Demethoxylation and hydroxylation of methoxy- and hydroxybenzoic acids by OH-radicals. Processes of potential importance for food irradiation

Article abstract of DOI:10.1139/v01-058

The hydroxylation process for methoxy- and hydroxy-benzoic acids (MBA, HBA) induced by γ-radiation is compared. 2-, 3-, and 4-methoxybenzoic acid as well as 3-hydroxybenzoic acid have been irradiated in N₂O and aerated solutions up to 1.5 kGy. The products were analyzed by HPLC. The results for 2- and 4-HBA have been taken from literature data. The OH^·-adduct distribution is generally the same for the hydroxy- as well as for the methoxy-benzoic acid isomers. With both 4-HBA and 4-MBA more than 65% C3-adducts and about 15% C4-adducts are formed, which could be proved by their reactions with K₃ Fe(CN)₆. Oxidation of the nonipso-adducts of 3-HBA and 3-MBA results in 84 and 87% of the corresponding phenols. Whereas in N₂O-saturated solutions only part of the OH^·-radicals leads to substrate decomposition, in the presence of air, the degradation of both kinds of compounds is equivalent to [OH^·]. The nonipso OH^·-adducts of the HBAs are converted into 68-77% hydroxylation products. With the MBAs, the hydroxylation process is ≤10%. This is attributed to different decay pathways of the peroxyl radicals, intermediates formed by O₂ addition to the OH^·-adducts. The hydroxyperoxycyclohexadienyl radicals of the HBAs decay mainly by HO₂^· elimination to the corresponding phenols, those of the MBAs decay predominantly by fragmentation of the benzene ring, yielding to nonidentified aliphatic products. The replacement of -OCH₃ by -OH is practically not influenced by the presence of oxygen, it increases in the sequence 3-MBA < 4-MBA < 2-MBA. For 2-MBA, yields of more than 15% are obtained. Both processes, hydroxylation as well as demethoxylation, might be of importance for the recognition of radiolytical changes in foodstuff.

Full text of DOI:10.1139/v01-058

394
Demethoxylation and hydroxylation of methoxy-
and hydroxybenzoic acids by OH-radicals.
Processes of potential importance for food irradiation
Birgit Gaisberger and Sonja Solar
Abstract: The hydroxylation process for methoxy- and hydroxy-benzoic acids (MBA, HBA) induced by γ-radiation is
compared. 2-, 3-, and 4-methoxybenzoic acid as well as 3-hydroxybenzoic acid have been irradiated in N2O and aerated
solutions up to 1.5 kGy. The products were analyzed by HPLC. The results for 2- and 4-HBA have been taken from liter-
C
ature data. The OH -adduct distribution is generally the same for the hydroxy- as well as for the methoxy-benzoic acid
isomers. With both 4-HBA and 4-MBA more than 65% C3-adducts and about 15% C4-adducts are formed, which could
be proved by their reactions with K3Fe(CN)6. Oxidation of the nonipso-adducts of 3-HBA and 3-MBA results in 84 and
C
87% of the corresponding phenols. Whereas in N2O-saturated solutions only part of the OH -radicals leads to substrate
C
decomposition, in the presence of air, the degradation of both kinds of compounds is equivalent to [OH ]. The nonipso
C
OH -adducts of the HBAs are converted into 68–77% hydroxylation products. With the MBAs, the hydroxylation process
is ≤10%. This is attributed to different decay pathways of the peroxyl radicals, intermediates formed by O2 addition to the
C
C
OH -adducts. The hydroxyperoxycyclohexadienyl radicals of the HBAs decay mainly by HO2 elimination to the corre-
sponding phenols, those of the MBAs decay predominantly by fragmentation of the benzene ring, yielding to nonidentified
aliphatic products. The replacement of -OCH3 by -OH is practically not influenced by the presence of oxygen, it increases
in the sequence 3-MBA < 4-MBA < 2-MBA. For 2-MBA, yields of more than 15% are obtained. Both processes,
hydroxylation as well as demethoxylation, might be of importance for the recognition of radiolytical changes in foodstuff.
Key words: γ-radiolysis, methoxybenzoic acids, hydroxybenzoic acids, phenolic acids, food components, reaction mech-
anisms, product analysis, HPLC analysis.
Résumé : On a fait une étude comparative des processus d’hydroxylation, induits par un rayonnement γ, des acides
méthoxy- et hydroxy-benzoïques (AMB, AHB). On a irradié les acides 2-, 3- et 4-méthoxybenzoïques ainsi que l’acide
3-hydroxybenzoïque en solutions saturées en N2O et en solutions aérées, jusqu’à 1,5 kGy. On a analysé les produits
par CLHP. Les résultats pour les acides 2- et 4-AHB correspondent aux données que l’on peut retrouver dans la littéra-
C
ture. La distribution de l’adduit OH est généralement la même pour les isomères des acides hydroxy- ainsi que mé-
thoxy-benzoïques. Avec les acides 4-AHB et 4-AMB il se forme dans chaque cas plus de 65% d’adduits C3 et environ
15% d’adduits C4 ce qui a pu être démontré par leurs réactions avec le K3Fe(CN)6. L’oxydation des adduits non ipso
des acides 3-AHB et 3-AMB conduit à la formation de 84 et de 87% des phénols correspondants. Lors des réactions
C
en solutions saturées en N2O, il n’y a qu’une partie des radicaux OH qui conduit à une décomposition du substrat; par
C
C
ailleurs, en présence d’air, la dégradation des deux types de composés est équivalent à la [OH ]. Les adduits de OH
non ipso des AHB sont transformés à 68–77% en produits d’hydroxylation. Avec les AMB, le processus
d’hydroxylation est ≤ 10%. Ceci est attribué à divers modes de décomposition des radicaux peroxyles qui se forment
C
de façon intermédiaire par l’addition de O2 aux adduits OH . Les radicaux hydroxyperoxycyclohexadiényles des AHB
C
se décomposent principalement par une élimination de HO2 conduisant aux phénols correspondants alors que ceux des
AMB se décomposent principalement par une fragmentation du noyau benzénique en conduisant à des produits alipha-
tiques non identifiés. Le remplacement d’un -OCH3 par un -OH n’est pratiquement pas influencé par la présence
d’oxygène; il augmente dans la séquence 3-AMB < 4-AMB < 2-AMB. Pour le 2-AMB, on obtient des rendements de
plus de 15%. Les deux processus, l’hydroxylation ainsi que la déméthoxylation, peuvent présenter de l’intérêt pour la
reconnaissance des changements radiolytiques dans les aliments de nature végétale.
