The importance of mesomerism in the termination of α-carboxymethyl radicals from aqueous malonic and acetic acids

Article abstract of DOI:10.1002/1521-3765(20010216)7:4<791::AID-CHEM791>3.0.CO;2-2

In the dioxygen-free aqueous malonic and acetic acid systems, α-carboxymethyl radicals are produced through hydrogen abstraction from the parent compound by radiolytically generated OH radicals. H abstraction from the CO₂^-/ CO₂H group followed by decarboxylation is a process of small importance (≤ 5 %). The α-carboxymethyl radicals terminate by recombination. Two types of recombination product are observed which are characterised by the formation of a C-C linkage or a C-O linkage. α-Carboxymethyl radicals are mesomeric systems. Their mesomeric state depends on the state of protonation and determines the proportion of the C-C-versus C-O-linked dehydrodimers they produce.

Full text of DOI:10.1002/1521-3765(20010216)7:4<791::AID-CHEM791>3.0.CO;2-2

FULL PAPER
The Importance of Mesomerism in the Termination of a-Carboxymethyl
Radicals from Aqueous Malonic and Acetic Acids
Wen-Feng Wang,[a, b] Man Nien Schuchmann,[a] Heinz-Peter Schuchmann,[a] and
Clemens von Sonntag*[a]
Abstract: In the dioxygen-free aqueous malonic and acetic acid systems, a-
carboxymethyl radicals are produced through hydrogen abstraction from the parent
compound by radiolytically generated OH radicals. H abstraction from the CO2 /
CO2H group followed by decarboxylation is a process of small importance (ꢀ5%).
The a-carboxymethyl radicals terminate by recombination. Two types of recombi-
nation product are observed which are characterised by the formation of a C C
linkage or a C O linkage. a-Carboxymethyl radicals are mesomeric systems. Their
mesomeric state depends on the state of protonation and determines the proportion
of the C C- versus C O-linked dehydrodimers they produce.
Keywords: acetic acid ´ carboxylic
acids ´ chemical oscillator ´ dimeri-
zation ´ malonic acid ´ radicals
the behaviour of the dicarboxymethyl radical, including its
termination reactions over a wide pH range. Two recombina-
tion products, ethane-1,1,2,2-tetracarboxylic acid (2,3-dicar-
boxysuccinic acid) and monomalonyl malonate (malonic acid
dicarboxymethyl monoester), have been observed upon
oxidation of malonic acid with CeIV in aqueous solution.[10]
From this it was inferred that two distinct species of malonic
acid-derived free radicals, both of them undergoing recombi-
Introduction
The oxidation of malonic acid to the dicarboxymethyl radical
by the CeIV ion has been recognised as a major elementary
step in the chemical oscillator known as the Belousov ± Zha-
botinsky reaction.[1±5] In general, dicarboxymethyl is produced
from malonic acid by the action of free radicals; in systems
which involve the reduction of a strong oxidant, for example,
bromate, various free radicals that attack malonic acid, such
as Br and BrO2 , play a role and will contribute to the
degradation of malonic acid.[2] These hydrogen-abstraction
processes further complicate the reaction-mechanistic imag-
ing of systems such as the Belousov ± Zhabotinsky reaction. In
addition, it has been recognised[4] that under conditions of
oxygenation, peroxyl radical reactions[7] must be taken into
account. Moreover, peroxyl radicals and molecular peroxidic
species may effect redox cycling and Fenton-like reactions[9] in
the case of transition metal-catalysed systems, though perhaps
not with cerium.
.
nation reactions, exist in these systems, namely, CH(COOH)2
.
.
.
and HO2CCH2 C(O)O in varying states of protonation
depending on the pH, either generated independently[10] or
tautomerically related.
The assumption that the lifetime of a carboxyl radical (here,
.
HO2C CH2 C(O)O ) is long enough for it to undergo
recombination reactions is in contradiction with what is
known of the behaviour of these radicals: the first-order decay
lifetime of aliphatic carboxyl radicals [cf. reaction (7) later]
has been estimated to be in the order of nanoseconds[11] (note
that otherwise the Kolbe electrolysis, which generates these
radicals at high local concentrations, could not be used as a
synthetic method for the preparation of alkanes).
