Ce4+-malonic acid reaction in the presence of O2. Reaction channels leading to tartronic and oxalic acid intermediates

Article abstract of DOI:10.1021/jp001005l

The effect of oxygen on the Ce⁴⁺-malonic acid reaction was studied in a semibatch reactor. That effect is important for the Belousov-Zhabotinsky (BZ) chemical oscillator. The Ce⁴⁺ reagent inflow and consequently the rate of the reaction itself was controlled by a peristaltic pump. The reaction products were analyzed by HPLC. With this technique two major oxidation pathways were identified. One is significant at high Ce⁴⁺ inflow rates only; the product of this channel is tartronic acid. The other pathway leading to oxalic acid is active at all flow rates but dominant when the feed is slow. A great part of oxalic acid is oxidized further to carbon dioxide and water. A reaction mechanism compatible with these findings is presented. A key step of this mechanism is the fate of the peroxymalonyl radical which is the first intermediate for both channels. It is proposed that at high Ce⁴⁺ concentrations a fast reaction of this intermediate with Ce⁴⁺ leads to tartronic acid. At low Ce⁴⁺ concentrations, however, the peroxymalonyl radical has a longer lifetime to decarboxylate before reacting with a second Ce⁴⁺ and giving oxalic acid this way. Two mechanistic schemes proposed for this low Ce⁴⁺ channel were tested with further HPLC and kinetic experiments. From the high and low Ce4+ channels it is only the low one which plays a significant role in oxygen-perturbed BZ systems. The effect of that channel and its intermediates on BZ oscillators is discussed briefly.

Full text of DOI:10.1021/jp001005l

9914
J. Phys. Chem. A 2000, 104, 9914-9920
Ce4+-Malonic Acid Reaction in the Presence of O2. Reaction Channels Leading to
Tartronic and Oxalic Acid Intermediates
La´szlo´ Hegedu1s, Horst-Dieter Fo1rsterling,*,† Ma´ria Wittmann, and Zolta´n Noszticzius*,‡
Fachbereich Chemie, Philipps UniVersita¨t Marburg, D-35032 Marburg/Lahn, Germany, and
Center for Complex and Nonlinear Systems and the Department of Chemical Physics,
Budapest UniVersity of Technology and Economics, H-1521 Budapest, Hungary
ReceiVed: March 16, 2000; In Final Form: August 14, 2000
The effect of oxygen on the Ce4+-malonic acid reaction was studied in a semibatch reactor. That effect is
important for the Belousov-Zhabotinsky (BZ) chemical oscillator. The Ce4+ reagent inflow and consequently
the rate of the reaction itself was controlled by a peristaltic pump. The reaction products were analyzed by
HPLC. With this technique two major oxidation pathways were identified. One is significant at high Ce4+
inflow rates only; the product of this channel is tartronic acid. The other pathway leading to oxalic acid is
active at all flow rates but dominant when the feed is slow. A great part of oxalic acid is oxidized further to
carbon dioxide and water. A reaction mechanism compatible with these findings is presented. A key step of
this mechanism is the fate of the peroxymalonyl radical which is the first intermediate for both channels. It
is proposed that at high Ce4+ concentrations a fast reaction of this intermediate with Ce4+ leads to tartronic
acid. At low Ce4+ concentrations, however, the peroxymalonyl radical has a longer lifetime to decarboxylate
before reacting with a second Ce4+ and giving oxalic acid this way. Two mechanistic schemes proposed for
this low Ce4+ channel were tested with further HPLC and kinetic experiments. From the high and low Ce4+
channels it is only the low one which plays a significant role in oxygen-perturbed BZ systems. The effect of
that channel and its intermediates on BZ oscillators is discussed briefly.
Introduction
products were tartronic acid (TA) and mesoxalic acid
(MOA):11
The Ce4+-malonic acid (MA) reaction is an important part
of the Belousov-Zhabotinsky (BZ) reaction which is a classical
system used to study oscillations and chemical waves.1,2 It is
well-known that oxygen can have a strong effect on both
oscillations3-7 and traveling waves8-10 in various BZ systems.
Despite these studies the chemical background of the oxygen
effect is still poorly understood. Only the first step of the
mechanism, the peroxymalonyl radical formation, was confirmed
directly by Neumann and co-workers.11,12 The aim of this work
is to learn more about that mechanism by identifying the
molecular intermediates of the aerobic Ce4+-malonic acid
reaction.
2MAOO‚ f TA + MOA + O2
(N2)
Later on, based on stoichiometric considerations mesoxalic acid
was changed to glyoxylic acid (GOA):12
2MAOO‚ f TA + GOA + CO2 + O2
(N3)
Although their chemical model worked rather well in simulating
the stoichiometry of the Ce4+ consumption, they had no direct
experimental evidence about the appearance of these organic
acids. In this study our aim was to get more information about
these processes by identifying the products using HPLC.
