Pm

Ea

b

h

�

�

�

�

a

ARRAA

KPPR

1

macac

if

0h

Journal of Hazardous Materials 243 (2012) 130– 138

Contents lists available at SciVerse ScienceDirect

Journal of Hazardous Materials

j our na l ho me p age: www.elsev ier .com/ locate / jhazmat

hotocatalytic degradation of paracetamol: Intermediates and total reactionechanism

dgar Moctezumaa, Elisa Leyvaa,∗, Claudia A. Aguilarb, Raúl A. Lunaa, Carlos Montalvob

Facultad de Ciencias Químicas, Universidad Autónoma de San Luis Potosí, Av. Manuel Nava #6, San Luis Potosí, S.L.P., 78290, MexicoFacultad de Química, Universidad Autónoma del Carmen, Calle 56 #4, Cd. del Carmen, Campeche, 24180, Mexico

i g h l i g h t s

The advanced oxidation of paraceta-mol with TiO2/UV was investigated.HPLC indicated the formationof hydroquinone, benzoquinone,p-aminophenol and p-nitrophenol.IR studies indicated that p-nitrophenol and p-aminophenolare also produced.PAM photocatalytic reaction mecha-nism includes an alternative deacy-lation pathway.

g r a p h i c a l a b s t r a c t

r t i c l e i n f o

rticle history:eceived 25 February 2012eceived in revised form 4 October 2012ccepted 5 October 2012vailable online 13 October 2012

a b s t r a c t

The advanced oxidation of paracetamol (PAM) promoted by TiO2/UV system in aqueous medium wasinvestigated. Monitoring this reaction by HPLC and TOC, it was demonstrated that while oxidation of para-cetamol is quite efficient under these conditions, its mineralization is not complete. HPLC indicated theformation of hydroquinone, benzoquinone, p-aminophenol and p-nitrophenol in the reaction mixtures.Further evidence of p-nitrophenol formation was obtained following the reaction by UV–vis spectroscopy.

eywords:aracetamolhotocatalytic degradationeaction mechanism

Continuous monitoring by IR spectroscopy demonstrated the breaking of the aromatic amide present inPAM and subsequent formation of several aromatic intermediate compounds such as p-aminophenol andp-nitrophenol. These aromatic compounds were eventually converted into trans-unsaturated carboxylicacids. Based on these experimental results, an alternative deacylation mechanism for the photocatalyticoxidation of paracetamol is proposed. Our studies also demonstrated IR spectroscopy to be a useful

xida

technique to investigate o

. Introduction

In recent years, there is a growing interest in the environ-ental relevance of pharmaceutical compounds in waters. There

re different sources for this type of pollution like waste from

hemical industries, disposal from household, and excretion afterdministration in humans and animals. Several pharmaceuticalompounds have been detected as minor pollutants in industrial

∗ Corresponding author. Present address: Department of Chemistry, UWO, Chem-stry Building, London N6A 5BA, Ontario, Canada. Tel.: +1 519 661 2166;ax: +1 519 661 3022.

E-mail address: elisa@uaslp.mx (E. Leyva).

304-3894/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.jhazmat.2012.10.010

tive mechanisms of pharmaceutical compounds.© 2012 Elsevier B.V. All rights reserved.

waste, surface and ground waters [1]. Many of these compoundsare extensively used as anti-inflammatories, analgesics, lipid reg-ulators, antibiotics, antiepileptics, antiseptics and disinfectants[2–5].

To avoid further accumulation of pharmaceutical compoundsin the aquatic environment, several research groups are utilizingadvance oxidation processes (AOP), where a highly reactive inter-mediate like HO• is generated, to achieve the destruction of thesecontaminants in water. The successful application of ozonation andother AOP (O3/H2O2, H2O2/UV, and H2O2/Fe+2/UV) for the oxida-

tion of pharmaceuticals and their metabolites has been reported[5–8].

An important AOP is the photocatalytic oxidation of organicpollutants by the use of a semiconductor catalyst that upon

zardo

pe(soohriodwhtss

T

e

h

h

h

H

e

O

H

4aitcaFifd1

rPdiidaomadsHw

smG

E. Moctezuma et al. / Journal of Ha

hotolysis with UV light undergoes a surface reaction in whichlectrons move from the valence band to the conduction bande−) leaving positive holes (h+) behind [9–13]. Both highly reactivepecies (e− and h+) initiate oxidation and reduction reactionsn the surface of a TiO2 catalyst. Organic compounds (D) can bexidized directly by positive holes (h+). In an aqueous suspension,+ reacts with water molecules and surface HO− groups to give HO•

adicals, which are known to be strong oxidizing species. Oxygens bubbled through the reaction mixture to trap free electrons inrder to inhibit the recombination step. Other reaction interme-iates contribute to the production of HO• radicals, which reactith organic pollutants and eventually mineralize them [11,14]. Itas been demonstrated by electron spin resonance spectroscopyhat HO• is the most abundant intermediate in aqueous TiO2uspensions [15]. The photocatalytic reaction mechanism could beummarized in the following cascade of reactions [9,12,16]:

iO2hv−→e− + h+ (1)

– + h+ → recombination (1a)

– + H2O → HO• + H+ (2)

– + HO– → HO• (3)

+ + D → D+ (Direct oxidation of organic compounds) (4)

