Ozone and hydroxyl radicals induced oxidation of glycine

Article abstract of DOI:10.1016/S0043-1354(98)00230-9

The study of the oxidation of glycine by ozone and hydroxyl radicals has been undertaken. Different experimental procedures involving the H₂O₂/UV process or ozone in the presence of hydrogen peroxide on one hand, and ozone alone on the other hand allowed to distinguish between the two oxidation pathways. It has been shown that nitrites and nitrates are observed byproducts resulting from molecular ozonation whereas ammonium ions constitute the only inorganic nitrogen species from hydroxyl radicals induced decomposition of glycine. Therefore, the ozonation process promoting radical pathways will favor ammonia production vs. nitrates. Among the organic compounds resulting from oxidation by hydroxyl radicals, oxalic acid was the major byproduct detected in the presence of oxygen. Oxamic acid and formic acid were also observed. In addition, the nitrogen mass balance indicated the formation of unidentified nitrogen byproducts derived from glycine. These nitrogen compounds could result from the interaction of oxygen with nitrogen radical, but, mainly, with α-amino radical of glycine. This reaction is very different from the molecular ozonation which directs the attack on the nitrogen functional group before the C-N bond cleavage of the amino acid molecule. The study of the oxidation of glycine by ozone and hydroxyl radicals has been undertaken. Different experimental procedures involving the H₂O₂/UV process or ozone in the presence of hydrogen peroxide on one hand, and ozone alone on the other hand allowed to distinguish between the two oxidation pathways. It has been shown that nitrites and nitrates are observed byproducts resulting from molecular ozonation whereas ammonium ions constitute the only inorganic nitrogen species from hydroxyl radicals induced decomposition of glycine. Therefore, the ozonation process promoting radical pathways will favor ammonia production vs. nitrates. Among the organic compounds resulting from oxidation by hydroxyl radicals, oxalic acid was the major byproduct detected in the presence of oxygen. Oxamic acid and formic acid were also observed. In addition, the nitrogen mass balance indicated the formation of unidentified nitrogen byproducts derived from glycine. These nitrogen compounds could result from the interaction of oxygen with nitrogen radical, but, mainly, with α-amino radical of glycine. This reaction is very different from the molecular ozonation which directs the attack on the nitrogen functional group before the C-N bond cleavage of the amino acid molecule.

Full text of DOI:10.1016/S0043-1354(98)00230-9

Wat. Res. Vol. 33, No. 2, pp. 433±441, 1999
1998 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
#
PII: S0043-1354(98)00230-9
0043-1354/98 $19.00 + 0.00
OZONE AND HYDROXYL RADICALS INDUCED
OXIDATION OF GLYCINE
Â
P. BERGER, N. KARPEL VEL LEITNER*, M. DORE and B. LEGUBE
*
M
Laboratoire de Chimie de l'Eau et de l'Environnement, UPRES A CNRS 6008, Ecole Supe
d'Ingenieurs de Poitiers, Universite de Poitiers, 40 Avenue du Recteur Pineau, 86022 Poitiers Cedex,
France
Â
rieure
Â
Â
(
First received December 1997; accepted in revised form May 1998)
AbstractÐThe study of the oxidation of glycine by ozone and hydroxyl radicals has been undertaken.
Di€erent experimental procedures involving the H /UV process or ozone in the presence of hydrogen
2
O
2
peroxide on one hand, and ozone alone on the other hand allowed to distinguish between the two oxi-
dation pathways. It has been shown that nitrites and nitrates are observed byproducts resulting from
molecular ozonation whereas ammonium ions constitute the only inorganic nitrogen species from hy-
droxyl radicals induced decomposition of glycine. Therefore, the ozonation process promoting radical
pathways will favor ammonia production vs. nitrates. Among the organic compounds resulting from
oxidation by hydroxyl radicals, oxalic acid was the major byproduct detected in the presence of oxygen.
Oxamic acid and formic acid were also observed. In addition, the nitrogen mass balance indicated the
formation of unidenti®ed nitrogen byproducts derived from glycine. These nitrogen compounds could
result from the interaction of oxygen with nitrogen radical, but, mainly, with a-amino radical of gly-
cine. This reaction is very di€erent from the molecular ozonation which directs the attack on the nitro-
gen functional group before the C±N bond cleavage of the amino acid molecule. # 1998 Elsevier
Science Ltd. All rights reserved
Key wordsÐglycine, hydroxyl radicals, ozonation, nitrate, ammonia, total nitrogen, amino acid,
advanced oxidation process
INTRODUCTION
with ozone has in most cases not been clearly men-
tioned. More recently, Le Lacheur and Glaze (1996)
The presence of organic nitrogen compounds in
have performed a complete study on the ozonation
waters originates mainly from humic substances but
also from total amino acids which can reach several
ozone oxidatively decarboxylates the amino acid to
micrograms per liter in surface waters. Within this
class, glycine, alanine, aspartic acid, glutamic acid
aldehydes and carboxylic acids. Under radical-pro-
and serine represent the major amino acids found
of serine. These authors concluded that molecular
yield nitrates, one and two carbon mixed functional
moting conditions, ammonia and byproducts with
three carbon atoms are found.
