CERAMICSINTERNATIONAL

Available online at www.sciencedirect.com

0272-8842/$ - sehttp://dx.doi.org/

nCorrespondinE-mail addre

Ceramics International 40 (2014) 7125–7132www.elsevier.com/locate/ceramint

Purification of titania nanoparticle thin films: Triviality or a challenge?

Urmas Joosta,n, Anu Saarvaa, Meeri Visnapuua, Ergo Nõmmistea, Kathriin Utta, Rando Saara,b,Vambola Kisanda,b

aInstitute of Physics, University of Tartu, Riia 142, 51014 Tartu, EstoniabEstonian Nanotechnology Competence Center, Riia 142, 51014 Tartu, Estonia

Received 11 November 2013; accepted 10 December 2013Available online 18 December 2013

Abstract

Although purification of titania nanoparticles is a very common process in laboratory practice the specific purification protocol is often chosenwithout scientific justification. In the present work the efficiency of three different and in laboratory practice widely used treatments is comparedfor removing organic contaminants from the surface of titania nanoparticles. The most effective purification treatment was annealing ofnanoparticles below 400 1C: annealing at higher temperatures causes besides purification the growth of nanoparticles.& 2013 Elsevier Ltd and Techna Group S.r.l. All rights reserved.

Keywords: A. Films; A. Sol–gel process; Nanoparticles; D. TiO2

1. Introduction

Titania (TiO2) has attracted attention as a perspectivematerial for many advanced applications. It has been studiedextensively as a promising photocatalyst [1], solar cell material[2], biocompatible material [3], material for anti-fogging andself-cleaning coating [4,5], etc. Nanostructure, crystal struc-ture, purity etc. all play important roles in many of titaniaapplications.

The efficiency of TiO2 photocatalytic properties dependsbesides its crystal structure also on the grain size [6]. Even theadhesion of living cells on titania films (extremely importantfor medical implants) depends on the nanostructure of thematerial [3]. The efficiency of a dye sensitized solar cell is aswell influenced by the nanostructure of the material, since thegrain size of the material directly influences how much dye canbe adsorbed on titania surface [7].

One of the ways to control the grain size of titania is bychanging the annealing temperature of the material. Withincreasing annealing temperature, the grain size starts toincrease, but also crystal phase change occurs during annealingfrom the photocatalytically active anatase to inactive rutile

e front matter & 2013 Elsevier Ltd and Techna Group S.r.l. All ri10.1016/j.ceramint.2013.12.047

g author. Tel.: þ372 580 17909.ss: urmas.joost@ut.ee (U. Joost).

crystal phase [8–11]. With annealing it is possible to get largercrystallites, but it has been shown that enhanced properties oftitania usually emerge with smaller particles/grains [3,6,7].Titania can as well exist in amorphous phase but amorphoustitania is not photocatalytically as active as anatase crystalphase [12]. Electron transfer is also faster between surfaceadsorbed dye and crystalline titania compared to amorphousmaterial [13] and conductivity of titania is enhanced with theevolution of its crystalline structure [14].For many applications it is feasible to synthesise titania

nanoparticles with controlled size and crystal phase andsubsequently manufacture the necessary object (electrode,functional coating etc.). There are numerous methods tosynthesise titania nanoparticles with different sizes and crystal-line phases. It is possible to obtain titania particles as small as4 nm with well-defined crystal structure using various methods[15–17]. With different methods titania consisting of anatase[15–19], rutile [18,19] or brookite [18] crystalline phase caneasily be obtained.Synthesis methods using organic media and/or organic

ligands might prove to be problematic in specific applicationsof titania due to excess amount of organic material adsorbed onthe surface of the synthesised nanoparticles. Depending on themethod used the organic material could be toxic, acidic orblocking the surface and making the material unusable for

ghts reserved.

U. Joost et al. / Ceramics International 40 (2014) 7125–71327126

the intended application. For this reason it is very important toinvestigate different methods of cleaning the synthesisedmaterial from organic residue.

