Synthesis of ionones on solid Br?nsted acid catalysts: Effect of acid site strength on ionone isomer selectivity

Article abstract of DOI:10.1016/j.cattod.2009.09.016

The effect of Br?nsted acid site strength on the liquid-phase conversion of pseudoionone to ionone isomers (α-, β- and γ-ionone) was studied on resin Amberlyst 35W, silica-supported heteropolyacid (HPAS) and silica-supported triflic acid (TFAS). Catalyst acidity was probed by temperature-programmed desorption of NH₃ coupled with infrared spectra of adsorbed pyridine. The initial pseudoionone conversion rate followed the order: TFAS > Amberlyst 35W ≈ HPAS. Synthesis of the three ionone isomers occurred via a common cyclic carbocation intermediate formed from the activation of the pseudoionone molecule on Br?nsted acid sites. Initial ionone mixtures containing a α:β:γ isomer distribution of about 40:20:40 were formed, irrespective of the acid site strength. But the ionone mixture composition changed with the progress of the reaction because γ-ionone was consecutively converted to α-ionone on HPAS and Amberlyst 35W, whereas the stronger acid sites of TFAS converted γ-ionone to β-ionone.

Full text of DOI:10.1016/j.cattod.2009.09.016

Catalysis Today 149 (2010) 267–274
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Synthesis of ionones on solid Brønsted acid catalysts: Effect of acid site
strength on ionone isomer selectivity
*
´
´
V.K. Dıez , C.R. Apesteguıa, J.I. Di Cosimo
Catalysis Science and Engineering Research Group (GICIC), INCAPE, UNL-CONICET, Santiago del Estero 2654, 3000 Santa Fe, Argentina
A R T I C L E I N F O
A B S T R A C T
Article history:
The effect of Brønsted acid site strength on the liquid-phase conversion of pseudoionone to ionone
Available online 3 November 2009
isomers (a-, b- and g-ionone) was studied on resin Amberlyst 35W, silica-supported heteropolyacid
(HPAS) and silica-supported triflic acid (TFAS). Catalyst acidity was probed by temperature-programmed
desorption of NH3 coupled with infrared spectra of adsorbed pyridine. The initial pseudoionone
conversion rate followed the order: TFAS > Amberlyst 35W ꢀ HPAS. Synthesis of the three ionone
isomers occurred via a common cyclic carbocation intermediate formed from the activation of the
Keywords:
Ionone isomers
Acid catalysis
Tungstophosphoric acid
Triflic acid
pseudoionone molecule on Brønsted acid sites. Initial ionone mixtures containing a
distribution of about 40:20:40 were formed, irrespective of the acid site strength. But the ionone mixture
composition changed with the progress of the reaction because -ionone was consecutively converted to
-ionone on HPAS and Amberlyst 35W, whereas the stronger acid sites of TFAS converted -ionone to
ionone.
a:b:g isomer
g
a
g
b-
ß 2009 Elsevier B.V. All rights reserved.
1. Introduction
significantly lower than those reported via the homogeneous
reaction (70–90%). A better approach to the homogeneous catalyst
performance was reached in our recent contributions [6,7], in
which we explored the synthesis of ionones on several solid acids
such as bulk tungstophosphoric acid (HPA), silica-supported HPA
containing between 18.8 and 58.5% HPA, Cs-HPA, zeolite HBEA,
SiO2–Al2O3 and Amberlyst 35W resin. Results showed that the
reaction can be efficiently promoted on catalysts containing a high
density of strong Brønsted acid sites. The highest ionone yield, 79%,
was obtained on a 58.5 wt.% HPA/SiO2 catalyst at 383 K and is
comparable to the best values reported in the literature for the
homogeneously catalyzed reaction using sulfuric acid.
Here, our research specifically focuses on the effect of Brønsted
acid site strength on ionone isomer selectivity. Previous work
using homogeneous catalysts showed that the ionone isomer
distribution depends on the nature, concentration and strength of
the liquid acid catalyst. For example, it has been reported that
Ionone isomers are fine chemicals widely used as pharmaceu-
ticals and fragrances. Specifically, -ionone is an important
precursor in the synthesis of vitamin A whereas - and -ionones
b
a
g
are in high demand in the fragrance industry because of their violet
and woody-fruity scent, respectively [1,2]. The current commercial
synthesis of ionones from acetone and citral uses homogeneous
catalysis and involves a two-step process, as illustrated in Scheme
1. Initially, the liquid-phase aldol condensation of citral with
acetone is catalyzed by diluted bases and forms selectively
pseudoionones. The consecutive cyclization of pseudoionones to
yield ionone isomers is catalyzed by liquid acids [3], which entails
environmental concerns because of corrosion and waste disposal.
