Reductive amination of cyclohexanone in the presence of cyclohexanol over zeolites Hβ and HY

Article abstract of DOI:10.1016/j.molcata.2010.01.013

Reductive amination of a mixture of cyclohexanone/cyclohexanol in varying proportions has been carried out in the gas phase over zeolites Hβ and HY. The products identified were cyclohexylamine, N-cyclohexylcyclohexanimine, 2-cyclohexen-1-ylcyclohexanone and 2-cyclohexylcyclohexanone. The product distribution during the experiments indicates that cyclohexanol does not undergo reductive amination over acid catalysts; it only forms a condensation product with cyclohexanone. The reaction rates were obtained from experimental data and fit to a kinetic model derived for this reaction. The fits show that this reaction follows a Langmuir-Hinshelwood pathway by the adsorption of both cyclohexanone and the NH₃ on the surface of the zeolite.

Full text of DOI:10.1016/j.molcata.2010.01.013

Journal of Molecular Catalysis A: Chemical 321 (2010) 15–21
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Reductive amination of cyclohexanone in the presence of cyclohexanol over
zeolites H␤ and HY
a,∗
b
c
a
Sharath R. Kirumakki , Maria Papadaki , Komandur V.R. Chary , N. Nagaraju
a
Department of Chemistry, St. Joseph’s College P.G. Center, 46 Langford Road, Bangalore 560027, India
Department of Environmental & Natural Resources Management, University of Ioannina, Seferi 2, Agrinio, GR 30100, Greece
Catalysis Division, Indian Institute of Chemical Technology, Tarnaka, Hyderabad 500007, India
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 13 May 2009
Received in revised form
Reductive amination of a mixture of cyclohexanone/cyclohexanol in varying proportions has been car-
ried out in the gas phase over zeolites H␤ and HY. The products identified were cyclohexylamine,
N-cyclohexylcyclohexanimine, 2-cyclohexen-1-ylcyclohexanone and 2-cyclohexylcyclohexanone. The
product distribution during the experiments indicates that cyclohexanol does not undergo reductive ami-
nation over acid catalysts; it only forms a condensation product with cyclohexanone. The reaction rates
were obtained from experimental data and fit to a kinetic model derived for this reaction. The fits show
that this reaction follows a Langmuir–Hinshelwood pathway by the adsorption of both cyclohexanone
and the NH3 on the surface of the zeolite.
2
9 December 2009
Accepted 21 January 2010
Available online 28 January 2010
Keywords:
Reductive amination
Zeolites
©
2010 Elsevier B.V. All rights reserved.
Cyclohexanone
Cyclohexanol
Reaction kinetics
1
. Introduction
Cyclohexylamine can undergo condensation reaction with cyclo-
hexanone to form N-cyclohexylcyclohexanimine (D). Apart from
the aminated products during the reaction cyclohexanone can
also form condensation products with cyclohexanone to form 2-
cyclohexen-1-ylcyclohexanone (E) and with cyclohexanol to form
2-cyclohexylcyclohexanone (F). Cyclohexanol may also undergo
dehydrogenation to cyclohexanone over a metal catalyst [10].
Amination of cyclohexanol/cyclohexanone mixture is important
because this mixture is the product of cyclohexane oxidation
[11,12]. Becker et al. have studied the gas phase reductive ami-
nation of cyclohexanol/cyclohexanone to cyclohexylamine and
aniline over Cu- and Ni-containing catalysts [13]. Reductive ami-
nation of cyclohexanol has also been studied over acidic and
A number of commercial processes are employed in amine syn-
thesis. Amines are produced by alkylation of organic halides with
ammonia, reductive amination of carbonyl compounds, hydrocya-
nation of alkenes followed by reduction, hydroamination of olefins,
etc. [1–4]. Amongst these processes reductive amination being a
single-step process, which employs inexpensive starting materials,
is the most practical and convenient approach. The transformation
of alcohols, ketones or aldehydes to the corresponding secondary
or primary alkylamines occurs in a single reactor without isolation
of the intermediary imines or hydroxyamines. Reductive amination
is generally a heterogeneous process [5]. It involves the reaction of
a carbonyl compound with ammonia and hydrogen over a catalyst,
+
Ni-containing metallosilicate catalysts [14]. Bronsted acid (H ) was
found to be inactive for this reaction.
◦
at temperatures ranging from 150 to 350 C and pressures of 18 to
2
00 atm.
