Advanced oxidation chemistry of paracetamol. UV/H(2)O(2)-induced hydroxylation/degradation pathways and (15)N-aided inventory of nitrogenous breakdown products.

Article abstract of DOI:10.1021/jo025604v

The advanced oxidation chemistry of the antipyretic drug paracetamol (1) with the UV/H(2)O(2) system was investigated by an integrated methodology based on (15)N-labeling and GC-MS, HPLC, and 2D (1)H, (13)C, and (15)N NMR analysis. Main degradation pathways derived from three hydroxylation steps, leading to 1,4-hydroquinone/1,4-benzoquinone, 4-acetylaminocatechol and, to a much lesser extent, 4-acetylaminoresorcine. Oxidation of the primary aromatic intermediates, viz. 4-acetylaminocatechol, 1,4-hydroquinone, 1,4-benzoquinone, and 1,2,4-benzenetriol, resulted in a series of nitrogenous and non-nitrogenous degradation products. The former included N-acetylglyoxylamide, acetylaminomalonic acid, acetylaminohydroxymalonic acid, acetylaminomaleic acid, diastereoisomeric 2-acetylamino-3-hydroxybutanedioic acids, 2-acetylaminobutenedioic acid, 3-acetylamino-4-hydroxy-2-pentenedioic acid, and 2,4-dihydroxy-3-acetylamino-2-pentenedioic acid, as well as two muconic and hydroxymuconic acid derivatives. (15)N NMR spectra revealed the accumulation since the early stages of substantial amounts of acetamide and oxalic acid monoamide. These results provide the first insight into the advanced oxidation chemistry of a 4-aminophenol derivative by the UV/H(2)O(2) system, and highlight the investigative potential of integrated GC-MS/NMR methodologies based on (15)N-labeling to track degradation pathways of nitrogenous species.

Full text of DOI:10.1021/jo025604v

Ad va n ced Oxid a tion Ch em istr y of P a r a ceta m ol. UV/H2O2-In d u ced
Hyd r oxyla tion /Degr a d a tion P a th w a ys a n d 15N-Aid ed In ven tor y of
Nitr ogen ou s Br ea k d ow n P r od u cts.
Davide Vogna,† Raffaele Marotta,‡ Alessandra Napolitano,† and Marco d’Ischia*,†
Department of Organic Chemistry and Biochemistry, University of Naples “Federico II”, Via Cinthia 4,
I-80126 Naples, Italy, and Department of Chemical Engineering, University of Naples “Federico II”,
Piazzale Tecchio, I-80126 Naples, Italy
dischia@cds.unina.it
Received February 8, 2002
The advanced oxidation chemistry of the antipyretic drug paracetamol (1) with the UV/H2O2 system
was investigated by an integrated methodology based on 15N-labeling and GC-MS, HPLC, and 2D
1H, 13C, and 15N NMR analysis. Main degradation pathways derived from three hydroxylation steps,
leading to 1,4-hydroquinone/1,4-benzoquinone, 4-acetylaminocatechol and, to a much lesser extent,
4-acetylaminoresorcine. Oxidation of the primary aromatic intermediates, viz. 4-acetylaminocatechol,
1,4-hydroquinone, 1,4-benzoquinone, and 1,2,4-benzenetriol, resulted in a series of nitrogenous and
non-nitrogenous degradation products. The former included N-acetylglyoxylamide, acetylaminoma-
lonic acid, acetylaminohydroxymalonic acid, acetylaminomaleic acid, diastereoisomeric 2-acetyl-
amino-3-hydroxybutanedioic acids, 2-acetylaminobutenedioic acid, 3-acetylamino-4-hydroxy-2-
pentenedioic acid, and 2,4-dihydroxy-3-acetylamino-2-pentenedioic acid, as well as two muconic
and hydroxymuconic acid derivatives. 15N NMR spectra revealed the accumulation since the early
stages of substantial amounts of acetamide and oxalic acid monoamide. These results provide the
first insight into the advanced oxidation chemistry of a 4-aminophenol derivative by the UV/H2O2
system, and highlight the investigative potential of integrated GC-MS/NMR methodologies based
on 15N-labeling to track degradation pathways of nitrogenous species.
In tr od u ction
by the need to improve existing methodologies for the
routine characterization of complex product mixtures in
toto; to establish spectra libraries of degradation products
for analytical purposes; and to verify the possible genera-
tion of toxic intermediates along the mineralization
pathway more hazardous than the parent pollutants.
Among the various organic pollutants that are cur-
rently under scrutiny as substrates for AOP, phenolic
compounds occupy a prominent position.4-7 UV/H2O2
treatment of phenol was reported to give as main
intermediates catechol, 1,4-hydroquinone, and a series
of dimeric species, such as phenoxyphenols, which even-
tually decomposed to organic acids such as muconic,
maleic, fumaric, D,L-malic, oxalic, and formic acid.8,9
Salicylic acid behaved similarly to afford catechol and 2,3-
dihydroxybenzoic acid, which in turn give rise to the same
pattern of carboxylic acids as phenol.8
Advanced Oxidation Processes (AOP) comprise a range
of water reclamation strategies aimed at achieving
complete mineralization of organic pollutants to CO2 and
H2O by the action of highly reactive oxygen species
generated by different techniques, including mainly
photochemical systems (UV/H2O2, photoassisted Fenton,
1-3
TiO2-photocatalyzed, UV/O3) and ozonation.
Whereas the applications of AOP to recalcitrant aro-
matic pollutants have expanded dramatically in recent
years, spurring extensive kinetic studies aimed at opti-
mizing operational parameters, considerable lacunae
remain as to the degradation mechanisms and the
structures, origin, and fate of ring-opened acyclic prod-
ucts. Insights into this relatively unexplored segment of
organic chemistry are hurdled by the need to contend
with bewilderingly complex patterns of products at
various degrees of oxygenation and degradation, often
present as intimate mixtures of highly polar stereo- and
regioisomeric species. Yet, pursuit of these issues is urged
Photocatalytic degradation of 1,4-hydroquinone was
found to yield a collection of oxocarboxylic acids via
hydroxy-1,4-benzoquinone and 1,2,4-benzenetriol.10
(4) Andreozzi, R.; Marotta, R. J . Hazard. Mater. 1999, 69, 303-
317.