O• + D

→ Doxidized (Oxidation of organic compounds by HO• radicals)

(5)

– + O2 → O2•– (6)

2•– + HO2

• + H– → H2O2 + O2 (7)

2O2 + e– → HO– + HO• (8)

Paracetamol (PAM, Fig. 1) also known as 4-hydroxyacetanilide,-acetamidephenol or acetaminophen is widely used as analgesicnd antipyretic agent. In contrast to aspirin, it does not cause gastricrritation and it is a component of more than one hundred medica-ions [17]. When it is used in excess or with alcohol, paracetamolould cause liver failure and even death [18]. This toxicity has beenttributed to the production of a highly toxic metabolite (NAPQI,ig. 1) formed by enzymatic oxidation in the liver. PAM occurrencen environment has attracted interest as a potential contaminantor waters. Concentration levels on the range of 1–6 �g/L have beenetected in European sewage treatment plant effluents and up to0 �g/L in water samples from natural sources in the USA [4].

An insight into the advanced oxidation chemistry of PAM waseported by Vogna et al. [19]. They investigated the oxidation ofAM with UV/H2O2 to determine the intermediate and final breakown products in this reaction utilizing a combination of analyt-

cal techniques such as GC–MS, HPLC, and NMR. In a subsequentnvestigation on ozonation and H2O2 photolysis of PAM, it wasemonstrated that both oxidative systems are able to destroy theromatic ring of the substrate but low mineralization in the orderf 30% was determined by TOC analysis [20]. Main reaction inter-ediates and products were identified for both systems by HPLC

nd GC–MS. To improve the amount of mineralization, the degra-ation of PAM has been investigated with several oxidative systemsuch as O3/UV with Fe+3 and Cu+2 as catalysts [21], electrogenerated2O2/UV with Fe+3 and Cu+2 as catalysts [22], and anodic oxidationith boron-doped diamond electrode [23].

The first advanced oxidation of PAM promoted by TiO2/UVystem in aqueous medium was investigated [24]. Continuousonitoring by several techniques, such as HPLC, TOC, ESI-MS andC–MS revealed that PAM removal was quite efficient under these

us Materials 243 (2012) 130– 138 131

reaction conditions. Fragmentation pathways were also proposedbased on MS analysis of aliquots taken at consecutive oxidationreaction times. Prior to MS analysis, all samples were submitted toa derivatization procedure to give the more stable trimethylsilylderivatives. In two subsequent studies, the degradation of PAM byTiO2 photocatalysis was also investigated to determine the optimaloperating conditions [25,26].

In this study, we investigated the photocatalytic oxidation ofPAM, a reagent extensively used in medicine, in the presence ofTiO2. This reaction was monitored by a combination of techniquessuch as UV–vis, IR and HPLC. The mineralization of paracetamolwas followed by total organic carbon analysis (TOC). Based on theseexperimental results a total reaction mechanism for the oxidationof PAM is proposed. In addition to the quinone and hydroquinoneoxidation pathway, involving the lost of acetamide, previouslyreported by others [19,20,24], an alternative deacylation oxidationroute is presented.

2. Experimental

2.1. Materials

PAM, p-aminophenol, p-nitrophenol, benzoquinone, hydro-quinone and 1,2,4-trihydroxybenzene were purchased fromAldrich. Other organic and inorganic acids and salts were purchasedfrom J.T. Bayer. Spectroscopic and chromatographic grade solventswere purchased from Mallinckrodt. Double distilled water, filteredthrough 0.45 �m HA cellulose acetate membranes (Millipore Corp.Bedford, MA) was used throughout.

2.2. Photocatalytic oxidation experiments

Photocatalytic oxidation experiments were carried out in a reac-tor system described in previous publications [12]. This unit isconfigured with a Pyrex glass tube reactor (400 mL) irradiated withfour 15 W UV light lamps (Cole Parmer E-09815-55, �max = 365 nm).It has a fan to cool down the reaction mixture to room temperature.For each set of experiments, 250 mL of a PAM standard solutionwere placed inside the glass reactor and slurried with 2 g/L of TiO2.This suspension was mixed with a magnetic stirrer. Pure oxygencan be bubbled through the reaction mixture at a constant rate of100 mL/min. Under these conditions, the initial pH of the solutionwas 7.9 ± 0.2, and it was measured with a 710A Orion pH meter.Samples were taken from time to time to monitor the progressof the reaction by different analytical techniques. Before analysis,all reaction mixtures were filtered through a 0.22 �m GV celluloseacetate membrane (Millipore Corp. Bedford, MA).

To identify the intermediate organic products generatedthrough photocatalytic degradation reactions, 300 mL of an aque-ous solution of PAM (300 ppm) were mixed with 2 g/L of TiO2 andirradiated in the photoreactor for several hours. The solution slurrywas kept under a constant flow of oxygen of 100 mL/min. A volume(100 mL) of mixture was removed at different times of reaction (2, 4and 6 h). The aqueous mixture was extracted three times with ethylacetate. The solvent was dried with anhydrous sodium sulfate andremoved in a rotary evaporator to yield an oil that was placed inKBr cells for subsequent analysis by FT-IR spectroscopy.