In this work, we demonstrate that the behaviour of
dicarboxymethyl radicals can be much more plausibly de-
scribed by a single free-radical species and by taking into
account its mesomeric states.
In the larger context as to the fate of dicarboxymethyl in
these malonic acid-containing systems, and also for the sake of
systematic free-radical chemistry, it is of interest to explore
[a] Prof. Dr. C. von Sonntag, Prof. Dr. W.-F. Wang, Dr. M. N. Schuchmann,
Dr. H.-P. Schuchmann
Max-Planck-Institut für Strahlenchemie
Stiftstrasse 34 ± 36, P.O. Box 101365
45413 Mülheim an der Ruhr (Germany)
Fax : (‡49)208-306-3951
E-mail: vonsonntag@mpi-muelheim.mpg.de
Experimental Section
[b] Prof. Dr. W.-F. Wang
Permanent address:
Shanghai Institute of Nuclear Research
P.O. Box 800Ð204, Shanghai 201800 (P.R. China)
Acetic acid (Merck, p.a.), malonic acid (Fluka), glycolic acid (EGA
Chemie), acetoxyacetic acid (Aldrich), tricarballylic (3-carboxyglutaric)
acid (Fluka), succinic acid (Merck), tartronic (hydroxymalonic) acid
Chem. Eur. J. 2001, 7, No. 4
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0704-0791 $ 17.50+.50/0
791

C. von Sonntag et al.
FULL PAPER
.
.
7
7
(Serva) and 1,1,2,2,-ethanetetracarboxylic acid tetraethyl ester (Aldrich)
were commercially available. 1,1,2,2-Ethanetetracarboxylic acid was ob-
tained by hydrolysis of the ester.[10] Monomalonyl malonate was synthes-
ised from malonic acid and hydroxymalonic acid by condensation, with
P2O5 as the dehydrating agent.[10] The crude product still contains malonic
and hydroxymalonic acid. The purity was sufficient for assignment
purposes.
are G( OH) ꢁ G(eaq† ˆ 2.9  10 , G(H ) ˆ 0.6  10
and
G(H2O2) ꢁ 0.7 Â 10 7 molJ . N2O was used to convert the
1
.
solvated electron into OH [reaction (2)]. The rate constants
.
eaq‡N2O ‡ H2O
!
OH ‡ OH ‡N2
(2)
of the reaction of the hydroxyl radical with malonic acid
(pKa ˆ 2.8 and 5.7) were re-determined by competition
kinetics with KSCN and by build-up kinetics of the dicarbox-
ylmethyl radical monitored optically at l ˆ 340 nm. We found
Aqueous solutions of the substrates (concentrations typically 1 Â
3
3
10 3 moldm for malonic acid and 3 Â 10 3 moldm for acetic acid) were
prepared in Milli-Q-filtered (Millipore) water and saturated with N2O. The
pH of the solutions was adjusted to the desired value with NaOH or
HClO4. g-Radiolyses were carried out in a panoramic 60Co-source at a dose
rate of 0.16 Gys 1 to total doses of up to 150 Gy (i.e., the conversion of the
substrate remained below 10%), except for the determination of the
secondary product 3-carboxyglutaric acid where the total dose was raised to
400 Gy (the dose of 1 Gy equals 1 J of radiation energy deposited in a
1
k3 ˆ 1.9  107 dm3 mol 1 s at pH 1.2 (free malonic acid) and
1
k3 ˆ 1.1  108 dm3 mol 1 s at pH 8.0 (malonic acid dianion).
The value for the malonic acid monoanion, k3 ˆ 4.2 Â
1
107 dm3 mol 1 s , was determined by build-up kinetics at
3
substrate mass of 1 kg, here 1 dm of solution).
pH 4.1. There is good agreement between the value of k3 for
undissociated malonic acid and the value reported,[13] while
the value of k3 for the malonic acid dianion is lower than that
reported.[13]
The OH radical reacts with malonic acid and acetic acid
largely by abstracting a carbon-bound H atom [reactions (3)
and (4)], and to a minor extent at the carboxyl group, either
by H-atom abstraction when the carboxyl group is proto-
nated, or by oxidation of the carboxylate [reactions (5) and
(6)].
