Hydroamination of 1,3-cyclohexadiene with aryl amines catalyzed with acidic form zeolites

Article abstract of DOI:10.1016/j.jcat.2006.01.007

The intermolecular hydroamination of 1,3-cyclohexadiene with aniline using zeolite catalysts was investigated. The reaction mechanism and the influence of amine basicity on the rate of reaction were studied. Zeolite H-BEA was the most active catalyst, whereas the incorporation of Zn²⁺ (Zn/H-BEA) led to decreasing catalytic activity, indicating that the reaction is catalyzed by Bronsted acid sites. Subtle shape selective effects determine the reactivity and selectivity of the zeolites.

Full text of DOI:10.1016/j.jcat.2006.01.007

Journal of Catalysis 239 (2006) 42–50
www.elsevier.com/locate/jcat
Hydroamination of 1,3-cyclohexadiene with aryl amines catalyzed
with acidic form zeolites
Oriol Jimenez a, Thomas E. Müller a,∗, Wilhelm Schwieger b, Johannes A. Lercher a
a
Lehrstuhl II für Technische Chemie, Technische Universität München, Lichtenbergstraße 4, D-85747 Garching, Germany
b
Lehrstuhl für Chemische Reaktionstechnik, Friedrich-Alexander-Universität Erlangen-Nürnberg, Egerlandstr. 3, D-91058 Erlangen, Germany
Received 17 October 2005; revised 23 December 2005; accepted 6 January 2006
Available online 9 February 2006
Abstract
The intermolecular hydroamination of 1,3-cyclohexadiene with aniline using zeolite catalysts was investigated. The reaction mechanism and
the influence of amine basicity on the rate of reaction were studied. Zeolite H-BEA was the most active catalyst, whereas the incorporation of
2+
Zn (Zn/H-BEA) led to decreasing catalytic activity, indicating that the reaction is catalyzed by Brønsted acid sites. Subtle shape selective effects
determine the reactivity and selectivity of the zeolites.
© 2006 Elsevier Inc. All rights reserved.
Keywords: Hydroamination; 1,3-Cyclohexadiene; Aniline; Substituent effects; Catalysis; Ion exchange; Zeolite; Characterization
1. Introduction
products. Other regioisomers are formed as byproducts [4].
Lanthanum complexes can efficiently catalyze the intramole-
cular hydroamination of aminodienes to nitrogen-containing
heterocycles [15]. Yoshifuji et al. [16] reported the use of pal-
ladium complexes to obtain a high product yield for the addi-
tion of anilines to 1,3-dienes. A recent study reported on the
use of a liquid–liquid two-phase system to catalyze the inter-
molecular hydroamination of terminal alkynes and dienes with
anilines [17]. The use of acids as co-catalysts can enhance the
catalytic activity of palladium in the addition of aniline to di-
enes [18]. A similar influence of Brønsted acids was also ob-
served for the reaction of vinylarenes with arylamines [19]. The
role of acid is still under investigation [20]; in most cases, the
addition of acid seems to prevent oligomerisation of the dienes,
as well as telomerisation of two or more olefin molecules with
one amine molecule [19,21].
Although a number of homogeneous catalysts for hydroami-
nation reactions are well known, only a few examples of het-
erogeneous catalysts have been reported. The amination of
isobutene with ammonia to tert-butylamine occurs over Re-
Y-zeolite with >90% selectivity [22]; however, this catalyst
suffers from rapid deactivation. BASF has developed an iron
silicate zeolite catalyst of the pentasil type that not only shows
>99% selectivity, but also affords a commercially acceptable
catalyst life [23,24]. Ion-exchanged zeolites [25,26] and immo-
Catalytic hydroamination is a chemical route with high syn-
thetic potential allowing the formation of a variety of nitrogen-
containing molecules by the direct addition of amines to alkenes
and alkynes [1–3]. Compared with other methods for the syn-
thesis of amines, enamines, and imines [4], hydroamination
offers one of the most attractive pathways to such mole-
cules. Conceptionally, the desired higher-substituted nitrogen-
containing products are formed in a single reaction step from
inexpensive alkenes and alkynes without the intrinsic formation
of side products [5,6]. Various transition metals are known to
be suitable as molecular catalysts for both inter and intramolec-
ular hydroamination of alkenes and alkynes [7]; for example,
titanium, ruthenium, palladium, and copper complexes have
been used as catalysts for intramolecular and intermolecular hy-
droamination of alkynes, allenes, and activated alkenes [8–14].
Intermolecular hydroamination of dienes is more difficult
than hydroamination of alkynes, and only few catalysts are
known. Nucleophilic addition of amines to 1,3-dienes with
sodium and alkyllithium salts provides primarily 1,4-addition
*
Corresponding author.
E-mail address: Thomas.Muller@ch.tum.de (T.E. Müller).
0021-9517/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2006.01.007

O. Jimenez et al. / Journal of Catalysis 239 (2006) 42–50
43
bilized zinc salts [27] have been used successfully for the cycli-
sation of 6-aminohex-1-yne, and immobilized transition metal
complexes have been used in the addition of 4-isopropylaniline
to phenylacetylene [28].