Cyclohexanone can be aminated in the gas phase over acidic
Shape selective zeolites, which catalyze alkylation of ammonia
with alcohols, have attracted significant attention for long years
[15–17]. This interest is due to their potential to replace the tra-
ditional silica-alumina catalysts over which bulky trialkylamines
can be formed, with zeolites, which are able to produce mono-
and dialkylamines selectively. Zeolites have also been applied to
the amination of phenol to aniline [18–20]. Zeolite ␤ and ZSM-5
displayed high activity for this reaction [19,20]. Zeolite ␤ demon-
strated higher activity than ZSM-5, which was attributed to its
relatively weak acidity [20]. Zeolite Y containing UO22+ cations has
been used as a catalyst for the amination of 1-octanol [21]. The ami-
nation of 1-octanol was tested on zeolites having MFI structure and
metal catalysts [6–8]. As can be seen in Scheme 1, cyclohex-
anone (A) reacts with ammonia to form cyclohexylideneimine (B),
which in turn is readily hydrogenated to cyclohexylamine (C) [9].
∗ Corresponding author. Current address: Ascend Performance Materials, F.M.
2
917, P.O. Box 711, Alvin, TX 77512, USA. Tel.: +1 281 228 4346;
fax: +1 281 228 4860.
E-mail addresses: srkiru@ascendmaterials.com, skirumakki@gmail.com
(
S.R. Kirumakki).
1
381-1169/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2010.01.013

1
6
S.R. Kirumakki et al. / Journal of Molecular Catalysis A: Chemical 321 (2010) 15–21
Scheme 1. Reaction pathways for the formation of products in the reductive amination of cyclohexanone/cyclohexanol mixture.
found to proceed through a nucleophilic substitution pathway [22].
However, the conversion achieved was not high. Although zeolites
have been successfully employed for the amination of alcohols,
the low pressure amination of cyclohexanol was found to result
in dehydrogenation products only; amination products were not
formed [23].
In order to overcome the limitations of the present genera-
tion of zeolite catalysts with respect to their activity during the
reductive amination of cyclohexanol, it is necessary to understand
the complex reaction sequences occurring during acid-catalyzed
aminations in the molecular sieve pores. Cyclohexanone is an inter-
mediate in the formation of cyclohexylamine from cyclohexanol
The reductive amination reactions were performed in a dif-
ferential down-flow fixed bed Pyrex glass reactor (length 35 cm,
diameter 1.5 cm) at atmospheric pressure. A known mass of the
catalyst (0.25–1.0 g) of mesh size 14–20 was introduced into the
reactor, the zone (∼5 cm) above the catalyst bed packed with quartz
grains being served as the preheater. The reactor was placed in a
tubular furnace with temperature control. A thermocouple placed
close to the catalyst bed was used to measure the reaction tem-
−
1
perature. The catalyst was activated under air flow (25 ml min
)
at 723 K for 2 h before the start of the experiment. The reaction
was carried out in the temperature range of 473–623 K. The liq-
uid feed consisting of a mixture of cyclohexanone/cyclohexanol
was charged into the reactor using a syringe pump (Braun Mel-
sugun, Germany) at a constant predetermined rate together with
NH3 and H2 gases whose flow rates were also constant during
the reaction. The liquid products were collected at the bottom
for 1 h using an ice-trap. Analyses of the products were done on
a GC–MS apparatus (HP 5973 Mass selective detector, column:
HP-1MS 15 m × 0.25 mm × 0.25 ␮m). The data were recorded after
1 h of steady-state operations. The products identified were cyclo-
hexylamine (C), N-cyclohexylcyclohexanimine (D), 2-cyclohexen-
1-ylcyclohexanone (E) and 2-cyclohexylcyclohexanone (F). The
major products being N-cyclohexylcyclohexanimine (D) and 2-
cyclohexen-1-ylcyclohexanone (E), their concentrations were
measured together with those of the unreacted cyclohexanol and
cyclohexanone. The total carbon balance was found to be >98%.
Reactions using pure cyclohexanol gave no reductive amination
products. They only resulted in dehydrogenation products in low
quantities, and thus conversions were calculated based on cyclo-
hexanone concentration only. The conversion and selectivity were
calculated based on the GC results.
[
14]. The aim of this work is to systematically study the reductive
amination of cyclohexanone/cyclohexanol mixtures over acidic
zeolite catalysts under atmospheric pressure in order to identify the
conditions under which the amination products C and D are formed.
To fulfill this objective three different types of zeolites with strong,
medium and weak acid active sites were employed in the current
study. More specifically, in the research presented here, reductive
amination of cyclohexanol/cyclohexanone mixtures over zeolites
␤, Y and ZSM-5 (Si/Al = 8, 25 and 30, respectively) were studied.
The kinetics of reductive amination of a starting mixture of opti-
mum proportion, shown later, over the aforementioned catalysts
was also studied and the overall rate of its reductive amination is
presented.