† Department of Organic Chemistry and Biochemistry.
‡ Department of Chemical Engineering.
(1) Andreozzi, R.; Caprio, V.; Insola, A.; Marotta, R. Catal. Today
1999, 53, 51-59.
(2) Pelizzetti, E.; Serpone, N. Homogeneous and Heterogeneous
Photocatalysis; D. Reidel Publishing Company: Boston, 1986.
(3) Legrini, O.; Oliveros, E.; Braun, A. M. Chem. Rev. 1993, 1446-
1454.
(5) Esplugas, S.; Gime´nez, J .; Contreras, S.; Pascual, E.; Rodriguez,
M. Water Res. 2002, 36, 1034-1042.
(6) Lunar, L.; Sicilia, D.; Rubio, S.; Perez-Bendito, D.; Nickel, U.
Water Res. 2000, 34, 3400-3412.
(7) Andreozzi, R.; Caprio, V.; Insola, A.; Marotta, R. Water Res. 2000,
34, 463-472.
(8) Scheck, C. K.; Frimmel, F. H. Water Res. 1995, 29, 2346-2352.
(9) Huang, C. R.; Shu, H. Y. J . Hazard. Mater. 1995, 41, 47-67.
10.1021/jo025604v CCC: $22.00 © 2002 American Chemical Society
Published on Web 07/26/2002
J . Org. Chem. 2002, 67, 6143-6151
6143

Vogna et al.
The chemical steps in the photocatalytic degradation
of 4-chlorophenol and 4-chlorocatechol have also been
delineated and a number of aldehydes and oxocarboxylic
acids have been identified, mainly by mass spectrometric
analysis.11
Very little is known about the advanced oxidation
chemistry of aminophenol derivatives.6,12 Among these
latter, paracetamol (acetaminophen, 4-acetylaminophe-
nol, 1), a widely used analgesic and antipyretic drug, and
an important material for the manufacturing of azo dyes
and photographic chemicals,13 has recently attracted
interest as a potential component of contaminated wa-
ters. Levels of 1 of up to 6 µg/L have been detected in
effluent waters from sewage treatment plants.14
Because of its hepatotoxicity15 and other side effects,
concern has been raised about the possible environmental
impact of 1 and its (bio)degradation products, which may
be toxic or hazardous even in trace amounts.
Although the mode of polymerization of 1, e.g. by the
peroxidase/H2O2 system, is established, 16,17 the oxidative
degradation pathways have remained a gap in the
chemistry of 1.
In this paper we report a survey of the advanced
oxidation chemistry of 1 and related compounds with the
UV/H2O2 system. By means of an integrated GC-MS and
NMR approach based on 15N labeling, it was possible to
elucidate the competing hydroxylation/degradation path-
ways of 1 and to track specifically the fate of nitrogenous
products.
F IGURE 1. Representative GC trace for the UV/H2O2 deg-
radation reaction of 1 as obtained at 20 min reaction time.
Main carboxylic acids identified (as silyl derivatives): hy-
droxyacetic (glycolic) acid (1), oxalic acid (2), malonic acid (3),
oxaloacetic acid (4), glyoxylic acid (5), maleic and fumaric acids
(6 and 9), succinic acid (7), tartronic (hydroxymalonic) acid (11),
ketomalonic acid (15), malic (hydroxysuccinic) acid (16), tar-
taric acids (19 and 23), R-ketoglutaric acid (22), and hydroxy-
ketosuccinic acid (24). Structural formulations were secured
by comparison of the chromatographic and mass spectrometric
behavior of the products with those of authentic compounds
or library matching exceeding 90%.
MS and a combination of 2D 15N, 1H, and 13C NMR
techniques.
P r od u ct An a lysis: GC-MS. In typical experiments,
reaction of 1 (1 mM) or its 15N-labeled derivative was
carried out with 100 mM H2O2 and UV radiation (254
nm) and GC-MS runs were performed at various inter-
vals of time. A silylation treatment prior to injection into
the gas chromatograph proved essential to allow for
analysis of highly polar polyhydroxylated and carbox-
ylated species. In all cases, complete silylation of both
alcoholic and carboxylic functions was observed. Omission
of the derivatization step resulted in chromatographic
traces of poor analytical value.
About 50% consumption of 1 was observed after 20
min, and this was selected as a suitable reaction time
for investigation of early intermediates and degradation
products thereof. No substantial changes of product
distribution were observed when the UV/H2O2 reaction
was carried in phosphate buffer at pH values in the range
5.5-8.0 rather than in water.
Resu lts a n d Discu ssion
Oxid a t ion Con d it ion s a n d An a lyt ica l Met h od s.
Advanced oxidation processes based on H2O2 with con-
comitant UV irradiation result in the production of OH
radicals18 and other reactive species that engage the
organic pollutants to produce small hydrophilic species
up to the level of inorganic materials (mineralization)
innocuous to human health.1-3 The system produces
complete destruction of organic pollutants resistant to
biological treatments, and is exploited for the cleanup of
wastewaters because of chemical efficiency (high extent
of mineralization) and practical convenience (low costs,
ease of operation).1 The aim of the present study was to
identify as many products as possible and to dissect
competing degradation pathways. To ease the analytical
difficulties associated with the identification of nitrog-
enous species, 15N-labeled 1 was prepared by a straight-
forward route and was subjected to degradation under
conditions identical with those used for the unlabeled
drug. Product characterization was carried out by GC-
Figure 1 shows a representative GC trace for the UV/
H2O2 degradation reaction of 1 as obtained at 20 min
reaction time.