The present work addresses the use of Beta, ZSM-5, Fauja-
site, and Mordenite as catalysts for the intermolecular hydroam-
ination of 1,3-cyclohexadiene with aryl amines. Zeolite H-BEA
in particular has been shown to be a good solid acid catalyst, es-
pecially for reactions involving bulky transition states, such as
isobutene/n-butene alkylation [29,30], and is in the focus of this
study. Special attention is given to the influence of amine basic-
ity on the rate of reaction.
distribution of the catalysts were determined by nitrogen ad-
sorption (Sorptomatic 1990 Series instrument) after activation
of the samples at 250 ◦C in vacuum.
Temperature-programmed desorption (TPD) profiles were
measured in a custom-built 6-port parallel setup. The catalysts
were pelletized, and a 20-mg sample was placed into each
quartz tube. The samples were activated at 450 ◦C in vacuum
(at 1 × 10−3 mbar) for 1 h, then cooled to 150 ◦C, and am-
monia was adsorbed at 3 mbar for 10 min. After saturation,
the samples were outgassed for 1 h to remove physisorbed am-
monia. Subsequently, the temperature was increased at a rate
of 10 ◦C min−1, with the desorption process monitored by MS
(Balzers QMS 200).
2. Experimental
For infrared (IR) spectroscopic measurements, a self-sup-
porting sample wafer was placed into a sorption cell and ac-
tivated at 450 ◦C in vacuum for 1 h. The sample was cooled
to 150 ◦C, and pyridine was adsorbed at 10−1 mbar for 1 h.
After saturation, the sample was outgassed at 150 ◦C for 1 h.
IR spectra of the activated sample were recorded in the re-
gion of 4000–400 cm−1 at a resolution of 4 cm−1 using a
Perkin–Elmer 2000 spectrometer. The sample was then heated
to 450 ◦C at a rate of 10 ◦C min−1, then outgassed for 1 h.
The temperature was subsequently reduced to 150 ◦C, and an-
other IR spectrum was obtained. The concentration of acid sites
was estimated from the intensity of the bands at 1544 and
1455 cm−1, assigned to pyridinium ions (Brønsted acid sites,
ε = 1.67 cmµmol−1) and coordinatively bound pyridine (Lewis
acid sites, ε = 2.22 cmµmol−1), respectively, using molar ex-
tinction coefficients reported previously [31].
2.1. General
All reagents were obtained from Aldrich and were used
as received. Zinc-exchanged zeolites were prepared by re-
peated ion exchange of the corresponding H-BEA zeolite (Süd-
chemie AG, T-4546, MA039 Hr99) in an aqueous solution of
Zn(CH3CO2)2 as described previously [26]. The material was
dried and calcined, and the metal loading was determined by
atomic absorption spectroscopy (AAS). The Zn/H-BEA zeo-
lites had a loading in the range of 0.03–0.54 mmolZn2+ g−1
catalyst. The H-MFI (H-MFI 220, EX 717 H1-C) and H-
MOR (H-MOR 90, SN 302 H/01) zeolites were supplied by
Südchemie. H-FAU zeolite (CBV 400) was obtained from Ze-
olyst International. H-BEA zeolites with different crystal sizes
were prepared at Friedrich-Alexander-Universität Erlangen-
Nürnberg.
Catalytic experiments were performed under inert nitrogen
atmosphere in a Radleys reaction carousel with 12 parallel re-
actors. The zeolite (0.25 g) was activated overnight at 200 ◦C
in vacuum. It was suspended in toluene (15 cm3), and the mix-
ture was heated to reflux at 111 ◦C. Aniline (91 µL, 1 mmol)
and 1,3-cyclohexadiene (196 µL, 2 mmol) were added. Sam-
ples (50 µL) for gas chromatography (GC) analyses were ob-
tained at regular intervals. GC analyses were performed on
a Hewlett–Packard HP 5890A gas chromatograph equipped
with a cross-linked 5% diphenyl–95% dimethyl–polysiloxane
column (30 m, Restek GmbH, Rtx-5 Amine). GC–mass spec-
troscopy (GC-MS) analyses were performed on a Hewlett–
Packard HP 5890 gas chromatograph equipped with an iden-
tical column and a mass selective detector (HP 5971A). Peak
areas were referenced to n-dodecane as an internal standard.
Reactions at temperatures above the boiling point of toluene
were performed in a 300-cm3 Parr autoclave. The apparent ac-
tivation energy was determined in the temperature range 110–
200 ◦C, and the reaction order in aniline was determined in the
concentration range 20–140 mmol L−1. To study the influence
of the aniline basicity, aniline was replaced with substituted ani-
lines.
Scanning electron microscopy images were obtained on a
JEOL 500 SEM microscope. Samples were outgassed for 1 day
and sputtered with gold. Images were taken by operating the
microscope at 23.0 kV.
3. Results and discussion
3.1. Catalyst characterization
Three samples of zeolite H-BEA with different particle sizes
but similar Si/Al ratios (11.6–14.9), as well as H-ZSM-5, H-
Mordenite, and H-Y, and a series of zinc-exchanged Zn/H-BEA
zeolites (zinc content 0.03–0.54 mmol g−1), were used in this
study. To facilitate understanding of the performance of the zeo-
lites in catalysis, the H-BEA, H-ZSM-5, H-Mordenite, and H-Y
samples were fully characterized beforehand (Table 1). A de-
tailed characterization of the Zn/H-BEA zeolites was recently
published [32].