Aliquot samples from the radiolysis of acetic acid were rotary evaporated
and subsequently trimethylsilylated with (bis-N,N-trimethyl)silyltrifluoro-
acetamide (BSTFA) in pyridine. The trimethylsilylated samples were
analysed by GC-MS (Hewlett-Packard 5971A) on a 12 m HP-1 column
(cross-linked methyl silicone gum). Glycolic, succinic, hydroxymalonic and
3-carboxyglutaric acids were identified by comparison of their mass spectra
with those of authentic material. 1,1,2,2-Ethanetetracarboxylic acid and
monomalonyl malonate were not derivatised adequately by this procedure
and cannot be analyzed in this manner.
The quantitative determination of the products was carried out with ion
chromatography (Dionex DX-2010i). Prior to injection, all samples were
adjusted to pH 3.0, because it has been observed that the response of the
acids can vary significantly depending on the pH of the injected solution.
.
.
.
OH (H ) ‡ CH2(CO2H)2 ! H2O (H2) ‡ CH(CO2H)2
(3)
(4)
(5)
(6)
Products from acetic acid: Succinic acid was analysed on an AS14 column
[4 Â 250 mm with 4 Â 50 mm AG14 pre-column, ASRS-ULTRA suppres-
sor, eluent: solution of Na2CO3 (9 Â 10 4 moldm 3) and NaHCO3 (8.5 Â
10 4 moldm 3), flow rate: 1 mLmin 1, retention time: 22 min]. Under these
conditions, glycolic acid has the same retention time as acetic acid
(3.5 min). Therefore, the former was determined on an ICE-AS1 column
.
.
.
OH (H ) ‡ CH3CO2H ! H2O (H2) ‡ CH2CO2H
.
.
OH ‡ CH3CO2H ! H2O ‡ CH3C(O)O
.
.
OH ‡ CH3CO2 ! OH ‡ CH3C(O)O
(10 Â 250 mm, without
a pre-column and suppressor, flow rate:
1.0 mLmin 1, eluent: water), together with succinic acid, since both show
practically the same retention times (4.7 versus 4.8 min) under these
conditions. The yield of glycolic acid was then obtained from the difference
between the results from the AS14 column and the ICE-AS1 column.
3-Carboxyglutaric acid was analysed on an AS16 column [2 Â 250 mm, 2 Â
50 mm AG16 pre-column, ASRS-ULTRA suppressor, eluent: NaOH
(3.5 Â 10 2 moldm 3), flow rate: 0.25 mLmin 1, retention time: 7.4 min].
The methylcarboxyl radical and the corresponding carboxyl
radical derived from malonic acid decompose rapidly [e.g.,
1
reaction (7), k7 ꢁ 1.3 Â 109 s ][11] and, thus, cannot participate
in the bimolecular termination reactions.
.
.
CH3C(O)O
!
CH3 ‡ CO2
(7)
Products from malonic acid: Monomalonyl malonate and ethane-1,1,2,2-
tetracarboxylic acid were analysed under the same conditions that were
used for 3-carboxyglutaric acid (AS16 column, retention times: 8.0 and
24 min). Under these conditions, an unidentified product was eluted
(7.0 min) on the tail of malonic acid (4.5 min). Since the monomalonyl
malonate reference material was not sufficiently pure for calibrational use,
irradiated samples were hydrolysed to malonic acid and hydroxymalonic
acid at 858C and pH 11 for 1 hour in order to quantify this compound. The
hydroxymalonic acid thus produced was analysed on a 4 Â 250 mm AS9
It is apparent that in both the malonic and acetic acid
systems the chemistry is largely determined by the formation
and the termination reactions of the a-carboxymethyl radical,
as indicated by the radiation-chemical yields of the products
derived from the radical (sum of G in the close vicinity of 3 Â
1
10 7 molJ , cf. Figures 1 and 2). The low yields of CO2
column with a 4 Â 50 mm AG9 pre-column, eluent: solution of Na2CO3
(1.8 Â 10 3 moldm
)
and NaHCO3 (1.7 Â 10 3 moldm 3), flow rate:
3
1 mLmin 1. Under these conditions the retention times of hydroxymalonic
acid and malonic acid were 15 min and 12 min, respectively.