During this work, the oxidation of glycine in
in natural waters. These compounds are of interest
to drinking water suppliers owing to the potentially
harmful byproducts resulting from chlorination but
aqueous solutions was investigated under speci®c
also from ozonation of these compounds. There is
experimental conditions allowing to clearly dis-
little information concerning the byproducts and
tinguish between molecular ozone and hydroxyl
the mechanism of ozonation of amino acids in aqu-
radicals impact. Hydrogen peroxide photolysis was
eous solutions. Although a number of studies report
used to study radical reactions in the presence and
the ozonation of amino acids, there is still some
in the absence of dissolved oxygen whereas ozona-
doubt concerning the origin of the formation of the
tion in the presence of carbonate/bicarbonate ions
inorganic end products observed. Studies on the
as radical scavengers led to the study of the direct
reaction of ozone. The ozone/hydrogen peroxide
oxidation process with di€erent hydrogen peroxide
doses was also investigated.
ozonation of glycine in aqueous solution showed
the formation of nitrate and ammonia, as well as
formic acid and carbon dioxide (Duguet et al.,
1980; Duguet et al., 1985). But the relative import-
ance of radical reaction pathway vs. direct reaction
Usually, the level of glycine ranges from 6 to
0% of the total dissolved amino acids for waters
3
of di€erent origins (Lytle and Perdue, 1981;
Chuboda et al., 1986; Malcolm, 1990; McKnight et
al., 1991; Dossier-Berne et al., 1994; Hubberten et
al., 1994).
*
Author to whom all correspondence should be addressed.
Tel: +33-5-49453915/49453916; Fax: +(33-5-
9453768, E-mail: nathalie.karpel@esip.univ-poi-
tiers.fr].
[
4
433

4
34
P. Berger et al.
Á
Glycine is a neutral amino acid which is charac-
terized by two acidity constants,
Table 2. Calculated competition values for OH radicals for ozona-
tion experiments in the presence of 0.2 M hydrogencarbonate ions.
QO =19.7 mmol h , pH = 8.3 (phosphate bu€er)
�1
3
pK
aꢀ‡H3N±CH2 ±COOH= H3N±CH2 ±COO
‡
�
ˆ 2:34 ꢀ258C†,
�1 �1 a
kOHÁ (l mol
s )
†
Reaction
ꢀO=Ogly†OHÁ
Á
Á
Á
7
9
6
OH +Gly
OH +Gly
1.7 Â 10
1.9 Â 10
8.5 Â 10
1
5.6
95
aꢀ‡H3N±CH2 ±COO =H2N±CH2 ±COO
�
�
ˆ 9:60 ꢀ258C†:
�
pK
†
�
3
OH +HCO
The rate constants for the reaction of hydroxyl rad-
7
a
From Buxton et al. (1988).
�1 �1
ical with glycine are about 1.7 Â10 l mol
s
for
9
�1 �1
the neutral form and 1.9 Â10 l mol
s
at pH 10 ously into the reactor to maintain the ratio (O /H O )
for the anionic form (Buxton et al., 1988). For the introduced equal to 2 mol/mol. Two sets of experiments
3
2
2
were carried out, one with no initial concentration of
H O and another with 75.5 mM before ozonation.
2 2
For the study of molecular ozone reactions, bicarbon-
ates (0.2 M) have been added in solutions as radical sca-
vengers. The relative contribution of the reactions
competitive to the consumption of glycine by ozone and
hydroxyl radicals are expressed by the competition value
O. This value O represents the product of the concen-
tration of the competitor with the rate constant towards
ozone reactions, the rate constants are respectively
�
5
1
�1
3
2
2.6 l mol
.08 Â10 l mol
s
for the neutral form and
s for the anionic form (Pryor et
�1 �1
al., 1984).
MATERIALS AND METHODS
Experimental procedure
the
(
considered
oxidant
:
at
pH
8
:
OOH ꢀor OO † ˆ ‰MŠ kOH ꢀor kO †). These values which
3
0
3
Aqueous solutions of glycine (Merck, minimal purity
9.7%) were prepared in ultra-pure water delivered by a
have been normalized to the reactions with glycine (Ogly
)
9
Millipore MilliRO-MilliQ system and bu€ered with phos-
show the favored reaction pathways. Tables 1 and 2 sum-
marize the calculations corresponding to the di€erent ex-
perimental conditions used in this study. Table 1 indicates
that in the presence of a high initial concentration of
hydrogen peroxide (experiment 2) the ozone reacts mainly
�
2
phate (pH = 8; ionic strength = 10 M).
