M A T E R I A L S C H A R A C T E R I Z A T I O N 8 7 ( 2 0 1 4 ) 7 4 – 8 5
Ava i l ab l e on l i ne a t www.sc i enced i r ec t . com
ScienceDirect
www.e l sev i e r . com/ loca te /matcha r
TiO2 coatings synthesized by liquid flame spray and
low temperature sol–gel technologies on autoclavedaerated concrete for air-purifying purposes
Anibal Maury-Ramireza,1, Juha-Pekka Nikkanenb,2, Mari Honkanenb,2,Kristof Demeesterec,3, Erkki Levänenb,2, Nele De Beliea,⁎aMagnel Laboratory for Concrete Research, Department of Structural Engineering, Faculty of Engineering and Architecture, Ghent University,Technologiepark Zwijnaarde 904, B-9052 Ghent, BelgiumbDepartment of Materials Science, Tampere University of Technology, Korkeakoulunkatu 6, FIN-33720 Tampere, FinlandcResearch Group EnVOC, Department of Sustainable Organic Chemistry and Technology, Faculty of Bioscience Engineering, Ghent University,Coupure Links 653, B-9000 Ghent, Belgium
Article history:Received 19 July 2013Received in revised form23 October 2013Accepted 27 October 2013
This article reports for the first time in literature the successful application of two TiO2
synthesis-coating technologies, i.e. low temperature sol–gel (LTS) and liquid flame spraying(LFS), to develop autoclaved aerated concrete with air-purifying properties. Significantdifferences in crystal and agglomerate sizes, specific surface areas and crystal phasecompositions have been observed in the synthesized photocatalytic coatings. These were,however, not reflected in their respective air purification performance as indicated bytoluene removal efficiencies around 60% (TOL inlet concentration of 14 ppmv, 23 °C, 50%relative humidity and 3 min gas residence time) for both coating types. The differentnano-scale characteristic effects indicate to compensate each other so that the overallphotocatalytic activities as observed in our test set-up and conditions do no differ for bothsynthesis-coating technologies. The toluene elimination rates reported here (approx.40 mg TOL·m−2·h−1) are significantly higher than that reported in previous research usinghigh temperature sol–gel techniques or commercial TiO2 embedded in a cementitiousmatrix. Moreover, LTS and LFS technologies cannot only be easily applied on precastmaterials during manufacturing, they can also be applied on existing buildings. Based onthese promising results and application potential, further research that systematicallyinvestigates the photocatalytic effect of different synthesis-coating parameters is highlyencouraged to further improve the knowledge about these technologies.
75M A T E R I A L S C H A R A C T E R I Z A T I O N 8 7 ( 2 0 1 4 ) 7 4 – 8 5
1. Introduction
The application of titanium dioxide photocatalysis on cemen-titious materials has received considerable attention in recentyears as this technology can efficiently degrade organic(e.g. toluene) and inorganic (e.g. NOx) air pollutants usinglight irradiation as the only energy source [1–6]. Furthermore,based on the hydrophilic properties produced by TiO2
photocatalysis, self-cleaning or easy to clean propertieshave been also reported [7–12]. TiO2 is a semiconductormaterial that when it is light irradiated with high energyphotons (higher than its band gap — Eg) promotes electrons(e−) from the valence band (VB — with electrons) to theconduction band (CB — without electrons). Due to theequilibrium in charges, this process also generates electronicholes (h+) in the valence band. A portion of the generatedelectron-hole pairs reaches the surface of the semiconductorand initiates reduction–oxidation processes with adsorbedoxygen (O2) and water (H2O) molecules from the surroundingair. Then, reactive oxygen species (e.g. UOH and O2U
−) respon-sible for the oxidation of different organic and inorganiccompounds as well as the hydrophilic effect are generated[13]. TiO2 can crystallize in three different structures: rutile(tetragonal), anatase (tetragonal) and brookite (orthorhom-bic). Among these crystalline forms, rutile is thermodynam-ically the most stable while brookite and anatase transformto rutile under heating. Transformation of anatase to rutile isa broadly studied mechanism and occurs at temperaturesbetween 550 °C and 1000 °C, while the less studied transfor-mation of brookite to rutile has been reported between 550 °Cand 600 °C [14,15]. The temperature of this transformationstrongly depends on the impurities or dopants present in thematerial as well as on the morphology of the sample [16].Differences among rutile, brookite and anatase crystals are inthe density and hardness but the most important forphotocatalytic applications is the difference between theirband gaps. Anatase, brookite and rutile have band gaps of3.19/388.7, 3.11/398.6 and 3.0/413.3 eV/nm, respectively [17].So far, the use of TiO2 as a mixed bulk additive in ce-mentitious materials is the most common TiO2 loadingmethod. Nevertheless, the bulk addition of TiO2 nanoparti-cles inside the cementitious matrix is not an ideal solutionbecause of the limited accessibility of light and pollutants tomost TiO2 particles. Furthermore, previous research indi-cates that also cement carbonation decreases the photocat-alytic activity in relation to atrazine degradation [18].Similarly, also inactivation of the TiO2 added in the masshas been reported in relation to degradation of toluene,probably due to the presence of hydrated products aroundthe TiO2 particles as they act as nucleation sites duringcement hydration [19].
Therefore, the development of innovative TiO2 loadingmethods such as through coatings is an interesting approach.As applied on top of the materials' surfaces, coatings couldfacilitate the access of pollutants (to be degraded) andphotons (activators) to the catalyst. Furthermore, coatingsoffer not only the potential to apply the photocatalyticproperties in a more efficient way, but also to widen theapplication possibilities to existing buildings or building
products. Numerous methods have been employed toachieve the synthesis of photocatalytic active TiO2 nano-powders. However, not many synthesis technologies havebeen directly applied on cementitious materials to producecoatings. Promising technologies for this purpose are liquidflame spraying and sol–gel synthesis based on metal-alkoxide (M-OR) precursors [20–23]. In the liquid flamespraying technology, the liquid precursor is injected into ahigh temperature flame (Tmax ±3000 K), where it evaporatesand nucleates to nanosized particles [24]. In the sol–gelsynthesis, metal alkoxides are dissolved into an organicsolvent to be followed by reactions with water. Mainreactions are hydrolysis and condensation. During hydroly-sis, the alkoxide reacts with water producing M\OH bonds.Partially hydrolyzed molecules can link together in thefollowing condensation reaction that occurs between OHgroups and finally precipitate as nanosized particles [25].Typical crystallization temperatures range from 350 to 450 °Cin this technology [26].
