Microporous and Mesoporous Materials 117 (2009) 233–242

Contents lists available at ScienceDirect

Microporous and Mesoporous Materials

journal homepage: www.elsevier .com/locate /micromeso

Synthesis and characterisation of novel titaniaimpregnated kaolinite nano-photocatalyst

Meng Nan Chong a, Vipasiri Vimonses a, Shaomin Lei c, Bo Jin a,b,d,*, Chris Chow d, Chris Saint b,d

a School of Chemical Engineering, The University of Adelaide, Adelaide, SA 5005, Australiab School of Earth and Environmental Sciences, The University of Adelaide, Adelaide, SA 5005, Australiac School of Resource and Environment Engineering, Wuhan University of Technology, Wuhan 430070, Chinad Australian Water Quality Centre, SA Water Corporation, Bolivar, SA 5110, Australia

a r t i c l e i n f o

Article history:Received 7 March 2008Received in revised form 16 June 2008Accepted 24 June 2008Available online 9 July 2008

Keywords:TiO2

KaolinitePhotocatalystsHeterocoagulationCongo redSol–gel

1387-1811/$ - see front matter � 2008 Published bydoi:10.1016/j.micromeso.2008.06.039

* Corresponding author. Address: School of EarthThe University of Adelaide, Adelaide, SA 5005, Austral+61 8 8303 6222.

E-mail address: bo.jin@adelaide.edu.au (B. Jin).

a b s t r a c t

Nano-sized titanium dioxide (TiO2) has received a great attention in the field of research and develop-ment as a promising photocatalyst to promote the degradation of organic contaminants in water. Oneof the key technical challenges involved in separation and recovery of the photocatalyst particles fromthe water treatment system makes this technology unviable as an industrial process. A novel titaniaimpregnated kaolinite (TiO2/K) photocatalyst was synthesized by a modified two step sol–gel method:hydrolysis of titanium(IV) butoxide and heterocoagulation with pre-treated kaolinite (K) clay. TheTiO2/K photocatalysts were characterised using X-ray diffraction (XRD), scanning electron microscopy(SEM), transmission electron microscopy (TEM) and BET specific surface area measurements (BET). Thephotocatalytic activity was evaluated by the degradation of Congo red in aqueous solution. The TiO2/Kphotocatalyst had a rigid porous layer structure and promising nano-size properties, and demonstratedan enhanced adsorption and photocatalytic ability for the removal of Congo red. The TiO2/K photocatalystcan be easily separated and recovered from the water treatment system. The TiO2/K photocatalyst isexpected to deliver a true engineering solution for an industrial water/wastewater treatment process.

� 2008 Published by Elsevier Inc.

1. Introduction

Heterogeneous photocatalysis employing nano-sized titaniumdioxide (TiO2) has been an attractive research and developmentsubject in water treatment owing to its proven capability in thedegradation of bio-recalcitrant organic contaminants [1–9]. Thehigh specific surface area (SSA) of the photocatalyst as a result ofits nanometer size has rendered its high efficiency in degradingindiscriminately any organic contaminants [10–12]. However,there are still some technical challenges that impede this noveltechnology towards large scale implementation. One of the mostprominent challenges is the post separation and recovery of thephotocatalyst particles after water treatment due to the very fineparticle size and hence the reusability of the particles [1–3,6]. Itwas reported that the small size (4–30 nm) of the TiO2 (DegussaP25) can easily form aggregates in suspension, resulting in a signif-icant reduction in both its effective surface area and photocatalyticefficiency [3].

Elsevier Inc.

and Environmental Sciences,ia. Tel.: +61 8 8303 7056; fax:

In recent years, numerous researches have been devoted insearching for the suitable substrates that could be used as animmobiliser to alleviate the problems involved in the post separa-tion and aggregation. The immobiliser substrates, including glassmatrix, Raschig rings, fibre optics, silica, stainless steel plates, syn-thetic membranes, and clay minerals were used in a number ofinvestigations [9,13–17]. Direct coating of the TiO2 on the internalsurface of the photoreactor has been studied [18,19]. The selectioncriteria for an ideal substrate for the immobilised-catalysts seem tobe a compromise of one another and depend on the reactor config-uration. For instance, inorganic immobilisers such as clay minerals(i.e., zeolites and montmorillonite) are meant for application insuspension reactor system, while stainless steel plates and mem-branes are suitable for continuous operation processes.

TiO2 supported on inorganic clay minerals has received wideattention. The natural structure and adsorption ability of the claymaterials can maintain large SSA, stability and consequently en-hance the photocatalytic efficiency of the photocatalysts [1,3,6,13,20–23]. This is because the photocatalytic reaction occurs on theparticles surface through the generated hydroxyl radicals (OH�).The enhanced adsorption provided by the clay minerals could directmore contaminants to the surface of the particles prior to photoca-talysis. Therefore, the selection for the suitable clay minerals is thekey step for the success in developing the immobilised-catalysts.

O

OH

Si

Al

Fig. 1. The crystal structure of kaolinite, viewed on the left hand side along b-axisand on the right hand side along a-axis. The c-axis is vertical. (Reproduction fromBarrer [3]).

234 M.N. Chong et al. / Microporous and Mesoporous Materials 117 (2009) 233–242

The physical properties of the immobilised-catalysts used in thesuspension reactor system have to be considered. For instance,montmorillonite shows a high capability in adsorption. However,the clay minerals may be swelled rapidly in the suspension system,resulting in particle sedimentation and poor hydrodynamics in thephotoreactor [3,13,24,25].

