tank reactor), namely the propylation of toluene [(CF3SO2)2N]<1.5 no catalytic activity was found. We
attribute this observation of a required “minimum
acidity” in the specific system AlCl3/[BMIM]
G
[(CF3SO2)2N] to the presence of different anion spe-
cies depending on the amount of AlCl3 added to the
matrix ionic liquid.[13] Only a molar ratio of AlCl3
greater than 1.5 forms anionic species (dimeric and
trimeric species are also assumed) acidic enough to
catalyse the propylation of toluene. Increasing the
amount of AlCl3 leads to an increased formation of
highly acidic species in the catalyst system accompa-
nied by a higher selectivity to m-cymene formation
(clearly beyond the analytical error).
Variation ofthe Total Amount ofCatalyst System
Since the amount of matrix ionic liquid [BMIM]
[(CF3SO2)2N] was kept constant in these experiments,
the total amount and volume of the catalyst phase
were changed with different amounts of AlCl3 added
to the catalyst system. The next set of experiments
aimed to make sure that the selectivity of the reaction
was not a function of the total volume of the catalyst
phase in the experiment. Therefore, further alkylation
reactions were carried out with a constant molar ratio
Figure 1. Isopropylation of toluene in ionic liquids.
720
www.asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 719 – 726

Ionic Liquid-Catalysed Propylation of Toluene
FULL PAPERS
Figure 2. Yield of products (based on propene) obtained with different AlCl3/
N
N
ionic liquid catalyst phase. Conditions: T=158C; reaction time: 15 min; propene=7 NL/h (0.4 mol/h); n(toluene)=0.27 mol,
n([BMIM][(CF3SO2)2N])=0.015 mol).
ACHTREUNG
Table 1. Influence of the total amount of catalyst system AlCl3/
N
G
distribution.[a]
m
[g]
A
[BMIM]
T
Yield [%]
Conversion
[%]
meta-
para-
ortho-
di.-substitut- tri-substitut-
cymene
cymene
cymene
ed
ed
1.6
4.4
8.9
11.8
11.1
11.4
18.2
18.2
18.1
21,7
21.4
20.5
26.7
31.1
32.8
5.1
8.3
5. 7
83.6
90.0
88.8
[a] Conditions: T=158C; reaction time: 15 min; propene=7 NL/h (0.4 mol/h); n(toluene)=0.27 mol.
AlCl3 to matrix ionic liquid [BMIM]
with different amounts of total catalyst system rated from the organic compounds after the reaction
(Table 1). time of 20 min and reused for a next alkylation reac-
As expected, we could clearly show that the ratio tion under identical conditions. In order to remove or-
of AlCl3 to [BMIM][(CF3SO2)2N] and thus the ionic ganic components completely from the ionic liquid,
[(CF3SO2)2N] but systemꢁs stability, the ionic catalyst system was sepa-
AHCTREUNG
liquidꢁs acidity is decisive for the selectivity change the catalyst was extracted three times with cyclohex-
observed (see Figure 1) and not the total amount of ane prior to reuse. The yields of products over the
acidic catalyst phase in the reactor.
five recycling experiments are shown in Figure 3.
Figure 3 shows an almost constant product distribu-
tion for the monoalkylated toluenes over all six
cycles. No loss of activity could be observed, so that
the application in a continuous rig appeared to be jus-
Recycling Experiments in Repetitive Batch Mode
The main precondition to apply the new catalyst tified. It should be noted that in this experiment the
system in a continuous liquid-liquid biphasic opera- molar ratio between toluene and propene was higher
tion is the sufficient stability of the catalytic system than in previous experiments resulting in slightly dif-
over time under the conditions of an intensive extrac- ferent selectivities with less formation of di- and tri-
tion of the acidic catalyst phase by the organic feed- substituted products. Increasing the molar ratio of tol-
stock/product phase. For a preliminary check of this uene with respect to propene also resulted in a higher
Adv. Synth. Catal. 2007, 349, 719 – 726
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
721

Viktor Ladnak et al.
