Photocatalytic Degradation of Poly(Acrylamide-co-acrylic Acid)

R. Vinu and Giridhar Madras*Department of Chemical Engineering, Indian Institute of Science, Bangalore, 560012, India

ReceiVed: March 04, 2008; ReVised Manuscript ReceiVed: April 29, 2008

Poly(acrylamide-co-acrylic acid) copolymers of different compositions were synthesized and characterized.The copolymers were statistical with a relatively high percentage of acrylamide units, as determined by 13CNMR. Reactivity ratios calculated by the Finemann-Ross and Kelen-Tudos methods showed that thecopolymers were random with a reactivity ratio of rAM ) 3.76 and rAA ) 0.28. The photolytic and photocatalyticdegradation of the copolymers and the homopolymers was conducted in the presence of combustion-synthesizednano anatase titania. The degradation of the copolymer in the presence of combustion-synthesized titania wassignificantly higher than that observed in the presence of commercial titania, Degussa P-25. The degradationwas modeled by using continuous distribution kinetics by following the time evolution of molecular weightdistribution. The degradation follows a two step mechanism, wherein the rapid first step comprises the scissionof weak acrylic acid units along the chain which is followed by the breakage of relatively strong acrylamideunits. The rate constants for the weak and strong links follow a linear trend with the percentage of acrylicacid and acrylamide in the copolymer, respectively. This linear variation can be correlated with a similartrend observed for the activation energies obtained for the pyrolytic degradation of the polymers.

Introduction

Poly(acrylamide-co-acrylic acid) [p(AM-co-AA)] is a watersoluble copolymer, which has nonionic acrylamide (AM) andanionic acrylic acid (AA) units. The water solubility of p(AM-co-AA) is due to the hydrophilic amide and carboxylic acidmoieties in its backbone, which impart sufficient polarity,charge, and hydrogen-bonding ability for hydration. P(AM-co-AA) is a classical copolymer the comonomer reactivity ratio ofwhich is very sensitive to pH. At highly acidic pH values, AAexists in its undissociated form, whereas at basic pH values, itexists as acrylate anion. It was shown1 that the reactivity ofAA is high at sufficiently low pH and low at high pH owing toits dissociation. Thus, by tweaking the pH, the sequence lengthdistribution of AA and AM can be tailored according to theirpotential application. Few reports are available on the synthesis,1–4

hydrolysis,5,6 and thermal degradation characteristics7 of p(AM-co-AA). In the past decade, the polychelatogenic properties ofp(AM-co-AA) in complexing different metal ions8 and itsflocculation efficiency in tannery waste water9 have beenreported. Hence, photodegradation of the copolymer assumesimportance in the disposal of the residual waste after the abovewater purification processes. Recently, spatially cross-linkedp(AM-co-AA) hydrogels with superior antiburn properties havebeen synthesized, and their swelling, network parameters, andrheological properties have been evaluated.10,11

Photolysis of the TiO2 semiconductor results in the generationof valence band holes and conduction band electrons, whichproduce hydroxyl (OH•) and superoxide radicals (O2

-•), whichoxidize the organic matter present in water. Recently, we havedeveloped combustion-synthesized nano anatase titania (CSTiO2),12,13 which has high specific surface area, high surfacehydroxyl content, and low band gap compared to commerciallyavailable Degussa P25 TiO2 (DP 25). It is well proven that CSTiO2 can degrade dyes12 and organics14 at a faster rate compared

to DP 25. CS TiO2 was also shown to catalyze the degradationof poly(bisphenol-A-carbonate),15 poly(ethyleneoxide) (PEO),and PAM16 more effectively than only ultraviolet radiation andDP 25. In the former, the mechanism of photocatalytic degrada-tion was formulated on the basis of the photo-Fries rearrange-ment and nonconcerted cage recombination. In the latter, thephotocatalytic degradation was attributed to the •OH radicalsattacking the R-carbon atom of both PEO and PAM, thusenabling the oxidative scission of the macroradical. Continuous-distribution kinetics provides an elegant way to model themacromolecular scission by assuming the molecular weight tobe a continuous variable. Polymer fission is assumed to takeplace by intramolecular hydrogen abstraction. The hydrogen-abstracted radical produced then degrades into a stable polymerradical of lesser molecular weight and another radical, whichagain undergoes degradation. The rate coefficients for thepolymer degradation, which are assumed to be linear functionsof chain length,17 are calculated from the moment solution tothe population balance equation (PBE).