Then, new industrial strategies for ionone synthesis from
pseudoionone cyclization demand the replacement of liquid acids
by solid catalysts.
However, very few papers have been published on the use of
solid acids for converting pseudoionone to ionone isomers. Lin
et al. [4] reported that cationic ion-exchange resins convert
concentrated sulfuric acid produces predominantly
while phosphoric acid forms essentially -ionone [3]. In contrast,
-ionone is the main isomer when Lewis acids such as BF3 are used
b-ionone [8,9]
a
g
pseudoionone to an ionone mixture containing
a
- and
b
-isomers;
[10]. To the best of our knowledge, there are not in the literature
papers dealing with the effect of the acid site density and strength
of solid catalysts on the ionone isomer selectivity. In this work, we
prepared solid Brønsted acid catalysts, namely resin Amberlyst
35W, silica-supported heteropolyacid and silica-supported triflic
acid, with the aim of correlating the catalyst acid site strength with
the resulting ionone isomer distribution. Our results show that the
total ionone yields were between 20 and 40%. Similar values were
reported by Guo et al. [5] employing sulfated and persulfated TiO2/
MCM-41. These ionone yields obtained on solid acids were
* Corresponding author. Tel.: +54 342 4555279; fax: +54 342 4531068.
E-mail address: verodiez@fiq.unl.edu.ar (V.K. Dı´ez).
relative concentration of a-, b- and g-ionone in the ionone mixture
0920-5861/$ – see front matter ß 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2009.09.016

268
V.K. Dı´ez et al. / Catalysis Today 149 (2010) 267–274
Scheme 1. Two-step process for ionone synthesis.
can be controlled by changing the catalyst Brønsted acid site
strength and the operative conditions, in particular temperature
and reaction time.
pyridine as probe molecule and a Shimadzu FTIR Prestige-21
spectrophotometer. Sample wafers were formed by pressing 20–
40 mg of the catalyst at 5 ton/cm2 and transferred to a sample
holder made of quartz. An inverted T-shaped Pyrex cell containing
the sample pellet was used. The two ends of the short arm of the T
were fitted with CaF2 windows. Sample wafers were evacuated 4 h
at 523 K for HPAS and 353 K for TFAS and then the catalyst
spectrum was recorded after cooling the sample at room
temperature. After that, admission of 0.12 kPa of pyridine and
adsorption at room temperature were performed, followed by
sequential evacuation at 298, 373, 423 and 473 K for HPAS and 298
and 353 K for TFAS. Spectra were recorded at room temperature
and the spectra of the adsorbed species were obtained by
subtracting the catalyst spectrum.
2. Experimental
2.1. Catalyst synthesis and activation
HPA/SiO2 (HPAS) and TFA/SiO2 (TFAS) catalysts with acid
contents of 58.5 and 8.2 wt.%, respectively, were prepared by
incipient wetness impregnation. Tungstophosphoric acid
(H3PW12O40ꢁxH2O, Merck, GR) and triflic acid (CF3SO3H, Sigma–
Aldrich, Reagent Grade) were added to a commercial SiO2 (Grace
2
˚
Davison, G62, 99.7%, 272 m /g, pore diameter = 122 A) using
aqueous solutions of HPA and TFA. The impregnated samples
were both dried at 353 K and then decomposed and stabilized at
523 K (HPAS) and 383 K (TFAS) for 18 h in N2 (40 cm3/min).
Amberlyst 35W resin pellets (Rohm and Haas) were crushed
2.3. Catalytic testing
The cyclization of pseudoionone (Fluka, >90%) was carried out
at 353 K under autogenous pressure (ꢀ250 kPa) in a batch PARR
reactor, using toluene as aprotic solvent and a catalyst/PS = 28–
56 wt.% ratio. Catalysts were pretreated ex situ in a N2 stream at
the calcination temperature for 2 h to remove adsorbed water.
After introducing the reactant and solvent the reactor was sealed
and flushed with N2 and then the mixture was heated up to the
final reaction temperature under stirring (300 rpm). Then the
catalyst was added to the reaction mixture to start the reaction.
