Microwave-assisted partial hydrogenation of citral by using ionic liquid-coated porous glass catalysts

Article abstract of DOI:10.1002/cssc.201100154

The microwave-assisted hydrogenation of citral (3,7-dimethylocta-2,6- dienal) to citronellal with molecular hydrogen as the reducing agent was investigated. Several polar and non-polar solvents were screened and imidazolium-based ionic liquids were applied as modifiers for the palladium-containing porous glass catalysts (Pd/TP). The best results were obtained with N-ethyl-N'-methylimidazolium dicyanamide, N-ethyl-N'- methylimidazolium acetate, or N-ethyl-N'-methylimidazolium trifluoroacetate, which were used to prepare supported catalysts with an ionic liquid layer (SCILL) on Pd/TP by wet-impregnation. The influence of pressure and temperature when using these ionic liquid-containing catalysts, as well as their long-term stabilities, were examined. Working with microwave-assisted heating, high yields of citronellal were achieved under mild conditions within short reaction times. Catalyst characterization was carried out by means of BET measurements, X-ray photoelectron spectroscopy (XPS) and thermo-gravimetric analyses. The influences of the ionic liquid layer were derived from experiments carried out before and after the reactions. Copyright

Full text of DOI:10.1002/cssc.201100154

DOI: 10.1002/cssc.201100154
Microwave-Assisted Partial Hydrogenation of Citral by
using Ionic Liquid-Coated Porous Glass Catalysts
Thomas Gallert,[a] Martin Hahn,[a] Martin Sellin,[a] Christine Schmçger,[a] Achim Stolle,*[a]
Bernd Ondruschka,[a] Thomas F. Keller,[b] and Klaus D. Jandt[b]
The microwave-assisted hydrogenation of citral (3,7-dimethy-
locta-2,6-dienal) to citronellal with molecular hydrogen as the
reducing agent was investigated. Several polar and non-polar
solvents were screened and imidazolium-based ionic liquids
were applied as modifiers for the palladium-containing porous
glass catalysts (Pd/TP). The best results were obtained with N-
ethyl-N’-methylimidazolium dicyanamide, N-ethyl-N’-methylimi-
dazolium acetate, or N-ethyl-N’-methylimidazolium trifluoroace-
tate, which were used to prepare supported catalysts with an
ionic liquid layer (SCILL) on Pd/TP by wet-impregnation. The in-
fluence of pressure and temperature when using these ionic
liquid-containing catalysts, as well as their long-term stabilities,
were examined. Working with microwave-assisted heating,
high yields of citronellal were achieved under mild conditions
within short reaction times. Catalyst characterization was car-
ried out by means of BET measurements, X-ray photoelectron
spectroscopy (XPS) and thermo-gravimetric analyses. The influ-
ences of the ionic liquid layer were derived from experiments
carried out before and after the reactions.
Introduction
Citral (3,7-dimethylocta-2,6-dienal, 1) is a naturally occurring
monoterpenoid that is applied in the synthesis of flavors, fra-
grances, pharmaceuticals, vitamins, and nutrients.[1] Hydroge-
nation of a,b-unsaturated aldehydes such as 1 is an approved
method for initial (de)functionalization in the production of
building blocks for the synthesis of specialty chemicals.[2] Allylic
alcohols (nerol, geraniol), saturated and unsaturated aldehydes
(citronellal, 2), as well as cyclic compounds (menthols, pule-
gols) can be formed by hydrogenation of 1 (Scheme 1) due to
the presence of different types of double bonds (i.e., conjugat-
ed and isolated carbon-carbon double bonds, a carbonyl
bond). However, parallel and consecutive reactions result from
the variety of products that may be formed initially. Thus, the
challenge in using hydrogenation catalysts is achieving differ-
entiation between the primary reaction pathways (formation)
and the suppression of consecutive reactions. This selectivity
depends on the catalyst metal, the support material, and the
auxiliary, which control formation of the target molecules.
Chatterjee et al. could achieve high yields of the unsaturated
alcohol geraniol (selectivity>98% at conversion>99%) by
using platinum on a zeolite support in supercritical carbon di-
oxide.[3] Cyclization and hydrogenation of 1 result in the forma-
tion of isopulegols and menthols, if metal catalysts (e.g., Ni) on
acidic supports (e.g., MCM-41, zeolite beta) are applied.[4,5]
Aumo et al. were able to transform citral to citronellal over Ni/
Al2O3 catalysts with selectivities of 93% at 90% conversion,[6]
whereas with a water-soluble PdCl2(TPPTS)2 complex (TPPTS=
trisodium tris(m-sulfonatophenyl)phosphine) a selectivity for
citronellal of 97% at 96% conversion could be reached using
Na2CO3 as a basic additive.[7]
[a] T. Gallert, M. Hahn, Dr. M. Sellin, Dr. C. Schmçger, Dr. A. Stolle,
Prof. Dr. B. Ondruschka
Institute for Technical Chemistry and Environmental Chemistry (ITUC)
Friedrich Schiller University Jena
Lessingstraße 12, 07743 Jena (Germany)
Fax: (+49)3641-948402
E-mail: achim.stolle@uni-jena.de
[b] Dr. T. F. Keller, Prof. Dr. K. D. Jandt
Institute of Materials Science and Technology (IMT)
Friedrich Schiller University Jena
Scheme 1. Reduced reaction scheme for the hydrogenation of citral (1), ne-
glecting hydrogenation of the carbonyl bond and cyclization processes. For
a complete reaction network see Ref. [8].