Methane and CO2 were scrubbed from the irradiated samples with argon as
the carrier gas[12] and measured by GC on a 3.2 m Porapak-9 capillary
column.
Results and Discussion
The free radical-generating system: Hydroxyl radicals are
generated by the radiolysis of water [reaction (1)]. The
radiation-chemical yields (G values) of the primary radicals
Figure 1. g-Radiolysis of malonic acid in N2O-saturated solution. Yields
ionising
*
(G values) of ethane-1,1,2,2-tetracarboxylic acid ( ), monomalonyl malo-
.
.
‡
H2O
eaq, OH, H , H , H2O2, H2
(1)
!
radiation
~
~
*
nate ( ), unknown product , sum ( ) of these three products.
792
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0704-0792 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 4

Mesomerism
791±795
assignment of structure 5 to the unidentified product could
not be established. It is observed that the proportion in which
products 2 and 4 together with the unidentified compound
(tentatively attributed to 5) are formed depends on the pH of
the solution (Figure 1). Because of the absence of a-hydrogen
in these radicals, disproportionation reactions are not ex-
pected to play a role in these systems. This view is supported
by the absence of hydroxymalonic acid from the products that
might be formed if the disproportionation occurred by
electron transfer.
Intermediate 3 could, in principle, undergo hydrolysis. This
would afford hydroxymalonic acid, besides malonic acid.
Since the former is not observed among the radiolysis
products, the ketonisation of 3 [reaction (11), Scheme 1] must
be favoured considerably over its hydrolysis. This situation
Figure 2. g-Radiolysis of acetic acid in N2O-saturated solution. Yields (G
~
*
*
values) of succinic acid ( ), glycolic acid ( ), sum ( ) of both products.
(Table 1) indicate that the forma-
tion of alkylcarboxyl radicals,
such as reaction (6), is not sig-
nificant. In the acetic acid system,
the formation of small amounts
of methane, derived from the
methyl radical produced in this
decarboxylation, was also detect-
ed (pH 3.5, G(CH4) ˆ 0.06 Â
1
10 7 molJ
;
pH 7.2, G(CH4) ˆ
0.01 Â 10 7 molJ ).
1
The malonic acid system: Three
products are formed, ethane-
1,1,2,2-tetracarboxylic acid, mono-
malonyl malonate and an un-
Scheme 1.
contrasts with the acetic acid system discussed below. The
formation of an unidentified product in addition to 3 and 4 has
already been reported in a UV-photolytic study of aqueous
malonic acid.[14] It remains unclear whether this is identical
known product. Their yields increase linearly with the dose
(data not shown). G values were calculated from the slopes of
the plot of the yield against the dose. The individual yields
vary considerably with pH; however, their sum is essentially
constant over the entire pH range studied (Figure 1). The
~
with the unknown species ( in Figure 1), especially since the
1
value of G(sum) is ꢁ3 Â 10 7 molJ , that is, a good material
analytical procedures differ (HPIC here, HPLC in ref. [14].)
The straightforward pK-type curves (Figure 1, inflexion
points at pH 5.8) shown by these three products imply that
their formation is associated with a property of the malonic
acid radical that is related to its second pKa value of 5.7,[6]
which appears to be the same as that of its parent compound,
pK2 ˆ 5.7. This aspect will be discussed below.
.
balance is obtained, based on the radiolytic yields of OH plus
.
H . Here, it has been assumed that the unknown species has
the same response factor as monomalonyl malonate.
Table 1. Carbon dioxide production (G values [10 7 molJ 1]) from malonic
.
and acetic acids by OH in g-radiolysis (N2O-saturated solutions).
pH
3.0
3.5
7.2
The acetic acid system: The proportion of the major product
succinic acid (7) to glycolic acid (10), increases again at high
pH (Figure 2). A reasonable material balance is obtained over
[a]
[a]
malonic acid
acetic acid
0.2
±
±
±
[a]
0.3
0.2
[a] ± not determined.