In this work, Australian kaolinite (K) clay was used as the TiO2

immobiliser. K (Fig. 1) belongs to the kandites mineral group and is

152.76°C

76.24°C399.9J/g

502.72°C

444.47°C239.5J/g

-6

-4

-2

0

2

0 200 400 600

0 200 400 600

Temperatur

Temperatu

Exo Up

1.299% Component 1 (0.3488mg)

On

497.70°C

86

88

90

92

94

96

98

100

102

104

106

108

110

Wei

ght (

%)

Hea

t Flo

w (W

/g)

a

b

Fig. 2. Differential thermal analysis and thermogravimetric profiles for raw kaolinite. (a)

a clay mineral with the chemical composition of Al2Si2O5(OH)4

[26,27]. It is a layered silicate mineral with one layer of octahedral,which reacts with one sheet (Si2)3(OH)2)n, resulting in a two-layersheet structure [24,27]. The siliceous side of K presents a surface ofoxygen, while the aluminous side provides a surface of hydroxylgroups. These double layers are then stacked upon one anotherwith the –OH groups of one such sheet against the oxygen of thenext sheets [28]. The interaction between the stacked layered inK are bonded covalently to each other, rather than Van der Waalsor electrostatic forces [24]. This interaction force makes K suitableas a structurally rigid substrate for supporting and immobilizingthe TiO2. The strong interaction forces make the immobilized par-ticles chemically stable from swelling and can endure high temper-ature of up to 950 oC (Fig. 2). To our knowledge, no investigationusing the K clay as an inorganic support for TiO2 in photocatalysiswas reported in the literature.

This study was to synthesize a novel titania impregnated kaol-inite (TiO2/K) nano-photocatalyst through a modified two step sol–gel method. A series of tests were carried out to characterise thesurface properties of the prepared photocatalyst. The photocata-lytic ability was evaluated through the degradation of Congo red

984.99°C

1144.20°C

800 1000 1200

800 1000 1200

e (°C)

re (°C)

11.83% Component 2(3 .178mg)

Residue:87.34% (23.46mg)

442.51°Cset temp:

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

0.12

[ — — —

— ] D

eriv

. Wei

ght (

%/ºC

)

Heat flow against temperature profile. (b) Weight loss against temperature profile.

M.N. Chong et al. / Microporous and Mesoporous Materials 117 (2009) 233–242 235

in the laboratory photoreactor system. The characterisation studyon the TiO2/K would provide useful information for better under-standing of the adsorption and degradation mechanisms, and therole of K in providing a structurally rigid and thermally stable sup-port, bonding interaction and photocatalytic enhancement for theTiO2/K photocatalysts.

2. Experimental set up

2.1. Materials

Titanium(IV) butoxide (tetrabutyl orthotitanate, purum gradeP97% gravimetric, Sigma–Aldrich), absolute ethanol (AR grade,Labserv Pronalys Australia) and sodium pyrophosphate (Analargrade, BDH VWR, England) were used as received. Raw K materialswere obtained from Unimin, Australia. Nitric acid (AR grade 69wt%, BDH VWR, England), sodium hydroxide (Analar grade, BDHVWR, England) and Congo red (C32H22N6Na2O6S2, Labchem AjaxFinechem, Australia) were prepared to a desired concentration bythe addition of Milli-Q water obtained from Barnstead NanopureDiamond Water ion exchange system with 18.2 M-cm resistivity.

2.2. Pre-treatments of kaolinite materials

The K clay was purified by a series of sedimentation, base treat-ment and outgassed at a temperature in accordance to the differ-ential thermal analysis (DTA–TG), as shown in Fig. 2a and b. TheK has a low cation exchange capacity of 15 meq/100 g. The suspen-sion with 20 g L�1 K was allowed to settle in a 20 cm high jar. Thesupernatant containing approximately <10 lm K was decanted.This supernatant was adjusted to pH 10 by 2 M NaOH, while theresultant mixture was stirred rigorously for 30 min. The K suspen-sion was then filtered, washed and dried at 70 oC for 4 h. The driedK samples were outgassed at 750 oC for 1 h. The particle size distri-bution for the pre-treated K particles is shown in Fig. 3.

2.3. Preparation of TiO2/K photocatalysts

The TiO2/K nano-photocatalysts were prepared by hydrolysisand condensation of titanium(IV) butoxide, heterocoagulationand subsequent heat treatment. These preparation procedures in-volved two steps before heat treatment, where the titania solwas first produced under an acid catalysed condition before het-erocoagulation with the base-treated K under constant hydrother-mal conditions at 37 oC. Twenty-five milliliter of titanium(IV)butoxide was mixed with 30 mL of absolute ethanol under vigor-ous magnetic stirring. This mixture was denoted as mixture A.

Fig. 3. The particle size distribution cu

The 60 mL diluted nitric acid (HNO3) was then added drop wiseinto the mixture A. The mixture A was stirred constantly untilthe dispersion became first milky white and subsequently trans-parent homogeneous sol with no visible precipitation. The stirringwas continued for 30 min until a transparent sol with certain vis-cosity was attained.