Lçbdergraben 32, 07743 Jena (Germany)
1654
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2011, 4, 1654 – 1661

Partial Hydrogenation of Citral by using Ionic Liquid-Coated Porous Glass Catalysts
Ionic liquids are compounds that consist exclusively of
Results and Discussion
anions and organic cations, are liquid below 1008C, have a
negligibly small vapor pressure, and lack a well-ordered crystal
structure.[9] In the last decades their tunable properties have
been used in a multitude of applications. For example,[10] they
have been used as materials for gas absorption, and as acid
scavengers in BASF’s BASIL process. In organic synthesis, their
use as alternative solvents has been established for a broad
range of reactions.[9b,11] Beside their solvent properties, influen-
ces on reaction pathways and/or catalytic effects are often ob-
served.[12]
The selective hydrogenation of 1 towards 2 according to
Scheme 1 was investigated by using a palladium catalyst on a
porous glass support (Pd/TP) as applied in catalytic hydrogena-
tion[22] as well as CÀC cross-coupling reactions, such as the
Suzuki–Miyaura[22b,23] or Mizoroki–Heck reactions.[23,24] The glass
support employed is a phase-separated borosilicate glass con-
sisting mainly of SiO2 with minor impurities of B2O3 and Al2O3,
as determined by EDX and XPS analyses. The specific surface
area SV, pore size dp, specific pore volume Vp, and particle size
d50 were determined to be 87 m2 gÀ1, 55 nm, 1.26 cm3 gÀ1, and
100–200 mm, respectively. Compared to other types of phase-
separated glass, silica, and alumina, the chosen material has
the best properties for application as the support material in
the palladium-catalyzed hydrogenation of 1. The catalysts
were prepared by wet-impregnation[22d] and exhibited a palla-
dium loading of 1 wt% (0.09 mmolgÀ1). The catalysts were
subsequently impregnated with various ionic liquids. The ab-
breviations and structures of cations and anions used in ionic
liquid combinations in this study are presented in Scheme 2.
Furthermore, acetate [OAc], iodide [I], thiocyanate [SCN], and
tetrafluoroborate [BF4] were applied as counter ions for the N-
alkyl-N’-methylimidazolium cations.
In addition, the negligibly low vapor pressure prevents sol-
vent loss due to evaporation, thus allowing more secure and
sustainable handling. Unfortunately ionic liquids are much
more expensive than classical solvents for laboratory applica-
tion. For that reason, much catalysis research is focused on
concepts where ionic liquids can be used sparingly by apply-
ing a small amount immobilized as a thin layer onto a support
material. Two different concepts can be found in literature:
(i) supported ionic liquid phases (SILP), and (ii) supported cata-
lysts with an ionic liquid layer (SCILL). In SILP systems, a homo-
geneous solution of a metal complex in an ionic liquid is de-
posited on the support surface.[13] This concept combines the
good performance of homogeneous catalysts with ease of sep-
aration, which allows applications in fixed-bed reactors.[13a,14]
Due to the excellent solubility of metal complexes in ionic liq-
uids, supplemental modifications are unnecessary. SCILL cata-
lysts comprise a classical supported catalyst, which is decorat-
ed with an ionic liquid by wet-impregnation or dip-coating.[8,15]
These systems are also known as supported ionic liquid cata-
lysts (SILCAs).[16]
Microwave-assisted heating has been applied for organic
synthesis (MAOS) in the laboratory[17] as well as in chemical in-
dustry.[18] Microwave irradiation heats the reaction mixture di-
rectly, uniformly, and more rapidly, resulting in shorter reaction
times and, often, fewer side reactions.[19] Due to more efficient
reaction control, less waste is produced and less energy is con-
sumed. Many papers describe the use of ionic liquids in combi-
nation with microwave heating.[20] Even small amounts of ionic
liquids added to microwave-transparent solvents are able to
accelerate chemical reactions due to strong interaction be-
tween ions and microwaves.[21] This increases the overall kinet-
ic energy, as manifested by the increased temperature of the
reaction mixture. This not only qualifies ionic liquids for appli-
cation as bulk solvents, but also as additives for efficient heat-
ing of non-polar solvents by microwave techniques.
Scheme 2. Abbreviations and formulas of cations (top line) and anions used
in the present study.
Catalyst characterization
Our aim was to investigate the combination of SCILL cata-
lysts with microwave-assisted heating for selective hydrogena-
tion of citral (1) to citronellal (2). Hence, a variety of ionic liq-
uids as well as reaction conditions (solvent, pressure, tempera-
ture) were examined. The catalysts were reused four times to
examine their stabilities, and the results were compared with
conventionally heated reactions. Catalyst characterization was
carried out by means of X-ray photoelectron spectroscopy
(XPS), thermogravimetric analysis (TGA), and Brunauer–
Emmett–Teller (BET) measurements.
The catalysts were prepared as follows: first, the porous glass
support was wet-impregnated with Pd(OAc)2 as the precursor;
second, the palladium was calcined, affording the precatalyst
comprising PdO particles on the glass surface;[22d] third, the
Pd/TP catalysts were transformed into SCILL catalysts by coat-
ing them with a specific amount of ionic liquid. At the begin-
ning of the reaction, an in situ reduction of the catalyst to the
correct oxidation state took place. Before and after the reac-
tions, the catalyst samples were characterized by BET measure-
ments, XPS, and TGA. Leaching of the ionic liquid layer was in-
vestigated by TGA and HPLC.
ChemSusChem 2011, 4, 1654 – 1661
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
1655

A. Stolle et al.
BET measurements
situ to the active species Pd0 at the beginning of the hydroge-
nation according to a loss of EB by 1 eV.[22d] Similar observations
were made for palladium species on the SCILL catalysts, which
suggests that there is no adverse effect on the reduction of
palladium from the additional ionic liquid layer. The high-reso-
lution scan in the range of 390–415 eV offered typical values
for EB of the N 1s core level (401 eV).[28] These were attributed
to the imidazolium nitrogen atoms of the cation [C2mim]. The
binding energy remained unaltered during the reaction, which
indicates that no interaction between the imidazolium cation
and the metal exists; this is probably due to repulsive electro-
static forces. Thus, the anion is coordinated to the metal rather
than the cation of the ionic liquid. In the case of dicyanamide
[DCA] containing CAT-1a, a second signal for nitrogen was de-
tected at 397 eV, which shifted to 398 eV during hydrogena-
tion. Due to the lower binding energies, a more electron-rich
environment must be present than in [DCA]. In contrast to the
imidazolium (401 eV), the shift for nitrogen in [DCA] confirms a
stronger influence of anions on the reaction. The same phe-
nomenon was observed while examining the [SCN]-containing
catalyst (N 1s: 397.0 to 398.0 eV and 401.0 to 401.3 eV). There-
fore, interactions of the anion with partially positively charged
palladium can be assumed, while cations would be repulsed.