It is known that in nitrogen atmosphere the primary malonyl
radicals (MA‚) react with each other to form 1,1,2,2-ethane
tetracarboxylic acid (ETA)13 and monomalonyl malonate
(MAMA).14 Under aerobic conditions malonyl radicals can react
not only with each other but also with oxygen. Both reactions
are fast but at high enough O2 concentrations the recombination
of the malonyl radicals becomes insignificant compared to their
reaction with O2. The first intermediate of the latter reaction,
as Neumann et al. proved by ESR measurements,11 is a
peroxymalonyl radical (MAOO‚):
Experimental Section
Materials. Commercial Products. Malonic acid (Fluka,
puriss.), tartronic acid (Heraeus), Ce(SO4)2‚4H2O (Merck, pro
analysi), oxalic acid (Merck, pro analysi), mesoxalic acid
(Sigma, pro analysi), PbO2 (pure), 0.025 M ferroin solution
(Fluka, pro analysi), Ce2(SO4)3 (Fluka, purum), 2-thiobarbituric
acid (Fluka, purum), ortho-phosphoric acid 85% (Merck, pro
analysi), NH4OH 25% (J. T. Baker), and H2SO4 (J. T. Baker)
were used as received. All solutions were prepared with doubly
distilled water.
MA‚ + O2 f MAOO‚
(N1)
Ferriin solution was freshly prepared from ferroin solution
as follows: 7.5 mL of 1.66 M H2SO4 solution and 0.035 g of
PbO2 powder were added into 5 mL of ferroin solution then
stirred for 1 h in an ice bath. The freshly prepared 0.01 M ferriin
(in 1 M H2SO4) solution was used immediately.
They also made suggestions on some disproportionation prod-
ucts of the peroxymalonyl radicals. First they proposed that the
* Authors to whom correspondence should be addressed.
† E-mail: foersthd@mailer.uni-marburg.de.
Ce4+-Malonic Acid Reaction in a Semibatch Reactor. A
30 mL three-necked vessel was used as a semibatch reactor.
‡ Fax: 36+1+4631896. E-mail: noszti@phy.bme.hu.
10.1021/jp001005l CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/14/2000

Ce4+-Malonic Acid Reaction in the Presence of O2
J. Phys. Chem. A, Vol. 104, No. 44, 2000 9915
One neck was used for the gas inlet, another for the reactant
inlet, and the third one for the gas outlet. A continuous inflow
of the reactant was established by a peristaltic pump. Backward
diffusion of air was prevented by bubbling the gas through water
after leaving the reactor.
the BZ reaction. No decomposition of these acids was observed
under these conditions. For further details see refs 13, 14, 17,
and 18.
Oxalic Acid Test. The 2 mL sample solution to be tested
was mixed first with 1 mL of 0.02 M Ce2(SO4)3 then with 1
mL of 25% NH4OH solution. The white precipitate (mixture
of Ce(III) hydroxide and Ce(III) oxalate) was centrifuged,
isolated, and washed with 0.005 M Ce2(SO4)3 solution twice.
The precipitate was dried at 100 °C. To liberate oxalic acid we
added 0.1 mL of 1 M H3PO4 solution and the mixture was boiled
over a free flame for half a minute. After cooling we added 0.2
mL of 25% NH4OH and evaporated the solution. Under these
conditions, oxalic acid forms ammonium oxalate which on dry
heating yields water and oxamide. In the last step approximately
0.1 g of 2-thiobarbituric acid was added to the sample and the
mixture was heated to 140-150 °C. If oxamide was present, a
brick-red condensation product resulted. This spot test is specific
for oxalic acid without any interference.19
The reaction of malonic acid with Ce4+ was examined in 1
M sulfuric acid medium at room temperature (22 ( 2 °C). We
used two solutions containing malonic acid and Ce4+, respec-
tively. Both reagents were dissolved in 1 M sulfuric acid. In all
experiments 5 mL of 0.1 M Ce4+ solution was added continu-
ously into 5 mL of 1 M malonic acid solution with a peristaltic
pump applying various pumping rates. The feed times were 1,
8, 30, 60, and 90 min, respectively. As malonic acid was in
great excess a steady-state Ce4+ concentration was reached
rapidly. In this state the rate of the reaction was determined
exclusively by the Ce4+ inflow. Consequently different steady-
state Ce4+ and malonyl radical concentrations could be estab-
lished by varying the Ce4+ input rate.
The reaction mixture and the reagent was bubbled continu-
ously with N2 or O2 according to the anaerobic or aerobic
conditions.
Ce4+-Malonic Acid Reaction with Added Oxalic or
Mesoxalic Acid. The reactions with added organic substrates
were performed as those with malonic acid except the malonic
acid solution contained oxalic or mesoxalic acid as well.
Results and Discussion
Analytical Results for the Ce4+-Malonic Acid Reaction.