2
. Experimental
Zeolites ␤, Y and ZSM-5 were used in the reductive amina-
tion of cyclohexanone/cyclohexanol. Proton forms of the zeolites Y
Si/Al = 25) and ZSM-5 (Si/Al = 30) were obtained from Conteka, The
(
Netherlands. The sodium form of zeolite ␤ (Si/Al = 8) was obtained
from United Catalysts India Limited. It was converted to its proton
3. Results and discussion
form via an ion exchange process using an aqueous 1 M NH NO3
4
Initial experiments were performed using different cyclo-
solution at 353 K for 12 h, which was then followed by calcination in
air at 823 K. BET surface area of the zeolite samples was measured
on a Micromeritics Pulse Chemisorb 2700 instrument using N2 as
the adsorbate. Acidity of the zeolites was measured by stepwise
temperature programmed desorption (STPD) of ammonia on the
same instrument; the relative final strengths of acidity were estab-
lished by bracketing the temperature of desorption. The measured
properties of the zeolites used are given in Table 1.
hexanone/cyclohexanol proportions in order to choose the
cyclohexanone/cyclohexanol ratio that result in the optimum
selectivity for the amination products for this reaction. All reac-
tions were carried out over 0.5 g of zeolite HY. The results are
shown in Table 2. The conversion of cyclohexanone increased
when its concentration in the feed decreased. A feed containing
only cyclohexanol did not yield any products under the cur-
rent reaction conditions. When only cyclohexanone was used
in the feed, its conversion was 40% and it increased to around
7
1% in a feed mixture containing 10 mol% cyclohexanone and
Table 1
Physicochemical properties of zeolites.
90 mol% cyclohexanol. The variation in the product selectivity was
quite interesting. When cyclohexanone was in excess in the feed,
2
Zeolite
BET surface area (m /g)
Acidity (mmol/g) NH3
2
-cyclohexen-1-ylcyclohexanone (E), was formed by the conden-
Weak (A)
Medium (B)
Strong (C)
sation of two cyclohexanone molecules; when cyclohexanol was
in excess, 2-cyclohexylcyclohexanone (F) was formed by the con-
densation of one molecule of cyclohexanol and one molecule of
cyclohexanone. Cyclohexylamine (C) was detected only when the
H␤
HY
HZSM-5
434
730
400
1.03
1.51
1.89
2.3
1.70
1.16
4.50
0.99
3.05

S.R. Kirumakki et al. / Journal of Molecular Catalysis A: Chemical 321 (2010) 15–21
17
Table 2
Conversion of cyclohexanone (A) and the product distribution during the reductive amination reaction using different % of cyclohexanone in a cyclohexanone/cyclohexanol
feed.
%
A in feed
Conv. A (%)
Selectivity (%)
Cyclohexyl amine (C)
N-cyclohexyl cyclohexanimine(D)
2-Cyclohexen-1-ylcyclohexanone (E)
2-Cyclohexyl cyclohexanone (F)
1
00
40
–
–
–
59.2
69.4
95.2
90
63
54
40.8
30.6
4.8
–
–
–
–
–
–
10
24
31
46
9
8
7
5
4
1
0
0
0
0
0
0
46.2
55.5
57.9
59.6
62.3
71.4
–
13
15
19
35
–
Reaction conditions: HY = 0.5 g, feed:NH3:H2 = 1:15:15, feed rate = 2 ml h−1, temperature = 523 K.
cyclohexanone concentration in the feed was 50% or less. Cyclo-
hexylamine reacts with excess cyclohexanone during the reaction
to form N-cyclohexylcyclohexanimine (D). The selectivities for
the condensation products E and F were the lowest when the
cyclohexanone concentration was 80%, which corresponds to a
cyclohexanone/cyclohexanol molar ratio of 8:2. This molar ratio
in the feed was used for further studies.
ture range. Beyond this temperature range cyclohexene formation
was observed. Since the variation in the conversion-selectivity pat-
tern is similar over all three zeolites, only zeolite HY was used to
optimize the other reaction conditions.
3.3. Effect of catalyst mass
The effect of zeolite catalyst HY mass was studied and the
results are shown in Fig. 3. The catalyst mass was varied from
3.1. Effect of catalyst type
0
.25 to 1.0 g. The reactions were performed at 523 K and cyclo-
The variation in the cyclohexanone conversion and the prod-
hexanone/cyclohexanol ratio of 8:2 (mol/mol) was employed. As
can be seen in Fig. 3, the conversion increased steadily from 19.5
to 80% as catalyst mass increases from 0.25 to 1.0 g. The yield of
D increased from 8.7 to almost 82% and yield of E decreased from
11 to 2.5% as catalyst mass increased from 0.25 to 1.0 g. This study
clearly shows that HY facilitates the formation of D and suppresses
that of E. The only products in measurable concentrations were D
and E. It is worth mentioning again that in the absence of a cata-
lyst the cyclohexanone conversion was 35% and the selectivity for
E was 96%.
uct selectivity during the reductive amination in the absence of
any catalyst (blank) and when zeolites H␤, HY and HZSM-5 are
employed were studied. Results are shown in Fig. 1. As can be seen
in the figure, zeolites H␤ and HY resulted in a higher conversion
than HZSM-5. Moreover, the selectivity for the reductive amina-
tion products was high on these two zeolites. The selectivity for D
was 95% over HY compared to 90% over H␤. It is interesting to note,
however, that in the absence of any catalyst the condensation reac-
tion producing E predominated. This indicates that the formation
of product E is a facile reaction, which is hindered by the zeolites.