On entering the coordination sphere of palladium, the electro-
chemical environments of the anion and the palladium are dis-
turbed, which could result in shifted binding energies with the
observed spectral changes as a consequence.
The porous glass support was coated twice during catalyst
preparation: First, with the active metal palladium and second,
with a layer of ionic liquid. Nitrogen physisorption was used to
investigate the influence of this process on textural properties.
The specific surface area of the bare support was 87 m2 gÀ1
(Table 1). Coating with palladium particles hardly decreased
this value, whereas coating the ionic liquid layer had a stron-
ger influence. Both preparation steps resulted in similar trends
regarding the pore volumes, in accord with previous findings
for these types of catalysts.[8,25,26] No pore volumes could be
calculated for the CAT-2b and CAT-3b samples due to incom-
plete hysteresis. This may be due to the diffusion of nitrogen
into the ionic liquid layer, disordering the adsorption–desorp-
tion equilibrium because of additional absorption mechanisms.
Table 1. Specific surface area (SV) and specific pore volume (Vpore) of the
employed catalysts.[a]
Entry
Composition[b]
SV
Vpore
[m2 gÀ1
]
[cm3 gÀ1
]
TP
porous glass
87
87
54
66
80
66
66
61
1.5
1.5
1.2
1,3
1.5
1.4
nd
nd
Pd/TP
porous glass+Pd
CAT-1a
CAT-2a
CAT-3a
CAT-1b
CAT-2b
CAT-3b
Pd/TP+10 wt% [C2mim][DCA]
Pd/TP+10 wt% [C2mim][OAc]
Pd/TP+10 wt% [C2mim][TFOAc]
Pd/TP+5 wt% [C2mim][DCA]
Pd/TP+5 wt% [C2mim][OAc]
Pd/TP+5 wt% [C2mim][TFOAc]
[a] SV and Vpore based on BET and BJH analyses, respectively. [b] Palladium
loading of the catalyst is 0.09 mmolgÀ1 (1 wt%). nd: not determined.
Thermogravimetric analysis
To control the stability of the SCILL catalysts, the ionic liquid
recovery rate was investigated by TGA. Fresh samples of CAT-
1a and CAT-2a were analyzed, used five times for hydrogena-
tion of 1 (4ꢁrecycling), and checked again. Prior to hydrogena-
tion, 11 wt% ionic liquid was found on both samples (Table 3).
The used CAT-1a catalyst still contained 11 wt% ionic liquid
whereas in the case of CAT-2a only 10 wt% or 9 wt% could be
detected. Thus, the layer of ionic liquid remained on the cata-
lyst surface and very little leaching occurred. These results
were supported by HPLC analysis of the reaction mixtures, in
which no ionic liquid was detected. With a detection limit for
the cation of 2.5 mgLÀ1 less than 1% of the ionic liquid layer
might have leached from the catalyst.[29]
X-ray photoelectron spectroscopy
XPS was not only used to identify oxidation states of the palla-
dium species, but also to retrieve information about the influ-
ence of the ionic liquid on the reaction. Therefore, high-resolu-
tion spectra of Pd 3d (330–355 eV) and N 1s (390–415 eV) core
levels were analyzed (Table 2). Prior to reaction, the binding
energies EB for Pd 3d5/2 and Pd3/2 of the neat catalyst Pd/TP
had similar values to those expected for Pd2+ species.[27] Ac-
cording to Schmçger et al., the PdO particles are reduced in
Table 2. Binding energies (EB) for Pd 3d and N 1s core levels of the cata-
lysts prior to and after application.
Table 3. Thermo-gravimetric analysis of the catalysts.[a]
[a]
[a]
Catalyst
EB Pd 3d5/2
EB Pd 3d3/2
EB N 1s[a]
[eV]
[eV]
[eV]
Catalyst
Recovery rate for ionic liquid
Before application
[wt%]
After 5th application[a]
[wt%]
Pd/TP
CAT-1a
336.1/335.3
335.5/334.0
341.3/340.3
341.0/339.5
397.3/398.3
400.8/400.5
401.0/400.3
401.0/401.0
CAT-1a
CAT-2a
11
11
11[b]/11[c]
10[b]/9[c]
CAT-2a
CAT-3a
335.8/335.0
335.5/334.5
341.0/340.0
340.8/339.5
[a] Reaction conditions: 7.5 mmol citral in 30 mL n-heptane, 200 mg cata-
lyst, 0.24 mol% Pd; p=0.2 MPa, T=508C, t=90 min. After each reaction
the catalyst was separated by filtration, washed with n-heptane and
dried. [b] Microwave-assisted heating. [c] Conventional heating.
[a] Before reaction/after reaction. Reaction conditions: 7.5 mmol citral in
30 mL n-heptane, 200 mg catalyst, 0.24 mol% Pd; p=0.2 MPa, T=508C;
t=90 min, microwave-assisted heating.
1656
www.chemsuschem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2011, 4, 1654 – 1661

Partial Hydrogenation of Citral by using Ionic Liquid-Coated Porous Glass Catalysts
Solvent selection
higher boiling point of n-heptane (b.p.=988C) allows a wider
temperature range for reaction conditions, while ease of prod-
uct separation is maintained. n-Heptane performed better than
the other solvents and thus was chosen for further experi-
ments.
The solvent is of certainly important when performing hydro-
genation reactions in liquid phase,. The solubilities of reactants
and gaseous hydrogen, as well as the absorption of microwave
irradiation and miscibility with ionic liquid, are essential for this
complex reaction system. Nonpolar solvents such as n-hex-
ane[15d,30] or n-dodecane[15c] are commonly used for reactions
with SCILL catalysts to avoid desorption of the ionic liquid
from the catalyst and its concomitant absorption or dissolution
into the solvent. To study the influence of microwave irradia-
tion, polar solvents (acetonitrile, DMF, ethyl acetate) were
tested along with nonpolar n-heptane and toluene (Table 4).