Semibatch Experiments in N2 Atmosphere. First, as a control,
the anaerobic reaction of Ce4+ with malonic acid13,14 was
repeated (Figure 1a) in the semibatch reactor. Independently of
the input rate only the recombination products of the malonyl
radicals, ETA and MAMA, appeared and their ratio was also
constant within the experimental error. The first step in the
anaerobic (and also in the aerobic) Ce4+-malonic acid reaction
is the formation of malonyl radicals. Malonyl radicals exist in
two forms: alkyl and carboxylato type radicals. These two types
are assumed to be in a fast equilibrium:
Rate Constant Measurements. The experiments were carried
out in 1 M H2SO4 medium. Two solutions were used: one
contained the organic acid and the other the Ce4+ reagent. Ce4+
reagent (1 mL) was injected into 140 mL substrate solution.
The temperature was controlled at 20 ( 0.05 °C in a thermo-
stated vessel with an optical path length of 10.8 cm. The reaction
was followed with a sensitive spectrophotometer15 measuring
the Ce4+ absorbance at 400 nm. The substrate solutions were
bubbled for half an hour with N2 or O2 before mixing the
reagents and the bubbling was continued during the reaction.
The applied concentrations (after mixing): 5 × 10-2 M malonic
acid + 10-4 M Ce4+; 10-3 M oxalic acid + 10-4 M Ce4+; 5 ×
10-5 M mesoxalic acid + 5 × 10-6 M Ce4+
.
Ferriin-Malonic Acid Reaction in N2 and O2 Atmosphere.
The ferriin-MA reactions were performed in a similar way as
with Ce4+. The only difference was that the reaction was carried
out as a batch experiment. A 5 mL volume of 1 M MA in 1 M
H2SO4 was added to 5 mL of 0.01 M ferriin solution while
bubbled with O2 or N2 gas stream continuously. The samples
were analyzed with HPLC after 20 min reaction time.
ETA is formed when two alkyl malonyl radicals recombine,
while MAMA is formed from one alkyl and one carboxylato
type radical. As in our experiments the ratio of the recombination
products ETA and MAMA was always constant, this supports
the assumption of the fast equilibrium between the two radical
species.
We remark that in Figure 1a most of the Ce4+ (about 84%)
produced ETA and MAMA but this is not the only sink of Ce4+
.
Analytical Methods
The remaining 16% is used in another reaction route20 leading
to formic acid and CO2 via carbene intermediates.
HPLC. HPLC experiments were performed with a Shimadzu
equipment (LC-10AS pump, CTO-10A column oven, SPD-10A
dual wavelength UV-detector, 8 nm bandwidth, 1 cm path
length, 8 µL sample volume). The sample was injected using a
Rheodyne 7010 injector with a 20 µL sample loop. The ion-
exchange column16 was working at 45 °C and a UV detector at
220 nm. All organic acids can be detected at this wavelength.
The eluent was 0.01 M H2SO4, flow rate 0.40 mL/min. As the
sulfuric acid concentration in the reaction medium (1 M) was
100 times more concentrated than in the eluent (0.01 M),
samples were diluted 100 times prior to the HPLC analysis.
The organic acid concentrations given in captions of Figures 1,
4, and 6 are for the original reaction mixture (that is, before the
100 times dilution for HPLC). The same HPLC conditions have
already been applied successfully in a series of previous
investigations to separate various organic acid intermediates of
Semibatch Experiments in O2 Atmosphere. Next the Ce4+
-
malonic acid reaction was investigated in the presence of
oxygen. The reaction was performed at various steady-state Ce4+
concentrations in the semibatch reactor. Figure 1b shows the
chromatogram of the reaction products in the case of relatively
high Ce4+ concentration (feed rate: 5 mL/1 min). Here, two
changes can be observed compared to the oxygen-free experi-
ment: (i) an additional peak appeared at tret ) 804 s and (ii)
the ratio of the peaks at 570 and 602 s (“MAMA and ETA
peaks”) has been changed. The peak at 804 s retention time
could be easily identified as tartronic acid as there is no possible
interference at that retention time. At the retention times of
mesoxalic and glyoxylic acids, the products of (N2) and (N3)
(638 and 968 s, respectively) no peak appeared. On the other
hand, as the ratio of the recombination products ETA and

9916 J. Phys. Chem. A, Vol. 104, No. 44, 2000
Hegedu¨s et al.
new product increasing the “ETA” peak at 602 s. Neither this
product nor the tartronic acid is present in the anaerobic
experiments showing that new reaction channels were opened
which can start with the reaction of malonyl radicals with O2.