3.4. Effect of time on stream
3.2. Effect of reaction temperature
The effects of time on stream on reductive amination at 523 K
The effect of temperature on this reductive amination reac-
tion over each catalyst was studied in the temperature range of
73–673 K. Results are shown in Fig. 2. As can be seen, the con-
version of cyclohexanone (closed triangular marks) increases with
temperature over all three zeolite catalysts. It increased from 45
to 64% over H␤, 51 to 69% over HY and 32 to 45% over HSM5. The
selectivity for D reached a maximum in the temperature range of
over all three catalysts employed in this research were studied. The
results obtained after 1 h of steady-state operation over HY zeo-
lite are shown in Fig. 4. As can be seen, the conversion decreased
steadily from 56% after 60 min time on stream to 26% after 240 min.
4
5
23–573 K with a concomitant decrease in the selectivity for E. The
selectivity for D was >90% over zeolites H␤ and HY in this tempera-
Fig. 2. Effect of temperature on reductive amination of cyclohexanol/cyclohexanone
over different catalyst types. H␤ (—), HY (—) and HZSM-5 (- - -). Reaction con-
ditions: feed = cyclohexanone/cyclohexanol 8:2 (mol/mol), temperature = 523 K.
Catalyst mass = 0.5 g, feed:NH3:H2 = 1:15:15, feed rate = 2 ml h . A (ꢀ) = conversion
of cyclohexanone, D (ꢁ) = N-cyclohexylcyclohexanimine, E (ꢂ) = 2-cyclohexen-1-
ylcyclohexanone.
Fig. 1. Effect of catalyst type on reductive amination of cyclohex-
anol/cyclohexanone. Reaction conditions: feed = cyclohexanone/cyclohexanol
−
1
8
feed
:2 (mol/mol), temperature = 523 K. Catalyst mass = 0.5 g, feed:NH3:H2 = 1:15:15,
−1
rate = 2 ml h
.
D = N-cyclohexylcyclohexanimine,
E = 2-cyclohexen-1-
ylcyclohexanone.

1
8
S.R. Kirumakki et al. / Journal of Molecular Catalysis A: Chemical 321 (2010) 15–21
Fig. 3. Effect of catalyst mass on reductive amination of cyclohex-
anol/cyclohexanone. Reaction conditions: feed = cyclohexanone/cyclohexanol
8
feed
:2 (mol/mol), temperature = 523 K. Catalyst = HY, feed:NH3:H2 = 1:15:15,
−1
Fig. 5. Effect of NH3 on reductive amination of cyclohexanol/cyclohexanone, over
rate = 2 ml h
.
A = cyclohexanone,
D = N-cyclohexylcyclohexanimine,
HY zeolite. Reaction conditions: feed = cyclohexanone/cyclohexanol 8:2 (mol/mol).
E = 2-cyclohexen-1-ylcyclohexanone.
−
1
Temperature = 523 K, HY catalyst mass = 0.5 g, feed:H2 = 1:15, feed rate = 2 ml h . A
ꢀ) = cyclohexanone, D (ꢁ) = N-cyclohexylcyclohexanimine, E (ꢂ) = 2-cyclohexen-1-
ylcyclohexanone.
(
The selectivity for D also decreased from 95 to 40% with a corre-
sponding increase in that for E. This implies that the active sites
on the zeolite HY were deactivated with time on stream. Similar
trends were also observed over zeolites H␤ and HZSM-5.
3
.6. Effect of H2
The effect of H2 on the reductive amination was stud-
−
1
ied by varying the H2 flow rate from 0 to 4 mol h . The
cyclohexanone/cyclohexanol feed rate was 2 ml h
−
1
3.5. Effect of NH3
(15 mmol
−
1
cyclohexanone h ), the feed to NH3 ratio was kept constant at
1:15 (mol/mol), and 0.5 g of zeolite HY was employed. Results
are shown in Fig. 6. The conversion of cyclohexanone increased
marginally from 52 to 57% with increasing H2 concentration from
1 to 4 mol h . An increase in the H2 flow rate above 1 mol h did
not produce any significant increase in the selectivity for D. The
selectivity was always greater than 80%.
The effect of NH3 on the reductive amination was studied by
−
1
varying the NH3 flow rate from 0 to 4 mol h . The cyclohex-
−1
anone/cyclohexanol feed rate was 2 ml h ; this corresponds to
−1
−1
−1
a cyclohexanone flow rate of 15 mmol h
and the feed to H2
ratio was kept constant at 1:15 (mol/mol). 0.5 g of zeolite HY
was used to catalyze this reaction. Results are shown in Fig. 5.