Ionic liquid selection
In the presence of an ionic liquid, a heterogeneous supported
catalyst will form a SCILL catalyst in-situ. Compared with unde-
corated support catalysts, the selectivity and activity can be
significantly influenced by coating a heterogeneous catalyst
with even a thin layer of ionic liquid.[15a,b,26] This effect can be
explained as follows: The ionic
Table 4. Conversion of citral (1) and selectivity to citronellal (2) after 90 min reaction time in the absence or
liquid is able to act as a co-cata-
lyst due to its intrinsic properties
and the effective concentrations
of substrates, intermediates, and
products can be changed by the
additional polar phase (physical
solvent effect).[38] The literature
contains several examples of
ionic liquids that were used to
form SCILL catalysts and tested
in the hydrogenation of 1. By
applying [C4mim][PF6][25,39] or
presence of 22 mg [C2mim][NTf2].[a]
Solvent
Absence
Conv. [%]
Presence
Conv. [%]
m [D][36]
tan d[37]
Sel. [%]
Sel. [%]
acetonitrile
DMF
ethyl acetate
n-heptane
toluene
3
37
51
>99
15
39
87
49
54
76
8
6
35
53
9
59
93
55
76
70
3.92
3.82
1.78
0
0.062
0.161
0.059
0.020[b]
0.040
0.38
[a] Reaction conditions: 7.5 mmol citral in 30 mL n-heptane, 200 mg catalyst, 0.24 mol% Pd; p=0.2 MPa, T=
508C; microwave-assisted heating; m: permanent dipole moment, tan d: loss tangent. [b] Value is for n-hexane.
Hydrogenation of 1 was carried out by using Pd/TP with or
without ionic liquid as an additive. The catalyst was less active
when using polar solvents, with the best results achieved in
ethyl acetate (conversion X=51%) and DMF (X=37%). The se-
lectivity (S) towards the target product 2 was good in DMF
(87%). Complete conversion of 1 was observed only with n-
heptane, which occurred within 90 min and at 0.2 MPa hydro-
gen pressure. Nevertheless, no trend could be recognized, as il-
lustrated by acetonitrile and toluene: in both solvents hardly
any reaction took place, although their polarities are complete-
ly different. Addition of [C2mim][NTf2] to the solvents resulted
in decreased X, due to the in situ formation of SCILL catalysts
and hindered mass transport.[31] However, a selectivity increase
occurs independently from the solvent applied. The remarka-
bly good performance in n-heptane seems to result from its
microwave transparency. Only the catalyst and the ionic liquid
as well as substrate and products are able to interact with mi-
crowaves. Generation of the reaction temperature afforded for
nonpolar solvents (n-heptane, toluene) a higher energy input,
since the number of molecular dipoles or ions that absorb mi-
crowaves are lower. Thus, superheating of the metallic cata-
lyst[32,33] or hot spots on the catalyst surface occur.[34,35] Howev-
er, those lead not to destruction of the catalyst but enhance
the heating up of the reaction solution in non-polar n-heptane.
In combination with the presence of coordinated charge carri-
ers (ionic liquid) it facilitates the possibility to operate the reac-
tion in a solvent which is known for its low microwave absorp-
tion ability As a result the catalyst will be activated in a more
direct and efficient way. The immiscibility with ionic liquids
supports the in situ formation and stability of SCILL catalysts,
leading to a higher selectivity. Compared to n-hexane, the
[BMPL][NTf2] (BMPL=N-butyl-N’-methylpyrrolidinium),[15d] citro-
nellal could be obtained as the main product and with
[C4mim][DCA] a selectivity of >99% was reached.[15d] To check
which ionic liquids are able to improve the formation of 2, a
variety of anion–cation combinations (Scheme 2) were added
to the reaction mixture. In ionic liquids, the chemical proper-
ties are determined mainly by the anion, whereas the physical
properties depend on the cation.[40] For that reason, two imida-
zolium cations with different alkyl side chains were combined
with several anions. Based on XPS studies, a stronger influence
of anions on the catalyst metal is observed (Table 2). We
assume that the polarity of the ionic liquid decreases with the
increasing length of alkyl side chains of the imidazolium
cation, so ethyl [C2mim] and n-butyl [C4mim] were chosen in
addition to the methyl group at the heterocycle. Further elon-
gation of the substituent’s chain length would result in dissolu-
tion rather than in stable deposition onto the catalyst. As seen
in the solvent tests, the conversion of 1 decreased with addi-
tion of [C2mim][NTf2]. Fortunately, the selectivity is improved
only by a few additives (Figure 1, 2): [C2mim][OAc], [C2mim]-
[TFOAc], [C2mim][DCA], and [C4mim][OAc] showed an in-
creased formation of 2 with increasing conversion of 1. In the
case of the halide and pseudo-halide containing ionic liquids
[C2mim][SCN], [C4mim][SCN], and [C4mim][I], the catalyst was
deactivated and no reaction took place. Even though [C2mim]-
[SCN] seemed to have a positive influence (S=89% at X=
39%) when used as an additive, its SCILL catalysts were no
longer active due to poisoning with sulfur. In high-resolution
XPS of thiocyanate-containing SCILL catalyst, shifts in EB of pal-
ladium from 336.8 to 337.5 eV (Pd 3d5/2) as well as from 341.8
to 343 eV (Pd 3d3/2) were observed, which indicate an oxida-
ChemSusChem 2011, 4, 1654 – 1661
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
1657

A. Stolle et al.
Figure 3. Citral (1; 7.5 mmol) conversion dependence on hydrogen pressure
in the hydrogenation to citronellal in 30 mL n-heptane (catalyst: 200 mg,
0.24 mol% Pd; see Table 1; T=508C, t=90 min; microwave-assisted heat-
ing).