In this research, nano-TiO2 particle based coatings forair-purifying purposes have been directly synthesized onautoclaved aerated concrete for the first time using liquidflame spray and low temperature sol–gel synthesis technol-ogies. A detailed characterization of the synthesized coatingsfrom nano- to macro-scale has been performed. At nano-scale, crystal and agglomerate sizes, specific surface areas,crystal phase compositions, band gaps and crystallinity havebeen determined for synthesized nano-TiO2 particles usingboth technologies. At macro-scale, the air cleaning potentialof the coatings applied on autoclaved aerated concretesamples has been evaluated at lab-scale by monitoringtoluene removal from air using a flow through operatedphotoreactor.
2. Materials and Methods
2.1. Autoclaved Aerated Concrete
High porosity of the substratematerial leads to a positive effectin thedegradation of tolueneandnitrogenmonoxide usingTiO2
photocatalytic cement based materials as reported by MauryRamirez et al. and Poon and Cheung, respectively [2,6]. There-fore, a substrate that satisfies this demand is selected in thisresearch. Autoclaved aerated concrete is a very porous andrough concrete. The open porosity of this material is previouslydetermined on 4 samples by the vacuum water saturationmethod described in the ASTM C1202-10 [27] and amounts to74.9 ± 2.9%. The roughness of the material is evaluated bymeans of distance measurements using the automated lasermeasurement system (ALM) described in [28], fromwhich theRa
factor presented in the BS 1134 is calculated [29]. In this case,eight profiles are obtained from each sample: 4 measurementsin longitudinal direction (measurement length: 80 mm) and4measurements in transversal direction (measurement length:60 mm). As a result, Ra, the arithmetical mean deviationof the profile from the center line reaches 70 ± 27 μm. Forthe air purification experiments, autoclaved aerated con-crete samples are cut to prisms with dimensions 100 mm ×80 mm × 30 mm.
76 M A T E R I A L S C H A R A C T E R I Z A T I O N 8 7 ( 2 0 1 4 ) 7 4 – 8 5
2.2. TiO2 Synthesis-Coating Technologies
2.2.1. Liquid Flame Spray (LFS)A liquid ready mixed feedstock of tetra-n-butyl orthotitanate(C16H36O4Ti > 98%, VWR) and 2-propanol (C3H7OH >99.5%,VWR) is atomized into micron sized droplets using a highvelocity hydrogen (H2) flow and fed into a turbulent hightemperature H2/O2 flame (Tmax ~3000 K). The concentration oftetra-n-butyl orthotitanate is 0.1 M. The flow of gases is40 L·min−1 H2 and 20 L·min−1 O2. The concrete samples areplaced in front of the flame at a fixed distance of 15 cm.Synthesized crystals are also collected by a cylindricalelectrostatic precipitator for further analysis. The appliedLFS technology is schematically illustrated in Fig. 1.
2.2.2. Low-Temperature Sol–Gel Synthesis (LTS)A low temperature TiO2 nanopowder sol–gel is prepared bydiluting 1.9 mLof tetra-n-butyl orthotitanate (C16H36O4Ti > 98%,VWR) in 6.25 mL of 2-propanol (C3H7OH > 99.5%, VWR). Afterstirring for 15 min at 50 °C, ion-exchanged water (90 ml) isadded. Then, the solution is further stirred in a reflux condenserreaction vessel at 50 °C and 98% relative humidity during 24 h.Finally, autoclaved aerated concrete samples are coated by dropapplication of the solution and further oven-drying at 50 °Cduring 24 h. Analogously to the LFS technology, to characterizethe synthesized coatings, also pure TiO2 is collected by filteringthe TiO2 solution followed by a drying process at 50 °C during24 h.
2.3. Characterization
2.3.1. Nano-scale
2.3.1.1. Specific Surface Area. The specific surface area ofthe synthesized TiO2 is obtained using a nitrogen adsorptiontest (Omnisorp 100cx, Coulter Electronics) operating under theBrunauer–Emmett–Teller (BET) method. Also, to compare thecrystal sizes with those obtained by TEM and XRD, the averagediameter of the synthesized crystals (dBET) is calculated usingEq. (1):
dBET ¼ 6� ρ� SSSAð Þ−1 ð1Þ
In this equation, ρ is the density (g·m−3) and SSSA is thespecific surface area (m2·g). The densities of anatase, brookiteand rutile are 3.9, 4.1 and 4.3 g·cm−3, respectively [17].
Liquid precursorfeed
Spray gun
H2/O2
Flame Substrate
Fig. 1 – Schematic representation of the applied liquid flamespray (LFS) technology.
2.3.1.2. Crystal Phase Composition. TiO2 crystal phases pres-ent in each synthesized coating are determined bymeans of anX-ray diffractometer (Kristalloflex D-500, Siemens) which usesmonochromatic CuKα radiation. Also, using the XRD smoothedpatterns and Eqs. (2) and (3), the relative content of the TiO2
crystal phases can be estimated.
WR ¼ AR 0:884AA þ ARð Þ−1 ð2Þ
WB ¼ kBAB kAAA þ kBABð Þ−1 ð3Þ
In these,WR is the mass fraction of rutile in the sample, AA
represents the integrated intensity of the anatase (101) peakand AR is the integrated intensity of the rutile (110) peak. WB
represents the mass fraction of brookite, AB is the integratedintensity of the brookite (121) peak, and kA and kB are coef-ficients, amounting 0.886 and 2.721, respectively [30]. Further-more, to compare the crystal sizes with those determined byTEM and BET, crystal sizes are calculated by the Scherrerformula (Eq. (4)).
t ¼ 0:9λ� B cosθð Þ−1 ð4Þ
In this, t is the size in nm, λ is the wavelength of X-rays innm, B is the full width at half maximum of the peak inradians, and θ is the Bragg angle in degrees [31].