The aromatic compounds that could be identified
included, besides unreacted 1 (peak 29), 4-acetylami-
nocatechol (peak 34, the main component), 1,4-hydro-
quinone (peak 12), 1,2,4-benzenetriol (peak 21), 1,2,4,5-
benzenetetraol (peak 27), and 4-acetylaminoresorcine
(peak 32, identified by straightforward analysis of the
EI-MS spectrum). In addition, a number of carboxylic
acids could be identified (see legend to Figure 1). Sodium
borohydride reduction of the mixture prior to analysis
resulted in only minor differences in GC traces at various
times. In particular, no appreciable change in 1,4-
hydroquinone levels was observed after reduction.
Structural assignments for the main nitrogenous spe-
cies, as deduced by ion trap MS data for both unlabeled
and 15N-labeled 1, are given in Table 1.
(10) Li, X.; Cubbage, J . W.; Tetzlaff, T. A.; J enks, W. S. J . Org. Chem.
1999, 64, 8509-8524.
(11) Li, X.; Cubbage, J . W.; J enks, W. S. J . Org. Chem. 1999, 64,
8525-8536.
(12) Andreozzi, R.; Caprio, V.; Insola, A.; Marotta, R. Water Res.
2000, 34, 463-472.
(13) Rao, E. K.; Sastry, C. S. Mikrochim. Acta 1984, 1, 313-19.
(14) Daughton, C. G.; Ternes, T. A. Water Res. 1998, 32, 3245-3260.
(15) O’Brien, P. J .; Khan, S.; J atoe, S. D. Adv. Exp. Med. Biol. 1991,
283, 51-64.
(16) Potter, D. W.; Miller, D. W.; Hinson, J . A. J . Biol. Chem. 1985,
260, 12174-12180.
(17) Potter, D. W.; Hinson, J . A. J . Biol. Chem. 1987, 262, 966-
973.
(18) Baxendale, J . H.; Wilson, J . A. Trans. Faraday Soc. 1957, 53,
344.
6144 J . Org. Chem., Vol. 67, No. 17, 2002

Advanced Oxidation Chemistry of Paracetamol
TABLE 1. Ma in Nitr ogen ou s P r od u cts in th e Ad va n ced Oxid a tion of [15N]1 Id en tified by GC-MS An a lysis.
Peak 10 was formulated as disilylated N-acetyl gly-
oxylamide hydrate on the basis of a (M - 15)+ peak at
m/z 262, a base peak at m/z 19119 akin to that observed
in the case of glyoxylic acid, and a fragmentation pattern
reminiscent of that of glyoxylic acid (see Supporting
Information).
Peak 25 displayed a molecular ion peak at m/z 391,
and significant fragmentation peaks at m/z 184 (vide
supra) and at m/z 200 (M - 191)+, denoting loss of the
TMSOdC-OTMS radical, consistent with a derivative
of 3-acetylamino-4,4-dihydroxy-2-butenoic acid (the hy-
drated form of 3-acetylamino-4-oxo-2-butenoic acid).
The product eluted under the small peak at tR 12.49
min (peak 14) gave a meaningful fragmentation spectrum
with an intense (M - 15)+ peak at m/z 302, a base peak
at m/z 200, denoting loss of COOTMS radical, and a peak
at m/z 184, taken as diagnostic of the 3-acetylaminopro-
penoic acid moiety. It was thus assigned the structure of
the disilyl derivative of acetylaminomaleic acid. Likewise,
the product for peak 18, one of the most abundant
intermediates along the degradation pathway, was iden-
tified as the trisilyl derivative of acetylaminohydroxy-
malonic acid, whereas peak 17 was regarded as being due
to the disilyl derivative of acetylaminomalonic acid.
Two main nitrogen-containing products at the four-
carbon level of degradation of the aromatic ring eluted
as peaks 28 and 30, and were formulated as diastereoiso-
meric 2-acetylamino-3-hydroxybutanedioic acids. Strong-
ly supportive of this assignment was a peak at m/z 189
characteristic of structures possessing a hydroxyacetic
acid moiety.
Peak 26 was due to a species displaying a (M - CH3)+
fragment at m/z 404, another significant fragmentation
peak at m/z 200 (M - 219)+, suggesting direct loss of the
TMSOCHCOOTMS radical from the molecular radical
cation (m/z 419), and the fragment at m/z 184. Accord-
ingly, it was assigned the structure of 3-N-acetylamino-
4-hydroxy-2-pentenedioic acid.
(19) Spiteller, G.; Spiteller, S. Biochim. Biophys. Acta 1998, 1392,
23-40.
J . Org. Chem, Vol. 67, No. 17, 2002 6145

Vogna et al.
TABLE 2. Selected Heter on u clea r Cor r ela tion Da ta fr om 2D-NMR Exp er im en ts on Cr u d e Mixtu r es Obta in ed by
Oxid a tion of [15N]1 a n d P r op osed Str u ctu r a l Assign m en ts.
compd
no.
1H-15N
HMBC
1H-13C
HMBC
1H
15N
13C
structure
1
1.75 (s)
22.5
110.8
172.9
172.9
acetamidea
oxalamidea
6.63 (d, J ) 86 Hz)
7.26 (d, J ) 86 Hz)
7.74 (d, J ) 88 Hz)
8.04 (d, J ) 88 Hz)
1.95 (s)
6.59 (d, J ) 8.4 Hz)
6.73 (m)
7.11 (dd, J ) 2.0, 2.1 Hz)
8.58 (s)
110.8
110.8
107.5
107.5
2
3
162.4
23.9
115.3
110.4
107.9
133.6
167.7
4-acetylaminocatechola
133.6, 145.0
107.9, 141.2
141.2, 110.4
133.6
133.6
8.94 (s)
9.56 (d, J ) 88 Hz)
133.6
141.2
171.6
4
2.0
24.5
78c
132.5
132.5
3-(acetylamino)-2,4-dihydroxy-
5.65
2-pentenedioic acid
9.62 (d, J ) 88 Hz)
132.5
119.5
116.3
5
6
4.51
72.2
82c
119.5
2-(acetylamino)-2,3-dihydroxy-
butanedioic acid
acetylaminoglyoxylic acid
8.09 (d, J ) 88 Hz)b
5.4 (m)
8.41 (d, J ) 88 Hz)b
a
b
Proton resonances were in good agreement with those of authentic samples. Deduced from the COSY spectrum. c Deduced from
HMQC.