The conversion of cyclohexanone increased with increase in NH3
concentration. It reached a maximum of 55% when NH concen-
3.7. Effect of feed rate
3
−1
tration was 2.5 mol h . Further increase in NH3 concentration led
to a decrease in the conversion. This may be due to zeolite acid
sites neutralization by excess NH3 molecules. The selectivity for
D constantly increased with NH3 concentration while the selec-
tivity for E dropped. This is presumably due to NH3 concentration
rate-dependence of the reaction resulting in the formation of D.
The effect of the feed rate on the reductive amination was stud-
ied by varying the cyclohexanone/cyclohexanol feed rate from 1 to
Fig. 4. Effect of time on stream in reductive amination of cyclo-
Fig. 6. Effect of H2 flow rate on reductive amination of cyclohexanol/cyclohexanone,
hexanol/cyclohexanone
feed = cyclohexanone/cyclohexanol
HY = 0.5 g, feed:NH3:H2 = 1:15:15, feed rate = 2 ml h
D (ꢁ) = N-cyclohexylcyclohexanimine, E (ꢂ) = 2-cyclohexen-1-ylcyclohexanone.
over
HY
8:2
catalyst.
(mol/mol),
Reaction
conditions:
over HY zeolite. Reaction conditions: feed = cyclohexanone/cyclohexanol 8:2
temperature = 523 K.
(mol/mol). Temperature = 523 K, HY catalyst mass = 0.5 g, feed:NH3 = 1:15, feed
−1
−1
.
A
(ꢀ) = cyclohexanone,
rate = 2 ml h
. A (ꢀ) = cyclohexanone, D (ꢁ) = N-cyclohexylcyclohexanimine, E
(ꢂ) = 2-cyclohexen-1-ylcyclohexanone.

S.R. Kirumakki et al. / Journal of Molecular Catalysis A: Chemical 321 (2010) 15–21
19
Fig. 8. Reductive amination of cyclohexanol/cyclohexanone: effect of
catalyst mass on reaction rate. H␤ (ꢁ), HY (ꢃ). Reaction conditions:
feed = cyclohexanone/cyclohexanol
feed:H2:NH3 = 1:15:15, feed rate = 4 ml h
8:2
.
(mol/mol).
Temperature = 523 K,
−
1
Fig. 7. Effect of feed rate on reductive amination of cyclohexanol/cyclohexanone,
over HY zeolite. Reaction conditions: feed = cyclohexanone/cyclohexanol 8:2
a differential (low-conversion) reactor was employed. The reactor
was fed with a cyclohexanone/cyclohexanol mixture (8:2 mol/mol)
(
mol/mol). Temperature = 523 K. HY catalyst mass = 0.5 g, H2:NH3 = 1:1, feed
−1
rate = 2 ml h
(
.
A
(ꢀ) = cyclohexanone,
D
(ꢁ) = N-cyclohexylcyclohexanimine,
E
−
1
at the rate of 4 ml h over 0.25 g of the catalyst. Conversion data
after 1 h of steady-state operation at 573 K were collected, and the
rate of the reaction in the differential reactor was calculated. The
data were used in the following equation to calculate the reaction
rate
ꢂ) = 2-cyclohexen-1-ylcyclohexanone.
− −1
ml h (7.5–37.5 mmol h of cyclohexanone) with the NH :H
3
1
5
2
−1
ratio kept constant at 2.25 mol h
each. 0.5 g of zeolite HY was
used. Results are shown in Fig. 7. The feed rate had an important
effect on the conversion as well as the selectivity levels in this
(
Xio − X ) · F
i
r = −
Catalyst mass (g)
−
1
reaction. Increase in the feed rate from 1 to 5 ml h caused a 30%
drop of cyclohexanone conversion (from 59 to 38%). Although NH3
and H2 were always kept in a substantial excess, a dramatic drop
of the selectivity for D was observed (from 95 to 56%) accompa-
nied by a concomitant rise in the selectivity for E (from 5 to 44%).
This indicates that as the contact time of the feed with the catalyst
decreases the conversion of cyclohexanone as well as the forma-
tion of D decreases whereas the formation of E increases. This again
underlines the importance of the catalyst in the formation of D.
where (Xio − X ) is the conversion of cyclohexanone and F = molar
i
flow rate of cyclohexanone.
Fig. 8 gives a plot of the reaction rate as a function of zeolite
mass. The mass of zeolites was varied in the range of 0.25–0.75 g,
a linear relationship exists between the mass and the rate of the
reaction. This indicates that diffusion limitations are not dominant.
It must also be noted that in the absence of catalyst 2-cyclohexen-1-
ylcyclohexanone (E) is formed whose dimensions are very similar
to D. Thus, in the analysis that follows we assume that the rate-
determining step is the chemical reaction step. Very small variation
3.8. Kinetic studies
in the conversion of cyclohexanone with H concentration indicates
2
that it plays only a minor role in this reaction, and thus the kinetics
should depend primarily on adsorbed cyclohexanone and NH3.
