Continuous gas-phase hydroformylation of 1-butene using supported ionic liquid phase (SILP)catalysts

Article abstract of DOI:10.1002/adsc.200600413

The concept of supported ionic liquid phase (SILP)catalysis has been extended to 1-butene hydroformylation. A rhodium-sulfoxantphos complex was dissolved in [BMIM] [n-C₈H₁₇OSO₃] and this solution was highly dispersed on silica. Continuous gas-phase experiments in a fixed-bed reactor revealed these SILP catalysts to be highly active, selective and long-term stable. Kinetic data have been acquired by variation of temperature, pressure, syngas composition, substrate and catalyst concentration. A linear dependency in rhodium concentration could be established over a large concentration range giving another excellent hint for truly homogeneous catalysis in the SILP system. Compared to former studies using propene, the SILP system showed significantly higher activity and selectivity with 1-butene as feedstock. These findings could be elucidated by solubility measurements using a magnetic microbalance.

Full text of DOI:10.1002/adsc.200600413

FULL PAPERS
DOI: 10.1002/adsc.200600413
Continuous Gas-Phase Hydroformylation of 1-Butene using Sup-
ported Ionic Liquid Phase (SILP) Catalysts
Marco Haumann,a,* Katrin Dentler,a Joni Joni,a Anders Riisager,b
and Peter Wasserscheida
a
Lehrstuhl für Chemische Reaktionstechnik (CRT), Universität Erlangen-Nürnberg, Egerlandstr. 3, 91058 Erlangen,
Germany
Fax : (+49)-9131-852-7421; e-mail: marco.haumann@crt.cbi.uni-erlangen.de
Department of Chemistry and Center for Sustainable and Green Chemistry, Building 206, Technical University of
b
Denmark, 2800 Kgs. Lyngby, Denmark
Received: August 17, 2006
Abstract: The concept of supported ionic liquid could be established over a large concentration
phase (SILP) catalysis has been extended to 1- range giving another excellent hint for truly homoge-
butene hydroformylation. A rhodium-sulfoxantphos neous catalysis in the SILP system. Compared to
complex was dissolved in [BMIM][n-C8H17OSO3] former studies using propene, the SILP system
and this solution was highly dispersed on silica. Con- showed significantly higher activity and selectivity
tinuous gas-phase experiments in a fixed-bed reactor with 1-butene as feedstock. These findings could be
revealed these SILP catalysts to be highly active, se- elucidated by solubility measurements using a mag-
lective and long-term stable. Kinetic data have been netic microbalance.
acquired by variation of temperature, pressure,
syngas composition, substrate and catalyst concentra- Keywords: hydroformylation; ionic liquids; kinetics;
tion. A linear dependency in rhodium concentration rhodium; supported catalysts
Introduction
cently been circumvented by us and other groups
using supported ionic liquid phase (SILP) systems.[4]
Among the biphasic catalysis concepts developed for In these systems, a thin film of ionic liquid is confined
the immobilisation of precious homogeneous catalysts on the surface of a highly porous solid by various
the use of ionic liquids has gained significant interest
over the last decade both in academia and industry.[1]
Ionic liquids consist entirely of ions and have ex-
tremely low vapour pressures which makes them at-
tractive as alternative solvents for homogeneous cat-
alysis. Their polar, non-aqueous nature allows the sta-
bilisation of ionic transition metal complexes as well
as complexes that are susceptible to hydrolysis. The
use of ionic liquids in catalysis has been reviewed re-
cently in a number of excellent papers.[2]
Biphasic ionic liquid-organic liquid systems general-
ly require a large amount of ionic liquid. Based on
economic considerations, this is unattractive since
ionic liquids will remain relatively expensive (com-
pared to ordinary organic solvents) even though they
are commercially available now.[3] In addition, the
high viscosity of ionic liquids can induce mass transfer
limitations if the chemical reaction is fast, causing
only a minor part of the ionic liquid and the precious
transition metal catalyst dissolved therein to take part
in the catalysed reaction. These drawbacks have re- phase (SILP) catalysis.
