Aldol Condensations Catalyzed by Basic Anion-Exchange Resins

Article abstract of DOI:10.1002/cctc.201600867

The aldol condensations of various aldehydes with ketones in the presence of anionic (basic) ion-exchange resins have been investigated. Both batch and continuous modes were studied and compared for the reaction of citral (a mixture of geranial and neral) with acetone to give ψ-ionone. Different reaction conditions were investigated, and the performances of five different ion-exchange resins were compared. The most stable resins could be used for 10 cycles in batch mode or 1000 min in continuous mode without a significant loss in activity or selectivity.

Full text of DOI:10.1002/cctc.201600867

Accepted Article
Title: Aldol Condensations Catalysed by Basic Anion Exchange Resins
Authors: Jan Schütz, Werner Bonrath, Yann Pressel, Erich Ferfecki, and
Klaus-Dieter Topp
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201600867
Link to VoR: http://dx.doi.org/10.1002/cctc.201600867
A Journal of
www.chemcatchem.org

10.1002/cctc.201600867
ChemCatChem
FULL PAPER
Aldol Condensations Catalysed by Basic Anion Exchange Resins
Werner Bonrath,[a] Yann Pressel,[a] Jan Schütz,*[a] Erich Ferfecki,[b] Klaus-Dieter Topp[c]
Abstract: Aldol condensations of various aldehydes with ketones in
the presence of anionic (basic) ion exchange resins have been
investigated. Both batch and continuous mode were studied and
compared for the reaction of citral (a mixture of geranial and neral)
with acetone to give y-ionone. Varying reaction conditions have been
investigated and the performance of five different ion exchange resins
compared. The most stable resins could be used for ten cycles in
batch mode or 1000 min in continuous mode without significant loss
in activity and selectivity.
Aldol condensations are catalysed by acids and bases.[2-6] The
most commonly applied catalysts, especially in industrial
applications are Brønsted bases. In the synthesis of y-ionone
these are most often potassium,[7] sodium[7-10] or barium
hydroxide[11,12]. The bases are often supported and applied in
liquid phase reactions. Strong bases, such as sodium ethoxide
and quaternary ammonium bases are also described as efficient
catalysts for this reaction type.[13,14] It was demonstrated that the
Aldol condensation can also be catalysed using weak or strong
basic ion exchange resins, such as supported secondary or
tertiary amines.[15-19] The range of basic catalysts for aldol
condensation is rather extensive. In addition to rare earth oxides
and other basic oxides,[20] phase transfer catalysts,[21] lithium
iodide,[22] KF on Al2O3,[23] and in the past fifteen years
hydrotalcites[24] have been successfully applied as catalysts.
Currently, citral is reacted on industrial scale with acetone to y-
ionone via an aldol condensation in a liquid bi-phasic process
applying a homogeneous base in aqueous phase. This process
has several disadvantages related to corrosion, disposal of spent
base materials and yields and selectivity which reach about
85%.[25]
Introduction
Aldol condensation reactions are fundamentally important
carbon-carbon bond forming reactions and are widely employed
in the synthesis of complex organic compounds. One example for
this transformation is the industrially significant condensation of
acetone with citral (a mixture of geranial and neral) to y-ionone
(Scheme 1). y-Ionone is an important intermediate in the
synthesis of isoprenoid compounds such as vitamin A, vitamin E
and of several carotenoids (b-carotene, lutein, cantaxanthin,
astaxanthin) as well as a valuable compound for flavours and
fragrance chemistry.[1]
Compared to solid acid catalysed processes the application of
solid base catalysts in industrial processes is far more limited.[26]
It is only recently that there has been renewed interest in solid
base catalysts. The new procedures and processes have various
advantages compared to liquid bases: less waste formation, easy
product and/or catalyst separation, and solvent recovery. Thus,
solid base catalysts offer often more environmentally benign and
more economical pathways for the production of fine chemicals.
The main problem of solid base catalysts, in contrast to solid acid
catalysts, is often stability which lowers the catalyst’s lifetime
significantly and limits their application.
O
O
O
Geranial
Neral
y-Ionone
O
Citral: mixture of geranial and neral
Here we describe our results obtained in a detailed study of
anionic (basic) ion exchange resins in the cross aldol
condensation of various aldehydes and ketones. We investigated
especially the aldol condensation of acetone and citral to y-
ionone.
Scheme 1. Aldol condensation of citral and acetone to y-ionone.