Figure 1. Conversion–selectivity plot for the hydrogenation (p=0.2 MPa,
T=508C) of citral (1; 7.5 mmol) to citronellal (2) in 30 mL n-heptane under
addition of [C2mim]-based ionic liquids (22 mg; catalyst: 200 mg Pd/TP,
0.24 mol% Pd; t=30, 60, and 90 min; microwave-assisted heating).
Figure 4. Citronellal selectivity dependence on hydrogen pressure in the hy-
drogenation of citral (7.5 mmol) in 30 mL n-heptane (catalyst: 200 mg,
0.24 mol% Pd; see Table 1; T=508C, t=90 min; microwave assisted heat-
ing).
Figure 2. Conversion–selectivity plot for the hydrogenation (p=0.2 MPa,
T=508C) of citral (1; 7.5 mmol) to citronellal (2) in 30 mL n-heptane under
addition of [C4mim]-based ionic liquids (22 mg; catalyst: 200 mg Pd/TP,
0.24 mol% Pd; t=30, 60, and 90 min; microwave-assisted heating).
containing layer seems to strongly decrease the citral concen-
tration, resulting in an inhibition of the reaction. To enhance
selectivity, hydrogen availability at the metal centers should be
low to avoid further hydrogenation. In the case of CAT-2 and
CAT-3, a linear dependence was observed, similar to the results
Burgener et al. obtained with their Pd/Al2O3 catalyst in a pres-
sure range of 3–19 MPa.[43] Except for CAT-2, a higher loading
of ionic liquid resulted in lower catalyst activity due to reduced
accessibility to the metal sites.
tion of PdII to PdIV.[22d,41] Further ionic liquid additives were
tested, however, they did not enhance the performance of the
catalyst. Therefore, SCILL catalysts were only prepared using
[C2mim][DCA] (CAT-1), [C2mim][OAc] (CAT-2), and [C2mim]-
[TFOAc] (CAT-3). The catalysts are differentiated by their ionic
liquid loading, indicated by either the suffix “a” (10 wt%) or
“b” (5 wt%; cf. Table 1).
Good to excellent selectivities could be achieved with the
tested catalysts at low pressure. Except for CAT-1b, lower selec-
tivities were observed using SCILL-catalysts loaded with only
5 wt% ionic liquid. By increasing the H2-pressure, the selectivi-
ty towards 2 remained constant for CAT-1 and CAT-2. Consecu-
tive hydrogenations were well inhibited, especially with CAT-2,
in spite of complete conversions of 1. This is because the ba-
sicity of the acetate anion[44] in the ionic liquid layer supports
the formation of 2.[7] Using CAT-3, increased pressure lead to a
loss of selectivity but an increase in conversion. Dihydrocitro-
nellal (4; Scheme 1) was the main side product formed due to
Pressure dependence
Pressure is an important parameter when working with gas-
eous hydrogen, as it determines the amount of dissolved H2 in
the reaction mixture (Henry’s Law).[42] In our experiments, the
H2 pressure was varied from 0.2–1.0 MPa while keeping the
temperature constant at 508C (Figures 3 and 4). For reactions
using CAT-1, an increase of pressure had almost no effect on
conversion of 1. Consequently, we assume the adsorption of 1
to be the rate-determining step. Hence, the [C2mim][DCA]-
1658
www.chemsuschem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2011, 4, 1654 – 1661

Partial Hydrogenation of Citral by using Ionic Liquid-Coated Porous Glass Catalysts
consecutive hydrogenations, however isomers of 2 could also
was observed. To the contrary, the selectivity of CAT-3b
be detected.
showed a clear dependence on the conversion level and thus
the temperature. It could not inhibit consecutive hydrogena-
tion, resulting in a significant loss of selectivity at conversion
levels above 80%.
Temperature dependence
At a H2 pressure of 0.4 MPa the temperature was varied from
30–908C. The conversion levels increased with temperature
reaching a maximum at 708C (Figure 5). Catalysts with 5 wt%
ionic liquid achieved similar or higher conversion levels than
the 10 wt% samples, as discussed above (Figure 3, 4). At 308C,
Recycling experiments
During experiments investigating the p/T-dependence of citral
liquid-phase hydrogenation, the highest conversions were ob-
served at medium pressure (0.4 MPa) and temperature (708C).
CAT-1a and CAT-2a were best suited for these conditions, since
they efficiently inhibit consecutive hydrogenation of 2. Also,
their long time-stabilities were tested in five hydrogenation re-
actions, including 4ꢁwork-up of the catalyst (Figure 7), and
were compared with experiments performed in a conventional
steel autoclave with PTFE-coating to avoid cross-contamination
(Figure 8). CAT-3 was not used in these experiments due to its
decreased selectivity towards 2 with increased conversion. The
selectivity achieved with both catalysts is consistently >95%,
Figure 5. Temperature dependence of citral (1; 7.5 mmol) conversion in the
hydrogenation to citronellal in 30 mL n-heptane (catalyst: 200 mg,
0.24 mol% Pd; see Table 1; p=0.4 MPa, t=90 min; microwave assisted heat-
ing).
CAT-1a and CAT-3a had low activities resulting in conversions
of 1 between 8% and 13%, whereas with CAT-2a acceptable
results could be achieved (X=57%). The selectivity regarding
the formation of 2 for CAT-1a and CAT-1b was consistently
higher than 90%, slightly increasing with temperature to 99%
and 98%, respectively (Figure 6). Also, with CAT-2a and CAT-2b
a selectivity increase with temperature could be obtained re-
sulting in S=98% and 97%, respectively, for catalysts at 908C.
In the case of CAT-3a, no effect of temperature on selectivity
Figure 7. Influence of catalyst (200 mg, 0.24 mol% Pd; Table 1) recycling on
conversion (full symbols) and citronellal selectivity (open symbols) during
the hydrogenation of citral (1; 7.5 mmol) in 30 mL n-heptane (p=0.4 MPa,
T=708C, t=90 min; microwave-assisted heating, catalysts were filtered
from the reaction mixture before reapplication).