2.3.1.3. Band Gaps. The corresponding band gap (Eg) of eachTiO2 synthesized coating is determined using UV–Vis Spec-trometry (CARY 300 UV–Vis, Varian) equipped with a diffusereflectance standard (SRS-99-010, Labsphere). In this, absor-bance (Abs) profiles are analyzed in a range of wavelengths (λ)between 200 and 800 nm at a scanning rate of 600 nm·min−1.Then, plotting (Abs × Energy)1/2 versus Energy (eV) anddrawing tangents to the curve, the band gaps can bedetermined. Energy, here representing the incident photonenergy, is obtained by using Eq. (5).
Energy eVð Þ ¼ 1240� λ−1 ð5Þ
2.3.1.4. Crystal and Agglomerate Sizes. In order to deter-mine the synthesized TiO2 crystal sizes, transmission electronmicroscope (TEM) images are acquired using a high resolutionanalytical TEM (JEM, JEOL Ltd.). Crystal size distributions areobtained using several TEM micrographs. The size of 300particles (n = 300) is measured for both LFS and LTS. Usingselected area electron diffraction (SAED) which is a crystallo-graphic experimental technique that can be performed withTEM, the existence of crystalline phases is verified. Further-more, to estimate the agglomerate sizes and their distribu-tion, field emission scanning electron microscopy (FE-SEM)images coupled with energy dispersive X-ray spectroscopy(EDS) analyses of the coated samples are performed on bothcross and top views (Zeiss, Ultraplus).
2.3.1.5. Crystallinity. In order to verify the crystallinity ofthe synthesized TiO2, differential scanning calorimetry (DSC)and thermogravimetry (TG) analyses are performed onboth amorphous and synthesized catalyst samples. In these
77M A T E R I A L S C H A R A C T E R I Z A T I O N 8 7 ( 2 0 1 4 ) 7 4 – 8 5
analyses, an amorphous sample is synthesized using thesame precursors and quantities as used in the LTS technol-ogy (Section 2.2.2). However, as determined in our previousstudies, the synthesis time selected (3 min) is short enoughto avoid the formation of crystalline TiO2 [22]. In thisresearch, only TiO2 synthesized at low temperatures (suchas that using LTS at 50 °C) will require the crystallinityanalysis as high temperature synthesized TiO2 (LFS) guaran-tees a complete crystallization around 3000 K. The temper-atures of the analyses range between 30 and 500 °C under adry nitrogen atmosphere. For the DSC and TG analyses, a DSC204 F1 Phoenix (Netzsch) and a simultaneous thermalanalyzer STA 6000 (Perkin Elmer) are used, respectively.
2.3.2. Macro-scale: Air Purification ExperimentsFor evaluating the air-purifying potential of the TiO2 coatedconcrete samples, the photocatalytic degradation of toluene(TOL) from air is investigated in a rectangular plexiglassflat-plate photoreactor (length: 20 cm, width: 10 cm, height:4 cm) using the test set-up described in [1,2]. Toluene isselected as a model pollutant in this research, given itsrepresentativeness for the broad group of industrial/trafficrelated volatile organic compound(s). Prior to all photocata-lytic degradation experiments, humidified TOL contaminatedair is passed through the reactor in the absence of UVirradiation for at least 12 h to obtain the desired relativehumidity (RH) and TOL gas–solid adsorption–desorptionequilibrium throughout the whole reactor, the latter beingexemplified by TOL in- and outlet concentrations not signif-icantly different from each other (t-test, significance levelα ≥ 0.05). Once the UV irradiation begins, the photocatalyticTOL degradation is carefully followed for at least 30 h, being10 times longer than advised in the Japanese Standard test: JISR 1701-3:2008 [32]. Then, TOL removal efficiency (nremoval),loading (LR) and elimination (ER) rates during the UVirradiated period are calculated as presented in Eq. ((6)–(8)).Finally, the UV irradiation is stopped and the TOL in- andoutlet concentrations are monitored for an additional periodof approximately 12 h.
In the last equations, Q is the total air flow passing throughthe reactor which is fixed at 10.2 L·h−1, A represents thegeometric area (0.008 m2) of the TiO2 coated concrete sampleexposed to the TOL-loaded gas and UV-light, and [TOL]in and[TOL]out are the average toluene concentrations during darkand UV-irradiated periods (μg·L−1) at the in- and outlets,respectively. Two sample repetitions for each coating typeand one reference sample (not coated) are evaluated.
UV irradiation is provided by a blacklight blue (BLB) 18 WUV lamp (340 < λ < 410; maximum emission at 365 nm,Philips), covered by a half cylindrical reflector and positionedjust above the 2.2 mm thick borofloat glass-plate closingthe reactor. Using potassium ferrioxalate actinometry, lightintensity at the catalyst surface was determined to be
2.3 mW·cm−2, being of the same order of magnitude as canbe expected during sunny days in Central Europe [3]. Therelative humidity and temperature of the TOL-loaded gas areregistered at 49.9 ± 1.7% and 23.0 ± 0.3 °C, respectively, bymeans of a TESTO 110 sensor device (TESTO NV). Usingsolid-phase microextraction (SPME) fibers (Supelco, 100 μmthick film of polydimethylsiloxane) at the in- and outlets ofthe reactor, chemical analyses of the gas are performed usinga 6890 Series Gas Chromatograph (Agilent), equipped with aflame ionization detector (FID, 250 °C) fed by 400 mL·min−1
air and 40 mL·min−1 hydrogen. A SPME inlet liner is installedand placed at a temperature of 200 °C. Separation is done on a30 m × 0.53 mm cross-linked methylsiloxane capillary col-umn with a film thickness of 5 μm (HP-1, Hewlett Packard).Equilibrated water–gas systems with a known toluene head-space concentration are used for calibration [33].