Although purification of titania nanoparticles is a verycommon process in laboratory practice the specific purificationprotocol is often chosen without scientific justification. Theaim of the present work is to compare efficiency of threedifferent and in laboratory practice widely used treatments forremoving organic contaminants from the surface of thenanoparticles.

2. Experimental

Titania nanoparticles were synthesised using a methoddescribed by E. Scolan et al. [15] with slightly modifiedparameters. Commercially available reagents were used, tita-nium(IV) butoxide (Ti(OCH2CH2CH2CH3)4) (Sigma-Aldrich,reagent grade), p-toluene sulfonic acid (PTSA) (Sigma-Aldrich, reagent plus), acetyl acetone (acac) (Sigma-Aldrich,reagent plus), butanol (Sigma-Aldrich) and deionised waterwere used as precursors. Since the compounds are sensitive towater, the solvent (butanol) was dried using CaH2 and distilledbefore utilization. The molar ratio between PTSA and titanium(IV) butoxide was set to 0.2, the molar ratio between acac andtitanium(IV) butoxide was set to 3 and the molar ratio betweenwater and titanium(IV) butoxide was set to 10. The reactionwas carried out overnight at reflux conditions. The mixturewas first cooled to room temperature and after that the solventwas evaporated with a rotary evaporator at 70 1C, the solidresidue was weighed and dispersed in acetone yielding auniform and stable 10 wt% colloidal solution.

Thin films were prepared by spin-coating colloidal solutionon Si(1 0 0) substrates in ambient atmosphere. The substrateswere cleaned prior to coating with ethanol to remove smalldust particles. The rotation frequency during spin-coating was3000 rpm and coating time was 0.5 min. The obtainedprecursor films were aged at room temperature in ambientconditions for 4 days. The purpose of such ageing was to allowthe remaining solvent to evaporate slowly in order to preventthe cracking of the films. After ageing the samples wereannealed in a Nabertherm L5/11/S27 furnace (in air atmo-sphere) in order to evaporate the remaining solvent andevaporate/burn off organic species. Similar methods for pre-paring thin titania films (not based on nanoparticles) have beenused by our group previously [20–23].

Three different and in laboratory practice widely usedtreatments were devised to remove organic contaminants fromthe surface of the nanoparticles: (i) annealing at differenttemperatures, (ii) annealing followed by washing with deio-nised water in an ultrasound bath for 10 min, and (iii)irradiation of nanoparticle film (aged, but not annealed) withUV light to degrade the organic molecules using photocatalyticproperties of titania.

At the beginning of annealing all samples were put into thefurnace and the temperature was increased to first set value of100 1C by using the mean ramping rate of 1.25 1C/min. Aftera 1-h dwelling time, one set of samples was removed from

the furnace and the temperature was increased to the nextdesired value (200 1C) by using a mean ramping rate 1.7 1C/min.This was followed by annealing for one hour at that temperatureand removal of another set of samples from the furnace. Theprocedure was repeated until the last sample was removed fromthe furnace after a 1-h annealing at 500 1C.Part of samples was washed with deionised water after

annealing. However films annealed lower than 200 1C weredestroyed during the washing (nanoparticles were washed ofthe substrate), only samples annealed at 200 1C or higher couldbe analysed after washing. Washing off the films wasconducted in deionised water for 10 min using 37 kHzfrequency at room temperature (Elmasonic P 30 H).UV light was used to photo-oxidize organic contaminants on

the titania nanoparticles (aged, but not annealed). Self-madesetup was used, a 6 W low pressure mercury lamp withluminous output of 1.7 W at 254 nm (according to datasheet[24]) was employed. Luminous flux at the sample surface wasmeasured with Delta Ohm HD2302.0 lux meter equipped withan UV C sensor. Luminous flux between 220 and 280 nm was60 W/m2. Photo-oxidation experiments were conducted inMemmert CTC 256 climate chamber at 70% of relativehumidity and 30 1C. It is stated that the most effectivehumidity conditions for photo-oxidation on titania surface isin the humidity range from 10 to 80% [25].X-ray photoelectron spectroscopy (XPS) was used for