Figure 8. Influence of catalyst (200 mg, 0.24 mol% Pd; Table 1) recycling on
conversion (full symbols) and citronellal selectivity (open symbols) during
the hydrogenation of citral (1; 7.5 mmol) in 30 mL n-heptane (p=0.4 MPa,
T=708C, t=90 min; conventional heating in steel autoclave, catalysts were
filtered from the reaction mixture before reapplication).
Figure 6. Temperature dependence of citronellal (2) selectivity during the
hydrogenation of citral (7.5 mmol) in 30 mL n-heptane (catalyst: 200 mg,
0.24 mol% Pd; see Table 1; p=0.4 MPa, t=90 min; microwave-assisted heat-
ing).
ChemSusChem 2011, 4, 1654 – 1661
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
1659

A. Stolle et al.
Standard preparation of the palladium catalyst:[22d]
reflecting a stable ability to inhibit consecutive hydrogena-
tions. While CAT-1a was deactivated, no significant loss in con-
version was observed using CAT-2a. In the case of CAT-2a,
nearly quantitative yields could be obtained in all five experi-
ments. Long time stability experiments in continuous flow re-
actors with SCILL-catalysts consisting of Pd/SiO2 and [C4mim]-
[DCA] maintained consistent activities for 78 h.[45] Due to its
consistent performances in batch trials, at least the acetate-
containing CAT-2 should be tested in further experiments, in-
cluding continuous flow applications. With a conventionally
heated steel-autoclave, similar results could be obtained by
preheating the reaction mixture under nitrogen atmosphere
(Figure 8). The time needed to reach the reaction temperature
would be longer than under microwave-assisted conditions,
but it might avoid the formation of unwanted side products
during the heating-up.
Pd(OAc)2 (0.09 mmol, 20 mg) was dissolved in dichloromethane
(50 mL) in a round-bottomed flask, and the porous glass support
(1 g) was then added to the mixture. After removal of the solvent
under vacuum, the catalyst precursor was calcined in air atmos-
phere for 2 h at 3008C in a muffle furnace (mls 1200 pyro; MLS
GmbH, Leutkirch, Germany), to afford a catalyst with a loading of
1 wt% Pd.
Standard preparation of CAT-x (see Table 1):[8]
1 g of the palladium catalyst was suspended in a solution of either
52 or 104 mg ionic liquid in 15 mL acetone and sonicated in an ul-
trasound bath for 10 min at room temperature. Afterwards, the sol-
vent was evaporated under vacuum (500 mbar) at 408C. The cata-
lysts had a loading of 5 wt% (CAT-xb series) or 10 wt% ionic liquid
(CAT-xa series).
Catalyst characterization
XPS spectra were recorded with a Quantum 2000 (PHI Co., Chan-
hassen, USA) apparatus with a focused monochromatic AlKa source
(1486.7 eV) for excitation. The pass energy was 58.7 eV, and the ref-
erence was the C 1s peak of the CÀC bond. Specific surface areas
(BET) and pore volumes (BJH) were determined using an Autosorb
(Quantachrome GmbH, Germany) with N2 as the adsorbent. Prior
to measurement the catalysts were tempered for 1 h at 1508C.
Thermal analyses were performed on a DTG-60 (Shimadzu, Japan)
with a Pt-cell and a heating rate of 10 KminÀ1 to a maximum tem-
perature of 6508C in a nitrogen atmosphere.
Conclusion
Microwave-assisted liquid-phase hydrogenation of citral was
applied using a catalyst consisting of palladium on a porous
glass support under mild conditions. After testing several sol-
vents, the best results were achieved in non-polar n-heptane.
In the next step a multitude of ionic liquids was added to the
reaction mixture to identify those with a positive effect on se-
lectivity towards citronellal. The Pd/TP catalyst was coated
with either [C2mim][DCA], [C2mim][OAc] or [C2mim][TfOAc]
to form SCILL-catalysts. By varying the reaction conditions, it
was possible to show that conversion of citral increased along
with increasing temperature and pressure. CAT-1a (10 wt%
[C2mim][DCA]) and CAT-2a (10 wt% [C2mim][OAc]) were able
to inhibit consecutive hydrogenations, thus achieving selectivi-
ties >95% and nearly quantitative yields. In contrast to the
catalysts just mentioned, CAT-3a (10 wt% [C2mim][TfOAc]) did
not eliminate the formation of dihydrocitronellal at high con-
version levels. In recycling experiments, CAT-2a remained
active for five runs while CAT-1a lost activity. However, both
catalysts could achieve excellent selectivity after five applica-
tions. In comparison with conventional heating, microwave-as-
sisted heating is faster and also applicable to SCILL-catalysts.
Regarding high throughput systems, a pre-heated convention-
al steel-autoclave can achieve similar results.
Standard reaction conditions
Citral (1, 7.5 mmol, 1.14 g), 100 mL hexadecane (internal standard),
the catalyst (200 mg; 0.24 mol% Pd), and n-heptane (30 mL) were
put into a quartz reactor vessel. The vessel was placed in the mi-
crowave reactor (type: multisynth, MLS GmbH, Leutkirch, Germany)
and connected to the hydrogen reservoir. The use of a gas inlet al-
lowed the direct flushing of dihydrogen into the reaction mixture.
The gas was continuously let off over a gas outlet valve to get a
uniform gas flow (0.2 mLminÀ1) through the reaction mixture. The
liquid-phase hydrogenations were carried out at various pressures
(0.2–1.0 MPa) and temperatures (30–908C). Samples were taken at
30 min intervals. The maximum power input was set to 400 W. The
hydrogenation process was observed by analyzing the samples
with GC-FID and GC-MSD.
Recycling experiments
After hydrogenation, the catalyst was separated from the reaction
mixture by filtration and washed with small amounts of n-heptane.
The spent catalyst system was air-dried before re-application. This
procedure was repeated three times (i.e., five reactions).