Mots clés : radiolyse γ, acides méthoxybenzoïques, acides hydroxybenzoïques, acides phénoliques, composants alimen-
taires, mécanismes réactionnels, analyse des produits, analyse par CLHP.
[Traduit par la Rédaction] Gaisberger and Solar 404
Received July 20, 2000. Published on the NRC Research Press Web site on April 30, 2001.
B. Gaisberger and S. Solar.1 Institute for Theoretical Chemistry and Structural Biology, University of Vienna, Althanstr. 14, A-
1090 Vienna, Austria.
1Corresponding author (telephone: +43-1-4277-52725; fax: +43-1-4277-52795; e-mail: sonja.solar@univie.ac.at).
Can. J. Chem. 79: 394–404 (2001)
© 2001 NRC Canada
DOI: 10.1139/cjc-79-4-394

Gaisberger and Solar
395
97%), sodium hydroxide (Fluka, 98%), K3Fe(CN)6 (Fluka,
98%), and phosphoric acid (Aldrich, 98%). Solutions were
made with triple-distilled water. To obtain oxidizing condi-
tions, the solutions were saturated with high purity N2O or
oxygen (5.0, Messer Griessheim) prior to irradiation.
Introduction
Hydroxybenzoic acids and their derivatives are ubiquitous
in plant material. The phenolic groups of these compounds
are often methoxylated or present as glycoside (1–3). Upon
irradiation of aqueous aromatic compounds, the main pri-
C
mary process is the electrophilic addition of OH -radicals to
Gamma radiolysis
activated ring positions, which can eventually result in
hydroxylation of the initial material. Such radiolytically
formed hydroxylated compounds can be used as markers for
irradiated food, as is the case with o-tyrosine, originating
from phenylalanine in protein rich food (4–7). The detection
of hydroxylation products from phenolic acid-food compo-
nents was proposed in 1989 as a promising chemical method
for the detection of irradiation treatment of fruits and vegeta-
bles (8, 9), but no marker compounds have been found so
far.
γ-Radiolysis experiments (Gammacell 220, MDS Nordion
International Inc., Kanata, Ontario, Canada) were carried out
at a dose rate of 160 Gy min–1, which was determined by
Fricke dosimetry using a radiation chemical yield of G-
value2 G (Fe3+) = 15.6 = 1.617 µmol J–1 (16).
Product analysis
Product analysis was carried out by high-performance liq-
uid chromatography (HPLC) using a Hewlett–Packard 1050
chromatograph (column: 125 × 4 mm Spherisorb ODS 2 RP-
18 (5 µm), Lichrosorb precolumn, flow rate 1.0 ml min–1)
equipped with a multiple wavelength detector. The mobile
phase consisted of water (Millipore® filtered) and methanol
(Promochem, HPLC grade), the binary gradient was water–
methanol 10–35%. (The gradient was 1–5 min: H2O:CH3OH =
90:10, 5–10 min: H2O:CH3OH = 85:15, 10–20 min:
H2O:CH3OH = 65:35 per volume). As for 3-HBA the mobile
phase was H2O:CH3OH = 90:10. The individual compounds
were detected by measuring their absorptions at 210, 235,
and 255 nm. Concentrations were determined by calibration
with standard solutions. The identity of the products was
confirmed by comparing the retention time and the UV spec-
tra to that of the corresponding standards. Irradiations were
carried out at least in duplicate, the HPLC analyses were re-
producible within 10%.
For methoxybenzoic acids, it has been reported that upon
irradiation in the presence of N2O and at pH ≤ 3, radical cat-
ions and phenoxyl radicals are generated. Radical cations are
C
formed from the nonipso OH -adducts by an acid-catalysed
water elimination. The phenoxyl radicals result from metha-
nol elimination from the ipso adduct carrying the methoxyl
group (10, 11). The work of these authors concentrated on
the detection of these radical intermediates by pulse
radiolysis and on the determination of methanol. There are,
however, no data concerning the formation of the phenolic
products from methoxylated benzoic acids. Upon radiolysis
of ortho- and para-hydroxybenzoic acid, the formation of
phenolic products has been found to be the major reaction
path (12–15).
This study focuses on a comparison of γ-radiation-
induced degradation and hydroxylation processes of benzoic
acids with free as well as methoxylated OH-groups, and its
importance for the detection of radiation-chemical changes
of phenolic components in food of plant origin. Priority was
given to the influence of oxygen on the product distribution.
Results and discussion
Radiolysis in the presence of N2O
In the first step of the investigations, the seven products
C
obtained as a consequence of the OH -radical reactions with
Experimental
the isomers of MBA (anisic acids) were studied. These com-
pounds were used as model substrates for derivatives of
hydroxybenzoic acids having the phenol group linked with
other residues. 3-HBA was investigated for comparison.