investigating the chemical state and elemental composition oftitania nanoparticle films after different treatments. XPSmeasurements were conducted using a surface station equippedwith an electron energy analyser (SCIENTA SES 100) and anon-monochromatic twin anode X-ray tube (Thermo XR3E2),with characteristic energies of 1253.6 eV (Mg Kα1,2 FWHM0.68 eV) and 1486.6 (Al Kα1,2 FWHM 0.83 eV). All XPSmeasurements were conducted in ultra-high vacuum (UHV)conditions. The binding energy scales for the XPS experimentswere referenced to the binding energy of Ti 2p3/2 (458.6 eV)photoemission line. To estimate overall atomic concentrationsof different compounds and elements average matrix relativesensitivity factors (AMRSF) procedure [26] and transmissionfunction of the instrument were used. Raw data were processedusing Casa XPS [27] software. Data processing involvedremoval of Kα and Kβ satellites, removal of background andfitting of components. Background removal was done usingTougaard background, for fitting Gauss–Lorentz hybrid func-tion was used (GL 70, Gauss 30%, Lorentz 70%) for best fit.However the absolute amounts of different compounds andelements have to be considered cautiously and are given tooutline trends and estimates only. Due to the possible surfaceregion deviation from chemical homogeneity in the workingrange of photoelectron spectroscopy (surface region withthickness up to three electron mean free paths), some signalsmight be amplified or supressed.The crystalline phases of titania were examined by measur-

ing room-temperature Raman spectra of the films prepared ona fused silica substrate using Renishaw micro-Raman setupequipped with 514 nm continuous mode argon ion laser,spectral resolution was about 1.5 cm�1.

U. Joost et al. / Ceramics International 40 (2014) 7125–7132 7127

The AFM measurements with the purpose of investigatingthickness of the films were conducted using Veeco AFM.Typically, tapping mode was utilized in order to provide anoptimal performance. OTESPA AFM tips (manufactured byBruker) were used.

The SEM images were collected and the elemental distribu-tions were determined using an SEM-FIB instrument (FEIHelios 600 setup equipped with an EDX detector, OxfordInstruments).

Optical transmission and reflection measurements wereconducted with a Jasco V-570 (UV/Vis/NIR) spectrometerequipped with a tungsten lamp and a photomultiplier.

Hydrodynamic diameter and size distribution of the particleswas measured in ethanol using DLS (dynamic light scattering)technique (Zetasizer Nano, Malvern Instruments).

Fig. 2. XPS C 1s spectra (hν¼1253.6 eV, scan step 0.5 eV) of non-heatedtitania nanoparticle film (marked as 25 1C) and a film annealed at 400 1C.

Fig. 3. Elemental composition (atomic %) of titania nanoparticle filmsannealed at different temperatures from room temperature to 500 1C.

3. Results and discussion

From the typical overview spectrum in Fig. 1 it can be seenthat the annealed nanoparticle films consist of titanium andoxygen as main chemical components and carbon and sulfur ascontaminants. Carbon and sulfur originate from the synthesisprocess of the nanoparticles since the synthesis is carried out inalcohol media with acac as a ligand and PTSA as a pHmodifier. Most of the solvent and other volatile substances areremoved during rotary evaporation but PTSA is not volatileenough at 70 1C to be removed during rotary evaporation andtherefore is the main remaining additive.

