The influence of reaction temperature on the oscillatory behaviour in the palladium-catalysed phenylacetylene oxidative carbonylation reaction

Article abstract of DOI:10.1039/b905444h

This paper reports the influence of reaction temperature on the occurrence and characteristics of pH oscillations that are observed during the palladium-catalysed phenylacetylene oxidative carbonylation reaction in a catalytic system (PdI₂, KI,

Full text of DOI:10.1039/b905444h

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/pccp | Physical Chemistry Chemical Physics
The influence of reaction temperature on the oscillatory behaviour in the
palladium-catalysed phenylacetylene oxidative carbonylation reaction
Katarina Novakovic,*a Ankur Mukherjee,a Mark Willis,a Allen Wrighta and
Steve Scottb
Received 18th March 2009, Accepted 30th June 2009
First published as an Advance Article on the web 4th August 2009
DOI: 10.1039/b905444h
This paper reports the influence of reaction temperature on the occurrence and characteristics of
pH oscillations that are observed during the palladium-catalysed phenylacetylene oxidative
carbonylation reaction in a catalytic system (PdI2, KI, air, NaOAc) in methanol. Isothermal
experiments were performed over the temperature range 10–50 1C. The experiments demonstrate
that oscillations occur in the range 10–40 1C and that a decrease in reaction temperature results in
an increase in the period and amplitude of the pH oscillations. Furthermore, it is observed that
during oscillations at any specific temperature, the time taken for pH to increase from a minimum
to a maximum value varies with respect to reaction time. However, the time required for the pH
to fall from maximum to new minimum is approximately constant with respect to the reaction
time and is a function of the reaction temperature.
temperature behaviours, tend to compensate for the influence
Introduction
of temperature on period length. Rabai and Hanazaki studied
The oscillatory nature of the palladium-catalysed phenyl-
temperature compensation in the oscillatory hydrogen peroxide–
thiosulfate–sulfite flow system.13 Inside the oscillatory region
acetylene oxidative carbonylation (PCPOC) reaction was
studied by Temkin and colleagues in a series of papers
at a fixed flow rate and input concentration they found that the
investigating the oxidative carbonylation of alkynes.1–4 It
amplitude of oscillations decreases gradually with increasing
was demonstrated that the PCPOC reaction in a catalytic
temperature while period length remains constant. While
system (PdI2, KI, O2, NaOAc in methanol) exhibits oscillations
average period length remained constant the high pH stage
in redox potential, pH and the rate of CO–O2 gas mixture
of
a period appeared to be longer and the low pH
consumption. Novakovic and colleagues5–7 were the first to
investigate the characteristics of the PCPOC reaction, when it
was carried out in a calorimeter. Novakovic et al. reported
simultaneous oscillations in pH and rate of heat evolution
(Qr). The oscillations in Qr were exothermic and no corres-
ponding endotherm was observed. The heat release was in
opposite phase with a pH fall and decreased as the pH
increased.5
stage becomes shorter with increasing temperature. Some
significant deviation from the average period was measured
along the borders of the oscillatory region. Nagao et al.14
reported an experimental observation of strongly non-
Arrhenius behaviour, namely temperature over-compensation
and compensation in the oscillatory oxidation of formic acid
on a polycrystalline platinum electrode in sulfuric acid
aqueous media. An increase in period accompanied
a
Of all the oscillatory reactions, the most documented is the
Belousov–Zhabotinsky (BZ) reaction.8 The BZ reaction
maintains a prolonged state of non-equilibrium leading to
macroscopic temporal oscillations and spatial pattern
formation that is very life-like.9 Belousov himself investigated
the influence of temperature on the period of oscillations in the
BZ reaction and he showed that the period decreases with
increasing temperature.10
Ruoff used the Brusselator11 to demonstrate that an
increase in rate constants belonging to positive feedback
reactions increases oscillation frequency (decrease period
length), while increasing rate constants of negative feedback
reactions will lower frequency (increase period length).12 Ruoff
concluded that for a certain set of activation energies positive
and negative feedback reactions will, due to their opposing
temperature increase which was much more pronounced for
low applied currents and weaker at high currents. The effect of
temperature and applied current on the oscillation amplitude
was reported to be more complicated. For the lowest
current investigated, the amplitude increased linearly with
temperature. They reported this increase to be less pronounced
for higher currents, and a slight decrease is observed at the
highest current value studied.
The PCPOC reaction is an oscillating system of higher
complexity than those studied previously and is of importance
as it presents a rare example of oscillatory behaviour in reactions
catalysed by metal complexes.1 Furthermore, it represents the
first example of complex molecules synthesised from relatively
simple reagents proceeding in a catalytic system in an oscillatory
mode. The products resulting from this reaction are shown in
Scheme 1. These are Z-2-phenyl-but-2-enedioic acid dimethyl
ester 1, 2-phenyl-acrylic acid methyl ester 2, E-3-phenyl-acrylic
acid methyl ester 3, E-2-phenyl-but-2-enedioic acid dimethyl
ester 4 and 3-phenyl-5H-furan-2-one 5.
a School of Chemical Engineering and Advanced Materials,
Newcastle University, UK. E-mail: katarina.novakovic@ncl.ac.uk
b Department of Chemistry, University of Leeds, UK
ꢀc
This journal is the Owner Societies 2009
9044 | Phys. Chem. Chem. Phys., 2009, 11, 9044–9049