Results and Discussion
[a]
Dr. W. Bonrath, Dr. Y. Pressel, Dr. Jan Schütz
DSM Nutritional Products
Five strongly basic anion exchange resins were applied as
catalysts in aldol condensation reactions. All resins have
trimethylamines covalently supported on the polymer resulting in
quaternary ammonium ions. Two resins are polystyrene based
and the other three polyacryl based (Table 1). One of the resins
is a gel-type, the others are macroporous. AMBERLYSTTM 4 is a
bifunctional resin. Next to the quaternary ammonium ions there
are secondary amines linked to the polymer which are
functionalized with an antioxidant group. Possible oxidative
Research and Development, Process Research
P.O. Box 2676, CH-4002 Basel, Switzerland
E-mail: jan.schuetz@dsm.com
[b]
[c]
E. Ferfecki
Dow Deutschland Anlagengesellschaft mbH
Industriestr. 1, 77836 Rheinmünster, Germany
Dr. K.-D. Topp
Dow Deutschland Vertriebs GmbH & Co. OHG
Am Kronberger Hang 4, 65824 Schwalbach, Germany
This article is protected by copyright. All rights reserved.

10.1002/cctc.201600867
ChemCatChem
FULL PAPER
degradation of the resin and oxidation of aldehydes to their
corresponding acids could thus be reduced.
We started our investigation with the aldol condensation of citral
and acetone to y-ionone. Acetone was applied in excess (from 1
to 8 equiv.) and in some cases alcohols were added as solvent.
The resins were recycled after a batch reaction and washed with
an alcohol prior re-use.
Table 1. Anion exchange resins investigated.
Moisture
holding
capacity
[%]
Conc. of
active sites
[eq/L]
Resin
Matrix
Type
Figure 1. Conversion of citral in batch reactions of the aldol condensation of
citral and acetone to y-ionone catalysed by anion exchange resins; 40 °C, 3 h,
AMBERLYSTTM
A26 OH
1.0
form)
(OH
(OH
(OH
(OH
polystyrene[a]
polyacryl
68
67
70
70
69
macroporous
gel
AMBERLYSTTM
1-4
and
AMBERLYSTTM
A26OH:
molar
ratio
acetone/citral/ethanol 4.0 : 1.0 : 77, AMBERLYSTTM 3, conc. and 4 conc.: molar
ratio acetone/citral/ethanol 8.4 : 1.0 : 0.7: ion exchange resin: 41 wt% with
respect to a mixture of citral and acetone.
AMBERLYSTTM
AMBERLYSTTM
AMBERLYSTTM
AMBERLYSTTM
1
2
3
4
1.0
form)
0.8
form)
polyacryl
macroporous
macroporous
macroporous
0.8
form)
polyacryl
polystyrene[a]
0.97 (OH
form)
[a] surface area: 30 m2/g, average pore diameter: 290 Å, total pore volume: 0.20
ml/g
The resins were remarkably active. The conversion of citral was
in general >95% in the first run (Figure 1). The selectivity to y-
ionone was >70% for AMBERLYSTTM 4 and around 60% for the
other resins (Figure 2). Running the reaction with small amounts
of solvent (ethanol or isopropanol) proved to be beneficial for
activity, selectivity, and lifetime of the resins. Under these reaction
conditions for up to 10 runs no activity and selectivity loss was
observed. Calculated on the active sites of the resins, the turn
over number (TON) for AMBERLYSTTM 3 and 4 with small
amounts of solvent, are 30 and 27 respectively (calculated for 15
and 17 runs respectively). The TON of AMBERLYSTTM 4 is likely
to be higher because the resin was still active after the 17th run.
Diluting the reaction mixture with more solvent led to a decrease
in activity, selectivity, and lifetime of the resins. Decrease in
selectivity is partly due to the formation of acetals from citral and
ethanol.
Figure 2. Selectivity (based on citral) of batch reactions of the aldol
condensation of citral and acetone to y-ionone catalysed by anion exchange
resins; 40 °C, 3 h, AMBERLYSTTM 1-4 and AMBERLYSTTM A26OH: molar ratio
acetone/citral/ethanol 4.0 : 1.0 : 77, AMBERLYSTTM 3, conc. and 4, conc.: molar
ratio acetone/citral/ethanol 8.4 : 1.0 : 0.7: ion exchange resin: 41 wt% with
respect to a mixture of citral and acetone.
The E/Z ratio of citral used in the experiments was about 63/37
(geranial/neral). This ratio changed in the obtained product y-
ionone slightly over the tested runs (Figure 3). The ratio was
monitored for AMBERLYSTTM 3 and 4 under the same reaction
conditions as described under Figure 1. The newly formed double
bond in position 3 of y-ionone was almost exclusively in the E-
configuration (98/2). The slight variations of the E/Z configuration
at position 5 of y-ionone are most likely due to a possible slightly
higher reactivity of geranial compared to neral. Under reaction
conditions favouring a thermodynamic rather than a kinetic
reaction course, all-E product is thermodynamically favoured.
With lower activity of the catalyst after each run this effect
becomes more pronounced forming slightly more 5E-y-ionone
than 5Z-y-ionone.
Oxidative degradation was not a major contributor to the
deactivation of the resins. AMBERLYSTTM 4 with antioxidant
groups did not outperform the other resins significantly.