Figure 1. Schematic representation of supported ionic liquid
Adv. Synth. Catal. 2007, 349, 425 – 431
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
425

Marco Haumann et al.
FULL PAPERS
methods such as, e.g., physisorption, tethering, or co- loading of the SILP catalyst was varied between 0.1
valent anchoring of ionic liquid fragments. Figure 1 and 0.9 wt% (mass Rh/total mass SiO2).
schematically illustrates such a catalytic SILP material
containing Rh complexes dissolved in the ionic liquid
film.
In a number of publications we have recently dem-
Variation of 1-Butene Partial Pressure
onstrated that the problem of mass transport limita- The 1-butene partial pressure was varied between 0.6
tion from the gas into the liquid phase can indeed be and 2.4 bar in the temperature range of 80 to 1208C
circumvented by using these SILP catalysts. More- at three different rhodium loadings, namely 0.1, 0.2
over, they allow the application of fixed-bed reactors and 0.3 wt%. The amount of 1-butene in each experi-
for simple continuous processing when applied in ment was well below stoichiometric conditions in
combination with gaseous reaction mixtures making order to avoid reaction rate limitation either by CO
the separation and catalyst recycling obsolete.[5] In the or H2. The residence time in each experiment was ap-
case of propene hydroformylation the use of 2,7-bis- proximately 17 seconds. The reaction order with re-
A
spect to 1-butene was determined from a differential
thene (sulfoxantphos) 1 modified rhodium complexes analysis by plotting ln reff against ln p1-butene and by
in [BMIM][n-C8H17OSO3] dispersed on silica resulted taking the gradient as the reaction order. Figure 2 ex-
in very active and highly selective Rh-1-SILP cata- emplifies the differential analysis for a rhodium load-
lysts. The catalyst lifetime exceeded 200 h time-on- ing of 0.1 wt%.
stream during which a slight decrease in activity was
In Table 1 the results of differential analysis from
observed. Since the selectivity towards n-butanal re- three different Rh loadings are compiled. The reac-
mained high at around 97%, it was concluded that tion rate of the SILP catalyst increased linearly with
possibly high-boiling side products (e.g., from an increasing p1-butene (e.g., entries 4 to 6). The average
aldol condensation consecutive reaction of the reaction order was found to be 1.0Æ0.1 and was in
formed product) accumulated in the ionic liquid film, good agreement with results from both literature
thus diluting the effective rhodium concentration. studies in homogeneous systems and our previously
Upon exposure of the SILP catalyst to 70 mbar reported results from SILP-catalysed propene hydro-
vacuum these heavy species could be removed and formylation. In the temperature range between 80
the initial activity was regained. Consecutive applica- and 1208C the TOFs increased significantly up to
tion of vacuum was possible and total lifetimes of 564 hÀ1 (entry 27). For a given temperature, the selec-
more than 700 h were obtained.[6]
tivity towards the linear aldehyde n-pentanal was nei-
In this paper we describe the extension of the SILP ther affected by p1-butene nor Rh loading (e.g., en-
concept to 1-butene hydroformylation using Rh-1- tries 1–3, 10–12, 19–21). For a given Rh loading, the
SILP catalysts. Although pure 1-butene is not an in- selectivity decreased slightly with increasing tempera-
teresting feedstockfor hydroformylation due to its
high price it can serve as a model for industrial C4
feedstocks. The most attractive C4 feedstockfor hy-
droformylation would be the so-called Raffinate II, a
mixture containing 1-butene (ca. 45%), cis-2-butene
and trans-2-butene (ca. 30%) and butanes (ca.
25%).[7] Furthermore, by using a larger 1-alkene
(compared to our earlier studies of propene hydrofor-
mylation) we wanted to see whether pore diffusion ef-
fects would begin to play a role in the SILP system.