Particle sizes were determined from scanning electron mi-
croscopy images. The three H-BEA samples (abbreviated as
BEA1, BEA2, and BEA3) consisted of crystallites with parti-
cle sizes of 0.15–0.20, 0.20–0.25, and 0.60–0.70 µm, respec-
tively (Fig. 1). However, closer inspection showed that each
particle was an agglomerate composed of much smaller pri-
mary particles (approximately 50–70 nm in diameter). Micro-
pore volume slightly decreased with particle size (0.135, 0.129,
and 0.112 ml g−1, respectively), indicating that fewer pores
can be accessed as the crystallites become larger. H-ZSM-5,
2.2. Physical and analytical methods
The Si/Al ratio was determined by AAS using a UNICAM
939 spectrometer. Surface area, pore diameter, and pore size

44
O. Jimenez et al. / Journal of Catalysis 239 (2006) 42–50
Table 1
Physicochemical properties of the H-Zeolites used in this study
a
b
c
d
e
Catalyst
Structure Ratio Pore diameter Particle size Surface area MSA
Micropore volume BAS
LAS
TPD
(mmol
2
−1
2
−1
−1
−1
−1
−1
g )
NH3
Si/Al (Å)
(µm)
(m g
)
(m g
191
)
(ml g
)
(mmol
g
)
(mmol
g
)
Py
Py
H-ZSM-5
MFI
45
45
5.1 × 5.5 [100] 0.40–0.50
359
418
0.084
0.21
0.09
0.35
0.33
5.3 × 5.6 [010]
H-Mordenite MOR
6.5 × 7.0 [001] 0.50–0.60
2.6 × 5.7 [001]
284
0.138
0.14
0.02
H-Y
FAU
BEA
2.7 7.4 × 7.4 [111] 0.40–0.50
527
541
415
391
0.209
0.135
0.22
0.17
0.31
0.14
0.77
0.63
H-BEA1
14.2 6.6 × 6.7 [100] 0.15–0.20
5.6 × 5.6 [001]
H-BEA2
BEA
BEA
11.6 6.6 × 6.7 [100] 0.20–0.25
5.6 × 5.6 [001]
624
679
359
348
0.129
0.112
0.19
0.16
0.14
0.14
0.41
0.30
H-BEA3
a
14.9 6.6 × 6.7 [100] 0.60–0.70
5.6 × 5.6 [001]
The crystallographic direction is given in square brackets.
b
c
d
e
BET surface area determined by N adsorption.
Micropore surface area (MSA) determined by N adsorption.
Brønsted acid site (BAS) concentration determined by pyridine (Py) adsorption.
Lewis acid site (LAS) concentration determined by pyridine adsorption.
2
2
Fig. 1. SEM Micrographs of the H-BEA zeolites used in this study.
H-Mordenite, and H-Y had particle sizes in similar ranges
(0.40–0.50, 0.50–0.60, and 0.40–0.50 µm, respectively).
The acidity of the zeolite samples was assessed by TPD of
ammonia and from infrared spectra after adsorption of pyri-
dine. The TPD traces of ammonia desorbing from the three
H-BEA zeolites are presented in Fig. 2. The traces show two
desorption maxima at ca. 340 and 560 ◦C. The low-temperature
peak is characteristic for desorption of ammonia from weak
acid sites, whereas the high-temperature peak is related to
desorption of ammonia from strong acid sites. Deconvolution
of the two peaks using a linear combination of Gauss func-
tions [33] has shown that 0.48, 0.28, and 0.22 mmol g−1 of
ammonia desorbs from the weak acid sites and 0.15, 0.13, and
0.08 mmol g−1 of ammonia desorbs from the strong acid sites
for the BEA1, BEA2, and BEA3 samples, respectively. From
the entire amount of ammonia adsorbed on BEA1, the overall
acid site density was calculated to 0.63 mmol g−1. With in-
creasing particle size, this value decreased to 0.41 mmol g−1
for BEA2 and to 0.30 mmol g−1 for BEA3. This strongly sug-
gests that the pore system is less accessible for ammonia mole-
cules at larger particle sizes. However, because ammonia sorbs
unspecifically [34] on Brønsted acid sites (BAS) and Lewis
acid sites (LAS), assignment of the low- and high-temperature
peaks is somewhat ambiguous. The overall acid site densities
for H-ZSM-5, H-Mordenite, and H-Y were in the same range
(0.35, 0.33, and 0.77 mmol g−1, respectively).
Fig. 2. TPD profiles for desorption of ammonia from H-BEA zeolites.
The IR spectra of activated zeolites exhibited several bands
for isolated and bridging OH groups (Fig. 3). The most intense
bands, attributed to external SiOH groups, were observed at
3739 cm−1 for BEA1 and at 3742 cm−1 for BEA2 and BEA3.
The inherent high concentration of defect sites in BEA leads
to a high concentration of internal silanol groups, which then
become clearly visible in the IR spectrum. These OH groups
frequently form hydrogen bonds, which give rise to the band

O. Jimenez et al. / Journal of Catalysis 239 (2006) 42–50
45
Fig. 4. Infrared spectra of the H-BEA zeolites after pyridine adsorption.