Exposure of water or dilute aqueous solutions to γ-radia-
Chemicals and preparation of the solutions
All chemicals were of the highest purity commercially
available: 4-methoxybenzoic acid (4-MBA) (p-anisic acid)
(Aldrich, 99%), 3-MBA (m-anisic acid) (Aldrich, 99%), 2-
MBA (o-anisic acid) (Fluka, 99%), 2-hydroxybenzoic acid
(2-HBA) (Merck, 98%), 3-HBA and 4-HBA (Merck, 98%),
3,4-dihydroxybenzoic acid (3,4-DHBA) (Aldrich, 97%), 2,3-
DHBA (Aldrich, 99%), 2,5-DHBA (Aldrich, 97%), 4-
hydroxy-3-methoxybenzoic acid (vanillic acid (VA))
C
C
tion leads to the primary species e −aq, OH , H , H2, and
H2O2. In the presence of N2O, the e −aq are converted into
C
C
OH (eq. [1]), yielding a G (OH ) = 5.5, which corresponds
to a concentration of 0.57 µmol J–1.
−
−
C
[1]
e aq + N2O → OH + OH + N2
(Aldrich,
97%),
3-hydroxy-4-methoxybenzoic
acid
C
By addition of OH to methoxylated benzoic acids, iso-
meric methoxyhydroxycyclohexadienyl radicals (demon-
strated for 4-MBA in eq. [2]) are formed, which after several
(isovanillic acid (iso-VA)) (Sigma, 97%), 2-hydroxy-3-
methoxybenzoic acid (3-methoxysalicylic acid (3-MSA))
(Aldrich, 97%), 4-MSA (Aldrich, 97%), 5-MSA (Aldrich,
[2]
2 G-value is the number of molecules reacting per 100 eV of absorbed energy. Conversion into SI-units: a G-value of 1 molecule per 100 eV
corresponds to a radiation-chemical yield of 0.10364 µmol J–1.
© 2001 NRC Canada

396
Can. J. Chem. Vol. 79, 2001
Fig. 1. Concentration–dose dependence curves obtained from ir-
radiated N2O-saturated aqueous solutions of 5 × 10–4 M of (a) 4-
MBA, (b) 3-MBA, and (c) 2-MBA at pH = 8 to 9. List of sym-
bols: (A) decrease of the substrates, (B) formation of iso-VA and
4-HBA, (C) formation of 3,4-DHBA, (D) formation of 3-HBA,
(E) formation of VA, (F) formation of 3-MSA, and (G) forma-
tion of 2-HBA.
reaction steps, may lead to hydroxylation and oxidative
demethoxylation.
The substrates degradation and product formations ob-
tained on radiolysis of aqueus N2O-saturated 5 × 10–4 M 2-,
3-, and 4-MBAs (pH 8–9) as a function of dose (50 Gy
steps) are presented in Figs. 1a–c. The irradiations were car-
ried out up to an absorbed dose that gives rise to about 50%
substrate decomposition. Representative chromatograms are
given in Figs. 2a–c.
Degradation
The initial degradation yields (Gi)3, are similar for 4-MBA
and 2-MBA (Gi = 3.9 and 4.0), which corresponds to about
C
70% OH -radical consumption, with the meta-isomer 91% of
C
the OH -radicals giving rise to decomposition. The linearity
of degradation with dose (up to 300 and 400 Gy for 4-MBA
and 3-MBA, respectively), and to more than 700 Gy for 2-
MBA, is given in Figs. 1a–c. The highest dose for a 50%
degradation is required for 4-MBA (0.9 kGy). The HBAs are
decomposed to a clearly smaller extent: Gi (-2-HBA) = 2.4
and Gi (-4-HBA) = 2.8 (14, 15). For 3-HBA a decomposition
yield of Gi (-3-HBA) = 2.1 was obtained, the 50% degrada-
tion is at 1.3 kGy (Table 1). The low initial degradation
yields of the HBAs indicate the occurrence of reactions lead-
ing to a reformation of the substrates.
Product formation
In the case of 4-MBA, two products could be detected by
HPLC (Figs. 1a and 2a): iso-VA resulting from R3, and 4-
C
HBA which has R4 as precursor (the OH -adduct formed by
C
attachment of OH to the ring carbon carrying the methoxyl
group). Such ipso-adducts have been found to decay by acid-
and base-catalyzed or uncatalyzed elimination of CH3OH
producing phenoxyl radicals, the rate constant at pH = 7 was
found to be k ≈ 5 × 104 s–1 (10, 11). In the case of radical R4
the 4-carboxyphenoxyl radical is formed (eq. [3]).
[3]
The yield of the demethoxylation product 4-HBA (Gi =
C
0.5), accounts for about 10% of the total OH -radicals. This
is more than twice as much as reported by O’Neill et al.
(10), who used methanol determinations for quantification.
C
The product arising from an OH attack on the meta-position
of 4-MBA, is only detectable in traces. The finding of hydro-
xylation products in ortho- and para-positions, and the lack
of products in meta-position to the ortho–para-directing
3 Initial G-values (Gi) are calculated from the linear part of dose–yield plots before secondary reactions occur.
© 2001 NRC Canada

Gaisberger and Solar
397
Fig. 2. Chromatograms of 5 × 10–4 M N2O-saturated aqueous solutions of (a) 4-MBA irradiated with 0.5 kGy at pH = 9.0, 255 nm;
(b) 3-MBA irradiated with 0.7 kGy at pH = 8.6, 235 nm; (c) 2-MBA irradiated with 0.7 kGy at pH = 8.5, 235 nm.
C
methoxyl group is in line with the electrophilic nature of the
33% of OH resulted in 3,4-DHBA with Gi = 1.8 (14, 15)
(Table 1).