*
the entire pH range studied (cf. in Figure 2). The inflexion
point is at pH ꢁ 5, not far from the pKa value of the acetic acid
radical for which values of ꢁ4.5 (pulse radiolysis, optical
detection)[13] and ꢁ4.2 (pulse radiolysis, conductometric
detection)[15] are reported.
The formation of succinic acid (7) is the result of
recombination of the acetic acid radical via its mesomeric
form 6a [reaction (12), Scheme 2]. The formation of glycolic
acid (10) is suggested to proceed via 8, the combination
product of 6a plus 6b [reaction (13)], followed by hydrolysis
[reaction (14), Scheme 2].
As already stated above, the carboxyl radical promptly
eliminates CO2 [cf. reaction (7)]. The formation of mono-
malonyl malonate 4 must, therefore, find an explanation other
than that offered previously.[10] This is provided by the
recognition that the malonic acid radical is capable of
mesomerism (radicals 1a and 1b), as shown in Scheme 1.
Recombination may, in principle, lead to the compounds 2
[reaction (8)] and 4 [reactions (9) and (11)], as well as to the
peroxide 5 [reaction (10)], as shown in Scheme 1, although the
Chem. Eur. J. 2001, 7, No. 4
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0704-0793 $ 17.50+.50/0
793

C. von Sonntag et al.
FULL PAPER
Scheme 3.
An explanation for the pH effect: With regard to the pH-
dependence of product formation, the inflexion point in the
acetic acid system is at pH ꢁ 5 and at pH ˆ 5.8 in the malonic
acid system. These values mirror the pKa values of the
corresponding radical, in the case of malonic acid that of its
second pKa value [equilibrium reactions (18) and (20),
Scheme 4]. The first pKa of the malonic acid radical [equili-
brium reaction (19)] is either considerably below two, or its
Scheme 2.
Interestingly, acetoxyacetic acid (9), the compound that one
expects to be produced provided that the ketonisation
reaction (15) (Scheme 2) proceeds at an appreciable rate in
competition with hydrolysis [reaction (14), Scheme 2], is not
formed under the present conditions, in contrast to the
ketonisation product 4 in the case of malonic acid [reac-
tion (11), Scheme 1]. The reference material, acetoxyacetic
acid, itself is only hydrolysed slowly under these conditions.
The reasons for the differing behavior of the enol esters
produced initially (8 undergoing hydrolysis to glycolic acid 10,
versus 3 undergoing ketonisation to 4) are not evident.
However, the main difference appears to lie in the fact that 3
Scheme 4.
ˆ
has an internal C C bond, while this bond is in a terminal
position in 8. In fact, enol ethers with the corresponding
structural characteristics show a drastic difference with
respect to the rate of hydrolysis.[16, 17] If, in the case of a
determination by pulse radiolysis has failed because of the
spectral similarities of the monoprotonated and fully proto-
nated radicals.[6] The fact that the pKa values of the parent and
of their corresponding radicals are found to be very close in
those cases for which the latter are observable, suggests that
the first pKa value of the malonic acid radical should also not
be far from that of the parent, namely pH 2.8. The absence of
any feature in this pH region (Figure 1) leads to the suggestion
that the reactivity pattern of the malonic acid radical does not
change at pK1, in contrast to the situation near pK2, at which
the mesomeric state of the malonic acid radical changes in
such a way that in the dicarboxylate form the expression of the
radical function in one of the carboxyl groups is essentially
suppressed as the carbonyl moiety of the latter loses its
identity in the dissociated state. As the second carboxyl group
becomes deprotonated, the free-radical function in the
malonic acid radical is frozen onto the methine carbon atom.
By analogy, in the acetate system the radical is highly
delocalised when protonated, but much less so in the anionic
state.
ˆ
terminal C C bond, hydrolysis occurred by protonation at
carbon, as one expects for the ketonisation of the enol esters,
the enol ether case would present an interesting analogy.