To examine the products of these channels the reaction
conditions should be chosen such that the O2-malonyl radical
reaction be dominant. If there is enough oxygen in the reaction
mixture this prerequisite is not a problem as malonyl radicals
form peroxymalonyl radicals with free oxygen rapidly.21 At
higher Ce4+-MA reaction rates, however, the rapid oxygen
consumption can drive the steady-state oxygen concentration
to a very low level and such conditions, together with the
increased malonyl radical concentration, favor the unwanted
radical-radical recombination reactions. This unwanted reaction
channel can be suppressed by decreasing the reaction rate, i.e.,
by adding the Ce4+ solution more slowly.
Figures 1c and 1d show the products of the Ce4+-malonic
acid reaction at lower reaction rates. In Figure 1c the flow rate
was 5 mL/8 min, and in Figure 1d it was 5 mL/30 min. Further
increase of the feed time had no effect, experiments with 30,
60, and 90 min feed times gave the same results. In Figure 1c
all the peaks appear that were present in Figure 1b, but their
heights have drastically changed. In Figure 1d the peak at 804
s disappeared, that is, no tartronic acid had been formed here.
Moreover, the peak of MAMA also disappeared, but the peak
at 602 s remained there. This peak cannot be ETA, however,
as the other recombination product is absent. The most likely
candidate is oxalic acid (OA) as its retention time is the same
as that of ETA. To check this hypothesis we performed
additional tests.
Qualitative Tests for Oxalic Acid. We found that while ETA
decomposes readily on heating, oxalic acid is rather stable
against it. Thus we kept the sample at 90 °C for 20 min and
then analyzed with HPLC again. We have performed this test
for all samples of anaerobic reactions and also of the aerobic
ones. Samples produced in N2 atmosphere, containing MAMA
and “real” ETA peaks, behaved as expected. After heating, the
real ETA peak nearly disappeared and peaks of its decomposi-
tion products (1,1,2-ethanetricarboxylic acid (ETRA) at 840 s
and succinic acid (SA) at 1240 s) appeared. The MAMA peak
also disappeared as at this temperature malonyl malonate
hydrolyses completely to tartronic and malonic acids. On the
other hand, in the aerobic case in samples produced with low
Ce4+ inflow (like Figure 1d) the peak at 602 s did not change
at all. This clearly shows that the product formed in the presence
of oxygen cannot be ETA but some more stable compound.
To get final evidence that this stable compound is oxalic acid
we have performed a color test which is specific for oxalic acid
with no interference (details are given in the Experimental
Section). This test was positive for the aerobic reactions and
negative for the anaerobic one. The conclusion is that at low
reaction rates the product of the Ce4+-malonic acid reaction
in the presence of oxygen is mainly oxalic acid.
Figure 1. HPLC chromatograms of the products of the Ce4+-MA
reaction in N2 or O2 atmospheres. A volume of 5 mL 0.1 M Ce4+ solu-
tion was added to 5 mL of 1 M MA solution with various feed rates.
(a) N2 atmosphere, feed rate: 5 mL of Ce4+ solution/1 min. The products
are ETA (1.47 × 10-2 M) and MAMA (6.30 × 10-3 M). (b) O2
atmosphere with high feed rate: 5 mL/1 min. The “ETA” and “MAMA”
peaks can still be detected but they are smaller and their ratio has been
changed. A new peak has also appeared at tret ) 804 s which corres-
ponds to tartronic acid (TA) (3.42 × 10-3 M). (c) O2 atmosphere with
an intermediate feed rate: 5 mL/8 min. Observe the decrease in the
tartronic acid (9.64 × 10-4 M) and MAMA peaks. (d) O2 atmosphere
with low feed rate: 5 mL/30 min. MAMA and tartronic acid peaks
have disappeared. The remaining peak at tret ) 602 s (at the former
place of ETA) is identified as oxalic acid (OA) (6.20 × 10-3 M).
Mechanistic Proposals. Before presenting the actual mecha-
nistic schemes it is useful to summarize the experimental
observations briefly. As Figure 1 shows, the product distribution
depends not only on the atmosphere but in the case of O2
atmosphere it also depends on the rate of the reaction. At high
reaction rates (e.g., in the case of batch experiments), a part of
the malonyl radicals might remove most of the oxygen creating
anaerobic conditions for the remaining ones. In this case the
already known anaerobic recombination channel leading to ETA
and MAMA can be also active. In the presence of oxygen,
however, two new reaction channels appear. One is active at
MAMA was always constant in all anaerobic measurements,
the peak ratio change is probably due to the appearance of some

Ce4+-Malonic Acid Reaction in the Presence of O2
J. Phys. Chem. A, Vol. 104, No. 44, 2000 9917
Figure 2. Mechanism of tartronic acid production from peroxymalonyl
radicals at high Ce4+ concentrations.
high Ce4+ inflow rates and leads to tartronic acid. The other is
active at all flow rates but it becomes dominant at lower steady-
state Ce4+ concentrations and gives oxalic acid. Below, we
present mechanistic proposals for these two new channels.