Furthermore, there were no significant changes observed if the
reaction was run under inert conditions or under air.
Insoluble compounds were found on the resins after the reaction
which were self-condensation products of citral and condensation
of several citral moieties with y -ionone.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201600867
ChemCatChem
FULL PAPER
Figure 5. Selectivity to y-ionone in batch reactions of the aldol condensation of
citral and acetone to y-ionone catalysed by AMBERLYSTTM A26OH with
various acetone equivalents; 40 °C, 3 h, molar ratio citral/ethanol 1.0 : 77.
Figure 3. 5E-y-Ionone formed in batch reactions of citral and acetone under the
reaction conditions described in Figure 1.
The spent catalyst (from the acetone:citral 4:1 experiment) was
further analysed after the 7th run. The total sites by weight
decreased from 4.7 (fresh AMBERLYSTTM A26OH) to 3.4 eq/kg.
This is due to insoluble compounds, most likely polycondensed
products from acetone and citral, which partly cover the resins
pores and particle surfaces.[27] From the total amount of active
sites (OH- form) only 6% remained after seven runs, suggesting
neutralization and/or Hofmann elimination[28] of the active sites.
The observed decrease in selectivity (Figure 2 and Figure 5) could
result from these polycondensed products on the resin’s surface
and pores which could not be washed off. Citral could further
condensate onto these products and thus lowering the selectivity
to y-ionone.
In order to study the influence of the acetone excess on the
activity and selectivity AMBERLYSTTM A26OH was used as
catalyst with one to eight equivalents of acetone with regard to
citral. The larger acetone excess increased the conversion of
citral (Figure 4). However, the lifetime of the catalysts was not
improved. With all tested acetone equivalents no conversion was
observed after the 7th run. Selectivity however improved with more
acetone due to less self-condensation of citral (Figure 5).
Treating used ion exchange resin with base should improve
activity in case of deactivation via neutralization. After five runs
the resin was treated with 5 M aqueous NaOH. Some activity
could be regained (Figure 6) but not the initial value suggesting
that deactivation is also occurring via different pathways such as
Hofmann elimination.
Figure 4. Conversion of citral in batch reactions of the aldol condensation of
citral and acetone to y-ionone catalysed by AMBERLYSTTM A26OH with
various acetone equivalents; 40 °C, 3 h, molar ratio citral/ethanol 1.0 : 77.
Figure 6. Reactivation of the ion exchange resin after the 5th run, conversion
and selectivity (based on citral) of batch reactions of the aldol condensation of
citral and acetone to y-ionone catalysed by AMBERLYSTTM A26OH; 40 °C, 3 h,
molar ratio acetone/citral/ethanol 4.0 : 1.0 : 77, ion exchange resin: 69 wt% with
respect to a mixture of citral and acetone.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201600867
ChemCatChem
FULL PAPER
The activity and selectivity of the ion exchange resins were
compared to some hydrotalcites. Applying the same reaction
conditions (one run) as described in Figure 2 for AMBERLYSTTM
3, conc. and 4, conc. lower activity and selectivity compared to
carbonyl group of the aldehyde by the carbanion generated from
methyl group of the ketone (see Scheme S1).[29] However, this
effect is strongly influenced by reaction temperature and reaction
time (see for example entry 4 and 5). Thus, bases with lower
basicity and longer reaction times and/or higher reaction
temperature favour linear products.
Weakly basic ion exchange resins such as AMBERLITETM IRA96
(with a dimethylamine group covalently supported on the polymer
resulting in tertiary amines.) were not basic enough to form the
required enolate (entry 2 and 6). Thus, no conversion was
observed.
In agreement to the results presented in Figure 4 and Figure 5 it
was in most cases beneficial for the selectivity to have a larger
excess of ketone (entry 3 vs. 4 and entry 10 vs. 11). Running the
reactions more concentrated resulted also in higher selectivity,
reducing possible side-reaction of citral with the alcoholic solvent
(entry 4, 14, and 16). 3,3,5-Trimethylcyclohexanone did not react
with citral (entry 15). Apparently steric hindrance does not allow
the basic attack on the a-proton. Reducing the steric hindrance
by removing methyl groups (to 3,5-dimethylcylohexanone and
cyclohexanone, entry 16 and 14) resulted in aldol condensation
product. In many cases small amounts of aldol addition product
(b-hydroxy ketones) formed as well.
AMBERLYSTTM
3
conc. and 4, conc. were observed.
Commercially available PURAL® MG 30, 50, and 70 (from Sasol,
aluminum magnesium hydroxy carbonates) were treated with
water in order to exchange the carbonate ion by a hydroxide ion.
After the treatment they were tested in the aldol condensation of
acetone and citral. The conversions were 38%, 18%, and 25%
respectively and the selectivity 34%, 8%, and 21% respectively.