The formation of small amounts of a secondary product,
3-carboxyglutaric acid (12), which has been identified by its
mass spectrum (trimethylsilylation, GC-MS) and by compar-
ison with the reference material, has been observed in the
acetic acid system (Figure 3). It is proposed that the precursor
Implications of the postulate of transient enol forms: The
C O-linked termination product (of the a-alkylcarboxy acid-
type) is more stable in the malonic acid than in the acetic acid
system. In the former case, an alkaline pH and an elevated
temperature are required to hydrolyse this product into the
indicator product to be analysed, namely, hydroxymalonic
acid. In the acetic acid system, on the other hand, glycolic acid
is observed directly at ambient temperature upon analysis of
the irradiated solution. This last observation, in conjunction
with the fact that the relevant reference material, acetoxy-
acetic acid, under these conditions undergoes hydrolysis
relatively slowly (i.e., on a timescale of many hours), suggests
that the C O-linked termination product is generated in the
enol form 8, which presumably hydrolyses more rapidly, as the
reaction scheme indeed demands (Scheme 2). The malonic-
Figure 3. Formation of 3-carboxyglutaric acid as a secondary product in
3
the radiolysis of N2O-saturated acetic acid (3 Â 10 3 moldm ) solutions (
*
,
*
pH 3.4; , pH 7.0). For comparison: at a dose of 50 Gy, the concentration of
the primary product succinic acid has reached values of ꢁ10 5 moldm 3 (cf.
Figure 2).
of this compound is the primary product 7 whose rate constant
.
for the reaction with OH [reaction (16), Scheme 3] is about
an order of magnitude larger than that of acetic acid.[18] The
recombination of the succinic acid radical (11) with the acetic
acid radical then gives rise to 12 [reaction (17), Scheme 3].
794
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0704-0794 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 4

Mesomerism
791±795
acid-derived compound would not suffer the same fate if, in its
case, carbonyl ± enol tautomerisation is sufficiently fast and if
the equilibrium strongly favours the carbonyl 4 over the enol 3
form.
Scheme 5.
The initial, though transient, presence of the C O-linked
recombination product in the enol form may give rise to
secondary products (the appearance of such enols may in fact
be of some relevance to minor behavioral details of malonic
acid/bromate-type chemical oscillators). This process would
be initiated by the addition of the a-carboxyalkyl radical to
On the other hand, dibromoacetic acid is also formed by
decarboxylation of dibromomalonic acid; in acidic solutions
this is a relatively slow process (t1/2 ꢁ 1 h in 1 molar H2SO4)[21]
which might, in principle, be capable of being differentiated
from a free-radical process by following the kinetics of its
accumulation.
ˆ
the C C bond and subsequent disproportionation or recom-
bination of this adduct radical. This may lead, for instance, to
a structure with a hemiacetalic function that is expected to
hydrolyse. In principle, a variety of products could be formed
in this manner. The more highly carboxylated such a
compound is, the longer is its ion-chromatographic retention
time, and therefore the greater the difficulty of detecting it at
small concentrations.
[1] R. J. Field, in Oscillations and Traveling Waves in Chemical Systems
(Eds.: R. J. Field, M. Burger), Wiley-Interscience, New York, 1985,
pp. 55 ± 92.
[2] L. Györgyi, T. Turanyi, R. J. Field, J. Phys. Chem. 1990, 94, 7162 ± 7170.
[3] Y. Gao, H.-D. Försterling, Z. Noszticzius, B. Meyer, J. Phys. Chem.
1994, 98, 8377 ± 8380.
[4] B. Neumann, S. C. Müller, M. J. B. Hauser, O. Steinbock, R. H.
Simoyi, N. S. Dalal, J. Am. Chem. Soc. 1995, 117, 6372 ± 6373.
[5] P. O. Kvernberg, E. W. Hansen, B. Pedersen, A. Rasmussen, P. Ruoff,
J. Phys. Chem. 1997, 101, 2327 ± 2331.
[6] M. N. Schuchmann, R. Rao, M. Hauser, S. C. Müller, C. von Sonntag,
J. Chem. Soc. Perkin Trans. 2 2000, 941 ± 943.
[7] C. von Sonntag, H.-P. Schuchmann, Angew. Chem. 1991, 103, 1255 ±
1279; Angew. Chem. Int. Ed. Engl. 1991, 30, 1229 ± 1253.