The High Ce4+ Reaction Channel. This aerobic channel leads
to tartronic acid which should be formed in a reaction between
a peroxymalonyl radical and a Ce4+. The proposed mechanism
is shown in Figure 2.
Actually, the scheme assumes two consecutive reactions with
Ce4+: the first one (not shown in the figure) generates the
peroxymalonyl radical, then the second one gives the tartronic
acid. This pathway is preferred by high enough Ce4+ concentra-
tions when other reactions of the peroxymalonyl radical (like
decarboxylation or chain transfer to malonic acid) are partially
suppressed by the Ce4+-peroxymalonyl radical reaction. These
suppressed reactions, however, are always present but they
dominate only at low Ce4+ concentrations.
The Low Ce4+ Reaction Channel. Routes (a) and (b).
Decreasing the Ce4+ input rate lowers the steady-state Ce4+
concentration in the semibatch reactor and the probability of
the Ce4+-peroxymalonyl radical reaction decreases accordingly.
Under these conditions the role of the slower, Ce4+ independent
decomposition routes of the peroxymalonyl radical becomes
more pronounced. As our experiments show, the low Ce4+
channel should lead to oxalic acid production. Two possible
routes lead to oxalic acid: routes (a) and (b) (see Figures 3a
and 3b).
In route (a) decarboxylation of the peroxymalonyl radical
takes place before the reaction with a second Ce4+. In the last
step a H+ ion promoted rearrangement of the peroxide bond
gives oxalic acid. Route (b) starts with a radical transfer
reaction: the attack of a peroxymalonyl radical results in a
malonyl radical and a hydroperoxy malonic acid molecule. As
the malonyl radical forms a new peroxymalonyl radical with
oxygen the process is a chain reaction producing hydroperoxy
malonic acid from oxygen and malonic acid. After a H+ ion
catalyzed rearrangement, hydroperoxy malonic acid is trans-
formed to mesoxalic acid which is oxidized by two Ce4+ to
oxalic acid.
We were not able to find mesoxalic acid in our HPLC
experiments. However, this does not exclude the possibility of
route (b) as the Ce4+-MOA reaction is probably very fast. To
check this and to decide whether route (a) or (b) is the dominant
one further experiments were necessary.
Figure 3. Two possible versions of the oxalic acid production from a
peroxymalonyl radical at low Ce4+ concentrations: (a) decarboxylation
route, (b) route starting with a radical transfer reaction producing
mesoxalic acid intermediate in a chain reaction.
Interpretation of the experiments in N2 atmosphere (Figure
4a-c) is rather straightforward: the more reactive substrates
(oxalic and mesoxalic acids) reduce a close to stoichiometric
part of Ce4+ and only the excess Ce4+ reacts with malonic acid.
The extent of the Ce4+-malonic acid reaction can be estimated
from the double peak of MAMA and ETA.
In Figure 4a, the initial Ce4+ concentration was 5 × 10-2
available for malonic acid should be reduced to 3 × 10-2
M
after mixing the reagents. In Figure 4b,c the Ce4+ concentration
M
and 10-2 M, respectively, assuming a stoichiometric reaction
with 10-2 M oxalic acid (Figure 4b) or 10-2 M mesoxalic acid
(Figure 4c). Thus the ratio of the double peaks should be 5:3:1
if in the oxalic acid-malonic acid and the mesoxalic acid-
malonic acid mixtures the more reactive substrates are oxidized
quantitatively to CO2 and water and only the remaining Ce4+
is used to oxidize the malonic acid. Really the experimental
ratio is very close to this theoretical prediction. A small deviation
from that can be detected in Figure 4c only, where some
remaining oxalic acid traces increase the ETA peak slightly.
Figure 4d-f shows analogous experiments such as in Figure
4a-c, but here in O2 atmosphere instead of N2. Now the results
are completely different. The main product of the reaction is
always oxalic acid. An initial addition of oxalic or mesoxalic
acid to the reaction mixture increases the final oxalic acid
concentration. Thus in the presence of oxygen the Ce4+ does
not consume all the initial oxalic acid indicating that O2
accelerates the overall Ce4+-malonic acid reaction which
produces oxalic acid. (Another possibility would be that O2
could slow somehow the Ce4+-oxalic acid reaction but ac-
cording to our experiments O2 has no effect on the rate of this
reaction.) Both routes (a) and (b) would be able to accelerate
the reaction but obviously route (b) should have a stronger effect.
Further Experiments Checking Routes (a) and (b). Semi-
batch Experiments with Mixed Substrates. First to test the role
of oxalic and mesoxalic acids in the mechanism of the Ce4+
-
malonic acid reaction we added these assumed intermediates
to the initial reagent mixture both in N2 and in O2 atmosphere
experiments. The products were analyzed by HPLC and the
results are displayed in Figure 4.