The experiments were carried out in a four-liters cylind-
rical batch reactor surrounded with an 208C isothermal
jacket. For the H O /UV oxidation process, this reactor
2 2
(internal diameter = 150 mm) was equipped with a low
Á
with hydroperoxide ions to yield OH radicals.
pressure mercury vapor lamp (Hanau NN 15/20) placed in
axial position in a Suprasil silica sleeve allowing an annu-
lar pathlength of 6.21 cm. This lamp emits a monochro-
matic radiation at 253.7 nm, and, at this wavelength, the
incident photonic ¯ux determined by actinometry with
hydrogen peroxide (Nicole et al., 1990) was found to be
Analytical methods: reactants
Glycine, in the range 0.05±1 mM, was analyzed by
HPLC with a spectro¯uorimetric detection (Detector F-
1050 Merck) after a ®fth dilution and derivatization with
orthophtaldialdehyde (O.P.A.). A 20 ml-sample prepared
�
6
�1
.
I
0
=5.9Â 10 Einstein s
with an automatic injector (AS 4000 Merck) was injected
Ê
Several experiments have been carried out in oxygen- on a C18 column (Delta Pak C18 Waters 5 mm, 100 A,
free solutions. These conditions were obtained by continu- 3.9Â 150 mm) eluted with methanol and water (30/70)
ous bubbling of nitrogen through a fritted glass ®tting at
the bottom of the reactor.
(Dossier-Berne et al., 1994).
Quanti®cation of ozone and hydrogen peroxide: For the
The same reactor was used for the study of the reactions analysis of hydrogen peroxide, the colorimetric method
involved during the O /H oxidation process. Ozone
previously described (Karpel Vel Leitner and Dore, 1994)
3
2
O
2
Â
gas, supplied by an oxygen ozonizer, was introduced from
the bottom of the reactor through two diametrally oppo-
site fritted glasses to improve ozone transfer in the aqu-
eous phase. For the applied ozone ¯ow rate of 15.7 mg
with titanium chloride as reagent was used. Dissolved
ozone was dosed according to the indigo trisulfonate
method (Bader and Hoigne, 1982). For the determination
Â
in the reactor's outlet and for the production, the ozone
gas was transferred in aqueous solution containing potass-
�
3
O min , a concentrated hydrogen peroxide aqueous sol-
1
1
�
ution ([H O
2 2
] = 0.102 mol l ) was introduced continu- ium iodide and then, iodine formed was quanti®ed using
sodium thiosulfate (International Ozone Association,
1
987).
Table 1. Calculated competition values for ozonation with hydro-
gen peroxide at pH 8 (phosphate bu€er)
Oxidation byproducts
�
1
�1
s )
Reaction
k (l mol
Experiment 1 Experiment 2
Organic acids were measured by HPLC. The column
used was either an ionic column (Supelcogel C-610H)
eluted with phosphoric acid in water (pH 2) or a C18 col-
umn (Nucleosil) with a solution of acetonitrile and n-octy-
lamine in water as eluent at pH 6.8. The detectors used
were a UV detector set at 210 nm (LDC Spectromonitor
3100) or a photodiode array detector (PDA 996 Waters)
connected to a software system (Millenium 2010 Waters).
This later detection mode allowed byproducts con®r-
mation by comparison of retention times and UV spectra
with standards.
ꢀ
O=Ogly†O3
a
b
O
O
O
3
3
3
+Gly
+Gly
+HO
32.6
5 b
1
160.2
45.7
1
�
a
2.08 Â 10
160.2
3 d
�
6 c
d
2
5.5 Â 10
2 Â 10
ꢀ
O=Ogly†OH
Á
Á
Á
Á
Á
7 e
9 e
9 e
7 e
OH +Gly
OH +Gly
1.7 Â 10
1.9 Â 10
7.5 Â 10
2.7 Â 10
1
2.8
1
2.8
d
5.3
�
�
d
OH +HO
OH +H
2
0.11
1.9
d
d
2
O
2
95
�
1
�1
Experiment 1: QH2 O2 =9.4 mmol h , QO =18.9 mmol h
.
,
Anions and cations were dosed by a capillary electro-
phoresis apparatus (C.I.A. Waters) equipped with an auto-
matic injector and a data acquisition system (Millenium
2010 software Waters). For nitrite and nitrate ions, the
3
�
1
Experiment 2: [H
2
O
.
]
2 o
=75.5 mM, QH2 O2 =9.4 mmol h
QO =
3
�
8.4 mmol h
+
1
1
a
�
�
�
Gly: H
3
N
±CH
2
±COO ; Gly : H
2
N±CH
2
±COO .
b
c
From Pryor et al. (1984).