Although many studies exist on the thermal and photodeg-radation of the homopolymers AM and AA,16,18–21 investigationson the degradation of p(AM-co-AA) are relatively scarce.7 Inthe present study, p(AM-co-AA) of different compositions havebeen synthesized by solution polymerization and characterizedby 13C NMR spectroscopy. The reactivity ratio of AM was kepthigh so that the effect of small amounts of AA units in thecopolymer on photodegradation can be evaluated. The mech-anism of chain scission based on the photogenerated hydroxylradical has been proposed, and the photodegradation wasmodeled by using continuous distribution kinetics. The degrada-tion rate coefficients of the homopolymers and of the copolymerof different compositions in the presence of UV radiation aloneand in the presence of two catalysts, CS TiO2 and DP 25, havebeen determined. The dependence of rate coefficients on pH ofthe reaction system has also been evaluated.

* Corresponding author. Tel: +91-80-2293-2321. Fax: +91-80-2360-0683. E-mail: giridhar@chemeng.iisc.ernet.in.

J. Phys. Chem. B 2008, 112, 8928–89358928

10.1021/jp801887t CCC: $40.75 2008 American Chemical SocietyPublished on Web 07/03/2008

Experimental Section

Materials. Poly(acrylic acid) (Mv ) 450 000) was purchasedfrom Sigma Aldrich. The monomers, AM (S. D. Fine Chem.,India) and AA (Merck, India) and the initiator potassiumpersulfate (S. D. Fine Chem., India) were purified prior topolymerization by standard techniques. Titanium isopropoxide(Lancaster, U.K.) and glycine (Merck, India) used in thepreparation of the catalyst were used as received. Double-distilled, Millipore filtered water was used for all purposes.

Polymer Synthesis and Characterization. P(AM-co-AA)was prepared by free radical polymerization in aqueous medium.Because a high percentage of AM units in the copolymer isdesired, copolymerization was carried out at a basic pH becauseAM is more reactive toward itself than toward AA at high pHvalues. Five different compositions (85/15, 75/25, 60/40, 50/50, and 30/70 by mol%) of AA and AM were taken in 30 mLculture tubes with 5% potassium persulfate as the initiator. Thetotal solution concentration was 1 mol L-1. The solutions wereinitially purged with nitrogen for 10 min and maintained at 50( 0.1 °C in a water bath for 2 h. After the reaction, thecopolymers were precipitated twice in methanol and dried at60 °C until constant weight was obtained. These were thenpowdered and used for degradation experiments. For all thecopolymers synthesized, the conversion was less than 5% forthe applicability of the copolymer equation.

Characterization of the copolymers was carried out by using13C NMR spectroscopy. The proton decoupled 13C spectra werecollected in AMX 400 MHz multinuclear FT-NMR spectrom-eter, with D2O as the solvent. The peak assignment for thevarious carbon atoms in the copolymer was as reported byZurimendi et al.6 The composition of the individual monomerunits in the copolymer was calculated from the integration areasof the resonance peaks corresponding to methylene carbon ofthe carboxylic and amide moiety appearing at δ (ppm) 34.54and 36.01, respectively. Figure 1 shows the dependence ofinstantaneous copolymer composition F on the initial comono-mer composition f. It is clear that the copolymers are statisticalwith a high percentage of AM units. Figure 2 shows theFinemann-Ross and Kelen-Tudos plots for calculating thereactivity ratios. The reactivity ratios using both methods wererAM ) 3.76 and rAA ) 0.28. The number average molecular

weight of the copolymers (henceforth indicated by the percent-age of AA units in the copolymer), C10.3%, C20.4%, C29.3%, C47.

8%, and C58.9% were 4.4 × 105, 4.2 × 105, 4.2 × 105, 4.6 × 105,and 4.2 × 105 g mol-1, respectively. PAM was synthesized bythe method reported elsewhere,16 and the number averagemolecular weight was 4.6 × 105 g mol-1.

Catalyst Preparation and Characterization. Nanosize ana-tase titania was prepared by solution combustion synthesis thatinvolves the combustion of aqueous solutions containing sto-ichiometric amounts of titanyl nitrate and glycine.12 A pyrexdish (300 mL) containing an aqueous redox mixture of sto-ichiometric amounts of titanyl nitrate (2 g) and glycine (0.8878g) in 30 mL of water was introduced into the muffle furnace,which was preheated to 350 °C. The solution initially undergoesdehydration, and a spark appearing at one corner spreadsthroughout the mass and yields anatase-phase TiO2.