C
OH -radical (17, 18). The production of the hydroxylation
C
product iso-VA is very low (Gi = 0.46, 8.4% of OH ). The
With 2-MBA, the replacement of the -OCH3-group by
-OH is more effective (Gi (2-HBA) = 1.3), it corresponds to
more than 20% of OH . This value is nearly five times
formation of the secondary product 3,4-DHBA can also be
C
C
observed (Fig. 1a). It results from the reaction of OH with
the primary final substances 4-HBA and iso-VA. In contrast
to 4-MBA, the hydroxylation of 4-HBA is quite efficient;
higher than that found by methanol determinations (10). The
yield of the expected hydroxylation products 2-methoxy-3-
© 2001 NRC Canada

398
Can. J. Chem. Vol. 79, 2001
C
Table 1. Initial yields of degradation (Gi), 50% degradation (kGy), and the hydroxylated products (Gi, percentage of OH ) produced by
C
reaction of OH with 500 µmol of the isomers MBAs (pH = 8) and HBAs (pH = 6) obtained in N2O-saturated solutions.
Decomposition
Hydroxylation total
Replacement -OCH3
Replacement -COO–
Substrate
(500 µmol)
C
C
C
Gi
50% degr. (kGy)
Gi
% of OH
<7
6.9
8.4
35
Gi
% of OH
Gi
% of OH
2-MBA
3-MBA
4-MBA
2-HBAa
3-HBA
4-HBAa
4.0
5.0
3.9
2.4
2.1
2.8
0.6
0.5
0.9
1.1
1.3
0.7
≈0.4
0.38
0.46
2.0
1.3
0.35
0.46
—
—
—
23
6.4
8.4
—
—
—
0
0
0
0
0
0
0.03
0
0.5
0
1.3
1.8
24
33
0.6
11
C
Note: G (OH ) = 5.5 or 0.57 µmol J–1.
aTaken from the literature (14, 15).
hydroxybenzoic acid and 2-methoxy-5-hydroxybenzoic acid
is very low (Figs. 1c and 2c). Since these compounds were
not available for calibration, their yields could only be esti-
mated to approximately Gi < 0.5. Comparing this result with
the hydroxylation of the corresponding HBA, it is again ob-
vious that in the case of the latter, the hydroxylation process
is distinctly higher (Table 1). For salicylic acid, an 83% con-
version of the decomposed acid (Gi (-2-HBA) = 2.4) into
2,3- and 2,5-DHBA (Gi = 1.6 and 0.4, respectively), has
frequently used to oxidize hydroxycyclohexadienyl radicals,
for example, from anisole (17) and phenol (18), to the corre-
sponding phenols. The rate constant for the electron transfer
from the isomeric methoxyhydroxycyclohexadienyl radicals
of anisole to K3Fe(CN)6 has been found to be k(ox)o,p = 2.4
× 109 M–1 s–1 for the ortho- and para-isomer and k(ox)m = 2
× 107 M–1 s–1 for the meta-isomer (17). The distribution of
the isomeric hydroxycyclohexadienyl radicals was deter-
mined for 4-MBA, 3-MBA, and 3-HBA. For 2-MBA this
process could not be carried out because of the lack of
C
been reported (13, 14). This selective reaction of OH en-
ables the detection of the hydroxylation products of salicylic
C
acid as an indicator for the appearance of OH radical in
vivo (19).
[4]
Irradiation of the meta-isomer of anisic acid produces
about 6.5% demethoxylation (Gi (3-HBA) = 0.35) and 7%
total hydroxylation (Gi (VA) = 0.24 and Gi (3-methoxy-
salicylic acid) = 0.14) (Figs. 1b and 2b). The compound 5-
hydroxylation products for calibration. The oxidation step is
demonstrated for the C3-adduct of 4-MBA (eq. [4]).
The detected oxidation products together with those re-
ported for 4-HBA, obtained by using various quinones as
oxidants (12), are summarized in Tables 2a and b.
C
methoxysalicylic acid, expected from the OH adduct in
para-position to -OCH3, could not be detected. Under the
conditions used, it has a retention time of 15.5 min. In this
time range there is, however, no compound detectable
(Fig. 2b). Radiolytic products of 3-HBA have been determined
The result clearly shows, that the MBAs and HBAs ex-
C
C
in the course of this work. Thirty-eight percent of the OH -
hibit a similar OH -adduct distribution. The yields of
radicals contribute to its degradation (Gi (-3-HBA) = 2.1),
the 50% decomposition dose is 1.3 kGy. The hydroxylation
products 2,3-, 2,5-, and 3,4-DHBA are formed with Gi =
0.37, 0.43, and 0.52, respectively (Figs. 3a and b).
hydroxylation products primarily formed on the nonipso
OH -adducts are practically the same. 4-MBA and 4-HBA
C
result in 68 and 69% total hydroxylation, respectively, with
3-MBA and 3-HBA 87 and 84% phenolic compounds are
produced, respectively, (Tables 2a and b). For anisole the
demethoxylation process has been found to be independent
of the presence of an oxidant (17), but it is influenced by pH
(11). Radiolysis of 4-MBA without oxidants at pH > 10
showed an increase of 4-HBA formation to 17%, which im-
proves the material balance to about 85%. 4-Methoxyphenol
and 3-methoxyphenol, the decarboxylation analogoues to
hydroquinone and resorcinol, could not be detected (Ta-
bles 2a and b). Nevertheless, in the case of 4-MBA, the re-
The total hydroxylation attributed to substrate degradation
is 60%, which is comparable to that of 4-HBA. In contrast to
2- and 4-HBA, no replacement of the carboxyl group is ob-
servable. This may be attributed to the meta-positioning of
the ortho–para-directing -OH-group vs. -COO–, which re-
C
duces the probability of OH -attachment at the carboxyl
group.