FULL PAPERS
Figure 3. Recycling experiments of the ionic catalyst system. Conditions: T=158C; reaction time=20 min; propene=7 NL/h
(0.4 mol/h); n(toluene)=0.46 mol, n([BMIM][(CF3SO2)2N])=0.015 mols), n([BMIM][(CF3SO2)2N]/AlCl3)=1/3.3.
A
N
ACHTREUNG
yield towards m-cymene in the monoalkylated prod- ionic catalyst solution in the organic phase. Two heat
ucts as can be seen by comparing the data with those exchangers in the loop ensured full temperature con-
presented in Table 1 and Figure 2.
trol. As the loop circulation flow is large compared to
the continuous inflow of feedstock and outflow of
products, the system in use was characterised by a res-
idence time distribution similar to a CSTR (continu-
ous stirred tank reactor). A more detailed residence
time distribution investigation revealed that the data
Continuous Experiments in a Loop Reactor with
Internal Phase Separation
To further check the technical potential of the new of a step function experiment could be well fitted to a
catalyst system [BMIM][(CF3SO2)2N]/AlCl3, the op- CSTR reactor showing 11% of short circuit flow and
AHCTREUNG
erational lifetime of the catalyst system was investi- 31% of the flow not fully back-mixed (e.g., due to
gated under the realistic conditions of a continuous dead volume in the separator).
operation in a loop reactor. In contrast to the semi-
Figure 5 shows product yield (with respect to pro-
batch experiments described above, the reaction is pene) and amounts of different products formed for a
carried out under pressure in the loop reactor to selected example of our continuous catalytic runs
ensure that propene is processed as a liquid phase over 18 h at 508C.
with no reactive gas phase present in the reactor. A
The yield of alkylated products reached a stable
schematic view and a picture of the applied loop reac- value after four hours time-on-stream (corresponding
tor system are shown in Figure 4, a detailed descrip- to about 7 residence times) and stayed constant for
tion of the operating method can be found in the Ex- the next 14 h. No deactivation of the catalyst was ob-
perimental Section.
served after 18 h of operation when the continuous
Prior to the catalytic reaction, the ionic catalyst so- reaction was liberally stopped. However, an interest-
lution was placed into the reactor loop. The organic ing shift in product selectivity was found during the
feedstock components were continuously fed and – first 8 h of operation with decreasing selectivity for
after passing an internal phase separator – the prod- meta-cymene (from 45% to 15% during the first 8 h)
ucts were continuously discharged. Thus, in contrast and a pronounced rise in the formation of ortho-
to the use of an external separation, all catalyst was cymene (from 10% to 20% within the first 10 h). Re-
always present in the loop reactor. This allowed direct markable, the product distribution remained essential-
investigation of the catalyst reactivity by analysis of ly unchanged between 10 h and 18 h time-on-stream
the isolated products. In the loop reactor, the liquid- with selectivity to monoalkylated products of about
liquid biphasic reaction mixture was circulated by a 55%. Compared to our semi-batch results the differ-
circulation pump with high flow rates (4 LminÀ1). ent initial selectivity can be explained by the much
Under these conditions the two static mixers in the higher initial molar ratio of toluene with respect to
reactor loop provided an efficient dispersion of the propene (the reactor was completely filled with tolu-
722
www.asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 719 – 726

Ionic Liquid-Catalysed Propylation of Toluene
FULL PAPERS
Figure 4. Continuous loop reactor used for the isopropylation of toluene with acidic ionic liquids of the type AlCl3/[BMIM]
G
[(CF3SO2)2N].
Figure 5. Continuous isopropylation of toluene in a liquid-liquid biphasic mode using the catalyst system AlCl3/[BMIM]
E
(propene)=8.7 mmolminÀ1; mole
˙
[(CF3SO2)2N] in a continuous loop reactor. Conditions: T=508C; p=40 baÀr1; mole flow n
À1
˙
˙
˙
flow n
A
dence time=32 min; AlCl3/
A
[(CF3SO2)2N]=2.5; m
G
N
ene when the reaction was started). In addition, we uted to some Al leaching. Interestingly, after 8 h time
wanted to investigate whether the observed change in on stream the system obviously stabilises with almost
selectivity with time over the first ten hours (in partic- constant selectivities with the following 10 h.
ular between 4 h and 10 h time on stream when the
product yield is constantly high) could also be attrib-
Adv. Synth. Catal. 2007, 349, 719 – 726
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
723

Viktor Ladnak et al.
