Chemical Engineering Journal 235 (2014) 264–274

Contents lists available at ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier .com/locate /ce j

Highly efficient magnetically separable TiO2–graphene oxide supportedSrFe12O19 for direct sunlight-driven photoactivity

1385-8947/$ - see front matter � 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.cej.2013.09.043

⇑ Corresponding author at: Environmental Engineering Laboratory, Departmentof Civil Engineering, Faculty of Engineering, University of Malaya, 50603 KualaLumpur, Malaysia. Tel.: +60 379677678; fax: +60 379675318.

E-mail address: saravananpichiah@um.edu.my (S. Pichiah).

Azrina Abd Aziz a, Yeu Harng Yau a, Gianluca Li Puma b, Carolin Fischer c, Shaliza Ibrahim a,Saravanan Pichiah a,d,⇑a Environmental Engineering Laboratory, Department of Civil Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysiab Environmental Nanocatalysis and Photoreaction Engineering, Department of Chemical Engineering, Loughborough University, Loughborough LE11 3TU, United Kingdomc NETZSCH-Geraetebau GmbH, Wittelsbacherstrasse 42, D-95100 Selb/Bavaria, Germanyd Nanotechnology and Catalysis Research Center (NANOCAT), University of Malaya, 50603 Kuala Lumpur, Malaysia

h i g h l i g h t s

� Natural energy active recoverableTiO2/GO/SrFe12O19 photocatalyst wassynthesized.� Resulted in fine band gap (�1.8 eV),

improved utilisation between 400–800 nm.� Achieved �98% mass recovery with

dominant ferromagneticsusceptibility.� Demonstrated superior solar photo

activity by degrading 2,4-DCPsuccessfully.

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

a r t i c l e i n f o

Article history:Received 15 June 2013Received in revised form 6 September 2013Accepted 7 September 2013Available online 19 September 2013

Keywords:TiO2

TiO2/GO/SrFe12O19

Magnetic separationSunlight2,4-DCP degradation

a b s t r a c t

A highly solar photoactive, magnetically separable, TiO2–graphene oxide supported SrFe12O19 (TiO2/GO/SrFe12O19) photocatalyst was synthesised via the solid reaction of silica (SiO2)-coated SrFe12O19 with TiO2

and GO, which were produced by a hydrothermal reaction and Hummer’s method, respectively. Severalaspects of the material chemistry of the prepared photocatalyst were explored: its crystallite phase, par-ticle size, surface morphology, inorganic elemental composition, adsorption–desorption hysteresis, BETsurface area, organic functional group, chemical state of surface, magnetic hysteresis, coercivity (Hci), sat-uration magnetisation (Ms), remanence (Mr), thermal property and visible light absorption analysis. Thesynthesised TiO2/GO/SrFe12O19 exhibited greater ferromagnetic properties (Hci: 2103 Oe; Ms: 3.406E�3 -emu g�1; Mr: 1.642E�3 emu g�1), which further enhanced its re-usability. The incorporation of GO andSrFe12O19 resulted in a drastic reduction in the bandgap energy (1.80 eV). Moreover, this incorporationcontributed for the higher visible light absorption. The photoactivity of TiO2/GO/SrFe12O19 was evaluatedunder direct sunlight for the degradation of 2,4-dichlorophenol (2,4-DCP). The degradation over a periodof 5 h suggested excellent photoactivity.

� 2013 Elsevier B.V. All rights reserved.

1. Introduction This process of heterogeneous photocatalysis was initially reported

The natural environment can be decontaminated from toxicchemicals by photocatalytic oxidation powered by solar light [1].

for the splitting of water with light energy for hydrogen (H2) pro-duction [2–5]. Titanium dioxide (TiO2), an n-type oxide semicon-ductor that has offered superior oxidation efficiency for manytoxic contaminants, is chemically inert and stable in many envi-ronments [6]. However, this semiconductor can be photo-activatedonly by photons in the ultraviolet (UV) region of the solar radiationspectrum (approximately 3–5% of total sunlight) due to its largeband gap (3.20 eV for anatase) [1,7,8].

A. Abd Aziz et al. / Chemical Engineering Journal 235 (2014) 264–274 265

The modification of TiO2 by single atomic layer of sp2 carbonatoms (graphene) can produce a novel electronic configuration sys-tem and a ‘‘dyade’’ structure, which favours the absorption of vis-ible and infrared solar photons (420–800 nm) [9–12]. Furthermore,graphene constrains the recombination of electron–hole (e�/h+)pairs and can increase the charge-carrier transfer rate of e� andthe amount of surface-adsorbed species through p–p bond interac-tions [6,13,14]. Nevertheless, graphene is poorly soluble in wateror polar organic solvents, which makes it difficult to deposit metalor metal oxide on its surface to synthesise graphene-based hybridmaterials [15]. Unlike graphene, graphene oxide (GO) is hydro-philic due to the oxygen-containing functional groups on the sheetsurface [16], which renders the GO a good candidate for supportingmetal or metal oxide particles. Moreover, the surface modificationof GO allows the individual GO to be suspended in both polar andnon-polar solvents. GO has been identified as a suitable candidatefor the charge-trapping layer of photoelectric devices. Due to theexcellent mobility of charge carriers and charge trapping, bothGO and graphene are excellent e- acceptors.

The separation, recovery and reuse of the catalyst from the reac-tor after treatment can be enabled by incorporating the photocat-alyst structure into alkaline earth hexaferrite with magneticproperties [17–19]. This material offers a potential approach forthe recovery and reusability of the photocatalyst [20–22]. Prevent-ing agglomeration and maintaining the durability conditions of thecatalysts is vital for any catalytic process [23]. Previous researchersused hard and soft ferromagnetic materials. For example, magne-tite (Fe3O4) core material, which is described as a soft ferromag-netic material, can straightforwardly be oxidised to maghemite(c-Fe2O3) and decrease the saturation magnetisation (Ms) by 12%after annealing treatment. Consequently, a solitary successful at-tempt has been published [24] that applied only the hard ferro-magnetic material to prevent the core oxidation from adverselyaffecting the separation capability. As a result, zincAnickel (ZnANi)ferrite has been used with the intent of magnetic agitation [25] andmagnetic recovery [26,27]. The principal use of a silica (SiO2) inter-layer in a magnetic photocatalyst was described by Chen and Zhaoin 1999 [28] and shortly elaborated by Beydoun et al. in 2001 [29].They indicated that the photoactivity of the magnetic photocata-lyst could be increased by adding a SiO2 interlayer deposited as ashell around the magnetic particles. The increase in the photocat-alytic activity was attributed to the prevention of the e�/h+ recom-bination rate between ferrite–TiO2 interfaces [28,29].