By similar reasoning, the product eluted under peak
31, featuring a (M - CH3)+ peak at m/z 492, a fragment
at m/z 273, conceivably via loss of the TMSOCHCOOTMS
radical, a fragment at m/z 260 suggesting a silylated
2-hydroxy-3-acetylaminopropenoic acid moiety, and the
peak at m/z 189, was formulated as deriving from 2,4-
dihydroxy-3-acetylamino-2-pentenedioic acid.
The remaining two nitrogenous species eluted under
the relatively retained peaks 35 and 33 were evidently
ring-opened products of 1. The former was assigned the
structure of a muconic-type cleavage product, 3-N-acetyl-
amino-2,4-hexadienedioic acid. The latter was apparently
a hydroxylated derivative, exhibiting very intense nitrogen-
containing fragmentation peaks at m/z 328, shared also
by peak 35, and at m/z 232.
HPLC and GC-MS analysis of the oxidation mixtures
after prolonged reaction times showed complete disap-
pearance of primary intermediates and the almost ex-
clusive presence of small fragments, e.g. acetic acid, oxalic
acid, and formic acid with TOC values of about 80%.
Formic acid from 1 mM 1 was less than 0.5 mM at 30
min reaction time and increased up to 1.15 mM after 80
min.
F IGURE 2. 1H,15N HMBC NMR spectrum of the oxidation
mixture of 15N-1 at 50 min reaction time. Letters denote cross-
peaks for identified species: a, acetamide; b, oxalic acid
monoamide; c, 4-acetylaminocatechol; d, 1; and e, imides.
Numbers in the 1H NMR spectrum mark major non-nitrog-
enous compounds: I, succinic acid; II, malonic acid; III, glycolic
acid; and IV, 2-oxo-3-hydroxybutanedioic acid.
P r od u ct Ch a r a cter iza tion : 15N, 1H, a n d 13C NMR.
Aliquots of reaction mixtures obtained by UV/H2O2
oxidation of 15N-labeled 1 were withdrawn at various
intervals of time, carefully lyophilized, and subjected to
15N NMR analysis. At 50 min reaction time the spectrum
revealed clusters of 15N resonances spread over the range
between δ 100 and 170, suggesting significant modifica-
tion of the chemical environment of the amide nitrogen
(δ 133.0 for 1). Proton signals showing well-discernible
cross-peaks with the nitrogen resonances were then
selected (Figure 2) and relevant homonuclear and het-
eronuclear one-bond and multiple-bond connectivities
were deduced on the basis of 1H,1H COSY, 1H,13C HMQC,
Besides some residual 1, 4-acetylaminocatechol ap-
peared prominently in the early stages and gave well-
1
recognizable H-15N correlations (Figure 2, region c). A
noticeable degradation product, which eluded GC-MS
analysis, was acetamide giving an intense nitrogen signal
at δ 110.8 correlating with two proton doublets at δ 6.63
and 7.26 (Figure 2, region a) due to the large 1H-15N
coupling constant and restricted rotation around the
C-N bond.
1
and H,13C HMBC spectra. The main nitrogenous prod-
ucts identified under different reaction conditions by
Another nitrogen signal at δ 107.5 exhibited one-bond
correlation with a pair of doublets at δ 7.74 and 8.04
(Figure 2, region b), suggesting an amide NH2 group. An
1
NMR analysis and the relevant H, 15N, and 13C chemical
shift data are reported in Table 2.
6146 J . Org. Chem., Vol. 67, No. 17, 2002

Advanced Oxidation Chemistry of Paracetamol
appreciable long-range correlation of the doublet at δ 7.74
with a quaternary carbon signal at δ 162.4 and the lack
of detectable cross-peaks in all other 2D correlation
spectra, together with mechanistic considerations, sug-
gested oxalic acid monoamide as a plausible candidate.
1
This conclusion was corroborated by comparison of H
and 13C NMR data with those of an authentic sample.
Two distinct clusters of nitrogen signals were apparent
1
at δ 161-163 and 168-170, which displayed J correla-
tion with an array of doublets between δ 9 and 11 and
long-range couplings with methyl signals around δ 2.2
(region e). These indicated N-H groupings experiencing
a relatively deshielding environment. Such a feature was
compatible with most of the ring breakdown structures
deduced from GC-MS analysis, in which the NH group
was linked to electron-withdrawing carboxyl-bearing
residues such as imides. For comparison, a δ value of 186
was predicted for the 15N resonance of the 3D energy-
minimized molecular structure of acetimide20 as a model
imide.
F IGURE 3. UV light irradiation of 1,4-benzoquinone in the
presence or in the absence of hydrogen peroxide. 1,4-Benzo-
quinone (∆) decay and 1,4-hydroquinone (0) formation by UV
light. 1,4-Benzoquinone (2) decay and 1,4-hydroquinone (9)
formation by UV/H2O2 system.
SCHEME 1
Main non-nitrogenous products identified by 1H,13C
one-bond and multiple-bond correlation were glycolic,
malonic, succinic, and 2-oxo-3-hydroxybutanedioic acid
(Figure 2, proton spectrum). Diastereoisomeric tartaric
acids as well as malic, fumaric, and tartronic acid
(hydroxymalonic acid), though supposedly present, could
not be identified unambiguously because their 1H and
13C signals fell in exceedingly overcrowded spectral
regions.