Without ion exchange PURAL® MG 30 showed 28% conversion
and 30% selectivity.
The aldol condensation of citral and acetone was extended to
further ketones (Table 2).
Remarkably AMBERLYSTTM A26OH catalysed unsymmetrical
ketones and citral predominantly to linear products (entry 3-11),
whereas with sodium ethanolate as base an almost 1:1 mixture of
linear and branched product formed (entry 7). The higher basicity
of the ethanolate is most likely the reason for the lower linear
selectivity, activating both the more reactive a-proton (at the
methyl group) and the less reactive one (at the methylene
group).[29] A further reason for the predominance of linear product
is the steric hindrance which kinetically favours an attack on the
This article is protected by copyright. All rights reserved.

10.1002/cctc.201600867
ChemCatChem
FULL PAPER
Table 2. Aldol condensations of citral and various ketones.
Aldol condensation product (branched/linear where
applicable)
Conversion
[%]
Selectivity
br./lin. [%]
Temp.
[°C]
Reaction
time [h]
Entry
Ketone [equiv.]
1
2
3
4
5
6
7
Acetone, 8
99
0
67
40
40
40
40
40
40
3
8
8
8
4
8
5
Acetone, 8[a]
-
2-Butanone, 4
2-Butanone, 8
2-Butanone, 8
2-Butanone, 8
2-Butanone, 6.3[f]
>99
>99
>99
0
24/37[b]
30/48[c]
23/66[d]
-
-10 to
22
96
46/49
8
2-Pentanone, 4
2-Pentanone, 4[h]
2-Pentanone, 4[h]
2-Pentanone, 8
81
74
80
94
4/13[g]
7/15[g]
7/14[g]
17/29
40
40
40
60
3
3
6
3
9
10
11
12
13
14
3-Pentanone, 4
3-Pentanone, 8
Cyclohexanone, 8[e]
>99
96
29[i]
30[j]
80[g]
40
40
40
24
24
24
>99
3,3,5-Trimethyl-
cylohexanone, 8
self-condensation of
citral[g]
15
16
40
40
24
24
no product formed
>99
64[g]
3,5-Dimethyl-
cylohexanone, 8[e]
Reaction conditions if not otherwise stated: solvent: ethanol, molar ratio citral/ethanol 1.0 : 77, 169 wt% AMBERLYSTTM A26OH with
regard to citral; the aldol condensations products y-ionone and 7,11-dimethyl-4,6,10-dodecatrien-3-one/ 3,6,10-trimethyl-3,5,9-
undecatrien-2-one (product from citral and 2-butanone) were analysed by wt%, all other products by area%, products were identified by
GC-MS; br. = branched aldol condensation product, lin. = linear aldol condensation product.
[a] Catalyst: AMBERLITETM IRA96 [b] Selectivity: 7% aldol addition product (linear and branched b-hydroxy ketones) [c] Selectivity: 5%
aldol addition product (linear and branched) [d] Selectivity: <1% aldol addition product [e] No solvent, 28wt% resin with regard to citral [f]
Catalyst: 0.5 equiv. NaOEt [g] Selectivity: <1% aldol addition product [h] Catalyst re-used from entry 8 and 9 respectively [i] Selectivity:
9% aldol addition product [j] Selectivity: 10% aldol addition product.
Further aldol condensations were performed with aldehydes other
than citral in presence of the resin AMBERLYSTTM A26OH in
order to study the scope of the catalyst (Table 3). Prenal was
reacted with 2-butanone (entry 1-3). Similarly to citral mainly
linear product was formed with an ion exchange resin as catalyst
while sodium ethanolate as catalyst resulted in almost a 1:1
mixture of linear and branched product (entry 7). The reaction rate
could be increased up by raising the reaction temperature from
40 to 60 °C (entry 1 and 2). However, basic ion exchange resins
are usually temperature sensitive because of increased Hofmann
elimination of the active amine groups at elevated temperatures.
The manufacturer (The Dow Chemical Company) of
AMBERLYSTTM A26OH recommends a maximum operating
temperature of 60 °C in order to keep the deactivation level low.
[30] It is therefore most desirable to either use basic ion exchange
resins at low reaction temperatures (requiring high activity of the
This article is protected by copyright. All rights reserved.

10.1002/cctc.201600867
ChemCatChem
FULL PAPER
resins) or to have a high temperature stable basic ion exchange
resin.
We also investigated the cross aldol condensation of
isovaleraldehyde with 2-butanone to give 6-methyl-3-hepten-2-
one and found that increasing the reaction temperature increased
conversion and yield (entry 6-8). At lower temperature significant
amounts of the aldol addition products formed (entry 6 and 7).
