Fenton, Photo-Fenton, H

2
O
2
Photolysis, and TiO
2
Photocatalysis for
Dipyrone Oxidation: Drug Removal, Mineralization, Biodegradability,
and Degradation Mechanism
Ardhendu Sekhar Giri and Animes Kumar Golder*
Department of Chemical Engineering, Indian Institute of Technology, Guwahati, Assam 781 039, India
*S Supporting Information
ABSTRACT: Dipyrone (DIPY), an analgesic drug, is very quickly hydrolyzed to 4-methylaminoantipyrine (4-MAA) in an acidic
solution. Batch study was conducted to compare the performance of dierent oxidation processes such as Fenton (FP), photo-
Fenton (PFP), UV/H
2
O
2
photolysis (UVP), and UV/TiO
2
photocatalysis (UVPC) for removal of 4-MAA from aqueous
solution. The degradation eciency was evaluated in terms of total organic carbon (TOC) reduction and enhancement of
biodegradability. Maximum 4-MAA removals of 94.1, 96.4, 74.4, and 71.2% were achieved in FP, PFP, UCP, and UVPC,
respectively, against mineralization of 49.3, 58.2, 47, and 24.6%. The proposed mechanisms suggest that the cleavage of three
methyl moieties followed by pyrazolinone ring breakage led to formation of various intermediates with low errors (0.88 to 0.11
g/mol). The intermediates primarily were hydroxylated and carboxylic derivatives. BOD
5
to COD (BOD, biochemical oxygen
demand; COD, chemical oxygen demand) ratio of 0.4 resulted from DIPY decomposition in all processes with highest
improvement in PFP (BOD
5
/COD 1.5). The collapse of iron(III)-chelates under UV irradiation gave higher biodegradability.
1. INTRODUCTION
Advanced oxidation processes (AOPs) are widely used for
removal of pharmaceutical and personal care products from
industrial and municipal wastewater.
1
AOPs undergo conditions
through dierent reacting systems such as homogeneous or
heterogeneous phases, in light or dark, etc. However, they have
common characteristics of formation of hydroxyl free radicals
(OH

).
2
It causes consecutive unselective degradation of organic
materials. Complete mineralization and/or oxidization of
contaminants occur even at very low concentration, and the
byproducts formed may also be environmentally nonhazardous.
3
The Fenton process (FP) is based on redox reaction between
Fe
2+
and H
2
O
2
generating OH

radicals,
4
whereas the photo-
Fenton process (PFP) occurs additionally under UVvis
irradiation. Heterogeneous semiconductor photocatalysis using
TiO
2
causes oxidation of organic pollutants via hydroxyl radicals,
OH