1,4-Hydroquinone gave mainly malonic acid, 1,2,4-
benzenetriol, tartaric acids, and 1,2,4,5-benzenetetraol,
whereas 1,2,4-benzenetriol afforded initially 1,2,4,5-
benzenetetraol and then malonic acid, tartronic acid,
malic acid, and tartaric acids.
1
Both H and 13C NMR spectra ruled out the presence
of significant levels of aldehyde products in the oxidation
mixture, consistent with the GC-MS data. 1,4-Hydro-
quinone and 1,4-benzoquinone were also below detection
limits.
Oxid a tion Beh a vior of Rea ction In ter m ed ia tes.
To determine the origin of the various breakdown species
identified in the mixture, established or putative aro-
matic intermediates, viz. 1,4-benzoquinone, 1,4-hydro-
quinone, 1,2,4-benzenetriol, and 4-acetylaminocatechol,
as well as 4-aminophenol, were separately subjected to
oxidation with the UV/H2O2 system under the same
conditions.
The behavior of 1,4-benzoquinone toward the UV/H2O2
system is worthy of discussion. At 1 mM concentration,
the compound was readily degraded in an oxygen-
independent manner with partial conversion to 1,4-
hydroquinone. The latter species attained maximum
concentration (ca. 0.5 mM) after about 15 min and then
decayed to trace levels after 40 min (Figure 3). GC-MS
analysis indicated, besides 1,4-hydroquinone as the main
product, small amounts of 1,2,4-benzenetriol, as well as
of malonic, maleic, malic, succinic, fumaric, tartaric (both
meso and D,L diastereoisomers), and 2-ketoglutaric acid.
In the absence of H2O2, UV irradiation of 1,4-benzo-
quinone resulted in a similar decay as in the presence of
H2O2, but under these latter conditions 1,4-hydroquinone
accumulated up to a concentration of ca. 0.5 mM (50%
yield) without substantial decomposition over 60 min. By
contrast, negligible 1,4-benzoquinone decay was observed
by exposure to H2O2 in the absence of UV radiation.
4-Acetylaminocatechol gave mainly malonic, 3-oxobu-
tanoic, tartronic, and tartaric acid (Scheme 1). At 1 mM
concentration, 4-aminophenol disappeared in about 30
min giving rise to some 1,4-hydroquinone, but no detect-
able 1,2,4-benzenetriol, and a range of non-nitrogenous
fragments. A remarkable feature of the GC trace was a
group of products eluted at higher retention times than
4-aminophenol, sharing molecular ion peaks at m/z 642,
or (M - 15)+ peaks at m/z 627. Their mass spectra were
strongly reminiscent of that of mucic acid (galactaric
acid), which, however, was absent in the mixture. These
species were tentatively formulated as isomeric tetrahy-
droxyhexanedioic acids and were not investigated fur-
ther.
Mech a n istic Issu es. There is general consensus that
UV/H2O2-promoted oxidations involve mainly OH radi-
cals produced by photolysis of H2O2.18 However, the
actual situation may be much more complex, due to
several concurrent oxidation pathways, e.g. reaction of
H2O2 with excited states of the substrate and intermedi-
ates, the intervention of (hydro)peroxyl radicals, and so
forth. In fact, under the experimental conditions adopted
in this study UV absorption by H2O2 is not sufficient to
rule out direct photolysis and other OH radical-indepen-
dent photochemical processes.21
Moreover a significant role for atmospheric oxygen was
ruled out by control experiments carried out under an
argon atmosphere, giving similar product patterns. tert-
(20) A value of δ 186 ppm was predicted for the 15N resonance of
(CH3CO)2NH. The experimentally measured 15N chemical shift value
of acetamide was used for estimation of the expected resonances.
(21) ꢀ254nm(1) ) 7420 M-1 cm-1, ꢀ254nm(H2O2) ) 18.6 M-1 cm-1
,
absorbed fraction (H2O2) ) ꢀ254nm(H2O2) × [H2O2]/(ꢀ254nm(H2O2) ×
[H2O2] + ꢀ254nm(1) × [1]) ) 0.20.
J . Org. Chem, Vol. 67, No. 17, 2002 6147

Vogna et al.
SCHEME 3
F IGURE 4. Time course of the decay of 1 and formation of
4-acetylaminocatechol and 1,4-hydroquinone by reaction with
the UV/H2O2 system. Left axis: solid line, decay of 1 (b);
dashed lines, experiments in the presence of 100 (0) or 200
mM (O) tert-butyl alcohol. Right axis: 4-acetylaminocatechol
(() and 1,4-hydroquinone (9).
SCHEME 2
SCHEME 4
Different mechanistic options can be envisaged for OH
radical-dependent para hydroxylation (Scheme 3). One
would involve a direct substitution reaction to give 1,4-
hydroquinone, by a mechanism similar to that reported
for salicylic acid,8 4-hydroxybenzoic acid,23 or 4-chloro-
phenol.11 The other pathway would envisage conversion
of the primary radical adduct to 4-acetylamino-4-hydroxy-
2,5-cyclohexadienone, which would decompose to yield
1,4-benzoquinone and acetamide.
In the third pathway (Scheme 3), oxidation of 1 may
proceed via H-atom abstraction leading to the semi-
quinone-like phenoxyl radical of 1, which would then
react with an OH radical or another oxygen species at
the para position. Alternatively, further oxidation of the
phenoxyl radical might lead to the corresponding qui-
noneimine, which could be susceptible to hydrolysis to
generate 1,4-benzoquinone. Typically, however, H-ab-
straction from a phenoxyl OH is 1-3 orders slower than
OH radical addition pathways and should contribute very
little to the product pattern.
To probe the quinoneimine path, recourse was made
to the peroxidase/H2O2 oxidation system, which produces
the quinoneimine of 1 in the presence of excess H2O2 by
a mechanism independent of OH radical formation.