9918 J. Phys. Chem. A, Vol. 104, No. 44, 2000
Hegedu¨s et al.
Figure 4. HPLC of the products of the Ce4-malonic acid reaction (a-c) in N2 and (d-f) in O2 atmosphere. Other experimental conditions were
like those in Figure 1d, except that beside malonic acid an initial oxalic acid concentration of 2 × 10-2 M was established in (b) and (e) or an initial
mesoxalic acid concentration of 2 × 10-2 M was applied in (c) and (f).
TABLE 1: Second-Order Rate Constant of the Reactions of
Different Organic Acids with Ce4+ in 1 M Sulfuric Acid
Solution at 20 °C
Experiments made with malonic acid are depicted in Figure
5. It can be seen that the logarithmic plot is not a straight line
in either O2 or N2. In the case of N2 this is due to the not
perfectly anaerobic conditions at the beginning. This increases
the reaction rate until the radicals remove the last traces of
oxygen. In O2 atmosphere the slope of the line is increasing
slightly because oxidation of oxalic acid (produced from malonic
acid) increases the reaction rate.
rate constant k (M-1 s-1
)
organic substrate
N2 atmosphere
O2 atmosphere
malonic acid
oxalic acid
mesoxalic acid
0.24
53
0.46
53
∼104
∼104
We can conclude that the slope of the logarithmic plot in O2
is about two times higher than in N2 atmosphere, i.e., the
reaction proceeds approximately two times faster in O2 than in
N2 atmosphere.22 This result suggests that route (a) should be
the dominant one and route (b) cannot be significant in the low
Ce4+ reaction channel. We present the following arguments to
support the above conclusion. The rate-determining step in the
Ce4+-malonic acid reaction is the first one which produces a
malonyl radical. In N2 atmosphere these radicals mainly
recombine to form ETA and MAMA. This way these radicals
do not consume additional Ce4+, while in O2 atmosphere they
are transformed to peroxymalonyl radicals which in route (a)
consume a second Ce4+ after decarboxylation. Thus route (a)
can explain the observed 2-fold acceleration22 in O2 atmosphere.
Moreover the experiments also show clearly that even if
mesoxalic acid were an intermediate it could not be detected in
the reaction mixture due to its very fast reaction with Ce4+
.
Thus while our experiments with mixed substrates gave some
information on the behavior of oxalic and mesoxalic acids in
the reaction system they did not clarify whether route (a) or (b)
is the dominant one. To answer that question we performed
kinetic experiments to determine the rate of the relevant Ce4+
organic acid reactions quantitatively.
-
Rate Constant Measurements. The second-order rate constant
of the Ce4+-organic acid substrate reaction was determined for
three substrates: malonic, oxalic, and mesoxalic acids both in
N2 and O2 atmosphere. The results are shown in Table 1. In all
experiments a great excess of the organic substrate was applied
and the pseudo first-order rate constant of the Ce4+ consumption
was measured. For further details see the Experimental Section.
On the other hand, route (b) predicts a more dramatic
acceleration in O2 atmosphere. For example, if each malonyl
radical would participate in only one chain transfer reaction

Ce4+-Malonic Acid Reaction in the Presence of O2
J. Phys. Chem. A, Vol. 104, No. 44, 2000 9919
Figure 5. Spectrophotometric traces of the Ce4+-malonic acid reaction
in oxygen and nitrogen atmosphere. “A” is the Ce4+ absorbance
measured at 400 nm. Initial conditions: [MA] ) 0.05 M, [Ce4+] )
10-4 M, [H2SO4] ) 1 M. (a) Linear, (b) logarithmic plots.
Figure 6. HPLC of the products of the ferriin-malonic acid reaction
(a) in N2 and (b) in O2 atmosphere. Initial conditions: [MA] ) 0.5 M,
[ferriin] ) 5 × 10-3 M, [H2SO4] ) 1 M. The products of the
reactions: (a) OA (3.10 × 10-4 M), MOA (4.46 × 10-4 M), TA (3.05
× 10-4 M), GOA (2.80 × 10-3 M); (b) OA (3.75 × 10-4 M), MOA
(4.05 × 10-3 M), TA (6.22 × 10-4 M), GOA (4.64 × 10-3 M).
before termination then the emerging mesoxalic acid would
consume two additional Ce4+ to form oxalic acid. Thus a 3-fold
acceleration would be expected and in the case of longer chains
this acceleration would be proportionally higher. The experi-
mentally found 2-fold acceleration suggests that route (b) cannot
play an important role here.
and low Ce4+ concentrations: the borderline should be some-
where between these concentrations that is around 10-3 M.
In BZ oscillators the Ce4+ concentration is usually below 10-3
M most of the time, thus the low Ce4+ channel is the only
important one in these systems. Furthermore tartronic acid, the
product of the high Ce4+ channel is much less reactive than
oxalic acid. Regarding these facts we conclude that it is the
low Ce4+ channel which is responsible for the oxygen effects
in the BZ oscillators.