The yields of degradation, and product formation for the
methoxy- and hydroxy-benzoic acids, are summarized in Ta-
ble 1. Hydroxylation is assigned to the total amount of prod-
C
maining 15% of OH presumably adds to the C-1 position;
this adduct probably decays by elimination of OH– forming
radical cations (eq. [5]).
C
C
ucts resulting from OH -adducts where OH is attached to
ring carbons not carrying methoxyl-, hydroxyl-, or carboxyl-
groups.
To elucidate whether the pronounced differences in
hydroxylation of hydroxy- compared to methoxy-benzoic ac-
ids could at least partly be attributable to a different pattern
[5]
C
of OH -radical addition to these substrates, irradiations in
the presence of K3Fe(CN)6 have been performed. K3Fe(CN)6
and the quinones p-benzoquinone and 2-methylbenzo-
quinone, which have a somewhat lower redox potential, are
C
Since the yields of the nonipso OH -adducts of methoxy-
and hydroxy-benzoic acids are equivalent (Tables 2a and b),
© 2001 NRC Canada

Gaisberger and Solar
399
C
Table 2a. Comparison of the yields of oxidation products from the primary OH -adducts from 4-MBA (500 µmol substrate, 100 µmol
K3Fe(CN)6, N2O) and 4-HBA (12).
Ring position
4-Methoxybenzoic acid
Product
4-Hydroxybenzoic acid
Product
C
C
% of OH
% of OH
C1
C2
C3
C4
4-Methoxyphenol
4-MSA
iso-VA
0
traces
68
10 (17a)
-78 (-85a)
-68
Hydroquinone
2,4-DHBA
3,4-DHBA
4-HBA
16
4
65
15b
100
69
4-HBA
Total yield
Total hydroxylationc
aAt pH > 10.
bTaken from the difference to 100%.
C
cResulting from the nonipso OH -adducts.
C
Table 2b. Comparison of the yields of oxidation products from the primary OH -adducts from 3-MBA and 3-HBA (500 substrate, 100
µmol K3Fe(CN)6, N2O).
Ring position
3-Methoxybenzoic acid
Product
3-Hydroxybenzoic acid
Product
C
C
% of OH
% of OH
C1
C2
C3
3-Methoxyphenol
3-MSA
3-HBA
Not detected
29
10
Resorcinol
2,3-DHBA
3-HBA
Not detected
27
16a
C4
C5
C6
VA
3-H-5-MBAb
5-MSA
29
3,4-DHBA
3,5-DHBA
2,5-DHBA
31
Not detected
Not detected
29
97
87
26
100
84
Total yield
Total hydroxylationc
aTaken from the difference to 100%.
b3-Hydroxy-5-methoxybenzoic acid.
C
cResulting from the nonipso OH -adducts.
their diverse hydroxylation can be attributed to the different
decay pathways of the OH -adducts. Principally, these tran-
sients can undergo unimolecular (eqs. [6] and [7]), and bi-
molecular reactions (eqs. [8], [9], and [10]):
In acid solutions in the case of methoxylated benzoic ac-
ids (R = CH3), the formation of radical cations (eq. [6]), has
been found to be the major path. Their production have been
followed for di- and tri-methoxylated benzoic acids in the
pH range 1–3 using optical and conductometric pulse
radiolysis (10). For HBAs (R = H, eq. [7]), an acid- and
base-catalyzed water-splitting takes place (12–15). During γ-
radiolysis the concentration of radicals is very low, and first-
order reactions are likely to proceed faster than second-order
ones. Thus the radicals formed initially may be altered by
first-order processes (eqs. [6] and [7]), before radical centers
are finally removed by second-order dismutation and dimeri-
zation (eqs. [8], [9], and [10]). Cross reactions (eq. [8]) be-
tween phenoxyl radicals (e.g., resulting from ipso-adducts,
eq. [3]) and hydroxycyclohexadienyl radicals, seem quite
probable because of the driving force exerted by the differ-
ent redox potentials. For the MBAs, phenoxyl radical pro-
duction is restricted to eq. [3]. Therefore hydroxylation- and
demethoxylation-products formed via eq. [8], are generated
in the same proportion as can be seen for 3- and 4-MBA (>6
and 8% each) (Table 1). In the case of HBAs, besides
hydroxylation, the starting compound is regenerated. The
latter is reflected in the lower initial degradation yield of the
HBAs (Table 1).
C
[6]
[7]
[8]
[9]
Radiolysis in the presence of SO4−C
Degradation rates of 4-MBA initiated by its radical cation
have not yet been reported. The generation of 4-MBA+C is
achieved by oxidation of 4-MBA with SO4−C (k13 = 3.5 ×
[10]
© 2001 NRC Canada

400
Can. J. Chem. Vol. 79, 2001
Fig. 3a. Concentration–dose dependence curves obtained from irradiated N2O-saturated aqueous solutions of 5 × 10–4 M 3-HBA at pH =
6.2. List of symbols: (A) decrease of 3-HBA, (B) formation of 3,4-DHBA, (C) formation of 2,5-DHBA, and (D) formation of 2,3-
DHBA. Fig. 3b. Chromatogram at 210 nm of 5 × 10–4 M N2O-saturated aqueous solution of 3-HBA irradiated with 1.2 kGy at pH = 6.2.
109 M–1 s–1) (20), which is formed by reduction of S2O28−
biphenylic products, which has been demonstrated for
halogenated benzoic acids (26). On the basis of these results,
the following decay reactions of 4-MBA+C are likely to pro-
ceed:
with e −aq in air-free solutions (k12 = 1.1 × 1010 M–1 s–1) (21).