ad
, and valence holes (h
+
) generated when the semi-
conductor is exposed to UV irradiation. TiO
2
photocatalysis
works at ambient conditions and may be induced by solar
irradiation.
5
UV photolysis of H
2
O
2
(UVP) and TiO
2
photo-
catalysis (UVPC) generally showa lesser extent of oxidation than
FP and PFP. Both homogeneous and heterogeneous catalytic
processes have shown promising results even at pilot-plant scale
in treatment of nonbiodegradable and toxic compounds.
Dipyrone (DIPY) is one of the most popular analgesics and
antipyretic drug. It is also known as Metamizole and Novalgin.
About one-third of it remains unchanged and can be found in
urinary excretion. DIPY is rapidly hydrolyzed into its main
metabolite, 4-methylaminoantipyrine (4-MAA). 4-MAA is
absorbed and biotransformed by enzymatic reactions.
6
The
metabolites of DIPY can be subdivided into two groups: (i)
decarboxylated metabolites and (ii) metabolites with a degraded
pyrazolinone moiety. The metabolic route is shown in Figure 1.
4-MAA is metabolized to 4-aminoantipyrine (4-AA) in human
liver via demethylation and further acetylated to acetylaminoan-
tipyrine (4-AAA) (Figure 1). 4-Formylaminoantipyrine (4-FAA)
is generated by an uncharacterized oxidation of an n-methyl
group of 4-MAA (Figure 1).
Received: July 16, 2013
Revised: January 1, 2014
Accepted: January 1, 2014
Published: January 1, 2014
Figure 1. Chemical structure of dipyrone drug (sodium [(2,3-dihydro-
1, 5-di methyl -3-oxo-2-phenyl -1-pyrazol -4-yl ) methyl ami no]-
methanesulfonate) and its metabolic pathways.
Article
pubs.acs.org/IECR
2014 American Chemical Society 1351 dx.doi.org/10.1021/ie402279q | Ind. Eng. Chem. Res. 2014, 53, 13511358
The metabolites of DIPY enter into the environment from
several anthropogenic sources such as pharmaceutical industry
and inrmary euent, water disinfection, and waste incineration
facilities. These compounds are not completely eliminated by
biological treatment, and thus their presence has been referenced
in sludge treatment plant euents and surface water at high
concentrations. Proper removal of DIPY and its metabolites
present in water and wastewater has an important role in the
prevention of diseases both in humans and animals. AOPs can be
employed for the detoxication of such compounds until the
biodegradability is improved to a level amicable for subsequent
biological treatment.
7
However, there is a lacuna of investigation
for the selection of suitable oxidation process(es) to achieve
biodegradability to such extent.
Therefore, in this study, four AOPs, i.e., FP, PFP, UVP, and
UVPC, are compared based on their performance for the: (i)
oxidation and mineralization of 4-MAA (primary metabolite of
DIPY), (ii) formation of degradation products, and (iii) variation
of biodegradability (BOD
5
/COD where BOD is the biochemical
oxygen demand and COD is the chemical oxygen demand). A
unied mechanism of 4-MAA oxidation is presented. It would
help to elucidate the eect of unreacted drug and degradation
products on the variation of biodegradability of dierent
oxidation processes.
2. MATERIAL AND METHODS
2.1. Reagents. Dipyrone (purity > 99%) was obtained from
Sigma Aldrich Chemical Ltd. (Hong Kong, China). The
chemical structure of DIPY is illustrated in Figure 1. DIPY is
hydrolyzed rapidly at low concentration. Almost complete
hydrolysis of DIPY to 4-MAA in a solution of 0.01 mM occurs
in 30 min at pH 2.5 and temperature 21 C.
8
Methanol (98% (v/
v) purity) and acetonitrile (99% (v/v) purity) of high
performance liquid chromatographic (HPLC) grade were
procured from Merck Specialties Pvt Ltd. (India). Ferrous
ammoniumsulfate heptahydrate (99%purity), sulfuric acid (98%
purity), titanium dioxide (TiO
2
; purity, 99% (w/w); crystal type,
rutile; specic surface area, 391 m
2
/g), H
2
O
2
(50% (v/v) purity),
silver sulfate (Ag
2
SO
4
), and K
2
Cr
2
O
7
(purity > 98%) were
obtained from Merck. Milli-Q water (model Elix 3, Millipore,
Billerica, MA, USA) was used to prepare all reagent and drug
solutions.
2.2. Analytical Techniques. High performance liquid
chromatography was used for the determination of 4-MAA
concentration. A 20 L sample was injected directly into a C
18
column of 4.6 mm inside diameter and 25 mm length. HPLC
instrument of Shimadzu, Japan (model LC-20AD) equipped
with an UVvisible detector was employed for the chromato-
graphic measurement. Methanol and water at the ow rate of 0.5
mL/min (80:20 (v/v)) was used as the mobile phase. The
scanning was performed at a xed wavelength of 254 nm. Liquid
chromatographytime-of-ight mass spectrometry (LC-TOF-
MS; Waters Q-Tof Premier & Aquity UPLC) system was
employed for the identication of drug fragments. The
chromatographic separation was performed on a YMC
(Wilmington, NC, USA) hydrosphere C
18
reverse phase column
(4.6 mm150 mm, 5 mparticle size) following a guard column
(4 mm 10 mm, 5 m particle size) using a mobile phase ow
rate of 0.8 mL/min at 25 C. The mobile phase was consisting of
H
2
O and acetonitrile with 0.1% (v/v) formic acid. A linear
gradient of 95 to 50% H
2
O was applied over 10 min. A 10 L
aliquot of both sample and calibration solution was injected from
an autoinjector. The samples were analyzed by electrospray
ionization (ESI) method in positive ion mode over the mass
range of 100400 amu. The parent ion was fragmented on the
basis of a suitable range of mass to charge (m/z) ratio. The most
prominent daughter ions were selected for further fragmentation.
Chemical oxygen demand was determined according to the
HACH method. A closed reux digester of HACH, Loveland,
CO, USA (model DRB 200) was used for digestion. Total
organic carbon (TOC) analyzer of O.I. Analytical, College
Station, TX, USA (model 1030C Aurora) was employed for the
measurement of TOCby nondispersive infrared method. A5 day
biochemical oxygen demand (BOD
5
) was found per the standard
method of analysis.
9
Dissolved oxygen (DO) was measured using
a precision dissolved oxygen meter of Hanna Instruments,
Smitheld, RI, USA (model HI 2400). Solution pH was
measured using a precision pH meter of Eutech Instruments,
Malaysia (model pH/ion 510).
2.3. Experimental Procedure. All experiments were
conducted in batch mode with continuous stirring. A 1000 mL
capacity cylindrical borosilicate vessel (i.d., 10.5 cm) was used as
the reactor system. A drug solution of 400 mL was taken for the
experimentation. The Fenton experiment was carried out with 50
mg/L DIPY (4-MAA) at room temperature (2325 C). The
initial concentration of DIPY was chosen based on the maximum
concentration reported in literature.
10
pH of the solution was
adjusted using 0.05 NH
2
SO
4
prior to addition of ferrous catalyst.
It was noted that the concentration of 4-MAA was almost
invariant in 2 h at pH 3.5 and temperature 25 C. Subsequently
the oxidation reaction was performed after 2 h of an initial
premixing period. Until and otherwise outlined, the eciency of
drug degradation and mineralization is reported with respect to
the concentration of 4-MAA.
A predetermined amount of Fe
2+
was added rst and mixed for
about 5 min at 260 rpmon a magnetic stirrer of Tarsons, Kolkata,
India (model Spinot 6020; stirring ba:, i.d., 0.8 mm; length, 40
mm). After that H
2
O
2
was added to 4-MAA/Fe
2+
solution. The
agitation was maintained at the same speed. A sample volume of
10 mL was withdrawn at selected time intervals. A 1 mL aliquot
of 0.1 N NaOH was immediately added into the sample to
terminate the oxidation reaction. Addition of NaOH increased
the solution pH at 12.3. Sludge was separated out by
centrifuging (Remi Instruments Ltd., Mumbai, India; model R-
23 8/06) at 2000 rpm for 30 min. Clear liquid was pipetted out
for pH, COD, TOC content, and drug concentration. Clear
liquid sample was heated at 70 C for the destruction of residual
H
2
O
2
(if any) prior to COD measurement. The supernatant was
ltered using 0.45 mcellulose lter (serial no. 08091ID0683) of
Pall India Pvt. Ltd. (Bangalore, India), before LC-MS analysis. 4-
MAA sorbed in sludge, if any, was determined by washing it in
distilled water followed by digestion (dissolution) in H
2
SO
4
(1
M).
In the case of PFP, the experiment was performed following
the same procedure in the presence of UV light. An UV lamp of 9
W from Hong Kong Jie Meng International Lighting Ltd.
Company, China (wavelength, 362 nm; intensity, 12 W/m
2
) was
employed for this work. The UVP experiment was conducted
with a xed concentration of H
2
O
2
under UV irradiation without
Fe
2+
. The UVPC test was carried out using TiO
2
photocatalyst.
The above experimental procedure was adopted for the UVPC
experiment with a xed amount of TiO
2
(1 g/L) only. All of the
experiments were performed in duplicate, and the average values
are reported.
Industrial & Engineering Chemistry Research Article
dx.doi.org/10.1021/ie402279q | Ind. Eng. Chem. Res. 2014, 53, 13511358 1352
3. RESULTS AND DISCUSSION
3.1. 4-MAARemoval and Mineralization. Concentrations
of Fe
2+
and H
2
O
2
and pH were varied from 1 to 3 mM, from 5 to
25 mM, and from 2 to 4, respectively. The results are shown in
Figures S1S3 of the Supporting Information. Drug removal
eciency increased with rise of one of these parameters, i.e., pH
and Fe
2+
and H
2
O
2
doses. It reached a maximum and then fell.
So, the catalytic eect of Fe
2+
was reduced at lower as well as at
higher concentration of a reactant. The best performance of
Fenton oxidation in terms of 4-MAA and TOC removal is
illustrated in Figure 2. PFP was carried out with the same
operating conditions. UVP was conducted with the same amount
of H
2
O
2
added in the Fenton reaction. In the case of the
heterogeneous catalytic reaction, catalyst dose and pH were
varied from 0.5 to 2.5 g/L and from 2 to 4.5, respectively. Their
eects on 4-MAA removal are in Figures S4 and S5 of the
Supporting Information. Increase in the TiO
2
dose from 1 g/L
had a negative eect on drug removal. Higher concentration
might cause catalyst agglomeration and hindrance to light
penetration.
11
pH determines the surface charge of the TiO
2
photocatalyst (pH
zpc
6.8). Lower pH favors 4-MAA
adsorption. However, too strong 4-MAA adsorption probably
decreased the absorption eciency of light quanta at lower pH.
The best performance of TiO
2
photocatalysis with these
experimental conditions are in Figure 2.
Temperature, solution volume, reaction time, and extent of
mixing were the same for all of the processes in order to compare
their eciencies for degradation of 4-MAA. The dynamics of 4-
MAA and TOC removal are shown in Figure 2a,b. The order of
percentage removal of 4-MAA at any time of the reactor was
found to be PFP > FP UVP > UVPC. Percentage removal of
drug achieved to 96.4, 94.1, 74.4, and 71.1%, respectively, in 45
min. Production of OH