[8] C. von Sonntag, H.-P. Schuchmann, in Peroxyl Radicals (Eds.: Z. B.
Alfassi), Wiley, Chichester, 1997, pp. 173 ± 234.
[9] I. L. Yurkova, H.-P. Schuchmann, C. von Sonntag, J. Chem. Soc.
Perkin Trans. 2 1999, 2049 ± 2052.
This leads back to the question of the nature of the
~
unknown product ( in Figure 1). Its ion-chromatographic
behaviour suggests that it might actually carry three carboxyl
groups rather than two, since its retention time (7.0 min)
under the present analytical conditions exceeds that of
another tricarboxylic reference compound, citric acid
(6.0 min), preceding (the also tricarboxylic) monomalonyl
malonate (8.0 min) by a similarly narrow margin. In any case,
the fact that its pH dependence parallels that of the C O-
linked recombination product monomalonyl malonate, indi-
cates that it owes its existence to either a C O or an O O
linkage.
[10] A. Sirimungkala, H.-D. Försterling, Z. Noszticzius, J. Phys. Chem.
1996, 100, 3051 ± 3055.
[11] J. W. Hilborn, J. A. Pincock, J. Am. Chem. Soc. 1991, 113, 2683 ± 2686.
[12] F. Weeke, E. Bastian, G. Schomburg, Chromatographia 1974, 7, 163 ±
170.
[13] M. Simic, P. Neta, E. Hayon, J. Phys. Chem. 1969, 73, 4214 ± 4219.
[14] I. Szalai, H.-D. Försterling, Z. Noszticzius, J. Phys. Chem. A 1998, 102,
3118 ± 3120.
[15] M. N. Schuchmann, H.-P. Schuchmann, C. von Sonntag, J. Phys.
Chem. 1989, 93, 5320 ± 5323.
[16] E. Staude, F. Patat, in The Chemistry of the Ether Linkage (Eds.: S.
Patai), Interscience Publishers, London, 1967, pp. 21 ± 80.
[17] I. Janik, P. Ulanski, J. M. Rosiak, C. von Sonntag, J. Chem. Soc. Perkin
Trans. 2 2000, 2034 ± 2040.
Importance of fragmentation reactions in the context of
malonic acid/bromate chemical oscillators: It has been seen
above that in the oxidation of malonic acid by the OH radical,
decarboxylation, that is, fragmentation of the carboxymethyl-
carboxyl radical, is practically absent. The data shown in the
Table 1 of ref. [19] implies that CeIV does not oxidise malonic
acid at the carboxyl group, but that the CO2 evolution is the
result of the decarboxylation of certain oxidation products of
malonic acid that are more reactive towards CeIV in this
respect. Attack at carboxyl group apparently occurs more
readily with methylmalonic acid for which the formation of
the decarboxylation product pyruvic acid has already been
observed.[5] Of course, for ketomalonic and oxalic acids,[20] this
pathway is the only one possible. The formation of dibromo-
acetic acid, which is an important (secondary) product in
addition to bromomalonic acid (primary product) in the CeIV-
catalyzed malonic acid/bromate system,[1] may also involve
this type of process [reactions (21) and (22) in Scheme 5],
which appears to be more facile in bromomalonic than in
malonic acid.[19]
[18] G. V. Buxton, C. L. Greenstock, W. P. Helman, A. B. Ross, J. Phys.
Chem. Ref. Data 1988, 17, 513 ± 886.
Â
Â
Â
[19] S. Nagygyöry, M. Wittmann, S. Pinter, A. Visegrady, A. Dansco, N. B.
Thuy, Z. Noszticzius, L. Hegedüs, H.-D. Försterling, J. Phys. Chem. A
1999, 103 ± 4885.
[20] J.-J. Jwo, R. M. Noyes, J. Am. Chem. Soc. 1975, 97, 5422 ± 5431.
[21] J. Oslonovitch, H.-D. Försterling, M. Wittmann, Z. Noszticzius, J.
Phys. Chem. A 1998, 102, 922 ± 927.
Received: June 7, 2000 [F2532]
Chem. Eur. J. 2001, 7, No. 4
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0704-0795 $ 17.50+.50/0
795