Catalysts were loaded as ground powders in which inter- and
intra-particle diffusional limitations were verified to be negligible.
Reaction products were periodically analyzed during the 6-h
reaction in a Varian Star 3400 CX gas chromatograph equipped
with an FID and a Carbowax Amine 30 M capillary column. Main
and sieved to retain particles between 180 and 480
mm. Then, the
resin was treated in N2 (40 cm3/min) at 373 K overnight.
2.2. Catalyst characterization
The HPA content in HPAS catalyst was determined via the
analysis of tungsten by UV spectroscopy in a Metrolab 1700 UV–vis
spectrometer. Sample was calcined in an oven at 1073 K to
transform H3PW12O40ꢁxH2O in tungsten oxide (WO3) and then
digested in an alkali solution. The solution containing WO3 was
finally analyzed by UV–vis technique. The TFA loading in TFAS was
determined by titration of sample protons. TFAS (0.3 g) was
suspended in 75 ml of an aqueous solution of KCl (0.03 M) to
release the triflic acid protons to the aqueous solution. The
suspension was stirred for 20 min and then titrated with a 0.06 M
KOH solution using phenolphthalein as acid–base indicator.
BET surface areas were measured by N2 physisorption at its
boiling point using an Autosorb Quantachrome 1-C sorptometer.
The acid site density (na) of HPAS catalyst was measured by
temperature-programmed desorption (TPD) of NH3 preadsorbed at
373 K. Sample (150 mg) was treated in He (40 cm3/min) at 523 K
for 1 h, cooled down to 373 K, and then exposed to a 1% NH3/He
stream during 15 min. Weakly adsorbed NH3 was removed by
flushing with He at 373 K during 1 h. Finally, the sample
temperature was increased from 373 to 1073 K at a rate of 10 K/
min in a He flow of 60 cm3/min. Desorbed NH3 was analyzed by
mass spectrometry (MS) in a Baltzers Omnistar unit.
reaction products were ionones (
ities (Sj, mol of product j/mol of PS reacted) were calculated as
Sj = Cj/ Cj where Cj is the concentration of product j. Product yields
j, mol of product j/mol of PS fed) were calculated as
a-, b- and g-isomers). Selectiv-
S
(
h
hj = Sj ꢂ XPS
,
where XPS is the pseudoionone conversion.
3. Results and discussion
3.1. Catalyst characterization
The surface areas and acid properties of the catalysts are
presented in Table 1. Incorporation of bulky HPA species to SiO2
caused a significant diminution of the SiO2 surface area, from 272
to 144 m2/g (HPAS sample), probably because of the formation of
an incipient three-dimensional HPA structure that partially blocks
the silica pores. The presence of the heteropolyacid structure in
The chemical nature of surface acid sites on HPAS and TFAS
catalysts was determined by Infrared Spectroscopy (IR) by using

V.K. Dı´ez et al. / Catalysis Today 149 (2010) 267–274
269
Table 1
Surface area and acid properties of HPAS, TFAS and Amberlyst 35W samples.
Catalyst
Surface area (m2/g)
Surface acid properties
Acid site density
Acid site nature by FTIR
a
na
(
m
mol/g)
nH+
(
mmol/g)
B (area/g)
L (area/g)
B/(B + L) (%)
TFAS
245
39
–
540b
5200d
610e
222c
–
450f
18c
–
56f
93
–
Amberlyst 35W
HPAS
–
144
566
89
B: Brønsted sites, L: Lewis sites.
a
By TPD of NH3.
b
By acid–base titration.
c
Data obtained after pyridine admission at 298 K and evacuation at 353 K.
From manufacturer information.
d
e
f
By UV spectroscopy analysis.
Data obtained after pyridine admission at 298 K and evacuation at 373 K.
HPAS sample after impregnation and calcination was confirmed by
comparing its XRD pattern with that of pure HPA (Fig. 1). In contrast,
impregnation of silica with less bulky triflic acid caused only a slight
decrease of the silica surface area, from 272 to 245 m2/g.