This question was addressed by detailed kinetic studies
of the continuous gas-phase reaction of 1-butene using
otherwise identical SILP systems as formerly investi-
gated for continuous propene hydroformylation.[6]
ture (e.g., entries 10–18). The average selectivity was
Results and Discussion
The standard SILP catalyst consisted of silica 100
(particle size 63 to 200 mm) coated with the ionic
liquid [BMIM][n-C8H17OSO3] (a=volume IL/total
pore volume=0.1) containing the rhodium-sulfoxant-
Figure 2. Differential analysis of 1-butene hydroformylation
using Rh-1-SILP catalysts. 1008C, 10 bar syngas (H2:CO=
1:1), residence time=17 s, L/Rh=10. Rh loading 0.2 wt%,
phos complex with an L/Rh ratio of 10. The rhodium ionic liquid loading 10 vol%.
426
www.asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 425 – 431

Continuous Gas-Phase Hydroformylation of 1-Butene
Table 1. 1-Butene hydroformylation using Rh-1-SILP catalysts at different partial pressures of 1-butene.[a]
FULL PAPERS
Entry
cRh [wt%]
0.1
Temperature [8C]
p1-butene [bar]
Conversion [%]
TOF [hÀ1]
n-pentanal [%]
n1-butene
0.9
1
2
3
4
5
6
7
8
9
80
80
80
100
100
100
120
120
120
0.6
1.2
2.4
0.6
1.2
2.4
0.6
1.2
2.4
1.4
1.3
1.3
6.4
6.4
5.8
7.7
7.0
6.6
17.0
31.1
64.6
79.1
158.2
285.6
94.8
172.9
324.4
99.9
99.9
99.9
98.1
98.2
98.0
97.9
97.8
97.7
average
99.9
99.9
99.9
99.9
98.0
98.0
97.6
97.7
97.6
average
98.4
98.2
98.2
97.8
98.0
97.8
97.7
97.5
97.1
average
0.9
0.9
0.9Æ0.0
10
11
12
13
14
15
16
17
18
0.2
80
80
80
100
100
100
120
120
120
0.6
1.2
2.4
0.6
1.2
2.4
0.6
1.2
2.4
1.6
1.3
1.8
4.3
6.1
6.4
9.8
10.9
10.0
9.6
1.1
15.6
43.3
25.5
72.3
152.3
59.1
130.0
238.9
1.3
1.0
1.1Æ0.2
19
20
21
22
23
24
25
26
27
0.3
80
80
80
100
100
100
120
120
120
0.6
1.2
2.4
0.6
1.2
2.4
0.6
1.2
2.4
5.6
5.1
5.6
18.5
18.8
19.6
35.5
31.2
32.1
24.7
44.5
97.5
81.3
164.8
344.8
155.7
273.8
563.5
1.0
1.0
0.9
1.0Æ0.1
[a]
Reaction conditions: p=10 bar, ratio H2:CO=1:1, residence time 17 s, L/Rh=10, ionic liquid loading=10 vol%.
found to be exceptionally high around 98% n-penta- and selectivity. Increasing the Rh loading had a pro-
nal under all conditions.
nounced effect on the reaction rate, which increased
almost linearly with cRh up to a rhodium loading of
0.3 wt% (entries 28 to 30). In this concentration
range the TOF of each SILP system remained un-
changed, indicating that each rhodium centre showed
Variation of RhodiumConcentration
At 1008C four different SILP catalysts with rhodium the same activity and no cluster formation occurred.
loadings of 0.1, 0.2, 0.3 and 0.9 wt% were studied. When using a rhodium loading of 0.9 wt%
Table 2 compiles the results with respect to activity (entry 31) the observed reaction rate showed a non-
linear behaviour. The TOF of 159 hÀ1 was significantly
lower compared to the rhodium loadings of 0.1 to 0.3
Table 2. 1-Butene hydroformylation using Rh-1-SILP cata- wt% (TOFs between 252 and 271 hÀ1). This reduced
lysts at different Rh loadings.[a]
activity might be caused by mass transport limitations
from the gas into the liquid phase, starting to compete
Entry cRh
reff
Conversion TOF
n-penta-
nal [%]
with the rate of reaction. The latter could be induced
by the very high Rh and – in particular – by the very
high ligand concentration leading to a strong increase
of ionic liquid film viscosity. In all cases the selectivity
was still high with 98% n-pentanal being formed so
that the formation of active, ligand free Rh clusters
appears unlikely. For the range of Rh loadings be-
tween 0.1 and 0.3 wt%, the reaction rate showed first
order dependency in rhodium concentration.