In this study, we reported the synthesis of a novel hybrid pho-tocatalyst made of GO nanofibers supported with TiO2 and stron-tium ferrite (SrFe12O19) magnetic particles. TiO2/GO/SrFe12O19 is amixture of TiO2 crystals, SiO2/SrFe12O19 prepared by liquid cata-lytic phase transformation and GO prepared by Hummer’s method[9,30]. The photocatalyst can harvest solar photons in the UV andvisible regions of the electromagnetic spectrum and can be mag-netically recovered and reused from an aqueous suspension. Thephotocatalytic activity was investigated by degrading a bio-recalci-trant and toxic pesticide, 2,4-dichlorophenol (2,4-DCP) under tran-sient sunlight irradiation.

2. Materials and methods

2.1. Materials

Tetrabutyl titanate (Ti(OC4H9)4), reagent grade 97%), nitric acid(HNO3, ACS reagent P90%), 2-propanol ((CH3)2CHOH, P99.7%),graphite nanofiber (C, >95% trace metals), potassium permanga-nate (KMnO4, 97%), hydrogen peroxide solution (H2O2, 30% (w/w)in H2O), sodium nitrate (NaNO3, P99%), strontium hexaferritenanopowder (SrFe12O19, 99.8% trace metals basis), sodium hexa-

metaphosphate crystalline ((NaPO3)6, 96%), sodium silicate solu-tion reagent grade (NaOHx(Na2OSiO3)y�zH2O), sulphuric acid(H2SO4, 99.999%), hydrochloric acid ACS reagent (HCl, 36.5–38%)and 2,4-dichlorophenol (Cl2C6H3OH, 99%) were purchased fromSigma Aldrich, Malaysia and used without further purification.Milli-Q water (P18.2 MX cm) was used in all experiments.

2.2. Synthesis of TiO2

Ti (OC4H9)4 was used as a precursor to the synthesis of TiO2 solvia partial hydrolysis and poly-condensation with H2O. Ti (OC4H9)4

was specifically chosen because it has the highest amount of alk-oxy groups, which can prevent oligomerisation and increase therate of hydrolysis and condensation. Approximately 3 mL ofHNO3 was added dropwise as a catalyst to enhance the hydrolysisprocess, and (CH3)2CHOH was used as the solvent. The Ti (OC4-

H9)4:H2O molar ratio was 1:2. The mixture then was stirred vigor-ously by a magnetic stirrer for over an hour. The resulting colouredsolution was calcined at 500 �C for an hour in a muffle furnace [31].

2.3. Synthesis of GO by Hummer’s method

Concentrated H2SO4 was added to a mixture of graphite nanof-ibers (1.0 g, 1 equiv wt) and NaNO3 (0.75 g, 0.5 equiv wt). The mix-ture was cooled to 0 �C. KMnO4, which acts as a strong oxidisingagent (3.0 g, 3 equiv wt), was slowly added in portions to maintainthe reaction temperature below 20 �C. The reaction was increasedto 35 �C and stirred for 30 min, after which Milli-Q water (175 mL)was slowly added. This process produced a large exothermic reac-tion, which raised the temperature to 98 �C. External heating wasintroduced to maintain the reaction temperature at 98 �C for anadditional 15 min. The mixture was then cooled in a water bathto ambient temperatures (�21 �C) for 10 min. An additional1.5 mL of Milli-Q water and 1.5 ml of 30% H2O2 were added, whichproduced another increase in the temperature. After air-cooling,the mixture was filtered by centrifugation (2500 rpm, 1 h) andwashed repeatedly with both 1 M HCl and Milli-Q water. Finally,the product was dried at 60 �C for 24 h to obtain GO.

2.4. Synthesis of TiO2/GO

GO (3 wt%) was dissolved in Milli-Q water, and TiO2 powder(1 wt%) was added to the GO solution. The mixture was sonicatedfor 1.5 h. The mixture was further stirred for 12 h at room temper-ature (without heating) to obtain a homogeneous solution. Theproduct was centrifuged (2500 rpm, 1 h) and dried at 100 �C.

2.5. Synthesis of SiO2/SrFe12O19 dispersion by liquid catalytic phasetransformation

Approximately 100 ml of a (NaPO3)6 aqueous solution (5%)was added to 150 mL of a SrFe12O19 dispersion. Initially, 3.62 gof SrFe12O19 was added to Milli-Q water and diluted to 150 mLto obtain the SrFe12O19 dispersion. Subsequently, 33 mL ofNa2O�3SiO2 solution (10%) were added to the mixed dispersionto obtain a 200 wt% SiO2/SrFe12O19 dispersion. The dispersionwas sonicated for 15 min in an ultrasonic water bath, followedby heating to 90 �C on a magnetic stirrer provided with a heater.The pH value of the dispersion was adjusted to ca. 10.0 by titrat-ing with a H2SO4 solution (5%) under vigorous stirring. Furtherstirring was carried out at 90 �C to obtain a viscous dispersion.A thin SiO2 layer was deposited on the SrFe12O19 nanoparticles.The SiO2-coated SrFe12O19 nanoparticles were repeatedly washedby centrifugation and re-dispersion in Milli-Q water to preventagglomeration [32].

a

c

d

e

b

A

R

R

R

R

R AR A

AR

A

G

G

R

R

RA

R

R AA

RR

SiO2 SrSr

Sr Sr Sr Sr

A

R

Sr Sr

R

RR

R A/Sr RSr

A: Anatase TiO2

R: Rutile TiO2

Sr: SrFe12O19

G: Graphite

Sr

A/A

Sr Sr Sr

A

G/SiO2/

Fig. 1. XRD spectra of (a) TiO2, (b) GO, (c) TiO2/GO, (d) SiO2/SrFe12O19 and (e) TiO2/GO/SrFe12O19.

266 A. Abd Aziz et al. / Chemical Engineering Journal 235 (2014) 264–274

2.6. Synthesis of TiO2/GO/SrFe12O19

Approximately 6 mL of the SiO2-coated SrFe12O19 dispersionand 1 g of TiO2/GO were mixed. A small portion of water was thenadded to alter the mixture to a paste. The mixture was sonicatedfor 15 min, dried, ground and annealed for 30 min at 400 �C to ob-tain the sunlight-driven and magnetically separable TiO2/GO/SrFe12O19 composite photocatalyst.