Fig. 3. Infrared spectra of the hydroxyl region of the H-BEA zeolites.
the concentration of pyridine adsorbed on the LAS of BEA2
was reduced by approximately 16%, to 0.12 mmol g−1, whereas
for BEA1 and BEA3, the concentration of adsorbed pyridine
molecules decreased by 50%, to 0.07 and 0.08 mmol g−1, re-
spectively. Pyridine molecules, which gave rise to the band at
1443 cm−1, were completely removed by outgassing at 450 ◦C.
This observation is consistent with assignment to weakly acidic
sodium cations.
tailing from 3700 cm−1 down to 3300 cm−1 [35]. The band at
3606 cm−1, observed at nearly the same intensity in all sam-
ples (0.062, 0.078, and 0.067 normalized absorbance units for
BEA1, BEA2, and BEA3, respectively), corresponds to bridg-
ing hydroxyl groups, which give rise to strong Brønsted acidity
of the material [36]. For BEA2, the additional band with low
intensity observed at 3780 cm−1 is assigned to AlOH groups of
extra-framework alumina.
After pyridine adsorption at 150 ◦C, the two bands at 3606
and 3780 cm−1 vanished completely. This shows that these OH
groups are fully accessible to pyridine. In contrast, the peaks
in the silanol region remained nearly unchanged after pyri-
dine adsorption. This confirms the weak acidity of the external
silanol OH protons [37]. The broad adsorption band at 3700–
3400 cm−1 also appears to be sensitive to pyridine adsorption,
albeit to a much lesser extent.
The ring vibrations of adsorbed pyridine were used to quan-
tify the acid site concentration (Fig. 4). The BAS were identi-
fied by the peak at 1544 cm−1 (pyridinium ions). Only minor
differences in the BAS concentration were observed for the
three samples (0.16–0.19 mmol g−1; see Table 1). Note that
the BAS concentration was independent of the zeolite particle
size. After desorption of pyridine by heating to 450 ◦C for 1 h,
the concentration of pyridinium ions decreased dramatically. In
BEA1, all pyridine was removed by this procedure, whereas for
BEA2 and BEA3, the BAS concentration decreased to 0.04 and
0.03 mmol g−1, respectively. Thus, BEA2 and BEA3 contained
higher proportions of strong BAS than BEA1.
3.2. Catalytic measurements
The addition of aniline to 1,3-cyclohexadiene was studied
as model reaction for hydroamination. The three samples of
H-BEA with different particles sizes, a series of Zn/H-BEA ze-
olites with varying zinc contents, and various H-zeolites with
different pore diameters were used as catalysts. All zeolites
except H-ZSM-5 catalyzed the reaction, providing cyclohex-2-
enyl-phenylamine (1) as the main product. More detailed analy-
sis of the product mixture indicated the formation of structural
isomers with differing double-bond positions (Fig. 5).
At short reaction times, the formal 1,4 and 1,2 addition prod-
ucts 1 and cyclohex-3-enyl-phenylamine (2) were formed in
parallel. Note that the 2,1 addition product is indistinguish-
able from the 1,4 addition product. At longer reaction times,
isomer 1 became the main product, and the concentration of
isomer 2 decreased (Fig. 6). In parallel, the concentration of
cyclohex-1-enyl-phenylamine (3) increased. In contrast, the
formation of the Schiff-base cyclohexylidene-phenylamine (4)
was not observed. This can be rationalized by a much lower
thermodynamic stability of isomer 4, probably caused by steric
conflict between the β-protons of the phenyl and cyclohexyli-
dene ring, which prevents parallel alignment of the phenyl ring
and the C=N π-system. At very long reaction times, increas-
ing amounts of di-(cyclohex-2-enyl)-phenylamine (5) were ob-
served, resulting from the addition of two 1,3-cyclohexadiene
molecules to one aniline molecule.
The LAS generally associated with accessible Al3+ cations
gave rise to a peak at 1455 cm−1 (coordinatively bound pyri-
dine). In the spectra of pyridine adsorbed on BEA1 and BEA3,
a further peak at 1443 cm−1 was observed. This peak is prob-
ably related to sodium cations remaining from the prepara-
tion process [38]. Quantitative analysis showed that BEA1
and BEA3 contained 0.68 and 0.66 wt% sodium, respec-
tively, whereas the sodium concentration in H-BEA2 was
0.01 wt%. From the combined intensity of the peaks at 1443
The reaction profile can be well described based on the
model shown in Fig. 7, which enables one to derive the rate
constants of the individual reaction steps. For the model, irre-
and 1455 cm−1, the same LAS concentration (0.14 mmol g−1
was measured for all three samples. After outgassing at 450 ◦C,
)

46
O. Jimenez et al. / Journal of Catalysis 239 (2006) 42–50
Fig. 5. Possible reaction products from the addition of aniline to 1,3-cyclohexadiene.
Note that the fitted lines (Fig. 6) appear to slightly underesti-
mate the overall reactivity with respect to the disappearance of
the reactant and the appearance of product 1, whereas there is a
minor overestimation of their concentration at the end of the re-
action. This suggests that the catalysts deactivate slightly with
reaction time. An accurate estimation of the deactivation was
beyond the scope of this study, however.