[wt%] [10À3 barsÀ1] [%]
[hÀ1]
28
29
30
31
0.1
0.2
0.3
0.9
7.7
6.6
271
263
252
159
98.0
98.0
97.9
97.6
15.0
20.6
27.7
13.2
18.1
24.3
[a]
Reaction conditions: T=1008C, p=10 bar, ratio H2:CO=
1 : 1, p1-butene =1.9 bar, residence time 16 s, L/Rh=10,
ionic liquid loading=10 vol%.
Adv. Synth. Catal. 2007, 349, 425 – 431
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
427

Marco Haumann et al.
FULL PAPERS
Variation of H2 and CO Partial Pressure
The effect of partial pressure of hydrogen and carbon
monoxide was studied using a standard Rh-SILP cata-
lyst with 0.2 wt% Rh loading. In the region of 0.4 to
2.1 bar partial pressure of CO a negative slope of
À0.1 was found as depicted in Figure 3.
In the region of 0.4 to 1.6 bar partial pressure of
H2, a positive trend was observed. This behaviour of
the SILP catalyst is in accordance with the hydrofor-
mylation mechanism described by Wilkinson.[8] Addi-
tion of hydrogen is assumed to be the rate-limiting
step for most systems, whereas the initial step of the
catalytic cycle is hampered by high partial pressures
of CO. The selectivity towards n-pentanal was not in-
fluenced by the syngas composition and an average
value of 98% was determined from the data.
Figure 4. Arrhenius plot for Rh-1-SILP-catalysed hydrofor-
mylation of propene and 1-butene. 10 bar syngas (H2:CO=
1:1), palkene =1.1 bar, residence time=17 s, L/Rh=10. Rh
loading 0.2 wt%, ionic liquid loading 10 vol%.
Variation of Temperature; Activation Energy
The standard SILP catalyst was studied in the temper-
ature range between 60 and 1308C for two rhodium
loadings: 0.2 wt% and 0.9 wt%. In Figure 4 the re-
sults from an Arrhenius analysis are shown. The
formal, effective activation energy EA,eff was calculat-
The empirical rate law for rhodium loading below
ed to be 63.8 kJmolÀ1 in the case of 0.2 wt% catalyst 0.3 wt% can be written as
loading. The value was in good agreement with the
previously determined EA,eff =63.2 kJmolÀ1 for pro- reff =keff · p1-butene · pÀCO0:1 · p0H:5 · cRh (for cRh <0.3 wt%).
2
pene hydroformylation.[6]
Both values indicate that the SILP system was op-
When using the SILP catalyst with a rhodium load-
erating in the kinetic regime and no severe pore diffu- ing of 0.9 wt% the EA,eff was calculated to be
sion influence occurred under these conditions. Inter- 55.2 kJmolÀ1. The lower value may indicate some ad-
estingly, the effective rate in butene hydroformylation ditional influence on the overall reaction rate. It is es-
was always higher than the one obtained in propene timated that the rate of the gas-film transition in the
hydroformylation.
catalyst pores may become an influencing factor at
the high catalyst and ligands loadings. Additionally,
the temperature dependent change in film viscosity
may contribute to the observed rate dependency on
temperature.
Comparison of Propene and 1-Butene
Hydroformylation
Remarkable, a clear enhancement in both activity and
selectivity can be noticed when comparing SILP-cata-
lysed hydroformylations of 1-butene and propene.