2.7. Characterisation

Structural/physical properties were characterised by X-ray dif-fraction (XRD) in terms of the crystal system and lattice. The crys-tallography was identified using a Bruker diffractometer (D8) witha Cu Ka emission of wavelength (k) 1.5406 Å. The angular 2h dif-fraction was varied between 10� and 80�, and the data collectionwas completed using a step size of 0.02� with a step time of 1 s.The XRD patterns were compared with the Joint Committee onPowder Diffraction Standards (JCPDS). The crystallite sizes werecalculated using the Debye–Scherrer equation.

To determine the morphology, orientation of materials andinorganic elemental compositions, field emission scanning electronmicroscopy (FESEM), scanning transmission electron microscopy(STEM) (Hitachi SU8000) equipped with an energy dispersiveX-ray (EDX) silicon drift detector (SDD) and high resolutiontransmission electron microscopy (HRTEM) (Carl Zeiss Libra 200FE) analyses were performed. Prior to HRTEM analysis, eachsample was suspended in ethanol (C2H6O), dispersed by sonication(Starsonic, 35) for 15 min and fixed on the lacey formvar carbon-coated copper grid.

The physical phenomenon (surface interaction) or so-calledphysisorption of the materials was investigated by a Brunauer–Emmet–Teller (BET) analysis. BET was performed using a Quanta-chrome Autosorb 6B. The sample was degassed for 5 h at 150 �C.

The paramagnetism and unpaired electron behaviour of TiO2

was evaluated by electron spin resonance (ESR) using a BrukerEMX plus at room temperature.

The organic compound content in the materials was determinedby Fourier transform infrared spectroscopy (FTIR). The spectrawere recorded on a Thermo Scientific iS10 from 500 to 3500 cm�1.

Raman scattering is a useful non-destructive tool to character-ise carbonaceous materials, particularly for studying ordered anddisordered carbon structures. Raman scattering was performedusing a Perkin Elmer Raman Micro-200 at a frequency of radiationof 4.0–7.5E14 Hz and wavelength range between 500 and 2500 nm.

The elemental analysis and quantification of elements were car-ried out using a Kratos X-ray photoelectron spectroscope (XPS) Ul-tra DLD. The binding energy (BE) of C1 s = 284.5 eV was used tointernally calibrate the energy scale.

The magnetochemistry and correlation between the magnetisa-tion and applied magnetic field intensity (MAH loop) were recordedby a MicroMag alternating gradient force magnetometer (AGM).

The mass changes and transformation energies of materialswere investigated by a thermo-gravimetric analysis and differen-tial scanning calorimetry (TGA/DSC) using STA 449 F3Jupiter�.

The optical absorption, especially the shift of the absorption edge,was evaluated using a UV–Vis (Shimadzu UV-2600) equipped with adiffuse BaSO4 reflectance integrating sphere (DRS) at ambient tem-peratures between 200 and 800 nm. The reflectance data were con-verted to the absorption coefficient F (R1) values according to theKubelka–Munk function to calculate the band gap energy.

2.8. Photoactivity evaluation

A stock solution of 1000 mg L�1 of 2,4-DCP was prepared.Approximately 250 ml of a working solution (50 mg L�1) of

2,4-DCP was prepared from the stock solution (1000 mg L�1). Typ-ically, 1 g of prepared photocatalyst was mixed in a batch quartzcylindrical reactor of 500 mL capacity. The suspension was con-stantly magnetically stirred under direct sunlight at the Universityof Malaya, Kuala Lumpur (latitude 101�390E and longitude 3�70N)from 0900 to 1400 h in June, 2012. The solar luminance was mea-sured using a LT Lutron LX-101 Light Meter, and the average lumi-nance over the duration of each run was calculated. The conversionof photon flux (lx) to power (W m�2) depends on the spectrum ofthe lamp. For sunlight (AM 1.5), 1000 W m�2 corresponds to �120000 lx [33]. Two millilitres of the suspension were withdrawn atgiven intervals, centrifuged at 10,000 rpm, filtered through a Milli-pore 0.25 lm and analysed by Ultra Performance Liquid Chroma-tography (Waters Acquity UPLC H-Class; Acquity UPLC BEH C18

column, 100 mm � 2.1 mm, 1.7 lm particle size; 60% acetonitrile(CH3CN), 40% H2O eluent). The TiO2/GO/SrFe12O19 photocatalystwas recycled for another two runs. The first run was carried outwith a prepared virgin photocatalyst under 769.9 W m�2 of sun-light. Run 2 was performed under 775.6 W m�2 of sunlight inten-sity with the recovered photocatalyst. The final cycle wasperformed under 487.2 W m�2 of sunlight intensity. The inconsis-tencies in the solar irradiation were due to its transient natureand local weather condition.

3. Results and discussion

3.1. Structural and morphological properties

The XRD spectra of the prepared photocatalysts are shown inFig. 1. The XRD spectra of TiO2 in Fig. 1a revealed the mixture ofthe TiO2 phases. The peaks appeared at 2h = 25.3�, 48.0�, 62.1�,62.7� and 70.3� and were ascribed to the [101], [200], [213],[204] and [220] Bragg reflections of the body-centred tetragonalTiO2 anatase phase with a fixed volume of 136.313 Å. The JCPDS cardnumber is 21-1272. Conversely, the peaks that appeared at2h = 27.5�, 36.2�, 41.3�, 44.1�, 54.4�, 56.8� and 69.2� were ascribedto the [110], [101], [111], [210], [211], [220] and [301] Braggreflections of the primitive tetragonal TiO2 rutile phase with a fixedvolume of 62.0516 Å. The JCPDS card number is 89-4920. Primitiverhombohedral graphite (JCPDS card No. 75-2978; fixed volume,17.4850 Å) was detected at an angle of 26.9� [002] and 43.3�[100], as shown in Fig. 1b. Fig. 1c shows the peaks that occurred inthe TiO2/GO sample. This figure again indicates the presence of

(a) (b)

(d)(c)

(f)(e)

(g)

GO Nanofiber

TiO2/GO/SrFe12O19

TiO2/GO/SrFe12O19

TiO2/GO/SrFe12O19

GO Nanofiber TiO2

SrFe12O19

O. D.

I.D.