radicals according to eqs 15 causes

degradation of 4-MAA as a function of time based on the
oxidation process.
+ = + +
=
+ +

k
Fe H O Fe OH OH
( 76 M s )
2
2 2
3
aq
1 1
(1)
+ +
=
+ +

hv
k
Fe(OH) Fe OH
( 3.1 10 M s )
2 2
3 1 1
(2)
+ =

h k H O 2OH ( 3.3 10 M s )
2 2
7 1 1
(3)
+ = + +
+ +
TiO (h ) H O TiO OH H
2 vb 2 ad 2 ad
(4)
+ = +
+
TiO (h ) OH TiO OH
2 vb ad 2 ad
(5)
4-MAA removal showed two distinct rate periods (Figure 2a),
i.e., initial faster drug removal followed by virtually constant rate
even though there was a notable amount of unreacted 4-MAA
present in solution in the case of UVP and UVPC. The transition
for these two processes lasted for a long period (2.515 min)
compared to the Fenton reaction with and without UV light
(2.510 min). It implies that OH

generation is predominant at
<2.5 min irrespective to the oxidation process. The rate of
formation of OH

falls because of H
+
scavenging (eq 6) and
H
2
O
2
autodecomposition (eq 7) with progress of reaction (>2.5
min). In the case of FP and PFP, drug removal of 73.5 and 83.2%
at 2.5 min of oxidation rose to 94.1 and 96.4%, respectively, in 45
min.
+ +
+
e OH H H O
2
(6)
+
+ +
H O H H O
2 2 3 2
(7)
Application of UV light during Fenton reaction generates excess
hydroxyl radicals, and ferric ions are reduced into ferrous leading
to further generation of OH

(eqs 1 and 2). Increase in gross

OH

concentration gave a bit higher degradation of 4-MAA in

PFP. It is evident from Figure 2a that UVP and UVPC are less
eective compared to the other two processes. UVP showed
higher removal eciency than UVPC. The rst process exhibits a
rate constant of nearly 25 times higher than the UV/TiO
2
process.
12
The drug removal achieved was 56.4 and 44.5%
within the rst 2.5 min in UVP and UVPC, respectively. It
increased to 74.4 and 71.1% in 45 min. OH