3. Results and Discussion
3.1. Nano-scale Characterization
3.1.1. Specific Surface AreaBecause photocatalysis is a surface-controlled process, thespecific surface area is a relevant information for character-izing the synthesized catalyst crystals. The specific surfacearea of the TiO2 synthesized crystals amounts 40 m2·g−1
and 271 m2·g−1 for LFS and LTS technologies, respectively.Furthermore, the average crystal sizes (dBET) calculated usingEq. (1) are approximately 35/40 nm (rutile/anatase) and 5.5–6.0 nm (brookite/anatase) in diameter for LFS and LTS,respectively. It has been stated in literature that decreasingparticle size increases band gap energy due to thequantum-size (Q-size) effect. Q-size effect occurs when theelectronic and optical properties of semiconductor particlesare dependent on the particle size. However, there is noagreement on the size range in which this effect is importantand about which is the optimal size of TiO2 crystals forphotocatalytic applications. For example, while Weller et al.reported that size quantization effects for semiconductorstypically occur between 2 and 50 nm in crystal diameters,Korkmann et al. stated that only crystal diameters between 2and 5 nm evidence these effects [34,35]. However, as con-cluded by Maury Ramirez, the optimum crystal size is alsorelated to the size of the molecule of the pollutant to bephotocatalytically removed [36]. In spite of this, as a generalrule, the crystal growth and decrease of surface area areunwanted processes in TiO2 synthesis.
3.1.2. Crystal Phase CompositionThe XRD pattern of the LFS synthesized particles collected bya cylindrical electrostatic precipitator is presented in Fig. 2a.Both anatase (A) and rutile (R) phases are detected in thiscatalyst. The mass fractions of anatase and rutile areestimated by Eq. (2) and their respective XRD pattern. So,anatase is present for 48% while rutile is about 52%. Anataseto rutile transformation occurs at temperatures between550 °C and 1000 °C [14,15]. However, calcination time forcomplete phase transformation is normally more than 1 h.Although the temperature of LFS is very high (approx. 3000 K),
0
20
40
60
80
100
120
140
160
180
20 30 40 50 60 70 80 90
AA AA AA AA
R (110)
AA AA
R
RAA AA AA AA AA AA
R
R
R
A(101)
25 27
A(101)
R(110)
Inte
nsity
/ co
unts
2 theta / degree
(a) (b)
Fig. 2 – XRD pattern of crystals synthesized by LFS (a) and LTS (b). The inset in the figure represents the smoothed XRD-patternbetween 20 and 30 (2θ/°) and A: anatase, R: Rutile, B: Brookite.
78 M A T E R I A L S C H A R A C T E R I Z A T I O N 8 7 ( 2 0 1 4 ) 7 4 – 8 5
the residence time of the injected precursor solutions in theflame is very short (few milliseconds) which explains theexistence of the anatase phase in the LFS synthesized catalyst.Determined with the Scherrer Equation (Eq. (4)) from the XRDpatterns, the crystal sizes are approximately 35 nm and 42 nmfor anatase and rutile, respectively. This is well consistent withthe crystal size of LFS synthesized catalyst estimated from thespecific surface area. Using the same technology but at differentspraying parameters,Mäkelä et al. reportedTiO2 crystalsmostlyformed by anatase of 10 nm [21]. The photocatalytic efficiencyof mixed-phase TiO2 photocatalysts consisting of anatase andrutile has been widely reported in literature [37–40]. Despitethe fact that single phase rutile is photocatalytically almostinactive, the mixture of anatase and rutile TiO2 has beenreported to exhibit higher photocatalytic activity than the soleanatase phase [39]. The explanation for this is a so-called syn-ergistic effect of anatase and rutile phases [41]. The presenceof rutile crystals has been found to create a structure whereelectron transfer from rutile to lower energy anatase latticetrapping sites under light irradiation leads to a stable chargeseparation, i.e., transfer of the produced electrons from rutile toanatase hinders the recombination of electrons and holes [39].
The XRD pattern of the TiO2 prepared by LTS is presentedin Fig. 2b. In this case, peaks of anatase (A) and a small peak ofbrookite (B) are detected in the synthesized catalyst. Brookiteis generally observed after hydrothermal treatment at lowtemperatures [42]. The mass fractions of anatase and brookiteestimated by Eq. (3) and XRD patterns are 90 and 10%, res-pectively. The diffraction peaks of the LTS synthesized cat-alyst are also wider than the peaks of the LFS synthesizedcatalyst. This is related to the smaller crystal size in the cata-lyst prepared at low temperature as compared to the catalystproduced by LFS technology. Determined also with the Scherrerformula (Eq. (4)) and the XRD patterns, the crystal sizes areapproximately 5 nm and 14 nm for anatase and brookite,respectively. The size of anatase crystals is well consistentwith the estimated crystal sizes calculated from the specificsurface area. Similar crystal sizes are also reported previouslyby Nikkanen et al. using this technology [22].
3.1.3. Band GapsAs indicated in Fig. 3, band gaps are 3.0 eV (413 nm) and 3.1 eV(400 nm) for LFS and LTS, respectively. Although these are notpure anatase, rutile or brookite crystals, these band gaps arein the same scale of magnitude than those reported by Li et al.for anatase, brookite and rutile crystals (3.19 eV/388.7 nm,3.11 eV/398.6 nm, and 3.0/413.3 nm, respectively) [17].
3.1.4. Crystal and Agglomerate SizesAfter analyzing several TEM images for both LFS and LTS, it isclear that crystals synthesized by LFS are significantly biggerthan the crystals synthesized by LTS (Fig. 4a–b). The biggerTiO2 crystals using LFS are most probably due to the rutileformation and grain growth of TiO2 synthesis at such a hightemperature (approx. 3000 K). Although crystal sizes rangebetween 10–90 nm and 2–11 nm for LFS and LTS, respectively,averages of individual TiO2 crystals show that the main sizesare between 40–50 nm (25%) and 5–6 nm (25%) for LFS andLTS, respectively (Fig. 5a–b). These ranges are in completeagreement with the crystal sizes determined from BET andXRD results.