(101)

(211)

(110)

0.2359 nm

0.335 nm

Fig. 2. (a) FESEM; (b) STEM bright field; (c) STEM dark field; (d) HRTEM; (e) lattice fringes and spacing; (f) selected area electron diffraction (SAED) ring pattern; (g) EDX ofTiO2/GO/SrFe12O19.

A. Abd Aziz et al. / Chemical Engineering Journal 235 (2014) 264–274 267

rhombohedral graphite, anatase and rutile TiO2 at specific peaks.The XRD spectra of SiO2/SrFe12O19 (Fig. 1d) show a wide wave peakbetween 17� and 30� of amorphous SiO2, which was intercalated inthe hybrid to avert the direct flow of the magnetic field created bySrFe12O19 into the TiO2. TiO2/GO/SrFe12O19 (Fig. 1e) showed a mix-ture of anatase and rutile TiO2, as well as SrFe12O19. The peaks ap-

peared at 2h = 32.4�, 34.2�, 37.1�, 42.5� and 63.1� and wereascribed to the [107], [114], [203], [206] and [220] Bragg reflec-tions of hexagonal SrFe12O19. The JCPDS card No. was 33-1340. Theanatase phase of TiO2 is well known to be more stable than the rutilephase due to its higher enthalpy of formation [32]. Moreover, thisform has been found to result in higher photocatalytic processes.

Fig. 3. Adsorption–desorption isotherm of prepared photocatalysts.

268 A. Abd Aziz et al. / Chemical Engineering Journal 235 (2014) 264–274

However, the rutile phase of TiO2 formed from the reduced stabilityof the semiconductor anatase phase in this study due to the anneal-ing treatment at 500 �C [31]. This process can contribute to the pho-tocatalysis system where the bulk-generated charge-carrier prevailsbecause only the holes (h+) sufficiently close to the surface will mi-grate before the recombination [7,34]. The presence of SrFe12O19

magnetic particles, on the other hand, is beneficial for easing the cat-alyst recovery. Furthermore, the lack of additional peaks for theTiO2/GO/SrFe12O19 suggests that impurities were not formed in thesynthesised photocatalyst. The crystallite grain size of TiO2 was25.4 nm. This value was calculated by employing the Debye–Scher-rer equation to the full width at the half maximum (FWHM) of thereflections for rutile TiO2 (27.5�). The nanosized TiO2 photocatalystresulted in a quantum effect. The electronic properties of solidschanged as the crystal size was reduced. In general, the quantum ef-fect will dominate and the physical properties will change when thephotocatalyst particles are smaller than 100 nm. Likewise, theabsorption spectra shift to a higher photon energy and develop a dis-crete character as the size of the particles decreases [32].

The FESEM (Fig. 2a) and STEM bright and dark field micrographsof TiO2/GO/SrFe12O19 (Fig. 2b and c) reveal that TiO2, GO and

Fig. 4. ESR spec

SrFe12O19 were well hybridised. The bright field micrograph ofTiO2/GO/SrFe12O19 (Fig. 2b) shows a densely packed structurecomposed of an aggregation of discoidal TiO2 together with theSrFe12O19 crystals, which were deposited on the GO nanofiber.The difference in the energy scattering ability, which depends onthe atomic number (Z) for each element, was demonstrated bythe dark field micrographs (Fig. 2c). The lighter element, carbon(C) (Z:6) contributed by the GO, was colourless in comparison tothe elements contributed by SrFe12O19 and TiO2. This findingproved that Sr (Z:38), Fe (Z:26) and Ti (Z:22) have different energyscattering abilities that depend on Z. A previous study by Aziz et al.,2012 indicated a positive correlation between the values of Z andthe energy scattering ability [32]. HRTEM micrograph fromFig. 2d indicates that the GO nanofiber had and internal diameter(I.D.) of approximately 10 nm and an outer diameter (O.D.) ofapproximately 100 nm. Additionally, the visible lattice fringesand their corresponding FFT images (Fig. 2e) indicate a (001)TiO2 facet with a lattice spacing of 0.236 nm. Conversely, a latticespacing of 0.335 nm indicates the lattice fringes of anatase TiO2.The selected area electron diffraction (SAED) ring patterns shownin Fig. 2f indicate that the TiO2/GO/SrFe12O19 was somewhatcrystalline. In addition, the SAED ring patterns show that theTiO2/GO/SrFe12O19 consisted of a mixture of anatase TiO2 reflectedby the (101) lattice plane and rutile TiO2 reflected by the (110)and (211) lattice planes.

The EDX spectra of the TiO2/GO/SrFe12O19 photocatalyst areillustrated in Fig. 2g. The TiO2/GO/SrFe12O19 consisted of Ti, C, O,Sr, Si and Fe. Each peak was particular to an element, and theywere distributed due to the heating at high temperatures duringthe preparation. The peak intensity indicates the concentration le-vel of the element in the photocatalyst. TiO2/GO/SrFe12O19 mainlyconsisted of 18.07 wt% Ti, 42.64 wt% C and 18.56 wt% O. Theamount of Ti indicates that the material was prepared correctlyand had a high photocatalytic efficiency. The minor elements in-cluded Si (6.39 wt%), Sr (0.58 wt%) and Fe (12.29 wt%). The signifi-cant amount of Fe and Sr encouraged a better recovery of thephotocatalyst with strong ferromagnetic properties.

3.2. BET surface area analysis

Fig. 3 shows the adsorption–desorption isotherms of TiO2, TiO2/GO and TiO2/GO/SrFe12O19. The figures indicate that the

tra of TiO2.

Fig. 5. (A) Functional group analysis of prepared photocatalysts, and (B) Raman spectra of GO.

A. Abd Aziz et al. / Chemical Engineering Journal 235 (2014) 264–274 269

synthesised photocatalysts are categorised as Type IV, commonlyassociated with the presence of mesoporosity. The lower pressureregion of graph is quite similar to Type II, which shows a roundedknee and explains the formation of a monolayer. After the knee, themicropores cease to contribute to the adsorption process. The lowslope region in the middle of isotherm indicates the first few mul-tilayers on the external surface, including meso- and macroporouslayers, before the onset of capillary condensation. The saturationlevel was reached at a pressure below the saturation vapourpressure. The hysteresis indicates capillary condensation in themeso- and macroporous state [32]. The H1 type hysteresis of thesynthesised photocatalysts indicates a cylindrical pore geometryof uniform size. The average pore size was found to be 19.45 nmwith a surface area of 35.40 m2 g�1 for TiO2. TiO2/GO had an aver-age pore size of 1.043 nm with a surface area of 27.29 m2 g�1. TiO2/GO/SrFe12O19 had an average pore size of 15.25 nm with a surfacearea of 70.87 m2 g�1. The small crystallite size might have in-creased the BET surface area of the TiO2/GO/SrFe12O19. This rela-tionship indicates that the number of TiO2/GO/SrFe12O19 pores inthe solid negatively correlates with the pore size and positivelycorrelates with the surface area and adsorption capacity. A highsurface area provides more active adsorption sites for the

photodegradation reaction but will enlarge the amount of sub-strate that comes in contact with TiO2/GO/SrFe12O19 by means ofadsorption. Furthermore, the most active oxidant (�OH) generatedby the photocatalyst, does not migrate very far from the active cen-tres of the TiO2/GO/SrFe12O19. Therefore, degradation will virtuallytake place on the surface of the photocatalyst.