radicals are
generated by photolysis of the peroxide bond (eq 3). It formed
on the surface of TiO
2
(eqs 4 and 5), and adsorbed 4-MAAon the
surface is oxidized. Hence, the reaction occurs at the diusion-
controlled regime in UVPC.
13
Figure 2. (a) Removal of 4-MAA in dierent AOPs with reaction time.
(b) Removal of TOC. Initial 4-MAA concentration, 50 mg/L; TOC =
23.4 mg/L; temperature, 25 C. FP: Fe
2+
, 2.25 mM; H
2
O
2
, 22.5 mM;
pH, 3.5. PFP: Fe
2+
, 2.25 mM; H
2
O
2
, 22.5 mM; pH, 3.5; UV light, 9 W.
UVP: H
2
O
2
, 22.5 mM; pH, 3.5; UV light, 9 W. UVPC: TiO
2
, 1.0 g/L;
pH, 2.5; UV light, 9 W.
Industrial & Engineering Chemistry Research Article
dx.doi.org/10.1021/ie402279q | Ind. Eng. Chem. Res. 2014, 53, 13511358 1353
The trend of mineralization was similar to that of drug
removal. The highest TOC removal took place in PFP. It was
very close in the cases of FP and UVP (Figure 2b). The initial
rapid mineralization was up to 2.5 and 5 min for FP and PFP.
However, it was not so distinct for the other two processes. TOC
removal of 28.4 and 42.78% was obtained in 2.5 and 5 min with
FP in addition to 53.6 and 56.0%being obtained in 2.5 and 5 min
for PFP. The rst stage of TOC removal was very fast due to
mineralization of three-methyl moieties. The slowsecond stage is
related to the opening and mineralization of pyrazolinone ring.
14
Usually, large molecular weight intermediates were either
mineralized or broken to lower molecular weight products like
Figure 3. Degradation pathways of 4-MAA and identication of intermediates were obtained in dierent oxidation processes at 10 min of oxidation
(photoreaction with an UV lamp of 9 W; FP, PFP, and UVP conducted at pH 3.5 and UVPC at pH 2.5): (a) Cleavage of 4-MAA; (b) degradation of D
2
(m/z = 279.11) continued from panel a; (c) D
5
(m/z = 176.07) cleavage continued from panel b. The exact mass to charge ratio is in parentheses.
Industrial & Engineering Chemistry Research Article
dx.doi.org/10.1021/ie402279q | Ind. Eng. Chem. Res. 2014, 53, 13511358 1354
oxalic and acetic acids.
15
Kavitha and Palanivelu indicated that
carboxylic acids are eliminated in PFP very quickly during the
rst stage of oxidation.
16
Bauer et al. reported continuous
formation as well as degradation of lightweight carboxylic acids
during UVPC.
17
There was a dierence of about 20% in drug
removal between FP and UVP. However TOC reduction was
almost the same. It is necessary to mention that H
2
O
2
(22.5 mM)
alone gave around 12.8% 4-MAA decomposition. However, only
UV irradiation did not show notable eect on drug removal.
There is a signicant role of iron-chelation on drug
mineralization. Iron(III)-chelate complexes are more stable in
PF because of lower possibility of further degradation.
18
PFP
would have higher ability to degrade such complexes under UV
luminesce. Knight et al. suggested that Fe(III)-humate
complexes are easily reduced to Fe(II) by H
2
O
2
in PFP.
19
In
addition, humic acids themselves are photodegraded through
formation Fe(III) complexes.
20
Klamerth et al. pointed out that
PFP could breakdown metal complexes of molecular acids and
iron through separation between heavy metal and its complexing
agent.
21
TOC removed in PFP mostly corresponds to oxidation
of hydrophobic components which predominate in natural
organic compounds present in rawwater.
22
It can be explained by
greater aromaticity of the hydrophobic fraction giving higher
reactivity toward oxidizing agents. The refractory hydrophilic
fraction is consisting of short-chain aliphatic amines, alcohols,
aldehydes, esters, ketones, aliphatic amides (<C
5
), polyfunc-
tional alcohols, carbohydrates, cyclic amides, and polysacchar-
ides.
22
In the present work, acidic digestion of sludge formed in
FP showed 2.74% 4-MAA appearance on sludge phase.
3.2. Mechanisms of Drug Degradation. Fenton, TiO
2
photocatalysis, and their allied processes are characterized by the
formation of a common oxidizing agent, i.e., OH

radical used for

the decomposition of organic compounds. However, the extent
of the degradation of parent compounds and the formation of the
intermediates largely depend on process parameters such as pH,
dose of Fe
2+
/H
2
O
2
and TiO
2
, etc. A unied mechanism is
developed to elucidate the possible routes of 4-MAA degradation
by OH

generated both in homogeneous and heterogeneous

catalytic processes. 4-MAA(m/z = 218.11) is formed quickly due
to the presence of a more water-soluble sulfonate group
(-SO
3
H). 4-MAA (D
1
) contains a pyrazolinone structure. It
was rst separated on a reverse phase HPLC column using an
isocratic mobile phase. Then the m/z ratio recorded in MS
spectra was veried using 4-isopropylantipyrine (m/z = 231.3) as
an internal standard.
23
The nitrogen atom instead of carbon in
the pyrazolinone moiety is the most preferred site of OH

attack
releasing aniline.
4
It was formed via hydroxylation of the aromatic
ring by addition of OH

as in Figure 1. N-atom is mainly released

in the form of N
2
and ammonia from the -NHNH- moiety at
around 70 and 7% of the stoichiometric amount.
24
The mass spectra acquired in dierent oxidation processes are
shown in Figures S6S9 of the Supporting Information. A total
27 (D
1
D
27
) intermediate products were detected in four
oxidation processes on the basis of m/z. The error in mass was
calculated from the dierence between the exact and proposed
structural mass of protonated ions. Fairly low error values were
obtained for most of the proposed compounds (Table S1 of the
Supporting Information). There are two most possible pathways
for the cleavage of 4-MAA as shown in Figure 3a. Path 1 shows
the route of formation of D
2
D
7
corresponding to the proposed
m/z values of 279.11, 250.07, 136.04, 176.07, 198.11, and 199.11,
respectively. Path 2 presents the formation route of D
8
D
10
with
m/z of 205.11, 189.06, and 225.11, respectively. D
2
was formed
by hydroxylation of a double bond in the pyrazolinone ring of D
1
due to nucleophilic character of the -electron. D
2
formed a
chelate compound (m/z = 346.07) with Fe
3+
due to the
availability of a lone electron pair on N-atom (Figure 3a).
The opening of the pyrazolinone ring was conrmed by the
detection of 1-acetyl-1-methyl-2-aminomethyl-oxamoyl-2-phe-
nylhydrazide (D
3
). This compound was yielded by addition of
two OH