The catalyst was characterized by various techniques suchas XRD, XPS, TEM, BET, TG-DTA, FT-IR, and UV spectros-copy, as described elsewhere.12,13 The XRD pattern of CS TiO2

was indexed to the pure anatase phase of TiO2 in the spacegroup I41/amd. The crystallite size determined by using theScherrer formula was 8 nm. XPS showed the presence of Ti in+4 state. TEM studies showed that the crystallites of TiO2 arehomogeneous with a mean size of 8 ( 2 nm, which agrees wellwith the XRD measurements. The BET surface area of thecatalyst was 240 m2 g-1, which is higher than that of DegussaP-25 (50 m2 g-1). Figure SI 1 in the Supporting Informationshows the UV-vis absorption spectra and the TG-DTA of CSTiO2 and DP 25. It is clear that CS TiO2 exhibits two opticalabsorption thresholds at 570 and 467 nm that correspond to bandgap energies of 2.18 and 2.65 eV, respectively. This issignificantly different from the band gap of Degussa P-25 (3.1eV) because of the carbide ion substitution for oxide ion of theform TiO2-2xCxRx, where R is the oxide ion vacancy.14 The TGthermograms of the catalysts indicate that there is a 15.6% totalweight loss for CS TiO2, indicating more surface hydroxylgroups, thus confirming the FT-IR study.

Photochemical Reactor. The photochemical reactor used inthe present study consisted of a jacketed quartz tube of 3.4 cmi.d., 4 cm o.d., and 20 cm length. A high-pressure mercury vaporlamp of 80 W (Philips, India) was placed inside the reactor.This reactor assembly was held concentrically inside a jacketedpyrex glass container of 5 cm i.d., 7.5 cm o.d., and 19 cm height.

Figure 1. Instantaneous copolymer composition (F) versus comonomerfeed composition (f).

Figure 2. Finemann-Ross plot (inset: Kelen-Tudos plot).

Photocatalytic Degradation of p(AM-co-AA) J. Phys. Chem. B, Vol. 112, No. 30, 2008 8929

Cold water was circulated in the annulus of both the reactorand the container to maintain the solution temperature below35 °C because excess temperature might deplete the dissolvedoxygen in the solution. The distance between the source andthe bottom of the vessel was 2 cm to aid better stirring by usinga magnetic stirrer. The lamp radiated predominantly at 365 nm(3.4 eV). The intensity and photon flux calculated by o-nitrobenzaldehyde actinometry22 were 4.64 × 10-6 einstein s-1

and 8.43 mW cm-2, respectively. The degradation was per-formed in an open system wherein the top surface of thephotoreactor was open to the atmosphere. This facilitatesatmospheric air to provide enough oxygen for oxidativedegradation of the polymer solution. Further details are providedelsewhere.12

Degradation Experiments in Solution. The synthesizedcopolymers and homopolymers were degraded in aqueoussolution at its natural pH, which was between 3.4 and 4.4. pH.It was monitored during the course of the reaction, and nosignificant changes in were observed. No degradation of thepolymers was observed without UV radiation. Every time, 100mL of 1 g L-1 solution was exposed to UV light with an optimalcatalyst concentration of 1 g L-1.12 The degraded polymersample collected at every time interval was centrifuged toremove the catalyst particles prior to analysis of molecularweight by HPLC-GPC.

GPC Analysis. The HPLC-GPC consisted of Waters 515isocratic pump, Rheodyne 7725 injector (sample loop, 200 µL),Waters Ultrahydrogel linear column measuring 7.8 mm × 300mm (maintained at 50 °C), Waters R 401 differential refracto-meter, and a data acquisition system. The eluent used was 0.1N NaNO3 at a flow rate of 1 mL min-1. The sample volumeinjected was 500 µL, and the chromatogram was converted tomolecular weight distribution by using PEO narrow standards(Polymer Standards, U.S.A.).