2.3. Analytical methods

For the determination of PAM and some intermediate com-

pounds, reaction mixtures at different irradiation times wereanalyzed by high performance liquid chromatography (HPLC) ina 600E Waters instrument equipped with a UV–vis Waters 490detector tuned at 242 nm. A Novapack-Phenyl column (4 �m,

132 E. Moctezuma et al. / Journal of Hazardous Materials 243 (2012) 130– 138

) and

3aseth45cE

Ush1t

3

3

dootH

F(fl

Fig. 1. Structure of paracetamol (PAM

.9 mm × 150 mm) was used to separate reactant and intermedi-te compounds. The mobile phase was a mixture of 60% aqueousolution (11 mM citric acid/0.055 mM EDTA) and 40% methanol. Thelute was delivered at a rate of 1 mL/min. For these operating condi-ions, the retention times for PAM, p-aminophenol, p-nitrophenol,ydroquinone and benzoquinone were 1.59, 1.61, 4.5, 1.44, and.7 min, respectively. The analytical error was controlled within%. The total organic carbon (TOC) was measured with a Shimadzuarbon analyzer model 5000A. IR spectra were obtained in a Perkinlmer 1600 FTIR instrument.

Some of the reaction samples were also analyzed byV–vis spectroscopy in a Shimadzu UV-2401 PC instrument in

tandard quartz cells. The spectra of standard solutions of PAM,ydroquinone, benzoquinone, p-nitroaniline, p-nitrophenol and,2,4-trihydroxybenzene were obtained for comparison with reac-ion mixtures.

. Results and discussion

.1. TOC and HPLC studies

An experiment was carried out to study the photocatalytic oxi-ation of PAM in aqueous solution (50 ppm) with TiO2 (0.2 g/100 mL

f solution), near UV light (�max = 365 nm) and a flow of 100 mL/minf pure oxygen. The composition changes in the reaction mix-ure were monitored by both HPLC and TOC (Fig. 2). ExperimentalPLC results show that PAM is quickly oxidized into other organic

ig. 2. Photocatalytic degradation of paracetamol with TiO2 and UV lightV = 250 mL, C0 = 50 ppm, TiO2 = 2 g/L, four UV lamps, �max = 365 nm, oxygenow = 100 mL/min).

main oxidation metabolite (NAPQI).

compounds. However, TOC results show that some intermediateorganic species remain in the solution for several hours. Compar-ing HPLC and TOC curves, it is apparent that PAM degradates muchfaster than intermediate species.

3.2. Influence of operating parameters on the photocatalyticoxidation of PAM

Various parameters such as amount of catalyst, concentration ofdissolved oxygen, and pH have a strong effect on the photocatalyticoxidation rate of a given organic compound. The photocatalyticdegradation of PAM (50 ppm) in aqueous solution with differentTiO2 catalyst concentrations (0.5, 1.0, 1.5, 2.0, 2.5, 3.0 g/L of solu-tion) was investigated. In the range studied, the effect of TiO2concentration was quite small and within experimental error (5%).These results are in agreement with other investigation on the opti-mal operating conditions for PAM photocatalytic oxidation in water[25].

The oxidation of PAM (50 ppm) in aqueous solution was studiedunder several reaction conditions (Fig. 3). In a first experiment,PAM was irradiated with near UV light (�max = 365 nm) and oxygenwas bubbled through the reaction. It was observed that 16.0 ± 4%of PAM was oxidized after 5 h of reaction indicating that pure

photochemical reactions play some role in the transformation ofPAM to other organic compounds. TOC analysis (Fig. 4) revealedthat PAM is not mineralized under these reaction conditions. Incontrast, when an air saturated aqueous solution of PAM was

Fig. 3. Photocatalytic oxidation of paracetamol under different reaction conditions.(V = 250 mL, C0 = 50 ppm, analysis of samples by HPLC) (�) UV light + oxygen flow,(©) UV light + oxygen saturated + catalyst (�) UV light + oxygen flow + catalyst.

E. Moctezuma et al. / Journal of Hazardous Materials 243 (2012) 130– 138 133

Fig. 4. Photocatalytic mineralization of paracetamol under different reaction con-ditions. (V = 250 mL, C0 = 50 ppm, analysis of samples by TOC) (�) UV light + oxygenflow, (©) UV light + oxygen saturated + catalyst (�) UV light + oxygen flow + catalyst.

iaItdi(Ucheo

Pitiatohh[obp

cptaibtca

Fig. 5. Effect of the initial pH of the solution on the photocatalytic oxidation of para-

of several aromatic compounds were subsequently analyzedby this technique to verify the identity of the components.It was interesting to realize that, in addition to the expected

Fig. 6. Photocatalytic oxidation of paracetamol and subsequent formation anddisappearance of intermediate organic products as a function of time (initial pH

rradiated with UV light in the presence of TiO2 without bubblingny oxygen; the conversion was up to 96% after 5 h of reaction.n this experiment, the oxidation reaction rate decreased withime since the amount of oxygen available to trap free electronsecreased [12]. As a consequence, the complete oxidation of PAM

s also arrested and only 59% of the original reactant is mineralizedFig. 4). When PAM oxidation was performed in the presence ofV light, catalyst and a steady flow of oxygen (100 mL/min), 100%onversion and 72% mineralization were achieved in only fourours of reaction. As expected for a photocatalytic process, a strongnhancement on the oxidation and mineralization reactions wasbserved in the presence of TiO2 (2 g/L).