As the second step of the synthesis process, the K particles weredispersed in 100 mL of Milli-Q water and the resultant K suspen-sion was placed in a three necked round bottom flask. This flaskwas then set in a water bath set at 37 oC. The K suspension wasmixed completely. The resultant homogeneous sol from the firststep of the synthesis was then added drop wise into the K suspen-sion with the aid of a separatory funnel. The mixture in the flaskwas then continuously stirred for 4 h before the mixture wascooled down to the room temperature. The resultant heterocoagu-lated mixture, denoted as mixture B, was then aged for 13–16 h,depending on the concentration of HNO3 used. The mixture Bwas filtered and washed repeatedly three times with Milli-Q water,and then dried at 65–70 oC for 2–4 h. A 3 h heat treatment was car-ried out at an increasing rate of 2 K/min to desired temperature.After calcination, the samples were cooled down to room temper-ature. The newly prepared photocatalyst of TiO2/K was stored inthe desiccators before use. For the bare TiO2, the K suspension stepwas taken off and the resultant mixture A was also stirred in thewater bath at 37 oC before being washed, filtered, dried and cal-cined under the same conditions as those stated above.

2.4. Characterisations of TiO2/K photocatalysts

The particle size distribution of the K sample was measuredusing a particle sizer (static light scattering laser diffraction)instrument (Malvern Mastersizer 2000), which covers the mea-surement range from 0.02 to 2000 lm. For this size measurement,the K particles were dispersed in water with sodium pyrophos-phate added with agitation and sonication until a stable dispersionwas achieved.

The differential temperature analysis (DTA) coupled with ther-mogravimetric analyser (TGA) (TA Instruments) were used to ana-lyse the weight loss of K as a function of elevated temperature. TheK sample was heat treated in the range of room temperature up to1200 oC at a heating rate of 10 oC/min under high purity nitrogengas to avoid any possible oxidation by atmospheric oxygen or air.

Scanning electron microscopy (SEM) images were obtainedusing the Nanoscope II Electrochemical electron microscope at anaccelerating voltage of 5 kV. The images of transmission electronmicroscopy (TEM) were obtained using Philips CM-100 TEM at anaccelerating voltage of 100 kV. The prepared TiO2/K photocatalysts

rve for the pre-treated particles.

236 M.N. Chong et al. / Microporous and Mesoporous Materials 117 (2009) 233–242

were suspended in ethanol (approximately 0.01% (w/v)), followedby ultrasonic treatment before being dropped onto the measure-ment copper grids of 2 mm in diameter. The particle size for thecoated TiO2 crystallites were estimated using the image analysissoftware analySIS�.

The Brunauer–Emmett–Teller (BET) specific surface area of theTiO2/K photocatalysts under different HNO3 concentration weredetermined by a micromeritics gas adsorption analyser (GeminiType 2375) at 77 ± 0.5 K in liquid nitrogen. Prior to the surface areameasurements, the sample vessels loaded with ca. 0.5–1.0 g werepre-treated in vacuum at 378.15 K and evacuation pressure of50 mTorr for overnight. The adsorption isotherms were then ana-lysed for the corresponding specific surface area using the BETequation.

X-ray diffraction (XRD) measurements were performed on aPhilips PW Diffractometer (Co X-rays k = 1.7902 Å) over the rangeof 5–90� 2h for the solid K powder samples. Samples were spikedwith 10 wt% of zinc oxide before analysis to facilitate thecalculation for the weight percent of the amorphous phase. The ba-sal distances, dL, were calculated from the Bragg reflections usingthe automated powder diffraction software. The average crystallitediameters of the particles, Dhkl, were obtained by using the Debye–Scherrer equation [Dhkl = ka/b cosh], where b is the line broadening(b = bs–bo, where bs and bo are the half-widths of the XRD peak ofthe standard), k is related to the crystalline shape and taken to be0.89 with Warren’s correction for instrumental broadening, and aand h are the radiation wavelength and Bragg angle, respectively.A standard for this instrumental measurement based on singlecrystal anatase TiO2 was used as a mean for calibration.

2.5. Photocatalytic experiments on Congo red

Congo red was used as a surrogate indicator to evaluate thephotocatalytic degradation ability of the newly synthesized TiO2/K nano-photocatalysts. Congo red, a secondary diazo dye, is a recal-citrant azo dye found in textile wastewater. It has been involved inseveral photocatalysis studies including those using TiO2 andhence is appropriate in making comparisons for the current study.The chemical structure of the Congo red is as shown in Fig. 4.

The photocatalytic degradation of the Congo red was performedin a crude box reactor system. The reaction solution was placed in abeaker and stirred using a magnetic stirrer. An 11 W UV light(Davis Ultraviolet, Australia) was positioned on the top of the bea-ker with an intensity of approximately 300 lW/cm2 on the surfaceof the swirling fluid. A simple sampling unit was connected to thebeaker where a 50 mL syringe was used to withdraw the reactionsolution into a connected screw cap vial for sample collection.The air ventilation in the reactor box was controlled using a vac-uum pump, where the air inside the box was replenished at a rateof 2 L/min through a 1 cm diameter opening on the both side of thebox.

The reaction solution contained 40 mg L�1 of Congo red. Thereaction solution was then stirred vigorously for 30 min on themagnetic stirrer to ensure homogeneous mixing. A triplicate ofthe reaction solution was then sampled where the absorption spec-

Fig. 4. Chemical structure of Congo red (a secondary diazo dye).

tra for the solution was scanned using UV–visible spectrophotom-eter (Helios Gamma, England). From the absorption spectra scanranging from 200 to 700 nm, it was found that the most intensepeak for the reaction solution was at 496.5 nm. Then a series ofabsorption spectra measurements were conducted for each of thequarterly diluted reaction solution in order to plot the absorbance– concentration profile. Each measurement for the diluted pointswas repeated three times to minimise measurement errors.