Apparently the energetically driven dehydration requires higher
reaction temperatures to reach equilibrium more on the side of the
condensed products. Increasing the excess of 2-butanone from 8
to 16 equivalents did not significantly change the outcome of the
reaction (entry 8-10). Running the reaction in the presence of
sodium ethanolate resulted in more branched product compared
to reactions catalysed by ion exchange resins. However, it is not
clear why more branched product is formed than linear. This is in
contrast to the results observed with citral (Table 2) and prenal
(Table 3, entry 1-3) and unsymmetrical ketones. The same
branched/linear preference was observed with heptanal and 2-
butanone (entry 24 and 25). Running the reaction under neat
conditions, in ethanol, or cyclohexane did not change the
branched/linear ratio significantly (entry 5, 8 and 12).
In the aldol condensation between heptanal and acetone an
increase in conversion was observed when the reaction
temperature was increased (entry 13 -17) whereas the selectivity
was stable. Increasing the excess of acetone did not change
significantly the outcome of the reaction (entry 17-19). Diluting the
reaction with ethanol increased the conversion but decreased
selectivity (entry 20 and 21) whereas in cyclohexanol the opposite
trend was observed (entry 23).
Cross aldol condensations with two aldehydes were tested with
isobutyraldehyde and formaldehyde (entry 26 and 27). The main
product was indeed the cross aldol condensation product prenal.
However, the selectivity was low due to aldol reactions and
condensations between the aldehydes themselves.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201600867
ChemCatChem
FULL PAPER
Table 3. Aldol condensations of various aldehydes and ketones.
Ketone/
Aldehyde
[equiv.]
Aldol condensation product
(branched/linear where
applicable)
Conv.
[%]
Selectivity (aldol
cond. products)
br./lin. [%]
Reaction
time [h]
Selectivity (aldol
add. products)
br.+lin. [%][a]
Entry
Aldehyde
Temp. [°C]
6
-
1
2-Butanone, 8
2-Butanone, 8
2-Butanone, 6.3[b]
2-Butanone, 4[c]
2-Butanone, 8[c]
2-Butanone, 8
2-Butanone, 8
2-Butanone, 8
2-Butanone, 12
2-Butanone, 16
2-Butanone, 6.3[b]
2-Butanone, 8[d]
Acetone, 8
>99
>99
97
>99
>99
76
83
97
97
98
>99
98
40
59
65
82
90
94
95
96
95
0
15/42
12/42
37/38
52/16
53/16
42/7
51/17
57/37
55/38
58/39
39/5
47/37
65
40
60
2
-
2
-10 to 22
40
5
-
3
24
8
17
17
40
23
5
3
2
7
5
-
4
40
5
30
8
6
40
8
7
60
8
8
60
8
9
60
8
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
-10 to 22
60
24
8
22
8
62
30
8
-
Acetone, 8
60
40
8
-
Acetone, 8
65
50
8
-
Acetone, 8
56
60
8
-
Acetone, 8
54
60
8
-
Acetone, 12
55
60
8
-
Acetone, 16
Acetone, 4[c]
41
40
8
-
Acetone, 8[c]
48
40
8
-
Acetone, 8[c,e]
Acetone, 8[d]
-
40
8
-
73
60
8
-
76
97
97
50
>99
2-Butanone, 4[c]
2-Butanone, 8[c]
Formaldehyde, 8[f]
Formaldehyde, 8[f]
25/4
37/5
52
40
8
16
22
-
40
8
40
24
24
45
60
-
Reaction conditions if not otherwise stated: no solvent: 50 wt% AMBERLYSTTM A26OH with regard to the aldehyde; all products were analysed by area% and
GC-MS; br. = branched aldol condensation product, lin. = linear aldol condensation product.
[a] aldol add. product: b-hydroxy ketones [b] Catalyst: 0.5 equiv. NaOEt, solvent: ethanol, molar ratio aldehyde/ethanol 1.0 : 30 [c] Solvent: ethanol, molar ratio
aldehyde/ethanol 1.0 : 77, 300 wt% AMBERLYSTTM A26OH [d] Solvent: cyclohexane, molar ratio aldehyde/cyclohexane 1.0 : 77, 50 wt% AMBERLYSTTM A26OH
[e] Catalyst: AMBERLITETM IRA96 [f] 37% formalin solution, no additional solvent, 300 wt% AMBERLYSTTM A26OH.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201600867
ChemCatChem
FULL PAPER
The aldol condensation of citral and acetone to y-ionone was also
studied in continuous reaction mode in a fixed-bed reactor.
Acetone was applied in access (9 equiv.) and isopropanol added
as additional solvent. Isopropanol was chosen as solvent to limit
the formation of acetals from the feed mix of citral and alcohol. In
Figure 7 five resins are compared with regard to the y-ionone
content in the reaction mixture. In contrast to batch mode (Figure
2 and Figure 3) in continuous mode AMBERLYSTTM A26OH and
AMBERLYSTTM 3 outperformed the other resins. This might be
due to the lower reaction temperature in continuous mode (22 °C
vs 40 °C). AMBERLYSTTM A26OH and AMBERLYSTTM 3 seem
to be more heat sensitive than the other resins which leads to
more deactivation by Hofmann elimination of the active centres of
the resins when the reaction is run at 40 °C. Oxidative degradation
of the resins did not play a significant role. AMBERLYSTTM 4 with
antioxidant groups did not perform better than the other resins.
