3
2
54
ALLAHVERDIEV, IRANDOUST, AND MURZIN
.71% K2O, 9.04% H2O, 2.42% CaO, 9.35% Al2O3 (by 18.11 � 20.13 � 7.42 and 24.60 A˚ , correspondingly showed
2
weight). The BET surface area of the catalyst was 14 m /g. a similar selectivity pattern. Moreover selectivities of these
The starting reactant (�-pinene) was purchased from zeolites were very much like our data on clinoptilolite,
Arizona Chemical OY (Finland), distilled, and kept under which has a much lower cell parameter. Therefore, pos-
vacuum. In all experiments 200 ml of �-pinene, contain- sible conformational diffusion effects that may exist play a
ing 97.29 wt% of �-pinene, 1.21 wt% of camphene, and minor role.
1
.49 wt% of limonene, was used.
The reaction was carried out in an autoclave under nitro-
gen pressure. The reaction mixture was stirred vigorously
RESULTS AND DISCUSSION
with a mixing speed of 1200 rpm. In a typical run 2 g of cata- Catalytic Activity
lyst with a fraction size of 50–100 �m and 200 ml of �-pinene
As noted above, kinetic studies of �-pinene isomeriza-
were charged in the reactor, which was flushed with nitro-
gen before charging for 0.5 h. A fresh portion of catalyst
was taken for each experiment. Experiments were carried
tion are rather sparse. The �-pinene consumption on TiO2
was reported to be of zero order (2). It was also supposed
(2) that the apparent reaction order in �-pinene isomeriza-
tion changesdepending on conversion. Two different values
of activation energy were reported, for the initial period,
20 kJ/mol, and for the final period, 168 kJ/mol (2).
Typical kinetic curves obtained in our experiments are
presented in Figs. 2–5. It was found that with increased
temperature the catalytic activity for the isomerization of
-pinene increased. Surprisingly pressure also had an in-
�
out under isobaric (1–25 bars) and isothermal (100–160 C)
dead-end conditions. Pressure and temperature were kept
�
constant with an accuracy of � 0.5 bar and � 0.5 C, respec-
1
tively.
A PC controlled the whole reactor. During the course of
the reaction, several samples were taken out of the reactor,
using a special sampling line. These samples were analyzed
by FID gas chromatography (Hewlett Packard 5890A with
�
fluence on reaction rate in the lower pressure region; after
5–10 bars the reaction rate was totally independent of ni-
0
.25 mm � 60 m CP-Wax 52 CB capillary column). Tem-
�
�
perature programming was applied (10 min 60 C, 2 C/min
�
�
�
trogen pressure. At high temperatures (above 135 C) no
from 60 to 180 C, and 5 min at 180 C). A PC also controlled
the GC, and the analytical peaks were calculated by means
of available software. Experimental errors were within 1% .
It is known that in three-phase catalytic reactions the
following processes occur: dissolving of gases in the liquid,
diffusion ofreactantsto and productsfrom the outer surface
of the catalyst, and diffusion in pores (18–21).
influence of pressure on reaction rate could be observed.
The influence of nitrogen pressure on reaction rate can-
not be explained by an increase of nitrogen solubility in
the reaction mixture, since the solubility is rather low and
the mole fraction of nitrogen in the liquid phase can be to-
tally neglected. At the same time catalyst selectivity was
not effected by a change in pressure. Dependence of ln of
The influence of external diffusion was determined fol-
lowing the published procedure (21), which was previously
applied to liquid-phase hydrogenation reactions. To this
end the external mass transfer coefficient was calculated us-
ingthe equation proposed in (22). Due to a rather high value
�
-pinene concentration on reaction time at different val-
ues of temperatures and pressures is presented in Fig. 6.
Detailed analysis of experimental data showed that the ki-
netics of �-pinene consumption can be described by first-
order kinetics. Kinetic parameters for the first-order kinetic
equation were determined using a least-squares method,
incorporated in a parameter estimation software. Depen-
dence of the first-order kinetic constant as a function of
temperature is presented in Fig. 7. This strongly supports
the assumption of first-order kinetics. The global activation
energy of �-pinene consumption was 80.9 kJ/mol.
�
9
2
of the �-pinene diffusion coefficient (� 10 m /s), small
particle size of catalyst, vigorous stirring (and thus a high
value of specific mixing power), and absence of diluent (e.g.,
high initial concentration of reactant), the rate of external
diffusion was sufficient. The influence of internal diffusion
for porous materials was estimated using the conventional
criteria reported in (23). The results of calculations showed
that under our conditions, the influence of internal mass
transfer was negligible. Experiments were also performed
with catalyst particle sizes of 0.075–0.1 and 1–2 mm. It was
Selectivity and Reaction Network
It was stated in (2) that the product composition re-
quite clear that activity and selectivity depended on particle mained unchanged until conversion reached 70 wt% . The
size, thus giving additional support to the choice of size of same author reported (3) that monocyclic terpenes consist
the clinoptilolite, which was applied in our kinetic studies. mainly of limonene and terpinolene. Secondary reactions
The general approach, that was used to verify the in- (e.g., transformations of camphene into tricyclene and fur-
fluence of mass transfer, however, cannot rule out com- ther isomerization of limonene) were thought to occur only
pletely possible conformational diffusion effects typical at high conversions.
for zeolites. On the other hand, dealuminated mordenite
As follows from our experimental data (Figs. 2–5), the
and Y zeolites (12), e.g., zeolites with a cell parameter main reaction products were camphene and limonene.