Then 100 mL of the reaction solution was introduced into thebeaker placed on the magnetic stirrer with 1 g of TiO2/K photocat-alysts. The suspension was allowed 30 min dark adsorption beforethe UV light was switched on. The samples were collected every1 h and centrifuged at 5000 rpm for 10 min. The supernatants werethen filtered using Millex VX filter (Millipore 0.45 lm) to ensurethe reaction samples were free of TiO2/K solids. Finally, the concen-tration of Congo red in the filtered supernatant was measured andcalculated using the UV–visible spectrophotometer (at 496.5 nm)and the absorbance – concentration profile. The sampling and sam-ple tests were carried out in triplicate.

3. Results and discussion

3.1. Pre-treatments of kaolinite

Natural clay minerals consist of surface bounded water, zeoliticwater, other chemicals and impurities [13,25]. These impuritiesmight affect the interlayer pore space and hence the total effectivesurface area that are necessary for the subsequent heterocoagula-tion process with titania sol [24]. Some foreign chemicals existingin the kaolinite might induce internal reactions during the processof the catalyst preparation, consequently affecting the propertiesand photocatalytic activities of the TiO2/K. [25]. An appropriatetreatment to remove these impurities from the raw kaolin clay isneeded.

The natural K materials were subjected to thermogravimetricanalysis (DTA–TG) in order to obtain the weight loss profile againstthe temperature gradient. The DTA–TG curves can be useful todetermine the maximum applicable outgassing temperature to re-move these impurities. It is important to note that outgassing Kmaterials beyond or above the maximum applicable temperaturecould cause an abrupt decrease in the pore volume owing to theirreversible collapses of the clay structure [24,25]. Both Fig. 2aand b show the thermal analysis – thermogravimetric (DTA–TG)curves. The first weight loss of the K as shown by the broad endo-thermic peak was found at 150 oC, which was attributed to removethe surface bound water. The second endothermic peak at 503 oCcorresponds to the dehydroxylation of the single layer of water be-tween each pair of silicate sheets in the K. DTA–TG curves showthat an exothermic peak onset exists at 985 oC, corresponding tothe maximum applicable outgassing temperature. This tempera-ture denoted the onset of the irreversible collapse to the K struc-ture and hence outgassing of the K particles should not be donebeyond or above this temperature. Therefore, the outgassing tem-perature was chosen at 750 oC for 1 h.

The BET specific surface area (SSA) measurements were carriedout to examine the effect of the chosen outgassing temperature onthe physical properties of the K. The result shows that the out-gassed K particles have a lower SSA than the K particles whichwere dried at 70 oC. The SSA of the outgassed K decreased from15.24 ± 0.02 m2/g to 14.54 ± 0.09 m2/g. This SSA decrease may bethe result of collapse in the oxide layer between the silicate sheetsin the K structure. The c spacing of the K structure shrank from10 Å to 7.2 Å when the single water layer was driven off. Barrer[24] reported that the tunnels in the outgassed K will be impededas a consequence of the outgassing treatment. This impedance only

M.N. Chong et al. / Microporous and Mesoporous Materials 117 (2009) 233–242 237

allows small polar molecules to replace the water layer but the finefibrous habit of the crystals sustains a large surface area [24].Hence, the resultant outgassed K particles present an enhanced ri-gid structure with large surface area that can minimise the possi-bility of chemical intercalation and promotes the reuse potentialfor the prepared photocatalysts.

3.2. Synthesis of titania sol

The synthesis of the titania sol was carried out by the modifiedtwo step sol–gel method. Titanium(IV) butoxide was used as thetitanium precursor in the initial gelation step of the acid catalysedsol–gel process. We found that the transparent titania sol was notformed using 1 M HNO3, which was the recommended HNO3 con-centration in previous reports [29,30]. Suh et al. [31] reported thatthe formation of a transparent titanium aerogel only occur in asmall range of HNO3 concentration. The treatment of titania coatedsubstrates with certain acid concentration can improve the photo-catalytic ability [4,32,33]. It is important to find a suitable HNO3

concentration to control the formation of titanium sol [5]. A studyon the impact of the HNO3 concentration on the preparation of thesol–gel and photocatalytic activity of the TiO2 has not been re-ported in the literature.

Fig. 5 indicates that the HNO3 concentration ranging from 0.05to 1 M has a varying impact on the formation of sol–gel. The opti-mal HNO3 concentration range of 0.25–0.30 M can contribute tothe formation of transparent titanium sol (i.e., denoted as cleargel (hard)) after aging for 13–16 h. The HNO3 plays an importantrole in controlling the rates of hydrolysis and condensation reac-tion of the titanium alkoxides precursor, so that the Ti–OH mono-mers can be slowly condensed associated with the growth ofstrong Ti–O–Ti polymeric network [31]. When an excessiveamount of HNO3 was used, the hydrolysed solution of titaniumprecursors exhibited an extensive degree of fluidity even afteraging for 13–16 h [34]. The excessive use of HNO3 can retard thecondensation process and further affects the monomer cluster –cluster growth to form the Ti–O–Ti polymeric network in the tita-nium sol [7,31,34]. However, once a low concentration of HNO3

was used, the gelation rate increased until a precipitation occurred.It was note that a suitable HNO3 concentration contributes to notonly the formation of suitable homogeneous transparent titaniumsol but also the gelation behaviour and textural properties. Fig. 5shows that the titanium sol exhibited four gelation behavioursand textural properties. A HNO3 concentration lower than the opti-mal range of 0.25–0.30 M resulted in increasing the gelation rateand forming a precipitated gel. Suh et al. [31] reported that the pre-

0 0.4

0Ti

mol/LH0.2

Fig. 5. The textural properties of titania sol un

cipitated gel was due to the presence of precipitates in the poly-meric gel network. The hard gel performed a low degree offluidity and has a negative impact on the subsequent heterocoagu-lation process with the pre-treated K particles. Therefore, the con-centration of 0.25–0.30 M HNO3 was used in the followingexperiments.