On the other hand, by means of systematic scaling FE-SEMand EDS analyses (from 100 μm to 100 nm) for both coatings(LFS and LTS), it has been found that both synthesizingtechnologies tend to stick TiO2 crystals together in roundagglomerates homogeneously distributed on the autoclavedconcrete surface. These can be confirmed by the cross and topview FE-SEM images and EDS analyses (Figs. 6a–d and 7a–d).However, bigger TiO2 agglomerates are visible for LTS com-pared to LFS (Fig. 6e–f). These agglomerates are over 100 nm indiameter while for LFS these are significantly smaller than100 nm. This makes sense as LTS produces smaller crystalswhich increases the agglomeration tendency compared to thelarger crystals synthesized by LFS. Although it is known thataggregation increases final catalyst particle size and reducescatalyst surface area, the relation between this physicalproperty and the photocatalytic activity has not been studiedin detail [43]. Main complications are derivate from the differ-ences in forces that cause the formation of the agglomerates [44].
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0 100 200 300 400 500 600 700 800 900
Abs
wavelength (nm)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0 100 200 300 400 500 600 700 800 900
Abs
wavelength (nm)
(a) (b)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0 1 2 3 4 5 6 7
(Abs
* E
nerg
y)^1
/2
Energy (eV)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0 1 2 3 4 5 6 7
(Abs
* E
nerg
y)^1
/2
Energy (eV)
Fig. 3 – Band gaps of synthesized catalysts by LFS (a) and LTS (b).
79M A T E R I A L S C H A R A C T E R I Z A T I O N 8 7 ( 2 0 1 4 ) 7 4 – 8 5
3.1.5. CrystallinityTiO2 synthesized by LTS proves to be a crystallized catalystfrom both the DSC and TG analyses (Fig. 8a-d). By comparingthe DSC analyses from the amorphous and LTS samples(Fig. 8a and c), the existence of a peak between 329 and 442 °Cis clear in the amorphous prepared material. This is the resultof an exothermal reaction due to the crystallization of anatasefrom amorphous sample. In LTS prepared sample, there is nocrystallization peak of anatase in the described range, whichindicates that the LTS prepared sample is already in the crys-talline form. However, there is a minor exothermic peakbetween 445 and 500 °C (Fig. 8c). This most probably indicatesthe crystallization of anatase from the brookite trace presentin the LTS sample. According to Bakardjieva et al., brookitetransforms to anatase during heat treatment around thistemperature range [45].
The crystallization enthalpy of the amorphous samplesis 221.5 J·g−1 which is in the same order of magnitude asdetermined by Xie et al. [46]. Furthermore, the bigger peakobserved between 30 and 110 °C in the amorphous samplecompared to the LTS catalyst sample is due to an endothermicreaction produced during the removal of the water retained inpartially dried catalyst. This can also be observed in the TGanalyses for both the amorphous and LTS samples (Fig. 8band d). So, the total weight loss of the amorphous sample
(a)
Fig. 4 – TEM and SAED (Selected Area Electron Diffracti
(approx. 20%) was higher than that one produced in the LTScatalyst sample (approx. 11%). This is most probably dueto the higher amount of physically adsorbed water in theamorphous sample compared to the LTS catalyst sample. Theamorphous sample has a significantly higher surface area(450 m2·g−1) compared to the LTS catalyst sample (271 m2·g−1).On theotherhand, TiO2 synthesized by LFSdoesnot require thiscrystallinity analysis as high temperature synthesis (approx.3000 K) guarantees complete catalyst crystallization.
3.2. Macro-scale Characterization: Air PurificationExperiments
As an illustration of a typical experiment, Fig. 9(a–c) shows theTOL in- and outlet concentrations measured as a function oftime during no UV-irradiated (dark) and UV-irradiated periodsfor a LTS, LFS and reference (non-coated) sample, respectively.Using Eqs. ((6)–(8)), toluene removal efficiencies (nremoval), load-ing and elimination rates (LR and ER) are calculated for theUV-irradiated periods. In Fig. 9(a–b) reproducible tolueneremoval efficiencies of about 60% are calculated with a TOLinlet concentration of 13.9 ± 0.4 ppmv and a 3 min gas resi-dence time for both synthesized coatings. In contrast, nosignificant difference between TOL in- and outlet concentra-tions is measured in the dark period (t-test, significance level
(b)
on) images of LFS (a) and LTS (b) synthesized TiO2.
Fig. 5 – Particle size distribution of TiO2 synthesized by LFS (a) and LTS (b), n = 300.
80 M A T E R I A L S C H A R A C T E R I Z A T I O N 8 7 ( 2 0 1 4 ) 7 4 – 8 5
α ≥ 0.05). Evaluating the reference sample at the same condi-tions of inlet concentration, gas residence time, RH, andtemperature clearly shows that no significant toluene removalis obtainedwith thismaterial, as exemplified by toluene in- andoutlet concentrations of 12.0 ± 0.5 and 12.6 ± 0.2 ppmv, respec-tively, during UV irradiation (Fig. 9c).