We determined the catalyst acid properties by combining TPD
of NH3 and IR spectroscopy of adsorbed pyridine. However, the low
stability of triflic acid (the boiling point of TFA is 435 K) hindered
the characterization of TFAS sample by TPD of NH3. On the other
hand, resin Amberlyst 35W could not be characterized either by
TPD of NH3 or FTIR of adsorbed pyridine due to its low
thermostability. Fig. 3 shows the NH3 TPD curve obtained on
HPAS sample. The NH3 desorption gave rise to several desorption
peaks at temperatures in the range of 500–900 K reflecting the
presence of surface acid species that bind NH3 with different
strengths. The total surface acid site density, na, was obtained by
deconvolution and integration of the NH3 TPD curve and is
On the other hand, we investigated the presence of triflic acid
2ꢃ
on TFAS sample by monitoring the IR SO2 vibration band at
1417 cmꢃ1, which is regarded as the characteristic band of
supported TFA [11]. In Fig. 2 we compare the FTIR spectra of TFAS
sample and pure SiO2. Clearly, addition of TFA to SiO2 gave rise to a
new IR band at 1417 cmꢃ1 arising from the stretching of the SO2
group. The formation of a broad absorption band at 3495 cmꢃ1
,
attributed to hydrogen-bonded TFA species to the silanols groups
of SiO2, and the consequent intensity decrease of the silanol group
band (3734 cmꢃ1) were also observed [11].
presented in Table 1. The proton content (nH+
measured on HPAS by UV spectroscopy (Table 1, column 4) was
similar to the corresponding na value ( mol NH3/g cat). This
, mmol H+/g cat)
m
agreement between na and nH+ values strongly suggests that the
HPA protons are completely accessible for NH3 adsorption in HPAS
sample.
The chemical nature and strength of surface acid sites of HPAS
and TFAS were determined from the IR spectra of adsorbed
pyridine after admission at 298 K and sequential evacuation at
increasing temperatures. The FTIR spectra of adsorbed pyridine
after evacuation at 373, 423 and 473 K on HPAS, and at 353 K on
TFAS sample are shown in Fig. 4a and b, respectively. TFAS and
HPAS catalysts showed the IR bands typical of pyridinium ion
formed on Brønsted acid sites, i.e. the ring vibrations modes at
1636, 1608, 1537 and 1485 cmꢃ1, and an additional weak band
arising from pyridine coordinated on Lewis acid sites at
Fig. 2. FTIR spectra of the SiO2 support after evacuation at 423 K, and of silica-
Fig. 1. XRD patterns of HPA and HPAS samples.
supported triflic acid (TFAS) catalyst after evacuation at 353 K.

270
V.K. Dı´ez et al. / Catalysis Today 149 (2010) 267–274
ꢄ1450 cmꢃ1 [12]. The fact that the HPAS sample displayed both
Brønsted and Lewis acid bands suggests that upon impregnation of
HPA on SiO2, in addition to the Keggin structure that accounts for
the Brønsted acid species, the heteropolyacid partially transforms
giving rise to lacunary or unsaturated species [13] of Lewis acid
character formed by interaction with the silica support. A measure
of the relative contribution of the Brønsted and Lewis acid sites on
HPAS and TFAS samples was obtained from Fig. 4a and b, by
integration of the bands at ꢄ1540 and ꢄ1450 cmꢃ1, respectively,
corresponding to the 19b ring vibrations of pyridine [14]; results
are presented in Table 1. Calculations show that Brønsted acid sites
represent about 90% of the total acid density on both samples
(Table 1, column 7). Moreover, we observed that the Brønsted/
Lewis acid site ratio on HPAS increased with the evacuation
temperature, thereby indicating that Brønsted acid sites are more
strongly acidic than the Lewis acid sites and retain pyridine up to
higher evacuation temperatures. In summary, characterization
using IR of adsorbed pyridine shows that HPAS and TFAS samples
are essentially solid Brønsted acid catalysts.
3.2. Catalytic testing
3.2.1. Catalyst activity and selectivity
Fig. 5 shows the evolution of ionone yields with reaction time
obtained on HPAS, TFAS and Amberlyst 35W samples at 353 K. The
local slopes of the ionone yield curves in Fig. 5 give the rate of
formation of each product at a specific PS conversion and reaction
time. From the curves of Fig. 5 we determined the initial ionone
formation rate (rI0ONONE, mmol/h g) by calculating the initial slopes
according to:
ꢀ
ꢁ
n0PS
W
dhIONONE
dt
rI0ONONE
¼
t¼0
where W is the catalyst load and n0PS is the initial molar amount of
PS. The obtained rI0ONONE values are given in Table 2. It is observed
that rI0ONONE on TFAS was about twice as high as on Amberlyst 35W
and HPAS. Moreover, if the rI0ONONE values obtained on TFAS and
Fig. 3. TPD profile of NH3 on HPAS sample. NH3 adsorption at 373, 10 K/min heating
rate.