For BEA2, isomer 1 (k1 = 0.94 (mmolL
)
−1 −1 h−1) was
formed three times faster than isomer 2 (k2 = 0.33 (mmol
L
)
−1 −1 h−1). Product 2 isomerised to 1 (k2−1 = 2.84 h−1) al-
most 10 times faster than it was formed. Isomer 1 isomerised
to 3 or reacted to the double-addition product 5 with rate con-
stants of k1−3 = 0.16 h−1 and k1−5 = 2.84 (mmolL
) ,
−1 −1 h−1
respectively. All zeolites except ZSM-5 followed the same re-
action sequence, in which the rate constant was higher for the
reaction 2 → 1 than for the formation of isomer 2.
Fig. 6. Reaction profile of the reaction between 1,3-cyclohexadiene and ani-
line using H-BEA as catalyst. The continuous lines describe the concentration
profile derived from the kinetic model given in Fig. 7.
The influence of potentially catalytically active Lewis acidic
cations, such as Zn2+, on the rate of intermolecular hydroam-
ination was explored for a series of ion-exchanged zeolites
Zn/H-BEA2 with zinc concentrations of 0.03–0.54 mmol g−1
.
versibility of the reactions and first order in all reactants were
assumed. In this respect, the reaction order for aniline was de-
The catalysts were initially selected because they showed excel-
lent catalytic properties for the intramolecular cyclisation of 6-
termined to 0.93 in the concentration range 20–140 mmol L−1
.
−1 −1 −1
Fig. 7. Kinetic model for the hydroamination of 1,3-cyclohexadiene with aniline using H-BEA2 as catalyst; units: k , k , k in (mmol L
)
h
; k , k
2−1 1−3
1
2
5
−1
in h
.

O. Jimenez et al. / Journal of Catalysis 239 (2006) 42–50
47
Fig. 8. Initial rate of formation of isomer 1 with a series of Zn/H-BEA catalyst
varying in zinc contents.
Fig. 9. Aniline conversion for the reaction between 1,3-cyclohexadiene and ani-
line catalyzed by different H-zeolites.
amino-1-hexyne [25,39]. The experiments were performed us-
ing the same substrate:catalyst ratio (in weight). Consequently,
the total amount of zinc in the reaction mixture was varied.
In contrast to our expectations, the initial catalytic activity
decreased with zinc loading. The initial rate of hydroamination
was 0.92 mmol g−Ca1t h−1 for the parent material and decreased to
To explore the potential role of the zeolite structure for
the addition of aniline to 1,3-cyclohexadiene, fully ammonia-
exchanged and -activated Brønsted acidic zeolites (without
metal cations at exchange positions) with different pore di-
ameters were examined. In particular, the catalytic activity of
H-ZSM-5, H-Mordenite, and H-Y was compared with that of
H-BEA zeolite. For all zeolites, the rate of reaction was reduced
(Fig. 9). No reaction was observed for H-ZSM-5.
0.59 mmol g−1
h
for Zn/H-BEA2 with a zinc concentration
−1
Cat
of 0.54 mmol g−1. This suggests that the zinc cations incorpo-
rated in BEA2 were catalytically not active or less active for
this reaction. Two regions of Zn2+ loading can be distinguished
with a linear decrease in the activity of the material (Fig. 8). For
the same series of zeolites, two regions of catalytic activity were
equally observed for the addition of aniline to phenylacetylene
and the cyclisation of 6-aminohex-1-yne, although the rate in-
creased with zinc loading [26].
The reactants are able to access the acid sites of all zeolites
used. Note, however, that the products face strong constraints
to diffuse out of the pores. This is especially pronounced for
H-ZSM-5, which has the smallest channel diameter. The larger,
but one-dimensional, pores of Mordenite led to low activity
with k1 and k2 values of 0.05 and 0.06 (mmol L
)
−1 −1 h−1, re-
spectively (Table 2), whereas 5, the product of double addition
of 1,3-cyclohexadiene to aniline, was not observed. H-Y zeo-
lite with a three-dimensional large pore system had nearly a
At low zinc concentrations, the zinc cations are exchanged
for protons, reducing the concentration of those BAS, which
exist so close to each other that the zinc cations can coordi-
nate to both sites [32]. Note the steeper decline of the catalytic
activity with increasing zinc concentration for these materials.
At higher zinc concentrations (ꢀ0.20 mmol g−1), the zinc is
present in the zeolite pores mainly as nanosized ZnO clusters,
which influence the Brønsted acidity of the zeolite to a lesser
degree [32].
The total concentration of LAS increased from 0.14 mmol
g−1 for the parent H-BEA zeolite (BEA2) to 0.20 mmol g−1
for the catalyst with the highest zinc loading [32]. In contrast,
the BAS concentration decreased with the zinc exchange de-
gree from 0.19 to 0.16 mmol g−1. The catalytic activity closely
followed the trend in the concentration of BAS. Thus, the
BAS seem to be responsible for the catalytic activity, whereas
the Zn2+ cations in the zeolite are inactive for the reaction.
It should be noted, however, that a molecular Brønsted acid
(trifluoromethanesulfonic acid) did not catalyze the reaction.