Table 1 summarizes the results obtained for both synthe-sized coatings (including 2 test repetitions). Removal efficien-cies of 59.8 ± 2.9% (ER = 37.5 mg TOL·m−2·h−1) and 61.2 ± 2.3%(ER = 43.3 mg TOL·m−2·h−1) for LTS and LFS synthesizedcoatings, respectively, indicate significant TiO2 photocatalyticactivity for degradation of toluene from air at the studiedconditions for both novel photocatalytic materials. Theadvantage of the LFS synthesized coating is the mixture ofanatase and rutile which is known to enhance the photocat-alytic activity by the synergistic effect of the combination ofthese two polymorphs. However, the specific surface area ofthe LFS catalyst crystals is only 40 m2·g−1 while the sameparameter amounts 271 m2·g−1 in the LTS catalyst particles.Consequently, the benefit of the LTS synthesized sampletowards the photocatalytic application is the relatively highspecific surface area. Although a systematic evaluation of thecoatings including different precursor quantities, synthesiz-ing temperatures and times is required to better understandthe link between nano-scale coating characteristics andphotocatalytic activity, the toluene removal efficiencies andrates obtained in this research are promising. Removal oftoluene from air using photocatalytic building materials hasbeen reported in some cases to be very low or even non-perceivable. For example, Maury-Ramirez et al. obtained atvery similar conditions only removal efficiencies from 1 to 7%(equivalent to elimination rates from 0.7 to 4.0 mg TOL·m−2·h−1,respectively) using sol–gel synthesized coatings (450 °C dur-ing 5 h) on different concrete types including autoclavedaerated concrete. These low efficiencies were attributed tothe reduction of the active surface and due to the presence ofionic species that contributed to charge recombination inthese synthesized sol–gel coatings [2]. Similarly, photocata-lytic removal of BTEX compounds from air using TiO2 (P25,Degussa) containing white Portland cement was investigatedby Strini et al. At concentrations of 2.5–3.7 μmol·m−3, BTEXremoval efficiencies between 5 and 54% (i.e. oxidation ratesof 0.2–1.3 μmol m−2·h−1) were noticed using 1% (by weight)
TiO2 containing cement. Thehighest photocatalytic activity wasobserved towards o-xylene degradation, followed by ethyl-benzene > toluene > benzene [47]. Using also TiO2 Degussa P25added concrete (5% and 10% on a weight basis), Chen et al. didnot observe toluene removal from air, neither when toluene(200–800 ppb) was dosed as a single pollutant nor when mixingtoluene at the same concentrationswith NO (400 ppb). Temper-ature, relative humidity and UV-A intensity amounted to 25 °C,50% and 10 W·m−2, respectively. As a general explanation, it wasconcluded that when nano-particles are embedded in cement-based materials, they become potential nucleation sites forcement hydration products. As the hydration reactions proceed,the products (e.g. calcium silicate hydrate and calcium hydrox-ide) gradually bind the individual nano-TiO2 together, forming adense coating on the TiO2 surface. Although the solid layerforms a stable support that protects the TiO2 particle fromabrasion or erosion, it may also weaken the photocatalyticactivity. In this case, the “protective” cementitious coatingcould absorb or block the photons that are supposed to hit thephoto-reactive TiO2 surface sites, reducing the quantum effi-ciency that is important for the oxidation of toluene molecules.It may also act as a diffusion barrier that inhibits the contactbetween the gas molecules and the active sites of TiO2. As aresult, the photocatalytic activity of TiO2 is significantlysuppressed and the removal of toluene cannot be observed inthe studied conditions [19].
4. Conclusions
This research presents the novel application of two differentTiO2 synthesis-coating methods on autoclaved aerated con-crete yielding photocatalytic activity towards toluene removalfrom air. Interestingly, although significantly different crystaltypes and sizes, specific surface areas, relative contents ofcrystalline phases, and agglomerate sizes are obtained for theTiO2 coatings synthesized by liquid flame spray (LFS) and lowtemperature synthesis (LTS), no significant differences areobserved towards toluene removal from air. Removal efficien-cies of 61.2% (ER = 43.3 mg TOL·m−2·h−1) and 59.8% (ER =37.5 mg TOL · m−2 · h−1) for LFS and LTS, respectively, wereobtained at the studied conditions. It seems that the differentnano-scale characteristic effects compensate each other so
(a) (b)
(c) (d)
(e) (f)
Fig. 6 – FE-SEM micrographs of LFS and LTS generated TiO2 coated concrete samples. (a and b) Cross sections, (c and d) topviews, (e and f) top views at higher magnification, (x) location of EDS analyses.
81M A T E R I A L S C H A R A C T E R I Z A T I O N 8 7 ( 2 0 1 4 ) 7 4 – 8 5
that the overall photocatalytic activities as observed in ourexperimental setup and conditions do no differ for bothcoating technologies. It is clear that both LFS and LTSsynthesis methods have their advantages in production ofphotocatalytic coatings. Synthesized catalyst particles usingLFS technology are characterized by their mixture composi-tion of rutile and anatase which is known to enhance thephotocatalytic activity, while catalyst particles synthesized byLTS are characterized by their significantly larger surface area.However, as there is the possibility that the effects of the
investigated characteristics are too small to observe, a sys-tematic evaluation including different precursor quantities,synthesizing temperatures and times is required to furtherimprove the knowledge about the effect of synthesis condi-tions on nano-scale coating characteristics on one hand, andabout the effect of these characteristics on the photocatalyticactivity on the other hand.
Although the full mechanisms are still not fully elucidated,this article reports for the first time in literature that both liquidflame spray and low temperature synthesis technologies are
CaC
O
Si Cl CaCa
Ti
Ti
CaC
O
Al
Si
Cl K
Ca
CaTi
CaC
O
NaSi
S CaKCl
Ti
Ti
CaC
O
NaAl
Si
SCl K
Ca
Ca Ti
(b)
(c) (d)
(a)EDS 1
EDS 1
EDS 2
EDS 2
Fig. 7 – EDS patterns from coated samples using LFS (a–b) and LTS (c–d). Location of these EDS analyses are indicated in Fig. 6(a–b).
-1.50E-01
-1.00E-01
-5.00E-02
0.00E+00
5.00E-02
1.00E-01
30 90 150 210 270 330 390 450
Peak:Area: -221.5 J/gMaximum : 383.1 C
DS
C (
mW
/mg)
Exo
Temperature
70
75
80
85
90
95
100
30 90 150 210 270 330 390 450
Wei
ght l
oss
(%)
Temperature
-1.50E-01
-1.00E-01
-5.00E-02
0.00E+00
5.00E-02
1.00E-01
1.50E-01
30 90 150 210 270 330 390 450
DS
C/ (
mW
/mg)
Exo
Temperature
85
90
95
100
30 90 150 210 270 330 390 450
Temperature
Wei
ght l
oss
(%)
(a) (b)
(c) (d)
Fig. 8 – DSC and TG curves of amorphous (a–b) and LTS (c–d) synthesized catalyst samples.
82 M A T E R I A L S C H A R A C T E R I Z A T I O N 8 7 ( 2 0 1 4 ) 7 4 – 8 5
0
5
10
15
20
0 10 20 30 40 50 60 70
0
5
10
15
20
0 10 20 30 40 50 60 70
TOL
concentration
(ppmv)
(a)
ηremoval= 57.8 %
UV-irradiationDark Dark
Time (h)
TOL
concentration
(ppmv)
UV-irradiationDark Dark
(b)
ηremoval= 62.8 %
TOL
concentration
(ppmv)
(c)
ηremoval~ 0 %
UV-irradiationDark Dark
Time (h)
Time (h)
Fig. 9 – Gas phase TOL inlet (•) and outlet (×) concentrations during an air purification experiment using an (a) LTS coatedsample, (b) LFS coated sample, and (c) reference sample.