Table 3 presents the experimental data for both sub-
strates at 100 and 1208C.
The SILP catalyst showed a 2.5-fold higher activity
when using 1-butene instead of propene at 1008C. At
1208C the activity was up 2.1 times, whereas the se-
lectivity also increased from ca. 95% to ca. 98% n-al-
dehyde (entries 32/33 and 34/35). Therefore, the solu-
bility of propene and 1-butene in the SILP system
was measured using a microbalance. The results for
Figure 3. Syngas variation in 1-butene hydroformylation
using Rh-1-SILP catalysts. 1008C, 10 bar syngas (H2:CO=
1:1), p1-butene =1.1 bar, residence time=17 s, L/Rh=10. Rh
loading 0.2 wt%, ionic liquid loading 10 vol%.
428
www.asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 425 – 431

Continuous Gas-Phase Hydroformylation of 1-Butene
FULL PAPERS
Table 3. Comparison between Rh-1-SILP-catalysed hydroformylation of propene and 1-butene.[a]
[b]
Entry
Substrate
psubstrate [bar]
Temperature [8C]
TOF [hÀ1]
C4H8/C3H6
n-aldehyde [%]
32
33
34
35
C3H6
C4H8
C3H6
C4H8
2.1
1.9
2.1
1.9
100
100
120
120
103
260
308
647
95.6
98.0
94.8
97.6
2.5
2.1
[a]
Reaction conditions: 10 bar syngas (H2:CO=1:1), residence time=17 s, L/Rh=10. Rh loading 0.2 wt%, ionic liquid load-
ing 10 vol%.
Ratio of activities for 1-butene and propene.
[b]
two different support materials are summarised in Conclusions
Table 4.
When using the SILP standard system prepared In the present workwe have extended the concept of
from a silica 100 support with an average pore diame- supported ionic liquid phase (SILP) catalysis to the
ter of 4 nm the difference in molar solubility between gas phase hydroformylation of 1-butene using contin-
propene and 1-butene resembled the observed differ- uous fixed bed reactors. The SILP catalyst showed the
ence in activity very well. In the pressure regime be- same characteristics of a truly homogeneous complex
tween 0 and 2.6 bar the 1-butene solubility in the as previously described for propene hydroformylation.
ionic liquid [BMIM][n-C8H17OSO3] was 2.4 times A first order in 1-butene partial pressure was ob-
higher compared to the propene solubility (entry 36). tained from differential analysis. The variation of rho-
A porous glass with a larger pore diameter of 30 nm dium loading showed a linear dependency of rate
was also studied (entry 37). Approximately the same with catalyst loading between 0.1 and 0.3 wt% and
ratio in solubility for 1-butene vs. propene was ob- for this regime a full rate law was established. The
served (entry 37). However, the molar solubility of first order with respect to rhodium indicated that the
both propene and 1-butene was up to 5.6 times higher reaction was not limited by mass transport from the
when using silica gel as support instead of porous gas into the liquid phase. The hydrogen partial pres-
glass (entry 40). This might give a hint that, in the sure was found to have a positive effect on the reac-
smallest pores of a SILP silica 100 system capillary, tion rate, whereas the carbon monoxide partial pres-
condensation occurs for both propene and 1-butene sure had a slightly negative one. All three observa-
as substrate. In an additional experiment the ionic tions are in accordance with the mechanism of homo-
liquid was changed from [BMIM][n-C8H17OSO3] to geneous Rh-catalysed hydroformylation. The activa-
[EMIM][n-C2H5OSO3] in order to study the influence tion energy was derived from an Arrhenius plot to be
of the alkyl chain length at the anion. The same solu- 63 kJmolÀ1. Since this value is in good agreement
bility ratio was observed as well (entry 38), but com- with literature data in both homogeneous and bipha-
pared with [BMIM][n-C8H17OSO3] the solubility for sic systems[9] it is assumed that the 1-butene hydrofor-
both 1-butene and propene was lower by factors of mylation using Rh-SILP catalysts is not influenced by
4.3 and 4.6, respectively (entry 40).