Oxidation of 1 (1 mM) with H2O2 (100 mM) in the
presence of horseradish peroxidase resulted in a fast
substrate consumption (90% within 1 min) leading to only
small amounts of 1,4-hydroquinone, but no 4-acetylami-
nocatechol, 4-acetylaminoresorcine, and 1,4-benzoqui-
none (Scheme 4).
Little or no fragmentation products were observed,
oligomeric and polymeric materials being the main
products. TOC values were near 100%, thus confirming
the lack of significant mineralization. This finding ruled
out UV-independent hydroxylation/degradation of the
Butanol, an established OH radical scavenger,22 de-
creased the rate of decay of 1 in a concentration-
dependent manner (Figure 4). UV or H2O2, separately,
induced little or no degradation of 1. These findings
indicated that OH radicals actually play a major role in
the oxidative degradation of 1 and other phenolic species,
e.g. 1,4-hydroquinone, and that direct photolysis is of
minor importance. By contrast, direct photolysis may
become an important contributory factor in the degrada-
tion of 1,4-benzoquinone and other quinonoid products
in the mixture. The formation of 1,4-hydroquinone by
photolysis of 1,4-benzoquinone without reductive treat-
ment has already been reported in the literature.10
The structures of the early aromatic products indicated
that three initial hydroxylation/oxygenation reactions
compete as the initial step in the degradation of 1
(Scheme 2). These reactions would involve attack of the
OH radical onto the aromatic ring of 1 leading to ortho,
meta, and para hydroxylation with respect to the OH
group.1H NMR monitoring of the reaction course in the
very early stages (keeping substrate consumption below
20%) indicated a formation ratio of acetamide/4-acetyl-
aminocatechol of ca. 1:1. In the reasonable assumption
that nearly all of the acetamide initially formed derives
from direct hydroxylation of 1, it could be concluded that
hydroxylation at the positions ortho and para (meta
hydroxylation is of minor relevance) proceed to compa-
rable degrees.
(22) Stoeffler, B.; Luft, G. Chem. Eng. Technol. 1998, 70, 1556-
1559.
(23) Matsuura, T.; Omura, K. Synthesis 1977, 173-184.
6148 J . Org. Chem., Vol. 67, No. 17, 2002

Advanced Oxidation Chemistry of Paracetamol
SCHEME 5
of 1,4-hydroquinone and 4-acetylaminocatechol. This
could explain why no significant levels of benzoquinone
were detectable by HPLC during degradation of 1. In
control experiments, 1,4-benzoquinone did not oxidize 1
or 4-acetylaminocatechol by direct redox reaction. Plau-
sible degradation pathways leading to nitrogenous break-
down products are sketched in Scheme 5. Such pathways
are merely indicative and are not intended to provide a
complete representation of the actual complexities of the
degradation process. OH radicals, because of their pro-
clivity for electron-rich aromatic compounds, could be
mainly engaged with the hydroxylation of phenolic spe-
cies, rather than of acyclic breakdown products, e.g.
carboxylic acids. 1,4-Hydroquinone (via 1,4-benzoqui-
none), 4-acetylaminocatechol, and 4-acetylaminoresorcine
would suffer further hydroxylation and/or one-electron
oxidation by OH radicals to give semiquinones and,
hence, quinones by disproportionation. These latter
would undergo muconic-type cleavage, and in part also
extradiolic cleavage, by the large excess of H2O2 through
a UV-assisted mechanism.24 The lack of detectable 4-ami-
nophenol in the mixtures indicated that deacetylation is
not a prominent route.
Oxalic acid monoamide, which accumulates together
with acetamide during the early stages of the degradation
of 1, could in principle arise by sequential OH radical-
induced hydroxylation steps of acetamide. However,
exposure of pure acetamide to the UV/H2O2 system led
to the formation of only negligible amounts of oxalic acid
monoamide (NMR evidence), thus ruling out a substan-
tial contribution of this route.
Con clu sion s
This paper provides the first detailed inventory of
intermediates and breakdown products formed by ad-
vanced oxidation of an aminophenol derivative. Besides
the practical implications of the research, these results
hold basic interest, as they expand the current knowledge
of the oxidation chemistry of 1, dissecting competitive
hydroxylation routes and the structural factors determin-
ing the final outcomes. Additional interest stems from
the potential of the combined GC-MS/NMR approach
based on 15N labeling to analyze nitrogenous degradation
products. The detection by 15N NMR of acetamide as a
main degradation product eluding GC-MS analysis is
paradigmatic in this regard. Finally, the generation of
relatively low amounts of aldehydes and carbonyl prod-
ucts by degradation of 1,4-hydroquinone and 1,2,4-
benzenetriol with the UV/H2O2 system compared to other
systems10 might be considered in assessing the ecological
relevance of the present study.
quinoneimine as a main reaction path of 1. Nonetheless,
the quinoneimine route may at least in part contribute
to 1,4-benzoquinone/1,4-hydroquinone formation.
To assess the relative importance of degradation
pathways downstream of the primary hydroxylation
steps, competition experiments were performed in which
4-acetylaminocatechol and 1,4-hydroquinone or 4-acetyl-
aminocatechol and 1,4-benzoquinone, with each compo-
nent at 0.5 mM concentration, were exposed to UV/H2O2
in the same mixture (Figure 5).
The results indicated a remarkable susceptibility to
oxidation of 1,4-benzoquinone, much greater than that
Exp er im en ta l Section
Ma ter ia ls. Paracetamol (1), glycolic acid, glyoxylic acid,
oxalic acid, oxalic acid monoamide, acetamide, 3-hydroxypyru-
vic acid, malonic acid, tartronic acid, ketomalonic acid, succinic
acid, maleic acid, fumaric acid, malic acid, oxalacetic acid, D,L-
tartaric acid, 2-ketoglutaric acid, mucic acid, 1,4-hydroquinone,
1,4-benzoquinone, 1,2,4-benzenetriol, p-aminophenol, 4-nitro-
catechol, phenol, 1,1,1,3,3,3-hexamethyldisilazane (HMDS),
chlorotrimethylsilane (TMSCl), and 30% hydrogen peroxide (w/
F IGURE 5. Decay of 4-N-acetylaminocatechol in the presence
of 1,4-benzoquinone or 1,4-hydroquinone by reaction with the
UV/H2O2 system: 4-N-acetylaminocatechol (2) vs 1,4-benzo-
quinone and (∆) 4-N-acetylaminocatechol (() vs 1,4-hydro-
quinone ()).