radicals to the double bond of the pyrazolinone ring of

D
1
followed by CCbond breaking. D
3
is found as a derivative in
the biotransformation of DIPY in humans.
25
The formation of D
4
was via an unstable intermediate with m/z = 220.03 because of
the presence of a diketo group more susceptible toward oxidation
in acidic medium.
23
The cleavage of NN bond in the hydrazine
group led to formation of N-phenylacetamide (D
4
) with m/z =
136.04 as well as an acetamide (Figure 3a). D
5
was originated in
PFP (Figure S7 of the Supporting Information) with low error
(0.031 g/mol) by hydroxylation of the carbonyl group followed
by CN bond breaking along with formation of acetic acid.
Attack of OH

radical to the carbonyl group due to its

electrophilic nature (i.e., electron decient center) caused aniline
formation with complete elimination of the pyrazolinone ring. D
6
was detected in FP, PFP, and UVP with very low error (0.04 g/
mol). It was originated by addition reaction between D
4
and
aniline on dehydration. Successive hydroxylation of aniline at
ortho- and para-positions originated a compound with m/z =
130.27 (Figure 3a). The -NH
2
group activates benzene nucleus
for the electrophilic substitution reaction due to +R eect
(mesmeric). D
7
was identied in PFP and UVP (Figures S7 and
S8 of the Supporting Information). The lone electron pair on the
N-atom of the D
6
compound acts as a Lewis base that appeals to
protons. Therefore, D
6
and D
7
are in equilibrium as shown in
Figure 3a.
4-Aminoantipyrine (D
8
, 4-AA) was formed with methanol in
FP on hydrolysis of D
1
. D
9
(m/z = 189.06) was detected in MS
spectra of FP. The loss of NH
3
from the pyrazolinone ring of the
previous compound yielded D
9
(error, 0.06 mg/L) because of a
shifting of a lone pair electron on N
1
. D
10
was found both in UVP
and UVPC(Figures S8 and S9 of the Supporting Information). A
similar explanation outlined for the formation of D
2
is also valid
here. D
2
was traced in mass spectra of UVP and UVPC (Figures
S8 and S9 of the Supporting Information). D
12
, D
14
-D
16
and D
19
formed in PFP were originated from D
2
(Figure 3b), whereas
only D
16
was formed in FP (Figure S6 of the Supporting
Information). It indicates formation of an iron-chelate complex
which is unstable under UV-luminance.
26
Carboxylic (Figure 3a)
and dicarboxylic acids (Figure 3c) are also known to form stable
iron complexes which inhibit the reaction with peroxide.
15
Carboxylic acids mainly exist as acetate and oxalate.
16
Oxalic acid
could form a iron(III)-oxalate complex as shown in Figure 3a.
Such a complex got stability through an extensive conjugation
with Fe(III) due to the availability of a vacant 3d orbital. In the
presence of UV light, its stability could loose because of UV light
absorption.
27
The ability of light absorption depends on electron
distribution of a molecule. Generally, the light absorption
capacity of sigma () bonds (CH, CC, CO, and OH) are
very less above 200 nm, whereas pi () bonds (carboxylic,
carbonyl group) and especially conjugated pi systems boost the
ability of a compound to absorb UV light at 254 nm by lowering
the energy barrier between ground and excited states.
28
Double
bond equivalent (DBE) was calculated by the summations of the
total number of pi bonds () and the residual ring(s). It is also
referred to as the degree of unsaturation.
29
PFP generated a
greater number of compounds having higher DBE (Table S1 of
Industrial & Engineering Chemistry Research Article
dx.doi.org/10.1021/ie402279q | Ind. Eng. Chem. Res. 2014, 53, 13511358 1355
the Supporting Information). Indeed, it can be seen from
Supporting Information Table S1 that compounds formed in
UVPC were consisting of lower DBE. The highest DBE of 9 was
calculated for both D
6
and D
7
.
There are also possibilities for cleavage of D
2
following path 3
and path 4 as shown in Figure 3b. Path 3 shows the route of
formation of D
11
and D
12
corresponding to m/z of 186.11 and
171.21, respectively. D
11
and D
12
both were formed from the
same intermediate (m/z = 164.11) via oxidation of D
4
originated
by hydrolysis of D
3
compound (path 1) as presented before. This
compound seems to be more stable in FP. Further attack of OH

at the -carbon of D
4
caused formation of D
11
in the case of PFP.
The carbonyl group is very susceptible to being oxidized at lower
pH (<3.5).
30
The valency of the carbon atom in D
12
is saturated
by two additional hydroxyl groups (-OH) due to the availability
of a partially vacant p-orbital in the N
1
-atom. Path 4 represents
the formation route of D
13
D
19
with m/z = 187.19, 152.19,
152.09, 166.09, 155.06, 183.11, and 212.02, respectively. The
structures of these compounds are proposed with low errors
(2.11 to 0.22 g/mol; Figure 3b). D
13
was formed by
hydroxylation of the -carbon of an intermediate with respect
to the N
2
-atom. Both D
14
and D
15
compounds have almost the
same mass (m/z = 152.19 and 152.09). Structures of these
compounds are dierent and have appeared from the same
intermediate with m/z = 194.08. The rst compound was
originated in PFP by decarboxylation (-CO
2
). The second was
the hydroxylation product with formation of NH
3
and CO
2
.
Identication of D
14
(Figure S7 of the Supporting Information)
is evidence of the opening of pyrazolinone ring by an alternative
route (Figure 3b). The proposed structure suggests oxidative
cleavage of NN bond to a multiple hydroxylated derivative.
OH