Experimental Section
General remarks
Citral, solvents, Pd(OAc)2, [C4mim][SCN], and [C4mim][OAc] were
purchased from Sigma–Aldrich. Citral used in this study consisted
of a 1:1 mixture of the E- and Z-isomer. [C4mim][BF4], [C4mim][I],
as well as [C2mim][OAc] and [C2mim][DEP] were purchased from
IoLitec and Solvent Innovation, respectively. All other ionic liquids
were purchased from Merck. Reagents and solvents other than
ionic liquids were used as received. The ionic liquids were dried
prior to their application.
Product analyses
Analyses of the reaction mixtures were carried out by GC-FID (HP
6890) and GC-MSD (HP 6890 N/5973 MS). GC-FID: HP 5, 30 mꢁ
0.32 mmꢁ0.25 mm, H2 5 psi, program: 358C (hold 1 min), 4 KminÀ1
up to 808C, 4.5 KminÀ1 up to 908C, 35 KminÀ1 up to 2808C (hold
for 3 min), injector temperature: 2508C, detector temperature:
1660
www.chemsuschem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2011, 4, 1654 – 1661

Partial Hydrogenation of Citral by using Ionic Liquid-Coated Porous Glass Catalysts
[16] a) J.-P. T. Mikkola, J. Aumo, D. Yu, T. Salmi, Catal. Today 2005, 105, 325–
2808C. GC-MSD: HP 5, 30 mꢁ0.32 mmꢁ0.25 mm, He 7 psi, pro-
gram: 558C (hold 1 min), 5 KminÀ1 up to1508C, 20 KminÀ1 up to
2808C (hold for 5 min), injector temperature: 2808C, EI (70 eV). The
products were identified by comparing both their retention times
and mass spectra with authentic samples. Conversions and selec-
tivities were calculated from the corrected peak areas using hexa-
decane as the internal standard.
330; b) J.-P. T. Mikkola, P. Virtanen, H. Karhu, T. Salmi, D. Y. Murzin, Green
Chem. 2006, 8, 197–205.
[17] a) C. O. Kappe, D. Dallinger, Mol. Diversity 2009, 13, 71–193; b) C. O.
Kappe, Chem. Soc. Rev. 2008, 37, 1127–1139; c) V. Polshettiwar, R. S.
Varma, Acc. Chem. Res. 2008, 41, 629–639; d) Microwaves in Organic
Synthesis (Ed.: A. Loupy), Wiley-VCH, 2006, Weinheim; e) C. Schmçger,
A. Stolle, W. Bonrath, B. Ondruschka, Curr. Org. Chem. 2011, 15, 151–
167.
[18] V. Polshettiwar, M. N. Nadagouda, R. S. Varma, Aust. J. Chem. 2009, 62,
16–26.
[19] C. O. Kappe, Angew. Chem. 2004, 116, 6408–6443; Angew. Chem. Int. Ed.
2004, 43, 6250–6284.
Acknowledgements
The authors thank B. Fꢀhndrich and N. Blaubach (both ITUC) for
thermogravimetric analyses, A. Schmidt (Institute for Physical
Chemistry, Friedrich-Schiller University Jena) for BET measure-
ments, R. Wagner (IMT) for conducting the XPS measurements,
and Dr. A. Stark (ITUC) for fruitful discussions. A.S. and B.O. ap-
preciate the cooperation of MLS GmbH (Leutkirch, Germany) for
providing and developing the microwave autoclave system.
[20] a) J. Hoffmann, M. Nꢄchter, B. Ondruschka, P. Wasserscheid, Green Chem.
2003, 5, 296–299;b) J.-M. Leveque, G. Cravotto, Chimia 2006, 60, 313–
320; c) E. Boros, K. P. Seddon, C. R. Strauss, Chimica Oggi 2008, 26, 28–
30.
[21] R. Martinez-Palou, J. Mex. Chem. Soc. 2007, 51, 252–254.
[22] a) M. Nꢄchter, B. Ondruschka, DE 102004033601, 2004; b) J. Freitag, M.
Hermann, M. Nꢄchter, B. Ondruschka, F. Schneider, A. Stolle, Opt. Appl.
2005, 35, 745–752; c) A. Stolle, C. Brauns, M. Nꢄchter, B. Ondruschka,
W. Bonrath, M. Findeisen, Eur. J. Org. Chem. 2006, 3317–3325; d) C.
Schmçger, A. Stolle, W. Bonrath, B. Ondruschka, T. Keller, K. D. Jandt,
ChemSusChem 2009, 2, 77–82; e) C. Schmçger, A. Stolle, B. Ondruschka,
W. Bonrath, WO2010020671A1, 2010.
[23] C. Schmçger, T. Szuppa, A. Stolle, F. Schneider, B. Ondruschka, A. Tied,
ChemSusChem 2008, 1, 339–347.
[24] J. Li, A. W.-H. Mau, C. R. Strauss, Chem. Commun. 1997, 1275–1276.
[25] J. Arras, D. Ruppert, P. Claus, Chem. Ing. Tech. 2009, 81, 2007–2011.
[26] J. Arras, E. Paki, C. Roth, J. Radnik, M. Lucas, P. Claus, J. Phys. Chem. C
2010, 114, 10520–10526.
Keywords: hydrogenation
chemistry · supported catalysts · porous materials
·
ionic liquids
·
microwave
[1] a) W. Bonrath, T. Netscher, Appl. Catal. A: Gen. 2005, 280, 55–73; b) W.
Bonrath, M. Eggersdorfer, T. Netscher, Catal. Today 2007, 121, 45–57.
[2] K. A. D. Swift, Top. Catal. 2004, 27, 143–155.
[3] M. Chatterjee, Y. Ikushima, T. Yokohama, M. Sato, Adv. Synth. Catal.
2008, 350, 624–632.
[27] K. Noack, H. Zbinden, R. Schlçgl, Catal. Lett. 1990, 4, 145–156.
[28] M. M. T. Khan, S. Srivastava, Polyhedron 1988, 7, 1063–1065.