Degradation Experiments using TGA. The pyrolytic deg-radation of the polymers was carried out in a Perkin-Elmer Pyris-Diamond thermogravimetric-differential thermal analyzer in thepresence of inert nitrogen at a flow rate of 150 mL min-1.Different heating rates (2-20 °C min-1) were employed to findthe activation energy of thermal decomposition calculated byFriedmann, Chang, and First Kissinger techniques.23

Theoretical Model. The experimental variation of the numberaverage molecular weight (Mn) without any catalyst and withCS TiO2 and DP 25 indicate that the reduction in Mn issignificant in the initial time period, and as time progresses,the reduction is moderate. Hence, it is logical to assume thatthere are weak links in the polymer chain that get depletedin the initial period, followed by the degradation of the stronglinks. The presence of weak and strong links has been observedfor the degradation of polystyrene.24 The conclusion on thepresence of weak and strong links is based on the comparisonof the degradation profiles of the copolymers with the respectivehomopolymers, as will be discussed later. Adsorption of thebyproducts of the degradation on the catalyst surface can alsoprevent further degradation. However, this is not the case herebecause similar trends of degradation are observed even for thephotodegradation of the copolymers without the catalyst.

The following continuous distribution model is based on theunderstanding that weak links are more susceptible to photo-degradation in the initial time period, after which the stronglinks degrade. A similar treatment has been applied for thethermal degradation of poly(styrene) in solution by Madras etal.,24 where the scission of the weak links was attributed to thehead-to-head or head-to-tail configuration of the phenyl groupand the photocatalytic degradation of poly(bisphenol-A-carbon-ate) by using CS TiO2 by Sivalingam and Madras,15 where thedegradation of weak links was attributed to photo-Fries rear-rangement and concerted cage recombination. No oligomericspecies, as a result of chain-end scission, was observed in theGPC chromatogram. The peak area corresponding to thepolymer mass concentration was constant for all the samples,suggesting a random scission of the copolymers. The reactionfor the binary cleavage of a copolymer of molecular weight xcan be written as24

p(x)98k(x)

p(x ′ )+ p(x- x ′ ) (I)

where, p(x) is the time-dependent molar concentration of thecopolymer. The total concentration of the copolymer is the sumof weak (w) and strong (s) links.

p(x, t)) pw(x, t)+ ps(x, t) (1)

Thus, at any time t, the total molar concentration of thecopolymer is given by

p(0)(t)) pw(0)(t)+ ps

(0)(t) (2)

The PBE for the random degradation of the copolymer in awell stirred batch reactor for the weak and strong links are15,24

Figure 3. Variation of Mn with time for the copolymers when degradedwith CS TiO2.

SCHEME 1: Mechanism of Formation of HydroxylRadicals from TiO2

8930 J. Phys. Chem. B, Vol. 112, No. 30, 2008 Vinu and Madras

∂pw(x, t)

∂t)-kw(x) pw(x, t)+

2∫x

∞kw(x ′ )pw(x ′ , t) Ω(x, x ′ ) dx’(3)

∂ps(x, t)

∂t)-ks(x) ps(x, t)+ 2∫x

∞ks(x ′ ) ps(x ′ , t) Ω(x, x ′ ) dx’

(4)

kw(x) and ks(x) are the strong and weak-link rate coefficients,which are linearly dependent on the molecular weight.17 Hence,kw(x) ) kwx and ks(x) ) ksx. Ω(x,x′) is the stoichiometric kernelthat defines the product size distribution when a polymer ofmolecular weight x′ breaks into two fragments of molecularweight x and (x′ - x). Because the mode of scission in thepresent case is random, Ω(x,x′) ) 1/x′.24

The two integro-differential eqs 3 and 4 are solved byapplying moment operation to the above equations defined by:

p(j)(t))∫0

∞x(j) p(x, t) dx (5)

The zeroth and first moment, which represent the total molarand mass concentrations, are obtained by setting j ) 0 and 1,respectively. The number (Mn) and weight average molecularweights (Mw) are defined as p(1)/p(0) and p(2)/p(1), respectively.

The PBEs after applying moment operation get transformedto

dpw(j)(t)

dt)-(j- 1

j+ 1)kw pw(j)(t) (6)

dps(j)(t)

dt)-(j- 1

j+ 1)ks ps(j)(t) (7)

The zeroth and first moment are given by

dpw(0)(t)

dt) kw pw

(1)(t) anddpw

(1)(t)

dt) 0 (8)

dps(0)(t)

dt) ks ps

(1)(t) anddps

(1)(t)

dt) 0 (9)

The rates of change of first moment for the weak and stronglinks show that their mass concentrations are constant, and hence

pw(1) ) pw0

(1) and ps(1) ) ps0

(1). The time rate of change of molarconcentration can be obtained by solving the two independentvariable ODEs with the corresponding initial conditions pw

(0)(t) 0) ) pw0

(0) and ps(0)(t ) 0) ) ps0

(0). The molar concentrationof the weak and strong links are given by

pw(0) - pw0

(0) ) kw pw(1)t (10)

ps(0) - ps0

(0) ) ks ps(1)t (11)

Adding eqs 10 and 11 and rewriting the expression in termsof total molar concentration and Mn yields

p(0)

p0(0)

- 1) (kw

pw(1)

p0(0)

+ ks

ps(1)

p0(0))t) (kw

′ + ks′ )t (12)

Mn0

Mn- 1) (kw

′ + ks′ )t (13)

The rate constants kw’ and ks’ can be found from the followingequation by assuming that the weak links degrade after a finitetime td; that is, no weak links are present after time td.