radical can easily attack at the N

1
position of the
pyrazolinone ring in the presence of a partially vacant p-orbital
of the N-atom.
31
D
16
and D
17
were originated by hydroxylation.
Oxidation of -carbon in D
17
with loss of hydroxylamine is
evidence of formation of D
16
. The -carbon in D
17
behaves like a
nucleophilic center because of the lone pair electron on the N
1
position shifted toward the terminal carbon atom through
conjugation. D
18
was formed from D
17
(Figure 3b). A possible
explanation similar to that for D
15
is applicable. The primary
alcohol group of the pyrazolinone moiety was oxidized to D
19
found in PFP. Waki et al. suggested that the primary alcohol in
acidic solution gives acid products on oxidation.
32
D
20
was
detected in UVPC with a low error (0.42 g/mol). Direct
hydroxylation of D
5
at the N
1
position with the loss of
methylamine justies its formation. D
20
was broken in three
dierent routes (paths 57; Figure 3c). Path 5 represents
formation of D
21
molecule by proton loss from D
21
. It is stable as
a salt form because of the slightly acidic nature of the -OH group.
D
22
compound was seen both in PFP and UVP yielded from
N-hydroxyl aniline derivative with formation of oxalic acid (path
6). It was ultimately transformed to CO
2
giving eective TOC
reduction. This aniline derivative by its para-position attack with
acetanilide, formed D
22
compound. Electrons are more localized
at the 4-position with respect to the N-substituted amine group
due to shifting of the lone pair electron and OH

attack at the
more electrophilic center of the carbonyl group. D
23
and D
24
were identied in PFP. The N
1
-atom has a vacant p-orbital to
accept the electron of OH

radicals (Figure 1; path 7). This

compound exists as a salt of sodium ion in alkaline medium (pH
> 8). It might have formed during pH adjustment at >12 for the
termination of oxidation reaction. Further oxidation of D
23
compound yielded the corresponding quinoneimine inter-
mediate (D
24
) which is generally unstable but it appeared in MS
spectra (Figure S7 of the Supporting Information). Hydrox-
ylation probably occurred at the para-position due to availability
of more electron density. The enol form of the D
24
compound is
less stable due to a positive charge on the N
1
-atom. It has a
greater tendency to be transformed into the quinone form,
15
i.e.,
the formation of D
25
compound obtained both in FP and UVP
having positive characteristics of the benzene nucleus.
D
26
and D
27
also appeared from D
25
by its successive
hydroxylation (Figure 3c). First the molecule was detected in
FP and UVP, whereas the second one was found in UVPC. D
27
is
alicyclic in nature and has resulted by hydroxylation. This
compound lost its aromaticity by substituting two -OH groups at
the 2- and 6-positions with respect to N,N-hydroxyl amine group.
3.3. Biodegradability: 4-MAA and Its Degradation
Products. The ratio of biochemical oxygen demand to chemical
oxygen demand (BOD/COD) is an important indicator to study
the nature of biodegradability.
33
Generally, wastewater contain-
ing pharmaceutical and personal care products are considered to
be reasonably biodegradable with BOD
5
/COD > 0.4.
34
The
initial BOD
5
and COD were 3.75 and 41.3 mg/L for an aqueous
solution with 50 mg/L DIPY. It implies the nonbiodegradable
nature of DIPY (BOD
5
/COD 0.1).
From Figure 2a it is evident that the extent of drug
decomposition was around 5% higher in PFP than FP.
Nevertheless, PFP resulted in notably higher biodegradability
(Figure 4). A BOD
5
/COD = 1.51 was achieved in PFP in
comparison to 0.62 in FP. It implies that the rst process
generated more (gross) biodegradable intermediates. UVPC
exhibited about 20% higher BOD
5
/COD improvement than
UVP (Figure 4). It was about 11%higher than the corresponding
drug decomposition. Comparatively lower BOD
5
/COD in UVP
and UVPCthan FP and PFP was largely due to unreacted 4-MAA
and intermediate products. Hyvonen et al. reported that a higher
extent of conjugation due to formation of chelates of three
dierent heavy metals, i.e., Cd(II), Hg(II), and Pb(II), and
(dicarboxyethoxy)ethyl]aspartic acid (BCA6) improves biode-
gradability.
35
Kummerer et al. reported that newly formed
Figure 4. Variation of BOD
5
/COD in dierent oxidation processes and
its initial value (initial 4-MAA, 50 mg/L; oxidation time, 10 min;
temperature, 25 C): with FP, 2.25 mM Fe
2+
, 22.5 mM H
2
O
2
, and pH
3.5; with PFP, 2.25 mMFe
2+
, 22.5 mMH
2
O
2
, pH3.5, and 9 WUVlight;
with UVP, 22.5 mMH
2
O
2
, pH3.5, and 9 WUVlight; with UVPC, 1.0 g/
L TiO
2
, pH 2.5, and 9 W UV light.
Industrial & Engineering Chemistry Research Article
dx.doi.org/10.1021/ie402279q | Ind. Eng. Chem. Res. 2014, 53, 13511358 1356
intermediates act as chelating agents that signicantly reduce
their biodegradability.
36
It can be seen from a to c of Figure 3 that D
2
, D
3
, D
14
, and D
23
could form chelates with Fe
3+
. D
1
, D
2
, D
6
, and D
25
traced in FP
could exhibit more biodegradability because of extended
conjugation. Therefore, the net eect of chelation and its
stability under UV irradiance and formation of intermediate
products and their dierent extent of conjugation resulted in
higher biodegradability in PFP. The biodegradable nature of a
number of intermediates as in the proposed structure is found
out from the earlier reports. It is summarized in Table S1 of the
Supporting Information.
4. CONCLUSIONS
Dipyrone at lower concentration (0.149 mM) was almost
completely hydrolyzed to 4-MAA within 2 h at pH 3.5 and
temperature of 25 C. Maximum 4-MAA removals of 94.1, 96.4,
74.4, and 71.2% were noted against mineralization eciencies of
49.3, 58.2, 47, and 24.6% in Fenton, photo-Fenton, H
2
O
2
photolysis, and TiO
2
photocatalysis, respectively. PFP showed
better mineralization eciency due to additional amount of OH