Kinetic studies of the amination reactions over zeolite catalysts
have been reported before [24–27]. According to these reports the
reaction is first order with respect to ammonia over Na-mordenite
For the reductive amination reaction, where the rate-limiting
step is the chemical reaction, the reaction rate can be described
by the equation using the Langmuir–Hinshelwood (L–H) model
and assumptions. The rate equation can be derived based on the
following (additional to L–H ones) assumptions:
[
24] as well as other H-zeolites [25]. Kogelbauer et al. employed
a Langmuir–Hinshelwood model to describe the kinetics of ami-
nation of methanol over Na-mordenite and Na,K-erionite catalysts
[
26]. Ilao et al. [28] suggested that a reaction of adsorbed methanol
and ammonia upon sequential adsorption and heating took place
over zeolite chabazite but not in zeolite mordenite. Based on this
a Langmuir–Hinshelwood mechanism was proposed for chabazite
and an Eley-Rideal one for mordenite. Baiker et al. [9] have used
the Langmuir model to describe the kinetics of Cu catalyzed ami-
nation of long-chain aliphatic alcohols. Becker et al. [13] have
performed kinetic experiments for the amination of the cyclohex-
anone/cyclohexanol mixture, which are indicative of a pseudo-first
order. They stated that up to 523 K, the rate-limiting step was
the chemical reaction; at higher temperatures however, the mass
transfer resistance predominated.
Becker et al. [13] have also stated that over supported Cu cat-
alysts, mass transfer resistance strongly influences the selectivity
of this reaction. They, moreover, stated that bulky products like N-
cyclohexylcyclohexanimine, (D) are preferred when mass transfer
limits the reaction rate. However this was not found to be the case
over zeolites H␤ and HY. HZSM-5 was not chosen for kinetic stud-
ies as this catalyst was not found to be very active in this reaction.
This may be because the smaller pore size in ZSM-5 may hinder
the formation of the products of this reaction. To study the kinetics
•
•
Hydrogen is not adsorbed or very weakly adsorbed on the zeolite
surface during the reaction.
Due to the excess of cyclohexanone in the feed and the basic
nature of cyclohexanone when compared to cyclohexanol, the
adsorption of cyclohexanol on the acid sites of zeolites is negligi-
ble.
For the reductive amination reaction the following elementary
steps are proposed:
C6H10O or (A) + S ꢄ C H10O–S or (A–S)
(I)
(II)
6
NH or (N) + S ꢄ NH –S or (N–S)
3
3
A–S + N–S → C6H10NH–S or (B–S) + H2O + S
B–S + H → C H NH –S or (C–S)
(III)
(IV)
(V)
2
6
11
2
C–S + C H O–S → C H11N–H10C –S or (D–S) + H O + S
6
10
6
6
2
D–S ꢄ D + S
(VI)

2
0
S.R. Kirumakki et al. / Journal of Molecular Catalysis A: Chemical 321 (2010) 15–21
It is moreover assumed that C H10O and NH3 react with the
6
catalyst site S via very fast steps, always in equilibrium to form
intermediates C H10O–S and NH –S, respectively. These interme-
6
3
diates react further according to reactions (III)–(V), to give adsorbed
products C H NH–S and C H NH –S, which subsequently react
6
10
6
11
2
to give adsorbed C H11N–H10C –S. C H11N–H10C6–S is subse-
6
6
6
quently desorbed via a very fast step also always in equilibrium.
As mentioned earlier in this section, the amination of cyclo-
hexanone, reaction (III), is considered the rate-limiting step.
Adsorption/desorption reactions (I), (II) and (VI) are considered to
be very fast and always in equilibrium.
C6H10O–S is also consumed via its reaction with cyclohexanol
for the production of C12H20O (F). However, at the selected temper-
ature and conditions of reaction no measurable amount of C12H20
F) was produced.
For reactions (I), (II) and (VI) which have been considered to be
O
(
Fig. 9. Reductive amination of cyclohexanol/cyclohexanone at 523 K. Fitting of data
to Eq. (10) for catalysts H␤ and HY.
always in equilibrium:
CAS = K C Cs
(1)
(2)
(3)
Table 3
A
A
Kinetic data for the reductive amination of cyclohexanone/cyclohexanol over zeo-
lites H␤ and HY.
CNS = KNCNCS
CDS = KDCDCS
−
4
−1 −1
Nonlinear regression, R2
Catalyst
ks (10 mol g
s
)
Ks
H␤
HY
1.48
1.62
0.87
1.02
0.921
0.936
where C is the concentration of substance i, and K its equilibrium
i
j
constant at 573 K. From a balance over the active catalyst sites Ctot
and subsequent replacement of concentrations of intermediates
from relations (1), (2) and (3)
where ks = kt/K2 and K = K /K.