Fig. 4. FTIR spectra of pyridine after adsorption at 298 K and evacuation at (a) 373, 423 and 473 K on HPAS and (b) at 353 K on TFAS. B: Brønsted; L: Lewis.

V.K. Dı´ez et al. / Catalysis Today 149 (2010) 267–274
271
Fig. 5. Ionone yield (hIONONE) as a function of time on HPAS, TFAS and Amberlyst
35W catalysts [T = 353 K, n0PS ¼ 0:009 mol, toluene/PS = 71 (molar ratio),
WHPAS = WAmberlyst 35W = 1.0 g, WTFAS = 0.5 g].
HPAS are determined per gram of acid loading, then the initial
ionone formation rate on TFAS results about 18 times higher in
comparison to HPAS. Results in Table 2 reveal that the initial
activity order for obtaining ionones is: TFAS > Amberlyst
35W ꢀ HPAS. This activity order probably reflects more differences
in the catalyst proton site strength than in the catalyst acid site
density. In effect, protons on TFA present the strongest acidity on a
Hammett acidity function basis (H0 = ꢃ14.6) [15], in contrast to
ꢃ13.2 [16] on HPAS and ꢃ2.65 [17] on Amberlyst 35W.
The PS conversion and ionone isomer yields (
hi , i: a-ionone, b-
ionone or -ionone) as a function of time for HPAS, TFAS, and
g
Amberlyst 35W at 353 K are presented in Fig. 6. The nonzero initial
slopes of the isomer yield curves indicate that the three ionones are
primary products, i.e. they are formed directly from PS. From the
initial slopes of the ionone isomer yield vs. time curves of Fig. 6 we
determined the initial formation rate of ionone isomers. Results are
shown in Table 2.
The three catalysts converted initially PS mainly to
a- and g-
Fig. 6. Ionone isomer yields (
h
IONONE) and PS conversion as a function of time on
ionones; the r0 and r0 values were, effectively, 2–3 times higher in
HPAS, TFAS and Amberlyst 35W catalysts. Reaction conditions as in Fig. 4.
a
g
all the cases than those of r0 . But the initial ionone isomer
b
distribution was similar on all samples of Table 1, giving an initial
a
:
b
:
g
isomer ratio of approximately 40:20:40 (Table 2). This result
Fig. 7 shows that on TFAS sample the contribution of
continuously increased with reaction time at the expenses of
ionone attaining ꢀ50% at the end of the run, while the
concentration of -ionone in the ionone mixture remained almost
constant (38%). In contrast, on Amberlyst 35W -ionone was
converted to -ionone so that the contribution of -ionone in the
ionone mixture increased from 41% at the beginning of the reaction
to 61% at the end of the 6-h run whereas the contribution of b-
b-ionone
suggests that the initial ionone isomer distribution does not
depend on the Brønsted acid site strength.
g-
Fig. 6 also shows that the
a
-isomer yield curve monotonically
a
grew with reaction time on the three catalysts. In contrast, the
isomer yield curve reached a maximum at about 80–90% PS
conversion on TFAS and Amberlyst 35W, thereby suggesting that
on these catalysts
g
g
a
a
g
-ionone is consecutively converted to the other
isomers. More insight on the ionone isomer distribution changes
with the progress of the reaction is inferred from Fig. 7. In fact,
ionone slightly increased. Finally, although on HPAS the ionone
isomer distribution did not change significantly during the 6-h run,
Table 2
Catalytic results on HPAS, TFAS and Amberlyst 35W catalysts.
Catalyst
Initial ionone formation rate (mmol/h g)
Initial ionone isomer distribution (%)
rI0ONONE
r0
r0
r0
a
b
g
a
g
b
TFAS
11.56
6.61
4.48
4.56
3.02
1.98
2.57
1.38
0.60
4.43
2.21
1.90
36.2
40.7
38.2
24.8
22.0
18.9
39.0
37.3
42.9
Amberlyst 35W
HPAS
Reaction conditions: T = 353 K, P =250 kPa, n0PS ¼ 0:009 mol, toluene/PS = 71 (molar ratio), WHPAS = WAmberlyst 35W = 1.0 g, WTFAS = 0.5 g.