These observations indicate that a reaction intermediate is sta-
bilized in the zeolite pores. In addition, LAS resulting from the
presence of aluminium in the material may play a role in the
catalytic cycle.
sixfold-higher reaction rate (k1 0.33 (mmol L
0.33 (mmol L
−1 −1 h−1) than H-Mordenite; however, the su-
percage of H-Y is sufficiently large to allow formation of the
double-addition product 5 [k5 1.11 (mmol L
−1 −1 h−1; 38%
)
−1 −1 h−1 and k2
)
)
yield after 24 h]. Zeolite BEA, on the other hand, has two inter-
connecting channel systems forming a smaller space than the
faujasite supercage. Consequently, the reactants and the desired
products can diffuse into and out of the pores, and the reac-
tion rate is higher (k1 and k2 0.33 (mmol L
product of the double addition is formed in lower amounts
[k5 0.59 (mmol L
−1 −1 h−1; 17% yield after 24 h]. The re-
)
−1 −1 h−1). The
)
sults suggest that the hydroamination reaction can be catalyzed
most efficiently in the pores of a zeolite with 12-membered ring
openings, such H-BEA zeolite, and that subtle shape selective
effects determine reactivity and selectivity.
The influence of particle size was studied for the series of
H-BEA zeolites with increasing particle sizes and similar BAS
concentrations. The rates of reaction decreased with particle
size (Fig. 10). Because the diameter of the micropores strongly
influences the activity, we exclude the possibility that the reac-
tion occurs exclusively at the pore entrance. The product for-

48
O. Jimenez et al. / Journal of Catalysis 239 (2006) 42–50
Table 2
Rate constants for single reactions steps and initial reaction rate for the hydroamination of 1,3-cyclohexadiene with aniline calculated on basis of the kinetic model
given in Fig. 7
a
Catalyst
Hydroamination
Isomerisation
Initial rate
r
ini
k
k
k
k
k
1
2
5
2−1
1−3
−1
−1
)
−1 −1 −1
−1 −1 −1
−1 −1 −1
−1
−1
(mmol g
h
((mmol L
)
h
)
((mmol L
)
h
)
((mmol L
)
h
)
(h
)
(h
)
Cat
H-ZSM-5
H-Mordenite
H-Y
H-BEA1
H-BEA2
H-BEA3
0.00
0.05
0.33
2.15
0.94
0.19
0.00
0.06
0.33
1.08
0.33
0.07
0.00
0.00
1.11
1.02
0.59
0.00
0.00
7.06
6.18
3.54
2.84
3.61
0.00
0.04
0.09
0.13
0.16
0.18
0.00
0.20
0.58
1.63
0.92
0.20
a
Initial rate of aniline consumption.
1,3-cyclohexadiene, is preferentially adsorbed. However, it is in
equilibrium with the free BAS and can be displaced by the ex-
cess of 1,3-cyclohexadiene in the reaction mixture.
According to Yang et al. [40] alkenyl carbenium ions can
be formed by direct protonation of dienes on acidic catalysts.
The carbenium ions are strongly stabilised in zeolite by in-
teraction with the BAS and more closely resemble an alkoxy
group. For 1,3-cyclohexadiene [41], coordination to one of the
BAS provides either a cation with delocalized charge (a) or the
less-stable cation b [42]. Subsequent nucleophilic attack of the
aniline on a or b leads to the intermediates a1 and b1, respec-
tively. Subsequently, one of the ammonium protons is trans-
ferred to the BAS, and the products 1 and 2 are desorbed. Two
possibilities exist for the nucleophilic attack during reaction
pathway a, providing the formal 1,4 and 2,1 addition products.
These two products are indistinguishable. Only reaction path-
way b accounts for formation of the 1,2 addition product. After
desorption, the primary reaction products 1 and 2 diffuse into
the bulk solution and are readsorbed before isomerisation. This
is also thought to occur on the BAS. An alternative reaction
sequence with hydroamination and isomerisation in succession
(without desorption and readsorption of the intermediates) can
be excluded based on the kinetic model.
Fig. 10. Aniline conversion for the reaction between 1,3-cyclohexadiene and
aniline catalyzed by H-BEA zeolites with different particle size.
mation constants for the BEA zeolites used in this study are
reported in Table 2. The rate constants k1 and k2 clearly de-
creased with particle size. This is attributed to lower diffusive
limitations in BEA1, whereas significant diffusion constrains
appear likely for BEA2 and BEA3. Product 5, which has a
much larger kinetic diameter, is probably formed exclusively
at the outer surface. The rate constant k5 also decreased with
particle size, confirming our conclusions on diffusion limita-
tions. Formation of product 5 was not observed for BEA3. In
contrast, isomerisation constants k2−1 and k1−3 are apparently
unaffected by particle size.
3.4. Influence of amine basicity
Because the reaction is acid-catalyzed, the basicity of the
reacting amine could drastically change the reactant–catalyst
interaction. Therefore, the influence of aniline basicity on the
hydroamination of 1,3-cyclohexadiene was studied. Anilines
with various substituents on the aromatic ring were used, par-
ticularly those with substituents in a para position to the NH2
group. This allows us to change the basicity of the aniline with-
out changing the minimum kinetic diameter of the molecule.