Oxygen and BZ Oscillators. The Role of “High” and “Low”
Ce4+ Reaction Channels in Oxygen-Perturbed BZ Systems. BZ
experiments are usually performed in batch or continuously fed
stirred tank reactors (CSTRs) where Ce4+ is generated in the
course of the reaction. How can we decide in this case whether
the “high” or the “low” Ce4+ channel is the dominant one? To
answer this question we first have to estimate the steady-state
Ce4+ concentrations in the semibatch experiments favoring the
high and low channels. The steady-state Ce4+ concentration in
The Ferriin-Malonic Acid Reaction in N2 and O2 Atmo-
sphere. As we could see in the low Ce4+ channel, route (b) can
play only a minor role. A direct evidence for the route (b)
mechanism would be the identification of its characteristic
the semibatch reactor [Ce4+
]
ss
can be calculated from the
following component balance:
intermediate mesoxalic acid. However, the extremely fast Ce4+
-
mesoxalic acid reaction prevents the identification of mesoxalic
acid traces in the reaction mixture. On the other hand we could
hope to find mesoxalic acid if ferriin is applied in the
experiments instead of Ce4+ because the oxidation of organic
substrates by ferriin is usually slow. Thus if the primary malonyl
radicals are generated with ferriin we can expect similar initial
reactions but further oxidation of the intermediates such as
mesoxalic acid can be avoided this way. Figure 6 shows HPLC
analysis of the products of the ferriin-malonic acid reaction in
N2 and in O2 atmosphere.
wF
[Ce4+]F ) [Ce4+]ss ‚ k[MA]
V
where wF is the volumetric flow rate of the feed, [Ce4+]F is the
Ce4+ concentration in the feed, V is the actual volume of the
liquid in the reactor, k is the second-order rate constant of the
Ce4+-MA reaction in the presence of O2, and [MA] is the actual
malonic acid concentration in the reactor:
V0
[MA] ) [MA]0 ‚
In N2 atmosphere the reaction is very slow; only traces of
various acids were found after 20 min reaction time. In O2
atmosphere, however, the reaction proceeds faster and the
mesoxalic acid (MOA) peak is 1 order of magnitude higher than
in N2. The peak of tartronic acid (TA) and glyoxylic acid (GOA)
also grows somewhat (about a factor of 2) in O2 atmosphere.
Glyoxylic acid indicates that probably route (a) also plays a
role here. (Glyoxylic acid can be produced when a decarboxy-
lated peroxymalonyl radical reacts in a chain transfer reaction
with malonic acid. The resulting peroxide type molecule can
give glyoxylic acid after a H+ ion-assisted rearrangement of
the chemical bonds.)
V
where [MA]0 is initial malonic acid concentration, and V0 is
initial volume of liquid in the reactor. (As the malonic acid is
in a high excess its consumption can be neglected.) Thus,
wF [Ce4+]F
[Ce4+]ss )
‚
V0
k[MA]0
With our experimental data the steady-state Ce4+ concentrations
calculated for Figures 1b and 1c are 3.2 × 10-3 M and 4 ×
10-4 M, respectively. This answers our question about the high

9920 J. Phys. Chem. A, Vol. 104, No. 44, 2000
Hegedu¨s et al.
and the ESF Program: “Reactor”. M.W. acknowledges the
support of the Bolyai Research Stipendium. The authors thank
V. Farkas for help in the experiments, E. Ko´kai, G. Taba, and
M. Koza´k for developing the details of the oxalic acid test, J.
Nyitrai for discussions, and especially a Referee who made
helpful mechanistic suggestions.
References and Notes
(1) Oscillations and TraVeling WaVes in Chemical Systems; Field, R.
J., Burger, M., Eds.; Wiley: New York, 1985.
(2) Chemical WaVes and Patterns; Kapral, R., Showalter, K., Eds.;
Kluwer: Dordrecht, The Netherlands, 1995.
(3) Field, R. J. In ref 1., p 65.
(4) Ruoff, P.; Noyes, R. M. J. Phys. Chem. 1989, 93, 7394.
(5) Chou, Y.-C.; Lin, H.-P.; Sun, S. S.; Jwo, J.-J. J. Phys. Chem. 1993,
97, 8450.
(6) Wang, J.; Hynne, F.; Sorensen, P. G.; Nielsen, K. J. Phys. Chem.
1996, 100, 17593.
(7) Treindl, L.; Ruoff, P.; Kvernberg, P. O. J. Phys. Chem. A 1997,
101, 4606.
(8) Zhabotinsky, A. M.; Mu¨ller, S. C.; Hess, B. Physica D 1991, 49,
47.
(9) Zhabotinsky, A. M.; Eager, M. D.; Epstein, I. R. Phys. ReV. Lett.
1993, 71, 1526.