Mn0

Mn- 1) kw

′ td + ks′ t (14)

Results and Discussion

Figure 3 shows the evolution of Mn with time when thecopolymers were degraded in the presence of CS TiO2. It isevident that almost 70% of the degradation occurs within a timeperiod of 20 min. A comparison of the Mn profiles suggeststhat PAM and PAA exhibit the slowest and the fastestdegradation, respectively. The degradation profiles of thecopolymers lie in-between the homopolymers and exhibit higherdegradation rates with increasing AA content. The initial fastdegradation of the copolymers can be due to AA, which formweak links in the copolymer chain compared to AM. Hence,we propose that AA units in the chain form weak links that arehighly unstable to UV radiation and degrade within the initial20 min. After this, AM units degrade at a slower rate, owing totheir strong links. This is also consistent with the observationof Smets and Hesbain5 for the acid catalyzed hydrolysis ofp(AM-co-AA), where the hydrolysis proceeds through two steps,in which the first step was 50 times faster than the second one.The mechanistic description of the chain scission is providedin the following section.

Mechanism of Photocatalytic Degradation. The mechanismof degradation of p(AM-co-AA) incorporates the faster depletionof AA units and the slower degradation of the AM units. UVirradiation of the copolymer solution without any catalyst resultsin the generation of polymeric R-radicals, which undergodegradation reactions, albeit with a slower rate. In the presenceof the catalysts, CS TiO2 and DP 25, hydroxyl radicals (•OH)are generated by the mechanism depicted in Scheme 1.12 WhenTiO2 is irradiated with UV light of energy greater than or equalto its band gap, valence band holes and conduction bandelectrons are produced. The valence band holes react with waterin the solution and hydroxyl anions and form hydroxyl radicals.The electron pathway proceeds with the formation of superoxideradicals, which react with protons to form hydrogen peroxideand hydroxyl radicals.

These hydroxyl radicals can react with the copolymer byhydrogen abstraction at the R and positions to the carboxylor amide moieties to form radicals. Ulanski et al.25 haveconfirmed by EPR spectroscopy that -to-R radical conversion

Figure 4. UV-vis absorption spectra of the copolymers duringdegradation.

Photocatalytic Degradation of p(AM-co-AA) J. Phys. Chem. B, Vol. 112, No. 30, 2008 8931

is significantly faster for PAA radicals at low pH values, whichcorrespond to the present experimental investigations. Further-more, for the PAM radicals, Grollmann and Schnabel26 havereported that the reactivity of the methine (R-) carbon to formradicals is four times that of the methylene (-) carbon and eighttimes that of amide nitrogen. Hence, it is reasonable to assumethat R radicals [P•] are the precursors of chain breakage. Oncethese primary radicals are formed, the next step is the formationof polymer peroxy radicals [POO•] by the reaction of [P•] withoxygen. This reaction is very fast and diffusion-controlled.27

These peroxy radicals can combine bimolecularly with anotherradical to form [POO-OOP]. This results in the formation of

polymer oxy radical [PO•] with the exclusion of oxygen. Thefinal step in photo-oxidative degradation is the formation ofscission products from the oxy radical. The scission producesa fragment radical and a nonradical fragment.

In the case of PAA degradation, Kaczmarek et al.21 haveconfirmed the carboxyl radical abstraction in the initial step byUV-vis absorption studies. This results in the formation ofcarbonyl group by the -scission reaction. This was alsoconfirmed in our experiments by the presence of a shoulder inthe 250-300 nm range in the UV-vis absorption spectra ofthe C20.4% and C29.3% copolymer solutions during degradation(Figure 4). This corresponds to the forbidden transition of the

SCHEME 2: Mechanism of Scission of AA Units in the Copolymer

SCHEME 3: Mechanism of Scission of AM Units in the Copolymer

8932 J. Phys. Chem. B, Vol. 112, No. 30, 2008 Vinu and Madras

π-π* orbital and is characteristic of the formation of aldehydeand ketonic moiety, which is consistent with the mechanism.