Concerning the pH, monitoring the photocatalytic oxidation ofAM by HPLC under three different conditions (pH 2, 7.9 and 10),t was observed that the optimal oxidation (Fig. 5) was achieved athe natural pH of the reaction mixture (7.9). Based on our exper-mental results, all the oxidation experiments were performedt this initial pH. Indeed, it has been reported in the literaturehat strongly acidic or basic conditions do inhibit the oxidationf PAM [25]. In slightly basic conditions (pH 7.9), there are moreydroxide ions present on TiO2 surface that are easily oxidized toydroxide radicals, which in turn oxidize paracetamol molecules27,28]. However, in strongly basic conditions (pH 10) the surfacef TiO2 is negatively charged and there is electrostatic repulsionetween the catalyst surface and the negatively charged (pKa = 9.5)aracetamol.

A controversial aspect in the photocatalytic oxidation of organicompounds is whether reactions occur in the homogeneous bulkhase or in the surface of the catalyst [29]. Based on the rela-ionship between oxidation rate and amount of organic compounddsorbed, it has been argued that oxidation at higher pH occursn the bulk aqueous phase [30]. In more recent publications, it haseen suggested that hydroxyl-like chemistry may occur at or nearhe surface of the catalyst with less rigorous requirements for spe-

ific adsorption modes than in a classical electron transfer between

substrate and TiO2 surface [31].

cetamol (V = 250 mL, C0 = 40 ppm, TiO2 = 2 g/L, four UV lamps, �max = 365 nm, oxygenflow = 100 mL/min).

3.3. HPLC studies

The photocatalytic oxidation of PAM at natural pH (7.9) wasinvestigated by HPLC to determine the intermediate organiccompounds formed in the reaction (Fig. 6). Standard solutions

7.9, V = 250 mL, C0 = 120 ppm = 0.793 mM, TiO2 = 2 g/L, four UV lamps, �max = 365 nm,oxygen flow = 100 mL/min).

134 E. Moctezuma et al. / Journal of Hazardous Materials 243 (2012) 130– 138

F(�

ipesta

3

apteriboacabi

dultiamsa

sdan

A sample of PAM was analyzed by IR (Fig. 10, line a) to identifythe characteristic bands [32]. A small alkane section is evidenced byan sp3 C H stretch band at 2882 cm−1. The presence of an aromaticring in the sample is evidenced by several bands in the spectrum.

ig. 7. Photocatalytic oxidation of paracetamol monitored by UV–vis spectroscopyinitial pH 7.9, V = 250 mL, C0 = 120 ppm = 0.793 mM, TiO2 = 2 g/L, four UV lamps,max = 365 nm, oxygen flow = 100 mL/min).

ntermediates hydroquinone and benzoquinone, p-nitrophenol and-aminophenol were also detected in the reaction mixtures. How-ver, p-aminophenol was only detected in trace amounts in theamples collected between 15 and 150 min of reaction. Based onhis observation, this particular reaction was also studied by UV–visnd IR spectroscopy.

.4. UV–vis spectroscopy studies

In order to investigate the formation and eventual disappear-nce of intermediate compounds in the reaction mixture, thehotocatalytic oxidation of PAM was monitored as a function ofime (Fig. 7). The UV–vis spectra of reaction mixtures at differ-nt reaction times present dramatic changes as the photocatalyticeaction proceeds. At the beginning, PAM presents two character-stic absorption bands with �max at 208 nm and 243 nm. The firstand is due to a �–�* transition and the latter to a n–�* transitionf the C O group. After short irradiation times (15–30 min), bothbsorption bands decreased in intensity. However, another bandharacteristic of p-nitrophenol [12] is observed to grow with a �max

t 320 nm. Upon extended photocatalysis, the three absorptionands decreased in intensity and eventually disappeared evidenc-

ng oxidation of aromatic compounds.For comparison, the UV–vis spectra of several possible interme-

iates in water were recorded (Fig. 8). It is well known that PAMndergoes oxidation to some intermediate organic compounds [19]

ike benzoquinone, hydroquinone and benzenetriol, with a charac-eristic UV–vis spectrum. However, the presence of any of thesentermediates in the photocatalytic reaction mixture could notccount for the observed band with a �max at 320 nm. This band isost likely due to the presence of p-nitrophenol since a NO2 sub-

tituent exerts a strong batochromic shift in the (n–�* transition)bsorption band characteristic of phenol [32–36].

Since UV–vis spectroscopy studies indicated the formation and

ubsequent degradation of p-nitrophenol upon photocatalytic oxi-ation of paracetamol in TiO2/UV, p-nitrophenol oxidation waslso investigated under the same reaction conditions. Indeed, a p-itrophenol aqueous solution gives a strong absorption band with

Fig. 8. UV–vis spectra of several possible intermediate reaction products of thephotocatalytic oxidation of paracetamol.

a �max at 320 nm (Fig. 9). Upon photolysis, this band decreases andanother band with a �max at 410 nm grows. Eventually, both bandsdecrease in intensity indicating chemical oxidation of aromaticstructure.