From Staehelin and Hoigne
�
electrolyte is made of OFM±OH (Waters) and sulfate
Â
(1982).
d
ions. The run potential was set on 12 kV with a migration
time of 50 s in hydrostatic mode. For ammonium ions de-
termination, the samples were eluted with the UV CAT 3
Value calculated from an average concentration of hydrogen per-
oxide.
From Buxton et al. (1988).
e

Radical induced oxidation of glycine
Waters) electrolyte at a ®xed pH value of 4.8. The mi-
435
(
RESULTS
gration time was 30 s with a run potential of 20 kV in the
hydrostatic mode and with a run sample of 5 kV in the
electromigration mode. Chloride ions are used as internal Oxidation of glycine with H O /UV process. In¯u-
2
2
standard for anions and lithium ions for cations. The
interference created by carbonate ions was removed by
treating the samples with a cationic resin (Dowex HCR-S)
before analysis.
Total and inorganic nitrogen in solutions were deter-
mined with a nitrogen analyzer Dorhmann DN 1900
based on chemiluminescence detection. Nitrates and
nitrites are converted to NO by chemical reduction using
ence of oxygen
Experiments have been carried out in the pre-
sence and in the absence of dissolved oxygen with
initial concentrations of hydrogen peroxide and gly-
cine equal to 1 mM. Figure 1 represents the results
concerning the evolution of nitrogen compounds
3
VCl as reducer at 808C in the presence of concentrated during irradiation. The photolysis of glycine by
sulfuric acid. This part constitutes the inorganic nitrogen
fraction. For organics, nitrogen is transformed to NO by
catalytic oxidation at 8508C in the presence of oxygen.
Thus total nitrogen is obtained by injecting the sample in
direct UV radiation was negligible under our exper-
imental conditions.
The action of hydroxyl radicals induced by the
the total nitrogen port so that the sample is ®rst oxidized photolysis of hydrogen peroxide leads to the for-
by the catalyst and then reduced by vanadium chloride. mation of ammonium ions as the major nitrogen
2
Finally, the reaction of ozone with NO leads to NO * in
byproduct. Neither nitrite nor nitrate ions were
formed. This was con®rmed by inorganic nitrogen
measurements which reveal concentrations close to
the detection limit of the analyzer (15 mM). In oxy-
an excited state which emits photons that are detected by
a photomultiplier. Note that according to this procedure,
ammonium ions are not considered as `inorganic nitro-
gen'.
Fig. 1. Evolution of nitrogen byproducts from oxidation of 1 mM glycine with 1 mM hydrogen per-
oxide (H
2
2
O /UV process). (a) In the presence of oxygen; (b) in oxygen-free solution (NT: total nitrogen,
NI: inorganic nitrogen).

4
36
P. Berger et al.
Fig. 2. Evolution of organic byproducts from oxidation of 1 mM glycine with 1 mM hydrogen peroxide
2 2
(H O /UV process). (a) In the presence of oxygen; (b) in oxygen-free solution.
gen-free solutions, the reaction leads to 1 mol of Ozonation of glycine
+
NH
4
/mol of glycine decomposed (Fig. 1(b))
3 2 2
Ozonation with hydrogen peroxide (O /H O
+
4
whereas only 0.75 mol of NH
/mol of glycine was
process). Perozonation experiments have been
carried out under two di€erent conditions of hydro-
gen peroxide concentration. In both cases ozone
and hydrogen peroxide have been introduced con-
tinuously. Table 1 (experimental procedure) sum-
marizes the contribution of the di€erent species for
the two experimental conditions used in this study.
During the ®rst experiment, the two reactives
formed when dissolved oxygen was present. So, in
the presence of oxygen, a small amount of nitrogen
byproducts of glycine (0.25 mol/mol of glycine
decomposed) are included in organic structures.
Organic byproducts of glycine have been tenta-
tively identi®ed by HPLC connected to the photo-
diode array detector. In the presence of oxygen, as
shown in Fig. 2(a), the oxidation leads to one
major byproduct assumed to be oxalic acid accord-
ing to its UV spectrum. Two other acids are
detected in small quantities: formic and oxamic
(
(
O
3
and were introduced in
[O ]/[H O ]) equal to 2 mol/mol with a
2 introduced
H
2
O
2
)
a ratio
3
2
zero initial concentration of hydrogen peroxide in
solution. For this experiment, Fig. 3 represents the
2
acids (oxamic acid: H N±CO±COOH) which have evolution of glycine and nitrogen inorganic bypro-
been spectrally identi®ed. In oxygen-free solutions, ducts. Results indicate that under these conditions
only one organic acid assumed to be oxamic acid is the nitrogen mass balance is complete. The de-
observed by HPLC. Figure 2(b) presents the evol- composition of glycine yields ammonia as the major
ution of this byproduct concentration. Furthermore inorganic nitrogen species and also nitrite and
results indicate that in the presence of oxygen the nitrate ions in equal amounts. Nitrite and nitrate
photolysis of hydrogen peroxide yields a better ions each account for 20% of the glycine removed
removal of glycine and a smaller H
composition than in oxygen-free solutions.