Table 1 – Average and standard deviation (SD) of inlettoluene concentrations ([TOL]in, ppmv), removal efficiencies(ηremoval, %), loading rates (LR, mg TOLm−2∙h−1), eliminationrates (ER, mg TOLm−2∙h−1), relative humidity (RH, %) andtemperature (T, °C), as measured during the evaluation ofboth LTS and LFS coatings (n = 2) during 30 h of UV-Airradiation.
83M A T E R I A L S C H A R A C T E R I Z A T I O N 8 7 ( 2 0 1 4 ) 7 4 – 8 5
promising for application on cementitious materials with theaim to obtain air-purifying properties on autoclaved aeratedconcrete. In relation to other TiO2 loading methods, thesetechnologies cannot only be easily applied on precast materialsduring manufacturing, they can also be applied on existingbuildings which is an important practical advantage. Moreover,the photocatalytic activity reported here is significantly higherthan that reported in [2,19,47], using high temperature sol–gelsynthesized coatings or commercial TiO2 (Degussa P25) embed-ded in the cementitious matrix at different concentrations.Potential application of these novel photocatalytic materials isin confined spaces such as canyon streets in densely populatedcities, where relatively high pollution levels can be foundbecause of vehicles exhaust and/or in case of proximity ofindustrial sources.
84 M A T E R I A L S C H A R A C T E R I Z A T I O N 8 7 ( 2 0 1 4 ) 7 4 – 8 5
Acknowledgements
The authors would like to thank Ghent University (Belgium)for the financial support given via the mobility grant (CWO)and the PhD grant (BOF 01W04308) that made this interna-tional cooperation research project possible. Similarly, wewould like to thank the Finish National Graduate School onNew Material and Processes for the financial support. Weare also grateful to Tampere University of Technology andGhent University personnel for technical support; especially,to Mikko Kylmälahti, Jarmo Laakso, Saara Heinonen andChristophe Walgraeve.
R E F E R E N C E S
[1] Maury-Ramirez A, Demeestere K, De Belie N. Photocatalyticactivity of titanium dioxide nanoparticle coatings applied onautoclaved aerated concrete: effect of weathering on coatingphysical characteristics and gaseous toluene removal.J Hazard Mater 2012;211–212:218–25.
[2] Maury-Ramirez A, Demeestere K, De Belie N, Mäntylä T,Levänen E. Titanium dioxide coated cementitious materialsfor air purifying purposes: preparation, characterization andtoluene removal potential. Build Environ 2010;45:832–8.
[3] Hüsken G, Hunger M, Brouwers HJH. Experimental study ofphotocatalytic concrete products for air purification. BuildEnviron 2009;44:2463–74.
[4] Chen J, Poon CS. Photocatalytic activity of titanium dioxidemodified concrete materials — influence of utilizingrecycled glass cullets as aggregates. J Environ Manag2009;90:3436–42.
[5] Demeestere K, Dewulf J, De Witte B, Beeldens A, VanLangenhove H. Heterogeneous photocatalytic removal oftoluene from air on building materials enriched with TiO2.Build Environ 2008;43:406–14.
[6] Poon CS, Cheung E. NO removal efficiency of photocatalyticpaving blocks prepared with recycled materials. Constr BuildMater 2007;21:1746–53.
[7] Maury-Ramirez A, De Muynck W, Stevens R, Demeestere K,De Belie N. Titanium dioxide based strategies to prevent algalfouling on cement based materials. Cem Concr Compos2013;36:93–100.
[8] Maury A, De Belie N. State of the art of TiO2 containingcementitious materials: self-cleaning approach. MaterConstrucc 2010;60:33–50.
[9] Fonseca AJ, Pina F, Macedo MF, Leal N, Romanowska-DeskinsA, Laiz L, et al. Anatase as an alternative application forpreventing biodeterioration of mortars: evaluation andcomparison with other biocides. Int Biodeter Biodegr2010;64:388–96.
[10] Maury Ramirez A, De Belie N. Evaluation of the algaecideactivity of titanium dioxide on autoclaved aerated concrete. JAdv Oxid Technol 2009;12:100–4.
[11] Giannantonio DJ, Kurth JC, Kurtis KE, Sobecky PA. Effects ofconcrete properties and nutrients on fungal colonization andfouling. Int Biodeter Biodegr 2009;63:252–9.
[12] Linkous CA, Carter GJ, Locuson DB, Ouellete AJ, Slattery DK,Smith LA. Photocatalytic inhibition of algae growth usingTiO2, WO3, and cocatalyts modifications. Environ Sci Technol2000;34:4754–8.
[13] Hoffmann MR, Martin ST, Choi W, Bahnemann DW.Environmental applications of semiconductorphotocatalysis. Chem Rev 1995;95:69–96.
[14] Hanaor DAH, Sorrell CC. Review of the anatase to rutile phasetransformation. J Mater Sci 2011;46:855–74.
[15] Li JG, Ishigaki T. Brookite → rutile phase transformation ofTiO2 studied with monodispersed particles. Acta Mater2004;52:5143–50.
[16] Allen NS, Edge M, Verran J, Stratton J, Maltby J, Bygott C.Photocatalytic surfaces: antipollution and antimicrobialeffects. In: Krug HF, editor. Nanotechnology volume 2:environmental aspects. Willey; 2008.
[17] Li JG, Li X, Sun X. TiO2 nanocrystals: synthesis and enhancedfunctionality (nanotechnology science and technology). NovaScience Publishers; 2010.
[18] Lackhoff M, Prieto X, Nestle N, Dehn F, Niessner R.Photocatalytic activity of semiconductor-modifiedcement-influence of semiconductor type and cement ageing.Appl Catal B Environ 2003;43:205–16.
[19] Chen J, Kou SC, Poon CS. Photocatalytic cement-basedmaterials: comparison of nitrogen oxides and tolueneremoval potentials and evaluation of self-cleaningperformance. Build Environ 2011;46:1827–33.
[20] Aromaa M, Keskinen H, Mäkelä JM. The effect of processparameters at liquid flame spray generated titaniananoparticles. Biomol Eng 2007;24:543–8.