It was observed that the photocatalysts prepared with differentHNO3 concentrations performed variable initial absorption abilityfor the Congo red. Fig. 6 shows that the TiO2/K prepared with0.28 M HNO3 contributed to an optimal 30% removal of Congored by initial adsorption. This adsorption ability is generated bythe strength of the polymeric gel network formed under differentmolarity of HNO3. This strength can be further enlarged whenthe TiO2/K is heat treated, resulting in TiO2/K particles with vari-able particle sizes and SSA [35–37]. The different network struc-ture of the TiO2/K can be proven through the direct comparisonof SEM images e.g., in Fig. 7a and b, which shows the surfaceimages of the prepared photocatalysts under the optimal 0.28 Mof HNO3 and the pre-treated K, respectively. The original prototypeof the K surface was in a flat sheet pattern. A highly porous struc-ture was observed after the attachment of TiO2 crystallites. The re-sults shown in Section 3.3 revealed that the K prepared under theoptimal molarity of HNO3 had a significant larger SSA of35.24 ± 0.11 m2/g than the pre-treated K (14.54 ± 0.09 m2/g), asstated in Section 3.1.

Robert et al. [38] stated that the resultant titanium sol shouldhave its pH adjusted to pH 3 with the addition of 1 M sodiumhydroxide (NaOH). We found that the pH value for the resultanttitania sol was 1.76. The use of NaOH to adjust the initial pH upto 3 likely contributed to the formation of large dense TiO2 crys-tallite after calcination. This was evidenced from the peak broad-ening in X-ray diffraction. Although the mechanism for theformation of the dense crystallites was unclear, it can be seen thatthe pH adjustment had an important impact on the formation ofdense crystallites, the porosity of the structure, and consequentlythe adsorption capacity of the TiO2/K. Thus it was proposed herethat the pH adjustment of the titania sol should not be carriedout to preserve the high specific surface area, but instead themolarity of HNO3 or other preparation conditions should bealtered.

It can be concluded that the titania sol required for the subse-quent heterocoagulation with pre-treated K can be synthesised un-der the optimal HNO3 range of 0.25–0.30 M without further pHadjustment with NaOH. This is certainly beneficial as the enhancedadsorption capability promotes the surface reaction and hence thephotocatalytic efficacy of the TiO2/K.

0.6 0.8 1NO3

PrecipitationGel Precipitated (Hard) Clear Gel (Hard)Very Clear Gel (Soft)No Gel

der different concentration of nitric acid.

Fig. 6. Initial adsorption tests of Congo red (40 ppm) for the photocatalysts prepared under different molarity of nitric acid. The other preparation conditions were constant;volume ratio Ti: EtOH: HNO3: 5:6:12. The K suspension loading used is 10% (w/v) (i.e., 10 g in 100 mL water). Calcination temperature is 600 oC.

Fig. 7. SEM images for sample K (a) After outgassed; (b) After coated with TiO2

(0.28 M HNO3).

238 M.N. Chong et al. / Microporous and Mesoporous Materials 117 (2009) 233–242

3.3. Heterocoagulation process

The TiO2/K nano-photocatalysts were synthesised using thepre-treated K materials. Heterocoagulation process constitutesthe second step of the preparation of the titanium sol. The titaniumsol was slowly added into the K suspension. This procedure isanticipated to promote interaction and attachment of the posi-tively charged titanium sol onto the slightly negatively chargedpre-treated K particles [13].

Through the heterocoagulation process, it was noted that thequantity of the K added in the suspension has a direct impact onthe property of the photocatalysts. The quantity of the settled Kparticles was independent of the amount of K in the suspension.Two separate phases were clearly distinguished in the heterocoag-ulated mixture, representing coated particles and independentlyprecipitated K particles.

K suspension was tested with three different K loadings duringthe preparation. At a low concentration of 5% (w/v) (i.e., 5 g of K in100 mL of Milli-Q water), no individual precipitation of K particleswas found in the heterocoagulated mixture after aging. Thisshowed that the K particles could interact fully with the titaniumsol without generating any excess K particles. However when thecoated particles were examined under the TEM, massive crystalstructures with relatively thick and inconsistent coated layer couldbe observed (Fig. 8a and b). It was believed that the bulkiness ofthe crystals led to a decrease in the total SSA and the reproducibil-ity of the synthesis method. Size quantification on the thickness ofthe coated layer was carried out by taking the images of a specificparticle by tilting the focus angles on the TEM. The inhomogeneityof coating layer has affected the thickness measurement in thiscase. Hence, a higher concentration of K in the suspension ispreferred.