View Article Online
(99.9+%), product number 34 860. The gas cylinders
(CO (99.9%) and air) were purchased from BOC.
Procedure
Methanol (400 ml) and PdI2 (430 mg, 1.19 mmol) were added
to the vessel and stirred at 250 rpm. The temperature was set
and maintained at the specified reaction temperature. After
stirring the mixture for 45 min, KI (37.39 g, 225 mmol) and
NaOAc (114 mg, 1.40 mmol) in 50 ml of methanol were added.
After a further 15 min, purging of the system commenced.
Both the CO and air were simultaneously introduced through
dip tubes at controlled flow rates of 50 ml minꢁ1. 20 min after
the purging started, phenylacetylene (6.2 ml, 56.5 mmol)
was added. In some experiments (30, 40, 50 1C) up to 80 g
(at 50 g minꢁ1) of methanol was added at a later stage to
compensate for evaporative loss.
Scheme 1 Reaction products.
These products have been isolated, purified by flash chromato-
graphy and compared to known standards using 1H NMR
spectroscopy.7
Novakovic et al.6 reported the influence of oscillations on
product selectivity as well as the dynamics of product formation.
When operating in an oscillatory pH regime product
formation is suppressed until oscillations occur, after which
there is selective formation of Z-2-phenyl-but-2-enedioic acid
dimethyl ester. When operating in a non-oscillatory pH mode
selectivity is poor and several other products are formed. For
the same initial conditions oscillatory and non-oscillatory
regimes were achieved by changing the amount of catalyst.
In this paper we demonstrate the influence of reaction
temperature on the oscillations in pH during the PCPOC
reaction and show that oscillations may occur over a wide
temperature range including the normal temperature of the
human body. Therefore temperature may be a potential
control variable that could be used to assist prospective
applications of this oscillatory system, e.g. as stimuli for
pH and temperature responsive smart materials, controlled
transport of nano-objects, creation of devices able to propel
themselves etc.15
The influence of temperature on the period and
amplitude of the pH oscillations
The first experiment was performed at 10 1C and as can be
seen in Fig. 1, several sets of oscillations were observed.
The first set commenced approximately 2030 min after the
beginning of the experiment and lasted for around 37 h.
Within the first set of pH oscillations two characteristic shapes
are present. The first is saw-like with the largest period of
110 min and amplitude of 2.1 pH units. Following this,
oscillations with progressive increases in pH were recorded.
Here, the largest period of oscillation observed was
approximately 140 min with amplitude of 2.8 pH units. The
next set of oscillations commenced following a sharp fall in the
pH at approximately 4255 min and lasted for approximately
46 h. The largest period of an oscillation recorded during this
set of oscillations was 182 min and this had amplitude of
3.1 pH units. Yet another set of oscillations was then recorded
after a new sharp fall in pH at approximately 7025 min. The
largest period of an oscillation recorded here was 81 min with
amplitude 0.7 pH units. 133 h from the beginning of the
experiment, although in oscillatory mode, the run was stopped
and pH recording terminated.
Experimental
Equipment and chemicals used
All experiments were performed in an HEL SIMULARt
reaction calorimeter. The reactor is a one litre double jacketed
glass vessel with a slightly narrowing diameter walls (average
90 mm). Agitation is achieved through stirring (0–600 rpm;
glass pitched blade impeller +45 mm;), precise temperature
control is achieved using an external cooling system with oil
flowing through the jacket together with a regulated electrical
heater (0–150 W) inside the vessel. In addition the reactor is
equipped with solid and liquid dosing units, a liquid sampling
unit, an inlet gas mass flow controller, a temperature probe
(Pt 100) and a pH electrode. Other recorded variables were
reaction temperature, pH, oil inlet and outlet temperatures,
and agitation speed. The reactor system was closed but
not sealed allowing close to atmospheric pressure to be
maintained.
For the second experiment the reaction was maintained at
20 1C. As can be seen in Fig. 2, at this temperature only one set
of oscillations is present. It is possible, therefore that the sets
of oscillations observed at 10 1C have now merged into one.
These oscillations started 2117 min after the beginning of
the experiment and lasted for approximately 72 h. Within the
oscillations two characteristic shapes may be distinguished.
The first is again saw-like with the largest period being 71 min
and a corresponding amplitude of 1.8 pH units. Following a
transition period, pH oscillations exhibiting progressive pH
increase are again observed. Here, the largest period recorded
was approximately 60 min with amplitude 2.4 pH units.
As can be seen in Fig. 3 (an experiment at 30 1C) and 4 (an
experiment at 40 1C) only a single set of oscillations was
observed at these temperatures. For the 30 1C experiment,
the oscillations started 928 min after the beginning of the
experiment and lasted for approximately 24 h. The largest
period of oscillation that was recorded was approximately
21 min with amplitude 0.8 pH units. For the 40 1C experiment,
All chemicals were purchased from Sigma Aldrich and used
as received: phenylacetylene (98%), product number 117 706;
palladium(II) iodide (99.99+%), product number 203 963;
potassium iodide (99.0+%), product number 221 945; sodium
acetate (99.0+%), product number 241 245; methanol
ꢀc
This journal is the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 9044–9049 | 9045