Carbon 1s photoelectron structure can be separated intoseven different sub-bands, sp2 hybridized carbon at 284.1 eV[28], sp3 hybridized carbon at 285.0 [29] and carbon attachedto different oxygen moieties: C–O at 286.5 eV [28,30], C=O at287.7 eV [30], O–C¼O at 288.3 eV [28,30,31], carbonates at290.1 eV [28,30], and π–π in the aromatic ring of PTSA at291.6 eV [28]. After deconvolution of the C 1s photoelectronstructure (Fig. 2) it can be seen that the dominating sub-bandcan be assigned to sp2 hybridized carbon. The only source ofsp2 hybridized carbon in the present experiment is the PTSAmolecule with its aromatic benzene ring. Some sp3 hybridizedcarbon is also present (one of the respective sources is methylgroup in the PTSA molecule) and traces of different oxygencontaining carbon species can be found.

Fig. 1. XPS overview spectrum (hν¼1253.6 eV, scan step 0.5 eV) of titaniananoparticle film annealed at 300 1C.

Three different strategies were devised to remove organiccontaminants from the prepared nanoparticle thin films: (i)annealing at different temperatures, (ii) annealing followed bywashing with deionised water in an ultrasound bath for 10 min,and (iii) irradiation of nanoparticle film (aged, but notannealed) with UV light to degrade the organic moleculesusing photocatalytic properties of titania.First strategy was annealing. The films were annealed at

100, 200, 300, 400, 500 1C in air atmosphere and after that thesurface composition of films was analysed with XPS, resultscan be seen in Fig. 3.It can be observed that the total carbon content decreases

during annealing drastically from 44 to 8%. The carbon contentdecreases in two steps, during annealing from room temperatureto 200 1C and from 300 to 500 1C. Also the composition of thecarbon changes during annealing as can be seen from Fig. 4.During annealing the amount of sp2 hybridized carbondecreases, but the amount of sp3 hybridized carbon does notchange until 400 1C. As mentioned before, the only source ofsp2 carbon in the present experiment is the PTSA molecule.

Fig. 4. The amount of different carbon species (atomic %) in titaniananoparticle films annealed from room temperature to 500 1C.

Fig. 5. Elemental composition (atomic %) of titania nanoparticle filmsannealed at different temperatures from room temperature to 500 1C and thenwashed in ultrasound bath with deionised water.

Fig. 6. XPS C 1s (hν¼1253.6 eV, scan step 0.5 eV) spectra of nanoparticlefilms annealed at 400 1C before and after washing with deionised water.

Fig. 7. Elemental composition of aged, but not annealed titania nanoparticlefilms irradiated with UV light. Luminous flux was 60 W/m2 and relativehumidity and temperature were 70% and 30 1C.

U. Joost et al. / Ceramics International 40 (2014) 7125–71327128

At low heating temperatures volatile components start toevaporate immediately, oxidation of organic compounds startsfrom 300 1C and higher annealing temperatures. Although theamount of oxygen bound carbon (C–O, C¼O and O–C¼O)starts to increase at 300 1C slightly, the oxidation of organiccompounds is not very extensive and the overall amount ofoxygen containing carbon species remains low as can be seenfrom Fig. 4. It can be concluded that evaporation is the mainmechanism of removal of organic contaminants in the wholetemperature range. The amount of sulfur impurity correlateswell with the overall amount of sp2 hybridized carbon(R2¼0.97) indicating the removal (i.e. evaporation) of thewhole molecule of PTSA during annealing rather than oxida-tion or decomposition of the molecule.

As demonstrated in Figs. 3 and 4, heat treatment at hightemperatures is effective in removal of organic compoundsfrom the titania nanoparticle film. Annealing the nanoparticlefilms for an hour at 500 1C decreases the carbon content to 8%,sulfur is completely removed and is not detectable in the XPSspectra. Exact determination and quantification of differentoxygen containing carbon species is complicated due to thesmall quantity of compounds present. A wide and featurelessshoulder at the higher energy side of the C 1s photo linecorresponds to a large number of different possible oxygen–carbon species (see Fig. 2) that can be found in the films.

It is important to note that in the context of present workannealing is used for cleaning nanoparticles. It is well knownthat annealing can change crystal structure of nanoparticles.However, as demonstrated in Fig. 11, very little change incrystal structure takes place during heating up to 400 1C.