Thus, it can be assumed that the para-substituted anilines have
similar diffusion rates inside the zeolite pores.
For BEA2, low apparent activation energies of 45 kJ mol−1
based on aniline and 38 kJ mol−1 based on the main product
of the reaction were determined in the temperature range 110–
200 ◦C. These relatively low activation energies are consistent
with pore diffusion limitations.
The reaction rate decreased linearly with the basicity (lower
pKb) of the substituted aniline (Table 3; Fig. 12). 2-Methoxyani-
line and parent aniline deviated from the correlation to lower
and higher rates, respectively.
According to the proposed mechanism, aniline and 1,3-cy-
clohexadiene compete for coordination to the BAS. Because
the substituted anilines are much more basic than 1,3-cyclo-
hexadiene, formation of the anilinium ions is preferred. How-
3.3. Proposed mechanism
The data obtained from the kinetic experiments suggest
the mechanism shown in Fig. 11 for the intermolecular hy-
droamination of 1,3-cyclohexadiene with anilines using ze-
olite H-BEA as catalyst. The simplest kinetic model would
start with a competitive adsorption between aniline and 1,3-
cyclohexadiene on the BAS. Aniline, being a stronger base than

O. Jimenez et al. / Journal of Catalysis 239 (2006) 42–50
49
Fig. 11. Proposed reaction mechanism for the intermolecular hydroamination of 1,3-cyclohexadiene with aniline catalyzed by H-BEA zeolites.
Table 3
Basicity (pK ) of the arylamines used and rate of reaction in the hydroamina-
b
tion of 1,3-cyclohexadiene
a
Aromatic amine
Substituent
pK
Initial rate
(mmol
Isomer
ratio
(%)
b
b
−1
−1
g
h
)
Cat
4-Methoxyaniline
4-Ethylaniline
4-Fluoroaniline
Aniline
2-Methoxyaniline
3-Methoxyaniline
4-Chloroaniline
Ethyl-4-aminobenzoate
para –OMe
para –Et
para –F
8.7 0.19
8.9 0.29
9.3 0.48
9.4 0.92
9.5 0.17
9.8 0.79
10.0 0.82
90/5/5
94/6/0
97/3/0
90/3/7
93/0/7
94/3/3
94/0/6
100/0/0
90/0/10
100/0/0
para –H
ortho –OMe
meta –OMe
para –Cl
para –COOEt 10.5 1.15
4-(Trifluoromethyl)aniline para –CF
4-Nitroaniline
11.2 1.07
13.0 1.80
3
para –NO
2
a
Fig. 12. Correlation between the initial rate of the addition of aniline to 1,3-cy-
clohexadiene and the basicity of the aniline.
Initial rate of aniline consumption.
After 24 h reaction time; ratio of isomers 1, 2, and 3, respectively.
b
to the shorter length of products 1 and 2 relative to the corre-
sponding products of the substituted anilines, thus providing for
a higher diffusion rate.
ever, it is assumed that the anilinium ion may be displaced from
the BAS by excess 1,3-cyclohexadiene. Less-basic amines form
less-stable anilinium ions and are more readily displaced. As-
suming equal diffusion rates, the linear correlation reflects the
concentration of sites a and b in the material.
4. Conclusion
Deviation of 2-methoxyaniline from the above relationship
is tentatively explained by the larger minimum kinetic diameter
of the N-(cyclohex-2-enyl)-2-methoxyaniline (6), which hin-
ders 6 from diffusing out of the pores. In addition, aniline with
a methoxy substituent in the ortho position may entail greater
constraints for the transition state during the nucleophilic attack
(partial shielding of the –NH2 group). In the same way, the rel-
atively high rate observed for the parent aniline might be due
Zeolite catalysts with 12-membered ring openings, such
as H-BEA, can efficiently catalyze the reaction between ani-
line and 1,3-cyclohexadiene. Generally, the rate is higher for
electron-poor (i.e., less-basic) anilines. Thus it seems appar-
ent that the acid-catalysed reaction between 1,3-cyclohexadiene
and (the much more basic) aliphatic amines is more difficult to
realize. Mechanistically, the key step seems to be the adsorp-
tion of 1,3-cyclohexadiene at the BAS and protonation to the

50
O. Jimenez et al. / Journal of Catalysis 239 (2006) 42–50
corresponding allyl or enyl cation. More-basic anilines adsorb
more strongly at BAS, leading to a lower concentrations of pro-
tonated 1,3-cyclohexadiene molecules and, consequently, lower
rates of reaction.
[19] M. Kawatsura, J.F. Hartwig, J. Am. Chem. Soc. 122 (2000) 9546.
[20] See, e.g., T.E. Müller, J.A. Lercher, V.N. Nguyen, AIChE J. 49 (2003)
214.
[21] M. Utsunomiya, J.F. Hartwig, J. Am. Chem. Soc. 126 (2004) 2702.
[22] K. Tanabe, W.F. Hölderich, Appl. Catal. A 181 (1999) 399.
[23] A. Chauvel, B. Delmon, W.F. Hölderich, Appl. Catal. 115 (1994) 173.