S
A
Ctot = C + C + CNS + CDS = C (1 + K C + KNCN + KDCD). Therefore,
S
AS
S
A A
The nonlinear curve fitting of the above kinetic equation was
performed utilizing the Levenberg–Marquardt method. A plot of
the reaction rate as a function of the concentration of cyclohex-
anone is shown in Fig. 9. As can be seen in Fig. 9 there is a very good
correlation between the fit and the experimental results. The rate
constants and the adsorption constants are given in Table 3. The
rate constant over HY is greater than that over H␤. The adsorption
constants indicate that the adsorption of cyclohexanone is slightly
stronger over HY than over H␤.
Ctot
+ K C + KNCN + KDCD
As mentioned earlier, the amination of cyclohexanone, reac-
tion (III), is considered the rate-limiting step. Thus the rate of A–S
destruction is:
CS = 1
(4)
A
A
dCAS
−
= rIII + rV
(5)
dt
rIII = kIIICASCNS
rV = kV CCSCAS
(6)
(7)
4. Mechanism
where r is the rate of elementary reaction j and k its rate coef-
According to Corbin et al. [29] during the amination of methanol,
j
j
ficient. Time is depicted as t. The concentration of C H11NH2 was
not measurable. Evidently reaction (5) is very fast but because the
the most probable mechanism for the methylamine formation is an
SN2 attack of a protonated methanol by weakly adsorbed ammo-
nia; i.e. the weak acid sites were found to be responsible for this
step. Kotrla et al. [30] have reported a correlation between zeo-
lite acidity and activity in the amination of acetone over zeolites.
Zamlynny et al. [14] have proposed a mechanism for cyclohex-
anol amination over Ni catalysts where cyclohexanol is assumed
to be dehydrogenated over Ni particles to cyclohexanone, which is
further aminated over acid sites to imine-like species and hydro-
6
concentration of C H11NH2 is practically zero, the contribution of
6
reaction (5) in the rate of destruction of A–S is negligible. In other
words rIII ꢀ rV and rIII + rV ≈ rIII. The rate, r, of A–S destruction thus
is:
2
kIIIK KNCtot C CN
2
S
A
A
r = kIIICASCNS = kIIIK KNC CNC
=
(8)
2
A
A
(
1 + K C + KNCN + KDCD)
A
A
0
In the conditions of the experiments the concentration of NH is
genated, again over Ni , to cyclohexylamine. In the present study
3
1
5 times higher than that of cyclohexanone. Therefore the change
the reductive amination of cyclohexanone is found to follow a L–H
mechanism where both cyclohexanone and NH3 are adsorbed on
adjacent Bronsted acid sites. The fact that both zeolites HY and H␤
are active for this reaction implies that the active sites are the
weak and the moderate acid and not the strong acid sites. This
is based on the assumption that since both these zeolites have a
12-membered pore opening, the difference in the pore structures
do not contribute to a difference in their catalytic activity for this
reaction. An SN2 type attack of NH3 leads to the formation of the
imine. N-cyclohexylcyclohexanimine (D) is the dimerization prod-
ucts of cyclohexylamine with cyclohexanone. This reaction route
has already been suggested by Hamada et al. [10]. The data col-
lected via the present research support Scheme 1 as the probable
reaction route. The variation in the selectivity for different prod-
ucts when the feed composition is varied can be explained based
on this pathway.
in NH concentration can be considered negligible and the product
3
KNCN can be taken as constant. Its concentration in the numera-
tor can also be incorporated in a overall constant as shown below.
Moreover weak product adsorption is expected from a good cata-
lyst surface so that the products can be easily desorbed. It is thus
assumed that adsorption of D is much weaker than that of A, so that
KD ꢁ K , and thus the above relation can be reduced to:
A
ktCA
r =
(9)
2
(
K + K C )
A
A
where kt = kIIIK KNC2totCN and K = 1 + KNCN, or by dividing numera-
tor and denominator by K , the final rate equation becomes:
A
2
ksCA
r =
(10)
2
(
1 + KsCA)

S.R. Kirumakki et al. / Journal of Molecular Catalysis A: Chemical 321 (2010) 15–21
21
5
. Conclusions
[7] N.S. Kozlow, L. Akhmetschina, Dokl. Akad. Nauk SSSR 85 (1952) 91.
[
8] E.J. Kaniewska, A.M. Kowalczyk, R.E. Modzelewski, W.K. Szelejewski, Stud. Surf.
Sci. Catal. 41 (1988) 387.
9] A. Baiker, W. Caprez, W.L. Holstein, Ind. Eng. Chem. Prod. Res. Dev. 22 (1983)
217.