pore diffusion effects inside the catalyst particle. Fur-
Table 4. Gas solubility of propene and 1-butene determined by magnetic suspension balance. Comparison of different sup-
port materials and types of ionic liquid.[a]
[b]
Entry
Support
Ionic liquid
Solubility1-butene [10À5 mol gÀ1]
Solubilitypropene [10À5 mol gÀ1]
C3H6/C4H8
36
37
38
39
40
SiO2
BO[c]
BO[c]
EE[e]
19.8
3.8
4.3
5.2
4.6
8.4
1.5
2.0
5.6
4.3
2.4
2.5
2.2
PG[d]
SiO2
SiO2/PG[f]
SiO2
BO
BO/EE[g]
[a]
Reaction conditions: 10 bar syngas (H2:CO=1:1), residence time=17 s, L/Rh=10. Rh loading 0.2 wt%, ionic liquid load-
ing 10 vol%.
Ratio of solubilitypropene/solubility1-butene
BO=[BMIM][n-C8H17OSO3].
PG=porous glass.
EE=[EMIM][n-C2H5OSO3].
Solubility ratio for silica and porous glass.
solubility ratio for [BMIM][n-C8H17OSO3] and [EMIM][n-C2H5OSO3]
[b]
[c]
[d]
[e]
[f]
.
[g]
Adv. Synth. Catal. 2007, 349, 425 – 431
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
429

Marco Haumann et al.
FULL PAPERS
detail previously.[6] The supported ionic liquid phase (SILP)
catalyst was filled into the reactor and the complete rig was
evacuated at room temperature. The rig was pressurised
with 50 bar helium and left under pressure for 30 min while
monitoring the pressure. If no pressure drop was observed,
the reactor was heated to the reaction temperature under
helium pressure. The complete set-up was evacuated and
flushed with helium three times before syngas and 1-butene
were fed into the system. 1-Butene was taken out of a reser-
voir in the liquid state and fed into a heated evaporator via
an HPLC pump to control the molar flow of the substrate.
Both carbon monoxide and hydrogen flows were adjusted
by means of mass-flow controllers (MFC, 5850 S series,
Brooks Instruments). The preheated gases were combined
with the propene in the mixing unit which was filled with
glass beads in order to ensure proper mixing and isothermal
conditions. The gas mixture could then either enter the reac-
tor or exit the system via a bypass. The reactor consisted of
a stainless steel tube (10 mm diameter, 220 mm length)
equipped with a bronze sinter plate for catalyst placement.
After the reactor, the gas mixture passed a 7 mm filter in
order to avoid decontamination of the tubing with catalyst
or solid particles. A back-pressure regulator valve (Samson)
was used to maintain the desired reaction pressure and
outlet gas flow. After the regulator valve, the gas stream
was split and a minor flow was passed through a 134 mL
sampling loop mounted on a gas chromatograph (HP 5890
II plus). Samples were taken at regular intervals by injecting
the volume of the sampling loop via a 6-port valve into the
GC.
thermore, the coefficient of 1-butene diffusion from
the gas into the ionic liquid phase is orders of magni-
tude lower than the gas diffusion coefficient of 1-
butene.[10,11]
At a rhodium loading of 0.9 wt% a lower TOF and
an activation energy of 55 kJmolÀ1 were observed, in-
dicating additional effects to pure chemical kinetics. It
is assumed that, under these conditions, the high
ligand concentration in the liquid film affects the vis-
cosity of the latter and decreases thereby the gas-film
mass transport rate.
Compared to the reaction of propene the SILP cat-
alyst exhibited significantly higher activity and selec-
tivity in 1-butene hydroformylation. The solubilities
of propene and 1-butene were measured and found to
match very well the observed difference in activity.
However, between two different supports and two dif-
ferent ionic liquids the solubility changed for both
propene and 1-butene in an analogous fashion. These
effects may be due to capillary condensation effects, a
point that will attract our specific attention in future
studies.