3.5. IR spectroscopy studies

Fig. 9. Photocatalytic oxidation of p-nitrophenol monitored by UV–vis spectroscopy(initial pH 7.9, V = 250 mL, C0 = 60 ppm, TiO2 = 2 g/L, four UV lamps, �max = 365 nm,oxygen flow = 100 mL/min).

E. Moctezuma et al. / Journal of Hazardo

Fig. 10. Photocatalytic oxidation of paracetamol monitored by FT-IR spectroscopy.ST

TCaatP3bocoaa1ss1i

trortw2s

degradation pathways for this oxidation were presented. Based

amples were taken at 0, 2, 4 and 6 h of reaction (initial pH 7.9, C0 = 300 ppm,iO2 = 2 g/L, four UV lamps, �max = 365 nm, oxygen flow = 100 mL/min).

wo sp2 C H stretch bands are located at 3102 and 3163 cm−1.haracteristic C C ring stretch bands are observed at 1611, 1507nd 1439 cm−1. Overtone combination bands appear between 1700nd 2000 cm−1. An out of plane sp2 C H bending at 805 cm−1 is par-icularly characteristic of para-substituted aromatic compounds.henol is evidenced by a broad O H stretching band between000 and 3300 cm−1. This band is rather broad due to hydrogenonding. Furthermore, a strong C O (aromatic) stretch vibration isbserved at 1227 cm−1. This band appears as a doublet since it isoupled to a characteristic C N (aromatic) stretching band that isbserved at 1263 cm−1. The amide functional group is evidenced by

sharp N H stretch at 3325 cm−1. An amide I band is observed as strong C O stretch at 1654 cm−1. An amide II band is observed at536 cm−1 and it is due to a combination of N H bending and C Ntretching vibrations. A weak band is observed at 3163 cm−1 in aecondary amide; it is attributed to a Fermi resonance overtone of536 cm−1 band. In addition, a very broad N H out of plane band

s observed at 684 cm−1.IR analysis (Fig. 10, line b) of an organic extract of the reac-

ion mixture after two hours of photocatalysis shows interestingesults. Only traces of the amide functional group ( NH C O) arebserved in the reaction mixture since its characteristic bands areather weak. In contrast, several C O stretch bands, characteris-ic of carboxylic acids, are observed in the region 1710–1727 cm−1

ith the expected broad band due to O H stretch in the region500–3300 cm−1. The sample still retains some aromatic characterince the bands assigned to the ring are still present. However, it

us Materials 243 (2012) 130– 138 135

is clear that some of the aromatic system has been replaced by amixture of trans-alkene compounds. These latter compounds givestrong C C H out of plane bending signals in the region between1000 and 1100 cm−1. Further evidence of the presence of a mix-ture of alkenes are the several C C stretch bands between 1600and 1660 cm−1. Two sp2 C H stretch bands (due to aromatics oralkenes) are observed at 3016 and 3067 cm−1. In addition, sev-eral sp3 C H stretch bands occur between 2800 and 3000 cm−1

indicating the presence of some alkane framework. The pres-ence of a small amount of aniline in the sample is evidenced byseveral bands. A weak N H stretching vibration is observed at3298 cm−1. A rather strong N H bending vibration of an aromaticamine is observed at 1460 cm−1. Two out of plane N H bendingat 666 and 702 cm−1 are characteristic of aniline. A strong C Nstretching band is observed at 1260 cm−1. This C N band usuallyoccurs at this frequency for aromatic C N bond since resonanceincreases the double-bond character between the aromatic carbonand the attached nitrogen atom. In the analysis of this sample, thepresence of a nitro-substituent is also evidenced by several vibra-tions. One band due to (NO2) asymmetric stretch at 1513 cm−1

and another band due to (NO2) symmetric stretch at 1371 cm−1.Furthermore, C N stretch at 869 cm−1 and aromatic C H out ofplane oscillations at 800 and 757 cm−1 are characteristic of p-nitrobenzene. The overtone region from 1700 to 2000 cm−1 and thestrong band at 800 cm−1 also indicate a para-substituted aromaticcompound.

IR analysis (Fig. 10, line c) of a sample extract after four hours ofirradiation indicated an increase in the amount of carboxylic acidssince a stronger C O stretching band at 1728 cm−1, characteris-tic of this functional group, is observed along with the broad OHstretching band between 3000 and 3400 cm−1. Larger amountsof trans-alkene compounds are present in the sample since theC C H out of plane bending signals in the 1000–1100 cm−1 regionincreased. Furthermore, the signal characteristic of an N H bend-ing vibration (1460 cm−1) of an aromatic amine decreases and thetwo bands (1513 and 1371 cm−1) due to NO2 group increase. Thislatter observation indicates that p-aminophenol is being replacedby p-nitrophenol since the sample still contains some aromaticcharacter.

A sample extract after six hours of irradiation showed veryfew bands upon IR analysis (Fig. 10, line d). No C O band due toa carboxylic acid is observed but very weak bands are observedaround 1600 and 1400 cm−1 that indicate the presence of somecarboxylate salts in the sample. Signals due to sp2 C H and sp3

C H stretch bands along with C C H out of plane bending bandsare still observed in the sample indicating some alkane and alkenecharacter in the sample.