Second strategy was annealing followed by washing withdeionised water in an ultrasound bath for 10 min. Washing oftitania nanoparticle films resulted in immediate removal ofsulfur and decrease in carbon content, but as can be seen inFig. 5 annealing at 200 1C and above prior to the washing hadlittle effect on the final composition of the films. Whencomparing the carbon content at 500 1C annealed films beforeand after washing (Figs. 3 and 5) it can be seen that the amountof carbon in the films decreases when the carbon content

U. Joost et al. / Ceramics International 40 (2014) 7125–7132 7129

before washing is high and even increases when the carboncontent before washing is low.

The carbon C 1s spectra at 400 1C annealed films before andafter washing (Fig. 6) differ significantly from each other. Thissuggests that the carbon contamination after washing withdeionised water origins from a different source. Probably thedeionised water is contaminated slightly with organic residueand due to the high surface area of titania nanoparticle films,

Fig. 8. The amount of different carbon species in aged, but not annealed titaniananoparticle films irradiated with UV light. Luminous flux was 60 W/m2 andrelative humidity and temperature were 70% and 30 1C.

Fig. 9. SEM images of 100, 300, 400, 500 1C ann

some of the organic compounds in the water will adsorb on thenanoparticles. Therefore carbon content of the films is notdecreased substantially during washing and is even increasedwhen the initial carbon content is low.Third strategy was irradiation of nanoparticle film (aged, but

not annealed) with UV light to degrade the organic moleculesusing photocatalytic properties of titania. During irradiationwith UV light the carbon content of the films (Fig. 7) decreasesslightly but the composition of the carbon compounds (Fig. 8)of the film changes remarkably. The amount of sp2 hybridizedcarbon decreases during 60 min of irradiation by UV light.There is no change in the amount of sp3 hybridized carbon butthe concentration of different oxygen containing carbonspecies increases. During photo-oxidation the PTSA aromaticring (only source of sp2 carbon in the present experiment) isattacked first but the photo degradation of resulting compoundsis slower and the overall carbon content of titania nanoparticlefilms decreases only slightly during 60 min of irradiation withUV light (Fig. 7). The amount of different oxygen containingcarbon species increases to a plateau indicating a steady rate ofoxidation and removal of the carbon contamination.Annealing of titania nanoparticle films until 300 1C does not

remarkably change the structure of the films but from 400 to500 1C particle growth can be observed as can be seen fromFig. 9. At 400 1C the particle size is slightly larger but at500 1C the particles have grown several times in diameter.

ealed titania nanoparticle films (not washed).

Fig. 10. Titania nanoparticle film thicknesses based on AFM measurementsaccording to the annealing temperature and following washing.

Fig. 11. Raman spectra of titania nanoparticle films annealed at differenttemperatures. The Raman spectrum of non-heated film (marked as 25 1C) isshown as well.

Fig. 12. Hydrodynamic diameter of titania nanoparticles, distribution is shownby the volume of the particles.

U. Joost et al. / Ceramics International 40 (2014) 7125–71327130

Washing in ultrasound bath with deionised water does notchange the surface morphology or damage the films in anyway and SEM images of the films are identical to the filmswithout washing treatment (images of washed films are notshown). Films annealed at 100 1C show distinctive aggregateson the surface which are probably the result of organic materialfrom inside the nanoparticle film. The aggregate disappear athigher annealing temperatures. Unfortunately it was impossi-ble to get adequate images of films without any heat treatment,and therefore cannot be shown in comparison.