In Figure 8 the two most stable resins AMBERLYSTTM 3 and
A26OH and the lower performing AMBERLYSTTM 4 are compared
by the yield of formed y-ionone. For the chosen reaction time the
yield was nearly constant with the resins AMBERLYSTTM 3 and
A26OH.
Figure 8. Yield of y-ionone in the continuous aldol condensation of citral and
acetone (fixed bed reactor) under the reaction conditions described in Figure 7.
AMBERLYSTTM A26OH was also studied as catalyst with different
amounts of acetone with regard to citral (Figure 9 and Figure 10).
There are no significant differences observed between four and
six equivalents of acetone. However with nine equivalents
acetone the deactivation of the catalyst is slowed down. The
experiment with one equivalent of acetone was carried out without
any additional solvent. Very fast deactivation was observed.
These results are in agreement with the ones observed in batch
mode (Figure 5 and Figure 6). Apparently the additional solvent
slows down deposition of polymeric by-products on the resin’s
surface and pores and also slows down Hofmann elimination
because of less thermal stress for the resins in the presence of
additional solvent.
Figure 7. Content of y-ionone in the continuous aldol condensation of citral and
acetone (fixed bed reactor), catalysed by anion exchange resins, 22 °C, molar
ratio acetone/citral/isopropanol 9 : 1 : 55; feed rate: 1 ml/min, inner diameter of
the reactor: 18 mm, WHSV=7.90 g(feed)/h/g(catalyst).
Applying AMBERLYSTTM 3, 4 and A26OH, the (5Z)/(5E) ratio of
y-ionone was constant (35:65). With AMBERLYSTTM A26OH, the
(3Z)/(3E) ratio was 2:98 for y-ionone isolated at the end of the
experiment. Thus, the thermodynamically favoured product
(3E,5E)-y-ionone was mainly formed. The (3Z)/(3E) ratio was not
determined for the other resins.
This observed Z/E ratios are in agreement with the results
obtained in batch mode.
Figure 9. Conversion of citral in the continuous aldol condensation of citral and
acetone (fixed bed reactor), catalysed by AMBERLYSTTM A26OH; 22 °C, molar
ratio acetone/citral/isopropanol 4,6 or 9: 1 : 55 and 1 : 1 : 0; feed rate: 1 ml/min,
inner diameter of the reactor: 18 mm, WHSV=7.90 g(feed)/h/g(catalyst).
This article is protected by copyright. All rights reserved.

10.1002/cctc.201600867
ChemCatChem
FULL PAPER
were concentrated in vacuo (40 °C, 40 mbar). The isolated orange oil was
analysed by GC (wt%). The yield was determined by considering all the
isomers (3E,5E)-y-ionone, (3Z,5E)-y-ionone, (3Z,5Z)-y-ionone and
(3E,5Z)-y-ionone).
Continuous aldol condensation of citral and acetone:
The continuous reactions were performed in a fixed bed reactor (Figure
11). The feed (citral, acetone and the solvent) were mixed and stored in a
bottle on a balance. The bottle was constantly purged with argon. A HPLC
(Gilson 305) pump conveyed the feed to the reactor. The reactor was kept
at room temperature (21-23 °C). The product flow was collected in an
argon purged bottle and the reaction was monitored by GC.
Figure 10. Content of y-ionone in the continuous aldol condensation of citral
and acetone (fixed bed reactor) under the reaction conditions described in
Figure 9.
Fixed-bed glass
reactor filled with ion
exchange resins
Conclusions
Gilson
pump
The cross aldol condensation of citral and ketones, especially
acetone to y-ionone was carried out in the presence of basic ion
exchange resins. The reactions were run in batch and continuous
mode. Over ten runs no significant loss in activity and selectivity
was observed for two of the tested resins. In continuous mode a
stable performance was noted for more than 15 h. A variety of
aldehydes and ketones was also successfully reacted to their
corresponding condensed aldol products. The tested systems
were superior to the state of the art. The long lifetime of the resins
and the successful application with a variety of substrates was so
far not yet reached. However, in all cases deactivation of the
resins was observed after several runs in batch mode. Analysis of
the spent resin revealed polycondensed products on the resin and
partly Hofmann elimination of the reactive centres of the resins. In
order to apply these resins in an industrial process their lifetime
has to be increased. Currently new resins are being developed
and application tests in our laboratories are ongoing.
Citral,
Products
acetone
and solvent
and solvent
Figure 11. Continuous experimental set-up.