C
The OH -radicals are scavenged by 2-methyl-2-propanol
(t-BuOH) (k11 = 6 × 108 M–1 s–1) (22) (eqs. [11], [12], and
[13]). Concentrations used were 500 µmol 4-MBA, 5 mmol
K2S2O8, and 100 mmol t-BuOH (pH = 6.5, Ar saturation).
Since k (e −aq + 4-MBA) = 2 × 109 M–1 s–1) (23) 5 mmol
[14]
[15]
K2S2O8 are sufficient to scavenge 98% of the e a−q
.
C
C
[11] (CH3)3COH + OH → H2O + (CH3)2C H2COH
[12] e a−q + S2O82− → SO4−C + SO42−
[13] SO4−C + 4-MBA → 4-MBA+C + SO24−
Radical cations of benzoic acid have been found to decay
by decarboxylation, giving rise to phenyl radicals, and (or)
by addition of water to hydroxycyclohexadienyl radicals. For
benzoate the distribution of these radicals is 70%:30% (24).
Phenyl radicals are known to decay by H-abstraction, for ex-
ample, from donors such as alcohols or by addition to an ar-
omatic system resulting in the formation of substituted
cyclohexadienyl radicals. The rate constants for both pro-
cesses are in the range of 106–107 M–1 s–1 (25). The
cyclohexadienyl radicals can be oxidized, thus giving stable
[16]
[17]
© 2001 NRC Canada

Gaisberger and Solar
401
C
Table 3. Initial yields of degradation (Gi), 50% degradation (kGy), and hydroxylated products (Gi, percentage of OH ), obtained by
radiolysis of 500 µmol aerated solutions of the isomers of MBAs (pH = 8) and HBAs (pH = 6).
Decomposition
Hydroxylation total
Replacement -OCH3
Replacement -COO–
Substrate
(500 µmol)
C
C
C
Gi
50% (kGy)
Gi
% of OH
Gi
% of OH
Gi
% of OH
2-MBA
3-MBA
4-MBA
2-HBAa
3-HBA
4-HBAa
2.7
2.8
2.9
2.6
2.7
2.7
1.1
0.9
1.1
1.4
1.4
1.0
<0.28
0.28
0.16
1.95
2.16
1.9
<10
10
5.7
69
77
68
0.46
0.18
0.32
—
—
—
16.4
6.4
10.6
—
—
—
0
0
0
0.45
0
0
0
0
16
0
18
0.5
C
Note: G (OH ) = 2.8 or 0.29 µmol J–1.
aTaken from the literature (14, 15).
If water addition (eq. [15]) is of importance, hydroxylated
compounds may be formed after fixing of the hydroxyl moi-
ety on the ring via e.g., eqs. [8] and [9]. Methoxyphenyl rad-
icals (eq. [14]) should result in anisole (eq. [16]), and (or) in
various methoxylated biphenyls, having e.g., eq. [17] as pre-
cursor. Since such biphenyl compounds were not available
as reference substances, their formation could not be con-
firmed by HPLC.
[21]
The degradation rate is rather slow (Gi (-4-MBA) = 0.6),
which corresponds to 22% of e −aq. No hydroxylation prod-
ucts could be detected, demonstrating that water addition to
the radical cation (eq. [15]) plays a minor role. Anisole is
not produced either, therefore, presumably, biphenyls formed
via eq. [17], can be expected to be the major products in the
decay process of the radical cations of 4-MBA. Since no
radiolytic products were detectable, the concentration vs.
dose plot, as well as the HPLC chromatogram, are not pre-
sented. The data were of similar quality to those in earlier
figures.
It was of particular interest whether the presence of oxy-
gen would enhance the formation of hydroxylation products
of the MBAs as was the case in the presence of the oxidant
K3Fe(CN)6 and as has been reported for 2- and 4-HBA (14,
15). As a consequence, this would enhance the probability
for using hydroxylated products of phenolic acids as markers
for irradiation treatment of food-vegetables, as has been sug-
gested previously (8, 9).
Aerated solutions of 500 µmol 2-, 3-, and 4-MBA as well
as 3-HBA (pH 8), have been irradiated in 50 Gy steps. The
substrate degradation and the concentration of the phenolic
products are presented in Figs. 4a–d as a function of dose.
Characteristic chromatograms are given for 3-MBA and 3-
HBA in Figs. 5a and b.
Radiolysis in the presence of oxygen
Oxygen is present in plant tissue, and, therefore, its influ-
ence on radiolytic product formation has to be taken into ac-
count. The substrates have been irradiated in aerated
solutions ([O2] = 0.25 × 10–3 M). In the presence of oxygen,
The initial yields of degradation, phenolic product forma-
tion, as well as the 50% decomposition dose for the isomers
of MBA and HBA are summarized in Table 3.
C
the reducing radicals H and e −aq are scavanged, forming
HO2 and O2−C
:
C
Decomposition
In contrast to the N2O saturated solutions in the presence
of air, no substrate reversion takes place. All OH -radicals
C
[18] H + O2 → HO2 k = 2.1 × 1010 M–1 s–1 (20)
C
C
contribute to the degradation of the hydroxy- and methoxy-
benzoic acids, indicating the addition of O2 to the OH -ad-
[19] e a−q + O2 → O2−C k = 2.0 × 1010 M–1 s–1 (20)
C
ducts. Their decomposition rates are comparable, a dose of
about 1 kGy is required for 50% degradation. 2-HBA and 3-
HBA exhibit a somewhat higher radiation resistance; their
half-life dose is 1.4 kGy.