In summary, IR studies indicated that PAM is oxidized by photo-catalysis and it is converted into other organic compounds. Afterseveral hours, the amide and other groups attached to the aromaticring are replaced by a mixture of carboxylic acids and alkenes. Aftertwo hours of irradiation, the aromatic amide has been replaced bya mixture of p-aminophenol and p-nitrophenol. Eventually, onlysmall amounts of alkane and alkene carboxylate salts are detectedin the reaction mixture.

3.6. Intermediates and reaction mechanism

In 2002, the advanced oxidation chemistry of PAM withUV/H2O2 was investigated by a combination of GC–MS, HPLC andNMR [19]. By means of these studies, the first detailed inventory ofintermediate and breakdown products was evidenced and the main

on the structures of the early aromatic intermediate compoundsit was proposed that three initial hydroxylation/oxidation reac-tions compete (Scheme 1). These reactions involve attack of HO•

136 E. Moctezuma et al. / Journal of Hazardous Materials 243 (2012) 130– 138

hanism

rhvt

aadfbdrdpfototbttotrmtuicd

TTccrHt[i

Scheme 1. Paracetamol oxidation mec

adical onto the aromatic ring of PAM to give ortho-, meta- or para-ydroxylation with respect to the OH group. NMR studies in theery early stages of the reaction indicated that hydroxylation athe meta-position is of minor relevance.

To account for some molecular fragments detected in the GC–MSnalysis of reaction mixtures, several degradation pathways werelso proposed for the nitrogen containing breakdown interme-iate products (Scheme 1). Due to the known proclivity of HO•

or electron rich aromatic compounds, it was proposed that thereakdown proceeded through further hydroxylation and oxi-ation of aromatic phenolic compounds [19]. It is important toealize that in this pioneering investigation p-aminophenol was notetected in the reaction mixtures and a deacylation degradationathway was ruled out. However, a remarkable fast degradationor p-aminophenol [19] and 1,4-benzoquinone [12] under thesexidation conditions has been demonstrated. These experimen-al results explain the fact that both intermediate compounds arenly detected in trace amounts in the photocatalytic degrada-ion of aromatic alcohols. Furthermore, it has been demonstratedy several investigations that hydroquinone oxidation proceedshrough different reaction pathways (Scheme 1), generating 1,2,4-rihydroxybenzene and benzoquinone and eventually a mixturef several dicarboxylic acids [33]. A subsequent investigation onhe oxidation of PAM promoted by ozonation and UV/H2O2 waseported [20]. Both oxidative conditions destroyed the PAM aro-atic ring with only partial conversion to carbon dioxide. Under

hese experimental conditions mineralization was obtained onlyp to 30 and 40%, respectively. In this study also, main reaction

ntermediates and products were identified under both oxidationonditions by HPLC and GC–MS. These studies confirmed the oxi-ation mechanism (Scheme 1).

In a more recent study, the advanced oxidation of PAM byiO2/UV in water was investigated [24]. Monitoring the reaction byOC, it was demonstrated that under these conditions PAM is effi-iently removed and converted into carbon dioxide. Based on thehemical structures of compounds detected by GC–MS, a reactionoute for PAM degradation under TiO2/UV system was proposed.

owever, this latter reaction mechanism was similar (Scheme 1)

o the one previously proposed for PAM oxidation under ozonation20] or UV/H2O2 system [19], which is also known to proceed withn situ generation of HO• radicals.

by UV/H2O2 [19,20] or UV/TiO2 [24].

In this study, monitoring the photocatalytic degradation ofPAM by HPLC, it was interesting to realize the formation of smallamounts of p-aminophenol and p-nitrophenol. UV–vis studies gavefurther evidence of the formation and eventual disappearance of p-nitrophenol in the reaction mixtures. Reaction extracts were alsoanalyzed by IR and the presence of small amounts of aromatictype intermediate compounds (p-aminophenol and p-nitrophenol)were also detected. Based on these experimental observations,we propose that a deacylation mechanism (Scheme 2) is anotherplausible not previously reported mechanism for photocatalyticoxidation of PAM. In fact, a deacylation pathway has been reportedto be a major route in the oxidation of the pesticide chlorotoluronthat has a similar structure to PAM with an amide (HN C O) func-tional group attached to an aromatic ring [34].

In photocatalytic processes, HO• has been suggested as the mainoxidizing species in alkaline medium. Moreover, HO• is known toreact mainly by addition to aromatic ring and not by abstraction ofH• from aniline [35]. Therefore, it is quite likely that p-nitrophenolis generated from p-aminophenol by means of an oxidative reac-tion sequence (Scheme 2). First, there is an HO• addition to thearomatic ring, followed by an HO− elimination and subsequentdeprotonation of the radical cation to generate an anilinyl radicalthat is easily oxidized to p-nitrophenol. This remarkable suscepti-bility of intermediate p-aminophenol to oxidize to p-nitrophenolunder photocatalytic conditions could explain the fact that no sig-nificant levels of this intermediate have been previously detectedin the oxidation of PAM [19].

The mechanistic pathway for the oxidation of para-substitutedphenols has been described in a number of contributions[12,33,36,37]. Since 4-chlorophenol leads to a significant numberof intermediate species, several parallel reaction pathways havebeen proposed for the oxidation of this aromatic compound [36,37].Previous studies in the photocatalytic oxidation of p-nitrophenol[12] have shown that there are two major chemical transforma-tions that compete. Hydroxylation in ortho-position to OH group inphenol with eventual formation of nitrocatechol and hydroxylationin para-position to OH group with subsequent NO2

• elimination

to form hydroquinone (Scheme 2). Indeed, our UV–vis studies(Fig. 9) indicated the formation and eventual disappearance of p-nitrocatechol (�max = 410 nm) in the photocatalytic degradation ofp-nitrophenol.