The suspension with 10% (w/v) and 15% (w/v) of the K weredetermined for controlling the thickness of the coated layer onthe K particles. This was based on the rationale that an increasein the K particles could provide more crystal nucleation sites forthe titanium sol, resulting in a decrease in the thickness of thecoating layer [39,40]. At 10% (w/v), a uniform layer of coatedTiO2 thickness was formed on the surface of the K particles. Thecoated layer was made up by small subunits of TiO2 crystals withan average particle size distribution of 7 nm (Fig. 8c and d),

Fig. 8. TEM images for photocatalysts prepared under different loading in kaolinite suspension. (a and b) 5% w/v; (c and d) 10% w/v; (e and f) 15% w/v.

M.N. Chong et al. / Microporous and Mesoporous Materials 117 (2009) 233–242 239

depending on the extent of the heat treatment condition. As a re-sult of the smaller subunits of TiO2 crystal, the formation of theuniform coating layer contributed to an increase in the SSA. Theseresults were shown in Section 3.4. A higher K concentration of(15% w/v) resulted in a relatively distant dispersion of crystalgrowth as presented in Fig. 8e and f, in which there were onlya few crystals formed on a K particle with inconsistent particlesize around 6–10 nm. This is due to an increase in the crystalnucleation sites from 10% (w/v) to 15% (w/v) that supersededthe fixed amount of titanium sol that will be attached onto theK particles. Therefore, 10% (w/v) of K suspension was chosen forall the subsequent experiments. A further study needs to be con-ducted to effectively determine the exact weight percentage incontrolling the coating film thickness, so that the optimal particlesize to SSA can be found.

3.4. Heat treatment of the photocatalyst

The heat treatment is a pre-requisite step for the transforma-tion of titanium sol into structural rigid and stable TiO2 crystallitesthat are attached firmly onto the K particles. This is because thecoated titanium sol is amorphous in nature, and further heat treat-ment could transform the coated TiO2 layer into the photoactivecrystalline phase. Many researchers reported that the temperaturefor the heat treatment could alter the crystal structure, SSA, parti-cle size, as well the photoactive phase of TiO2 crystalline [2,35–37].

In the prepared photocatalysts, the heat transformation ofamorphous into the TiO2 crystalline involves the shrinkage and col-lapse of the polymeric Ti–O–Ti gel network onto the K particles[2,31]. During the course of shrinkage, the Ti–O–Ti polymeric net-work is believed to form an interaction bonding with the siliceous

Table 1The characteristics of the titania/kaolinite nano-photocatalysts prepared underdifferent degrees of heat treatment

Calcinationtemperature (oC)

Photoactivephase (A/R)

BET specific surfacearea (m2/g)

Particlediameter (nm)

500 Anatase 39.23 ± 0.10 6.13550 Anatase 37.87 ± 0.10 6.50600 Anatase 35.24 ± 0.11 7.09650 Anatase/rutile 25.51 ± 0.12 12.76700 Anatase/rutile 24.04 ± 0.12 18.32

240 M.N. Chong et al. / Microporous and Mesoporous Materials 117 (2009) 233–242

side of K that presents a surface of oxygen and resulting in the for-mation of Ti–O–Si bonding. Gates et al. [41] reported that the pres-ence of such bonding is highly beneficial in influencing the stabilityof the prepared photocatalysts and its application for the utilisa-tion in catalysis. Besides this, it was also reported that the exis-tence of such Ti–O–Si bonds could suppress the phasetransformation of anatase to rutile [2]. This is highly beneficial,as most of the self-synthesized TiO2 is highly photoactive in itsanatase form. All these have shown the influences of K, apart fromproviding a structural rigid support. However, the existence ofsuch bonding is highly influenced by the preparation conditions,mainly the calcination temperature. In order to synthesise a highlyphotoactive photocatalyst that compromises the essential criteriasuch as the particle size, SSA and crystalline phase, an extensivestudy needs to be conducted to govern the correlation betweeneach of the criteria.

Fig. 9 shows the correlation between the particle sizes and SSA,and the calcination temperature. The size of the TiO2 crystallitescoated on the K particles has direct relationship with the calcina-tion temperature. When the calcination temperature increased

Fig. 9. The correlation between specific surface area (m2/g) wit

Fig. 10. X-ray diffraction (XRD) patterns for the titania/kaolinite nano-photocatalysts und(E) 700 oC. A: Anatase; R: Rutile.

from 500 oC to 700 oC, the particle size increased from approxi-mately 6 nm to 18 nm. The increase in the particle size which isdue to the formation of agglomerates and accompanying crystalgrowth (as evidenced from SEM and XRD peak) led to a reductionin the SSA of the prepared photocatalysts. The SSA changed slightlyfrom 39.23 ± 0.10 m2/g to 35.24 ± 0.11 m2/g between 500 oC and600 oC. An abrupt decrease in the specific surface area to25.51 ± 0.12 m2/g was observed when the particles were treatedat 650 oC.

h calcination temperature (oC) and particle diameter (nm).

er different heat treatment conditions. (A) 500 oC; (B) 550 oC; (C) 600 oC; (D) 650 oC;

Fig. 11. The kinetics for the photo-degradation of 40 ppm of Congo red in the reactor system.