(24) Sawaky, Y.; Foote, C. S. J . Am. Chem. Soc. 1983, 105, 5035-
5040.
J . Org. Chem, Vol. 67, No. 17, 2002 6149

Vogna et al.
w) were obtained from commercial sources and used as
obtained. 1,2,4,5-Benzenetetraol was prepared by reduction of
2,5-dihydroxy-1,4-benzoquinone with sodium hyposulfite ac-
cording to a reported procedure.25 All other chemicals were of
the highest grade commercially available. Doubly glass-
distilled water was used throughout this study. Horseradish
peroxidase (donor: H2O2 oxidoreductase, EC 1.11.1.7) type II
(167 U/mg, RZ E430/E275 ) 2.0), bovine liver catalase (H2O2
oxidoreductase, EC 1.11.1.6), and [15N]NaNO2 (>99%) were
used.
out at 70 eV. Data were processed with Saturn WS data
analyses software. Temperature program: 80 °C, hold time 1
min; from 80 to 150 °C, rate ) 7 °C/min; 150 °C, hold time 5
min; from 150 to 200 °C, rate ) 7 °C/min; 200 °C, hold time 5
min. The inlet, transfer line, and manifold were taken at 250,
170, and 80 °C in that order. The acquisition started 4 min
after the injection (solvent delay 4 min), and was set in scan
mode in the range 40-650 amu. The threshold was fixed at 1
and sampling was 0.8 scans/s. For sample preparation 5-mL
aliquots were taken at different times from the oxidation
reaction of 1, [15N]1, 1,4-benzoquinone, 1,4-hydroquinone,
4-acetylaminocatechol, 1,2,4-benzenetriol, and 1,4-aminophe-
nol, lyophilized, and treated with HMDS (200 µL), anhydrous
pyridine (200 µL), and TMSCl (50 µL).10 The resulting mixtures
were shaken for 1 min and then centrifuged to separate the
precipitate prior to injection into the chromatograph.
Syn th esis of [15N]1. To solution of phenol (488 mg, 5.3
mmol) and Na15NO2 (731 mg, 10.6 mmol) in water (25 mL)
was added concentrated sulfuric acid (1 mL) under vigorous
stirring in an ice-bath. After 15 min, the yellow precipitate
formed was collected by filtration and dried. The solid thus
obtained (659 mg) was treated with zinc powder (2.0 g) in
glacial acetic acid (25 mL) and the mixture was taken under
reflux for 1 h. After removal of solid materials, the mixture
was taken to dryness under reduced pressure. The residue
dissolved in 0.2 M carbonate buffer at pH 10 (30 mL) was
treated with Na2S2O4 (5 g) and, after 15 min, the mixture was
filtered and added with acetic anhydride (2 mL). After 15 min,
the mixture was extracted with ethyl acetate (3 × 20 mL) and
the combined organic layers were washed with brine, dried
over sodium sulfate, and taken to dryness. The residue (259
mg) was purified by silica gel column chromatography with
chloroform/methanol 95:5 as the eluant to afford the title
compound (180 mg, 22% yield) homogeneous to TLC (chloroform/
methanol 90:10, Rf 0.72). 1H NMR (DMSO-d6), δ (ppm) 1.96
(s), 6.66 (d, J ) 8.0 Hz), 7.32 (dd, J ) 8.0, 1.6 Hz), 9.15 (s),
9.65 (d, J ) 88 Hz); 13C NMR (DMSO-d6) δ (ppm) 23.8 (CH3),
115.0 (CH), 120.0 (CH), 131.6 (C), 153.2 (C), 167.6 (C); 15N
NMR (DMSO-d6) δ (ppm) 133.0. GC-MS analysis (O-TMS
derivative) tR 21.71 m/z 224 (M+, 100). [Found C, 63.09; H,
5.90; N, 9.88; C8H915NO2 requires C, 63.14; H, 5.96; 15N 9.86.]
1
Meth od s. H, 13C, and 15N NMR spectra were recorded at
400.1, 100.6, and 40.5 MHz, in that order. An instrument fitted
with a 5 mm 1H/broadband gradient probe with inverse
geometry was used. For 15N NMR experiments delay values
up to 10 s were used. Experiments used were standard Bruker
implementations of gradient-selected versions of inverse (1H
detected) heteronuclear multiple quantum coherence (HMQC)
and heteronuclear multiple bond correlation (HMBC) experi-
ments. The HMBC experiments used a 100 ms long-range
coupling delay. 15N chemical shifts were referenced to [15N]-
urea in DMSO-d6 at δ 76.97 ppm, relative to NH3 (liquid, 298
K) at 0.0 ppm (15N NMR).26 DMSO-d6 was used as solvent.
TOC determinations were performed with a total organic
carbon analyzer. Irradiation experiments were carried out in
an annular reactor (420 mL) thermostated at 25 °C equipped
with a 17 W low-pressure mercury monochromatic lamp (254
nm).
Calculations of chemical shift values were carried out on
energy-minimized molecular structures with a Gaussian 01
program.27
HP LC An a lyses. Analytical HPLC was carried out with
an instrument equipped with a diode array detector. 0.1 M
formic acid pH 3/CH3CN (95:5 v/v) was used as the eluant for
determination of 1, 1,4-benzoquinone, 1,4-hydroquinone, and
4-acetylaminocatechol with use of an octylsilane coated column
(4.6 × 250 mm, 5 µm particle size) and a flow rate of 0.7 mL/
min. Identification of reaction products was made by compari-
son of the chromatographic behavior (detection wavelength 254
and 280 nm) and UV spectra with those authentic samples.