E. Moctezuma et al. / Journal of Hazardous Materials 243 (2012) 130– 138 137

acetam

4

icnecawImaeapttaaic

A

00

R

[

[

[

[

[

[

[

[

[

Scheme 2. Alternative deacylation par

. Conclusions

Our HPLC studies indicated the photocatalytic oxidation of PAMn a water solution by TiO2/UV system to be a highly efficient pro-ess. However, TOC studies indicated that total mineralization isot easily achieved. HPLC analysis of a reaction mixture at differ-nt reaction times indicated the formation of several intermediateompounds such as hydroquinone, benzoquinone, p-aminophenolnd p-nitrophenol. The formation of this latter aromatic compoundas also evidenced following the reaction by UV–vis spectroscopy.

R studies indicated that upon photocatalytic degradation of aro-atic amide (PAM) some aromatic compounds (p-aminophenol

nd p-nitrophenol) are produced and eventually mineralized. Thesexperiments demonstrate that PAM oxidation occurs also throughn alternative deacylation mechanism involving the formation of-aminophenol that is easily oxidized to p-nitrophenol, which inurn is converted into hydroquinone and nitrocatechol. These lat-er intermediates are oxidized to low molecular weight carboxyliccids with some alkane and alkene character. Our investigationslso demonstrated IR spectroscopy to be a useful technique tonvestigate oxidative degradation mechanisms for pharmaceuticalompounds.

cknowledgements

Financial support from CONACyT (grants 28253-U and CB-2008-1-103532) and from UASLP (grants PIFI 2010 and C10-FRC-07-3.04) is gratefully acknowledged.

eferences

[1] C.G. Daughton, Environmental stewardship and drugs as pollutants, The Lancet360 (2002) 1035–1036.

[2] K. Kümmerer, Pharmaceuticals in the Environment: Sources, Fate and Risks,Springer-Verlag, Berlin, Heidelberg, 2008.

[

ol oxidation mechanism by UV/TiO2.

[3] C. Zwiener, F.H. Frimmel, Oxidative treatment of pharmaceuticals in water,Water Res. 34 (2000) 1881–1885.

[4] D.W. Kolpin, E.T. Furlong, M.T. Meyer, E.M. Thurman, S.D. Zaugg, L.B. Barber,H.T. Buxton, Pharmaceuticals, hormones, and other organics waste water con-taminants in U.S. streams, 1999–2000: a national reconnaissance, Environ. Sci.Technol. 36 (2002) 1202–1211.

[5] J.P. Bound, N. Voulvolis, Pharmaceuticals in the aquatic environment-a com-parison of risk assessment strategies, Chemosphere 56 (2004) 1143–1155.

[6] M. Ravina, L. Campanella, J. Kiwi, Accelerated mineralization of drug Diclofenacvia Fenton reactions in a concentric photo-reactor, Water Res. 36 (2002)3553–3560.

[7] T.A. Terness, J. Stuber, N. Herrmann, D. McDowell, A. Ried, M. Kampmann, B.Teiser, Ozonation: a tool for removal of pharmaceuticals, contrast media andmusk fragrances from wastewater, Water Res. 37 (2003) 1976–1982.

[8] M.M. Huber, S. Canonica, G.Y. Park, U. Von Gunten, Oxidation of pharmaceut-icals during ozonation and advanced oxidation processes, Environ. Sci. Technol.37 (2003) 1016–1024.

[9] H. de Lasa, B. Serrano, M. Salaices, Photocatalytic Reaction Engineering,Springer, New York, USA, 2005.

10] O. Legrini, E. Oliveros, A.M. Braun, Photochemical processes for water treat-ment, Chem. Rev. 93 (1993) 671–698.

11] E. Leyva, E. Moctezuma, M.G. Ruiz, L.M. Torres-Martínez, Photodegradationof phenol and 4-chlorophenol by BaO-Li2O-TiO2 catalysts, Catal. Today 40 (4)(1998) 367–376.

12] E. Moctezuma, E. Leyva, G. Palestino, H. de Lasa, Photocatalytic degradation ofmethyl parathion: reaction pathways and intermediate reaction products, J.Photochem. Photobiol. A: Chem. 186 (2007) 71–84.

13] E. Pelizzetti, N. Serpone, Homogeneous and Heterogeneous Photocatalysis, D.Reidel Publishing Company, Boston, USA, 1986.

14] P.F. Schwarz, N.J. Turro, S.H. Bossmann, A.M. Broun, A. Wahab, H. Dürr, A newmethod to determine the generation of hydroxyl radicals in illuminated TiO2

suspension, J. Phys. Chem. B 101 (1997) 7127–7134.15] C.D. Jaeger, A.J. Bard, Spin trapping and electron spin resonance detection of

radical intermediates in the photodecomposition of water at TiO2 particulatesystems, J. Phys. Chem. 83 (1979) 3146–3152.