M.N. Chong et al. / Microporous and Mesoporous Materials 117 (2009) 233–242 241

The abrupt decrease in SSA found at 650 oC may be due to thephase transformation of the crystalline TiO2 [31,35]. Although thephase transformation of anatase to rutile has been widely reported,there is rarely a discussion on the structural change in TiO2 latticeassociated with the SSA change. A temperature for phase transfor-mation from anatase to rutile was reported to be in a range between550 oC and 800 oC, depending on the preparation conditions andchemicals used [2]. We found that the phase transformation of crys-talline anatase to rutile occurred at 650 oC, which was evidencedfrom the XRD results shown in Fig. 10. Beydoun et al. [2] reportedthat the anatase to rutile phase transformation involves the ruptureof Ti–O bonds. Thus, the breakage of the Ti–O bonds at 650 oC canlead to a small SSA as the Ti lattice structure collapses. Apart fromthis detrimental effect on the SSA, the heat treatment at 650 oC alsoaffects the cracking of the Ti–O–Si bonds, as discussed earlier, andalters the thermal stability of the photocatalysts (Table 1). It canbe concluded that the optimal heat treatment for the phase trans-formation of 600 oC is needed to maintain the highly photoactivepure anatase phase with justifiable particle size and specific surfacearea for photocatalysis.

3.5. Photocatalytic degradation of Congo red

The photocatalytic degradation of Congo red was used as themodel reaction to evaluate the photocatalytic ability of the newlysynthesized TiO2/K photocatalysts. Congo red, a secondary diazodye is known to be a recalcitrant dye that presents a complicationin its removal from various sewage and industrial treated waste-water [8,42,43]. It was proposed that the photocatalytic oxidationof Congo red commences with the oxidation on both of the azoN@N bonds, resulting in the decolourisation and liberation ofmolecular nitrogen [42,43].

Various studies on photocatalytic degradation of Congo redhave been conducted using TiO2 prepared from different synthesismethods [8,14,42–45]. Wahi et al. [43] prepared nine nano-sizedTiO2 photocatalysts with different particle size, SSA and anataseto rutile phase composition for the photo-degradation of Congored. Their outcomes showed that the anatase nanorods preparedusing a surfactant concentration of 0.17 mmol L�1 tetraethylammonium hydroxide performed a complete degradation of14 lmol L�1 in 10 h.

In this study, the synthesized TiO2/K photocatalyst was appliedfor the photocatalytic degradation of 40 ppm (57.4 lmol L�1) ofCongo red. Fig. 11 shows that both the initial adsorption and pho-tocatalytic degradation of the Congo red was enhanced using theTiO2/K. Approximately 30% colour removal was due to the initial

adsorption of the Congo red by the photocatalysts. The bare TiO2

(without immobilisation on K) demonstrated low photo-degrada-tion ability. This can be explained by the fact that formation ofdense crystallites impeded up swirling and mixing in the reactorsystem. The TiO2/K photocatalysts performed a high efficiency inthe crude reactor system to degrade 40 ppm Congo red in 6 h. Itis important to note that the TiO2 coated on the K can improvethe homogeneity mixing by partitioning the titanium sol onto itssurface, and provides a structural support to enhance adsorptioncapability, photocatalytic efficiency and ease of separation.

4. Conclusion

A reproducible modified two step sol–gel method was devel-oped to synthesise a novel TiO2/K nano-photocatalyst in this study.The systematic manipulation in addition of the HNO3 to establishthe optimal concentration in the process not only provides titaniasol with optimal gel behaviours but a highly porous coating layerthat enhances adsorption ability. The TiO2/K photocatalyst syn-thesised using 10% (w/v) pre-treated K had a constant coatingthickness of 7 nm. The variation in the temperature of the subse-quent heat treatment contributed to the TiO2/K photocatalystswith different physical properties in terms of particle size, specificsurface area and the photoactive phase. The optimal heat treat-ment was suppressed at 600 oC. This was proposed to be primarilydue to the appearance of rutile phase at 650 oC, which resulted inthe rupture of Ti–O bonds and abrupt decrease in SSA.

The TiO2/K photocatalysts had desirable porous layer and nano-size structure and promising properties to demonstrate a highadsorption and degradability for the removal of Congo red. Impor-tantly, the immobilised TiO2/K particles can be easily separatedand recovered from a water treatment system. All this makesTiO2/K nano-particles a novel photocatalyst for a viable industrialwater/wastewater treatment process.

Acknowledgments

This work was supported by the Australian Research CouncilLinkage Grant (LP0562153) and Australian Water Quality Centrethrough the Water Environmental Biotechnology Laboratory(WEBL) at the University of Adelaide.

References

[1] A. Arques, A.M. Amat, L.S. Juanes, R.F. Vercher, M.L. Marin, M.A. Miranda,Journal of Molecular Catalysis A: Chemical 271 (2007) 221–226.

242 M.N. Chong et al. / Microporous and Mesoporous Materials 117 (2009) 233–242

[2] D. Beydoun, R. Amal, Materials Science and Engineering B 94 (1) (2002) 71–81.[3] A. Bhattacharyya, S. Kawi, M.B. Ray, Catalysis Today 98 (2004) 431–439.[4] B. Cheng, J.G. Yu, X.Q. Zhang, X.J. Zhao, Journal of Materials Science Letters 22

(2003) 967–970.[5] H. Choi, Y.J. Kim, R.S. Varma, D.D. Dionysiou, Chemistry Materials 18 (2006)

5377–5384.[6] Z. Ding, H.Y. Zhu, G.Q. Lu, P.F. Greenfield, Journal of Colloid and Interface

Science 209 (1998) 193–199.[7] D.D. Dunuwila, C.D. Gagliardi, K.A. Berglund, Chemistry Materials 6 (1994)

1556–1562.[8] S. Erdemoglu, S.K. Aksu, F. Sayilkan, B. Izgi, M. Asilturk, H. Sayilkan, F. Frimmel,

S. Gucer, Journal of Hazardous Materials, 2007, doi:10.1016/j.jhazmat.2007.11.087.

[9] J. Fernandez, J. Kiwi, J. Baeza, J. Freer, C. Lizama, H.D. Mansilla, Journal ofApplied Catalysis B: Environmental 48 (3) (2004) 205–211.