The nanoparticle films shrink during annealing (Fig. 10), atfirst the films were E150 nm thick but after annealing at500 1C the thickness was only E90 nm. When annealed at1001 for an hour the films shrink to 140 nm. The thicknessdoes not change remarkably when the annealing temperature isfurther increased to 200 1C and starts to decrease again from300 1C and higher annealing temperatures. The films shrinkdue to the removal of organic compounds. The initial shrinkingis due to the removal of volatile compounds, e.g. residualbutanol. This is supported by XPS data (Fig. 3) that show arapid decrease in carbon content. Washing the films inultrasound bath with deionised water had also a slight effecton the thickness compared to the annealed film. When filmswere annealed at lower temperatures then washing decreasedthe film thickness by removing some of the organic com-pounds but in case of higher annealing temperatures (400 and500 1C) the film thicknesses were increased slightly comparedto the unwashed films. This is due to the chemi- andphysisorption of water on the nanoparticles. Also increase ofthe content of organic compounds in the films is in princiblepossible.

Raman studies show that the nanoparticle films consist ofanatase as bands were observed at 144 (Eg), 197 (Eg) and 397(B1g) cm

�1 which correlate well with the Raman active modesof anatase [32,33]. The band originally around 150 cm�1

shifts at higher annealing temperatures to lower values whichis consistent with the work of Li Bassi et al. [33] where avisible shift of 144 cm�1 band to higher values in case ofsmall nanoparticles is shown. From Fig. 11 it can be observedthat the band at 150 cm�1 does not start to shift until 400 1C

but at 500 1C it has already shifted to 144 cm�1 as reported inliterature for bulk anatase [32,33]. Raman data correlates wellwith SEM images in Fig. 9, that show particle growth for 400and 500 1C annealed films.The particle size and size distribution were also measured

prior to the preparation of nanoparticle films using DLStechnique. As can be seen from Fig. 12 the size of the particlesis less than 10 nm and the particles are almost monodisperse.

4. Conclusion

In the present work different purification methods for titaniananoparticle films were studied. Three different methods werecompared: (i) annealing at different temperatures, (ii) anneal-ing followed by washing with deionised water in an ultrasoundbath, and (iii) irradiation with UV light to degrade the organicmolecules using photocatalytic properties of titania. The mosteffective was annealing at high temperatures, the amount ofcarbon in the film dropped from 35 to 8% and sulfur wascompletely removed. Unfortunately annealing at 400 1C andhigher will cause the nanoparticles to grow, and the surface

U. Joost et al. / Ceramics International 40 (2014) 7125–7132 7131

area of the films will decrease and properties of the nanopar-ticle films will change.

Washing the annealed films using ultrasound in deionisedwater destroyed films annealed at lower temperatures than200 1C (nanoparticles were removed from the substrate).However washing did not alter the surface morphology offilms annealed at 200 1C and higher temperatures. The carboncontent of films annealed at low temperatures decreased but ifthe films had already relatively low carbon content the carboncontent increased after washing with deionised water. Theincrease in the carbon content of the films was caused by traceamounts of different carbon containing organic residue in thedeionised water. Sulfur was always removed with the washingprocedure.

UV irradiation also slowly decreased both the carbon andsulfur content of the nanoparticle films. Aromatic ring ofPTSA molecule was attacked first in the process of photo-oxidation, but the photodegradation of the resulting com-pounds was slower and the carbon content decreased only by5% during 60 min of UV irradiation.

Acknowledgement

Financial support by the Estonian Ministry of Education andResearch (target-financed theme IUT2-25), Estonian ScienceFoundation (grants 8216 and 8737), Estonian NanotechnologyCompetence Center (EU29996), ERDF projects (“IRGLASS”3.2.1101.12-0027, “TRIBOFILM” 3.2.1101.12-0028, “Nano-Com” 3.2.1101.12-0010, “Mesosystems: Theory and Applica-tions” TK114, „High-technology Materials for SustainableDevelopment“ TK117), Graduate School “Functional Materi-als and Technologies” (European Social Fund project1.2.0401.09-0079), Development Fund of University of Tartu,are gratefully acknowledged. The authors are grateful to MissMeeri Lembinen for help with AFM setup.

References

[1] M. Keshmiri, M. Mohseni, T. Troczynski, Development of novel TiO2

sol–gel-derived composite and its photocatalytic activities for trichlor-oethylene oxidation, Appl. Catal., B 53 (2004) 209–219.