3.3. ESR, FTIR, Raman and XPS analysis

Fig. 4 shows the ESR spectra of the synthesised TiO2 recorded atroom temperature. A strong ESR signal recorded at g = 1.979 indi-cates the presence of bulk Ti3+. Generally, the commercial TiO2 failsto have a strong signal at g = 1.979. The presence of Ti3+ on the TiO2

surface can trap photogenerated electrons (e�) and leave unpairedcharges behind to promote the photoactivity. Hence, Ti3+ promotesthe effective segregation of the e- and cavity and interface chargetransfer. It also lowered the probability of compounding the cavity,but it increased the photocatalytic performance. In addition, oneminor peak for the g-value at 2.004 corresponds to the presenceof O2�. The ESR spectra resulted in a complex pattern becausethe synthesised TiO2 contained anatase and rutile phases.

Fig. 6. XPS analysis of (a) Ti 2p, (b) C 1s, (c) Si 2p, (d) Sr 3d and (e) Fe 2p.

270 A. Abd Aziz et al. / Chemical Engineering Journal 235 (2014) 264–274

The FTIR spectra of GO in Fig. 5A showed characteristic bandsbetween 900 and 4000 cm�1. The broad band between 3000 and3500 cm�1 is the stretching vibration of the OAH bond. The CAOstretching vibration at 1076 cm�1 and vibration of the peroxidegroup at 950 cm�1 were also observed. A less intense band at1271 cm�1 was attributed to the breathing vibration of the epoxygroup. The peak at 1577 cm�1 was attributed to the stretchingvibration of the C@C in the GO. Each group that was present in

the GO was also present in the TiO2/GO/SrFe12O19. However, theamounts were marginal.

A Raman G band at 1425 cm�1 and a typical D band at1332.8 cm�1 were observed for the GO sample (Fig. 5B). The Gband is characteristic of pristine graphite. It was assigned to theE2g phonon of the C sp2 atoms. The D band was attributed to thebreathing mode of the r-point phonons of A1g symmetry, whicharose from local defects and disorders, particularly the defects

b

c

a

(A)

(B)

TiO2/GO/SrFe12O19

Magnetic bar

Fig. 7. MAH hysteresis loop of (a) SrFe12O19, (b) SiO2/SrFe12O19 and (c) TiO2/GO/SrFe12O19.

Fig. 8. TGA/DSC of (a) TiO2; (b) TiO2/GO; (c) TiO2/GO/SrFe12O19.

A. Abd Aziz et al. / Chemical Engineering Journal 235 (2014) 264–274 271

located at the edges or in samples. The presence of the D band indi-cates that GO is highly disordered with a high density of defects.We observed that G band of the prepared GO shifted towards alower wavenumber and that the D band was more intense. Thischange might be due to amount of oxidising agent used to preparethe GO. Furthermore, this oxidation degree of graphite signifies thestructure of the prepared carbon material [35,36]. However, thisoxidation degree does not significantly influence the photocata-lytic activity of the composite.

The fully scanned XPS spectra indicate that the elements Ti, C,Si, Sr and Fe were present in the TiO2/GO/SrFe12O19. The XPS ofthe Ti 2p, C 1s, Si 2p, Sr 3d and Fe 2p core levels were presentin the following forms: (1) Ti 2p region (450–470 eV); (2) C 1s re-gion (280–293 eV); (3) Si 2p region (98–105 eV); (4) Sr 3d region(129–137 eV) and (5) Fe 2p region (708–732 eV). The peaks werefitted to the spectra using a Gaussian–Lorentzian peak shape afterthe background was subtracted. As shown in Fig. 6a, the peaksobserved at 459 eV (Ti 2p3/2) and 465.1 eV (Ti 2p1/2) correspondedto Ti4+. This finding indicates an active compound for theenhancement of visible light photocatalytic activity. The BE peakat 284.6 eV in Fig. 6b can be attributed to the graphitic carbonCAC, C@C and CAH bonds. Furthermore, the BE peak observedat 98.7 eV and 103.2 eV for the Si 2p (Fig. 6c) corresponds tothe elemental Si and SiO2. Fig. 6d and e show the presence ofelemental Sr at BE values of 134.2 eV for Sr 3d. Fe2O3 wasdetected at a BE value of 711.7 eV for the Fe 2p3/2. The magneticproperties of Sr and Fe2O3 help to separate and recover theTiO2/GO/SrFe12O19 after use.

3.4. AGM analysis

The magnetochemistry of the photocatalyst showed a clearhysteresis behaviour in response to an applied magnetic field,which is shown in Fig. 7. Pure SrFe12O19 (Fig. 7A(a)) showed awide hysteresis in the MAH loop with an Oe coercivity (Hci) of3637, which is significantly higher compared to other ferritenanoparticles. As a consequence, a high Hci, remanence (Mr) andsaturation magnetisation (Ms) are retained in the material.Generally, ferrite materials with a high Hci value are called

Fig. 9. (A) UV–vis absorption spectra of (a) TiO2, (b) SiO2/SrFe12O19, (c) TiO2/GO and(d) TiO2/GO/SrFe12O19; (B) Tauc plot of bandgap energy for (a) TiO2, (b) SiO2/SrFe12O19, (c) TiO2/GO and (d) TiO2/GO/SrFe12O19.