Experimental Section
General Remarks
All syntheses were carried out using standard Schlenktech-
Gas Chromatography
niques under argon (99.999%). Rh
N
E
and MeOH (HPLC grade) were purchased from Aldrich
and used without further purification. The synthesis of sul-
foxantphos ligand 1 was carried out according to literature
procedures.[12] Silica gel 100 (63 to 200 mm) was purchased
from Merckand was thermally pre-treated at 450 8C for
24 h. Porous glass (PG) was obtained from VitraBio GmbH
and was pre-treated at 4508C for 24 h. Carbon monoxide
(99.7%), hydrogen (99.99%) and 1-butene (99.5%) were
purchased from Linde. The ionic liquids were purchased
from Solvent Innovation GmbH. The water content was
measured with a Karl Fischer titration and the ionic liquids
were dried under high vacuum at 508C overnight.
The conversion of 1-butene as a function of process condi-
tions was measured using the on-line GC technique. An HP
5890 GC equipped with a Pona column (50 m, 0.2 mm diam-
eter, 0.25 mm coating) and a flame ionisation detector (FID)
pre-calibrated for 1-butene, n-pentanal and isopentanal (al-
lowing the peakareas to be transferred into 1-butene con-
version) was applied: Injector temperature 1508C, split ratio
43:1, helium carrier gas flow 2.4 mLminÀ1, detector temper-
ature 2508C. To detect possible high boiling by-products
(heavy species), the following temperature program was
used: initial temperature 508C, initial time 5 min, heating
ramp of 508C minÀ1, final temperature 1508C, final time
3 min, cooling ramp 508C minÀ1, final temperature 508C.
SILP Catalyst Preparation
Microbalance Measurements
Rh(CO)2(acac) (0.0512 g, 0.2 mmol) was dissolved in 20 mL
A
dried MeOH and stirred for 10 min. The ligand sulfoxant-
phos (1.57 g, 2 mmol) was added and the orange solution
was stirred for another 10 min. Afterwards, 1 mL (1.06 g) of
ionic liquid was added to the solution. After stirring for
30 min, 10 g calcinated silica were added and the solution
was stirred for 60 min. The MeOH was removed under
vacuum and a pale red powder was obtained. The supported
ionic liquid phase catalyst was stored under argon until fur-
ther use.
The solubility of propene and 1-butene in the ionic liquid
films of the SILP catalysts was measured using a magnetic
suspension balance from Rubotherm Präzisionsmesstechnik
GmbH. Samples were analysed at 808C and in the pressure
range between 0 and 9.0 (propene) and 0 and 3.5 bar (1-
butene). Densities of the two gases were taken from the
NIST data base.
Acknowledgements
Kinetic Experiments
The continuous reactor set-up used in our laboratories for
the SILP-catalysed hydroformylation has been described in
The authors thank Dipl.-Ing. (FH) Helmut Gerhard for sup-
port regarding the GC analysis and Dipl.-Ing. Daniela Rege-
430
www.asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 425 – 431

Continuous Gas-Phase Hydroformylation of 1-Butene
FULL PAPERS
lein for help with catalyst and ligand synthesis. Prof. Wolf-
gang Arlt, Dipl.-Ing. Alexander Buchele and Dipl.-Ing.
Marta Macakova are gratefully acknowledged for help with
the magnetic microbalance measurements. Furthermore, we
would like to thank Prof. Rasmus Fehrmann (DTU, Lyngby,
Denmark) for many inspiring and fruitful discussions about
the SILP concept.
tions and Processes, (Eds. R. D. Rogers, K. R. Seddon),
ACS Symposium Series, Vol. 902, Chap. 23, 2005, p
334.
[5] a) A. Riisager, R. Fehrmann, S. Flicker, R. van Hal, M.
Haumann, P. Wasserscheid, Angew. Chem. Int. Ed.