[2] G.M. Kiema, M.J. Colgan, M.J. Brett, Dye sensitized solar cellsincorporating obliquely deposited titanium oxide layers, Sol. EnergyMater. Sol. Cells 85 (2005) 321–331.

[3] T.J. Webster, R.W. Siegel, R. Bizios, Osteoblast adhesion on nanophaseceramics, Biomaterial 20 (1999) 1221–1227.

[4] R. Wang, K. Hashimoto, A. Fujishima, M. Chikuni, E. Kojima,A. Kitamura, M. Shimohigoshi, T. Watanabe, Light-induced amphiphilicsurfaces, Nature 388 (1997) 431–432.

[5] R. Wang, K. Hashimoto, A. Fujishima, M. Chikuni, E. Kojima,A. Kitamura, M. Shimohigoshi, T. Watanabe, Photogeneration of highlyamphiphilic TiO2 surfaces, Adv. Mater 10 (1998) 135–138.

[6] Z. Zhang, C. Wang, R. Zakaria, J.Y. Ying, Role of particle size inNanocrystalline TiO2-based photocatalysts, J. Phys. Chem. B 102 (1998)10871–10878.

[7] E. Stathatos, P. Lianos, C. Tsakiroglou, Highly efficient nanocrystallinetitania films made from organic/inorganic nanocomposite gels, Micro-porous Mesoporous Mater 75 (2004) 255–260.

[8] M.C. Carotta, M. Ferroni, V. Guidi, G. Martinelli, Preparation andcharacterization of nanostructured titania thick films, Adv. Mater 11(1999) 943–946.

[9] K.R. Zhu, M.S. Zhang, J.M. Hong, Z. Yin, Size effect on phase transitionsequence of TiO2 nanocrystal, Mater. Sci. Eng., A 403 (2005) 87–93.

[10] T.B. Ghosh, S. Dhabal, A.K. Datta, On crystallite size dependence ofphase stability of nanocrystalline TiO2, J. Appl. Phys. 94 (2003)4577–4582.

[11] M. Maeda, T. Watanabe, Effects of crystallinity and grain size onphotocatalytic activity of titania films, Surf. Coat. Technol 201 (2007)9309–9312.

[12] S. Sakthivela, M.C. Hidalgo, D.W. Bahnemann, S.U. Geissen,V. Murugesan, A. Vogelpohl, A fine route to tune the photocatalyticactivity of TiO2, Appl. Catal., B 63 (2006) 31–40.

[13] H. Nishikiori, W. Qian, M.A. El-Sayed, N. Tanaka, T. Fujii, Change intitania structure from amorphousness to crystalline increasing photo-induced electron-transfer rate in dye-titania system, J. Phys. Chem. C 111(2007) 9008–9011.

[14] H. Nishikiori, N. Tanaka, T. Kitsui, T. Fujii, Photocurrent observed indye-doped titania gel, J. Photochem. Photobiol., A 179 (2006) 125–129.

[15] E. Scolan, C. Sanchez, Synthesis and characterization of surface-protectednanocrystalline nitania particles, Chem. Mater 10 (1998) 3217–3223.

[16] M. Niederberger, M.H. Bartl, G.D. Stucky, Benzyl alcohol and transitionmetal chlorides as a versatile reaction system for the nonaqueous andlow-temperature synthesis of crystalline nano-objects with controlleddimensionality, J. Am. Chem. Soc 124 (2002) 13642–13643.

[17] M. Niederberger, M.H. Blart, G.D. Stucky, Benzyl alcohol and titaniumTetrachlorideA versatile reaction system for the nonaqueous and low-temperature preparation of crystalline and luminescent titania nanoparti-cles, Chem. Mater 14 (2002) 4364–4370.