The ion exchange resin (6.0 g, previously stirred in 10 ml ethanol for 15
min) was placed in a vertical fixed-bed glass reactor (length 25 cm, 18 mm
inner diameter). The level of a siphon was adjusted so that the solvent
level in the reactor was just over the catalyst bed (h = 12 cm, V = 30 ml).
A solution of citral, acetone and isopropanol (ratio: 9 : 1 : 55) was fed into
the reactor with a Gilson pump (1 ml/min, WHSV = 7.9 h-1). Samples were
taken regularly, concentrated in vacuo (40 mbar, 40 °C) and analysed by
GC (wt%). After 18 hours feeding, the solvent of the collected solution was
evaporated in vacuo (40 mbar, 40 °C) and the isolated yellow oil analysed
by GC (wt%). The yield was determined by considering all the isomers
(3E,5E)-y-ionone, (3Z,5E)-y-ionone, (3Z,5Z)-y-ionone, and (3E,5Z)-y-
ionone).
Experimental Section
All reactions were carried out under argon. Solvents, reagents, and ion
exchange resins were used as received.
y-Ionone isolated at the end of the experiment was analysed by 1H-NMR
to determine the (3Z)-y-ionone: (3E)-y-ionone ratio (2:98) with the
integration of the proton signals at position 4 of y-ionone (7.10-7.50 ppm).
All ion exchange resins were purchased from Rohm and Haas/ The Dow
Chemical Company. AMBERLYSTTM 1-4 are experimental ion exchange
resins from Rohm and Haas/ The Dow Chemical Company. PURAL® MG
30, 50, and 70 were purchased from Sasol.
1H NMR (400 MHz, CDCl3, 20 °C, TMS) of (3E,5E)-y-ionone:
d=1.61 (s, 3H; (CH3)2C), 1.68 (s, 3H; (CH3)2C), 1.91 (s, 3H; CH3CCH2),
2.17 (br. s, 4H; CH2CH2), 2.27 (s, 3H; CH3CO), 5.07 (br. s, 1H; CH(8)),
6.00 (d, 3J(H,H)=12 Hz, 1H; CH(5), 6.08 (d, 3J(H,H)=18 Hz, 1H; CH(3)),
7.43 (dd, 3J(H,H)=9 Hz and 3J(H,H)=15 Hz, CH(4)).
Standard procedure for the aldol condensation of citral and acetone
in batch reactions:
A basic ion exchange resin (3.00 g) was stirred in ethanol (50 ml) for 15
min. The resin was filtered under argon, ethanol (50 ml, 77 equiv.) was
added, and the reaction mixture heated to 40 °C. Acetone (2.59 g, 44.33
mmol, 4 equiv.) and citral (1.78 g, 11.08 mmol, 95%, 1 equiv.) were added.
After 3 h the reaction mixture was filtered under argon and the filter cake
was washed twice with ethanol (2 x 25 ml). The combined mother liquors
1H-NMR (400 MHz, CDCl3, 20 °C, TMS) of (3Z,5E)-y-ionone:[31]
1.61 (s, 3H, (CH3)2C), 1.68 (s, 3H, (CH3)2C), 1.86 (s, 3H, CH3CCH2), 2.18
(m, 4H, CH2CH2), 2.23 (s, 3H, CH3CO), 5.09 (br. s, 1H, CH(8)), 5.93 (d,
This article is protected by copyright. All rights reserved.

10.1002/cctc.201600867
ChemCatChem
FULL PAPER
1H, 3J(H,H)=11.5 Hz, CH(5)), 6.73 (t, 1H, 3J(H,H)=11.5 Hz, CH(3)), 7.19
(d, 2H, 3J(H,H)=11.5 Hz, CH(4)).
[23] Y. Lin, C. Gong, A. Han, Xiangtan Daxue Ziran Kexue Xuebao, 1991,
14(1), 82-5.
[24] for example: M. J. Climent, A. Corma, S. Iborra, A. Velty, J. Mol. Cat. A:
Chem. 2002, 182-183, 327-342; S. Abello, F. Medina, D. Tichit, J. Perez-
Ramirez, X. Rodriguez, J. E. Sueiras, P. Salagre, Y. Cesteros, Y., Appl.
Catal. A: Gen. 2005, 281(1-2), 191-198; F. Winter, A. J. van Dillen, K. P.
de Jong, Chem. Commun. 2005, 31, 3977-3979.
Acknowledgements
We gratefully acknowledge Jonathan Medlock, Claude Furer, and
Francesco Pace for fruitful discussions and assistance in carrying
out several experiments.
[25] W. Dobler, N. Bahr, K. Breuer, A. Kindler (BASF), WO2004041764, 2004.
[26] K. Tanabe, W. F. Hölderich, Appl. Catal. A: Gen. 1999, 181, 399.