272 A. Abd Aziz et al. / Chemical Engineering Journal 235 (2014) 264–274

‘‘hard ferromagnetic’’ materials. This material shows good mag-netic properties that are very stable against magnetic propertylosses [32]. The intercalation of SiO2 into the SrFe12O19

(Fig. 7A(b)) widened the hysteresis in the MAH loop andincreased its magnetic properties (Hci = 3889 Oe, Mr = 8.567E�3

emu g�1 and Ms = 0.01481 emu g�1). However, the Hci, Mr andMs of the TiO2/GO/SrFe12O19 were lower (Hci = 2103 Oe,Mr = 1.642E�3 emu g�1 and Ms = 3.406E�3 emu g�1) and thehysteresis in the MAH loop narrowed (Fig. 7A(c)) due to thedeposition of paramagnetic TiO2 and GO. TiO2 has a very smallmagnetic susceptibility because the e� have unpaired spins,which decreases the interparticle interactions and increases theparticle separation. Such detrimental effects, which are causedby the dipole–dipole interactions between the nanoparticles,contribute to the magnetic anisotropy of the TiO2 material(mixture of antiferromagnetism and ferromagnetic interaction)[37]. Nevertheless, the relatively high Hci of the TiO2/GO/SrFe12O19

implies that the particles lacked superparamagnetic properties.The photocatalyst can therefore be easily recovered from asuspension by applying an external magnetic field, as shown inFig. 7B [38]. Moreover, it can be almost completely recoveredwith a minimum loss for many cycles without diminishing boththe magnetic property and photoactivity. The properties of thismaterial are highly desirable for the treatment of contaminatedwaters.

3.5. Thermal property

Fig. 8a shows a mass loss of �1% between 80 and 120 �C of TiO2

due to the removal of physisorbed water molecules, which resultedin an endothermic effect. At 240 �C, an initial stage or pre-pyrolysisof the oxygen (O2) that contacted functional groups was observed,while the structural water molecule (OAH) was removed at 400 �C.A negligible exothermic effect was observed between 600 and800 �C with an enthalpy of �27.1 J g�1, which was clearly attribut-able to the phase transition of TiO2 from anatase to rutile. Beyond800 �C until 1000 �C, the carbon-containing functional groups werecovered either as CO or CO2, and these functional groups weredominated by the pyrolysis reaction above 1000 �C. The findingwas similar for other composites, but the temperatures wereslightly shifted due to the influence of the materials employed asa composite. The TGA/DSC curve of TiO2/GO is depicted in Fig. 8band shows a weight loss between 700 and 1300 �C. The weight losswas also observed below 700 �C, but the loss was minor. Fig. 8cshows two overlapping endothermic effects (total enthalpy of153 J g�1) with peak temperatures at 934 and 984 �C. A third endo-thermic peak was observed at 1263 �C. Similarly to other compos-ites, the low temperature range was associated with a weight lossdue to the release of moisture; higher temperatures were associ-ated with pre-pyrolysis, structural water removal, the removal ofcarbon-associated functional groups and pyrolysis in the respec-tive temperature zone. At higher temperatures, the thermal effectsand weight loss were attributed to the release of oxygen (O2) andhydroxides (OAH) from the GO. These losses were caused by therelease of moisture and organic impurities at lower temperaturesand the release of the hydroxyl group from the GO at temperaturesabove 800 �C. The endothermic effects were observed for all thesamples from lower (�100 �C) to higher temperatures (�1263 �C)with varied enthalpies, as shown in Fig. 8a–c.

3.6. UV–vis absorption spectra and band gap energy

The UV–vis absorption spectra of the synthesised photocata-lysts were compared and are depicted in Fig. 9A. The synthesisedTiO2 (Fig. 9A(a)) shows a sharp edge at approximately 425 nm,whereas TiO2/GO (Fig. 9A(c)) and TiO2/GO/SrFe12O19 (Fig. 9A(d))display an obvious red shift of approximately 275 nm at theabsorption edge. SiO2/SrFe12O19 (Fig. 9A(b)) enhances the visiblelight absorption by shifting the edge to approximately 375 nm atthe absorption edge. This shift indicates that the presence of GOand the dispersion of SrFe12O19 particles in the TiO2 matrix en-hanced the excitation of TiO2/GO/SrFe12O19 in the visible light re-gion. We ascribed the enhanced photocatalytic performance ofTiO2/GO/SrFe12O19 under a visible light source (sunlight) to theband gap narrowing of the photocatalyst. Fig. 9B shows plots ofthe Kubelka–Munk remission function (i.e. relationship of [ahm]1/2

with the photon energy) corresponding to each spectrum. Theseplots indicate that the band gap of the synthesised TiO2 was2.90 eV (Fig. 9B(a)). The band gap of TiO2/GO/SrFe12O19 in the spec-trum was significantly reduced to 1.80 eV (Fig. 9B(d)). This phe-nomenon should occur due to the direct interaction between theC and Ti atoms on the TiO2/GO/SrFe12O19 surface. This interactionrevealed that the synthesised TiO2/GO/SrFe12O19 had lowered theband gap energy between the conduction (CB) and the valenceband (VB), which means that the excitation occurred at a muchlower energy obtained from the visible spectrum of electromag-netic radiation [32].

3.7. Photoactivity evaluation

The photoactivity was evaluated for the photodegradation of2,4-DCP under sunlight with pristine catalysts as well as after

Fig. 10. Photocatalytic activity of 2,4-DCP; Run 1 under 769.9 W m�2, Run 2 under 775.6 W m�2 and Run 3 under 487.2 W m�2.

A. Abd Aziz et al. / Chemical Engineering Journal 235 (2014) 264–274 273

one and two further consecutive cycles of catalyst recovery and re-use (Fig. 10). The 2,4-DCP was completely degraded by the TiO2/GO/SrFe12O19, while partial degradation was observed when theTiO2 and TiO2/GO were used. The 2,4-DCP was completely de-graded after 180 min for virgin TiO2/GO/SrFe12O19 under769.9 W m�2 of sunlight irradiation (Run 1). After the recycle, theexperiments were carried under slightly higher irradiation(775.6 W m�2). The TiO2/GO/SrFe12O19 was completely degradedafter 150 min (Run 2). The sunlight intensity for the last cycle,Run 3, was varied to 487.2 W m�2. Although the TiO2/GO/SrFe12O19

was recycled twice, the degradation still occurred and completedafter 5 h.

The superior photoactivity of TiO2/GO/SrFe12O19 might have re-sulted from the dual effect of the presence of an anionic carbona-ceous material, the GO and the higher surface area of the TiO2/GO/SrFe12O19 [39–41]. Although the conductivity of the preparedGO was quite low (based on Raman spectra observed, Fig. 5B), itnevertheless played a role as an electron acceptor that marginallycontributed to the interfacial electron-transfer process from TiO2,thus hindering the recombination of charge-carriers [42,43]. Inaddition, the GO is predicted to enhance the adsorption of reactingspecies due to the presence of abundant functional groups at theedge or on the surface of GO.