2
O
2
global de- whereas the yield of ammonia reaches 60% of the
amino acid decomposed.

Radical induced oxidation of glycine
437
3 2 2
Fig. 3. Evolution of the amine group from oxidation of 0.98 mM glycine with the O /H O process.
The second experiment was performed to pro- formic and oxalic acid are formed during molecular
mote the production of hydroxyl radicals (Table 1). ozonation. However, the concentration of these two
Thus hydrogen peroxide was introduced initially in later compounds reaches maximum values of 55
the solution ([H O ] =75.5 mM) and also added and 40 mM respectively i.e. only 14% and 4% of
2
2 0
continuously during ozonation.
glycine decomposed. Under these conditions, the
Under this condition, the ozone dose required for consumption of 1 mmol of glycine required about
a given abatement of glycine is high owing to com- 4.2 mmol of ozone and no ozone residual was
Á
petitive reactions of H
2
O
2
for OH radicals and detected in solution during this experiment.
ozone. The removal of glycine does not yield the
formation of nitrite nor nitrate ions. Ammonium
ions, which accounted for 22% of the oxidized gly-
Discussion
Hydroxyl radicals induced oxidation. It is useful
cine, were the only inorganic nitrogen byproduct to note that similar results were obtained relative to
observed (Fig. 4). Oxamic acid was detected in the nature of the inorganic nitrogen species formed
small amounts (<1 mM) but neither oxalic acid nor on one hand with the H O /UV system and on the
2
2
formic acid were present. In addition to oxamic other hand with the O /H O oxidation process in
acid, other nitrogen organic compounds must be the presence of a high concentration of H O . This
3
2
2
2
2
produced.
con®rms the radical nature of the oxidizing species
Molecular ozonation. The study of the action of responsible for glycine decomposition in this later
molecular ozone in the presence of hydrogenocar- process. In oxygen-free solution, the hydroxyl rad-
bonate ions in solution shows that the oxidation of ical resulting from the photolysis of hydrogen per-
glycine leads mainly to nitrate ions (Fig. 5) without oxide leads to quantitative conversion of nitrogen
formation of ammonia. Oxamic acid but above all containing glycine into ammonia. When oxygen is
Fig. 4. Evolution of nitrogen compounds from oxidation of 1.25 mM glycine with the O
in the presence of a 75.5 mM hydrogen peroxide initial concentration.
3 2 2
/H O system

4
38
P. Berger et al.
Fig. 5. Evolution of the amine group and organic byproducts during ozonation of 1.1 mM glycine in
the presence of 0.2 M hydrogenocarbonate ions. (a) Nitrogen compounds; (b) organic compounds.
present, ammonia, which is still the only inorganic dation produces nitrites and nitrates which are not
species detected, is produced in smaller amounts in- observed from H /UV and O /H experiments.
2
O
2
3
2 2
O
dicating the formation of nitrogen organic bypro-
ducts.
When the amino group is unprotonated, electron
abstraction from the amine nitrogen followed by re-
Hydrogen abstraction is believed to be the com- arrangement with decarboxylation to form the
strongly reducing a-aminoradical is another usual
proposed pathway (Monig et al., 1985),
mon initial step for hydroxyl radical reactions.
According to the literature, hydrogen abstraction
can occur either on amine nitrogen or on a C±H
bonds.
Considering the attack of the amine nitrogen, in
the presence of oxygen in solution, peroxyl radicals
are formed. Their decomposition into nitroparan,
ꢀ
4†
oxime and substituted hydroxylamine from primary However, under the pH conditions of this study,
amines was shown to occur (Jayson et al., 1955).
even if the anionic form of the glycine molecule is
not negligible, the presence of two hydrogen atoms
in a of the amino group is unfavorable to decarboxy-
lation reactions (Monig et al., 1985). Thus, the for-
mation of formylamine byproducts seems unlikely.
Other reactions, more probable, consist of an
:
2
R±CH2 ±NHO2 ��� 4R±CH2 ±NO2 ‡ R±CH.NOH
‡
H2O
ꢀ1†
or
:
:
2
R±CH2 ±NHO2 ��� 42R±CH2 ±NHO ‡ O2
ꢀ2† hydrogen abstraction on the a C-H bonds (Neta et
al., 1970), particularly when the amino group is
positively charged (Simic et al., 1971).