FULL PAPERS
Figure 6. Al leaching from the ionic catalyst system during the long-term experiment as calculated from the results of ICP-
AES analyses of the separated organic phases after different times-on-stream.
ICP-AES Analysis to Check for Al Leaching from
the Catalyst System
may be relatively labile with a chance to release neu-
tral molecules, such as, for example, Al-
A
Some Al leaching could be convincingly confirmed by anions have degraded over time, the system stabilises,
accompanying ICP-AES analyses in which the Al indicating the presence of very robust anionic, acidic
contents of the isolated organic product phases were species in the system at acidities around
monitored over time-on-stream. The graph in Figure 6 AlCl /
N
N
has been calculated from these ICP-AES results by far above the minimum acidity of 1.5 which is re-
expressing Al leaching data as decreasing quired to catalyse the reaction (see Figure 2), the
AlCl /
A
ACHTREUNG
ionic liquid catalyst phase.
The curve shows clearly that Al leaching is relative- changed selectivity compared to the initial higher
ly significant in the first 8 h of the experiment while acidity.
only negligible amounts of aluminium-containing
It is quite illustrative to present the relative product
compounds are lost into the organic phase between distribution of the three monoalkylated products in a
8 h and 18 h time-on-stream. The latter fact is remark- triangle diagram (Figure 7). The arrow in the Figure
able taking into account that in our experiment about indicates the change in the ratio of the three cymene
30 mL ionic liquid are extracted every hour of the ex- products over the first 8 h reaction time.
periment with more than 1 litre of organic phase.
At the beginning of the alkylation reaction – where
Thus between 8 h time-on-stream and 18 h time-on- the high toluene to propene ratio largely determines
stream the catalyst layer is extracted with about 10 the selectivity – the mixture of monoalkylated prod-
litres organic reaction mixture with only negligible ucts happens to be near to the thermodynamic equi-
amounts of Al leaching detected in the ICP-AES librium. However, the stable catalytic system obtained
analyses. This proves impressively the operational sta- after 8 h time-on-stream operates clearly under kinet-
bility of our catalyst system AlCl /
[(CF3SO2)2N] once a certain ratio of AlCl3 to matrix ised for the subsequent 10 h time-on-stream was
ionic liquid [BMIM][(CF3SO2)2N] and thus a certain found to be 1:1.6:1.8 for m-cymene/p-cymene/o-
3 [BMIM] ic control. The ratio of the three cymene isomers real-
AHCTREUNG
ionic liquid acidity is reached. The change in leaching cymene under our operation conditions.
behaviour over reaction time is explained by the fact
that the initial catalyst system AlCl3/[BMIM]
G
[(CF3SO2)2N]=2.5 represents a complex mixture of
acidic anionic species. Some of these anionic species
724
www.asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 719 – 726

Ionic Liquid-Catalysed Propylation of Toluene
FULL PAPERS
Experimental Section
General Remarks
All chemicals – except cyclohexane and toluene – were ob-
tained from commercial sources (Fluka, Aldrich, Merck)
and used without further purification. Cyclohexane and tolu-
ene (BASF) were dried using a solvent treatment rig from
Glass Contour and stored under argon prior to use. All
other chemicals were also stored under argon. The ionic
liquid [BMIM][(CF3SO2)2N] was prepared and characterised
G
according to well-known procedures[14] and dried overnight
at 808C under high vacuum. Its water content was con-
firmed to be below 20 ppm prior to application by Karl–
Fischer titration using a Metrohm 756 KF Coulometer with
a Hydranalꢂ Coulomat AG reagent. All synthetic reactions
were performed in dried glassware under a static pressure of
argon. Liquids and solutions were transferred with syringes.