16] M.I. Litter, Heterogeneous photocatalysis transition metal ions in photocat-alytic systems, Appl. Catal. B: Environ. 23 (1999) 89–114.

17] G.G. Graham, K.F. Scott, Mechanism of action of paracetamol, Am. J. Ther. 12(1) (2005) 46–55.

18] E.J. Corey, B. Czakó, L. Kürti, Molecules and Medicine, John Wiley & Sons, New

Jersey, USA, 2007.

19] D. Vogna, R. Marotta, A. Napolitano, M. d’Ischia, Advanced oxidation chem-istry of paracetamol. UV/H2O2-induced hydroxylation/degradation pathwaysand 15N-aided inventory of nitrogenous breakdown products, J. Org. Chem. 67(2002) 6143–6151.

1 zardo

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[36] X. Li, J.W. Cubbage, T.A. Tetzlaff, W.S. Jenks, Photocatalytic degradation of4-chlorophenol. 1. The hydroquinone pathway, J. Org. Chem. 64 (1999)

38 E. Moctezuma et al. / Journal of Ha

20] R. Andreozzi, V. Caprio, R. Marotta, D. Vogna, Paracetamol oxidation from aque-ous solutions by means of ozonation and H2O2/UV system, Water Res. 37 (2003)993–1004.

21] M. Skoumal, P.I. Cabot, F. Centellas, C. Arias, R.M. Rodríguez, J.A. Garrido, E.Brillas, Mineralization of paracetamol by ozonation catalyzed with Fe2+, Cu2+

and UVA light, Appl. Catal. B 66 (2006) 228–240.22] I. Sires, J.A. Garrido, R.M. Rodríguez, P.I. Cabot, F. Centellas, C. Arias, E. Bril-

las, Electrochemical degradation of paracetamol from water by catalytic actionof Fe2+, Cu2+, and UVA light on electrogenerated hydrogen peroxide, J. Elec-trochem. Soc. 153 (1) (2006) D1–D9.

23] E. Brillas, I. Sires, C. Arias, P.L. Cabot, F. Centellas, R.M. Rodríguez, J.A. Garrido,Mineralization of paracetamol in aqueous medium by anodic oxidation with aboron-doped diamond electrode, Chemosphere 58 (2005) 399–406.

24] I. Dalmázio, T.M.A. Alves, R. Augusti, An appraisal on the degradation of para-cetamol by TiO2/UV system in aqueous medium. Product identification by gaschromatography–mass spectrometry, J. Braz. Chem. Soc. 19 (1) (2008) 81–88.

25] L. Yang, L.E. Yu, M.B. Ray, Degradation of paracetamol in aqueous solutions byTiO2 photocatalysis, Water Res. 42 (2008) 3480–3488.

26] L. Yang, L.E. Yu, M.B. Ray, Photocatalytic oxidation of paracetamol: dominantreactants, intermediates, and reaction mechanisms, Environ. Sci. Technol. 43(2009) 460–465.

27] S.R. Zheng, Q.G. Huang, J. Zhou, B.K. Wang, A study on dye photoremoval in TiO2

suspension solution, J. Photochem. Photobiol. A: Chem. 108 (1997) 235–238.28] C. Galindo, P. Jacques, A. Kalt, Photodegradation of the aminoazobenzene acid

orange 52 by three advanced oxidation processes: UV/H2O2, UV/TiO2 andVIS/TiO2. Comparative mechanistic and kinetic investigations, J. Photochem.Photobiol. A: Chem. 130 (2000) 35–47.

[

us Materials 243 (2012) 130– 138

29] C. Minero, F. Catozzo, E. Pellizetti, Role of adsorption in photocatalyzed reac-tions of organic molecules in aqueous TiO2 suspensions, Langmuir 8 (1992)481–486.

30] M.I. Franch, J.A. Ayllón, J. Peral, X. Domènech, Photocatalytic degradation ofshort-chain organic diacids, Catal. Today 76 (2002) 221–233.

31] X. Li, J.W. Cubbage, W.S. Jenks, Variation in the chemistry of the TiO2-mediated degradation of hydroxy- and methoxybenzenes: electron transferand HOads

• initiated chemistry, J. Photochem. Photobiol. A: Chem. 143 (2001)69–85.

32] G.M. Lampman, D.L. Pavia, G.S. Kriz, J.R. Vyvyan, Spectroscopy, Brooks/Cole,Cengage Learning, Canada, 2010.

33] E. Moctezuma, B. Zermeno, E. Zarazua, L. Torres-Martínez, R. García, Photo-catalytic degradation of phenol with Fe-titania catalysts, Top. Catal. 54 (2011)496–503.

34] W. Bahnemann, M. Muneer, M.M. Haque, Titatium dioxide-mediated photo-catalyzed degradation of selected organic pollutants in aqueous suspensions,Catal. Today 124 (2007) 133–148.

35] M. Canle, J.A. Santaballa, E. Vulliet, On the mechanism of TiO2-photocatalyzeddegradation of aniline derivatives, J. Photochem. Photobiol. A: Chem. 175(2005) 192–200.

8509–8524.37] X. Li, J.W. Cubbage, W.S. Jenks, Photocatalytic degradation of 4-chlorophenol.

2. The 4-chlorocatechol pathway, J. Org. Chem. 64 (1999) 8525–8536.