:
2
R±CH2 ±NHO ��� 4R±CH.NOH
Á
The main pathway relative to OH induced oxi-
dation of glycine would then lead to ammonia and
‡
R±CH2 ±NHOH: ꢀ3†
However, the presence of such compounds seems glyoxylic acid according to reactions proposed in
unprobable under our conditions since their oxi- the literature (Weeks and Garrison, 1958) (Fig. 6).

Radical induced oxidation of glycine
439
Fig. 6. Two di€erent pathways for the radical oxidation of glycine in the presence and in the absence
of oxygen (Weeks and Garrison, 1958).
Fig. 7. Mechanism for the reaction of glycine with molecular ozone according to the results obtained in
this study and based on some reaction pathways proposed by Laplanche and Martin (1991) and Bailey
(1982). (A) Ozonide form intermediates; (B) electrophilic attack on nitrogen; (I) oxidation on the
carbon with or without decarboxylation; (II) dehydratation; (V) oxidation on N with or without
decarboxylation; (VI) oxidation on N; (VII) oxidation on the carbon; (IX) dehydratation.

4
40
P. Berger et al.
In the presence of oxygen, peroxyl radicals are di€erent reactions involved has been studied else-
formed (Abramovitch and Rabani, 1976). where (Berger and Karpel Vel Leitner, 1998b).
Simultaneously to imino acetic acid formation, the
Molecular ozonation. Concerning the ozonation
hydroperoxyl radical release followed by H
pro- experiments, decreasing hydrogen peroxide promot-
2
O
2
duction can explain the weak evolution of this reac- ing contribution gives rise to nitrite and nitrate ions
tive under such conditions. Thus, after a 200 min appearance. When molecular ozone reactions are
irradiation time, the global quantum yield for favored (i.e. when hydrogen peroxide is decreased
hydrogen peroxide photolysis corresponding to the and radical scavengers such as carbonates are
experiments carried out in the presence and in the added), the yield of nitrates increases whereas
ammonia proportion decreases.
In literature (Elmghari-Tabib et al., 1982; Bailey,
absence of dissolved oxygen averaged respectively
0
.5 and 0.8.
1
982; Laplanche et al., 1985), reaction sequences of
2
The formation of oxamic acid (H N±CO±
the attack of molecular ozone on amines were pro-
posed, based on organic byproducts identi®cation.
It was therefore suggested that the breakdown of
the carbon±nitrogen bond and the formation of
inorganic nitrogen would occur after direct action
of ozone on nitrogen leading to production of
hydroxylamine, oxime or amine oxide, and by an
oxidation of the carbon in a position leading to
amides and formylamine. All these suggested reac-
tions were shown to involve decarboxylation.
COOH) identi®ed in this work also indicates that a
minor route (Berger and Karpel Vel Leitner, 1998a)
can consist in the oxidation of the carbon in a pos-
ition into a carbonyl group after hydrogen abstrac-
tion. This reaction could involve the organic
peroxyl radical formation which decomposes into
the oxamic acid in accordance with mechanism pro-
posed by Von Sonntag and Schuchmann (1991) for
the oxidation of unsubstituted aliphatic acids. This
mechanism would explain the formation of other
In the case of glycine, the intermediate resulting
from ozone addition would either release oxygen to
give monohydroxylamine (Fig. 7; route B then IV
and V) or yield oxidation of the carbon in a (route
A) via a probable ozonide form intermediate as
suggested by Laplanche and Martin (1991) for
amino acids. However, oxamic and oxalic acids
observed in our study indicate that these reactions
can occur without decarboxylation. Therefore as
presented in Fig. 7 with R = ±COOH, monohy-
droxylamine, dihydroxylamine and oximes with two
carbon atoms can be intermediates in the formation
of the byproducts identi®ed in this study.
unidenti®ed nitrogen compounds and also the stoi-
+
chiometry of NH production (only 0.75 mol of
4
+
4
NH formed/mol of glycine decomposed).
Oxalic acid identi®ed in this study resulting from
a reaction mechanism involving deamination before
decarboxylation could arise from the C±N bond
cleavage of the oxamic acid molecule. But following
the hydrolysis of imino acetic acid, the oxidation of
glyoxylic acid by hydroxyl radicals is a more prob-
able decomposition pathway (Fig. 6). Formic acid
which was also detected could therefore originate
from the reaction of glyoxylic acid with residual
hydrogen peroxide under these pH conditions.
Under oxygen-free conditions neither oxalic acid
nor formic acid formation, which requires oxygen,
was observed. In this case, the low decomposition
of glycine is due to the regeneration of glycine
when imino acetic acid is formed.
Since the reaction of molecular ozone with
�1 �1
ammonia is very slow (kO =NH =20 l mol
s ;
Hoigne et al., 1985) it can not become oxidized sig-
Â
3
3
ni®cantly under our experimental conditions (at pH
�1 �1
8
, kO3=NH3 ±NH 4
‡
=1 l mol
s ; Hoigne et al., 1985).