272
V.K. Dı´ez et al. / Catalysis Today 149 (2010) 267–274
Fig. 8. Ionone isomer distribution and PS conversion as a function of time on HPAS
sample [T = 383 K, n0PS ¼ 0:009 mol, toluene/PS = 71 (molar ratio), WHPAS = 1.0 g].
Fig. 9. Arrhenius plots for the formation of a-, b- and g-ionones from pseudoionone
on HPAS [T = 343–383 K, n0PS ¼ 0:009 mol, toluene/PS = 71 (molar ratio),
WHPAS = 1.0 g].
remained almost constant. This qualitative evolution of ionone
isomer distribution with reaction time is similar to that observed
on Amberlyst 35W at 353 K in Fig. 7.
From the ionone isomer yields vs. time curves obtained on HPAS
at 343, 353, 363, 373 and 383 K, we determined first the initial
formation rates and then the apparent activation energy, Ea, for
each individual ionone isomer. The apparent activation energies
for the formation of ionone isomers were determined using an
Fig. 7. Ionone isomer distribution as a function of time on HPAS, TFAS and
Amberlyst 35W catalysts. Dotted line: reaction time at 100% PS conversion.
Reaction conditions as in Fig. 4.
Arrhenius-type function (ri0 ¼ AeꢃE =RT , where A is a constant
a
containing the pre-exponential factor) by plotting ln ri0 values as a
function of 1000/T, Fig. 9. The Ea values obtained for the synthesis
some isomer interconversion can be observed. In summary, results
of Figs. 6 and 7 show that the kinetics of ionone isomer
interconversion is affected by the catalyst surface proton strength.
To explore the effect of temperature on catalyst activity and
selectivity, we performed additional catalytic tests on sample
HPAS varying the reaction temperature between 343 and 383 K.
Specifically, we carried out catalytic tests at 343, 353, 363, 373, and
383 K. Fig. 8 illustrates the results obtained at 383 K. At this
temperature, PS was completely converted in 2 h. The contribution
of
respectively. These values indicate that an increase in reaction
temperature on HPAS would promote more the formation of
ionone than - and -ionones in agreement with our experimental
results. For example, the initial contribution of the -isomer in the
ionone mixture on HPAS increased from 38.2% at 353 K to 44.8% at
a-, b- and g-ionones from PS were 24, 19 and 15 kcal/mol,
a
-
b
g
a
383 K, while for the same temperatures
42.9 to 35.7%.
g-ionone decreased from
of
a
-ionone continuously grew at the expense of the
g
-isomer with
-isomer
The calculated kinetic parameters A and Ea for the formation of
the three ionone isomers were plotted in Fig. 10 as ln A vs. Ea in
the progress of the reaction, while the contribution of the
b

V.K. Dı´ez et al. / Catalysis Today 149 (2010) 267–274
273
hydrogen fluoride. They postulated that medium strength liquid
acids promote the PS protonation forming an enol carbonium that
leads to
acids may protonate the cyclic double bond of
carbonium ion that leads to the formation of
authors [3] noted that medium strength liquid acids such as
phosphoric acid form essentially -ionone.
a
-ionone by rearrangement. Consecutively, strong liquid
-ionone forming a
-ionone. Other
a
b
a
Based on the results of Figs. 6–8, we present in Scheme 2 the
reaction pathways involved in the synthesis of ionone isomers
from PS on HPAS, TFAS and Amberlyst 35W catalysts. We postulate
that the PS molecule is initially activated on surface Brønsted acid
sites and forms a common cyclic intermediate for the consecutive
ionone isomer formation. The synthesis of the three ionone
isomers from PS via the same cyclic carbocation intermediate is
consistent with the compensation effect observed for the initial
formation of ionone isomers on HPAS sample (Fig. 10). As noted
previously [23–25], the compensation effect is frequently verified
when
a catalyst promotes a group of reactions involving
compounds of analogous molecular structures, sharing the same
mechanism and transition state. The common cyclic intermediate
in Scheme 2 contains H (two), H (one) and H (three) protons
Fig. 10. Compensation effect in ionone isomer formation on HPAS.
a
b
g
that have to be detached to form
a-,
b- and
g
-ionone, respectively.
order to determine if the experimental data obey a Cremer–
Constable relation [18,19]
We observed here (Fig. 6) that the three ionone isomers are formed
directly from PS, i.e. they are primary products.