[10] L. Jing, X. Sun, B. Xin, B. Wang, W. Cai, H. Fu, Journal of Solid State Chemistry177 (2004) 3375–3382.

[11] M.S. Lee, S.S. Hong, M. Mohseni, Journal of Molecular Catalysis A; Chemical 242(2005) 135–140.

[12] H. Yang, R. Shi, K. Zhang, Y. Hu, A. Tang, X. Li, Journal of Alloys and Compounds398 (2005) 200–202.

[13] K. Mogyorosi, I. Dekany, J.H. Fendler, Langmuir 19 (2003) 2938–2946.[14] R. Molinari, F. Pirillo, M. Falco, V. Loddo, L. Palmisano, Chemical Engineering

and Processing 43 (2004) 1103–1114.[15] S. Takeda, S. Suzuki, H. Odaka, H. Hosono, Thin Solid Films 392 (2) (2001) 338–

344.[16] Z. Xiong, Y. Xu, L. Zhu, J. Zhao, Langmuir 21 (2005) 10602–10607.[17] W. Yan, B. Chen, S.M. Mahurin, E.W. Hagaman, S. Dai, H. Overbury, Journal of

Physical Chemistry B 108 (9) (2004) 2793–2796.[18] L. Zhang, Y. Zhu, Y. He, W. Li, H. Sun, Journal of Applied Catalysis B:

Environmental 40 (4) (2003) 287–292.[19] D. Zhao, J. Zhou, N. Liu, Materials Science and Engineering A 431 (2006) 256–

262.[20] I. Ilisz, A. Dombi, K. Mogyorosi, A. Farkas, I. Dekany, Applied Catalysis B:

Environmental 39 (2002) 247–256.[21] K. Mogyorosi, A. Farkas, I. Dekany, Environmental Science and Technology 36

(2002) 3618–3624.[22] T. Mori, J. Suzuki, K. Fujimoto, M. Watanabe, Y. Hasegawa, Journal of Sol–Gel

Science and Technology 19 (2000) 505–510.

[23] Y. Xu, C.H. Langford, Journal of Physical Chemistry B 101 (1997) 3115–3121.[24] R.M. Barrer, Zeolites and Clay Minerals as Sorbents and Molecular Sieves,

Academic Press Incorporation Ltd, London, 1978.[25] H.I. Inyang, S. Bae, G. Mbamalu, S.W. Park, Journal of Materials in Civil

Engineering 19 (1) (2007) 84–90.[26] H.Z. Ma, B. Wang, Journal of Hazardous Materials: B 136 (2006) 365–370.[27] B. Wang, L. Gu, H. Ma, Journal of Hazardous Materials 143 (2007) 198–205.[28] Q. Zhuo, H. Ma, B. Wang, L. Gu, Journal of Hazardous Materials 142 (2007) 81–

87.[29] B.L. Bischoff, M.A. Anderson, Chemistry Materials 7 (1995) 1772–1778.[30] K.P. Kumar, J. Kumar, K. Keizer, Journal of American Ceramic Society 77 (5)

(1994) 1396–1400.[31] D.J. Suh, T.J. Park, Chemistry Materials 8 (1996) 509–513.[32] J.C. Yu, J. Yu, J. Zhao, Applied Catalysis B: Environmental 36 (2002) 31–43.[33] J.C. Yu, J. Yu, W. Ho, Z. Jiang, L. Zhang, Chemistry Materials 14 (2002) 3808–

3816.[34] M. Guglielmi, G. Carturan, Journal of Non-Crystalline Solids 100 (1988) 16–30.[35] N. Negishi, K. Takeuchi, Materials Letters 38 (2) (1999) 150–153.[36] Y. Tanaka, M. Suganuma, Journal of Sol–Gel Science and Technology 22 (2001)

83–89.[37] B. Xia, H. Huang, Y. Xie, Materials Science and Engineering B 57 (2) (1999)

150–154.[38] D. Robert, J.V. Weber, Journal of Materials Science Letters 18 (1999) 97–98.[39] B. Aiken, E. Matijevic, Journal of Colloid and Interface Science 126 (22) (1998)

645–649.[40] I.U. Haq, E. Matijevic, Journal of Colloid and Interface Science 192 (1) (1997)

104–113.[41] B.C. Gates, H. Knozinger, Advances in Catalysis, Elsevier Science Book Edition,

2004.[42] C. Guillard, J. Disdier, C. Monnet, J. Dussaud, S. Malato, J. Blanco, M.I.

Maldonado, J.M. Herrmann, Applied Catalysis B: Environmental 46 (2003)319–332.

[43] R.K. Wahi, W.W. Yu, Y. Liu, M.L. Mejia, J.C. Falkner, W. Nolte, V.L. Colvin,Journal of Molecular Catalysis A: Chemical 242 (2005) 48–56.

[44] I.K. Konstantinou, T.A. Albanis, Applied Catalysis B: Environmental 49 (2004)1–14.

[45] Y. Yang, Q. Wu, Y. Guo, C. Hu, E. Wang, Journal of Molecular Catalysis A:Chemical 225 (2005) 203–212.