[29] A. Stark, P. Behrend, O. Braun, A. Mꢄller, J. Ranke, B. Ondruschka, B. Jas-
torff, Green Chem. 2008, 10, 1152–1161.
[30] J.-P. T. Mikkola, P. P. Virtanen, K. Kordꢅs, H. Karhu, T. O. Salmi, Appl. Catal.
A: Gen. 2007, 328, 68–76.
[4] A. F. Trasarti, A. J. Marchi, C. R. Apesteguia, J. Catal. 2007, 247, 155–165.
[5] Y. Nie, S. Jꢂnicke, G.-K. Chuah, Chem. Eur. J. 2009, 15, 1991–1999.
[6] J. Aumo, J. Warna, T. Salmi, D. Yu. Murzin, Chem. Eng. Sci. 2006, 61,
814–822.
[7] K.-C. Tin, N.-B. Wong, R.-X. Li, X.-J. Li, J. Mol. Catal. A: Chem. 1999, 137,
113 – 119.
[31] J. Arras, D, Ruppert, P. Claus, Appl. Catal. A: Gen. 2009, 371, 73–77.
[32] A. Stadler, C. O. Kappe, Org. Lett. 2002, 4, 3541–3543.
[33] D. Bogdal, M. Lukasiewicz, J. Pielichowski, A. Miciak, S. Bednarz, Tetrahe-
dron 2003, 59, 649–653.
[34] X. Zhang, D. O. Hayward, Inorg. Chim. Acta 2006, 359, 3421–3433.
[35] W. L. Perry, J. D. Katz, D. Rees, M. T. Paffet, A. K. Datye, J. Catal. 1997,
171, 431–438.
[36] Handbook of Chemistry and Physics, 78th ed. (Ed.: D. R. Lide), CRC Press,
Boca Raton, 1998, pp. 15–14–15–18.
[37] C. O. Kappe, A. Stadler, Microwaves in Organic and Medicinal Chemistry,
Wiley-VCH, Weinheim, 2005, p. 13.
[8] J. Arras, M. Steffan, Y. Shayeghi, D. Ruppert, P. Claus, Green Chem. 2009,
11, 716–723.
[9] a) C. Reichardt, Solvents and Solvent Effects in Organic Chemistry, 3rd ed.,
Wiley-VCH, Weinheim, 2003, pp. 322–324; b) H. Olivier-Bourbigou, L.
Magna, D. Morvan, Appl. Catal. A: Gen. 2010, 373, 1–56.
[10] R. Giernoth, Angew. Chem. 2010, 122, 2896–2901; Angew. Chem. Int. Ed.
2010, 49, 2834–2839.
[11] a) J. Dupont, R. F. de Souza, P. A. Z. Suarez, Chem. Rev. 2002, 102, 3667–
3692; b) T. M. Potewar, S. A. Ingale, K. V. Srinivasan, Tetrahedron 2007,
63, 11066–11069.
[12] a) J. Mo, L. Xu, J. Xiao, J. Am. Chem. Soc. 2005, 127, 751–760; b) E. J.
Leonard¼o, J. de Oliveira Feijꢃ, S. Thurow, G. Perin, R. G. Jacob, C. C. Sil-
veira, Tetrahedron Lett. 2009, 50, 5215–5217; c) C. Wang, W. Guan, P.
Xie, X. YYun, X. Hua, Y. Wang, Catal. Commun. 2009, 10, 725–727.
[13] a) A. Riisager, K. B. Eriksen, P. Wasserscheid, R. Fehrmann, Catal. Lett.
2003, 90, 149–153; b) A. Riisager, P. Wasserscheid, R. Fehrmann, S. Flick-
er, R. Van Hal, M. Haumann, Angew. Chem. 2005, 117, 826–830; Angew.
Chem. Int. Ed. 2005, 44, 815–819; c) C. Van Doorslaer, J. Wahlen, P. Mert-
ens, K. Binnemans, D. De Vos, Dalton Trans. 2010, 39, 8377–8390.
[14] A. Riisager, P. Wasserscheid, R. van Hal, R. Fehrmann, J. Catal. 2003, 219,
452–455.
[15] a) M. H. Valkenberg, C. DeCastro, W. F. Hçlderich, Appl. Catal. A: Gen.
2001, 215, 185–190; b) C. P. Mehnert, R. A. Cook, C. Dispenziere, M. Afe-
worki, J. Am. Chem. Soc. 2002, 124, 12932–12933; c) U. Kernchen, B.
Etzold, W. Korth, A. Jess, Chem. Ing. Tech. 2007, 79, 807–819; d) J. Arras,
M. Steffan, Y. Shayeghi, P. Claus, Chem. Commun. 2008, 4058–4060;
e) A. Jess, W. Korth, B. Etzold, WO 2007124896, 2007.
[38] Y. Gu, G. Li, Adv. Synth. Catal. 2009, 351, 817–847.
[39] P. Virtanen, H. Karhub, K. Kordꢅs, J.-P. T. Mikkola, Chem. Eng. Sci. 2007,
62, 3660–3671.
[40] N. V. Plechkova, K. R. Seddon, Chem. Soc. Rev. 2008, 37, 123–150.
[41] G. Corro, C. Cano, J. L. G. Fierro, J. Mol. Catal. A: Chem. 2010, 315, 35–
42.
[42] Catalysis from A to Z, Vol. 2 (Eds.: B. Cornils, W. A. Herrmann, M. Muhler,
C.-H. Wong), Wiley-VCH, 2007, Weinheim, p. 639.
[43] M. Burgener, R. Furrer, T. Mallat, A. Baiker, Appl. Catal. A: Gen. 2004, 268,
1–8.
[44] R. Lungwitz, M. Friedrich, W. Linert, S. Spange, New J. Chem. 2008, 32,
1493–1499.
[45] N. Wçrtz, J. Arras, P. Claus, Appl. Catal. A: Gen. 2011, 391, 3319–324.
Received: March 22, 2011
Published online on August 8, 2011
ChemSusChem 2011, 4, 1654 – 1661
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
1661