GC analyses were carried out using a Varian CP 3900
FID. The analytical error has been checked to be below 2%
relative to the measured values. Conditions: column:
50 m”0.25 mm CPSIL PONA CB DF =0.5 mm; temperature
profile: 110–2508C, 15 iso, 158CminÀ1, 15 min iso; injector
temperature: 2508C; eluent: H2; volume flow: 1 mLminÀ1.
All detected products were given in yields of o-, m- , p-
cymene or grouped as di- or trisubstituted products. The
combined yield of all indicated products and product frac-
tions equals the overall product yield.
Figure 7. Product distribution of the monoalkylated products
during the long-term experiment; the curved arrow indicates
the change with time; the indicated value for the thermody-
namic equilibrium is taken from ref.[16]
Conclusions
1H and 13C NMR spectra were recorded in DMSO on a
Jeol 400 MHz spectrometer.
Our results show clearly that the immobilisation of
Lewis acids in ionic liquids offers a new and very at-
tractive method for continuous, liquid-liquid biphasic
Friedel–Crafts alkylation reactions. The miscibility
gap of the aromatic products with the catalyst phase
combined with a significant density difference be-
tween catalyst phase and product mixture (ca.
0.6 g/mL) enable very effective product separation in
the internal gravity separator of a loop reactor. The
ICP-AES Analysis
ICP-AES analyses were performed using a Perkin–Elmer
Emissions-Spectrometer Plasma 400. For sample prepara-
tion, the organic sample (~2 g) was two times extracted with
50 mL distilled water. The combined aqueous phases were
then analytically investigated.
catalyst system AlCl /
A
G
Preparation ofthe Catalyst
applied in a 18-h continuous run in which more than
18 litres of reaction mixture were processed in the re-
actor with yields greater than 90% with respect to the
stoichiometric limiting component propene. The acidi-
ty of the catalyst system decreased in the first 8 h
from 2.5 to 2.1 due to Al leaching into the organic
phase. This change in acidity resulted in a pronounced
change in selectivity to the different cymene products.
The calculated amount of AlCl3 was added in portions
slowly to the liquid [BMIM][(CF3SO2)2N] at room tempera-
G
ture under strictly inert conditions to avoid any effect of the
atmospheric moisture. The flask was heated with an oil bath
to 808C and the contents were vigorously stirred until the
AlCl3 was completely dissolved.
Batch Experiments
Interestingly,
AlCl3/[BMIM]
the
system
stabilised
at
an
A
ACHTREUNG
The alkylation reactions were carried out in 250-mL three-
maining 10 h of the experiments in which stable activ- neck flasks, immersed in an oil bath with an automatic tem-
perature controller. The flasks were equipped with a gas
inlet, a reflux condenser and magnetic stirrer. The reactants
were introduced into the flask under strictly inert conditions
in order to avoid any effect of the atmospheric moisture.
After the completion of the reaction, the organic layer con-
taining the products and unreacted reactants was separated
from the ionic liquids layer by decantation. The products
were washed with water, dried over Na2SO4 and analysed by
gas chromatography.
ity and stable selectivity could be demonstrated.
Future work will deal with detailed kinetic studies of
a slightly modified system under continuous operation
as well as attempts to maximise the catalytic activity
with respect to Al equivalents applied. In this way we
would like to gain more insight into Lewis acid-cata-
lysed, liquid-liquid biphasic, ionic liquid processes to
further evaluate the technical applicability of such
systems.
For the recycling experiments the ionic liquid was three
times extracted with cyclohexane in order to remove all or-
Adv. Synth. Catal. 2007, 349, 719 – 726
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
725

Viktor Ladnak et al.
FULL PAPERS
ganic products. Afterwards the ionic liquid was re-injected
into the flask and the experiment was repeated as described
before.
from the loop reactor the remaining alumina and hence the
remaining acidity of the ionic liquid in the reactor could be
easily calculated.