Â
It means that the observed nitrates result from the
oxidation of the nitrogen atom of glycine before the
C±N bond cleavage occurs. Thus, oxamic acid, the
only nitrogen byproduct observed, resulting from
the oxidation of the carbon a of glycine would yield
nitrates or nitrites which react rapidly with ozone
The whole reactions involved in the decompo-
sition pathway of glycine by hydroxyl radicals leads
to ammonia as observed by Le Lacheur and Glaze
+
(
4
1996) for serine. At pH 8, the form NH , unreac-
tive towards hydroxyl radicals, predominates over
NH . Furthermore, the reaction rate constant of
5
�1 �1
3
(
kO =NO =3.7 Â10 l mol
�
s ; Haag et al., 1984)
3
2
Á
OH
with
NH3
being
relatively
small
(Hoigne and
to form nitrates. Like oxamic acid, the unidenti®ed
nitrogen organic intermediates are ®nally oxidized
into nitrates.
7
�1 �1
(
kOHÁ=NH =1.7±8.7Â 10 l mol
Â
Bader, 1978), no reaction with NH ) many solutes,
s
3
+
4
even in low concentrations, may protect ammonia
from oxidation due to competitive reactions
towards hydroxyl radicals. Neither nitrites nor
nitrates are formed from hydroxyl radical oxidation
CONCLUSION
The study of the oxidation of glycine showed
of glycine. Total nitrogen analysis shows a constant that the nature of the inorganic nitrogen byproducts
content in solution.
di€ers between ozonation and hydroxyl radicals
The presence of high concentrations of hydrogen induced reactions. The action of molecular ozone
peroxide (i.e. 75.5 mM in our experiment) induces a which leads to nitrates demonstrated that even
low production of ammonium ions with a constant when the carbon atom is ®rstly oxidized, the C±N
nitrogen balance. The role of this compound in the bond cleavage never occurs before nitrogen oxi-

Radical induced oxidation of glycine
441
inorganic compounds in water ± III. Inorganic com-
pounds and radicals. Water Res. 19(8), 993±1004.
Hubberten U., Lara R. J. and Kattner G. (1994) Amino
acid composition of seawater and dissolved humic sub-
stances in the Greenland Sea. Mar. Chem. 45, 121±128.
dation. In the case of hydroxyl radicals, the attack
on the amino group of the molecule seems unfa-
vored at neutral pH. Hydrogen abstraction on the
a-carbon, followed by deamination and ammonia
production, is assumed to involve reactions with International Ozone Association (1987) Iodometric
method for the determination of ozone in a process gas.
IOA Standardization Committee ± Europe ± 001/87 (F).
ozonation on one hand and of ammonia from hy-
oxygen. The formation of nitrates from molecular
Jayson G. G., Scholes G. and Weiss J. (1955) Chemical
droxyl radical processes on the other hand could be
used in ozonation processes as a mean of evaluating
the hydroxyl radicals participation compared with
ozone.
action of ionising radiations in solution. Part XIV. The
action of X-rays (200 kV) on some aliphatic amines in
aqueous solution with particular reference to the for-
mation of oximes. J. Chem. Soc. 2594±2600.
Karpel Vel Leitner N. and Dore
gene dissous dans le mecanisme de de
l'acide formique en solution aqueuse par irradiation UV
en presence de peroxyde d'hydrogene. J. Chim. Phys.
1, 503±518.
Abramovitch S. D. and Rabani J. (1976) Pulse radiolytic Laplanche A. and Martin G. (1991) Action de l'ozone sur
Â
M. (1994) Roà le de l'oxy-
Á
Â
Â
composition de
Â
Á
REFERENCES
9
investigations of peroxy radicals in aqueous solutions of
acetate and glycine. J. Phys. Chem. 80(14), 1562±1565.
Â
Bader H. and Hoigne J. (1982) Determination of ozone in
water by the indigo method; a submitted standard
method. Ozone Sci. Eng. 4, 169±176.
les composes azotes moleculaires. In Ozone et Ozonation
  Â
des eaux, coordinator W.J. Masschelein. Tec. et Doc.-
Ed. Lavoisier.
Laplanche A., Le Cloirec C. and Martin G. (1985)
Ozonation de compose
ferents oxydants sur la phenylalanine. Proceeding
Wasser Berlin 69±93.
s azotes. Action comparee de dif-
  Â
Bailey P. S. (1982) Ozonation of nucleophiles. In
Ozonation in Organic Chemistry, ed. H. H. Wasserman,
Â
Vol. II, ch. VII, pp. 155±201. Academic Press Inc., New Le Lacheur R. M. and Glaze W. H. (1996) Reactions of
York.
ozone and hydroxyl radicals with serine. Environ. Sci.