The SiO2 interlayer inhibited the diffusion of Sr ions into theTiO2, thereby preventing detrimental effects on the performanceof the photocatalyst. The SiO2 also limited the electronic interac-tions between the TiO2 and SrFe12O19 heterojunction while pre-venting the SrFe12O19 from acting as an e�/h+ recombinationcentre. Furthermore, the narrowed band gap energy could effec-tively reduce the recombination of photogenerated charge carriersand enhance the photocatalytic activity. The photogenerated e�

from CB was transferred to the trapping sites in the presence ofsunlight irradiation. This transfer of e� to lattice trapping siteshelps to effectively separate the charge carriers. These trappingsites also benefit by mostly preventing the recombination andfacilitating the charge separation, which activates the catalyst.The trial runs were carried out by separating the catalyst under astrong magnetic field. A mass recovery of �98% was achieved witha minor loss. This loss was expected due to the nanoparticle size.Overall, the present study revealed the outstanding photoactivityfor a sustainable development.

4. Conclusions

The above results demonstrate that the prepared novel TiO2/GO/SrFe12O19 photocatalyst can harvest visible spectrum lightand sunlight for the photodegradation of organic species whileexhibiting enhanced catalyst recovery properties. The degradationof 2,4-DCP concentration was maximised under both bright anddiffused sunlight for both photocatalysts. The results of the presentprepared photocatalyst laid a pathway for a green and sustainablecatalyst design. Thus, the TiO2/GO/SrFe12O19 photocatalyst couldserve as a new generation of photocatalyst for the complete, eco-nomical and productive treatment of toxic wastewater.

Acknowledgments

This work was supported by the University of Malaya ResearchGrant, UMRG (RG091/10SUS) and Postgraduate Research Grant,PPP (PV092/2011A). The first author is grateful to the Ministry ofEducation Malaysia (MOE) for providing financial support.

References

[1] J.S. Lee, K.H. You, C.B. Park, Highly photoactive, low bandgap TiO2 nanoparticleswrapped by graphene, Adv. Mater. 24 (2012) 1084–1088.

[2] D. Zhang, X. Yang, J. Zhu, Y. Zhang, P. Zhang, G. Li, Graphite-like carbondeposited anatase TiO2 single crystals as efficient visible-light photocatalysts,J. Sol–Gel. Sci. Technol. 58 (2011) 594–601.

[3] L. Zhao, X. Chen, X. Wang, Y. Zhang, W. Wei, Y. Sun, M. Antonietti, M.-M.Titirici, One-step solvothermal synthesis of a carbon@TiO2 dyade structureeffectively promoting visible-light photocatalysis, Adv. Mater. 22 (2010) 3317–3321.

[4] J. Liu, H. Bai, Y. Wang, Z. Liu, X. Zhang, D.D. Sun, Self-assembling TiO2 nanorodson large graphene oxide sheets at a two-phase interface and their anti-recombination in photocatalytic applications, Adv. Funct. Mater. 20 (2010)4175–4181.

[5] H.N. Kim, T.W. Kim, I.Y. Kim, S.-J. Hwang, Cocatalyst-free photocatalysts forefficient visible-light-induced H2 production: porous assemblies of CdSquantum dots and layered titanate nanosheets, Adv. Funct. Mater. 21 (2011)3111–3118.

[6] J. Du, X. Lai, N. Yang, J. Zhai, D. Kisailus, F. Su, Hierarchically ordered macro-mesoporous TiO2-graphene composite films: improved mass transfer, reducedcharge recombination, and their enhanced photocatalytic activities, ACS Nano5 (2010) 590–596.

[7] G. Li, K.A. Gray, The solid-solid interface: explaining the high and uniquephotocatalytic reactivity of TiO2-based nanocomposite materials, Chem. Phys.339 (2007) 173–187.

274 A. Abd Aziz et al. / Chemical Engineering Journal 235 (2014) 264–274

[8] G. Liu, Y. Zhao, C. Sun, F. Li, G.Q. Lu, H.-M. Cheng, Synergistic effects of B/Ndoping on the visible-light photocatalytic activity of mesoporous TiO2, Angew.Chem. Int. Ed. 47 (2008) 4516–4520.

[9] D.C. Marcano, D.V. Kosynkin, J.M. Berlin, A. Sinitskii, Z. Sun, A. Slesarev, L.B.Alemany, W. Lu, J.M. Tour, Improved synthesis of graphene oxide, ACS Nano 4(2010) 4806–4814.

[10] P.H. Wöbkenberg, G. Eda, D.-S. Leem, J.C. de Mello, D.D.C. Bradley, M.Chhowalla, T.D. Anthopoulos, Reduced graphene oxide electrodes for largearea organic electronics, Adv. Mater. 23 (2011) 1558–1562.

[11] L. Wang, K.-Y. Pu, J. Li, X. Qi, H. Li, H. Zhang, C. Fan, B. Liu, A graphene-conjugated oligomer hybrid probe for light-up sensing of lectin and E. coli, Adv.Mater. 23 (2011) 4386–4391.

[12] K.K. Manga, S. Wang, M. Jaiswal, Q. Bao, K.P. Loh, High-gain graphene-titaniumoxide photoconductor made from inkjet printable ionic solution, Adv. Mater.22 (2010) 5265–5270.

[13] H. Zhang, X. Li, Y. Li, Y. Wang, J. Li, P25-graphene composite as a highperformance photocatalyst, ACS Nano 4 (2009) 380–386.

[14] Y.H. Ng, I.V. Lightcap, K. Goodwin, M. Matsumura, P.V. Kamat, To what extentdo graphene scaffolds improve the photovoltaic and photocatalytic responseof TiO2 nanostructured films?, J Phys. Chem. Lett. 1 (2010) 2222–2227.

[15] Y. Gao, X. Pu, D. Zhang, G. Ding, X. Shao, J. Ma, Combustion synthesis ofgraphene oxide–TiO2 hybrid materials for photodegradation of methyl orange,Carbon 50 (2012) 4093–4101.

[16] D.R. Dreyer, S. Park, C.W. Bielawski, R.S. Ruoff, The chemistry of grapheneoxide, Chem. Soc. Rev. 39 (2010) 228–240.