C
[20] H+ + O2−C X HO2 pK = 4.8
The rates of reactions of superoxide radicals with organic
substances have been found to be rather slow (27), thus
Product formation
C
OH -radicals remain the most reactive species. As the main
Upon radiolysis of the HBAs in the presence of air, a con-
siderable amount of phenolic products has been found. For
salicylic acid: 2,3-DHBA (Gi = 0.85) and 2,5-DHBA (Gi =
1.1) (13, 14), this corresponds to a conversion of 69% of the
difference to the N2O conditions, the addition of oxygen to
C
the primarily formed OH -adducts is to be expected. The
peroxyl radicals formed thereby may undergo various decay
C
processes. The most ubiquitous are HO2 or O2−C elimination,
C
OH -radicals into hydroxylation of the substrate. In the case
especially by those radicals which carry the peroxyl radical
function in ortho-position to the H-atom to be eliminated
(28).
of 4-HBA, 3,4-DHBA is formed with Gi = 1.9 (i.e., 68% of
OH ) (12, 14). 3-HBA yields three hydroxylation products:
C
2,3-DHBA (Gi = 0.61), 2,5-DHBA (Gi = 0.77), and 3,4-
C
C
This reaction course is demonstrated on the C3-OH -ad-
ducts of 4-MBA and 4-HBA:
DHBA (Gi = 0.77), in total this amounts to 77% of the OH -
radicals. Whereas the conversion factor of OH to
C
© 2001 NRC Canada

402
Can. J. Chem. Vol. 79, 2001
Fig. 4. Concentration–dose dependence curves obtained from irradiated aerated aqueous solutions of 5 × 10–4 M of (a) 4-MBA, (b) 3-
MBA, (c) 2-MBA, and (d) 3-HBA at pH = 7–9. List of symbols: (A) decrease of the substrates, (B) formation of 4-HBA, (C) forma-
tion of iso-VA, (D) formation of 3,4-DHBA, (E) formation of 3-HBA, (F) formation of VA, (G) formation of 3-MSA, (H) formation of
3,4-DHBA, (I) formation of 2-HBA, (J) formation of 3,4-DHBA, (K) formation of 2,5-DHBA, and (L) formation of 2,3-DHBA.
hydroxylation was 0.3 ± 0.06 for the HBAs in N2O, in
aerated solutions its value is 0.75 ± 0.05, which confirms the
proceeding of eq. [21] as major reaction pathway at
radiolysis of HBAs in aerated solutions.
were detectable, VA and 3-MSA with GI = 0.14 each. The
product with -OH in para-position to the -OCH3 group (5-
methoxysalicylic acid), is not formed, as was the case in
N2O-saturated solutions (Table 3). In addition, experiments
in oxygen-saturated solutions ([O2] = 1.25 × 10–3 M) have
been carried out. Even a level of oxygen which was five
times higher did not influence the yield of the hydroxylation
products.
For the MBAs, the presence of air is not reflected in an
C
enhancement of OH conversion, either into hydroxylation or
into the oxidative replacement of the methoxyl group (Ta-
bles 1 and 3). The yields of both product types about 10%,
except the demethoxylation of 2-MBA, which is >15%. The
hydroxylation product for 4-MBA (iso-vanillic acid) is
formed with Gi = 0.16. The products expected for 2-MBA,
2-methoxy-3-hydroxybenzoic acid and 2-methoxy-5-
hydroxybenzoic acid, which were not available for calibra-
tion, could be estimated to be <10% (Fig. 4c). From the
meta-isomer 3-MBA, only two of three possible substances
C
The rate constants of the OH -adducts of the MBAs and
HBAs with oxygen are not known, but it can be assumed
that they are comparable. For anisole and benzoate this rate
constant has been found to be k = 8 × 108 M–1 s–1 and k =
2 × 108 M–1 s–1, respectively, (28, 29). Thus it can be de-
C
duced that the rate constant for oxygen addition to the OH -
adducts of the MBAs is sufficiently large to ensure a quantitative
© 2001 NRC Canada

Gaisberger and Solar
403
Fig. 5. Chromatograms of 5 × 10–4 M aerated aqueous solutions of (a) 3-MBA irradiated with 1.0 kGy at pH = 8.4, 235 nm; (b) 3-
HBA irradiated with 1.2 kGy at pH = 6.9, 210 nm.
conversion into peroxyl radicals, as is the case with the
HBAs. The decay, however, takes different courses. As
could be shown for benzene and substituted benzenes, if O2
instead of K3Fe(CN)6 is used as oxidant, the conversion of
the hydroxycyclohexadienyl radicals into the corresponding
phenols is no longer quantitative (29, 30). There are mainly
the presence of oxygen, the product partition in aerated and
N2O solutions is comparable (Tables 1 and 3).
Conclusions
C
two competitive decay pathways for hydroxycyclohexa-
Although the primarily formed OH -adducts are present in
C
dienylperoxyl radicals: (i) HO2 elimination leading to phe-
the same concentration for both the hydroxy- and methoxy-
benzoic acids, their final hydroxylation product yields are
nols, and (ii) a formation of endo-peroxidic structures,
which in turn are converted into peroxyl radicals and subse-
quently undergo fragmentation (29–31). Apparently frag-
mentation is the dominant reaction path with the MBAs.
This is confirmed by the fact that no aromatic products are
detectable by HPLC. Presumably aliphatic aldehydes and ac-
ids are the radiolytic products.