Scheme 2a,b depicts the mechanism of photodegradation ofAA units in the initial 20 min time period. In the scission ofthe polymer oxy radical [PO•], two situations can arise. Whenthere is an AA unit next to the primary radical, the fragmentradical can either be an R-radical of the subsequent AA or AMunit. This is based on the sequence length distribution of thedyads consisting of AA and AM units. From statistical calcula-tions using the reactivity ratios and feed compositions,28 it isevident that the probability of the formation of AA-AM orAM-AA dyad is greater than that of AA-AA dyad. For theC47.8% copolymer, P(AM-AM) is 30%, P(AA-AA) is 20%,and P(AA-AM) is 50%. It is also found that P(AA-AM)increases with and increase in AA content in the copolymer.Hence, it is more probable to find an AM unit next to AA, andtherefore, the formation of an AM R-radical is more possible.

The degradation mechanism after the initial period is givenby Scheme 3. The polymer oxy radical of PAM degrades bytwo processes. The first one is the scission fragment radical

Figure 5. Variation of [(Mn0/Mn) - 1] for the homopolymers andcopolymers (a) without any catalyst, (b) with CS TiO2, and (c) withDP-25.

Figure 6. Variation of weak-link rate coefficient, kw with percentageAA content.

Figure 7. Variation of strong-link rate coefficient, ks with percentageAM content.

Photocatalytic Degradation of p(AM-co-AA) J. Phys. Chem. B, Vol. 112, No. 30, 2008 8933

produced by the abstraction of a proton from the adjacent AMunit. Because the probability of the formation of AM-AM dyadis high (67% for C18.7%), proton abstraction occurs from theAM unit only. The second one is the hydrolysis of the amidefunctional to form carboxylic acid with the abstraction of NH3.

Effect of Catalysts. Figure 5a-c shows the variation of[(Mn0/Mn) - 1] with the time of irradiation for the homo- andcopolymers without the catalyst, in the presence of CS TiO2

and DP 25, respectively. The strong and weak-link ratecoefficients calculated from eq 14 suggest that CS TiO2 is 40and 60% more efficient compared to the uncatalyzed andcatalyzed DP 25 systems, respectively. It is found that for allthe copolymers, the rate of photodegradation with DP 25 isslightly less than that of the uncatalyzed system. This may bedue to the fact that DP 25, with its 37 nm anatase and 90 nmrutile grains,29 may block the available photoenergy incidenton the polymer species because of its high dispersion charac-teristics. Hence, because of the reduced radiation flux densityin presence of DP 25, lower photoactivity is observed.

The high photoactivity of CS TiO2 compared to the uncata-lyzed and DP 25 systems is due to the high surface hydroxylcontent and reduced band gap of CS TiO2,13 compared to thatof DP 25. In addition to the above observations, the specificsurface area of CS TiO2 is five times that of DP 25, and hence,

CS TiO2 shows a marked increase in photoactivity comparedto DP 25 for the photodegradation of PAM, p(AM-co-AA), andPAA. Furthermore, in a previous work,14 it was proved that thehigher photocatalytic efficiency of CS TiO2 is unaltered by theheat treatment of the catalyst to 400 °C that reduces the surfacearea of the catalyst to 63 m2/g and increases the particle size to16 nm.

Effect of Copolymer Composition. Figures 6 and 7 showthe linear variation of the weak-link and strong-link ratecoefficients with the percentage AA and AM content in thecopolymer, respectively. It is clear that the weak link coefficientincreases with an increase in AA content in the copolymer,which is due to the high photoinstability of AA units. However,the strong-link rate coefficient reduces with increase in the AMcontent. It was earlier suggested that after 20 min, the degrada-tion is mostly due to the scission of the AM units, which aremore stable to photooxidation compared to AA units. Theobserved reduction in rate coefficient can be attributed to thelower probability of finding the AM-AA dyad with an increasein AM content in the copolymer. Hence, for low AM content,it is more probable to find a subsequent AA unit, which candegrade at a faster rate because of its inherent photoinstability.This is supported by the TG weight loss profiles of the polymers.All the polymers exhibit three distinct stages of weight losscorresponding to dehydration, decarboxylation, and/or inter- andintramolecular imidization and main chain scission.19,30 It isevident from Figure SI 2 in the Supporting Information thatthe inclusion of AA units in the copolymer results in fasterdegradation with AA homopolymer exhibiting maximum weightloss.