View Article Online
Fig. 1 pH oscillations at 10 1C.
Fig. 2 pH oscillations at 20 1C.
the oscillations commenced after 1115 min and lasted for
Discussion and conclusions
approximately 11 h. Once established the largest period of
an oscillation in pH was approximately 12 min with amplitude
0.6 pH units.
This work has demonstrated that oscillations in pH may be
achieved when the PCPOC reaction is performed at various
temperatures within the range 10–40 1C. Experiments below
10 1C were not performed as a suitable pH probe was not
available and it was found through experimentation that at
50 1C oscillatory behaviour ceased. The characteristics of the
The final experiment was conducted at 50 1C. Fig. 5
shows the on-line recorded pH for this experiment and
as may be observed from this figure oscillations were not
present.
ꢀc
This journal is the Owner Societies 2009
9046 | Phys. Chem. Chem. Phys., 2009, 11, 9044–9049

View Article Online
Fig. 3 pH oscillations at 30 1C.
Fig. 4 pH oscillations at 40 1C.
to a maximum value varies throughout the experiment while
the time taken for pH to decrease is characteristic of temperature
and at a given temperature it remains constant regardless of
the period of the oscillation (see Table 1, column pH fall and
Fig. 6).
In Fig. 6, the oscillations have been aligned to demonstrate
that at a specific temperature the time taken for the pH to fall
is equal. To further validate this observation a non-isothermal
experiment was performed—see Fig. 7. In this experiment the
operating temperature was set at 40 1C for the first 2600 min of
the experiment after which it was reduced to 20 1C.
Fig. 5 Run at 50 1C.
oscillations obtained from the experimental runs are
summarised in Table 1.
As can be seen in Fig. 7b and c once the temperature is
reduced from 40 to 20 1C there is an increase in both amplitude
(from 0.1 to 0.3 pH units) and period (from 11 to 55 min) of
the recorded oscillations. It is also apparent that within the
oscillations observed at 40 and 20 1C the time taken for pH
to decrease remains approximately 2.5 min and 9 min,
respectively (Fig. 7d). This confirms the observations made
after the isothermal runs, namely that: (a) the decrease in
It may be seen from Table 1 that an increase in temperature
results in a decrease in the maximum period and amplitude
recorded during the pH oscillations. At the same time it may
be noticed that the pH interval within which oscillations
occur widens at lower temperatures. In addition, within one
oscillation, the time taken for pH to increase from a minimum
ꢀc
This journal is the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 9044–9049 | 9047

View Article Online
Table 1 Summary of the results obtained for the isothermal experiments in the temperature range 10–50 1C
pH/pH units
Time/min
Within oscillations
Oscillations
Started Stopped
Oscillations
Start
End
Max period
pH fall
T/1C
Min–max range
Max amplitude
a
10
20
30
40
50
3.78
3.13
1.21
1.61
—
2.0–6.0
2.0–4.8
1.21–3.68
1.61–3.70
—
3.1
2.4
0.8
0.6
—
2030
2117
928
1115
—
8000a
6440
2380
1755
—
182
71
21
12
19
9
5
2.5
—
4.3
3.68
3.70
—
—
a
Stopped recording during oscillations.
Fig. 7 Non-isothermal run commencing at 40 1C with temperature
reduced to 20 1C at the later stage of the run. (a) pH of the overall run;
(b) pH oscillations in more detail at 40 1C; (c) pH oscillations in more
detail at 20 1C; (d) randomly chosen oscillation at 40 and 20 1C
aligned.
Fig. 6 pH oscillations taken at random times during the 10, 20, 30
and 40 1C experiments.
temperature results in an increase in period and amplitude of
oscillations and (b) the time that it takes for the pH to fall
within one oscillation is a function of the temperature. It may
be noted (Table 1, column pH fall) that the time for the pH to
fall within one oscillation approximately halves for every 10 1C
rise in temperature, suggesting that the rate determining step
causing the pH fall has an activation energy of the order of
5 ꢂ 104 kJ kmolꢁ1 based on Arrhenius behaviour.
ꢀc
This journal is the Owner Societies 2009
9048 | Phys. Chem. Chem. Phys., 2009, 11, 9044–9049