[27] P. Lecocq, Report from The Dow Chemical company, 2007.
[28] D. V. Banthorpe, E. D. Hughes, C. Ingold, J. Chem. Soc. 1960, 4054-
4087.
Keywords: aldol condensation, solid catalysts, ion exchange
resin, AMBERLYSTTM, citral, y-ionone, flavour compounds
[29] M. J. Climent, A. Corma, S. Iborra and A. Velty, Green Chem. 2002, 4,
474–480.
[30] Product Data Sheet of AMBERLYSTTM A26OH, The Dow Chemical
Company.
[1]
J. Panten, H. Surburg, Flavors and Fragrances, 2. Aliphatic Compounds.
In Ullmann’s Encyclopaedia of Industrial Chemistry, online ed. Wiley-
VCH, Weinheim, Germany, 2015, DOI: 10.1002/14356007.t11_t01.
[31] E. J. Corey, M. Sodeoka, Tetrahedron Lett. 1991, 32, 7005-7008.
Available
online:
http://onlinelibrary.wiley.com/doi/10.1002/14356007.t11_t01/abstract
(accessed on 26 August, 2016).
[2]
[3]
Y. Ono, J. Catal. 2003, 216, 406-415.
F. Figueras, M. L. Kantam, B. M. Choudary, Curr. Org. Chem. 2006, 10,
1627-1637.
[4]
[5]
H. Hattori, Chem. Rev. 1995, 95, 537-558.
A. T. Nielsen, W. J. Houlihan, The Aldol Condensation. In Organic
Reactions, online ed. Wiley-VCH, Weinheim, Germany, 2011, DOI:
10.1002/0471264180.or016.01.
Available
online:
http://
http://onlinelibrary.wiley.com/doi/10.1002/0471264180.or016.01/abstrac
t (accessed on 23 June, 2016).
[6]
C. H. Heathcock, The Aldol Reaction: Acid and General Base Catalysis.
In Organic Reactions, 1st ed. (Eds.: B. M. Trost, I. Fleming), Pergamon
Press: Oxford, England, 1991.
[7]
[8]
W. Dobler, N. Bahr, K. Breuer, A. Kindler (BASF), WO 2004041764, 2004.
L. Qing-Ning, G. Jian-Sheng, T. Qin, S. Xiao-Ling, Z. Zu-Xin (Shanghai
Institute of Technology), CN 1202065, 2002.
[9]
Q. Wang, Z. Guo, G. Jiang, Huaxue Shijie 1992, 33(10), 449-451.
[10] Y. Shi, Q. Gu, Huaxue Fanying Gongcheng Yu Gongyi 1993, 9(4), 353-
359.
[11] Z. Ma, Y. Kan, Y. Zheng, Y. Zheng, Tianran Chanwu Yanjiu Yu Kaifa
1997, 9(3), 59-64.
[12] Z. Ma, Y. Kan, Jingxi Huagong 1996, 13(4), 16-17.
[13] E. Galera, A. Zabza, Bulletin de l'Academie Polonaise des Sciences,
Serie des Sciences Chimiques 1977, 25(8), 615-625.
[14] S. Huang, H. Wen, C. Liu, Z. Wu, Tianran Chanwu Yanjiu Yu Kaifa 1993,
5(1), 21-25.
[15] N. Zhang, G. Liang, Lizi Jiaohuan Yu Xifu 1991, 7(2), 142-146.
[16] B. Szmidt, M. Beldowiez, G. Wishniewski, J. Góra, U. Antezak, D.
Kalemba, W. S. Brud, PL 147748, 1986.
[17] J. Góra, D. Kalemba, Pollena 1985, 29(11-12), 236-43.
[18] W. Bonrath, H. Fleischhauer, W. F. Hölderich, J. Schütz (DSM),
WO2008145350, 2008.
[19] D. J. W. Chong, F. H. L. Chong, J. Latip, Green Process Synth 2016, 5,
297-304.
[20] M. Sigl, C. Miller, W. Dobler, M. Haake (BASF), WO 2003047748 A1,
2003; G. Kaibel, C. Miller, W. Dobler, T. Dirnsteiner, M. Sigl, H. Jansen,
B. Kaibel (BASF), WO 2003047747, 2003; M. J. Climent, A. Corma, S.
Iborra, A. Velty, J. Mol. Cat. A: Chem. 2002, 182-183, 327-342; R.
Fischer, N. Goetz (BASF), DE 2148610, 1973; R. Fischer, N. Goetz, H.
Mueller (BASF), DE 2150992, 1973.
[21] S. Yu, Hunan Shifan Daxue Ziran Kexue Xuebao, 1991, 14(4), 325-9.
[22] R. G. Kelleher, M. A. McKervey, P. Vibuljan, J. Chem. Soc., Chem.
Comm. 1980, 11, 486-8.
This article is protected by copyright. All rights reserved.