More interestingly, the activation energy of thermal decom-position for the polymers calculated for the final stage (420-600°C) corresponding to main chain scission decreases linearly withan increase in AA content in the copolymer (Figure 8). Thissuggests that the thermal energy barrier for degradation de-creases with an increase in AA content and hence fasterdegradation. Figures SI 3 and SI 4 in the Supporting Informationshow the Friedmann and Chang plot for the pyrolytic degrada-tion of the polymers at a heating rate of 5 °C min-1. Althoughthe activation energy calculated by the above single heating ratemethods are close to each other, activation energy by theKissinger method is nearly twice as large for all the polymers,although the linear variation with the percentage compositionof AA unit is the same in all the methods. The aboveobservations are further confirmed by the glass transitiontemperature (Tg) of the copolymer, reported by Klein andHeitzmann,2 where the Tg of the copolymers decreases with anincrease in AA content. This means that the mobility (due tobond rotation) and hence the flexibility of the chain increasewith an increase in AA content in the copolymer. Thus, oxygendiffusion occurs more rapidly, resulting in the higher degradationrates observed for the copolymers with higher AA content.

Effect of Solution pH. In order to ascertain whether pH hasany role in the rate of degradation of the homopolymers andcopolymers evaluated thus far, the rate coefficients wereevaluated at neutral and basic pH regimes. Figure 9 shows thevariation of the weak- and strong-link rate coefficients for PAM,PAA, and C47.8% copolymer degraded without any catalyst. Atneutral pH, the rate coefficient is slightly higher for PAM andthe copolymer compared to that of the natural (acidic) pHcondition. The rate of degradation of PAA is doubled at neutralpH, and at basic pH regime, the rate is the lowest for all thepolymers. This is consistent with the observation of Gurkaynaket al.20 for the high-temperature degradation of PAA in solution,

Figure 8. Variation of thermal activation energy with percentage ofAA content.

Figure 9. Effect of solution pH on the rate constants for the degradationof PAM, PAA, and copolymer C47.8%.

8934 J. Phys. Chem. B, Vol. 112, No. 30, 2008 Vinu and Madras

where high rates observed at low pH were attributed to thepolarization induced in the carbon-carbon bond by the un-charged hydroxyl substituents. Hence, our investigation of therate coefficients reflect the highest possible photocatalyticdegradation of the copolymers.

Conclusions

P(AM-co-AA) has been synthesized by free radical copo-lymerization with high reactivity of AM units. To our knowl-edge, this is the first work that reports the photolytic andphotocatalytic degradation of p(AM-co-AA) with titania cata-lysts. The degradation mechanism based on the hydroxyl radicalattack, hydrogen abstraction from the R-carbon and -scission,has been proposed. The copolymer exhibits a peculiar mode ofdegradation, wherein AA and AM units form weak and stronglinks that photodegrade at a faster and slower rate, respectively.The degradation proceeds at a faster rate with the inclusion ofmore AA units in the chain. The strong- and weak-link ratecoefficients are found to be linear functions of the compositionof comonomers, which is analogous to the linear reduction ofthe thermal activation energy with increasing AA content inthe copolymer. CS TiO2 exhibits a higher rate of degradationfor all the copolymers compared to the systems without anycatalyst and with DP 25. This observed high photoactivity ofCS TiO2 can be attributed to its high specific surface area andhigh surface hydroxyl content.

Acknowledgment. The authors thank the NMR ResearchCenter, Indian Institute of Science, for the analysis of thecopolymer samples. R.V. thanks the General Electric Foundationfor a scholarship. G.M. thanks the department of science andtechnology for financial support and Swarnajayanthi fellowship.

Supporting Information Available: SI 1, UV-vis absorp-tion spectra of CS TiO2 and Degussa P 25 TiO2. (inset: TGprofiles of the catalysts); SI 2, normalized TG traces for thecopolymers and homopolymers at a heating rate of 5 °C min-1

in N2 atmosphere at a flow rate of 150 mL min-1; SI 3,Friedmann plot for the copolymers and homopolymers at aheating rate of 5 °C min-1 in N2 atmosphere at a flow rate of150 mL min-1; SI 4, Chang plot with a unit order (n ) 1) ofdecomposition for the copolymers and homopolymers at aheating rate of 5 °C min-1 in N2 atmosphere at a flow rate of150 mL min-1. This material is available free of charge via theInternet at http://pubs.acs.org.