View Article Online
Table 2 Partial pressure and solubilities of the gases
Solubility/mol dmꢁ3
10 1C 40 1C
Partial pressure/kPa
10 1C 40 1C
7.3 38.3
Methanol
CO
O2
—
—
Scheme 2 Overall reaction forming product 1.
44.3
11.4
37.7
29.5
7.5
25.2
0.0035
0.0104
—
0.0025
0.0057
—
This would suggest therefore that CO absorption is rate
limiting at 40 1C.
N2
This conclusion is further supported by the observation that
a reduction in agitation speed from 550 rpm to 400 rpm in an
otherwise identical experiment resulted in fewer oscillations.5
Although no experimental rate data are available for
temperatures below 40 1C, the solubility of CO will be
significantly higher while the chemical reaction consuming
the CO would be expected to be slower. This may explain
the similarity in the observations at 10 1C and 20 1C in this
work with the effect of increased agitation speed previously
reported, i.e. existence of several sets of oscillations.
For any given oscillation in an isothermal run, the time for
pH to fall from its maximum to its minimum is approximately
constant and independent of phenylacetylene and the main
product (Z-2-phenyl-but-2-enedioic acid dimethyl ester)
concentration.5 This apparent pseudo-zero order behaviour
may be related in some way to the physical rates of gas liquid
mass transfer (essentially a zero order process) rather than
chemical reaction rates.
The solubility of pure CO in methanol is estimated
(using BatchCADt) to be 0.00805 mol dmꢁ3 (at 10 1C) and
0.00868 mol dmꢁ3 (at 40 1C) whilst the solubility of pure O2 is
0.0921 mol dmꢁ3 (at 10 1C) and 0.0754 mol dmꢁ3 (at 40 1C)
and the vapour pressure of methanol is 7.2 kPa (at 10 1C) and
35.6 kPa (at 40 1C). Assuming ideal gas behaviour the partial
pressures of the gases comprising the mixture in the reactor
head space together with their calculated solubilities are given
in Table 2.
Acknowledgements
The authors wish to acknowledge Mrs Julie Parker for her
assistance in the experimental study.
Therefore, although the solubility of pure CO increases with
temperature its effective solubility falls significantly due to the
exponential increase in the partial pressure of the methanol
vapour. Using previously published data5 the highest rate
of consumption of phenylacetylene during oscillations is