Technol. 30, 1072±1080.
Lytle C. R. and Perdue E. M. (1981) Free, proteinaceous,
and humic-bound amino acids in river water containing
high concentrations of aquatic humus. Environ. Sci.
Technol. 15(2), 224±228.
Berger P. and Karpel Vel Leitner N. (1998a) Hydroxyl
radicals oxidation of the amino group depending on the
oxidation degree of the a-carbon. Environ. Sci. Technol.,
submitted.
Berger P. and Karpel Vel Leitner N. (1998b) Irradiation
of aqueous solution of glycine: H
radiolysis. J. Chim. Phys. submitted.
2
O
2
/UV process and g Malcolm R. L. (1990) The uniqueness of humic substances
in each soil, stream and marine environments. Anal.
Buxton G. V., Greenstock C. L., Helman W. P. and Ross
A. B. (1988) Critical review of rate constants for reac-
tions of hydrated electrons, hydrogen atoms and hy-
droxyl radicals (OH /O ) in aqueous solution. J. Phys.
Chem. Ref. Data 17(2), 513±886.
Chim. Acta 232, 19±30.
McKnight D. M., Aiken G. R. and Smith R. L. (1991)
Aquatic fulvic acids in microbially based ecosystems:
results from two desert lakes in Antartica. Limnol.
Oceanogr. 36(5), 998±1006.
Á
Á�
Chuboda J., Hejzlar J. and Dolezal M. (1986) Microbial Monig J., Chapman R. and Asmus K. D. (1985) E€ect of
Á
polymers in the aquatic enviroment ± III. Isolation from
river, potable and underground water and analysis.
Water Res. 20(10), 1223±1227.
the protonation state of the amino group on the OH
radical induced decarboxylation of amino acids in aqu-
eous solution. J. Phys. Chem. 89, 3139±3144.
Dossier-Berne F., Panais B., Merlet N., Cauchi B. and
Â
Legube B. (1994) Analyse des acides amines dissous
Á
totaux dans les eaux a potabiliser en cours de traite-
ment. Environ. Technol. 15, 901±916.
Neta P., Simic M. and Hayon E. (1970) Pulse radiolysis of
aliphatic acids in aqueous solution. III. Simple amino
acids. J. Phys. Chem. 74(6), 1214±1220.
Nicole I., De Laat J., Dore M., Duguet J. P. and Bonnel
Â
Duguet J. P., Jaulin L., Aurelle Y., Roques H. and
Legeron J. P. (1980) Experimental study of ozone action
Â
on glycine. Ozone Sci. Eng. 2, 105±122.
C. (1990) Use of UV radiation in water treatment:
measurement of photonic ¯ux by hydrogen peroxide
actinometry. Water Res. 24(2), 157±168.
Duguet J. P., Brodard E., Dussert B. and Mallevialle
J. (1985) Improvement in the e€ectiveness of ozonation
of drinking water through the use of hydrogen peroxide.
Ozone Sci. Eng. 7, 241±258.
Pryor W. A., Giamalva D. H. and Church D. F. (1984)
Kinetics of ozonation. 2. Amino acids and model com-
pounds in water and comparisons to rates in nonpolar
solvents. J. Am. Chem. Soc. 106, 7094±7100.
Elmghari-Tabib M., Laplanche A., Venien F. and Martin Simic M., Neta P. and Hayon E. (1971) Pulse radiolytic
G. (1982) Ozonation des amines dans l'eau. Water Res.
6, 223±229.
Haag W. R., Hoigne
investigation of aliphatic amines in aqueous solution.
Int. J. Radiat. Phys. Chem. 3, 309±320.
Â
J. and Bader H. (1984) Improved Staehelin J. and Hoigne J. (1982) Decomposition of ozone
1
Â
ammonia oxidation by ozone in the presence of bromide
ion during water treatment. Water Res. 18(9), 1125±
in water: rate of initiation by hydroxide ions and hydro-
gen peroxide. Environ. Sci. Technol. 16(10), 676±681.
Von Sonntag C. and Schuchmann H. P. (1991) The eluci-
dation of peroxyl radical reactions in aqueous solution
with the help of radiation-chemical methods. Angew.
Chem. Int. Ed. Engl. 30, 1229±1253.
1128.
Hoigne
Â
J. and Bader H. (1978) Ozonation of water: kin-
etics of oxidation of ammonia by ozone and hydroxyl
radicals. Environ. Sci. Technol. 12(1), 79±84.
Hoigne
Â
J., Bader H., Haag W. R. and Staehelin J. (1985)
Rate constants of reactions of ozone with organic and
Weeks B. M. and Garrison W. M. (1958) Radiolysis of
aqueous solutions of glycine. Radiat. Res. 9, 291±304.