On the other hand, our results on the initial ionone isomer
formation rates in Table 2, show that on TFAS, HPAS and Amberlyst
35W samples the PS conversion forms initially ionone mixtures of
ln A ¼ mEa þ c
The resulting linear correlation in this semilogarithmic plot is
indicative of the existence of a ‘‘Compensation Effect’’ in the
kinetics of ionone isomer formation on HPAS.
similar composition, containing about 40%
and 40% -ionone, which is close to the statistical composition
= 33:17:50). This result strongly suggests that the initial
a-ionone, 20% b-ionone
g
(a:b:g
selectivity of ionone isomers does not depend significantly on the
strength of surface Brønsted acid sites. In contrast, the ionone
isomer interconversion during the progress of the reaction
3.2.2. Reaction mechanism
The reaction pathways of PS cyclization to ionone isomers have
been studied almost exclusively using liquid acid catalysts. Royals
[20] investigated the cyclization of PS using different liquid acids
such as H2SO4, H3PO4, Cl3CCOOH, and ClCH2COOH. They observed
that the composition of the ionone mixture depends on the acid
strength. In fact, the proportion of
strength of the acid while formation of
when using weakly acidic media. Consistently, other authors have
depended on the acid site strength. In fact,
isomerized to -ionone on HPAS and Amberlyst 35W (Figs. 7
and 8), while the stronger acid sites of TFAS converted -ionone to
-ionone (Fig. 7).
Analyzing the chemical structure of the ionone molecules
(Scheme 1), it can be assumed the following stability order:
-ionone. The -isomer is highly stable due to the
molecule extended conjugated system and therefore is not
isomerized to - or -ionones. In contrast, taking into account
g-ionone was
a
g
b
b
-ionone increased with the
a
-ionone was predominant
b
>
a
>
g
b
reported that concentrated sulfuric acid produces mainly
[8,9]. Markovich et al. [21,22] studied the synthesis of
using trifluoroacetic acid, fluorosulfonic acid and anhydrous
b
b
-ionone
-ionone
a
g
Scheme 2. Reaction network for the synthesis of ionone isomers from pseudoionone on Brønsted acid catalysts.

274
V.K. Dı´ez et al. / Catalysis Today 149 (2010) 267–274
the instability and reactivity of
C55C bond in this molecule, conversion to the other ionones with
the progress of the reaction is expected.
g
-ionone because of the exocyclic
ionone isomer interconversion may be controlled by changing the
Brønsted acid site strength, temperature and reaction time.
Acknowledgements
In summary, our results show that the ionone isomer mixture
composition can be controlled by changing the Brønsted acid site
strength, temperature and reaction time. The initial PS conversion
forms ionone mixtures with nearly statistical composition,
irrespective of the Brønsted acid site strength. By increasing the
´
´
Authors thank the Agencia Nacional de Promocion Cientıfica y
´
Tecnologica (ANPCyT), Argentina (Grant PICT 14-11093/02),
CONICET, Argentina (Grant PIP 5168/05) and the Universidad
Nacional del Litoral, Santa Fe, Argentina (Grant CAI+D 007-040/05)
for the financial support of this work. They also thank H. Cabral for
technical assistance.
temperature and/or the reaction time,
- or -ionone, depending on the Brønsted acid site strength. Final
ionone mixtures containing essentially - and -ionone may be
obtained at 353 K on TFAS (mixture of about 50% -ionone and 40%
-ionone) and on Amberlyst 35W (mixture of about 60% -ionone
g-ionone is transformed to
a
b
a
b
b
References
a
a
and 30%
ionone and 20%
b
-ionone) or at 383 K on HPAS (mixture of about 80%
a
-
[1] Ullmann’s Encyclopedia of Industrial Chemistry, 6th ed., 2002 (electronic).
[2] E. Brenna, C. Fuganti, S. Serra, P. Kraft, Eur. J. Org. Chem. (2002) 967.
[3] H. Hibbert, L.T. Cannon, J. Am. Chem. Soc. 46 (1924) 119.
[4] Z. Lin, H. Ni, H. Du, C. Zhao, Catal. Commun. 8 (2007) 31.
[5] D. Guo, Z.-F. Ma, Q.-Z. Jiang, H.-H. Xu, Z.-F. Ma, W.-D. Ye, Catal. Lett. 107 (2006)
155.
b-ionone).