estimated to be 9 ꢂ 10ꢁ5 mol dmꢁ3 minꢁ1 at 40 1C. Under
steady state conditions the rate of absorption of a gas A into a
liquid in which a slow chemical reaction consumes the gas at a
rate RA is given by:
References
1 A. V. Malashkevich, L. G. Bruk and O. N. Temkin, J. Phys. Chem.
A, 1997, 101(51), 9825–9827.
2 S. N. Gorodskii, A. N. Zakharov, A. V. Kulik, L. G. Bruk and
O. N. Temkin, Kinet. Catal., 2001, 42(2), 251–263.
3 S. N. Gorodskii, E. S. Kalenova, L. G. Bruk and O. N. Temkin,
Russ. Chem. Bull., 2003, 52(7), 1524–1543.
4 O. N. Temkin and L. G. Bruk, Kinet. Catal., 2003, 44(5), 601–617.
5 K. Novakovic, C. Grosjean, S. K. Scott, A. Whiting, M. J. Willis
and A. R. Wright, Chem. Phys. Lett., 2007, 435, 142–147.
6 K. Novakovic, C. Grosjean, S. K. Scott, A. Whiting, M. J. Willis
and A. R. Wright, Phys. Chem. Chem. Phys., 2008, 10, 749–753.
7 C. Grosjean, K. Novakovic, S. K. Scott, A. Whiting, M. J. Willis
and A. R. Wright, J. Mol. Catal. A: Chem., 2008, 284, 33–39.
8 A. F. Taylor, Prog. React. Kinet. Mech., 2002, 27, 247–325.
9 http://staff.science.nus.edu.sg/~parwani/c1/node65.html.
10 A. M. Zhabotinsky, Chaos, 1991, 1(4), 379–386.
kl(effective solubilityA ꢁ [A]) = aRA
(1)
where a represents the stoichiometric coefficient and [A] is the
concentration of a gas A. The mass transfer coefficient, kl, for
the reaction vessel and impeller speed used in this work,
has been measured as approximately 0.07 minꢁ1 16
From
.
11 I. Prigogine and R. Lefever, J. Chem. Phys., 1968, 48, 1695–1700.
12 P. Ruoff, J. Interdiscipl. Cycle Res., 1992, 23(2), 92–99.
13 G. Rabai and I. Hanazaki, Chem. Commun., 1999, 1965–1966.
14 R. Nagao, I. R. Epstein, E. R. Gonzales and H. Varela, J. Phys.
Chem. A, 2008, 112, 4617–4624.
15 A. J. Ryan and R. A. L. Jones, Mater. Today (Oxford, UK), 2008,
11(7–8), 20–23.
16 R. Worth, K. Novakovic, M. J. Willis and A. R. Wright, The
Importance of Mass Transfer in Gas–Liquid Reactions, Chemspeed
Technologies, Northampton, UK, 2009, www.chemspeed.com.
eqn (1) and overall reaction as described in Scheme 2, the
concentration of CO may be estimated as:
2 ꢂ 9 ꢂ 10ꢁ5 mol dmꢁ3 minꢁ1
½COꢃ ¼ 0:0025 mol dmꢁ3
ꢁ
0:07 minꢁ1
½COꢃ ¼ ꢁ7:1 ꢂ 10ꢁ5 mol dmꢁ3
ð2Þ
ꢀc
This journal is the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 9044–9049 | 9049