[18] D. Reyes-Coronado, G. Rodriguez-Gattorno, M.E. Espinosa-Pesqueira,C. Cab, R. de Coss, G. Oskam, Phase-pure TiO2 nanoparticles: anatase,brookite and rutile, Nanotechnology 19 (2008) 145605.

[19] A. Vioux, Nonhydrolytic sol�gel routes to oxides, Chem. Mater 9(1997) 2292–2299.

[20] V. Kisand, U. Joost, V. Reedo, R. Pärna, T. Tätte, J. Shulga, A. Saar,L. Matisen, A. Kikas, I. Kink, Influence of the heating temperature on theproperties of nickel doped TiO2 films prepared by sol–gel method, Appl.Surf. Sci 256 (2010) 4538–4542.

[21] R. Pärna, U Joost, E. Nõmmiste, T. Käämbre, A. Kikas, I. Kuusik, I. Kink,M. Hirsimäki, V. Kisand, Effect of different annealing temperatures andSiO2/Si(1 0 0) substrate on the properties of nickel containing titania thinsol–gel films, Phys. Status Solidi A 209 (2012) 953–965.

[22] R. Pärna, U. Joost, E. Nõmmiste, T. Käämbre, A. Kikas, I. Kuusik,M. Hirsimäki, I. Kink, V. Kisand, Effect of cobalt doping and annealingon properties of titania thin films prepared by sol–gel process, Appl. Surf.Sci 257 (2011) 6897–6907.

[23] U. Joost, R. Pärna, M. Lembinen, K. Utt, I. Kink, M. Visnapuu,V. Kisand, Heat treatment and substrate dependant properties of titaniathin films with high copper loading, Phys. Status Solidi A 210 (2013)1201–12212.

[24] Osram AG, ⟨http://www.osram.com/osram_com/products/lamps/specialty-lamps/ultraviolet-lamps/puritec-hns/index.jsp⟩, (11.11.2013).

[25] L. Frazer, Titanium dioxide: environmental white knight?, EnvironHealth Perspect. 109 (2001) 174–177.

[26] M.P. Seah, I.S. Gilmore, S.J. Spencer, Quantitative XPS: I. Analysis ofX-ray photoelectron intensities from elemental data in a digital photo-electron database, J. Electron. Spectrosc 120 (2001) 93–111.

[27] N. Fairley, CasaXPS version 2.3.12, www.casaxps.com (11.11.2013).[28] V. Datsyuk, M. Kalyva, K. Papagelis, J. Parthenios, D. Tasis, A. Siokou,

I. Kallitsis, C. Galiotis, Chemical oxidation of multiwalled carbonnanotubes, Carbon 46 (2008) 833–840.

[29] J. Diaz, G. Paolicelli, S. Ferrer, F. Comin, Separation of the sp3 and sp2

components in the C1s photoemission spectra of amorphous carbon films,Phys. Rev. B: Condens. Matter 54 (1996) 8064–8069.

[30] B.P. Payne, M.C. Biesinger, N.S. McIntyre, X-ray photoelectron spectro-scopy studies of reactions on chromium metal and chromium oxidesurfaces, J. Electron. Spectrosc. Relat. Phenom 184 (2011) 29–37.

U. Joost et al. / Ceramics International 40 (2014) 7125–71327132

[31] S. Beverly, S. Seal, S. Hong, Identification of surface chemical functionalgroups correlated to failure of reverse osmosis polymeric membranes,J. Vac. Sci. Technol. 18 (2000) 1107–1113.

[32] I.R. Beattie, T.R. Gilson, Single crystal Laser Raman spectroscopy, Proc.R. Soc. London, Ser. A 307 (1968) 407–429.

[33] A. Li Bassi, D. Cattaneo, V. Russo, C.E. Bottani, E. Barborini, T. Mazza,P. Piseri, P. Milani, F.O. Ernst, K. Wegner, S.E. Pratsinis, Ramanspectroscopy characterization of titania nanoparticles produced by flamepyrolysis: the influence of size and stoichiometry, J. Appl. Phys 98(2005) 074305.