(10) Ungva´rai, J.; Nagy-Ungva´rai, Zs. ACH - Models in Chemistry
1998, 135, 393.
(11) Neumann, B.; Mu¨ller, S. C. J. Am. Chem. Soc. 1995, 117, 6372.
(12) Neumann, B.; Steinbock, O.; Mu¨ller, S. C.; Dalal, N. S. J. Phys.
Chem. A 1997, 101, 2743.
(13) Gao, Y.; Fo¨rsterling, H.-D.; Noszticzius, Z.; Meyer, B. J. Phys.
Chem. 1994, 98, 8377.
Figure 7. Skeleton mechanism of the oxygen-mediated Ce4+ oxidation
of malonic acid. The “high” Ce4+ channel leading to tartronic acid is
on the right-hand side, while the other pathways leading to oxalic acid
and CO2 represent the “low” Ce4+ channel with two parallel routess
routes (a) and (b).
(14) Sirimungkala, A.; Fo¨rsterling, H.-D.; Noszticzius, Z. J. Phys. Chem.
1996, 100, 3051.
(15) Fo¨rsterling, H.-D.; Pachl, R.; Schreiber, H. Z. Naturforsch. 1987,
42a, 963.
(16) The column used for the separation of organic acids was a Merck,
Polyspher OA KC cation exchanger column with a length of 30 cm and
diameter of 7.8 mm which was proposed by Merck for our special
requirements. The column could resolve all the organic acids relevant for
the present research except for the oxalic acid-ethane tetracarboxylic acid
pair.
To present detailed mechanistic proposals for the ferriin-
malonic acid reaction would be too early at this stage, however.
Our aim here was only to show that the products and
consequently also the mechanism can be different for Ce4+ and
ferriin oxidants.
Summary and Conclusions for O2-Perturbed BZ
Oscillators.
(17) Oslonovitch, J.; Fo¨rsterling, H.-D.; Wittmann, M.; Noszticzius, Z.
J. Phys. Chem. A 1998, 102, 922.
(18) Szalai, I.; Fo¨rsterling, H.-D.; Noszticzius, Z. J. Phys. Chem. A 1998,
102, 3118.
(19) Feigl, F.; Anger, V. Spot tests in organic analysis; Elsevier:
Amsterdam, 1966; p 457.
(20) Nagygyo˜ry, Sz.; Wittmann, M.; Pinte´r, Sz.; Visegra´dy, A.; Dancso´,
A.; Thuy, N. B.; Noszticzius, Z.; Hegeduˆs, L.; Fo¨rsterling, H.-D. J. Phys.
Chem. A 1999, 103, 4885.
(21) Schuchmann, M. N.; Rao, R.; Hauser, M.; Mu¨ller, S. C.; Sonntag,
C. J. Chem. Soc., Perkin Trans. 2000, 2, 941.
As a summary Figure 7 presents an integrated picture showing
all possible routes of the reaction mechanism. For oxygen-
perturbed Ce4+ oscillators the main stream is the vertical one
flowing through the oxalic acid intermediate via route (a). In
these oscillatorss when the autocatalytic reaction is “switched
off”sthe main stream produces oxalic acid affecting strongly
the reaction dynamics. That effect is probably due to the oxalic
acid-hypobromous acid reaction23 generating some inhibitory
bromide. This could also explain the somewhat surprising
observation that oxygen can keep a BZ oscillator in its reduced
steady state.4 In oxygen-perturbed ferroin catalyzed systems
mesoxalic acid can play a similar role. It was observed that in
these systems mesoxalic acid has a strong effect on oscillations
and chemical waves.24
(22) Note: In our explanation we neglect further reaction of Ce4+ with
the oxalic acid product. As the initial concentration of Ce4+ is 10-4 M the
maximum oxalic acid concentration would be 5 × 10-5 M. Regarding the
rate constant in Table 1 and the malonic acid concentration which is 0.05
M in this case, the rate of the Ce4+-oxalic acid reaction can be maximum
30% of the rate of the Ce4+-malonic acid reaction. Thus in O2 atmosphere
roughly 20% of the Ce4+ consumption is due to another reaction. But the
same is true in N2 atmosphere: about 16% of the Ce4+ is consumed in
nonrecombinative but CO2-producing reactions.20 Thus a theoretical 1:2
ratio for the rates in N2 and O2 atmosphere is still an acceptable estimation.
(23) Knoller, Y.; Perlmutter-Hayman, B. J. Am. Chem. Soc. 1955, 77,
3212.
Acknowledgment. This work was partially supported by
OTKA (T-030110), FKFP (0287/1997) grants, by the Deutsche
Forschungsgemeinschaft, the Fonds der Chemischen Industrie
(24) Kru¨ger, F.; Nagy-Ungva´rai, Zs.; Mu¨ller, S. C. Physica D 1995,
84, 95.