Quantitation was carried out by comparing integrated peak
areas with external calibration curves.
For determination of formic acid, acetic acid, and oxalic acid,
aliquots of the oxidation mixture of 1 (1 mL) were periodically
withdrawn and added in a screw-capped vial to a solution (2
mL) of 1,2-phenylenediamine (22 mM) in 1% sulfuric acid.
After being heated for 1 h at 100 °C, samples were neutralized
with NaOH and analyzed by HPLC with use of a hexylsilane-
coated column (2.0 × 25 mm, 10 µm particle size), with
detection set at 254 nm. Phosphate buffer (0.1 M)/acetonitrile
(90:10) containing 5% methanol was used as the mobile phase,
at a flow rate of 0.5 mL/min.
Syn th esis of 4-Acetyla m in oca tech ol. 4-Nitrocatechol
(500 mg, 3.22 mmol) in 0.2 M carbonate buffer at pH 10 (40
mL) was added to Na2S2O4 (5.6 g) at room temperature under
stirring. After 15 min acetic anhydride (0.72 mL) was added
and, after an additional 15 min, the mixture was extracted
with ethyl acetate (3 × 30 mL). The combined organic layers
were washed with brine, dried over sodium sulfate, and taken
to dryness to afford the title compound as a colorless solid (480
mg, 89% yield) that was stored under vacuum to avoid
darkening in air. 1H NMR (DMSO-d6) δ (ppm) 1.95 (s), 6.59
(d, J ) 8.4 Hz), 6.73 (dd, J ) 8.4, 2.0 Hz), 7.11 (d, J ) 2.0 Hz),
8.58 (s), 8.94 (s), 9.56 (s); 13C NMR (DMSO-d6) δ (ppm) 23.9
(CH3), 107.9 (CH), 110.4 (CH), 115.3 (CH), 131.6 (C), 141.2
(C), 145.0 (C), 167.7 (C); GC-MS analysis (O-TMS derivative)
tR 26.41 m/z 311 (M+, 100). [Found C, 57.51; H, 5.37; N, 8.42;
C8H9NO3 requires C, 57.48; H, 5.43; N, 8.38.]
GC-MS An a lyses. GC-MS analyses were carried out on a
GC instrument coupled with an ion trap detector using a 30
m cross-bond 5% diphenyl-95% dimethylpolysiloxane column
(0.25 mm i.d., 0.25 µm df). Helium was the carrier gas with a
1 mL/min flow rate. Electron impact ionization was carried
(25) _Anslow, W. K.; Raistrick, H. J . Chem. Soc. 1939, 1446-1454.
(26) Levy, G. C.; Lichter, R. L. Nitrogen-15 Nuclear Magnetic
Resonance Spectroscopy; J ohn Wiley & Sons: New York, 1979.
(27) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.;
Stratmann, J r. R.; Burant, E. J . C.; Dapprich, S.; Millam, J . M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .;
Barone, V.; Mennucci, B.; Cossi, M.; Adamo, C.; J aramillo, J .; Cammi,
R.; Pomelli, C.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Morokuma,
K.; Salvador, P.; Dannenberg, J . J .; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J . B.; Ortiz, J . V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J .; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J .; Keith,
T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.;
Gill, P. M. W.; J ohnson, B.; Chen, W.; Wong, M. W.; Andres, J . L.;
Gonzalez, C.; Head-Gordon, M.; Replogle, E. S.; Pople, J . A. Gaussian
01, Development Version (Revision A.01); Gaussian, Inc.: Pittsburgh,
PA, 2001.
Oxid a tion Rea ction : Gen er a l P r oced u r e. Solutions of
the appropriate substrate (1.0 mM) and hydrogen peroxide
(100 mM) in water (420 mL) or 0.1 M phosphate buffer, pH
5.5, 7.0, or 8.0, were irradiated under stirring. Aliquots of the
mixture were taken periodically and subjected to analysis
(HPLC, GC-MS, and NMR) with or without treatment with
NaBH4 as reported.10 When required, the reaction was carried
out with purging of the solution with argon for at least 30 min
prior to addition of hydrogen peroxide and throughout the
reaction course. Control experiments were carried out for all
substrates under the same reaction conditions but without
irradiation. In some experiments, tert-butyl alcohol at 100 or
200 mM was included in the oxidation mixture of 1. In another
experiment, oxidation of 1 (1 mM) was carried out with
6150 J . Org. Chem., Vol. 67, No. 17, 2002

Advanced Oxidation Chemistry of Paracetamol
peroxidase (8.5 U/mL) and H2O2 (100 mM) and the reaction
was halted by addition of catalase at various time intervals.
University for NMR facilities and Dr. Italo Giudicianni
for technical assistance.
Ack n ow led gm en t. This work was carried out in the
frame of the EC project “Ecotoxicological assessments
and removal technologies for pharmaceuticals in waste-
waters” REMPHARMAWATER (No EVK1-CT-2000-
00048), and is also part of the research program
“Tecniche Integrate di decontaminazione”, sponsored by
Regione Campania. GC-MS facilities were available at
the Department of Chemical Engineering, University
of Naples “Federico II”. We thank the “Centro Interdi-
partimentale di Metodologie Chimico-Fisiche of Naples
1
Su p p or tin g In for m a tion Ava ila ble: 1H, 13C, 15N, H,1H
COSY, H,13C HMQC/HMBC, and H-15N HMBC NMR spec-
tra of the oxidation mixture of 1 in DMSO-d6; MS data of main
GC peaks and structural interpretation for nitrogenous species
eluting under peaks 10, 14, 17, 18, 25, 26, 28, 30, 31, 33, and
35; and structures for recurrent and peculiar fragments of MS
spectra. This material is available free of charge via the
Internet at http://pubs.acs.org.
1
1
J O025604V
J . Org. Chem, Vol. 67, No. 17, 2002 6151