2005, 44, 815; b) A. Riisager, R. Fehrmann, M. Hau-
mann, P. Wasserscheid, Eur. J. Inorg. Chem. 2006, 4,
695.
[6] A. Riisager, R. Fehrmann, M. Haumann, B. S. K.
Gorle, P. Wasserscheid, Ind. Eng. Chem. Res. 2005, 44,
9853.
[7] F. Obenaus, W. Droste, J. Neumeister, Butenes, in: Ull-
mannꢀs Encyclopedia of Industrial Chemistry, Wiley-
VCH, Weinheim, 2002.
[8] a) J. F. Young, J. A. Osborn, F. A. Jardine, G. Wilkin-
son, J. Chem. Soc., Chem. Commun. 1965, 131; b) D.
Evans, J. A. Osborn, G. Wilkinson, J. Chem. Soc. (A)
1968, 3133; c) D. Evans, G. Yagupsky, G. Wilkinson, J.
Chem. Soc. (A) 1968, 2660.
[9] a) L. A. Gerritsen, W. Klut, M. H. Vreugdenhil, J. J. F.
Scholten, J. Mol. Cat. 1980, 9, 265; b) C. Yang, Z. S.
Mao, Y. Wang, J. Chen, Catal. Today 2002, 74, 111;
c) R. M. Desphande, Purwanto, H. Delmas, R. V.
Chaudhari, Ind. Eng. Chem. Res. 1996, 35, 3927; U. J.
Jµuregui-Haza, E. Pardillo-Fontdevila, P. Kalck, Catal.
Today 2003, 79–80, 409.
References
[1] H. Olivier-Bourbigou, Catalysis in Nonaqueous Ionic
Liquids, in: Multiphase Homogeneous Catalysis, (Eds.:
B. Cornils et al.) Wiley-VCH, Weinheim, 2006, pp 407–
603.
[2] a) P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed.
2000, 39 3772; b) C. M. Gordon, Appl. Catal. A: Gener-
al 2001, 222, 101; c) P. J. Dyson, Transition Metal Chem.
2002, 27, 353; d) D. Zhao, M. Wu, Y. Kou, E. Min,
Catal. Today 2002, 74, 157; e) J. Dupont, R. F. de Souza,
P. A. Z. Suarez, Chem. Rev. 2002, 102, 3667.
[3] Selection of ionic liquid suppliers: a) Acros Organics
(www.acros.com); b) Fluka (www.fluka.com); c) Merck
(www.merck.com); d) Sigma–Aldrich (www.sigma-al-
drich.com); e) Solvent Innovation (www.solvent-inno-
vation.com); f) Strem (www.strem.com); g) Wako
(www.wako-chem.co.jp).
[4] a) C. P. Mehnert, R. A. Cook, N. C. Dispenziere, M.
Afeworki, J. Am. Chem. Soc. 2002, 124, 12932; b) A.
Riisager, K. M. Eriksen, P. Wasserscheid, R. Fehrmann,
Catal. Lett. 2003, 90, 149–153; c) A. Riisager, P. Was-
serscheid, R. van Hal, R. Fehrmann, J. Catal. 2003, 219,
252–255; d) A. Riisager, R. Fehrmann, P. Wassersc-
heid, R. van Hal, in: Ionic Liquids IIIB: Fundamentals,
Progress, Challenges, and Opportunities – Transforma-
[10] a) D. Camper, C. Becker, C. Koval, R. Noble, Ind.
Chem. Eng. Res. 2006, 45, 445; b) D. Morgan, L. Fergu-
son, P. Scovazzo, Ind. Chem. Eng. Res. 2005, 44, 4815.
[11] R. H. Perry, D. W. Green, in: Perryꢀs Chemical Engi-
neersꢀ Handbook,7th Edn., McGraw-Hill, New York,
1998.
[12] P. Wasserscheid, R. van Hal, A. Bçsmann, Green
Chem. 2002, 4, 400.
Adv. Synth. Catal. 2007, 349, 425 – 431
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.asc.wiley-vch.de
431