[24] V. Taglieber, W.F. Hölderich, R. Kummer, W.D. Mross, G. Saladin, US
4929758 (1990), to BASF AG.
Acknowledgments
[25] J. Penzien, T.E. Müller, J.A. Lercher, Microporous Mesoporous Mater. 48
(2001) 285.
[26] J. Penzien, C. Haeßner, A. Jentys, K. Köhler, T.E. Müller, J.A. Lercher,
J. Catal. 221 (2004) 302.
[27] V. Neff, T.E. Müller, J.A. Lercher, J. Chem. Soc., Chem. Commun. 8
(2002) 906.
The authors thank Xaver Hecht and Martin Neukamm for
experimental support.
References
[1] T.E. Müller, M. Beller, Chem. Rev. 98 (1998) 675.
[2] P.W. Roesky, T.E. Müller, Angew. Chem. 115 (2003) 2812.
[3] F. Alonso, I.P. Beletskaya, M. Yus, Chem. Rev. 104 (2004) 3079.
[4] J. Seayad, A. Tillack, C.G. Hartung, M. Beller, Adv. Synth. Catal. 344
(2002) 795.
[5] T.E. Müller, in: I.T. Horváth (Ed.), Encyclopedia of Catalysis, Wiley, New
York, 2002, p. 492.
[6] R. Taube, in: B. Cornils, W.A. Herrmann (Eds.), Applied Homogeneous
Catalysis with Organometallic Compounds, vol. 1, VCH, Weinheim, 1996.
[7] For mechanistic background see, e.g., H.M. Senn, P.E. Blöchl, A. Togni,
J. Am. Chem. Soc. 122 (2000) 4098.
[8] J.R. Petersen, J.M. Hoover, W.S. Kassel, A.L. Rheingold, A.R. Johnson,
Inorg. Chim. Acta 358 (2005) 687.
[9] L.K. Vo, D.A. Singleton, Org. Lett. 6 (2004) 2469.
[10] B. Kalita, K.M. Nicholas, Tetrahedron. Lett. 46 (2005) 1451.
[11] A. del Zotto, W. Baratta, A. Felluga, P. Rigo, Inorg. Chim. Acta 358 (2005)
2749.
[12] N.T. Patil, N.K. Pahadi, Y. Yamamoto, Tetrahedron. Lett. 46 (2005) 2101.
[13] D.P. Klein, A. Ellern, R.J. Angelici, Organometallics. 23 (2004) 5662.
[14] L.M. Lutete, I. Kadota, Y. Yamamoto, J. Am. Chem. Soc. 126 (2004) 1622.
[15] S. Hong, T.J. Marks, J. Am. Chem. Soc. 124 (2002) 7886.
[16] T. Minami, H. Okamoto, S. Ikeda, R. Tanaka, F. Ozawa, M. Yoshifuji,
Angew. Chem. 113 (2001) 4633.
[17] J. Bodis, T.E. Müller, J.A. Lercher, Green Chem. 5 (2003) 227.
[18] O. Löber, M. Kawatsura, J.F. Hartwig, J. Am. Chem. Soc. 123 (2001)
4366.
[28] S. Breitenlechner, M. Fleck, T.E. Müller, A. Suppan, J. Mol. Catal. A 214
(2004) 175.
[29] L. Borreto, M.A. Camblor, A. Corma, J. Perez-Pariente, Appl. Catal. 82
(1992) 37.
[30] K.S.N. Reddy, B.S. Rao, V.P. Shiralkar, Appl. Catal. 95 (1993) 53.
[31] C.A. Emeis, J. Catal. 141 (1993) 347.
[32] J. Penzien, A. Abraham, J.A. van Bokhoven, A. Jentys, T.E. Müller,
C. Sievers, J.A. Lercher, J. Phys. Chem. B 108 (2004) 4116.
[33] D.M. Roberge, H. Hausmann, W.F. Hölderich, Phys. Chem. Chem. Phys. 4
(2002) 3128.
[34] C. Costa, J.M. Lopes, F. Lemos, F. Ramoa Ribeiro, J. Mol. Catal. A 144
(1999) 221.
[35] J.P. Marques, I. Gener, P. Ayrault, J.C. Bordado, J.M. Lopes, F. Ramoa-
Ribeiro, M. Guisnet, Microporous Mesoporous Mater. 60 (2003) 251.
[36] J.B. Higgins, R.B. La Pierre, J.L. Schlenker, A.C. Rohrman, J.D. Wood,
G.T. Kerr, W.J. Rohrbaugh, Zeolites 8 (1988) 446.
[37] R.J. Gorte, Catal. Lett. 62 (1999) 1.
[38] J.A. Lercher, G. Ritter, H. Vinek, J. Colloid Interface Sci. 106 (1985) 215.
[39] J. Penzien, T.E. Müller, J.A. Lercher, J. Chem. Soc., Chem. Commun. 18
(2000) 1753.
[40] S. Yang, J.N. Kondo, K. Domen, Catal. Today 73 (2002) 113.
[41] S. Spange, S. Adolph, R. Walther, Y. Zimmermann, J. Phys. Chem. B 107
(2003) 298.
[42] J. Pawlas, Y. Nakao, M. Kawatsura, J.F. Hartwig, J. Am. Chem. Soc. 124
(2002) 3669.