[17] W. Fu, H. Yang, L. Chang, B. Hari, M. Li, G. Zou, Anatase TiO2 nanolayer coatingon strontium ferrite nanoparticles for magnetic photocatalyst, Colloid. Surf. A:Phys. Eng. Aspects 289 (2006) 47–52.

[18] W. Fu, H. Yang, M. Li, L. Chang, Q. Yu, J. Xu, G. Zou, Preparation andphotocatalytic characteristics of core–shell structure TiO2/BaFe12O19

nanoparticles, Mater. Lett. 60 (2006) 2723–2727.[19] W. Fu, H. Yang, M. Li, M. Li, N. Yang, G. Zou, Anatase TiO2 nanolayer coating on

cobalt ferrite nanoparticles for magnetic photocatalyst, Mater. Lett. 59 (2005)3530–3534.

[20] V. Polshettiwar, R.S. Varma, Green chemistry by nano-catalysis, Green Chem.12 (2010) 743–754.

[21] S. Shylesh, V. Schünemann, W.R. Thiel, Magnetically separable nanocatalysts:bridges between homogeneous and heterogeneous catalysis, Angew. Chem.Int. Ed. 49 (2010) 3428–3459.

[22] A.-H. Lu, E.L. Salabas, F. Schüth, Magnetic nanoparticles: synthesis, protection,functionalization, and application, Angew. Chem. Int. Ed. 46 (2007) 1222–1244.

[23] S. Xuan, W. Jiang, X. Gong, Y. Hu, Z. Chen, Magnetically separable Fe3O4/TiO2

hollow spheres: fabrication and photocatalytic activity, J. Phys. Chem. C 113(2009) 553–558.

[24] D. Beydoun, R. Amal, G. Low, S. McEvoy, Occurrence and prevention ofphotodissolution at the phase junction of magnetite and titanium dioxide, J.Mol. Catal. A: Chem. 180 (2002) 193–200.

[25] W.L.K. Iv, J. Drwiega, D.W. Mazyck, S.-W. Lee, W. Sigmund, C.-Y. Wu, P. Chadik,Magnetically agitated photocatalytic reactor for photocatalytic oxidation ofaqueous phase organic pollutants, Environ. Sci. Technol. 39 (2005) 8052–8056.

[26] D.G. Shchukin, A.I. Kulak, D.V. Sviridov, Magnetic photocatalysts of the core–shell type, Photochem. Photobiol. Sci. 1 (2002) 742–744.

[27] D.G. Shchukin, E.A. Ustinovich, D.V. Sviridov, A.I. Kulak, Titanium and ironoxide-based magnetic photocatalysts for oxidation of organic compounds andsulfur dioxide, High Energ. Chem. 38 (2004) 167–173.

[28] F. Chen, J. Zhao, Preparation and photocatalytic properties of a novel kind ofloaded photocatalyst of TiO2/SiO2/–Fe2O3, Catal. Lett. 58 (1999) 245–247.

[29] D. Beydoun, R. Amal, J. Scott, G. Low, S. McEvoy, Studies on the mineralizationand separation efficiencies of a magnetic photocatalyst, Chem. Eng. Technol.24 (2001) 745–748.

[30] W.S. Hummers, R.E. Offemen, Preparation of graphitic oxide, J. Am. Chem. Soc.80 (1958) 1339.

[31] A.A. Aziz, C.K. Cheng, S. Ibrahim, M. Matheswaran, P. Saravanan, Visible lightimproved, photocatalytic activity of magnetically separable titaniananocomposite, Chem. Eng. J. 183 (2012) 349–356.

[32] A.A. Aziz, K.S. Yong, S. Ibrahim, S. Pichiah, Enhanced magnetic separation andphotocatalytic activity of nitrogen doped titania photocatalyst supported onstrontium ferrite, J. Hazard. Mater. 199–200 (2012) 143–150.

[33] H. Pettersson, T. Gruszecki, Long-term stability of low-power dye-sensitisedsolar cells prepared by industrial methods, Sol. Energy Mater. Solar Cells 70(2001) 203–212.

[34] D.C. Hurum, K.A. Gray, T. Rajh, M.C. Thurnauer, Recombination pathways inthe Degussa P25 formulation of TiO2: surface versus lattice mechanisms, J.Phys. Chem. B 109 (2005) 977–980.

[35] K. Karthikeyan, V. Murugan, Y. Kyusik, K.S.-J., The chemical and structuralanalysis of graphene oxide with different degrees of oxidation, Carbon 53(2013) 38–49.

[36] C. Chao, C. Weimin, L. Mingce, Z. Baoxue, W. Yahui, W. Deyong, F. Yujie,Synthesis of visible-light responsive graphene oxide/TiO2 composites with p/nheterojunction, ACS Nano 4 (2010) 6425–6432.

[37] S.D. Sharma, D. Singh, K.K. Saini, C. Kant, V. Sharma, S.C. Jain, C.P. Sharma, Sol–gel-derived super-hydrophilic nickel doped TiO2 film as active photo-catalyst,Appl. Catal. A: Gen. 314 (2006) 40–46.

[38] Y. Ao, J. Xu, S. Zhang, D. Fu, Synthesis of a magnetically separable compositephotocatalyst with high photocatalytic activity under sunlight, J. Phys. Chem.Sol. 70 (2009) 1042–1047.

[39] F. Kim, L.J. Cote, J. Huang, Graphene oxide: surface activity and two-dimensional assembly, Adv. Mater. 22 (2010) 1954–1958.

[40] G.L. Puma, A. Bono, D. Krishnaiah, J.G. Collin, Preparation of titanium dioxidephotocatalyst loaded onto activated carbon support using chemical vapordeposition: a review paper, J. Hazad. Mater. 157 (2008) 209–219.

[41] C. Minero, F. Catozzo, E. Pelizzetti, Role of adsorption in photocatalysedreactions of organic molecules in aqueous TiO2 suspensions, Langmuir 8(1992) 481–486.

[42] T.-D. Nguyen-Phan, V.H. Pham, E.W. Shin, H.-D. Pham, S. Kim, J.S. Chung, E.J.Kim, S.H. Hur, The role of graphene oxide content on the adsorption-enhancedphotocatalysis of titanium dioxide/graphene oxide composites, Chem. Eng. J.170 (2011) 226–232.

[43] I.V. Lightcap, T.H. Kosel, P.V. Kamat, Anchoring semiconductor and metalnanoparticles on a two-dimensional catalyst mat. Storing and shuttlingelectrons with reduced graphene oxide, Nano Lett. 10 (2010) 577–583.