4. Conclusions
[6] V.K. Dı´ez, C.R. Apesteguı´a, J.I. Di Cosimo, Catal. Lett. 123 (2008) 213.
[7] V.K. Dı´ez, B.J. Marcos, C.R. Apesteguı´a, J.I. Di Cosimo, Appl. Catal. A: Gen. 358
(2009) 95.
[8] K. Steiner, H. Ertel, H. Tiltscher, US Patent 5,453,546, to Hoffmann-La Roche Inc.
(1995).
[9] U. Rheude, U. Horcher, D. Weller, M. Stroezel, US Patent 6,288,282, to BASF
Aktiengesellschaft (2001).
[10] G. Ohloff, G. Schade, Angew. Chem. Int. Ed. 2 (1963) 149.
[11] A. de Angelis, C. Flego, P. Ingallina, L. Montanari, M.G. Clerici, C. Carati, C. Perego,
Catal. Today 65 (2001) 363.
[12] C. Morterra, A. Chiorino, G. Ghiotti, E. Fisicaro, J. Chem. Soc. Faraday Trans. I 78
(1982) 2649.
[13] I.V. Kozhevnikov, K.R. Kloetstra, A. Sinnema, H.W. Zandbergen, H. van Bekkum, J.
Mol. Catal. A: Chem. 114 (1996) 287.
[14] B. Bachiller-Baeza, J.A. Anderson, J. Catal. 228 (2004) 225.
[15] C.K. Jørgensen, Naturwissenschaften 67 (1980) 188.
[16] T. Okuhara, M. Misono, in: R.B. King (Ed.), Encyclopedia of Organic Chemistry,
John Wiley and Sons, 1994.
[17] R. Bringue´, M. Iborra, J. Tejero, J.F. Izquierdo, F. Cunill, C. Fite´, V.J. Cruz, J. Catal. 244
(2006) 33.
[18] F.H. Constable, Proc. Roy. Soc. (London) 108 (1925) 355.
[19] E. Cremer, Z. Phys. Chem. 144 (1929) 231.
[20] E.E. Royals, Ind. Eng. Chem. 38 (1946) 546.
[21] Y.D. Markovich, A.V. Panfilov, A.A. Zhirov, A.T. Kirsanov, L.A. Gorbach, K.A. Tara-
skin, Pharm. Chem. J. 32 (10) (1998) 557.
[22] Y.D. Markovich, A.V. Panfilov, Y.N. Platunov, A.A. Zhirov, S.I. Kosenko, A.T. Kirsa-
nov, Pharm. Chem. J. 32 (11) (1998) 603.
[23] G.C. Bond, M.A. Keane, H. Kral, J.A. Lercher, Catal. Rev. Sci. Eng. 42 (3) (2000) 323.
[24] M. Boudart, Chem. Eng. Prog. 57 (1961) 33.
[25] A.K. Galwey, Adv. Catal. 26 (1977) 247.
The liquid-phase synthesis of ionone isomers (a-, b- and
g
-
ionone) from pseudoionone cyclization is efficiently achieved on
solid Brønsted acid catalysts such as Amberlyst 35W, silica-
supported triflic acid (TFAS) and silica-supported heteropolyacid
(HPAS). Catalyst activity depends on the surface acid site strength
so that the initial activity order for obtaining ionones is
TFAS > Amberlyst 35W ꢀHPAS.
The pseudoionone molecule is initially activated on surface
Brønsted acid sites and forms a common cyclic intermediate for
the consecutive ionone isomer generation. This cyclic carboca-
tion intermediate contains three different kinds of protons that
upon direct detachment lead to
products.
a-, b- or g-ionones as primary
Selectivity to the three ionone isomers is also strongly
dependent on the acid site strength. Under initial conditions,
pseudoionone transformation gives ionone mixtures with approxi-
mately statistical composition (40%
40% -ionone), regardless of the catalyst Brønsted acid site
strength. However, with the progress of the reaction -ionone,
the least stable isomer, is isomerized to -ionone on HPAS and
Amberlyst 35W, and to -ionone on the stronger acid sites of TFAS.
Therefore, binary mixtures of - and -ionone can be obtained
by properly selecting operational conditions since the kinetics of
a-ionone, 20% b-ionone and
g
g
a
b
a
b