formation and collapse of iron complexes under UV irradiation

probably because of enhanced decarboxylation. However, in the
case of TiO
2
photocatalysis more hydroxylated products were
identied that resulted from direct attack of OH

ad
. A total of 29
intermediate products appeared in the mass spectra within the
mass to charge ratio of 100400 in four oxidation processes from
the same parent molecule. Pyrazolinone ring was degraded
preceded by cleavage of methyl moieties. The proposed
mechanism implies that most of the intermediates were formed
by pyrazolinone ring degradation. The order of biodegradability
in dierent oxidation processes was found to be PFPFP > UVP
> UVPC.

ASSOCIATED CONTENT
*S Supporting Information
Table S1 and Figures S1S9 containing additional information
on intermediate products, experimental optimization of CIP
decomposition, and mass spectra as outlined in the text. This
material is available free of charge via the Internet at http://pubs.
acs.org.

AUTHOR INFORMATION
Corresponding Author
*E-mail: animes@iitg.ernet.in. Tel.: +91-361-2582269. Fax: +91-
361-258-2292.
Notes
The authors declare no competing nancial interest.

ACKNOWLEDGMENTS
We gratefully thank the Indian Institute of Guwahati (India), for
providing the research fellowship to Mr. Ardhendu Sekhar Giri
and necessary research facilities to the Department of Chemical
Engineering and Central Instruments Facility (CIF).