The gas phase reductive amination of a mixture of cyclo-
[
hexanone/cyclohexanol in varying proportions has been carried
out over zeolites H␤ and HY. The product distribution during
the experiments indicates that cyclohexanol does not undergo
reductive amination over acid catalysts; it only forms a conden-
sation product with cyclohexanone. The product distribution and
cyclohexanone conversion were greatly influenced by the cyclo-
hexanone/cyclohexanol ratio. The highest selectivity for reductive
amination was seen when the cyclohexanone:cyclohexanol mole
ratio was 8:2. The reaction rates were obtained from experimen-
tal data and fit to a kinetic model derived for this reaction based on
the Langmuir–Hinshelwood pathway. This reaction proceeds by the
[
[
10] H. Hamada, Y. Kuwahara, T. Sato, K. Wakabayashi, Bull. Chem. Soc. Japan 60
1987) 55.
(
11] B.R. Bluestein, J.M. Nelson, B. Lawrence, US Patent 3.551. 487 (1970), Witco
Chemical Company.
[12] N. Goetz, L. Hupfer, W. Hoffmann, M. Baumann, EP 0 050 229 B1 (1982) BASF
AG.
13] J. Becker, J.P.M. Niederer, M. Keller, W.F. Hoelderich, Appl. Catal A: Gen. 197
(2000) 229.
[14] V. Zamlynny, L. Kubelkova, E. Baburek, K. Jiratova, J. Novakova, Appl. Catal. A:
Gen. 169 (1998) 119.
15] A.B. van Geysel, W. Musin, in: B. Elvers, S. Hawkins, G. Schultz (Eds.), Ullmann’s
Encyclopedia of Industrial Chemistry, vol. A16, 5th ed., VCH, Weinheim, 1985,
p. 535.
16] M.G. Turcotte, T.A. Johnson, in: J.J. Kroschwitz (Ed.), Kirk Othmer Encyclopedia
of Chemical Technology, vol. 2, 4th ed., John Wiley & Sons, New York, 1992, p.
[
[
[
adsorption of both cyclohexanone and the NH on the surface of the
3
zeolite. The adsorbed species interact to give the condensation and
reductive amination products.
369.
[17] L.D. Pesce, W.R. Jenks, in: J.A. Kent (Ed.), Riegel’s Handbook of Industrial Chem-
istry, 9th ed., Van Nostrand Reinhold, New York, 1992, p. 1109.
[
[
[
18] M. Yamamoto, N. Takamiya, S. Murai, Nippon Kagaku Kaishi (1974) 2135.
19] C.D. Chang, W.H. Lang, Eur. Patent 62542A1 (1982) to Mobil Oil Corp.
20] N. Katada, S. Ijima, H. Igi, M. Niwa, Stud. Surf. Sci. Catal. 105 (1997) 1227.
References
[
[
1] J. March, Advanced Organic Chemistry, 4th ed., Wiley, New York, 1992, p 768
and references therein.
2] J.P. Collman, B.M. Trost, T.R. Veroeven, in: G. Wilkinson, F.G.A. Stone (Eds.), Com-
prehensive Organometallic Chemistry, vol. 8, Pergamon Press, Oxford, 1982, p.
[21] A. Azzouz, D. Nibou, B. Abbad, M. Achache, J. Mol. Catal. 68 (1991) 187.
[22] C. Dume, W.F. Hoelderich, Appl. Catal. A: Gen. 183 (1999) 167.
[23] L. Brabec, J. Novakova, L. Kubelkova, J. Mol. Catal. 94 (1994) 117.
[24] F.J. Weigert, J. Catal. 103 (1987) 20.
8
92, and references therein.
[25] D.T. Cheng, L. Zhang, J.M. Kobe, C. Yi, J.A. Dumesic, J. Mol. Catal. 93 (1994) 337.
[26] A. Kogelbauer, C. Grundling, J.A. Lercher, J. Phys. Chem. 100 (1996) 1852.
[27] A. Kogelbauer, C. Grundling, J.A. Lercher, Stud. Surf. Sci. Catal. 84 (1994) 1475.
[28] M.C. Ilao, H. Yamamoto, K. Segawa, J. Catal. 161 (1996) 20.
[29] D.R. Corbin, S. Schwarz, G.C. Sonnichsen, Catal. Today, 37 71 (1997).
[30] J. Kotrla, D. Nachtigallova, V. Bosa c¸ ek, J. Novakova, Phys. Chem. Chem. Phys. 1
(1999) 2613.
[
[
3] T.E. Muller, M. Beller, Chem. Rev. 98 (1998) 675.
4] M.S. Gibson, in: S. Patai (Ed.), The Chemistry of Amino Groups, Interscience,
New York, 1968, p. 61.
5] A.W. Heinen, J.A. Peters, H. van Bekkum, Eur. J. Org. Chem. 13 (2000)
2
[
[
501.
6] P.S. Sabatier, A. Mailhe, Comptes Rendus 153 (1912) 1204.