The yields for the products resulting from -OCH3 replace-
ment did not change in aerated solutions. This independence
of demethoxylation on the presence of an oxidant is in line
with the data obtained for anisole (17). The dominant reac-
tion for the ipso-adducts carrying the methoxyl group is a
prompt methanol elimination under formation of phenoxyl
radicals. Since the decay of the latter is not influenced by
C
distinctly different. Whereas 68–77% of the OH -adducts of
the HBAs are converted to hydroxylation products in the
presence of air, only about 10% hydroxylation occurs with
C
the OH -adducts of the MBAs. This fact is attributable to
different decay pathways of the intermediate peroxyl radical.
The detection of hydroxylation products of phenolic acids
has been suggested as a promising method for the identifica-
tion of irradiated food of plant origin. Free HBAs, in the
nonmethoxylated form, will produce sufficient amounts of
hydroxylation products. If the OH-groups of phenolic acid
derivatives are, however, linked to other residues (e.g., pres-
ent as ether, ester, or glycoside) the hydroxylation process
can expected to be of less significance.
© 2001 NRC Canada

404
Can. J. Chem. Vol. 79, 2001
12. R.F. Anderson, K.P. Patel, and M.R.L. Stratford. J. Chem. Soc.
Faraday Trans. 1. 83, 3177 (1987).
13. Z. Maskos, D. Rush, and W.H. Koppenol. Free Radicals Biol.
Med. 8, 153 (1990).
14. F. Swoboda. Msc. Thesis. University of Vienna, Austria. 1996.
15. F. Swoboda and S. Solar. Radiat. Phys. Chem. 56, 291 (1999).
16. H. Fricke and E.J. Hart. In Radiation dosimetry. Edited by F.H.
Attix, W.C. Roesch, and E. Töchilin. Academic Press, New
York. 1966. p.167.
The oxidative replacement of the -OCH3 by -OH amounts
to between 10–20%. Such addition–elimination reactions
may be of importance for e.g., the release of glycoside resi-
dues. Preliminary radiolysis results from chlorogenic acid
(3-(3,4-dihydroxycinnamoyl)quinic acid) showed the forma-
tion of caffeic acid (3,4-dihydroxycinnamic acid), indicating
the occurrence of this process.
17. S. Steenken and N.V. Raghavan. J. Phys. Chem. 83, 3101
(1979).
18. N.V. Raghavan and S. Steenken. J. Am. Chem. Soc. 102, 3495
(1980).
19. M. Grootveld and B. Halliwell. Biochem. J. 237, 499 (1986).
20. P. Neta, V. Madhavan, H. Zemel, and R.W.Fessenden. J. Am.
Chem. Soc. 99, 163 (1977).
21. P. Neta, R.E. Huie, and A.B. Ross. J. Phys. Chem. Ref. Data.
17, (1988).
22. G.V. Buxton, C.L. Greenstock, W.P. Helman, and A.B. Ross. J.
Phys. Chem. Ref. Data. 17, (1988).
23. B. Gaisberger. Msc. Thesis. University of Vienna, Austria.
1998.
24. H. Zemel and W. Fessenden. J. Phys. Chem. 82, 2670 (1978).
25. V. Madhavan, R.H. Schuler, and R.W. Fessenden. J. Am.
Chem. Soc. 100, 888 (1978).
26. I.A. Jankovic, L.R. Josimovic, and S.V. Jovanovic. Radiat.
Phys. Chem. 51, 293 (1998).
27. B.H.J. Bielski, D.E. Cabelli, R.L. Arudi, and A.B. Ross. J.
Phys. Chem. Ref. Data. 14, 1041 (1985).
28. C. von Sonntag and H.-P. Schuchmann. In Peroxyl radicals.
Edited by Z.B. Alfassi. John Wiley & Sons Ltd., Chichester.
1997. p. 173.
29. X. Fang, X.-M. Pan, A. Rahmann, H.-P. Schuchmann, and C.
von Sonntag. Chem. Eur. J. 7, 423 (1995).
30. X.-M. Pan, M.N. Schuchmann, and C. von Sonntag. J. Chem.
Soc. Perkin Trans. 2, 289 (1993).
Acknowledgement
The financial support of this work by the Oesterreichische
Nationalbank, Jubiläumsfonds, project JFP 6749, is grate-
fully acknowledged.
References
1. K. Herrmann. Fortschr. Chem. Org. Naturst. 3, 73 (1978).
2. Römpp Lexikon Lebensmittelchemie. Edited by G. Eisenbrand
and P. Schreier. G. Thieme Verlag Stuttgart, New York. 1995.
3. H.-D. Belitz and W. Grosch (Editors). In Lehrbuch der
Lebensmittelchemie. 3rd. ed. Springer-Verlag, Berlin. 1987.
p. 494.
4. N. Chuaqui-Offermanns and T.E. McDougall. J. Agric. Food
Chem. 39, 200 (1991).
5. C. Krach, G. Sontag, S. Solar, and N. Getoff. Z. Lebensm.
Unter.-Forsch. A204, 417 (1997).
6. C. Krach, G. Sontag, and S. Solar. Food Res. Int. 32, 43
(1999).
7. C.H. McMurray, E.M. Stewart, R. Gray, and J. Pearce
(Editors). Detection methods for irradiated foods. The Royal
Society of Chemistry, Cambridge. 1996. p. 303.
8. M. Grootveld and R. Jain. Radiat. Phys. Chem. 34, 925 (1989).
9. M. Grootveld and R. Jain. Free Radical Res. Commun. 5, 271
(1989).
10. P. O’Neill, D. Schulte-Frohlinde, and S. Steenken. J. Chem.
Soc. Faraday Discuss. 63, 141 (1977).
11. S. Steenken and P. O’Neill. J. Phys. Chem. 81, 505 (1977).
31. G. Merga, H.-P. Schuchmann, M. Rao, and C. von Sonntag. J.
Chem. Soc. Perkin Trans. 2, 1097 (1996).
© 2001 NRC Canada