[21] Mäkelä JM, Hellsten S, Silvonen J, Vippola M, Levänen E,Mäntylä T. Collection of liquid flame spray generated TiO2
nanoparticles on stainless steel surface. Mater Lett2006;60:530–4.
[22] Nikkanen JP, Huttunen-Saarivirta E, Kanerva T, Pore V, KiveläT, Levänen E, et al. Photoactive TiO2 nanopowder synthesizedat low temperature without catalyst. J Ceram Sci Technol2011;02:97–102.
[23] Nikkanen JP, Kanerva T, Mäntylä T. The effect of acidity inlow-temperature synthesis of titanium dioxide. J CrystGrowth 2007;304:179–83.
[24] Keskinen H, Mäkelä J, Aromaa M, Keskinen J, Areva S,Teixeira C, et al. Titania and titania-silver nanoparticledeposits made by liquid flame spray and their functionalityas photocatalyst for organic- and biofilm removal. Catal Lett2006;111:127–32.
[25] Gopal M, Moberly Chan WJ, De Jonghe LC. Room temperaturesynthesis of crystalline metal oxides. J Mater Sci1997;32:6001–8.
[26] Švadláka D, Shánělová J, Málek J, Pérez-Maqueda LA, CriadoJM, Mitsuhashi T. Nanocrystallization of anatase inamorphous TiO2. Thermochim Acta 2004;414(2):137–43.
[27] American Society for Testing and Materials. ASTM C 1202-10Standard test method for electrical indication of concrete'sability to resist chloride ion penetration. Annual Book ofASTM Standards; 2004 (04.02).
[28] De Belie N, Monteny J, Beeldens A, Vincke E, Van Gemert D,Verstraete W. Experimental research and prediction of theeffect of chemical and biogenic sulfuric acid on differenttypes of commercially produced concrete sewer pipes. CemConcr Res 2004;34:2223–36.
[29] British Standards Organization. BS 1134 Britishstandard method for the assessment of surface texture; 1990(Parts 1 & 2).
[30] Zhang H, Banfield JF. Understanding polymorphic phasetransition behaviour during growth of nanocrystallineaggregates: insights from TiO2. J Phys Chem B2000;104:3481–7.
[31] Cullity BD. Elements of X-ray diffraction. USA:Addison-Wesley Publishing Company; 1967.
[32] Japanese Standards Association. JIS R 1701-3 fine ceramics(advanced ceramics, advanced technical ceramics) — testmethod for air purification performance of photocatalyticmaterials — part 3: removal of toluene. Japanese IndustrialStandard JIS; 2008 1–8.
[33] Van Durme J, Dewulf J, Sysmans W, Leys C, Van LangenhoveH. Efficient toluene abatement in indoor air by a plasmacatalytic hybrid system. Appl Catal B Environ 2007;74:161–9.
85M A T E R I A L S C H A R A C T E R I Z A T I O N 8 7 ( 2 0 1 4 ) 7 4 – 8 5
[34] Korkmann C, Bahnemann DW, Hoffmann M. Preparation andcharacterization of quantum-size titanium dioxide. J PhysChem-us 1998;92:5196–201.
[35] Weller H, Eychmuller A. Photochemistry andphotoelectrochemistry of quantized matter: properties ofsemiconductor nanoparticles in solution and thin-filmelectrodes. In: Neckers DC, Volman DH, Von Bunau G, editors.Advances in photochemistry, vol. 20. JohnWileyandSons; 1995.
[36] Maury Ramirez A. Cementitious materials with air-purifyingand self-cleaning properties using titanium dioxidephotocatalysis. Doctoral thesis Belgium: Faculty ofEngineering and Architecture. Ghent University; 2011-2012.
[37] Yu J, Yu H, Cheng B, Zhou M, Zhao X. Enhancedphotocatalytic activity of TiO2 powder (P25) by hydrothermaltreatment. J Mol Catal A Chem 2006;253:112–8.
[38] Bakardjieva S, Subrt J, Steng V, Dianez MJ, Sayagues MJ.Photoactivity of anatase-rutile TiO2 nanocrystalline mixturesobtained by heat treatment of homogeneously precipitatedanatase. Appl Catal B Environ 2005;58:193–202.
[39] Hurum DC, Agrios G, Gray KA. Explaining the enhancedphotocatalytic activity of Degussa P25 mixed-phase TiO2
using EPR. J Phys Chem B 2003;107:4545–9.[40] Bacsa RR, Kiwi J. Effect of rutile phase on the photocatalytic
properties of nanocrystalline titania during the degradationof p-coumaric acid. Appl Catal B Environ 1998;16:19–29.
[41] Ohno T, Sarukawa K, Tokieda K, Matsumura M. Morphologyof a TiO2 photocatalyst (Degussa, P-25 consisting of anataseand rutile crystalline phases. J Catal 2001;203:82–6.
[42] Li G, Gray KA. Preparation of mixed-phase titanium dioxidenanocomposites via solvothermal processing. Chem Mater2007;19:1143–6.
[43] Prafulla KJ. Titanium dioxide: applications, synthesis andtoxicity (chemistry research and applications). Nova SciencePub; 2013.
[44] Teleki A, Wengelerb R, Wengeler L, Nirschl H, Pratsinis SE.Distinguishing between aggregates and agglomerates offlame-made TiO2 by high-pressure dispersion. PowderTechnol 2008;181–3:292–300.
[45] Bakardjieva S, Stengl V, Szatmary L, Subrt J, Lukac J, Murafa N,et al. Transformation of brookite-type TiO2 nanocrystals torutile: correlation between microstructure and photoactivity.J Mater Chem 2006;16:1709–16.
[46] Xie H, Zhang Q, Xi T, Wang J, Liu Y. Thermal analysis onnanosized TiO2 prepared by hydrolysis. Thermochim Acta2002;381:45–8.
[47] Strini A, Cassese S, Schiavi L. Measurement of benzene,toluene, ethylbenzene and oxylene gas phasephotodegradation by titanium dioxide dispersed incementitious materials using a mixed flow reactor. ApplCatal B Environ 2005;61:90–7.