Continuous Experiments[15]
The acidic ionic liquid, typically about 45 g of References
AlCl3/[BMIM][(CF3SO2)2N]=2.5, was injected in the loop
A
ACHTREUNG
reactor that was completely evacuated for several hours and
filled with inert gas (N2). The aromatic compound (diluted
with cyclohexane) was introduced first in the liquid phase
by HPLC pump (toluene/cyclohexane: 0.4–40 mLminÀ1) so
that the reactor was totally filled with liquid. The reaction
started with the liquid introduction of propene by HPLC
pump (propene: 0.01–10 mLminÀ1). A back-pressure regula-
tor ensured a constant pressure (0–60 bar). The total volume
of the continuous rig was determined to be 285 mL, showing
11% of short circuit flow and 31% of the flow not fully
back-mixed (e.g. due to dead volume in the separator). Two
static mixers (Sulzer SMX DN6) dispersed the ionic liquid
[1] P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed. 2000,
39, 3772–3789.
[2] P. Wasserscheid, T. Welton (Eds.), Ionic Liquids in Syn-
thesis, 1st edn, Wiley-VCH, Weinheim, 2002, pp 1–356.
[3] S. A. Forsyth, J. M. Pringle, D. R. MacFarlane, Aust. J.
Chem. 2004, 57, 113–119.
[4] T. Welton, Coord. Chem. Rev. 2004, 248, 2459–2477.
[5] A. A. Fannin, D. A. Floreani, L. A. King, J. S. Landers,
B. J. Piersma, D. J. Stech, R. L. Vaughn, J. S. Wilkes
and J. L. Williams, J. Phys. Chem. 1984, 88, 2614.
[6] M. Earle, in: Ionic Liquids in Synthesis, (Eds.: P. Was-
serscheid, T. Welton), Wiley-VCH, Weinheim, 2003,
pp 174–213.
with the organic reactants.
A circulation pump (1–
4.5 LminÀ1) provided enough energy for a fine dispersion,
while heat exchangers controlled the reaction temperature
(15–958C). At the gravimetric phase separator, a diffuser
absorbed out the kinetic energy of the dispersed liquids so
that the ionic liquid settled down and were recirculated in
the loop while a part of the organic products was discharged
from the loop reactor.
[7] N. Brausch, A. Metlen, P. Wasserscheid, Chem.
Commun. 2004, 1552–1553.
[8] a) H. Xin, Q. Wu, M. Han, D. Wang, Y. Jin, Appl.
Catal. 2005, 292, 354–361; b) Z. Zhao, W. Qiao, Z. Li,
G. Wang, L. Cheng, J. Mol. Catal. 2004, 222, 207–212.
[9] X. Sun, S. Zhao, R. Wang, Chin. J. Chem. Eng. 2004,
12, 658–661.
Since the loop circulation flow was larger compared to
the continuous inflow of feedstock and outflow of products,
the system in use was characterised by a residence time dis-
tribution similar to a CSTR (continuous stirred tank reac-
tor). Hence, the most part of the organic phase is being re-
circulated several times before leaving the reactor.
After a certain time a quantitative amount of the products
was neutralised by washing with distilled water and analysed
by gas chromatography.
In order to obtain the amount of alumina which was dis-
charged from the reactor, the water from the neutralising
process was analysed with a Perkin–Elmer Emissions Spec-
trometer Plasma 400. Since no ionic liquid was discharged
[10] Z. Zhao, W. Qiao, Z. Li, G. Wang, L. Cheng, J. Mol.
Catal. 2004, 222, 207–212.
[11] Y. Xiao, S. Malhotra, J. Mol. Catal. 2005, 230, 129–133.
[12] L. Jia, Y. Wang, H. Chen, Y. Shan, L. Dai, Reaction
Kinetics and Catalysis Letters 2005, 86, 267–273.
[13] N. Brausch, Dissertation, Universität Erlangen-Nürn-
berg, 2006.
[14] P. Bonhꢃte, A.-P. Dias, N. Papageorgiou, K. Kalyana-
sundaram, M. Grätzel, Inorg. Chem. 1996, 35, 1168–
1178.
[15] P. Wasserscheid, M. Eichmann, Catal. Today 2001, 66,
309–316.
[16] C. Perego, P. Ingallina, Catal. Today 2002, 73, 3–22.
726
www.asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 719 – 726