References and Notes

(1) Cabaness, W. R.; Lin, T. Y.; Parkanyi, C. J. Polym. Sci., Part A:Polym. Chem. 1971, 9, 2155–2170.

(2) Klein, J.; Heitzmann, R. Makromol. Chem. 1978, 179, 1895–1904.(3) Shawki, S. M.; Hamielec, A. E. J. Appl. Polym. Sci. 1979, 23, 3155–

3166.(4) Liu, X.; Chen, D.; Yue, Y.; Zhang, W.; Wang, P. J. Appl. Polym.

Sci. 2006, 102, 3685–3690.(5) Smets, G.; Hesbain, A. M. J. Polym. Sci. 1959, 40, 217–226.(6) Zurimendi, J. A.; Guerrero, S. J.; Leon, V. Polymer 1984, 25, 1314–

1316.(7) Maurer, J. J.; Harvey, G. D. Thermochim. Acta 1987, 121, 295–

306.(8) Rivas, B. L.; Pooley, S. A.; Soto, M.; Geckeler, K. E. J. Polym.

Sci., Part A: Polym. Chem. 1997, 35, 2461–2467.(9) Jha, A.; Agrawal, S.; Mishra, A.; Rai, J. P. Iran. Polym. J. 2001,

10, 85–90.(10) Turan, E.; Caykara, T. J. Appl. Polym. Sci. 2007, 106, 2000–2007.(11) Samchenko, Y. M.; Ul’berg, Z. R.; Komarskii, S. A.; Kovzun, I. G.;

Protsenko, I. T. Colloid J. 2003, 65, 78–83.(12) Sivalingam, G.; Nagaveni, K.; Hegde, M. S.; Madras, G. Appl.

Catal., B 2003, 45, 23–38.(13) Nagaveni, K.; Hegde, M. S.; Ravishankar, N.; Subbanna, G. N.;

Madras, G. Langmuir 2004, 20, 2900–2907.(14) Nagaveni, K.; Sivalingam, G.; Hegde, M. S.; Madras, G. EnViron.

Sci. Technol. 2004, 38, 1600–1604.(15) Sivalingam, G.; Madras, G. Appl. Catal. A 2004, 269, 81–90.(16) Vijayalakshmi, S. P.; Madras, G. J. Appl. Polym. Sci. 2006, 100,

3997–4003.(17) Madras, G.; McCoy, B. J. AIChE J. 1998, 44, 647–655.(18) Caulfield, M. J.; Qiao, G. G.; Solomon, D. H. Chem. ReV. 2002,

102, 3067–3083.(19) McNeill, I. C.; Sadeghi, S. M. T. Polym. Degrad. Stab. 1990, 29,

233–246.(20) Gurkaynak, A.; Tubert, F.; Yang, J.; Matyas, J.; Spencer, J. L.;

Gryte, C. C. J. Polym. Sci., Part A 1996, 34, 349–355.(21) Kaczmarek, H.; Kaminska, A.; Swiatek, M.; Rabek, J. F. Ang.

Makromol. Chem. 1999, 261/262, 109–121.(22) Willett, K. L.; Hites, R. A. J. Chem. Educ. 2000, 77, 900–902.(23) Wang, X. S.; Li, X. G.; Yan, D. J. Appl. Polym. Sci. 2000, 78,

2025–2036.(24) Madras, G.; Smith, J. M.; McCoy, B. J. Polym. Degrad. Stab. 1997,

58, 131–138.(25) Ulanski, P.; Bothe, E.; Hildenbrand, K.; Rosiak, J. M.; von Sonntag,

C. J. Chem. Soc., Perkin Trans. 2 1995, 13–22.(26) Grollmann, U.; Schnabel, W. Polym. Degrad. Stab. 1982, 4, 203–

212.(27) Rabek, J. F. Photodegradation of polymers: Physical characteristics

and applications; Springer: Berlin, 1996.(28) Odian, G. Principles of polymerization, 3rd ed.;John Wiley:

Singapore, 2002.(29) Zhang, Z.; Wang, C. C.; Zakaria, R.; Ying, J. Y. J. Phys. Chem. B

1998, 102, 10871–10878.(30) Van Dyke, J. D.; Kasperski, K. L. J. Polym. Sci., Part A 1993, 31,

1807–1823.

JP801887T

Photocatalytic Degradation of p(AM-co-AA) J. Phys. Chem. B, Vol. 112, No. 30, 2008 8935