REFERENCES
(1) Moldovan, Z. Occurrences of pharmaceutical and personal care
products as micropollutants in rivers from Romania. Chemosphere 2006,
64, 18081817.
(2) Torres, R.; Sarria, V.; Torres, W.; Peringer, P.; Pulgarin, C.
Electrochemical treatment of industrial wastewater containing 5-amino-
6-methyl-2-benzimidazolone: toward and electrochemical-biological
coupling. Water Res. 2003, 37, 713.
(3) Malato, S.; Blanco, J.; Vidal, A.; Richter, C. Photocatalysis with
solar energy at a pilot-plant scale: An overview. Appl. Catal. B: Environ.
2001, 37, 115.
(4) Pignatello, J. J.; Oliveros, E.; MacKay, A. Advanced oxidation
processes for organic contaminant destruction based on the Fenton
reaction and related chemistry. Crit. Rev. Environ. Sci. Technol. 2006, 36,
184.
(5) Malato, S.; Fernandez-Ibanez, P.; Maldonado, M. I.; Blanco, J.;
Gernjak, W. Decontamination and disinfection of water by solar
photocatalysis: recent overview and trends. Catal. Today 2009, 147, 1
59.
(6) Oliveira, R.; Almeida, M.; Santos, L.; Madeira, L. Experimental
design of 2,4-dichlorophenol oxidation by Fentons reaction. Ind. Eng.
Chem. Res. 2006, 45, 12661276.
(7) Morais, J. L.; Zamora, P. P. Use of advanced oxidation processes to
improve the biodegradability of mature landfill leachates. J. Hazard.
Mater. 2005, 123, 181186.
(8) Ergun, H.; Fratarelli, D. A. C.; Aranda, J. V. Characterization of the
role of physicochemical factors on the hydrolysis of dipyrone. J. Pharm.
Biomed. Anal. 2004, 35, 479487.
(9) APHA. Standard Methods for the Examination of Water and
Wastewater, 20th ed.; American Public Health Association: Washington,
DC, USA, 1998.
(10) Huber, M. M.; Gobel, A.; Joss, A.; Hermann, N.; Loffler, D.;
Mcardell, C. S.; Ried, A.; Siegrist, H.; Ternes, T. A.; Von Gunten, U.
Oxidation of pharmaceuticals during ozonation of municipal wastewater
effluents: A pilot study. Environ. Sci. Technol. 2005, 39, 42904299.
(11) Sohrabi, M. R.; Ghavami, M. Taguchi experimental design used
for Nano photo catalytic degradation of the pharmaceutical agent
Aspirin. J. Chem. Pharm. Res. 2008, 153 (3), 12351239.
(12) Jayson, G. G.; Parsons, B. J. Oxidation of ferrous ions by
perhydroxyl radicals. Trans. Faraday Soc. 1972, 68, 236242.
(13) Petrovic, M. M.; Bianca, F. S.; Aleksandra, J. R. Oxidative
transformation of fluoroquinolone antibacterial agents and structurally
related amines by manganese oxide. Chemosphere 2011, 85, 13311339.
(14) Gomez, M. J.; Martnez-Bueno, M. J.; Lacorte, S.; Fernandez-
Alba, A. R.; Aguera, A. Pilot survey monitoring pharmaceuticals and
related compounds in a sewage treatment plant located on the
Mediterranean coast. Chemosphere 2007, 66, 9931002.
(15) Leonidas, A.; Perez-Estrada, S. M.; Ana, A. R. Degradation of
dipyrone and its main intermediates by solar AOPs Identification of
intermediate products and toxicity assessment. Catal. Today 2007, 129,
207214.
(16) Kavitha, V.; Palanivelu, K. The role of ferrous ion in Fenton and
photo-Fenton processes for the degradation of phenol. Chemosphere
2004, 55, 12351243.
(17) Bauer, R.; Waldner, G.; Fallmann, H.; Hager, S.; Klare, M.;
Krutzler, T.; Malato, S.; Maletzky, P. The photo-fenton reaction and the
TiO
2
/UV process for wastewater treatmentNovel developments.
Catal. Today 1999, 53, 131144.
(18) Trovo, A. G.; Raquel, F. P.; Nogueira, A. A.; Carla, S. A.
Photodegradation of sulfamethoxazole in various aqueous media:
Persistence, toxicity and photoproducts assessment. Chemosphere
2009, 77, 12921298.
(19) Knight, R. J.; Sylva, R. N. Spectrophotometric investigation of
iron (III) hydrolysis in light and heavy water at 25C. J. Inorg. Nucl.
Chem. 1975, 37, 779783.
(20) Fukushima, M.; Tatsumi, K.; Nagao, S. Degradation Character-
istics of Humic Acid during Photo-Fenton Processes. Environ. Sci.
Technol. 2001, 35, 36833690.
(21) Klamerth, N.; Malato, S.; Aguera, A.; Fernandez-Alba, A. Photo-
Fenton and modified photo-Fenton at neutral pH for the treatment of
emerging contaminants in wastewater treatment plant effluents: A
comparison. Water Res. 2013, 47, 833840.
(22) Buchanan, W.; Roddick, F.; Porter, N.; Drikas, M. Fractionation
of UV and VUV pretreated natural organic matter from drinking water.
Environ. Sci. Technol. 2005, 39, 46474654.
Industrial & Engineering Chemistry Research Article
dx.doi.org/10.1021/ie402279q | Ind. Eng. Chem. Res. 2014, 53, 13511358 1357
(23) Ojha, A.; Rathod, R. Quantification of 4-methylaminoantipyrine,
the active metabolite of dipyrone, in human plasma. Bioanalysis 2009, 1,
293298.
(24) Calza, P.; Sakkas, V. A.; Medana, C.; Baiocchi, C.; Dimou, A.;
Pelizzetti, E.; Albanis, T. Photocatalytic degradation study of diclofenac
over aqueous TiO
2
suspensions. Appl. Catal. B: Environ. 2005, 67, 197
205.
(25) Wessel, J. C.; Matyja, M.; Neugebauer, M.; Kiefer, H.; Daldrup,
T.; Tarbah, F. A.; Weber, H. Characterization of oxalic acid derivatives as
new metabolites of metamizol (dipyrone) in incubated hens egg and
human. J. Pharmaceut. Sci. 2006, 28, 1525.
(26) Zepp, R. G.; Faust, B. C.; Hoigne, J. Photoproduction of hydrated
electrons from natural organic solutes in aquatin environments. Environ.
Sci. Technol. 1992, 26, 313319.
(27) Zhou, W.; Moore, D. E. Photochemical decomposition of
sulfamethoxazole. Int. J. Pharm. 2011, 110, 5563.
(28) Hesse, M.; Meier, H.; Zeeh, B. Spectroscopic Methods in Organic
Chemistry, 2nd ed.; Thieme: New York, 2008; pp 1517.
(29) Finar, I. L. Stereochemistry and the Chemistry of Natural Products,
Vol. 2, 5th ed.; Holloway: London, 2001; pp 89.
(30) Fabre, H.; Eddine, N. H.; Bressolle, F.; Mandrou, B. Stability
indicating assay for dipyronePart I. Analyst. 1982, 107, 6166.
(31) Sykes, P. A guidebook to mechanism in organic chemistry, 6th ed.;
Fellow of Christs College: Cambridge, U.K., 2005; pp 320322.
(32) Waki, K.; Zhao, J.; Horikoshi, S.; Watanabe, M.; Hidaka, H.
Photooxidation mechanism of nitrogen-containing compounds at
TiO
2
/H
2
O interfaces: an experimental and theoretical examination of
hydrazine derivatives. Chemosphere 2000, 41, 337342.
(33) Tekin, H.; Bilkay, O.; Selale, S.; Tolga, H. Use of Fenton oxidation
to improve the biodegradability of a pharmaceutical wastewater. J.
Hazard. Mater. 2008, 136, 258265.
(34) Tang, W. Z.; Tassos, S. Oxidation kinetics and mechanisms of
trihalomethanes by Fentons reagent. Water Res. 1997, 31, 11171125.
(35) Hyvonen, H.; Orama, M.; Arvela, R.; Henriksson, K.; Saarinen,
H.; Aksela, R.; Paren, A.; Jakara, I.; Renvall, I. Studies on three new
environmentally friendly chelating ligands. Appita J. 2006, 59, 142149.
(36) Kummerer, K.; Al-Ahmad, A.; Mersch-Sundermann, V.
Biodegradability of some antibiotics, elimination of the genotoxicity
and affection of wastewater bacteria in a simple test. Chemosphere 2000,
40, 701710.
Industrial & Engineering Chemistry Research Article
dx.doi.org/10.1021/ie402279q | Ind. Eng. Chem. Res. 2014, 53, 13511358 1358